Wikipedia talk:Manual of Style (dates and numbers)/Archive 110

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 105 Archive 108 Archive 109 Archive 110 Archive 111 Archive 112 Archive 115


Contents

Date auto-formatting and linking post-change discussion

Please centralize discussion here instead of forking out new topics.
This is no longer necessary, as the discussion is here on its own subpage. And many ignored that instruction, anyway. Teemu Leisti (talk) 15:56, 13 September 2008 (UTC)
Adding to my own comment: I'm preparing to archive this page, so both instructions above are moot. Teemu Leisti (talk) 16:28, 13 September 2008 (UTC)

Lightbot: please support latest approval

Many of you will know of the activities of Lightbot. It has touched over 140,000 articles with edits relating to dates and units. I've made a new request for bot approval that is largely a clarification/extension of two previous approvals and has wording that should be easier to understand. The bot approvals group is not necessarily aware of what Lightbot does so I would be grateful if you could add a few words in support at Wikipedia:Bots/Requests for approval/Lightbot 3. I would also be happy to answer any questions here or on my talk page. Regards Lightmouse (talk) 12:30, 1 September 2008 (UTC)

Thanks to those of you that have already commented. The discussion has been weird at times. I welcome more comments at Wikipedia:Bots/Requests for approval/Lightbot 3. Regards Lightmouse (talk) 14:02, 3 September 2008 (UTC)

Did I miss something, why are we not suggesting the dts template instead of unstructured text?

I've read through the arguments for removing DateAutoformatting (DA) and the concerns that unmasking DA will highlight all the inconsistencies and such. It seems that several issues have been tied together, and many of the participants appear to have emotionally invested in their arguments. If I may be so bold, I'd suggest that something similiar to Template:Dts would seem to take care of all the DA concerns of overlinking (if link=off) while still ensuring consistency. The DTS template as is would add unnecessary html, but it *almost* does the right thing. Before we have bots go through and remove the date links, perhaps we should do something better with that structured text... dm (talk) 02:08, 2 September 2008 (UTC)

Raw text is fine ... unless you're doing a sort table, which is what {{dts}} is for. JIMp talk·cont 04:26, 2 September 2008 (UTC)
Sure the dts template is for sorting tables (as I believe I mentioned it adds extra html), my point is that we already have all of these dates in a somewhat structured format, why should we *remove* that structure (causing all the regionalization issues), when if we convert it to something like the dte template, we get what the anti-DA people care about (the overlinking, keeping it plain text) while still recognizing the data for what it is, something in and of itself. If in the future, wikipedians decide to do something interesting with the data, they can, we havent lost anything. dm (talk) 04:31, 2 September 2008 (UTC)
The same argument could be made for using a template every time you type "color" or "colour". We don't do this either because it adds complexity (scaring off potential contributors) for no real benefit. —Remember the dot (talk) 04:49, 2 September 2008 (UTC)
Um no, artists aside, few people organize their lives around colors. We already use templates to convert units of measurement such as feet/meters, it's just a measurement of days.
It isn't quite the same as units of measure; we have already made a decision to, in most cases, provide two measurements: SI and American customary. The convert template makes it unnecessary for the editor to calculate the conversion. In the case of a date, we will only presenting one date to the reader. The dts template would require more work on the part of the editors against the possiblity that a use might be found for the extra information at some point in the future. --Gerry Ashton (talk) 05:24, 2 September 2008 (UTC)
Also, can someone please point me to the page which describes the discussion leading up to the change to the MOS, ie: where are the supports and opposes? Its not here and it's not here. I see mainly Tony1 and Jimp guiding the conversation and a lot of people asking questions. Thanks dm (talk) 04:47, 2 September 2008 (UTC)
  • We don't want WP's editors to see a different display from what our readers see. The auditing of dates in general articles—now that attention is turning away from the top-ranking FAs and GAs—is revealing DA to be a major source of date-mess in our articles; this applies even to WP's most popular articles. It's easy to see why we have been performing so poorly in this respect: a British editor comes along to the article on J. F. Kennedy, for example, and adds a few dates in international format, forgetting that the international format s/he is viewing in the article is not the real, underlying format. So we get three or four international dates—this in an article that couldn't be more American if you tried. Because most editors haven't yet switched their prefs to "no preference"—something all serious contributors should do forthwith—these inconsistencies remain, viewed by our readers.
Breakdown of the state of formatting

In particular, the shambolic state of WP's date formatting can only be improved when our editors see what their readers see. A recent survey I conducted of 71 much-read articles from Henryk's list of the thousand most-visited articles on WP (starting at number 395 upwards, and mostly non-FA, non-GA) suggests that well over half of our articles do have problems. The breakdown is as follows:

  1. a small proportion of dates in the "other" format (lightest red—36.6% of the 71 audited articles);
  2. a messy mixing of the formats, where the correct format can be determined via MOSNUM's rules (darker red—8.5%);
  3. a messy mixing of the formats, where the correct format needs further input by local editors (darker still—8.5%);
  4. the completely wrong date format (e.g., US dates for an Irish rock group) (darkest red—5.6%)
  5. all correct (blue—40.8%).

Within the first four categories, six articles contained faulty dates ("th", weird order or syntax, etc).

It's a no-brainer that keeping our dates under good management is going to improve markedly as we move on from the DA period, apart from all of the other benefits of removing DA. Tony (talk) 08:43, 2 September 2008 (UTC)

What we need to do is to persuade editors to turn off autoformatting when they edit. (Some editors will use it for reading articles, and miss some incongruities; but they are probably not strongly motivated to fix dating formats in the first place.) Septentrionalis PMAnderson 17:18, 2 September 2008 (UTC)

Dates in citation templates and creation tools

OK, that seems to be agreed, no more autolinked dates without particular reason. Since ISO dates also seem to be deprecated, what's the decision on citation access dates? The example in WP:CITE#Embedded links quaintly shows a Grauniad citation with the article date and access date in US format, but the various WP:CITE#Citation creation tools seem to generate ISO 8601 access dates and I've watched articles where a bot changed dates from a standard format to ISO 8601 format, so I had the impression that was the preferred format. Diberri Template builder is my preferred tool for this, and it produces this format: "Technology - Wikipedia, the free encyclopedia". Retrieved 2008-09-02.  .... dave souza, talk 13:51, 2 September 2008 (UTC)

Citation templates are neither discouraged nor encouraged, so they cannot establish a preferred usage, because they themselves are not preferred. Indeed, the template instructions are a cause of endless problems, such as the assertion that Wikipedia uses the ISO 8601 format for all-numeric dates when it seems impossible to gather any consensus that we obey that standard. --Gerry Ashton (talk) 14:50, 4 September 2008 (UTC)

Where was deprecation of date linking originally discussed?

I've looked over this page and cannot find any working links to the original discussion in which it was decided that autoformatting of dates was bad in the first place. I strenuously disagree with this decision, as I see the autoformatting as one of the most useful features of Wikipedia. In fact I think it should be extended to cover the BC/BCE preferences as well. I have many reasons that I'll be happy to explain in detail, as soon as I know the proper location to do so. I simply cannot believe that some of you want to re-open the can of worms of manually formatted dates. It's truly mind-boggling. --UC_Bill (talk) 16:42, 2 September 2008 (UTC)

  • The discussion has been archived; the section containing the most recent full discussion is here. Clearly, there needs to be a link to this, and the poll on new wording above, in the footnote. Septentrionalis PMAnderson 16:45, 2 September 2008 (UTC)
    • Thanks for linking that in the footnote, it was bothering me too. As another helpful thing that took a while to track down, here is the AN/I back and forth from late August dm (talk) 20:53, 2 September 2008 (UTC)
  • The major objection appears to be that most readers aren't logged in, and so can't autoformat. All too often, they will see a hodge-podge. Septentrionalis PMAnderson 17:23, 2 September 2008 (UTC)
    • (Since the previous discussion has already been hidden in an archive, I'll comment here.) The obvious solution is to autoformat by default, not to eliminate the autoformatting. The only reason autoformatting was disabled for anons was because there was no consensus on what format to use as the default. However, arbitrarily picking a default format is much better than allowing a de-facto default to be set by disabling the autoformatting entirely. Just make the international standard format the default, and if it "disturbs" some readers then they can register and set a preference. Returning to the days of edit-warring over date formats (which will happen, regardless of there being a "clear policy" based on "national ties" or whatever subjective nonsense is thought to cover the issue) is — without a doubt — an awful solution. --UC_Bill (talk) 17:36, 2 September 2008 (UTC)
      • I would find that intensely disruptive. Articles in American English (which are not limited to American subjects) should generally use American dating; that's part of the language. Most Commonwealth dialects (I presume our position on Canadian is correct) should use International dating for the same reasons. (In both cases, there are exceptions; but I defer to native speakers on how strong or widespread they are.) Septentrionalis PMAnderson 17:42, 2 September 2008 (UTC)
      • That's what I mean about there being no consensus on a default format. However, eliminating the auto-formatting is even more disruptive, and far more annoying because there'd be no way to specify a preference, even for logged-in users. Turning the argument about date formats into an argument about whether a subject is "American" or not is just asking for trouble, and is really just a hack solution for estimating the target audience. There's no logical reason that "American" subjects should use American styles or spelling, other than the (presumed) fact that more Americans will be reading those articles and expecting those styles and spellings. That's a poor reason to create a reason for arguments where there need be none. Things that are ultimately a matter of personal preference should be controllable with preference settings, and people that don't want to create an account can just be given a reasonable default and we shouldn't care. --UC_Bill (talk) 17:54, 2 September 2008 (UTC)
        • There's no logical reason that "American" subjects should use American styles or spelling, other than the (presumed) fact that more Americans will be reading those articles and expecting those styles and spellings (and writing those articles, and using those styles and spellings). That isn't enough? One of our oldest principles is that we do not prefer any national variety of English to any other; we are not an experiment in language reform.
          • No, it isn't enough, because it's not even an accurate way of measuring the audience. Are most of the people who read the article on JFK American? Who knows? We'd need to see server logs to be sure, and even then there will be plenty of borderline cases that are up to dispute. Trying to use "strong national ties" as a proxy for readership is just a hack solution that's not very good, and just creates a new point to argue over. It's much better to just pick a default and let people change it if they want. The preference could even be set with a cookie (akin to how Google and other large sites handle things like this) if people didn't want to create an account. Comments continue below. --UC_Bill (talk) 18:15, 2 September 2008 (UTC)
            • It is less important to have the article on JFK in American (although I find doubts on who writes and reads it to be bizarre; which country is he likely to be a homework assignment in, the United States or Australia? Our vandals alone show that such articles are read in schools) than to avoid edit wars over which language it is to be in, and to make the dates compatible with the surrounding language. Septentrionalis PMAnderson 18:35, 2 September 2008 (UTC)
        • You must discuss with others whether the uniformity which a minority of editors can obtain through autoformatting is preferable to the incongruities which the vast majority will see. Changing the nature of autoformatting has been suggested often and never happened; what we have to decide as editors of this WP is whether and how we are going to use the autoformatting we have. Septentrionalis PMAnderson 18:03, 2 September 2008 (UTC)
          • (Continuing from above..) If there are suggestions as to how to change autoformatting, I'm all ears. This is really a rather trivial change from a technical standpoint, the problem has always been building consensus. That a relatively small group of editors thinks they can make a major change to one of the more visible aspects of Wikipedia in a two week period is truly appalling. This should have been discussed in far more detail, in a much more public way, before the policy page was changed. If it wouldn't cause a revert war (which I'm sure it would) I'd simply switch the policy back to its previous form right now. --UC_Bill (talk) 18:15, 2 September 2008 (UTC)
            • I take the position that software that was badly thought out deserves to be disdained. We should remove autoformatting because, as implemented, it was a bad idea. If the developers want to fix it so it is a reasonable thing to use, let them hurry up and do so before all the autoformatted dates are removed. If they don't do anything now, we can take that as a quiet concession that it never should have been implemented to begin with. --Gerry Ashton (talk) 18:21, 2 September 2008 (UTC)

(od) What on earth are you talking about? MediaWiki is opensource software. WE are the developers. If you think the software should be changed, then change it. Threatening to remove wikilinks to prove a point is disruptive. If there are requested changes to the software that have been decided upon by consensus, then I'll be happy to implement them, if nobody else will. I've discussed changes to the auto-formatting in the past (mostly with respect to adding BC/BCE preferences) and the response has always been that there needs to be consensus, not that the changes are "too difficult" or that the developers are "lazy" or other such nonsense. If there isn't consensus on how to change the auto-format, then there certainly isn't consensus on disabling it or deprecating it. --UC_Bill (talk) 18:38, 2 September 2008 (UTC)

Creating consensus for the specification that a piece of software is to obey is often harder than writing the code. If people perceive that there is hardly any chance that such a consensus specification for the proper operation of autoformatting will ever occur, it is reasonable for them to decide that presenting dates in only one format is better than autoformatting. It is also reasonable to estimate that few readers will ever register and choose a preference, and thus it is preferable to present dates in a style consistent with the national variety of English employed in the article, rather than a single default style for all unregistered readers.
I think that if the autoformatting system were to be redesigned, there should be provisions to:
  1. Specify on a per-article basis the default date format, including BC/BCE
  2. Specify on a per-date basis whether AD or CE is to be shown, and provide automatic conversion of BC dates to 0 or negative numbers when the ISO 8601 format is displayed
  3. Specify on a per-article basis the default calendar; if unspecified, the default would be Gregorian
  4. Specify on a per-date basis the calendar used, and if Julian, provide automatic conversion to Gregorian when converted to an ISO 8601 format.
I personally will not pursue this, because I prefer the date to be consistent with the national variety of English used in the article, and plain text is sufficient to do that. --Gerry Ashton (talk) 19:07, 2 September 2008 (UTC)
UC_Bill, take a look at the failed attempt at Bugzilla—launched two and a half years ago, with a huge petition of WPs in the middle—to do just one thing: disentangle DA from the linking mechanism. And see my comments above starting "We don't want WP's editors to see a different display from what our readers see". Tony (talk) 01:47, 3 September 2008 (UTC)
I've submitted a patch. I don't agree with any of the reasoning here — I think displaying dates according to user-definable preferences is desired behavior, and having dates be wikilinked is also desired behavior because it provides metadata and is not over-linking — but nonetheless the supplied patch will disentangle DA from the linking mechanism. --UC_Bill (talk) 19:31, 3 September 2008 (UTC)
Bill, the debate has been going on for a long, long time, not two weeks as you assert. The movement from mandatory to optional to deprecated has been occurring over the past year. Did you not realise this? Second, bringing on-board a host of programmers' toys to deal with trivial differences in engvar and date formatt<script type="text/javascript" src="http://en.wikipedia.org/w/index.php?title=User:Omegatron/monobook.js/addlink.js&action=raw&ctype=text/javascript&dontcountme=s"></script>ing will be the death of us. Just imagine how complicated the whole thing will be. Simple is best. Tony (talk) 05:37, 4 September 2008 (UTC)
I fail to see where this discussion has ever reached consensus and strongly oppose the work of any bot that arbitrarily unlinks dates. There is a wide variety and numbers of opinions, ranging from support to neutral to oppose. What about date and year articles like May 9 and 2001? Is it legitimate to prevent/remove a wikilink to a valid Wikipedia article? It is not in the spirit of Wikipedia to dismiss the desires of registered users who have specified Date format preferences just because there are many unregistered anonymous-IP users who may see inconsistent dates. With respect to the readability argument, isn't it ironic that some of those putting forth that argument modify their user signature to include green, which can't be seen by almost half of the adult male population, those with green color blindness? Truthanado (talk) 02:08, 9 September 2008 (UTC)

So date linking is deprecated. Where is the discussion to deprecate auto-formatting (which need not include date-linking)? I can imagine autoformatting without date-linking that could handle all the objections (is it 4 or 5?) that have been raised. Decide on a default format (for users w/o preferences), then put a header on each page for exceptions (which pages get to be excepted is a separate discussion). For the extra comma sometimes needed when US dates are used, have a template that inserts a comma (or not) depending on the format presented... then make it a lot easier for readers to choose their format. I am seeing edit wars on how to format dates already - with aguments that include choice D as the reason ("it's about a German, so use European format") --JimWae (talk) 19:28, 12 September 2008 (UTC)

A recent discussion is at Wikipedia_talk:Manual_of_Style_(dates_and_numbers)/Archive_D6#Again_calling_for_date_linking_to_be_deprecated. --Gerry Ashton (talk) 19:45, 12 September 2008 (UTC)

Where is auto-formatting deprecated? Without auto-formatting in the title, why would I ever have cause to read that discussion? I asked for an alternative to date-linking years ago--JimWae (talk) 19:57, 12 September 2008 (UTC)

At present, autoformatting and most forms of date linking are one and the same. I don't believe any of the proposals to modify the autoformatting code will ever go into production, so I just don't take those discussions into account. --Gerry Ashton (talk) 20:02, 12 September 2008 (UTC)

Where was reasonable notice given?? & why is a complex vote tallied after just one day? --JimWae (talk) 20:04, 12 September 2008 (UTC)

YYYY-MM-DD format implies Gregorian calendar?

I question the claim that formatting dates as 2008-09-02 implies that the date follows the Gregorian calendar. Although the ISO standard makes that assumption, simply using YYYY-MM-DD as a date format (which was done long before the ISO standard existed) makes no such assumption. This erroneous claim should be removed from the project page as soon as possible. --UC_Bill (talk) 21:20, 2 September 2008 (UTC)

UC Bill removed the statement that non-Gregorian dates should not be written in the ISO 8601 format, saying in his edit summary "simply being in that format does NOT imply the date is gregorian)."
Perhaps simply being in the YYYY-MM-DD format does not imply a date is Gregorian, but as far back as 2004 this guideline described the format as being ISO 8601. Furthermore, the date and time preference window shows the option as "2001-01-15T16:12:34"; the presence of the "T" makes it unmistakeably ISO 8601 format. The Cite template instructions also describe it as ISO 8601. Therefore it seems clear that the English Wikipedia regards dates written in the YYYY-MM-DD format as being ISO 8601 dates, and that standard requires the use of the Gregorian calendar.
If Wikipedia does not want to use the ISO 8601 standard, it should make a very clear and prominent statement as to exactly what the YYYY-MM-DD format does mean, and every statement equating the YYYY-MM-DD format with ISO 8601 format for use within Wikipedia should be expunged. So UC Bill, either your edit was wrong, or was nowhere near bold enough. --Gerry Ashton (talk) 21:23, 2 September 2008 (UTC)
(ec) There are two debateable claims in Gerry's post.
  • The ISO standard says dates before 1582 are in the proleptic Gregorian calendar if there is an agreement that to read them that way. The implication from ISO to Gregorian is therefore uncertain.
  • The encyclopedia (in article space) exists independently of all other spaces; that's a design feature, so that we can be mirrored. Therefore references in template space and talk spaces do not imply anything about article space; article space means what it says, and 2008-09-02 isn't ISO unless it says so. Septentrionalis PMAnderson 21:32, 2 September 2008 (UTC)
    • In fact, this page says Dates before the adoption of the Gregorian calendar on 15 October 1582 are normally given in the Julian calendar. If the cite templates make dates ISO, this makes them Julian; so Gerry's position is paradox-ridden. Septentrionalis PMAnderson 21:43, 2 September 2008 (UTC)
Just to point out some issues if the YYYY-MM-DD dates are NOT ISO 8601, then many points become undefined. For example, should years before 1000 have leading zeros? Should months less than 10 have leading zeros? Should years before the common era be written with BCE or should negative years be used? Is there a year 0? Should we just say that the format is only loosely defined, and all interpretation issues will have to be resolved by examining the context of the article? --Gerry Ashton (talk) 21:31, 2 September 2008 (UTC)
And doubtless the handful of dates in that format in article space before 1970 (indeed, before the foundation of WP) vary in those regards. Therefore, yes, it is loosely defined. Septentrionalis PMAnderson 21:34, 2 September 2008 (UTC)

I'll respond to some of PMAnderson's points. He wrote "The ISO standard says dates before 1582 are in the proleptic Gregorian calendar if there is an agreement that to read them that way. The implication from ISO to Gregorian is therefore uncertain."

The standard says and "This International Standard uses the Gregorian calendar for the identification of calendar days" and also "The use of this calendar for dates preceding the introduction of the Gregorian calendar (also called the proleptic Gregorian calendar) should only be by agreement of the partners in information interchange."(p. 8) I conclude that to comply with both statements, partners wishing to interchange information who have no agreement amongst them should just not express dates before the introduction of the Gregorian calendar, or should express such dates in a different format.

The partners in Wikipedia would be the editors and the readers. In the spirit of the Shrink wrap contract, we could decide that an agreement with the readers exists so long as we give them sufficient warning about the meaning of early ISO-8601 dates. If we don't feel such an agreement has been formed, the ISO 8601 format should just not be used before 15 October 1582.

  • No, the ISO 8601 standard should not be used. The format should not be entered into WP source code, and it would be nice if it were not autoformatted; but I can't do that. Septentrionalis PMAnderson 22:09, 2 September 2008 (UTC)

PMAnderson also wrote "Therefore references in template space and talk spaces do not imply anything about article space". I agree with respect to talk space, but I disagree with template space. When interpreting what appears in a template, it is reasonable to suppose the editor obeyed the tempate instructions. I also disagree with respect to Wikipedia: space. For example, the MOS tells us that the decimal marker is the full stop, not the comma. --Gerry Ashton (talk) 21:59, 2 September 2008 (UTC)

I would agree about the transcluded part of a template; but template instructions (especially, when, as here, they make no difference to what editors actually put; the ISO instuctructions relate to access dates, which do not go back before 2001) are another matter.
The decimal point is again another matter. MOS is entitled to say the decimal marker is the full stop because it normally is in English. Granted MOS says a lot of stupid things it is not entitled to say, but that's WP:OTHERSTUFFEXISTS. Septentrionalis PMAnderson 22:06, 2 September 2008 (UTC)

But I agree that, while autoformmatting exists, this is a real problem. I don't think it's as serious as the others, but it's worth mentioning, which is the chief issue here. Septentrionalis PMAnderson 22:11, 2 September 2008 (UTC)

Concerning PMAnderson's statement that the instructions of Cite templates make dates Julian, I disagree. One interpretation always applies: since access dates apply only to technology that did not exist before 1583, access dates can safely be expressed in ISO 8601 fomat. For other dates, there are two interpretations:
  1. If there is no agreement with readers to use ISO 8601 dates for pre-Gregorian dates, and considering that the Cite instructions say that YYYY-MM-DD formatted dates ARE ISO 8601, pre-Gregorian dates should be expressed in a different format and autoformatting should be prevented. If the only way to prevent autonformatting is to not use a cite template, so be it.
  1. If there is an agreement with readers to use ISO 8601 dates for pre-Gregorian dates, state in the text of the References section that the dates are, if necessary, in the proleptic Gregorian calendar, and convert any Julian dates that might appear in the actual paper sources into the proleptic Gregorian calendar.
A comment in this guideline that certain dates are generally in the Julian calendar does not overcome an explicit assertion to the contrary. --Gerry Ashton (talk) 22:12, 2 September 2008 (UTC)
No, no: it's this page that makes dates Julian. Septentrionalis PMAnderson 22:15, 2 September 2008 (UTC)
My point is that a statement in an article about which calendar was used to write the date overcomes any statement in this guideline about what calendar a date is generally in. For example, the article Mesoamerican Long Count calendar contains the sentence "Using a modified vigesimal tally, the Long Count calendar identifies a day by counting the number of days passed since August 11, 3114 BCE (proleptic Gregorian)." Clearly, in that article, August 11, 3114 BCE is certainly in the proleptic Gregorian calendar, the contents of WP:MOSNUM not withstanding. Clearly in an environment where a number of competent partners had adopted ISO 8601 and agreed to use it for pre-Gregorian dates, merely using that format would be an explicit statement that the date was in the (proleptic) Gregorian calendar. Unfortunately, the degree of agreement in Wikipedia is in doubt, and the competency of the partners varies. --Gerry Ashton (talk) 22:32, 2 September 2008 (UTC)
Precisely; and when there is no date stated, the style customary at the time and place referred to is the conventional second choice. Given the lack of firm agreement, the ISO standard for proleptic Gregorian is a distant third. Septentrionalis PMAnderson 22:46, 2 September 2008 (UTC)

Proposal. The ISO date format should not be used for Julian calendar dates. These should be expressed as DD-MONTH-YYYY or MONTH-DD-YYYY. Note that the name of the month should be spelled out in full to avoid ambiguity. For important dates, for Julian dates used after 1582, for wikilinked Julian dates put the Gregorian date in brackets after the Julian dates - just as we put metric quantities after Imperial.Filceolaire (talk) 07:30, 3 September 2008 (UTC)

Filceolaire proposes two new date formats that are seldom, if ever, seen in English documents: using hyphens as separators rather than spaces or commas together with a spelled-out month. We already have more date formats than we know what to do with; there is no reason to add two more. Also, I don't think we usually need to list both dates, because I think most people interested enough in history to care about a difference of 10 days or so that happened 400 years ago probably already know how to convert; they just need to know know which system they are starting with. More obscure calendars should give a conversion into either Julian or Gregorian. --Gerry Ashton (talk) 14:40, 3 September 2008 (UTC)
I agree with Gerry. Teemu Leisti (talk) 05:39, 5 September 2008 (UTC)
I don't see any problems using ISO 8601-formated dates before 15 October 1582. This is one of many dates seen from an international perspective. For example Denmark and Norway adopdet the gregorian calendar in 1 March 1700 ... See Gregorian_calendar#Timeline. We should use the ISO 8601-format in all templates (as used in the citation templates today). The standard says "This International Standard does not assign any particular meaning or interpretation to any data element that uses representations in accordance with this International Standard. Such meaning will be determined by the context of the application." 1 Scope and "The use of this calendar for dates preceding the introduction of the Gregorian calendar (also called the proleptic Gregorian calendar) should only be by agreement of the partners in information interchange." (3.2.1 The Gregorian calendar).
1. Proleptic: In all date parameters we should use a precise definition. It seems logical to use proleptic Gregorian calendar-dates (but this is an other discussing than what format to use).
2. Format: What ever is agreed upon on 1. it's important to use a standard way of giving dates. Then the format YYYY-MM-DD is quite obvious since it's adopted by most countries, even the US Implementation of the ISO 8601 Standard Around The World ?
Nsaa (talk) 15:28, 6 September 2008 (UTC)
The statement "This International Standard does not assign any particular meaning or interpretation to any data element that uses representations in accordance with this International Standard. Such meaning will be determined by the context of the application." must be read in conjunction with the statement at the beginning of the Scope section on the same page: "This International Standard is applicable whenever representation of dates in the Gregorian calendar . . . ." So non-Gregorian dates are entirely outside the scope of the standard, and every non-Gregorian date is a non-ISO 8601 date. "The use of this calendar for dates preceding the introduction of the Gregorian calendar (also called the proleptic Gregorian calendar) should only be by agreement of the partners in information interchange" means that users of the standard should not exchange any date before the introduction of the Gregorian calendar unless they reach an agreement to do so. If the ISO 8601 format were the only format available, the information exchange partners would just have to avoid writing about anything that happened before the introduction of the Gregorian calendar. --Gerry Ashton (talk) 18:03, 6 September 2008 (UTC)
Even amongst this august assemblage of date-representation wonks, we find a statement like "Denmark and Norway adopdet the gregorian calendar in 1 March 1700 ... See Gregorian_calendar#Timeline." Is it clear to anyone from that statement whether that 1 March 1700 represented is in the (Gregorian), (Julian) or (unspecified calendar)? It is clearly necessary to have an accepted way of representing the calendar used, even if editors rarely need make use of it. The principle is simple: Say What You Mean. LeadSongDog (talk) 18:56, 10 September 2008 (UTC)

Recently added disadvantage of DA

Expressing a date as YYYY-MM-DD may imply that it is in the ISO 8601 standard and so that it is Gregorian. It is undesirable to express dates bfore the acceptance of the Gregorian calendar in this format. Conventionally formatted dates from that era will normally be in Julian. Wikilinking such dates will cause them to be autoformatted into ISO 8601 for some users, which would constitute a false assertion they are Gregorian. Therefore such dates should never be linked.

  1. MOSNUM says above that there are only two standard date formats (international and US). This appears to allow them if unlinked.
  2. Why tell people that they may never be linked when we've already said that no dates should be linked? Tony (talk) 01:38, 3 September 2008 (UTC)
Right. Seems like it was just an oversight. — SMcCandlish [talk] [cont] ‹(-¿-)› 05:49, 4 September 2008 (UTC)
The passage quoted by Tony1 is one of three disadvantages given for date autoformatting. The statement that autoformatting of dates is now discouraged is just before these disadvantages. What is the problem? --Gerry Ashton (talk) 01:55, 3 September 2008 (UTC)
  • Exactly how 2008-09-03 can only be ISO or Gregorian I cannot understand. China, Japan, Hungary, and other countries use 2008-09-03 without it being ISO. Just because ISO is the only date format in big endianess known to some people it doesn't make it the only possibility for 2008-09-03! NerdyNSK (talk) 12:58, 3 September 2008 (UTC)
China and Japan have never been on the Julian calendar, so it must be Gregorian - you can't express "year of the rat" in ISO format.RockyMtnGuy (talk) 17:14, 3 September 2008 (UTC)
Some good advise I once heard is "don't write to be understood, write so you can't be misunderstood". Of course, 2008-09-03 is almost certainly Gregorian because it is contemporary. A date like 1752-06-15, standing alone, might be either. The author might or might not have intended to follow the ISO 8601 standard. The problems when such a date appears in a Wikipedia article are
  1. We have provided some ambiguous notice, but not a clear and strong notice, to our readers and editors that the English Wikipedia follows the ISO 8601 standard. Thus some, but not all, readers and editors will think that it is, or should be, a Gregorian date
  2. The English Wikipedia provides no notice to suggest it isn't governed by ISO 8601, nor does it suggest what rules should be followed in formatting such dates if they aren't ISO 8601.
  3. Wikipedia articles are copied to other places in many different ways, so any notice we might give (either way) is likely to become disassociated with the text of the article (unless it is contained within the text of the article).
The safest way to prevent misinterpretation would be to ban the use of YYYY-MM-DD dates before 14 September 1752 in the English Wikipedia, since that is the date England and the British Empire adopted the Gregorian calendar. At the time, England and the British Empire accounted for the majority of English-speakers. This would also avoid most formatting issues; common sense and ISO-8601 coincide for years greater than 999. The only formatting issue I can think of that would come up after 1752 is whether single-digit days and months must be padded on the left with a zero. --Gerry Ashton (talk) 14:59, 3 September 2008 (UTC)
The only safe way is for the author to state which calendar a date is on if there is some doubt. For instance, the Russian October Revolution happened in November 1917 on the Gregorian calendar because Russia was still on the Julian calendar at that point in time, so that fact should be noted - and is. The most confusing case is Nova Scotia, which switched to the Gregorian calendar in 1605, reverted to the Julian calendar in 1710, and then changed to the Gregorian calendar again in 1752. The other Canadian provinces changed calendars in various different years so, you can't generalize on any particular rule - except that if you use ISO format, it should always be on the Gregorian calendar.RockyMtnGuy (talk) 17:14, 3 September 2008 (UTC)
That would be imposing bad writing where it is unnecessary, an endemic problem with this page. The convention, understood by all literate readers, that an article will use the dating format common at the time and place being written about will normally suffice. The only major exceptions are in Northern Europe (and North America) between 1582 and the local adoption of the Julian calendar - because contemporary usage itself varied. Septentrionalis PMAnderson 18:09, 3 September 2008 (UTC)
PManderson, please provide a source, such as a writing manual or book about how to write history, that backs up your opinion. Such a source would be better if it described how to write for general audiences, rather than for historians. Also, there are many parts of the world that went from using a local calendar (neither Julian nor Gregorian) to using the Gregorian without ever using the Julian calendar. Surely one would not use the local calendar when writing in English, unless there was doubt among scholars about how to convert from the local calender to Gregorian. --Gerry Ashton (talk) 18:19, 3 September 2008 (UTC)
I said format. Countries who did not use the Roman months at all before the colonial era are another question entirely. If I come across a ready reference on the matter, I will consult it. Septentrionalis PMAnderson 18:52, 3 September 2008 (UTC)
For clarity, PMA, do you really posit that it is bad writing to identify the calendar used? LeadSongDog (talk) 21:40, 3 September 2008 (UTC)
Yes. Eschew needless words. Septentrionalis PMAnderson 23:55, 3 September 2008 (UTC)
If it's not clear which calendar is being used, words used to make it clear are not needless. Teemu Leisti (talk) 05:38, 5 September 2008 (UTC)
So we'll just replace all the history articles with "Men are born, suffer and die"? I'd guess there might be some pushback on that.LeadSongDog (talk) 22:09, 8 September 2008 (UTC)

Date autoformatting without wikilinks

Since UC Bill has already submitted a patch for this bug that will autoformat dates without having to wikilink them, that eliminates objection #2 (overlinking) and since objection #3 is just nonsense (the order in which the date elements appear has nothing to do with the calendar being used — writing "the October Revolution occurred on 1917-10-25 (Julian calendar)" is no more or less confusing than writing "the October Revolution occurred on 25 October 1917 (Julian calendar)") that leaves only objection #1 as a valid complaint. Rather than throwing the baby out with the bath water (i.e. eliminate autoformatting entirely), I think a better solution would be to simply use a default format for dates, either across Wikipedia or on a per-article basis. Then, everybody will see the same format unless they explicitly decide to change it, in which case we should honor their choice. Taking away the choice is not a good option. Since Tony1 is objecting to the submitted patch (despite having indirectly asked for it) I think we should build consensus around a technical solution to this issue, and have the system administrators of Wikipedia accept the submitted patch. --Sapphic (talk) 02:28, 4 September 2008 (UTC)

  • Problem 3 is more serious than you make out; ahd requiring (Julian date) in all pre-Revolutionary Russian articles is a real cost. But I've added problem 4: Idiom is "October 25, 1917, was the first day of the October Revolution...." and "25 October 1917 was the first day of the October Revolution..." Dealing with this second comma, which should be present in "On 25 October 1917,... " requires either complex and unidiomatic rules for writing dates, or an output program which understands English syntax. Septentrionalis PMAnderson 02:46, 4 September 2008 (UTC)
Most people aren't even aware of the existence of the Julian calendar, let alone the dates on which it was abandoned (or readopted) in various countries. If there is any chance for confusion, even for dates in the more distant past (which more knowledgeable folk would immediately recognize as pre-Gregorian), then the calendar should be specified. That's just common sense — we're writing an encyclopedia for the layperson, not a textbook for historians. As for the formatting requirements, using <nowiki> is a perfectly reasonable way of enforcing whatever formatting you like, if the autoformatting screws things up. The only debatable issue here is what the default date format should be for non-logged-in users. --Sapphic (talk) 02:57, 4 September 2008 (UTC)
This is absurd. By that reasoning, we should write "Columbus discovered America on October 12, 1492 (Julian calendar)" and "Caesar was assasinated on March 15, 44 BC (Julian calendar)". It would be better to delete this page and destroy autoformatting than to write so badly. Septentrionalis PMAnderson 03:03, 4 September 2008 (UTC)
It is neither absurd, nor is it relevant to the point of this section, which is to discuss autoformatting of dates sans-wikilinks. The first mention of a date in any article referring to a calendar other than the one that the overwhelming majority of readers will be familiar with should contain a reference (perhaps in the form of a footnote, if text seems too obtrusive) explaining that the dates are according to some other calendar. That's the last I have to say on that, here. Unless you have some point about calendars that's actually relevant to date autoformatting, please discuss it in one of the other sections. --Sapphic (talk) 03:10, 4 September 2008 (UTC)
PMAnderson is right to raise this issue and to stress that it is important to the precision of how we present chronological facts, and the succinctness with which we express them. However, it sits beside at least two other major issues: (1) the increasingly obvious fact that because DA prevents WPians from viewing what is displayed for their readers, they regularly insert inconsistently formatted dates; and (2) the utter triviality of which goes first, month or day. These are no-brainers, I put it to you. Tony (talk) 03:22, 4 September 2008 (UTC)
I agree that the first issue is important, and although I have a different opinion on what the solution should be, my main point was that it is completely irrelevant to the date autoformatting. Most people see 1492-10-12 and 12 October 1492 as the same date written two ways, not as one proleptic Gregorian date and a different Julian date. Some people might remember "something about the calendar changing" from history class, but most would be hard-pressed to give any specifics, either for their own country or in general. I think the right thing to do is to have a footnote (or in-article text, in the more interesting cases where it's more relevant, like with October Revolution) at the first non-Gregorian date mentioned in an article, and take the opportunity to educate people a little. But regardless, I don't think it's relevant to most people's perception of YYYY-MM-DD format.
Now, my main point was that we could answer objection #1 by deciding on a default date format — which is effectively what we're doing anyway, by disabling the autoformatting by unlinking dates. Except we're also taking away the ability to specify a preference at all, even for logged-in users. If we just come up with some reasonable way of specifying a default, then we just apply the patch and we can unlink any dates we want, but still give logged-in users (who may just be regular readers who want to customize their settings, style, etc.) the ability to override the default date format, if they want. I'm thinking something along the lines of how the DEFAULTSORT thing works. That should work just fine with the parser cache. The parser cache is why we can't do something even better like use the reader's browser's locale setting. --Sapphic (talk) 03:56, 4 September 2008 (UTC)

I took the trouble to read the Julian calendar" article that PMAnderson linked to when he wrote "This is absurd. By that reasoning, we should write . . . 'Caesar was assasinated on March 15, 44 BC (Julian calendar)'. " [emphasis added] That article contains this fascinating sentence (with added emphasis): "A curious effect of this is that Caesar's assassination on the Ides (15th day) of March fell on 14 March 44 BC in the Julian calendar." I also checked in The Oxford Companion to the Year by Blackburn and Holford-Strevens; they don't provide an algorithm to convert roman to proleptic Julian dates, but they do make it clear there was immense confusion between 45 BC and AD 8. --Gerry Ashton (talk) 03:54, 4 September 2008 (UTC)

Yes, that is discussing the calendar (and making the point that the Julian calendar actually in effect in AUC 710 appears to be a day off from the proleptic Julian calendar as presently instituted.) I have no objection to this; I would object to the same wording in Julius Caesar, however.
What Julius Caesar does do is to date the assasination to the Ides of March, and refer to the Roman calendar; this is a trifle complex and flamboyant, but I presume at some point they quoted "The Ides of March have come." Mentioning the Julian calendar in the same passage would be chaos. Septentrionalis PMAnderson 14:17, 4 September 2008 (UTC)

I would like to describe for Sapphic a situation that I think is analogous to writing 1750-06-15 Julian calendar. Imagine a notary public acts in the state of Texas in the summer; the legal time in effect is Central Daylight Time. The notary dates the certificate "July 4, 2008, Greenwich Mean Time". So which prevails, the normal customary time for legal documents in Texas, or the notary's notation about the time zone used? The notary has created a mess which should have been avoided. --Gerry Ashton (talk) 04:00, 4 September 2008 (UTC)

Actually, we can use the browser's locale setting, even with the parser cache. We just need to have the preference-checking code look at the browser settings before deciding which cached copy of the page to use. Then there would be no need to decide on page-specific defaults, since we'd know what locale preferences each reader has, which would be far better than trying to use "strong national ties" as a proxy for that information. The point, after all, is to format the dates according to the preferences of the reader, and the only reason we've even been discussing the topics of the articles at all is because we had been thinking that's the best we could do. But we can do better. So here's my proposal:

Proposal

1. Put a hold on date unlinking. There's no rush on this, and it can be done in an automated way after we've implemented the rest of the changes described below.

2. Apply the submitted patch to disentangle date autoformatting from wikilinking. This will address one of the major objections to the current autoformatting: overlinking.

3. Create, submit, and apply another patch to set a reasonable default format for non-logged-in users, possibly based on the browser's locale setting. This will guarantee that users see a consistent date format across the entire site, and that the format they see is the one they prefer (or at least the one that their browser is configured to use.)

4. Resume date unlinking, but in an automated way.

I think that should address all of the major concerns with date autoformatting, and in as least-disruptive manner as possible. Thoughts? --UC_Bill (talk) 15:42, 4 September 2008 (UTC)

First, this proposal presumes that most readers want to see the date format the same, no matter what the style of the article is. Some readers might prefer to have the date format match the style of the article, that is, month-first with "color" and day-first with "colour".
Second, the proposal says to stop in the anticipation that one patch will be accepted and another will be created and accepted. Some of us might feel that there is little chance that these patches will ever go into production, and be tired of waiting.
Third, I can't really understand the proposal because it uses the word wikilinking too much. Please provide examples of what dates, in the edit window, would look like before and after step 4. --Gerry Ashton (talk) 15:58, 4 September 2008 (UTC)
Indeed, I confirm Gerry's points. My experience with such technical expeditions is that they hit snags at many points and take a long time to come to nothing. Sorry to be a pessimist; I'm not usually so. Bill, you still haven't addressed the major concern that to manage dates on WP, editors need to see what their readers see. The idea of defaults based on locale is full of issues: which parts of Canada will have which format? What about US military articles that their authors intend to be in international format? It's just all too complicated and likely to provoke yells of protest from afar. That is one reason that the community is embracing the simplest approach, which, in this case, happens to eschew the artefacts of computer programming. Finally, do you really care that much about whether month or day comes first? Tony (talk) 16:07, 4 September 2008 (UTC)
The first assumption is true. Some readers may want the date format to change from article to article, but they will be very much in the minority. Most people prefer consistency (in fact that's why there's an argument about the date format in the first place) and find changes in style jarring. Second, it is disruptive to unlink dates, because it breaks a feature that many people (myself included) really like. Date autoformatting was introduced to solve a very real problem (check the archives going back several years if you don't believe me) and though it was implemented in a less than ideal way, that can be fixed. Third, the examples are that dates will be autoformatted in the same way they are now, except without having to make them links. If you write 4 September 2008 in an edit box, it will show up as September 4, 2008 if that's your preference (specified either by logging in and selecting that format, or by using a browser with a locale setting that indicates that format.)
Since the first point is the most urgent, I'll explain more why this is important. By unlinking dates, we already have to decide on what date format to use. Why not simply put the wikilinked dates into that format, then? It will have the same effect (making the format consistent for non-logged-in users) but won't annoy users who have specified a preference. The only downside is that the dates will still be links — but that's precisely the issue addressed by the second point of the proposal. So here is the modified version of the first point:

1. Put a hold on date unlinking. There's no rush on this, and it can be done in an automated way after we've implemented the rest of the changes described below. While we're waiting to implement the remaining steps, we can continue to standardize the date format while retaining the link for autoformatting purposes and this should meet with less resistance from other editors than widespread unlinking would.

That make things clearer? --UC_Bill (talk) 16:12, 4 September 2008 (UTC)

UC Bill wrote "Third, the examples are that dates will be autoformatted in the same way they are now, except without having to make them links. If you write 4 September 2008 in an edit box, it will show up as September 4, 2008 if that's your preference (specified either by logging in and selecting that format, or by using a browser with a locale setting that indicates that format.)"
That will never work. No computer on earth (or elsewhere in the solar system) can figure out whether September 4, 2008 should be reformatted for a reader when the date is in running text and is not marked up. Examples of cases when it should not be reformatted are
  1. direct quotes
  2. articles about date formats
  3. groups of words and numbers that have the same character sequence as dates, but are not actually dates. --Gerry Ashton (talk) 16:50, 4 September 2008 (UTC)

Tony1, editors do not need to see what readers see, we just need to know that readers see something consistent. That goal can be satisfied by using a reasonable default format rather than manual enforcement.

Gerry, the <nowiki> tags work as expected with the new patch, so editors that need to override the autoformatting for whatever reason can do so (just as they can use the | within a link to do so with the link-based autoformatting now.)

Some of the points being raised here aren't actually valid goals, they're work-arounds based on flaws in the old system. The problem isn't that editors might see something different than readers, it's that readers were seeing inconsistent formats within the same article, and the autoformatting made it difficult to fix that. If we fix the inconsistency with a technical solution, the problem goes away and we don't need to see what readers see. Similarly, the suggestion of tying date format to the subject of the article was a workaround based on the fact that we had no way of knowing the preferences of unregistered users. There are other workarounds that are much more reliable, such as the browser locale settings. If an American reader visits the article on Charles Dickens, the dates should be displayed in American format (assuming that's what the reader prefers) — it's just that the existing system has no way of doing that, so we were stuck with per-article standards. We can do better than that.

Even if the goal is to eliminate autoformatting, it still makes more sense to standardize the date formats first and to disable autoformatting later (which is a trivial one-line change to the configuration file for MediaWiki) than to do it in the disruptive manner it's currently being done. --UC_Bill (talk) 17:07, 4 September 2008 (UTC)

UC Bill wrote "Gerry, the <nowiki> tags work as expected with the new patch, so editors that need to override the autoformatting for whatever reason can do so (just as they can use the | within a link to do so with the link-based autoformatting now.)"
This says "work", present tense. Does that mean the patch you have already submitted does autoformatting on dates that have no markup?
Changing from no markup meaning leave the date alone to no markup meaning autoformat the date means that an unknown number of dates (and character sequences that resemble dates) in existing articles will break, and there is no automated way to find them. That is not good. --Gerry Ashton (talk) 17:29, 4 September 2008 (UTC)
Yes, the patch for "Provide preference-based autoformatting for unlinked dates" does autoformatting on dates that have no markup. That's what was requested. You raise a good point about suddenly autoformatting dates that may have intentionally been left unlinked to prevent it. So here's a modified proposal:

More general proposal

1. Stop unlinking dates. This is disruptive activity that can be handled in a better way.

2. Gather data on dates (format, linked or not, etc.) in current articles.

3. Place <nowiki> tags around currently unlinked dates. This will not harm anything, and will prevent the submitted patch from unexpectedly autoformatting previously unformatted dates.

4. Apply the patch. At this point, we can go past merely deprecating the linking of dates, and outright forbid it (except where the date really should be linked for information purposes, obviously.)

5. Resume unlinking. At the same time, the date formats can be standardized according to whatever criteria has been deemed appropriate. At this point, we can evaluate the level of disagreement over the standards.

6. Decide whether to keep autoformatting or not. It can be disabled with one change to the configuration settings, or kept in place if the changes detailed above have satisfied the consensus.

I suggest that a WikiProject be formed to handle many of these tasks, and that bots be employed to take care of much of the tedious work of gathering data, unlinking dates, and standardizing date formats. I'd be happy to help out with that work, as I'm sure several other editors would. --UC_Bill (talk) 18:47, 4 September 2008 (UTC)
I have started a project to address these issues. Please feel free to join the project and to contribute to its further development. --UC_Bill (talk) 19:44, 4 September 2008 (UTC)
Is this conversation now gonna continue in the project or here? — OwenBlacker (Talk) 22:20, 4 September 2008 (UTC)
I think everything except step 1 should be discussed on the project talk page. I think it's important that we stop unlinking dates for the time being, since it's disruptive and there are other points to address first. The text of MOS should be changed to indicate that date autoformatting is being re-evaluated, and that editors should refrain from either mass linking or mass de-linking of dates, until a consensus has been reached on what to do. Once a consensus has been reached, the rest of the plans (possibly including the permanent disabling of date autoformatting) can be carried out rather easily. --UC_Bill (talk) 22:26, 4 September 2008 (UTC)
      • Support I think focusing on step 1 (stop unlinking dates) makes sense. I believe UC_Bill is on the right track and will be able to address the stated goals of the anti-DA group, namely to remove the linked dates and ensure the *default* format for non-editors is consistent. As a beneficial side-effect, it removes the need to edit war over country formats. Personally, I suggest we go down the template path similar to {{dts2}}, since that will be more obvious to editors while still doing the right thing for non-editors and editors alike. dm (talk) 22:48, 4 September 2008 (UTC)
  • Oppose I oppose the creation of software that will inevitably create errors by autoformatting things that ought not to be autoformatted, just so some people will see their preferred date order. I see no need to address this one preference when so many other preferences, such as units of measure, spelling, and idiomatic expressions, can't be addressed (nor should they be). Since the anticipation of this software is the only reason not to delink articles, I say go ahead and delink them. --Gerry Ashton (talk) 23:06, 4 September 2008 (UTC)
Gerry, that's not what we're discussing here. We're simply saying that we should hold off on any more date unlinking until we have a more complete plan in place. That plan may still ultimately end with us disabling date autoformatting, but that is something we can discuss on the new project page. Even if the goal is to eliminate date autoformatting, it still doesn't make sense to go around unlinking dates right away. Doing that will just cause conflicts that can be avoided if we're more systematic about this. --UC_Bill (talk) 23:33, 4 September 2008 (UTC)
UC Bill's patch does not require linked dates to work. Date delinking won't prevent the patch from being implemented (other reasons might). If date delinking were to proceed at once, and later UC Bill's patch went into production, all that would happen would be a relatively small proportion of our readership (those who have accounts and select a date preference) would temporarily see some dates in something other than their preferred format. I just feel that the people who selected a preference for dates (as the system exists today) have backed a losing horse, and it's time for them to toss their bet receipts in the trash. --Gerry Ashton (talk) 04:30, 5 September 2008 (UTC)
  • Support a temporary halt to date unlinking, so that consensus on a long-term plan can be reached. --Sapphic (talk) 02:46, 5 September 2008 (UTC)
Guys, we've been long-terming for a long time. You've come in right at the end of a process that has evolved over more than two years. There are fatal flaws in the proposed tech solution, just as in DA itself, the removal of which has been met by positive reaction, sometimes enthusiasm, by the community. There is insufficient information about the proposal. Nothing has been tested, and is sure to encounter significant snags. More importantly, none of you has addressed the basic issue: there WAS no problem in the first place. Tony (talk) 03:14, 5 September 2008 (UTC)
Are you insane? There very much WAS a problem that DA solved. Go look through the archives. In any event, unlinking should NOT be the first step in disabling DA. It should be part of the cleanup process that takes place AFTER DA is turned off. It's very easy to disable DA, that option is built-in to the MediaWiki software. Since you obviously can't build consensus for doing that, you're trying to take the back-door approach and effectively disable it by unlinking all the dates. You keep talking about your actions being based on a "community decision" but I don't see anything like that here — I see a lot of people who can't agree on the approach to take. Which is why some of us are saying to back off and approach this the right way. --Sapphic (talk) 03:25, 5 September 2008 (UTC)
  • In defense of Tony, I am quite certain he is sane. He has however, been active here for years and has seen consensuses repeatedly develop here by those who were participating at one time or another. And then, when a decision is finally made to actually do something that corrects past mistakes, fresh blood comes here. Why? Not surprisingly, those changes in articles brings in experienced editors (but who are relative newcomers here) exhibiting some military-grade WTF?!? reactions. Clearly, there are a number of editors who oppose this. But a “consensus” on Wikipedia does not mean that 100% of editors are in complete agreement—and it never did. It only means that this issue was deliberated and debated for quite some time, editors’ arguments were considered, and a general consensus arrived at. And that consensus? That autoformatting was a bad idea in the first place that only swept editorial problems under the rug rather than really fix anything. Greg L (talk) 03:42, 5 September 2008 (UTC)
  • And when they show up to protest, a broader consensus needs to be formed; that's exactly what MOS is so bad at. Septentrionalis PMAnderson 22:30, 5 September 2008 (UTC)
  • Oppose For the reasons stated above by Tony. This is a simple, simple issue (choose a date format most suitable for the subject matter), made more complex than it needs to be largely because it’s just human nature to dislike change and even more like human nature (and quite understandable) to not want one’s pre-existing work screwed with. But an utterly ridiculous amount of arguments and battles have raged over this silly issue. The growing realization is that the autoformatting tools were an ill-thought-out tool that made “pretty” results only for editors and didn’t benefit regular I.P. users (99.9% of Wikipedia’s readership) one twit. Worse, the tools were masking editorial problems in our articles because we editors couldn’t see the problems (but I.P. users could see them). And the second problem with the auoformatting is it added needless links to lists of random trivia that more often than not, had nothing whatsoever to do with the article. In Kilogram, since it was a French invention and didn’t have strong ties to the U.S., I chose to format the dates like “7 April, 1799.” You know how many editors over the last year have changed the date format or complained about it? Or even commented on it? Zero. Zip. Nada. Not even some 14-year-old American junior high student. No one gives a crap about this issue except for we editors. We’ve got to put and end to this argument and get on with life; the issue is too trivial to bother wasting Wikipedia server storage space on. Greg L (talk) 03:28, 5 September 2008 (UTC)
  • Did you even read what was being discussed, or did you just give a knee-jerk reaction? The ONLY issue up for debate in this section is whether to stop the unlinking of dates, until a better plan can be reached. It's fine if you're against DA (I think you're ignorant of the history of Wikipedia in that case, but that's just me) but the FIRST step should NOT be unlinking the dates. That's the cleanup after DA is disabled. --Sapphic (talk) 03:40, 5 September 2008 (UTC)
Sapphic, forgive my cynicism about change at the programming end: I've learned through hard and long experience not to trust any suggestions that tech-change will happen any time in the next decade, or that it will happen eventually in a functional way. Sorry. No way: that's how we allowed the DA debacle to happen in the first place. Editors here on the ground need to take charge of the matter on the local level, and that's exactly what they're doing.
Indeed, Greg. It's striking how much heat the DA thing has generated when you look at how superbly well we've learned to manage the engvar issue at WP—indeed, engvar is one of the project's most outstanding administrative achievements—to bring together the whole English-speaking world and largely avoid in-fighting about which variety to use where. This, I believe, has resulted from (1) significantly improved wording in the MoS engvar guideline, (2) a realisation by editors that, "hey, it's fine to apply this binary choice to article-consistent spelling and heck, the varieties are so damned similar", and (3) an evolving maturity in the way people interact on Wikis, which were new and raw at the time the DA urchin was concocted. Now we see the same plot evolving, several years too late, for date formatting—also binary in which the difference between the two versions is utterly trivial.
Right now, we need to sort out the non-anglophone-country issue for the choice of date format, which should have been done years ago, but which was never addressed because autoformatting concealed the issue from us (but not our readers). Just another reason that we all need to see the same text that our readers do. And ... it's less work.
Oh, and one more thing: do remember that WikiMedia's software serves hundreds of sites all over the world, not just WP. This may partly explain why there's a reluctance to change anything lightly at that end of the chain. But individual editors can still act for the good of their readers ... Tony (talk) 03:50, 5 September 2008 (UTC)
  • Sapphic: Probably a bit of both. Fortunately, I caught the essence of your proposal (stop the unlinking of dates), and voted properly (for me) on that salient point. The links are worse than worthless in my opinion as they result in over-linked articles and direct readers to long, tedious, random lists of trivia. If I’m reading an article on particle physics, and read that Enrico Fermi discovered something on 16 October, I might be hoping to learn more about Fermi’s discovery when I click on the link. I’d be quite disappointed to go to a huge list of trivia, which includes such fascinating events as “[On this date in] in 1925 Angela Lansbury, English-born actress [is born].” (*yawn*) Greg L (talk) 03:52, 5 September 2008 (UTC)
  • Comment I think we all need to take a look at WP:CIVIL and WP:OWN again, there's no need for name calling or sanity questioning in either direction. However, I think there's a strong reaction from people that unlinking dates was *not* the best way to go about this either. I read quite a ways back on the entire thread, and one of the oppose people (Anderson?) said that mass unlinking would get exactly this sort of reaction, so no surprise. What we have now is a proposal to halt the mass unlinking while more editors show up and get involved. Hopefully out of this comes a disentangling of the issues so that we can address them separately. Sorry if it's not the result you would have picked, but I think it's the wiki way.dm (talk) 03:59, 5 September 2008 (UTC)
No, the issues have been well disentangled, churned, widely debated for a long long time. You are trying to scupper a reform that you're personally uncomfortable with (I've no idea why) by implying that the issues have not been thoroughly debated—please try to understand that they have. I can point you to lots of evidence if you wish, and lots of evidence that the community wants the reform. I don't think PMAnderson would want to be counted as an "oppose"; he has expressed a more nuanced view. "I think there's a strong reaction from people that unlinking dates was *not* the best way to go about this either." There has been a strong reaction from a few isolated locations, and many of the original people who were opposed have changed their tune over the past two months. On the contrary, the reactions have been either positive or neutral. Selected comments are gathered here, but this has not been updated to include more recent enthusiastic support by WPians out there. Dmadeo, I have a sneaking suspicion that you're stalking me, but I'll readily withdraw this comment and apologise if it's untrue. Tony (talk) 04:53, 5 September 2008 (UTC)
Wow. Stalking? Is that AGF, have I said something about you? I think I'm going to walk away from this discussion for a few days, I thought I was trying to bring a little bit of civility back into this discussion, but perhaps that's not what the consensus wants.... dm (talk) 05:12, 5 September 2008 (UTC) [Dmadeo has explained the misunderstanding on my talk page. All is resolved on that front.]
  • Also oppose. I especially oppose a change to the software that would start changing the way unmarked-up dates would be displayed. As the proposal's point 3 says, "Place <nowiki> tags around currently unlinked dates. This will not harm anything, and will prevent the submitted patch from unexpectedly autoformatting previously unformatted dates." At least unlinking dates is an explicit procedure, and will have the effect of making the date being displaed exactly the way the editor wants to, while the proposed change to the software will make unmarked-up articles display differently without any editor having touched the articles in question. The workaround, placing the <nowiki> tags around all currently unlinked dates, will not be obvious to editors. This is too draconian.
And by the way, try to keep your calm, people. Calling people insane, accusing them of stalking, etc., are not conducive to a civil, productive debate. Teemu Leisti (talk) 05:36, 5 September 2008 (UTC)
  • Support. Stop delinking of dates now. I for one want my dates to be autoformatted. I don't want to read "September 5, 2008" even in an article about an American subject and I'm sure many Americans don't want to read "5 September 2008" even in articles about non-American subjects. Until this is resolved, please stop delinking. -- Necrothesp (talk) 15:06, 5 September 2008 (UTC)
Necrothesp, may I point out two things? First, your readers out there, British and North Americans all, have been viewing both formats—the raw ones we key in—all along. Your personal preferences are noted, but many editors are keen to place the readership above those personal concerns. That is a commendable and decent approach, I believe, and underpins the professionalism that makes WP a powerful authority on the Internet. Second, I presume you're British. Many British newspapers use the so-called US date format. My own daily newspaper does. Most of its readers probably don't even notice that this is different from the prevailing international format used in my country. On the other side, the US military uses international date format; a quick browse through the thousands of US military articles will show that this is reflected in our articles. Try this list of articles, for example, although where DA is still used in those US articles, you'll have to select "no preference" for your dates ("my preferences") to see what your readers see. Please let me know your reactions. Tony (talk) 15:25, 5 September 2008 (UTC)
  • Wow! Necrothesp, your response speaks directly to the heart of the problem with autoformatting. Editors probably represent only 0.1% of Wikipedia’s readership. That means 99.9% of our readership don’t have accounts and haven’t set their user prefs. And what does that mean? That date formats like [[2005-05-05]], which produces a wonderful, gorgeous, spelled-out date like 5 May 2005 for you, produces a crappy looking 2005-05-05 for virtually our entire readership. Even the more “benign” formats, like [[May 5]], [[2005]] in an article on France gives *you* what you want to see (sorry, that’s not the litmus test here) but produces a wholly inappropriate date format (May 5, 2005) for virtually everyone else in a European-related article. Do you think that is a good thing? I don’t. These God-forsaken tools have only been masking editorial problems the entire time we’ve been using them.

    We editors have got to wake up to the fact that Wikipedia does not exist to produce custom content that only we editors can see so we can enjoy a privileged version of Wikipedia and marvel at our handiwork and pat ourselves on our own backs for a job well done. If we’re not looking at the exact same content that regular IP users see, we’re just sweeping editorial problems (like I illustrated above) under the rug. We need to fix these shortcomings in our articles and the first step towards that end is to unformat dates.

In an article on the speed of light, how does a link to random trivia like this got to do with anything?
Beyond the issue of *formating* of dates, I just can’t see the wisdom of cluttering up articles with even more links if all they do is take readers to random lists of historical trivia that have nothing whatsoever to do with the subject matter at hand. If I were writing an article on the speed of light, I might link to meter, for instance. Such topical links properly invite exploration and learning. When articles are over‑linked, we’re just numbing readers to links, such as when we add a link within speed of light that takes the reader to an article that says…

[On this date in] 1600 - Tokugawa Ieyasu defeats the leaders of rival Japanese clans in the Battle of Sekigahara, which marks the beginning of the Tokugawa shogunate, who in effect rule Japan until the mid-nineteenth century.

By the way, that is a real date link that is in the speed of light article. Just because one can add links to an article, doesn’t mean one should. If readers want to read a mind‑numbing list of random trivia that occurred on Oct. 21st, they can type it into the search field. When we over‑link, we just turn articles into a giant blue turd.
This issue is resolved. The general consensus is that the proper course of action is to improve Wikipedia’s articles for the maximum benefit of our IP readership—not us. The arguments of other editors here, which were along the lines of your argument above (“I want a special view of the data just for me ‘cause I don't want to read ‘September 5, 2008’ even in an article about an American subject”) had been considered and discredited. Greg L (talk) 15:48, 5 September 2008 (UTC)
I have to say that I've never, ever seen a date on Wikipedia formatted as [[2005-05-05]]. Have you? And can I ask you to drop the patronising tone please - what I am objecting to is the sudden delinking without any form (that I've noticed) of public discussion beyond discussion on a talkpage that very few people are actually going to look at (so not really any form of "general consensus"). Yes, I am aware we write for our readers - strange as it may seem, somebody who has contribued as many articles as I have to the project is intelligent enough to realise the point of said project and how the whole linking thing works. "The arguments of other editors here, which were along the lines of your argument above (“I want a special view of the data just for me ‘cause I don't want to read ‘September 5, 2008’ even in an article about an American subject”) had been considered and discredited" - I should reread that statement and see just how arrogant and patronising it sounds! Who has "discredited" them? You? Autoformatting was for all of us to make articles easier to read for all of us, not just for me as you so patronisingly claim. -- Necrothesp (talk) 19:26, 5 September 2008 (UTC)
  • “I have to say that I've never, ever seen a date on Wikipedia formatted as [[2005-05-05]]. Have you?”  Uhm… okaaaaay… try logging out so you look like a regular IP user; that’s sorta my point. In your 15:06, 5 September 2008 post above, you wrote “I don't want to read ‘September 5, 2008’ even in an article about an American subject…” so I’m confident given that you are so strident on this topic, that you don’t routinely work and play on Wikipedia while logged out. That practice masks this particular format and all the other formatting issues I’m trying to point out here.

    So you can understand, here is some sample text. Take a look at it while logged in and while logged out.

Do you see now? To save other editors the effort of having to log out, here is how the above was coded:

During the French Revolution, the Estates-General convened in Versailles on [[May 5]], [[1789]] and lengthy speeches by Necker and Lamoignon… and then on [[2005-05-05]] a memorial was built on the site…

And here is what virtually every single user of Wikipedia sees (because they aren’t *special* like we registered editors):

During the French Revolution, the Estates-General convened in Versailles on May 5, 1789 and lengthy speeches by Necker and Lamoignon… and then on 2005-05-05 a memorial was built on the site…

But for someone logged in with “Euro-style” preferences like Necrothesp likes, here’s what they see:

During the French Revolution, the Estates-General convened in Versailles on 5 May 1789 and lengthy speeches by Necker and Lamoignon… and then on 5 May 2005 a memorial was built on the site…

It should be quite clear that both date formats 1) are inappropriate for use in an article on the French Revolution, and 2) they appear lousy or awkward for the vast majority of Wikipedia’s readership, and 3) editors who are logged in and have their user pref settings adjusted to their suiting can’t even see where these problems exist! As editors who are largely responsible for this stuff, we’ve been marching along all fat, dumb & happy, totally oblivious to what we’ve been making the vast majority of our readership look at. As I stated above, we editors should never use tools that allow us to look at regular editorial content in a way that is at all different from what everyone else has to look at. The only preferences settings that affect what we see should be date offsets for when edits are made—that sort of thing.

As for your “what I am objecting to is the sudden delinking without any form (that I've noticed) of public discussion…”, please see the 19:37, 5 September 2008 post below by 86.44.27.255.

And as for my patronizing tone, it is not a personal attack, nor have I been uncivil to you, so you may complain all you want but you are wasting your time when you presume to tell me how I may think or express my thoughts. The patronizing tone comes out, apparently, when I deal with a registered editors, who has clearly set his user preferences so his dates are Euro-formatted, and then writes about how he has never seen a date formatted like [[2005-05-05]]. Is this concept of what “you” see v.s. what “virtually everyone else” sees getting through to you yet? This realization of what we’ve been doing with these autoformatting tools has sunk in to a sufficient number of editors now. Hopefully, you will understand that what you are seeing is not what IP users see—but should. Greg L (talk) 21:23, 5 September 2008 (UTC)

Many editors who use the citations using the cite or citation templates write the dates in the [[2005-05-05]] format. Sometimes the square brackes are included, sometimes not. --Gerry Ashton (talk) 19:42, 5 September 2008 (UTC)
  • Oppose. The third step is the most unappealing and potentially problematic. The proposal says that it will "not harm anything", but I'd say instead that it will not harm anything if executed correctly - otherwise, it will get really disruptive. However, I can't agree more with the first step: do not unlink dates yet. The final setup must be decided first. If anything, wikilinks say "this is a date, and you're allowed to change its format" (it is not a quote, etc.), which could be useful to bots once we decide what (if anything) to do e.g. with the existing wikilinked ISO 8601 dates. GregorB (talk) 18:20, 5 September 2008 (UTC)

Unlinking

What's with the new wording? Editors can't form a consensus to unlink dates on the relevant article talk pages? Some Wikiproject which (no disrepect but) I've never heard of and which currently has 5 members gets to tell us when it's ok? Very strange. It's depreciated. Of course editors can unlink dates provided they don't edit-war about it, use the talk page if reverted etc. Where is the discussion that led to this bizarre wording? 86.44.27.255 (talk) 06:14, 5 September 2008 (UTC)

It seemed to be a recent addition without discussion so i have reverted. 86.44.27.255 (talk) 06:21, 5 September 2008 (UTC)

Correction: without consensus. 86.44.27.255 (talk) 07:36, 5 September 2008 (UTC)
This has been discussed for years, and widely. Please see the rest of the talk page before firing off a salvo. — SMcCandlish [talk] [cont] ‹(-¿-)› 21:41, 6 September 2008 (UTC)
Better advice would be to tell me to include a diff to whatever the hell it is i am talking about [1]. 86.44.21.173 (talk) 11:43, 9 September 2008 (UTC)

December 09, 2004

Expecting that no one will disagree, I've inserted a note about not doing this. In the process of auditing the dates of many articles, I've found quite a few examples of this. Please say so here if there's a problem. Diff. Tony (talk) 09:07, 5 September 2008 (UTC)

Good point. — SMcCandlish [talk] [cont] ‹(-¿-)› 21:41, 6 September 2008 (UTC)

When/how did this happen

Whoa, whoa. Hang on a minute. I've only just become aware of this when people started to delink dates from articles on my watchlist. This is a huge change which seems to have only been discussed by a few parties who are particularly interested in dates. It's the biggest change to editing procedure I remember in my four years of editing, but where were the attempts to alert other editors to the discussion? Where was the prominent note on the front page? I don't agree with it and I'm sure many other people don't either. Let's have a bit more open discussion before a small group of people start trying to tell all other editors how to edit. -- Necrothesp (talk) 13:47, 5 September 2008 (UTC)

  • This has been debated for years, Necrothes. And it is finally coming to an end. The “linking” of dates is unwise and the autoformatting of dates was a terribly unwise decision made two years ago to resolve editwarring. Because of autoformatting of dates, 99.9% of Wikipeida’s readership (regular IP users) were often seeing unsuitable date formats in articles. Further aggravating the problem is we editors couldn’t even see the problems because our preferences settings were blinding us to this fact. The first step in fixing these articles is to unformat (delink) the dates. Then we simply write out the date in a method that is most suitable for the topic. The current MOSNUM wording…

Articles clearly written in a national variety of English should generally use the more common date format for that nation; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format.

…is just fine; there should be absolutely no editwarring over the issue. See my above, expanded response (with the fascinating picture of Tokugawa Ieyasu) for more info on just why autoformatting was—and remains to this day—a piss-poor set of tools that never should have been made in the first place. Greg L (talk) 17:11, 5 September 2008 (UTC)
Necrothesp, I appreciate that you were unaware of any discussion, but discussion on depreciating autoformatting of dates, already longstanding, drew in a large number of editors on this page through July/early August as it moved towards consensus. It was then posted to the Village pump and to WP:Centralized discussions through August.[2] Additionally, several Wikiprojects were individually made aware of it, as were a large number of individual article writers. Thus, autoformatting was depreciated. 86.44.27.255 (talk) 19:37, 5 September 2008 (UTC)
Not to mention the fact that de-linking was even the subject of a "successful" petition to the developers that the developers simply ignored, in 2006. This is by no means some sudden change. — SMcCandlish [talk] [cont] ‹(-¿-)› 21:40, 6 September 2008 (UTC)

Throwing an idea on date format

Alright, we've got a bunch of hoo-haa going on about everything. A quick summary of it is

A) People use different formats in the world
B) Editors come from around the world.
C) Editors can't agree on which to use.
D) Autoformat is a nice solution, but deficient because readers can't get the benefit.
E) Overlinking is evil and unecessary.

So, how about we do this.

1) Use ISO dates format in autoformat everywhere, unless problematic (and I'm not talking about ISO meaning, just the format, so you'd say something like 1054-08-12 (Julian Calendar) if you want to point out its a Julian date, or however it's currently done).
2) Autoformat is on for EVERYONE, the default option is the international format (which I don't really like, but it's the international format so that should be the default option).
3) Registered readers/editors can specify a preference for non-international format (like right now).
4) Autoformat dates display unlinked.
5) To write month/day, write an "empty year" ISO date such as -09-04 to produce and internationally format date.

This fixes seeing a date format such as 2008-05-15 if you are a non-registered user. This fixes overlinking. This eliminates the need for complicated and slippery slope rules such as "For Canadian topics, use canadian formats, for American topics use American format, etc.". And this fixes once and for all inconsistent date formating. It would probably require some modification of wiki software and other technical considerations but let's not worry about that. What I'm asking if we could do it somehow, would this be a (the?) viable solution?

Headbomb {ταλκWP Physics: PotW} 21:19, 5 September 2008 (UTC)

  • I’m all for many variations on this theme Headbomb. But the essential element of it requires that autoformatting be made to work for regular IP users. This requires geolocation based on IP addresses. There is nothing particularly difficult about figuring out what country a reader hails from; commercial Web sites routinely track raw statistics on visitors, such as where they live, what operating system they use, what browser they use, etc. If you go to LiveLeak, little adverts will say something like “Meet hot chicks who live in [your city]” because LiveLeak’s servers drill down in the requesting I.P. address so they know where precisely where the ISP is located. But Wikipedia’s servers and the infrastructure and tools aren’t currently set up to pull off this stunt. Further, it seems unrealistic to expect that such a substantial change at the very underpinnings of how Wikipedia’s servers operate will be produced any time soon for us. So then the question becomes “what do we do with these autoformatted dates in the mean time?” Since your proposal would require re-writes of articles (just as de-autoformatting does), we might as well go fix the problem now for the tools that are currently available.

    Further, if/when I.P.-based geolocation ever does become available, we can use the information to make other tools. Personally, I don’t mind looking at either February 10, 2008 or 10 February 2008 so I can barely relate to what the big deal is. But it is more disruptive to my reading when I encounter “realise” and “colour” and “caesium”. With geolocation of the requesting reader, we could have {{dialect|us|commonwealth|realize|realise}}, and {{dialect|us|commonwealth|trunk|boot}}. Now that would cut down on “brain interrupts” in my reading. For that matter, such a tool could be used to write {{dialect|us|commonwealth|The [[Boston Tea Party]] was a great thing to get those greedy bastard Brits off their backs.|The [[Boston Tea Party]] was put on by a bunch of ungrateful malcontents.}}. Greg L (talk) 22:09, 5 September 2008 (UTC)

Writing "1054-08-12 (Julian Calendar) " on a wikimedia-powered server is equivalent to writing "$99.95 (pounds sterling)". This is because of Meta:Help:Date formatting and linking which specifically states that for purposes of autoformatting, the YYYY-MM-DD format is ISO 8601. ISO 8601 says you can't write "1054-08-12 (Julian Calendar)" for two reasons.

    1. The use of any year outside the range 1583 through 9999 requires mutual agreement among the partners interchanging date information, and no such agreement exists between Wikipedia editors and readers.
    2. ISO 8601 dates must use the Gregorian calendar.
--Gerry Ashton (talk) 22:10, 5 September 2008 (UTC)
  • All too complicated. I'm keen that there not be mass conversions of US format right now, except where it's wrongly used in articles with strong ties to a non-North-American anglophone country, or defies the engvar. We just have to bite the bullet and make a decision about Proposals 1–3 above. Tony (talk) 04:47, 6 September 2008 (UTC)
  • Ok, I don't buy one bit the argument of "figuring out I.P.s", but perhaps others do. We don't need to figure out the IPs. You give international format by default. If for some reason, people want a non-international format on a non-international topic, there could (again throwing an idea) be a "page argument" that would say "here autoformat gives the December 31, 2006 format by default".

For example written something like

{{daf|american=yes}}

Jimbob was an accountant born on [[1943-05-06]] and died on [[1986-09-13]]. He ate a cat on [[-03-02]].
He met his wife on [[1973-02-02]].

[...]

would yield (unless you are a registered user)

Jimbob was an accountant on May 6, 1943 and died on September 13, 1986. He ate a cat on March 2 of every leap year. He met his wife on February 2, 1973.
[...]
  • Second about ISO. This is not about ISO dates meaning a non-ISO date. This is about an input format, which is outputted in a non-ISO format (at least by default). Writing for input [[1042-04-03]], will output 3 April 1042. If the date can somehow be confused with the ISO date 1042-04-03, then I don't see what the problem is with written [[1042-04-03]] (Julian Calendar) which will display as 3 April 1042 (Julian Calendar). All the instances of ISO dates requiring a "ISO interpretation" on Wikipedia should not be autoformatted anyway, so I really don't see any problem on the ISO front from the reader's perspective.

Headbomb {ταλκWP Physics: PotW} 07:07, 6 September 2008 (UTC)


This also have the advantage of being incredibly bot-friendly. Headbomb {ταλκWP Physics: PotW} 07:08, 6 September 2008 (UTC)
But we are an encyclopaedia that anyone can edit. That basically means that people will enter dates in their preferred format, especially if they are writing text rather than filling in templates. I really don't like the idea of stalking new editors and yelling at them until they do things as per the MoS. If they write dates in the "wrong" format for an article, then someone can come along later and fix it up. Or a bot can trawl through articles and do it. Making things as easy and as logical as possible for new editors also makes them feel comfortable about contributing. --Pete (talk) 08:53, 6 September 2008 (UTC)
But anyone can still edit it. People will enter dates in their preferred format as much as they do now (basically everyone except those aware of and willing to follow the MOSNUM). Someone (or a bot) will clean up after them, and everyone will be happy. You don't have to yell at anyone more than you already have to. Or maybe that is what you were trying to argue and somehow thought you were arguing something else. Headbomb {ταλκWP Physics: PotW} 11:17, 6 September 2008 (UTC)
Oh and also it would fix the conflict between reference date format and main article format, and template could be made to follow international format by default and to respond to the "page argument" when it is there. Headbomb {ταλκWP Physics: PotW} 11:17, 6 September 2008 (UTC)

Some preliminary work shows that only about 28% of dates are linked, and only about 0.04% are in the format advocated by Headbomb. (I'm skeptical of the 0.04%; it seems like all those accessdate parameters in cite templates should push that higher.) I don't see how it is feasible to process all those articles manually or semi-automatically. Doing it automatically would not allow dates to be left alone when they ought to be, such as within quotes. --Gerry Ashton (talk) 15:00, 6 September 2008 (UTC)

Gerry, I'm wobbly about the stats and suggest that we seek details of the sample and methodology first. Tony (talk) 16:31, 6 September 2008 (UTC)

I don't see what's is the problem with quotes. Can't bots be coded to leave stuff within quotes alone? Headbomb {ταλκWP Physics: PotW} 16:32, 6 September 2008 (UTC)

There are two issues. First, there are different ways of marking quotes. There are single and double quote marks, both regular and curly. They may be mixed in with each other in the case of nested quotes, along with apostrophes, inch symbols, foot symbols, minutes of arc symbols, and arcsecond symbols. The various quote templates might be used. The <blockquote> tags might be used. Other methods of indentation might be used, because the blockquote mechanisim resembles what my neighbors' horses leave on the road.
The second issue is that some dates other than those in quotes should not be altered. They might occur in snippits of programming language. They might be occur in titles of publications. I'm there sure are other places I have not thought of. And of course, strings of characters may occur that look like dates, but are not dates at all. --Gerry Ashton (talk) 16:47, 6 September 2008 (UTC)
  • Headbomb, perhaps it would be easier for you to understand how your proposal of “just default all dates to international format” would fly like a wet noodle if I were to instead propose that we default all dates to U.S. style. While I don’t have strong feelings one way or another about date formats (I’m American but typically use internationally formated dates in articles not strongly associated with the US), I am, at least, realistic in looking at the amount of bickering that has taken place over this issue and can see that there are divergent views.

    This is all just silliness to me. I truly can not fathom why so much discord has arisen over this date formatting issue. Either format reads smoothly enough for me that I don’t suffer (!) brain interruptions as I read. Neither format is ambiguous in the least. I shit you all not; I’m not at all trying to come across here like a “high road” sorta person, but I truly can not understand how the hell there could possibly be so much bickering here over dates. Would someone please explain this phenomenon to me? The existing guidelines on MOSNUM seem sensible enough to me: if it’s a US-related article, US-style dates, otherwise, international. What’s so hard about that? No special tools for newbie editors to learn. If there’s a dispute, anyone can just be advised to just go to WP:MOSNUM, read the rules, and then simply type out the dates that are most suitable for the subject matter and don’t link to mind-numbing trivia. Is that all so “Well… Duhhhhhh that it’s too simple and can’t possibly fly here? Why does this have to be so complex? Greg L (talk) 18:52, 6 September 2008 (UTC)

It's a behavioural issue. Some editors prefer American dates and American units of measurement. I'm beginning to think that there's not much point in continuing to argue about where to draw the line. Just have one format for everything and stick to it. International format because we save about a bazillion commas that way. And we're an international project. --Pete (talk) 00:46, 7 September 2008 (UTC)
Quoting Greg L:

The existing guidelines on MOSNUM seem sensible enough to me: if it’s a US-related article, US-style dates, otherwise, international.

Paraphrasing my earlier post:

International by default, US style if US centered topic.

I don't see the difference. Headbomb {ταλκWP Physics: PotW} 15:16, 7 September 2008 (UTC)

Making controversial changes without consensus

I haven't checked as to who changed the section on date formats and national ties, but I've restored it to the version that has prevailed for nine months. Discussion is ongoing, and the proposal to link date format to variety of English has not gained consensus. Could we please find a solution on the talk page instead of changing text that is flagged as disputed? I think that we are making progress, with various problems being explored in an atmosphere of reasonably genteel co-operation, but it doesn't help when people change the text to suit their personal opinions on a controversial matter. If people want to edit-war instead of discuss major changes, perhaps we should seek mediation? --Pete (talk) 05:50, 6 September 2008 (UTC)

It is Skyring who is edit-warring for an extreme minority opinion. He need only read the section shortly above to see GregL's express approval of the text I have restored; the text to which he reverted is certainly not consensus. If this continues, dispute resolution may indeed be in order. Septentrionalis PMAnderson 16:22, 6 September 2008 (UTC)
Pete, your actions are not helping; please keep it on the talk page for the moment. I know that you feel strongly about the issue, but Anderson is right in this case. Tony (talk) 16:29, 6 September 2008 (UTC)
Concur with PMAnderson & Tony1 (no disrepsect to Pete; it's just a matter of keeping the pot from boiling over). — SMcCandlish [talk] [cont] ‹(-¿-)› 21:24, 6 September 2008 (UTC)
SMcCandlish's comments above about the "current" MOSNUM wording alerted me to the fact that it had been changed. It's not my preferred wording either, but best to put it back the way it's been for nine months until we find a consensus. --Pete (talk) 22:59, 6 September 2008 (UTC)
header 1 header 2 header 3
row 1, cell 1 row 1, cell 2 row 1, cell 3
row 2, cell 1 row 2, cell 2 row 2, cell 3

Invitation to join the discussion at the tennis project page

Regardless of your views, you are invited to join this discussion at the tennis project page concerning the linking of years and whether a project-specific consensus can override the Manual of Style of this subject. This discussion also concerns the operation of bots. Thanks for your time! Tennis expert (talk) 06:05, 7 September 2008 (UTC)

Who made DD/MM/YYYY the "international" date format?

As far as I can tell, the International Organization for Standardization made "YYYY-MM-DD" the international date format.

  • "DD/MM/YYYY" is the British date format,
  • "MM/DD/YYYY" is the American date format,

and ne'er the twain shall meet. The former might also be considered the European date format, but not the Asian date format since the Chinese and Japanese don't use it. As for Canadian date formats, the Canadian government's English language editorial style guide says that for alphanumeric dates,

  • either MMMM DD, YYYY (e.g. February 14, 2008)
  • or DD MMMM YYYY (e.g. 14 February 2008)

is acceptable since neither is ambiguous and anybody who speaks English can understand them. However, for all-numeric dates,

  • neither DD/MM/YYYY
  • nor MM/DD/YYYY

are acceptable since they are mutually ambiguous, and the

  • ISO standard YYYY-MM-DD (e.g. 2008-02-14).

is the only acceptable format since it is the only one that is unambiguous.RockyMtnGuy (talk) 03:10, 5 September 2008 (UTC)

  • Sound logic. However, there are two issues as far as our articles are concerned: firstly, consistency of date formats should be the focus of our efforts, not really 'which one'. Secondly, ISO couldn't look more awkward in an article. Ohconfucius (talk) 09:46, 5 September 2008 (UTC)
  • You're quite correct, ISO 8601 is the only language-independent unambiguous standard date format in existence. To summarize - its general advantages are:
  1. Language-independent
  2. Unambiguous
  3. Supports sorting
  • But when one takes a look at its specific advantages for use in English Wikipedia, #1 becomes irrelevant, #2 also (no extra edge over, say "September 5, 2008"), and #3 is easily solved by templates. GregorB (talk) 09:59, 5 September 2008 (UTC)
I think people writing "DD/MM/YYYY" above, like in the map caption, don't mean anything by the slashes, and rather really mean the alphanumeric "DD MMMM YYYY". I, as a European (Swede, to be exact), would always assume that "04/05/06" is in April 2006 (slashes are used by Americans), that "04-05-06" and "040506" are both in May 2004 (shortened ISO formats), and that "04.05.06" is in May 2006 (typical European format). That's not to say that any of these formats should be used in Wikipedia (other than in quotations and the like), but just pointing out that there's probably nobody advocating the use of "04/05/2006" to mean 4 May 2006, or thinking that this format is in common use. I definitely agree that if there is a reason to use an all-numeric date format in an article, full ISO standard is the only way to go, but there should hardly ever be such a reason. -- Jao (talk) 10:08, 5 September 2008 (UTC)
-->there's probably nobody advocating the use of "04/05/2006" to mean 4 May 2006, or thinking that this format is in common use.<--
  • Unfortunately, the British do use slashes in their dates, i.e. they use "04/05/2006" to mean 4 May 2006, whereas the Americans use it to mean April 5, 2006. This is a fundamental source of ambiguity.
  • The shortened ISO date "04/05/06" is completely ambiguous, because the British use it to mean 4 May 2006, the Americans to mean April 5, 2006, and the ISO date means either 6 May 2004. Also, it might mean 4 May 1906, April 5, 1905, or 6 May, 1904. Therefore you should always use 4-digit year numbers.
  • All-numeric dates are useful in tables and templates, and if you use the ISO format you can sort them without special software.
The British and the Americans are both very stubborn peoples, so you can't expect either one of them to agree to use the other's date format. The use of slashes doesn't make any difference to them. However, if you use dashes it almost certainly will mean ISO format.RockyMtnGuy (talk) 16:14, 5 September 2008 (UTC)
I have discovered that the identification of the YYYY-MM-DD format with ISO 8601 is part of the Wikimedia software.[3] Granted, this identity has been poorly communicated in Wikipedia; indeed, Wikipedia lacks a mechanism for effectively communicating any important writing convention to readers. But turning attention to this thread, there are only two ISO 8601 formats for communicating a day, month, and year:
  1. The basic format, "Example: 19850412"
  2. The extended format, "Example: 1985-04-12" (from ISO 8601:2004 p. 12)
To say or imply that 85-04-12 has any connection to an ISO standard is an error. --Gerry Ashton (talk) 16:35, 5 September 2008 (UTC)
The 85-04-12 format was allowed by earlier versions of ISO 8601 (back in the 1980s and 1990s) but it was dropped from the 2000 version onwards. Using only two digits for year information was the source of the Y2K Problem and ISO moved to only allow a full four-digit year to be used from then on. -- 2008-09-12 08:20 (UTC) —Preceding unsigned comment added by 79.73.229.72 (talk)
  • I will put this here, since I have just entered the discussion. There are FOUR formats accepted in Canada & from what I see in the proposed guideline, ONLY two will be accepted even for Canadian articles. A few articles have always used YYYY-MMMM-DD format. As the guidelines stand, they require choosing one of only two. I think the guideline should be rewritten to acknowledge that other specific formats are also acceptable. --JimWae (talk) 00:10, 7 September 2008 (UTC)
JimWae, I imagine they are (where lower case letters indicate leading zeros are not used and MMMM is a spelled-out month):
    • dD MMMM yyyY
    • MMMM dD, yyyY
    • YYYY-MMMM-DD
So what is the fourth format? --Gerry Ashton (talk) 00:42, 7 September 2008 (UTC)

Actually, FIVE, counting these as three:

  • YYYY-MM-DD
  • YYYY-MMM-DD
  • YYYY-MMMM-DD --JimWae (talk) 00:46, 7 September 2008 (UTC)

So, do any proposed policies NOT completely eliminate YYYY-MMMM-DD??? --JimWae (talk) 08:10, 7 September 2008 (UTC)

Question about date autoformatting

Now I've started reading over this issue, and am getting a bit confused, I see a section in the WP:MOSNUM page about autoformatting and a link to WP:Autoformatting that seems to contradict parts of itself. Could each party fill in what they think should be WP's guideline on date presentation, as it relates to autoformatting in the below sections. One user to a section, a new section for each proposal, please. MBisanz talk 21:41, 30 August 2008 (UTC)

No, too many editors are running around creating diverse places to present their views. I'm ignoring this place. --Gerry Ashton (talk) 21:51, 30 August 2008 (UTC)

Indeed. The issue was decided and has wide support and little opposition. This is just churning. Tony (talk) 02:56, 31 August 2008 (UTC)
Can someone please point to the decision that was made with wide support and little opposition? dm (talk) 06:13, 2 September 2008 (UTC)
Yes, and I do wish that chaotic archiving system made it easier to find things. It's here. Evidence of wider approval is here. Tony (talk) 08:15, 2 September 2008 (UTC)

There is some opposition located here [4]. Daytrivia (talk) 21:38, 3 September 2008 (UTC)

A fresh start on date autoformatting (DA)

Having looked through the edit wars over the Date autoformatting subpage and noting the points made in disagreement, as well as having followed the issue here for quite a while and participated in it occasionally, I’d like to offer a “rough draft” revision of the DA subpage that to my mind’s eye seems to capture as much of everyone’s concerns as I best I can. I’ve also done so with an eye for what appears to be the relative degree of consensus for the various issues; as a rule, I’ve selected “softer” words instead of “harder” ones where a consensus is least evident. That is why I’ve used “discouraged” in preference to “deprecate” (which seems to often get misinterpreted as “forbidden” and taken as a call for a crusade for their removal – for which there isn’t such a strong consensus), and introduced “deprecate” in a more limited sense later on. I’ve also purposefully chosen to go with a lengthier version than the concise, bulleted one currently displayed. I’ve done so because this “deprecation” of usage will be better tolerated and less abused if explained, rather than just summarized.

I’m offering this as a fresh starting ground. I believe – and several people have noted – that there was no express language for the consensus to deprecate DA. Instead, the needed language was built through direct edits of the MOSNUM page, and it led to edit warring. The proper place for sparring, as we all know, is here on the Talk page.

The linking of dates purely for the purpose of autoformatting is strongly discouraged, because the use of these tools results in more disadvantages than advantages. This feature can only be seen by a tiny minority of Wikipedia’s readership: those registered editors who have chosen to configure their date preferences (My preferences → Date and time → Date format). This obscures the existence and extent of conflicting date styles in an article that is otherwise readily apparent to the majority of readers, thereby hindering correction of the problem.

Furthermore, the resulting links normally go to lists of historical trivia which usually have little or nothing to do with the subject of the article. The use of these formatting tools therefore tends to produce overlinked articles. All links from articles should be of high value to the reader; that is, following them should genuinely help the reader understand the topic more fully. Overlinking can negatively affect overall readability and appearance in articles which already have numerous high-value links.

In addition, dates from before the introduction of the Gregorian calendar should not be expressed in the ISO 8601 format, which was designed for use with the Gregorian calendar. Conventionally formatted dates from the pre-Georgian era will normally be in the Julian calendar format, and being wikilinked and autoformatted into would constitute a false assertion they are Gregorian dates.

To avoid disruption, however, this deprecation should not be taken as license for wholesale removal of existing links from articles currently employing them extensively. Such removal from a given article should follow a consensus to do so among the editors of that page.

The functions provided by these tools nonetheless remain available and are described at Date autoformatting.

Please critique it in the spirit in which it’s offered and with an eye toward compromise. Askari Mark (Talk) 22:40, 30 August 2008 (UTC)

I would say wikilink ISO 8601 rather than just ISO. I would also say "...extensively for Gregorian dates..." (addition bolded)because I think it is a good idea to conduct a campaign for the sole purpose of removing autoformatting from articles that use autoformatting on Julian dates. Ideally such a campaign would also insure the calendar used in the article is explicitly stated. --Gerry Ashton (talk) 23:01, 30 August 2008 (UTC)
Thanks, Gerry, good catch! That's the sort of input I'm looking for. I've incorporated your recommendations. Askari Mark (Talk) 23:37, 30 August 2008 (UTC)
The automatic removal from Julian dates is faintly silly. Stating the style explicitly on articles about the Middle Ages is utterly unnecessary; it would be bizarre to use Gregorian before it existed. Autoformatting was intended to switch back and forth between 30 August and August 30; I suspect most people use it that way; and that's perfectly safe. Anyone who chooses to autoformat to ISO does so at his own risk (and, like Woodstone, she may not care about Renaissance history). Please make that by consensus too. Septentrionalis PMAnderson 00:34, 31 August 2008 (UTC)
"Stating the style explicitly on articles about the Middle Ages is utterly unnecessary" is bullshit. Not everyone who would like to read an article about the middle ages knows about the convention for which calendar to use, nor do they necessarily know when the Gregorian calendar went into force in any given place. Also, I cannot find the design criteria for the autoformatting "feature" so I do not know what the intentions of its designers were. I would be interested in seeing a record of what they were thinking. --Gerry Ashton (talk) 00:58, 31 August 2008 (UTC)
That had been my understanding, too, PMA – until I began reading the recent debate over it and Gerry and others began indicating that it may not be so. Do you know someone sufficiently familiar with the code to tell us one way or the other? I’m copacetic with it either way; one of the reasons I wrote it in as a separate parenthetical note is so it can easily be removed if need be. Askari Mark (Talk) 01:08, 31 August 2008 (UTC)
See #How did we get here? on this page. Looking at the past versions of MOSNUM confirms that autoformatting was intended to settle dating disputes between the two conventional alphanumeric formats. In accordance with this, I am removing the parenthetical. Septentrionalis PMAnderson 20:55, 31 August 2008 (UTC)
Actually, although it shows as a blue link, it would not work for me and there is no such heading on this page. I also looked in the D6 archive, but found nothing there. Askari Mark (Talk) 22:41, 31 August 2008 (UTC)
My apologies: #Where did this come from?. Septentrionalis PMAnderson 21:17, 2 September 2008 (UTC)

This is an attempt to scuttle a genuine reform of a major formatting function that almost no one likes, by slowing its pace so that it will take decades rather than a few years, and will forever be patchy in it application.

"To avoid disruption, however, this deprecation should not be taken as license for wholesale removal of existing links from articles currently employing them extensively. Such removal from a given article should follow a consensus to do so among the editors of that page. (Exception: Date autolinks should be removed whenever they are found being used with non-Gregorian dates.)"

  • Tony, I agree; you engage in outrageous and unfounded attacks when you do have a shadow of an argument. But your claim is a falsehood: I think that if we don't push things right away, we will be rid of autoformatting in a year. Going full speed ahead now will be damaging to the encyclopedia and is the only thing at all likely to delay this reform for decades. Septentrionalis PMAnderson 20:45, 31 August 2008 (UTC)

The argument seems to rely on (1) mixing up date warring and ISO with the decision here and elsewhere to get rid of DA, and (2) some kind of unstated personal discomfort at bold change. However, the current debate about the choice of format for third-party countries is a response to an issue that has always been there for our readers. There is no relationship with the current "edit wars" over choice of date format, except that the sudden realisation that DA has shielded WPians from seeing the mess of inconsistency and wrong global choices of format; this has concentrated people's minds on it, which is a good thing.

So, work out what you want with ISO dates, and decide on a guideline (or not) for the third-party format for dates, but don't try to confuse these issues with the removal of DA.

And why refer to "deprecation" in the quote above while pussy-footing about it in the opening sentence and requiring a long, convoluted generation of consensus at every single article to achieve the goal? You can't have it both ways. (And the bit about Gregorian dates begs the larger question while introducing muddy logic.)

As for "To avoid disruption", any bold improvement to WP could be framed as "disruption", and it is idle and manipulative to frame it thus without evidence of widespread opposition. There is none: show me, please, the raft of reversions of the date auditing that has been conducted over the past few weeks. And please face up to the strong—not mild—agreement of so many people, here and on the other pages I linked to during the debate. It keeps trickling in, like this:

Contributors should revile this attempt at sabotage. The proposed change does nothing to improve what is now a straightforward, concise statement. We need to grow up. Tony (talk) 02:51, 31 August 2008 (UTC)

Wikilinking dates was a kludge brought in to settle the date wars of some years ago. In all the time since then, no better method has been found. However, it is a kludge, and increasingly obvious. The two major disadvantages are:
  1. It fills an article full of useless links. Click on a date link and most of the time you are taken to a page that has zero relationship with the original article.
  2. It doesn't work for the 99% of Wikipedia users who are not registered with user accounts and date preferences.
We are writing an encyclopaedia, and our readers must be our main focus. We are presenting information to the wide world. so I'm totally in favour of removing date autoformatting. We get to see what the readers see, and we make the reader's experience better by removing a sea of useless links.
However, we now have to deal with renewed outbreaks of date wars, as editors, no longer shielded from date formats, see raw dates in a format they personally don't like. I think that for the removal of autoformatting to work, we've got to work out a solid solution on date formats. --Pete (talk) 05:37, 31 August 2008 (UTC)
Agree strongly with Tony and Pete. It isn't "disruption" to alter Wikipedia articles to bring them in line with what we've agreed is the right style; the disruption is tryin to stall this process in spite of the lack of any substanial arguments against it. Also I don't see that the current solution we have on date formats is in any way unsolid.--Kotniski (talk) 16:41, 31 August 2008 (UTC)
We shall see. I don't think it's unsolid; I think that authorizing mass implementation is a recipe for making, and keeping, this controversial, and that doing so will slow down acceptance immensely; festina lente, "make haste slowly", is very old and very good advice. It is consensus here to get rid of autoformatting is a good idea - if we proceed, page by page, for a few months, we will probably persuade the rest of Wikipedia to join this consensus. By that time other editors will have heard of the change, and (with luck and tact) most of them will agree with us or not care. At that point, stronger language and actions may be in order - or, if we convince enough people, unnecessary. If every editor thought it was the normal thing to delink dates, we'd be done without bots. Septentrionalis PMAnderson 20:40, 31 August 2008 (UTC)

Tony, I ask you to assume good faith and to make no assumptions about my motives or intents – simply ask me and I will answer your questions. I have no agenda other than to forward this process with minimal disruption. Period. To accuse me of trying to “sabotage” the effort is uncalled for, utterly lacking in evidence, and frankly a personal attack since it maligns my motives. This is exactly the sort of behavior I mentioned earlier as discouraging participation by other editors. As PMA notes, you’re perfectly capable of offering logical arguments sans ad hominems.

What I have attempted to do is to capture the actual degree of consensus for the various elements, starting with the text as I found it, and taking into consideration the preceding discussions and edit comments on changes (leading up to and during the edit warring that got the guideline page locked down). What I originally offered was a good-faith starting point for working toward a final text that enjoys general consensus – while avoiding edit-warring. FWIW, it’s not my favored version and I little care what the final wording is, as long as it enjoys consensus.

As regards the ISO statement, it’s included because it was in the locked-down text. I wasn’t aware that there is still an argument over whether ISO should or should not be used with non-Gregorian dates. As I noted to PMA, my impression has always been that all the DA does is take care of handling the day and month (and American-style comma use) per a user’s preferences. If that’s all it does, then I frankly don’t see how it can screw up Julian dates. In that case, DA has nothing to do with whether dates are Gregorian or Julian, and the whole paragraph (and parenthetical note on DA removal from Julian dates) should be excised – and I will be happy to do it; just provide me the facts.

As for a crusade to slaughter all remaining DAs, there has never been consensus for it. Moreover, a consensus for it has never yet been sought in the foregoing debates; the only time the issue has been raised has been by editors appalled by actions of the raft of crusaders. Frankly, I see such a mass removal as disruptive, have previously noted so, and less disruptive approaches have been suggested. In fact, nobody disagreed with me and you yourself thought it a good idea! In any case, I have no love for DA and wouldn’t mind if they all disappeared overnight – however, such a move has no consensus. If you wish to seek one for aggressive removal and achieve it, hey – this guideline can be changed to say so! Right now, though, there’s consensus only for “soft” deprecation, not “hard”.

Pete, I’m not arguing for DA. That they’re not needed and mostly unwanted by or unavailable to readers is a pretty solid consensus. I agree with you that “for the removal of autoformatting to work, we've got to work out a solid solution on date formats.” I would further suggest we should not set off on any mass DA removal effort until after consensus on that is achieved. However, I think you will agree that whatever formula is developed, it can be worked into the summary text I’ve proffered.

PMA, thanks for being the voice of reason! Askari Mark (Talk) 23:40, 31 August 2008 (UTC)

Whist I agree with Andserson's urge for a soft start to this depreciation, I oppose having it enshrined in the guideline. I'd say that our best bet would be, for now, to seek consensus on talk pages first. By doing this we should expect to raise awareness of the evils of date linking instead of inspiring knee-jerk reactions against the depreciation. We've got a very good consensus here for depreciation of date autoformatting via linking but we've got that consensus here. Our challenge is to spread it by reasoned arguement, we've got loads of strong arguement, but arguement does not necessarily win when put up against the knee-jerk reaction, which doesn't lightly bear retraction. Nonetheless, we don't need the MOS telling us to go easy. I think all of us are well wikied enough not to go about disrupting the place. The approach should be soft, the guidance must be strong. What we certainly don't need is a suggestion that date delinking requires a stronger than usual consensus. Bold edits are okay:

Consensus is typically reached as a natural and inherent product of the wiki-editing process; generally someone makes a change or addition to a page, and then everyone who reads the page has an opportunity to either leave the page as it is or change it.

It is not required that consensus be sought first as Askari Mark's wording would have us think. No, let's take it slow to start but not have the guidelines worded such as to keep it slow. JIMp talk·cont 18:43, 1 September 2008 (UTC)
If you're the only editor who cares, you are consensus. At worst, this would commend asking on talk pages, and then doing the work after about 24 hours if there is no objection; it might be compatible with straight BRD. But the opportunity to tell other editors why DA is a bad idea is the most important part of the process; if msot of the editors were convinced, the problem would solve itself. Septentrionalis PMAnderson 16:56, 2 September 2008 (UTC)

Why is this just about "autoformatting"? Wouldn't it be much simpler to say, "Dates in wikipedia should not be blue-linked unless there is an important reason to do so, for example a discussion of the September 11 attacks." -- Ssilvers (talk) 01:48, 3 September 2008 (UTC)

Jimp, the reason I’ve been encouraging a soft start is two-fold: most obviously there’s the desire to avoid major disruption over the removal of DAs, but the second reason is that there has never been an attempt to query consensus for mass delinking. Some people here want that to happen (and to be able to read that into the new wording), but most of the discussion concerning it has been raised on stopping people who were going ahead and doing it. Remember, my aim was to capture what I understand the developed consensus to be. If the addition is unwanted, it can be removed, but at the least it made for summary of what was and was not explicitly in the consensus (and could have been useful for prompting discussion). As for “enshrining” it, well, DA has been “gospel” for a couple years; leaving this caution in for a few months (and removing it when consensus says it’s no longer necessary) is quite “kosher” IMHO.
In any case, it seems many folks here really would rather indulge in our local MOS version of the Bavarian face-slapping contest than actually work together collegially, so I see no need to continue with this particular effort. Askari Mark (Talk) 04:06, 3 September 2008 (UTC)
Where is this major disruption you talk of? Tony (talk) 04:44, 3 September 2008 (UTC)
Here is some hard evidence against the fear of disruption: ever since the guidance was changed to no longer recommend autoformatting, I've been routinely unlinking dates whenever I'm working on any article. I've done this with hundreds of articles, and I've had three or four questions, where I've explained my reasoning and it's been accepted, and two violent objectors who weren't prepared to discuss it reasonably at all, just reverted everything I did; and as for all the other articles, no response, pro or anti. These figures don't suggest to me that there's a significant danger of disruption from objections to large-scale unlinking. Colonies Chris (talk) 12:53, 3 September 2008 (UTC)
Where is the consensus for rapid mass delinking, Tony? Colonies Chris, when the deletion of Fair Use articles began, there was only some grumbling, too, as editors only became aware of it as it affected their watchlisted articles; as it continued to logroll, the complaints and anger grew apace, and quite a few productive editors left over it. Today, whenever Betacommand/Betacommandbot ends up on AN/I, there’s a huge pile-on on him that results mostly from the antipathy he garnered implementing the removals. Which is truly a better approach – quick mass deletion (for which no consensus has been sought) and the risk of mass disruption, or going a little slower (maybe even taking the time to gain the consensus) and minimizing – or even avoiding – same? Askari Mark (Talk) 15:03, 5 September 2008 (UTC)

I was one of the heavy date editors until a month or so ago. I hit the 2008 Atlantic Hurricane Season which I look at for reasons of interest and noticed that NONE of the dates were linked. I spent the last month looking for this article. I think that it would be useful to provide a way for dates to be listed in a reader's preferrable method, but I agree that linking everything that happened on 9 May to a list of bizarre events that occured that day is not useful. However, it would be useful to use a similar but different symbol to force all date to the format and possibly even calendar requested by the user.Autkm (talk) 05:40, 6 September 2008 (UTC)

Another person here disagreeing with the delinking after seeing mass delinking on watchlisted articles.
"pre-Georgian era" (1714/1830) should probably be changed to "pre-Gregorian era" (1582/1923).
Will anyone mistake post-Gregorian dates for Julian dates because they're not formatted in ISO 8601?
The massive delinking is causing middle-endian dates to show for people who've set up their date formatting otherwise, making it harder for some people to understand the encyclopedia's articles: "Until the software can how about not delinking, format unlinked dates?" (translated: "How about not delinking until the software can format unlinked dates?"). -- Jeandré, 2008-09-06t10:45z
Before addressing your complaint, could I ask you to clarify whether the second and third points are connected with the removal of date autoformatting, or are a separate matter? What are "middle-endian" dates? Tony (talk) 12:00, 6 September 2008 (UTC)
Endianness#Other meanings: little endian (least important info first) 31 December 2008; big-endian (most important info first) 2008-12-31 (and alphabetical also = chronological); middle-endian or mixed-endian (mixed up) December 31, 2008. I'm not sure what you asked before that. -- Jeandré, 2008-09-06t12:28z
Re User:Tony1/Information on the removal of DA#Community attitudes:
1: Kinda agree, I'm a programmer.
2: Agree, have dates autoformatted without links before massive unlinking.
3: Disagree, I'm fine with changing things if the changes are making WP a better encyclopedia. I've been editing WP since 2002 and things have changed drastically, nd if for the better I'm fine with it.
4: Not really, tho linked dates help scanning (other links to important articles help more) the kind of information in encyclopedias; as does sections, italics, bold text, lists, tables, images, captions, etc..
5: Kinda, I prefer RFC 3339 (ISO 8601) tho there's practically zero support for using that outside of refs. I don't like little-endian dates tho they're "bad as not as, middle-endian dates" (translation "not as bad as middle-endian dates"). My preferences are now being ignored with the massive delinking. Like the vast majority of people who read English, it's not my first language - I have to now read articles with middle-endian dates making reading and comprehension slower and harder. Compare how some language say numbers: 123 is not said 1 hundred 2wenty 3, but 1 foo 3 bar 2baz - making things very hard, expecially if your preference to have it said 1 hundred 2wenty 3 instead is removed. -- Jeandré, 2008-09-06t12:55z
Jeandré—thanks for explaining the endy thing, which makes perfect sense in retrospect. I'm sorry to be appear daft, but I can't locate the points you're responding to in your numbers 1–5. Thus, all I can provide by way of a rejoinder at this stage is that your personal preferences are noted, but our readers have to view the raw formatting. Forgive me, but they are all I care about! Tony (talk) 13:08, 6 September 2008 (UTC) Ah, now I get it: the five bullets Jeandré is responding to are my take on why various people might disagree with the deprecation of DA. Sorry, I am daft. Tony (talk) 13:11, 6 September 2008 (UTC)

Date format choice

Since autoformatting dates by enclosing them in square-brackets is now discouraged/deprecated, agreement on the format of dates to be used becomes even more important. The discussion has gone on for a long while, and more participants are needed to arrive at a reliable consensus. Teemu Leisti (talk) 01:06, 4 September 2008 (UTC)

Note: there was a template here earlier, commented out now to prevent this section from being pointed to from the RFCstyle page once this page is archived, which I will do after saving this change. Teemu Leisti (talk) 16:28, 13 September 2008 (UTC)

Three versions have been proposed:

  • Articles on topics with strong ties to a particular English-speaking country should generally use the more common date format for that nation; articles related to Canada may use either format consistently.
    • Present text, widely approved here.
  • Articles on topics with strong ties to a particular English-speaking country should generally use the more common date format for that nation; articles related to Canada may use either format consistently.
Articles related to other countries that commonly use one of the two acceptable formats above should use that format.
  • This is what we used to have, and what Skyring reverts to.
  • Articles clearly written in a national variety of English should generally use the more common date format for that nation; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format.
    • Woodstone's idea discussed here. This would make articles in American use American dating; articles in a Commonwealth English (except Canadian) use international. So far nobody objects to this.

Comments? Septentrionalis PMAnderson 00:45, 31 August 2008 (UTC)

Perhaps if you were a little more truthful, we might find a solution agreeable to all. The "straw poll" you created was roundly condemned for being confusing. You left out my proposed wording, which received more support than yours. There are further examples in previous sections. I can understand why some people have strong feelings about date formats, but that's all the more reason to keep things cool. Quite obviously, what we don't have is consensus on this point, which is why I keep restoring the original wording, which at least had the benefit of remaining untouched for nine months. Perhaps we could look at things on which we do agree and build up consensus that way? --Pete (talk) 01:35, 31 August 2008 (UTC)
Just catching up on all the date change stuff so I may be a little behind here. But if this is going to change it might be a good rule of thumb to follow conventions similar to what are used at WP:ENGVAR for variant spellings. The biggest problem I see with eliminating the auto-formatting as it stands right now is that there are a bunch of articles where despite recommendations that date formats are a mismatch of ISO and other formats. This is seems especially common in reference sections where the date format may be prescribed vs the body text. If auto-formatting through links is not the best approach perhaps it might be possible to request an extension to wiki-media to provide an automatic formatting syntax that does not require links. PaleAqua (talk) 05:20, 31 August 2008 (UTC)
The conventions at WP:ENGVAR have long been cited for date formats, and should be the basis for settling disputes. The guts of it is "strong national ties to a topic", so we pick the variety of English to suit the topic. The Washington article is written in American English, the London article in British English, the Canberra article in Aussie English, complete with differences in spelling, punctuation and syntax. By and large, date formats follow the variety of English, and there is little disagreement over whether an article relating to the English-speaking world should use day-month year or month-day-year.
Problems arise when writing about topics linked to countries where English is not spoken. Obviously there is not a variety of English we can turn to for articles on Albania, Vietnam, Angola or Cuba. The convention is that the variety chosen by the first major contributor stands, and again, that has worked very well.
However, for non-English-speaking nations, date formats are independent of the language. Spain uses day-month-year, the Philippines month-day-year, and China the year-month-day ISO format. Because this is the English-language Wikipedia, we use the English-language month names for dates in articles for these countries, so the local language is not a factor.
In fact, the language is not a factor even for English-speaking nations. We know what date format each nation uses, and we use them. The variety of language does not determine the date format, nor vice versa.
What is useful in the WP:ENGVAR guidelines is the phrase about "strong national ties". If a nation uses one date format in preference to another, then we should use that date format in articles that have a strong national tie. The exception is for ISO dates, which are awkward when used in written text, though useful in tables and templates.
The parallel I draw is with units of measurement. We give priority in writing to miles over kilometres in nations that use miles in preference to kilometres, and vice versa. The language is immaterial. What counts is what is used in the country we are writing about.
One point repeatedly made is that there is no confusion between 7 December 1941 and December 7, 1941. Even if a reader prefers one format over the other, there is no doubt at all as to which day in history we are talking about.
I think that removing the language qualifier from the existing wording is all we need to do:
  • Articles on topics with strong ties to a particular country should generally use the more common date format for that nation; articles related to Canada may use either format consistently.
I'd like to think that this is just common sense, but ironically I find more useful wisdom in WP:NOCOMMON, where it says, "Wikipedians come from diverse ethnic, religious, political, cultural and ideological backgrounds and have vastly different beliefs on everything from science to shoe shopping." That's the crux of the matter. We need to be tolerant, understanding and embracing of diversity, and if we write an article about an Italian or an American or a Angolan, it is common courtesy to use the forms of their country, to the best of our ability. --Pete (talk) 06:12, 31 August 2008 (UTC)
I find much to recommend Anderson's principle of aligning engvar with date format (with an explicit note that Canada-related articles may use either date format). It's simple, clear, and can leave a small crack for non-compliance where there are good reasons not to comply.
Venezuela apparently uses US English but international date formats. But you know what? I couldn't give a dump if US date formats are used in that article, because the article is written in English, not Spanish, and is for people who can read English; if a US editor happened to start the article, good on him or her—that should be the end of the matter. Sorry, Pete, I also don't care what system Italy uses, nor what system it did use in 1850, if that's the topic context. I'm keen to decide this issue and move on, so I call now for objections to this course of action. Unless someone comes up with a significant problem, I intend to insert this into MOSNUM in about two days' time. Tony (talk) 06:53, 31 August 2008 (UTC)
Venezuela uses US English???? I thought they used Spanish. For non-English-speaking nations, it is generally going to be a matter of personal opinion as to which version they use. I need only point to Japanese English, commonly known as Engrish, which defies categorisation. You might not care which date format a country uses, but for many people, not least the Wikipedians living in that country, it is important. We are trying, I hope, to find a solution for those who care. Those Wikipedians who just go with the flow aren't the ones who generate conflict. I think it is important that we short-circuit future unpleasantness. --Pete (talk) 07:12, 31 August 2008 (UTC)
Pete, Sandy was pretty insistent that this is the case, where English is used in Venezuala. But Sandy's in transit at the moment, so can't comment. Tony (talk) 09:23, 31 August 2008 (UTC)
But they speak Spanish in Venezuela. Any English is, like Engrish in Japan, going to be highly variable, depending on the background and the preference of the speaker. Hard, if not impossible, to find an objective answer for each country, especially if it's dependent on asking your friends for their opinions. And, more importantly, quite irrelevant as to the date format actually used in the country. If we are to make a choice on date format to use on a country by country basis, why not use the format the country uses, instead of basing our choice on the individual spelling preference of speakers of a minority non-official language? --Pete (talk) 10:06, 31 August 2008 (UTC)
And the dating format in Venezuelan Spanish should concern the Spanish Wikipedia, not us. Septentrionalis PMAnderson 20:32, 31 August 2008 (UTC)
If we are writing articles about Venezuela, we should use the format common in that country, so long as it can be expressed in English. We use the units of measurement common in Venezuela, we use the Venezuelan spelling of personal and place names, including diacritical marks not generally used in American English such as Hugo Chávez Frías, so why should we impose American Dating format on them if they don't use it?. Your proposal is out of line with established wikipractice. --Pete (talk) 21:37, 31 August 2008 (UTC)
But this would effectively mean WP:ENGVAR rules encroaching on articles that are normally not subject of WP:ENGVAR (e.g. articles about Venezuela) - hardly an established wikipractice itself. Personally, I find the English date format of "dd month yyyy" a bit odd; "month dd, yyyy" feels more natural to me, although this is not a common format in the country where I live (Croatia). Other editors might feel the same. Why should I (or anyone else) be compelled to write "dd month yyyy" in English articles about Croatia? GregorB (talk) 12:41, 1 September 2008 (UTC)
What format does Croatia use? Using that format seems the best and fairest way. --Pete (talk) 10:29, 3 September 2008 (UTC)
The format is dd. month yyyy, or dd.mm.yyyy (numeric). This would correspond to the "international" date format in English. But if you took a survey of the articles in the Category:Croatia (of which, one might presume, most were written by Croatian editors), you would likely find that the "international" format is not too popular: Croatian editors could have used it if they felt it was the best, most natural format - yet they didn't. GregorB (talk) 13:41, 3 September 2008 (UTC)

Where did this come from?

The former versions of this page are quite clear. There was a controversy between two factions who wanted to use February 11, 1958 and 11 February 1958 respectively. Nobody advocated 1958 February 11; 1958-02-11 is even not mentioned as being one of the alternatives, although it is discussed later.

"To put an end to this debate, in July 2003, a new MediaWiki feature known as "dynamic dates" was implemented. This allows users to select in their user preferences which date format they prefer to see. Dates written in any of the above three formats are automatically converted to the user's preferred format as the page is displayed. The default is to leave it as written, although it may be slightly cleaned up and standardised."

That should be definitive. Anybody who wants to go back and see exactly when this language was introduced should feel free to do so. Septentrionalis PMAnderson 20:21, 31 August 2008 (UTC)

That does clarify what the editors of the Manual of Style were thinking. What the developers who implemented the "solution" were thinking is another matter. It would be interesting to know if autoformatting of ISO 8601 dates was done right from the beginning, or something tacked on later. --Gerry Ashton (talk) 21:02, 31 August 2008 (UTC)
As phrased, this suggests that ISO was an add-on, not yet implemented; the "three formats" mentioned above are September 2, 2008, 2 September 2008, and 2008 September 2. It might be possible to tell by seeing when the chart now in WP:Autoformatting was substituted for the text in the link. Septentrionalis PMAnderson 16:27, 2 September 2008 (UTC)

Finally, can we arrive at a solution, please?

Here's a sequential four-test process to decide (in most cases, only the first one or two will be needed). They would replace the last two subsections in "Full date formatting", that is, "Strong national ties to a topic" and "Retaining the existing format".

Which format to use [See newer version in green below (Tony1)]

Each of the following four tests should be applied in sequence until the format for an article is determined. However, this may be disregarded where are good reasons to do so (e.g., the US military usage of international format) and there is local consensus for this.

  1. Does the article have strong ties to an anglophone country? If not, or Canada-related, or if related to more than one anglophone country ...
  2. Which variety of English is the article written in? If unclear, or Canada-related ...
  3. What is the existing formatting in the article? If significantly mixed ...
  4. Use the original format by consulting the edit history.

NB, ties to "more than one anglophone country" are exemplified by a British actor who spent most of their career in the US. I've just audited the dates in such an article, actually.

"Support", "Object" with reasons, or "Comment", please:

  • Support: Might be tedious sifting through to determine engvar. However, if that is unclear, you just go with the existing date format, or if it's significantly mixed, go with the format first inserted into the article. It's easy. Tony (talk) 06:53, 31 August 2008 (UTC)
  • Object: Here's my test: Which date format does the country actually use? That's a lot simpler and more accurate than linking it to the variant spelling of a minority non-official language. --Pete (talk) 07:30, 31 August 2008 (UTC) [Pete, I moved this down from the top and tagged it "Comment", since its chronology was confusing. I hope that's OK. (Tony1)]

*Support: Wikipedians living in a foreign country may come to English WP to read an article, yet it seems to me that if they see the date format in their native format, whether day-month-year or month-day-year, it is secondary (if even that) to the substance of the article itself. They read WP to learn about a subject and can understand either formats dating convention. The validity of the information is what is most important. The style of the article aides the reader in following the information in a logical presentation. I believe as long as there is continuity in style throughout the article, it matters little to them whether they read 7 August 2008 or August 7 2008. They both mean the same date. The insistence of the importance of seeing nation specific dating conventions does not bear out. I have read many, many articles related to non-English speaking countries that use the month-day-year style and have remained stable in that style for MONTHS after editing. If an outcry truly existed to adhere to a non-English speaking nations dating convention, wouldn't those non-English speaking nation readers have edited the article to conform? The argument presented for this convention would suggest, yes, a very strong YES. But alas, these articles remain unchanged as written. Where is this outcry from non-English speaking readers, and why aren't they here standing up for their belief? I see ONE native English speaking editor carrying the torch of a phantom conflict.--«JavierMC»|Talk 08:37, 31 August 2008 (UTC)

I withdraw from this merry-go-round. It's ridiculous to try and leave an opinion/argument when a response like the one below by Pete is made to refute it. The discussion was about whether to adhere to a non-English speaking countries dating convention, and now we have auto-formatting thrown in the mix when it in itself is being debated for general removal as useless in continuity of style for 90% of the users. --«JavierMC»|Talk 19:48, 31 August 2008 (UTC)
Autoformatting was introduced in order to end conflict over date formats. See Pmanderson's notes above. Remove autoformatting and we return to the previous situation, along with an increased potential for conflict and disruption. You can hardly argue that it's been quiet for the years editors have been shielded from raw dates and so it's going to continue that way when we return to the old method. Looking at recent ArbCom cases, I'd say that Wikipedians have become more antagonistic rather than less. --Pete (talk) 21:59, 31 August 2008 (UTC)
    • I believe as long as there is continuity in style throughout the article, it matters little to them whether they read 7 August 2008 or August 7 2008. The reason why we had date autoformatting in the first place is because a lot of Wikipedians cared very strongly about this precise point and autoformatting was introduced to end the bickering. Hardly a phantom conflict. --~~
  • Basically support but I don't see how it's supposed to differ substantially from the existing wording.--Kotniski (talk) 16:30, 31 August 2008 (UTC)
  • Support in principle, since it is the same effect as Woodstone's proposal (the difference from the present language is that both proposals will institute national dating format on articles written in a national variety of English and the present wording won't); but this is going around Robin Hood's barn to achieve the same effect. Septentrionalis PMAnderson 20:27, 31 August 2008 (UTC)
    • The language of the existing proposal is Articles clearly written in a national variety of English should generally use the more common date format for that nation; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format. This is, when combined with the strong national ties clause of WP:ENGVAR, equivalent to the above four tests. But if people want to make things more explicit, my only objection is length. Septentrionalis PMAnderson 21:04, 31 August 2008 (UTC)
  • Comment. I'm just puzzled as to why we have to be so proscriptive. Can't we keep it simple? Keep them in the same format, use an appropriate format to the article topic and if you don't know what that is use the format used by the first contrib unless consensus dictates otherwise. Yes, it likely is similar to what you have up there, but what you have up there is formalised too rigidly and doesn't allow for consensus to decide. Consensus should always decide, that's the content mechanism we use and is a founding principle. I'm fairly relaxed that consensus won't enforce British formatting on US articles or vice versa. Hiding T 23:31, 31 August 2008 (UTC)
    • Comment. The problem is in the "appropriate format for the article topic" part. Some editors want to lock in existing styles, so if an article on (say) former French president Jacques Chirac uses month day year format although France uses day month year format, the article cannot be changed. That's why we have the rigmarole about varieties of English in the proposal above, to exclude using the actual format used by a country. --Pete (talk) 00:33, 1 September 2008 (UTC)
      • If they can get a damn strong consensus of Wikipedians that says fine, we all agree, then that's what should happen. How likely do you think that is? And if there's no consensus then we use an appropriate format to the article topic. Consensus comes first: ALWAYS. In this case trust that consensus will do the right thing. I do. I appreciate that this conflicts with some people's desire to ensure every article looks the same, but that was never our goal and contradicts the principles upon which Wikipedia was established. That isn't to say that such a desire is a bad idea, it is just that it shouldn't get in the way of the main thrust of Wikipedia: letting anyone edit. Hiding T 08:29, 1 September 2008 (UTC)
        • Well, there's always someone coming along after to tidy things up. That's work I like doing. --Pete (talk) 11:29, 1 September 2008 (UTC)
  • Generally Support. I tend to agree, though, with Hiding that it seems overly proscriptive. The international style is common in certain circles in the U.S. with extensive contact with the “outside world”. With the modern military aerospace articles I’m most active in, I typically use it in preference to the U.S. style, even if it’s a U.S. topic – and have never had a complaint. In fact, I normally use it as well as in my professional work, as do many of my peers. It’s not that big a problem. In any case, I see two problems with the formula so far. First, it is that it’s overly focused on U.S. vs. Commonwealth; second, it ignores the large number of articles dealing with topics that are nationally “neutral” (e.g., science).
A simpler approach might be something along the lines of: Articles on topics with strong ties to a particular country should generally use the date format most widely used in that country. In cases such as Canada, where both formats are common, either format may be used in a given article, although that choice should be employed consistently throughout. In other cases, the type of format initially used in the article should be used thereafter. Askari Mark (Talk) 00:19, 1 September 2008 (UTC)
  • Comment. Looking further at the Jacques Chirac article, it's written using different date formats, different varieties of English - it has both "favor" and "honour", as well as "ize" and "ise" endings - and I don't know what variety of English France as a whole uses - French English, presumably - so it would fall through to the fourth rule above, in which case we look at the third edit, of 22 April 2002, in which is birth date is given as "Novermber 29, 1932" (sic). There must be vast numbers of articles in such a situation, where they were begun as stubs long before we worked out our current rules, and where development has proceeded using whatever style the editor thought was a good thing at the time. --Pete (talk) 00:50, 1 September 2008 (UTC)
  • Comment - I've only glanced at this debate, but I have to say I'd rather see a single date format used across all of en-Wikipedia. I prefer the day-month-year format (despite living in the US for most of my life) as there are technical advantages to it as well as being easy to get used to once you've been around it for a while. SharkD (talk) 04:40, 1 September 2008 (UTC)
Responses to Hiding, Mark Askari and Anderson: It seems that a little latitude for local editors to choose between the two standard formats as they see fit might be in order. Why not add the second sentence here to the lead, then: "Each of the following four tests should be applied in sequence until the format for an article is determined. However, this may be disregarded where are good reasons to do so (e.g., the US military usage of international format) and there is local consensus for this choice." Anderson, I do believe it's still shorter than the two sections it would replace—or at least not significantly longer. Its advantage is that it provides a simple sequence of tests that is likely to avert disputes. The sequence itself embodies the priorities. What do you think? Tony (talk) 04:39, 1 September 2008 (UTC)
Sounds good to me, but, with apologies to Tony, I've always wanted to do this: you might want to add a word into the suggested sentence for better flow. And no, I won't let the door hit me... Hiding T 08:29, 1 September 2008 (UTC)
That additional license is already included in the short form in my !vote: it uses should generally be for that reason. The four tests are unnecessarily long; the first two, which are the most important, simply repeat ENGVAR; we can do that by invoking it explicitly. Septentrionalis PMAnderson 16:18, 2 September 2008 (UTC)
  • Nearly support, there should be latitude to use the so-called international format in fields where that is the norm, such as articles related to the American military. Since the MoS is only a guideline, I think there is always latitide to depart in unusual circumstances, but military articles are not unusual. --Gerry Ashton (talk) 18:52, 1 September 2008 (UTC)
This map is based on cultural, political and geographic links between English and non-English speaking countries.
  • Fifty-Fifty - while the idea is good for places like the UK, I think (remembering a comment here) the simplest way to do it for non English-speaking countries would be for Africa, Europe, Oceania, the Caribbean and for most of Asia to use BrE and for the USA, Central and Southern America, Japan and China to use AmE (see map on right). ChrisDHDR 11:00, 2 September 2008 (UTC)
  • Comment: Oppose: I fear it's going to be a never-ending swamp of editing, reverting, and counter-reverting, if to determine the proper format for an article, one has to go into a minutiae of rules, exceptions, and finding out who used which date format first.
Instead, I propose we keep it simple: prefer the international format in all cases, except in articles strongly linked to the US, or any other English-speaking country that uses the M-D-Y format. Even if a non-English-speaking country X uses some other format, this is the English-language wikipedia, so we shouldn't be unduly concerned about that format.
I realise that Wikipedia is US-dominated, and that here as elsewhere, America's weight behind standards almost unique to itself makes things more complicated than strictly necessary. However, since Wikipedia is an international institution, with almost half of the en.wikipedia.org's contributions coming from outside the US, and the subject material covering the whole wide world and beyond, I think it should strive to use international standards as far as possible, in both measurement units and in language. This would make it simpler to determine which format to use in a certain article.
So, what I would propose for the use of date formatting: The international date format ("31 December 1999") should be used by default. Articles on topics with strong ties to the United States, ones concerning subjects where the international format is not the standard, should use the US format ("December 31, 1999").
And not to forget: Do not use date autoformatting, nor other techniques that make dates into links, unless the linked date has particular significance upon the context. Teemu Leisti (talk) 12:47, 2 September 2008 (UTC)
(Comment changed to Oppose. Teemu Leisti (talk) 21:03, 2 September 2008 (UTC))
  dd/mm/yyyy
  dd/mm/yyyy and yyyy/mm/dd
  yyyy/mm/dd
  mm/dd/yyyy
  mm/dd/yyyy and dd/mm/yyyy
  mm/dd/yyyy,dd/mm/yyyy,yyyy/mm/dd
Use the International DD-MM-YYYY format by default, unless the article is about something which has ties to another calendar format, in which case use that format, so for science articles we use international, for US articles we use American format, for EU articles we use international format, and for China articles we use whatever they use in China. NerdyNSK (talk) 01:40, 3 September 2008 (UTC)
I take DD-MM-YYYY to mean that I would format 4 July 1776 as 04-07-1776. That format is unacceptable because it is too easy to interpret as 7 April 1776. --Gerry Ashton (talk) 01:45, 3 September 2008 (UTC)
Sorry I mean DD-MMMM-YYYY. NerdyNSK (talk) 01:55, 3 September 2008 (UTC)
Both DD-MM-YYYY and DD-MMM-YYYY are NOT the international date format. Only YYYY-MM-DD is the international data format. Most of the Far-East already uses that by default. --2008-09-12 08:33 (UTC)
  • Support. As I understand it, this is more or less WP:ENGVAR applied to dates. GregorB (talk) 12:59, 2 September 2008 (UTC)

Discussion

Two (arguably three) of these tests amount to declaring dating format to be part of WP:ENGVAR and then applying it. This is repetitious. I have put

*Articles clearly written in a national variety of English should generally use the more common date format for that nation; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format; see WP:ENGVAR.

as representing the consensus above, without repetition. If someone really wants to write out the tests explicitly, that's fine by me; the only objections would be whether the shorter form is more readable, and whether the long form allows the little latitude that Tony, correctly, advocates above. Septentrionalis PMAnderson 16:34, 2 September 2008 (UTC)

Yep, I have to agree that it's more digestible and looks less bureaucratic. However, it needs to cover instances where the variety of English is unclear, which I'e done in the second bullet. The third bullet is just a release clause where local custom has consensus.

Which format to use

  • Articles clearly written in a national variety of English should generally use the more common date format for that country; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format; see WP:ENGVAR.
  • Where the variety of English is unclear and there is no clear tie to a particular English-speaking country, use the format chosen by the first major contributor to the article.
  • A format may occasionally be used regardless of these guidelines where there is good reason and this is supported by local consensus (e.g., US military usage of international format).
Editors are reminded that edit-warring over date formats is utterly unacceptable; issues concerning individual articles should be brought to WT:MOSNUM.

I'm assuming that the support for the four-test version above applies to this version, which differs only in the addition of marginal details. Thanks to Anderson, and to Hiding for initial conceptual input. Tony (talk) 02:23, 3 September 2008 (UTC)

This seems substantially good to me, and seems to follow consensus well. However, the first half of the first sentence seems to me to be saying the same thing as the second sentence – after all, all articles on WP.en are necessarily written in some “national variety of English”. Also, Canada isn’t the only English-speaking country in which more than one format is used; Canada is simply the largest and most noteworthy. Wouldn’t it be simpler to phrase the first bullet as
“Articles on topics with strong ties to a particular English-speaking country should use the date format most common for that country; see WP:ENGVAR. Where two formats are common, such as in Canada, either may be chosen, but this should be used consistently throughout the given article.”
Askari Mark (Talk) 03:24, 3 September 2008 (UTC)
  • I also substantially agree. My only quibbles with Tony's draft would be length, again, and that we prefer an established style to first contributor when both exist: If an article has been stable in a given style, it should not be converted without a style-independent reason. Where in doubt, defer to the style used by the first major contributor.
  • Not all articles are written in national varieties of English: some successfully evade the points of difference; all too many are a hodge-podge: some Scottish, some American, some Australian. The latter should be cleaned up; but in the meantime, we should allow them to use either format consistently.
  • My choice for a second paragraph would be Where the variety of English is unclear and there is no clear tie to a particular English-speaking country, use the existing dating format if one has been established in the article; if not, use the one chosen by the first major contributor to the article. Doubtless Tony can condense this. Septentrionalis PMAnderson 01:59, 4 September 2008 (UTC)
  • I'm also not sure about the final clause:issues concerning individual articles should be brought to WT:MOSNUM: do we want all that discussion here? This talk page is long enough already. Septentrionalis PMAnderson 02:01, 4 September 2008 (UTC)
I don't mind your suggestion for the second bullet, much as I hate to add another layer, and will insert it into Proposal 2 in the new section below. I'm hoping that the guidelines that emerge will minimise the uncertainty such that people rarely have to come here to ask for expert advice.

No consensus

I can't see any consensus above. A lot of the support votes are weak and heavily caveated. I repeat the point I made earlier. With the removal of date autoformatting, we need to get our position on this issue right. Whether a country uses English or not is immaterial. What matters is the date format used in the nation, and everything falls into place. U.S. articles use month-day-year, UK articles use day-month-year, articles with a strong tie to nations using one of these formats use that format, and articles with a multinational focus or linked to a nation using ISO year-month-day format use whatever the first major contributor used. That's simple and clear and we don't need to argue about whether Japan uses American Engrish or British Engrish. --Pete (talk) 10:23, 3 September 2008 (UTC)
That' funny—it looks like reasonable, although not overwhelming consensus for something the proposal. On the other hand, people have largely rejected your crusade, Pete: the articles are written in English, not Croatian, and date formats used in Croatian text are absolutely irrelevant to English-speakers. Tony (talk) 12:04, 3 September 2008 (UTC)
The only English language part of date formats is the names of the months. English-speaking folk use January, February and so on. I have no idea what names Croatian folk use for their months, but we won't be using them here. Or in articles about Croatia. Date format doesn't depend on language. As we can see for ourselves, with the English speaking world using two different formats. 1/2/2003 means different things in different countries speaking the same language. What matters is what format is used in the country, not the language they speak. This is a discussion about date format, surely? Not language?
It's not consensus. Consensus isn't a matter of counting hands and seeing whether there are more Ayes than Noes. See WP:CONSENSUS. Let me quote from that page:
  • In the case of policy and process pages a higher standard of participation and consensus is expected than on other pages.
  • Developing consensus requires special attention to neutrality - remaining neutral in our actions in an effort to reach a compromise that everyone can agree on.
  • a poll (if one is even held) is often more likely to be the start of a discussion than it is to be the end of one.
  • Minority opinions typically reflect genuine concerns, and the logic may outweigh the logic of the majority.
Looking at the various "votes" cast, I see just one whole-hearted Support vote. Every other editor has expressed reservations and misgivings. However, looking at GregorB's vote, he says, "this is more or less WP:ENGVAR applied to dates". Except it isn't. The guts of ENGVAR is that we use the variety of English that an English-speaking nation uses when we write an article about that nation. I don't think that there is any dispute over which date formats we use for English-speaking nations. Canada is a special case and all are agreed that either format is acceptable. The dispute is over which date format to use for non-English-speaking nations. Trying to decide whether Croatian English (whatever that might be) or Japanese Engrish is closer to American English or British English is something that seems to me to be a difficult and pointless task.
Looking at what variety of English an article is written in is another concept entirely. You might have two different articles about French cities, one written in American English, one written in British English, depending on who wrote them. As France is not an English-speaking nation, either variety is acceptable. More likely, you'll get a mix of styles, reflecting the various contributions by various editors. I've already highlighted the Jacques Chirac article, written using a colourful variety of American and British spellings. And date styles.
Let me list some comments from the poll above, which you claim shows a consensus.
  • use an appropriate format to the article topic --Hiding
  • Articles on topics with strong ties to a particular country should generally use the date format most widely used in that country. --Askari Mark
  • I'd rather see a single date format used across all of en-Wikipedia. I prefer the day-month-year format (despite living in the US for most of my life) --SharkD
  • there should be latitude to use the so-called international format in fields where that is the norm, such as articles related to the American military.' --Gerry Ashton
  • the simplest way to do it for non English-speaking countries would be for Africa, Europe, Oceania, the Caribbean and for most of Asia to use BrE and for the USA, Central and Southern America, Japan and China to use AmE (see map on right) --ChrisDHDR
  • The international date format ("31 December 1999") should be used by default. Articles on topics with strong ties to the United States, ones concerning subjects where the international format is not the standard, should use the US format ("December 31, 1999"). --Teemu Leisti
  • Use the International DD-MMMM-YYYY format by default, unless the article is about something which has ties to another calendar format, in which case use that format. --NerdyNSK
That's seven opinions that basically say the same thing - link the date format to the topic or the nation, not the variety of English. Add in my opinion, and you've got eight voices out of the thirteen editors who participated in the survey.
In fact, I really like Askari Mark's suggested wording, which seems to sum up the mood very well:
  • 'Articles on topics with strong ties to a particular country should generally use the date format most widely used in that country. In cases such as Canada, where both formats are common, either format may be used in a given article, although that choice should be employed consistently throughout. In other cases, the type of format initially used in the article should be used thereafter.'
My only addition would be to exempt countries such as China, which uses YYYY-MMM-DD format. That format is fine for tables and templates, not so good for written text. In such cases we should use whichever format came first. --Pete (talk) 00:07, 4 September 2008 (UTC)
The DD-MMMM-YYYY format is NOT "the international format". Please stop calling it that. Every time you say "International Format" I am going to assume you are talking about the YYYY-MM-DD format. --2008-09-12 08:40 (UTC)
"Link the date format to the topic or the nation, not the variety of English." Exactly. And only pay attention to the nation if it's an English-speaking nation; otherwise, use the international format by default.
To me, this is similar to the choice of using metric vs. US customary. "International" means exactly that; prefer that format over others.
I also agree with what you say about consensus. As there are only a few people participating in this debate (I myself stumbled upon it by happenstance a couple of weeks ago), why don't we add one of those requests-for-help that someone used a little while ago? Teemu Leisti (talk) 00:40, 4 September 2008 (UTC)
Just did it myself. Teemu Leisti (talk) 01:06, 4 September 2008 (UTC)
I partially support the version in the box. I would replace the second bullet point by this: Where the variety of English is unclear and there is no clear tie to a particular English-speaking country, use the international format. The article subject should be the only criterion for determining the date format. Teemu Leisti (talk) 23:18, 3 September 2008 (UTC)
This violates a very long-established principle of Wikipedia: The English Wikipedia has no general preference for a major national variety of the language.. Septentrionalis PMAnderson 02:11, 4 September 2008 (UTC)
We're not talking about varieties of English here. All varieties of English use the same month names, spelt the same way. We're talking about date formats. Using the date format appropriate to the topic or the nation seems like plain common sense, and a course best calculated to avoid causing offence/offense. --Pete (talk) 02:47, 4 September 2008 (UTC)
Shorter Skyring: Do it the way I do; it's only common sense, Next you'll be telling our American co-editors that they should have had sense enough not to rebel against George III. They have a different definition. Septentrionalis PMAnderson 02:54, 4 September 2008 (UTC)
Heh. You know, I think the boot is on the other foot. George III should have done everything he could to have held on to the Thirteen Colonies. He should never have allowed such foolish and unjust laws. He should have listened to and acted on the advice of the colonial leaders to keep them happy members of the Empire. It would have made the Napoleonic Wars and the First and Second World Wars very different. Better yet, they might never have happened. I'm wholeheartedly on the side of the colonists. They did the right thing in a difficult situation, and they established the first of the modern liberal democracies. I've stood in the hall at the Archives and looked on the Declaration of Independence, thinking of those brave and far-sighted men.[5] I've got facsimiles bought from the gift shop of the Declaration and the Constitution, and they are treasured possessions of mine. Because the Americans fought for liberty, we Australians gained our independence without a bloody struggle. We and the world have much to thank America for.
But we're talking about date formats. Not the American Revolution. Not the way we spell honor. Please keep your mind on the task, and if you disagree with me and your fellow editors, please be civil about it. That's just common courtesy. --Pete (talk) 03:31, 4 September 2008 (UTC)
Date formats and the ways we spell honour are equally distracting if the unexpected one shows up without explanation; I trust this sentence demonstrates itself. We have guidance on such subjects: don't mess with an established style. Please stop your crusade. Septentrionalis PMAnderson 03:44, 4 September 2008 (UTC)

Call for comments: two (now three) possible solutions

By Tony1: I believe that Pete is overstating the objections to the engvar proposal. Mark Askari, for example, has given in-principle support to it, yet he's counted as an objector. Pete and Teemu are proposing much more radical change, which would be capsized by significant objection among North American editors. I've been bold and put what I think is Pete's proposal into words as Proposal 1, which is surprisingly short and simple. I've restated the previous "engvar" proposal as Proposal 2 below it, with a slight adjustment per PMA. I'm keen to resolve this soon—this should have been done years ago.

With respect, I don't support the "international format for all but U.S. articles" proposal. It would lead to unrest. I've put forward a number of proposals, all along similar lines. I'm astonished that you haven't noticed. I'll take the liberty of adding my own (Proposal 3), based on the existing wording by Mark Askari. --Pete (talk) 05:06, 4 September 2008 (UTC)
Proposal 3 says nothing about new articles that are not country-related, and where the existing usage is mixed or otherwise unclear; can you address this, please? Tony (talk) 05:24, 4 September 2008 (UTC)
Articles on topics with strong ties to a particular country should generally use the date format most widely used in that country. Surely that covers new articles as well as existing. It's pretty much the same wording as has been in place for nine months and presumably new articles have appeared during that time. Where formats are mixed or unclear, I'd assume that a wikignome or bot will come along and sort it all out in due course. I wouldn't get too wrapped around the axles about it. --Pete (talk) 01:01, 6 September 2008 (UTC)

PROPOSAL 1: International date format except for North-America-related articles

Which format to use

Articles should use the international date format, except that those with clear ties to the United States should normally use the US format, and those with clear ties to Canada may use either format.

-- What do you mean by International format? Are you talking about DD MMMM YYYY (sic) or do you mean YYYY-MM-DD? Please clarify. --2008-09-12 08:44 (UTC)

PROPOSAL 2: Determine by the variety of English; if unclear, use the existing format, or the initial choice of format

Which format to use

  • Articles clearly written in a national variety of English should generally use the more common date format for that country; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format; see WP:ENGVAR.
  • Where neither the variety of English nor the tie to a particular English-speaking country are clear, use the existing date format; if this is unclear, use the one chosen by the first major contributor to the article.
  • A format may occasionally be used regardless of these guidelines where there is good reason and this is supported by local consensus (e.g., US military usage of international format).

PROPOSAL 3: Determine by national preference. If general or ISO, use existing format

Which format to use

  • Articles on topics with strong ties to a particular country should generally use the date format most widely used in that country.
  • In cases such as Canada, where both formats are common, either format may be used in a given article, although that choice should be employed consistently throughout.
  • Where a nation uses ISO (numeric year-month-day) format, or there is no strong tie to a particular nation, the type of format initially used in the article should be used thereafter.


My comments: Proposal 1 has three considerable advantages over the existing Proposal 2, and in the case of No. 3, the proposal by Teemu above:

  1. It's beautifully short and simple, and is likely to be understood and remembered by WPians at large. The twists and turns have disappeared, and the chain of contingency measure where engvar or existing formatting are unclear can go.
  2. The decision is always a one-step process on the basis of the title and the opening, not potentially four-step where an article is not US-related, involving the analysis of the text for the first sign of US or non-US English (sometimes you read and read and still there's no clue), and if that doesn't work, determining whether the formatting is mostly one or the other, or is in a mess, in which case you have to hunt through the edit history to determine which format was used initially.
  3. It side-steps the can of worms in having to determine which system a non-anglophone country uses, or did use at the time of the topic (modern China? modern Iran? medieval France? ancient Egypt?).
  4. It resonates with the international aims, scope and reach of WP (although this is a minor advantage, I think).

There are three significant disadvantages to Proposal 1:

  1. It is likely to offend some North American editors.
  2. It renders many many articles in breach because they use US format but lack "clear ties to the United States or Canada". (However, the human-supervised automated removal of DA would address this at no marginal cost, over time.)
  3. It means that many articles will be written in US English but will have international dates, which may be perceived as inconsistent. (However, this doesn't appear to raise eyebrows in many Canadian articles, given the considerable similarity of Canadian and US spelling, and in some US military articles, e.g. one I audited yesterday and did not touch. Is this a big deal?) Tony (talk) 04:22, 4 September 2008 (UTC)

Comments, "supports", "objects"

I ask for your forbearance in again commenting on, or better, stating your preference for one of these two solutions. It's a matter we should have decided many years ago.

  • Support Proposal 2. On reflection, I prefer Proposal 1 for its despite the practical simplicity of Pr. 1. Tony (talk) 04:22, 4 September 2008 (UTC)
  • My problem is with how to define 'strong ties' as opposed to 'clear ties'. The existence of multiple 'strong' ties in this globalised world for many subjects would imply much ambiguity and confusion. This rules out option 2, for me. Under option 3, we would find ourselves using ISO in articles related to China, where ISO is closest to the default way of looking at dates. However, let us not forget that this is English wikipedia, where the appearance of dates in this third (and uncommon for English-speakers) format would appear very peculiar indeed. So I too, would favour proposal 1 for its simplicity and unambiguity. Ohconfucius (talk) 05:49, 4 September 2008 (UTC)
    • Proposal 3 specifically excludes ISO dates. Perhaps the wording needs to be tweaked to make this clearer? --Pete (talk) 10:18, 4 September 2008 (UTC)
      • Clause 2 under option 2 is a bit of a cop-out IMHO, and could still result in edit wars if a contributor comes in and offers a major re-write and expansion of an article which was previously ambivalent. Under proposal 3, Chinese articles could well see a tendency to opt out of one or other by going for yyyy mm dd, the standard for China - not exactly ISO, but was what I was getting at. Ohconfucius (talk) 02:22, 5 September 2008 (UTC)
  • Proposal 2 is alright, because it ensures that the date formats should match the English variety used. But does this really need so much discussion and angst? We're never going to use dates like "07/02/2008" owing to their inherent ambiguity. So long as we use the full name of the month, there's very little confusion likely to result from the use of either date system in an article. This doesn't need to be turned into a US vs Everyone Else battle. - Mark 06:05, 4 September 2008 (UTC)
  • I'd go for Proposal 2. It is basically an application of WP:ENGVAR, an already well-established guideline. The current situation with date formats is probably closest to Proposal 2, while Proposal 1 is furthest from the current practice. Proposal 1 would introduce "consistency", but it would work at the expense of WP:ENGVAR: it is tantamount to preferring "color" over "colour" or vice versa. GregorB (talk) 06:46, 4 September 2008 (UTC)
  • Proposal 2. For reasons excellently expounded by Gregor.--Kotniski (talk) 07:33, 4 September 2008 (UTC)
  • Proposal 3. For ease and simplicity. The variety of English is immaterial when considering date formats, and difficult or impossible to determine for articles on non-English-speaking nations. Use the correct format for the nation. --Pete (talk) 10:16, 4 September 2008 (UTC)
  • Prefer Proposal 1, then Proposal 3, then Proposal 2. This is mostly on the basis of simplicity. Proposal 1: it is easiest to have a default with a small set of exceptions (the other two options do not have a default). Proposal 3: it is easier to determine a country of an article (sometimes just from the title) than the tests in Proposal 2. I have been trying to stay out of this debate but I decided to express my preference now. Frankly, I don't care strongly whatever you decide ... even no requirement for consistency would be fine with me (that would be the zero option with no visible change for most readers), as long as the links are removed. Lightmouse (talk) 13:41, 4 September 2008 (UTC)
  • Proposal 2. In addition to the reasons at #Finally, can we arrive at a solution, please? (mostly the analogy of ENGVAR); either of the others would involve massive changes, which are likely to be both controversial and error-prone. Pete's talk page already contains evidence of both, and that's just one editor. If Proposal 2 seems too long, we can leave out the second paragraph, which is only an application of our heading; also, this is clearly set forth in the next subsection now; we could just leave it where it is. Septentrionalis PMAnderson 14:25, 4 September 2008 (UTC)
  • Proposal 2. Requires fewest changes to existing articles, and avoids the requirement for editors to figure out the date format of a non-English speaking country (which may turn out to be unusable if the year comes first, or the country has no consistent way of writing Gregorian dates). --Gerry Ashton (talk) 14:35, 4 September 2008 (UTC)
  • I support Proposal 3, or failing that Proposal 2. This isn't a dictatorship. Deb (talk) 17:40, 4 September 2008 (UTC)
  • Proposal 2: support for same reasoning as GregorB --«JavierMC»|Talk 23:27, 4 September 2008 (UTC)
  • Not surprisingly, I support Proposal 1, whose pros and cons were well summarised by Tony. In my view, simplicity is the most significant of these. (It would be even simpler to have a rule that says "Use the international format for all articles.", but that's an unrealistic aim.) I would perhaps add a point similar to one used in Proposal 2: "The international format may be used in US-realted articles where there is good reason and this is supported by local consensus (e.g., US military usage of the international format)."
And remember, this is just about the order of the day and the month; if we can get this finally nailed down, we can then proceed with beautifying Wikipedia by unlinking dates. Teemu Leisti (talk) 01:12, 5 September 2008 (UTC)
  • AFAICT, the two issues revolve around dates, but are independent. There is nothing stopping the de-linking of dates in American and British articles. Only Canadian and a few other ones are ambivalent. Ohconfucius (talk) 09:19, 5 September 2008 (UTC)
  • Comment Looking at Canadian English, I find that there is only one variety of English used (apart from minor regional variations). So why does Proposal 2 allow two date formats? It's because Canada uses two different formats. Clearly date format is not linked to the variety of English. It is linked to the country. Could those editors supporting Proposal 2 explain this? And if we name Canada for special treatment, then why not name other nations? --Pete (talk) 10:52, 5 September 2008 (UTC)
  • For me, it's because Canada is an English-speaking country, so its dates are English dates. I see no reason to attempt to imitate the format of a country's non-English dates when writing English.--Kotniski (talk) 11:16, 5 September 2008 (UTC)
But Canada doesn't have "English" dates. Canada uses both American and British dates. The date formats allowed by Proposal 2 for Canada don't link date format to the variety of English, they link it to the country.
Pete, tell me, what date format was used in medieval France? And what about current practice in arabic, which doesn't use the same numerical symbols (even though what we use are ironically called arabic numerals)? Tony (talk) 12:29, 5 September 2008 (UTC)
Again, I wouldn't get too wrapped around the axles about mediaeval France - or any other subject where there are no clear links to a given date format. After a while it gets ridiculous, like trying to work out what date format you would link to articles written in (say) Thai English. --Pete (talk) 01:09, 6 September 2008 (UTC)


What format should be used for Bastille Day? Lightmouse (talk) 14:54, 5 September 2008 (UTC)

  • Le 14 juillet, which makes international the closest, I guess. ;-) Ohconfucius (talk) 16:18, 5 September 2008 (UTC)

So which proposal are you relying on for that? Lightmouse (talk) 16:24, 5 September 2008 (UTC)

I think we most all concur that the key unresolved issue is with what date format to use when there’s no natural tie to US/UK/Canada. Essentially these come down to the following: Option 1 imposes the “international” date format as the common default except where the US and Canada are concerned; Option 2 recommends following WP:ENGVAR and where that doesn’t apply, the following the choice of the first editor to make one; Option 3 calls for whichever is used in that country and initial use for “neutral” topics.
While Option 1 certainly offers the benefit of simplicity and brevity, I’m afraid it begs for reopening the ENGVAR conflict between those who would prefer it and those who might prefer a US-favoring alternative – e.g., “Articles should use the US date format, except that those with clear ties to other English-speaking nations should normally use the international format, and those with clear ties to Canada may use either format.” (After all, they can point out, Wikipedia.en is domiciled in the U.S.) It’s a dead-end debate. Option 3 is the most open-ended, and in general I personally have no real trouble with it, it has the distinctive drawback of requiring editors to know what date format is – or formats are – preferred in a given country. If it weren’t from these debates, I wouldn’t know China uses an ISO-style (I suspect its origins may predate ISO), nor would I know how to determine this since I neither read nor speak the language. In effect, when it comes to application, this option incurs the complication of requiring additional research before plunking down a date – which can be a really difficult challenge for non-English-speaking countries.
Option 2 offers the benefit of a tie-in to WP:ENGVAR, a fairly simple, already well-known and widely accepted rule, so it doesn’t require learning a new set of rules; moreover, it was established as the result of minimizing edit warring, and so might be able to extend that promise here to date formatting. However, the wording of Proposal 2 is only one way of expressing Option 2, and it is one I’ve earlier taken exception to in that it requires expertise in recognizing English dialects. Let’s face it, most editors can be expected to differentiate between “American” and “British” English, but Scottish or Australian from “British”? This is an unnecessary complication for date formatting. Accordingly, I’d like to offer a variant of Proposal 2 to incorporate some language from Proposal 3 that gets around all of this and keeps it simple. Askari Mark (Talk) 21:15, 5 September 2008 (UTC)

PROPOSAL 2A: Determine by the variety of English; if unclear, use the existing format, or the initial choice of format

Which format to use

  • Articles on topics with strong ties to a particular English-speaking country should generally use the date format most widely used in that country. (See WP:ENGVAR.) Where two formats are common, such as in Canada, either may be chosen, but this should be used consistently throughout the given article.
  • In all other cases, either the American or the international style may be chosen; the type of format initially used in the article should be used thereafter.
  • A format may occasionally be used regardless of these guidelines where there is good reason and this is supported by local consensus (e.g., US military usage of international format).

I agree with you that Proposal 1 might have problems. But Proposal 3 definitely not more difficult to apply than Proposal 2. The default assumption that everyone seems to agree on is that a country will use dmy unless we discover otherwise. It is extremely useful to have a default. In a less than a minute, using Proposal 2, I can search just by title and create a list of 1000 articles that need a date audit into consistent dmy. Using Proposal 3, I have to examine each article in detail. It might take me ten minutes to puzzle over english-variants (some of which I can't even detect as variations because they look normal to me) and then 20 minutes investigating edit history. Lightmouse (talk) 22:02, 5 September 2008 (UTC)

Again, I would like to work in established formats per the header and the section on Retaining the existing format (which we can just leave as it is). Otherwise we get the problem about "he cheated back in 2002" that has repeatedly come up in Talk:Yoghourt. But this can be dealt with easily if the principle has consensus. Septentrionalis PMAnderson 22:27, 5 September 2008 (UTC)
If we stick with "established formats", then Wikipedia never changes. The idea was to stop people changing dates expressed as BC and AD to BCE and CE. And back again. That caused a lot of disruption, eventually reaching ArbCom level. The "without good reason" exception to such changes allows format changes where there is a strong link to a particular format. An article on scientific methods in archaeology would be expected to use "BCE/CE", and an article about Jesus Christ, BC/AD. We've reached a similar duality in date formats for use in text. It's either 1 January 1901 or January 1, 1901. Using the YMD format 1901-01-01 in text looks odd in written English, though it works fine in templates and references. To my mind, where there is a strong link to a country, using the same date format as that actually used in the country makes a lot of sense. We don't need to differentiate between English-speaking countries and the rest. And, as noted, Canada speaks English, but uses both formats, so the language link breaks down in at least one case. We're not going to use French or Chinese or Arabic month names, so there is no reason at all to tie date format to language.
Correct me if I'm wrong, but I view "locking in existing formats" as shorthand for "battling to retain American date format", because most editors were American when Wikipedia was young, and that situation remains. American editors, in creating an article, will naturally use the date format they are best familiar with, even if it is inappropriate for the topic. Most editors don't go off and digest the Manual of Style before making an edit - they just do it. It's pretty easy to find articles on British subjects that use American dates, and that's because the initial editor was an American, or thought that American format was the norm. That quintessentially British poet, Rudyard Kipling, for example. Rupert Brooke, Elizabeth Barrett Browning, Kingsley Amis. Did they act in knowing opposition to the MoS, or did they just do what they thought was best? The MoS is really for those experienced editors who care about things like presentation and professionalism and consistency, and I resent it being used to try to push one format over another when both are valid or (as in the case of articles on European topics) the one being "locked out" is actually the better choice. --Pete (talk) 01:36, 6 September 2008 (UTC)
  • In support of Skyring’s post above, I’d hate to see editwarring, but the primary rule should be that date formats are article-appropriate. If some American editor managed to be the first major contributor on some particularly European topic, why not change the dates to make them fit better with the subject matter? It seems this “locking in” business is the product only of trying to keep editors from behaving like kindergardeners. If we really have to do that, then so be it. If not, let’s let articles improve. Greg L (talk) 02:26, 6 September 2008 (UTC)

Greg, we lock in engvar for articles unrelated to an anglophone country, and it works very well. I see no tension between US or international date format for an article on China or Egypt, or for that matter medieval France. I like Mark Askari's Proposal 2A, because it's probably easier, where there's doubt, to quickly go back to the first date inserted than to hunt through hoping to find the word "color/our" or some such. Tony (talk) 04:58, 6 September 2008 (UTC)

  • I see. Very well. Greg L (talk) 00:45, 7 September 2008 (UTC)
Askari Mark wrote: "Option 2 offers the benefit of a tie-in to WP:ENGVAR, a fairly simple, already well-known and widely accepted rule, so it doesn’t require learning a new set of rules; moreover, it was established as the result of minimizing edit warring, and so might be able to extend that promise here to date formatting. However, the wording of Proposal 2 is only one way of expressing Option 2, ... Accordingly, I’d like to offer a variant of Proposal 2 to incorporate some language from Proposal 3 that gets around all of this and keeps it simple."
Proposal 2B is almost there. However, I'm still not convinced we should have to look at initial use to determine later date use. This will make for inconsistent date styles even closely related articles, just based on the accident of what format the first date was in.
Edit-warring can be mitigated by the existence of a clear, widely-supported rule in the style guide: one only needs to point to it.
Furthermore, as I wrote earlier, Wikipedia is an international institution, and its subject material covers everything possible. So, I think it should strive to use international standards as far as possible. And as Greg L wrote earlier today: "In support of Skyring’s post above, I’d hate to see editwarring, but the primary rule should be that date formats are article-appropriate. If some American editor managed to be the first major contributor on some particularly European topic, why not change the dates to make them fit better with the subject matter? It seems this “locking in” business is the product only of trying to keep editors from behaving like kindergardeners. If we really have to do that, then so be it. If not, let’s let articles improve."
It seems we're close to reaching a solution here; let's not blow the opportunity to get it right once and for all.
So, I make Proposal 2B by varying Proposal 2A's second point:

PROPOSAL 2B: If the subject is closely tied to an English-speaking country determine by the country; otherwise, or if unclear, use the international format.

Which format to use

  • Articles on topics with strong ties to a particular English-speaking country should generally use the date format most widely used in that country. (See WP:ENGVAR.) Where two formats are common, such as in Canada, either may be chosen, but this should be used consistently throughout the given article.
  • In all other cases, use the the international style.
  • A format may occasionally be used regardless of these guidelines where there is good reason and this is supported by local consensus (e.g., US military usage of international format).
So, which is it? 2B or not 2B? Teemu Leisti (talk) 09:26, 6 September 2008 (UTC)
I'm prepared to support this one as a compromise, though what about the Philippines, which use U.S. format? --Pete (talk) 06:53, 7 September 2008 (UTC)
Thanks. If the Philippines the US format is the one generally used in the Philippines, then, per the first bullet point, that would be the format to be used in articles concerning the Philippines. Teemu Leisti (talk) 07:38, 7 September 2008 (UTC)
Mmmm, but they speak Spanish in the Philippines. --Pete (talk) 08:31, 7 September 2008 (UTC)
But the article on Philippines asserts that Spanish was discontinued as an official language in 1973, and "Since then, the two official languages are Filipino and English." Teemu Leisti (talk) 13:23, 7 September 2008 (UTC)
Ah. What do I know of the Philippines? Thanks for that! I should have checked Wikipedia, hey? --Pete (talk) 17:26, 7 September 2008 (UTC)
  • Support Proposal 3 with suggestions
    • Proposal 1
      • I'm confused by the wording
        • "North-America" could be confused with Canada.
        • What does "related to" mean?
          • Instead of "related to," I suggest "an overriding association with"
    • Proposal 2.
      • The last two words of the first sentence have a back reference to "that" country. Which country is "that" country?
      • This proposal seems to suggest, for example, that an article written about Chicago in a non-American style should remain in that style. I'm a Texan and I'll betcha you're glad I ain't plannin' on writin' an article about Cambridge, England.
      • Instead of "tie to," I suggest "overriding association with"
      • What is "local" consensus? The interested editors? The people geographically local to the topic?
      • Instead of "existing format," I suggest "prevailing format"
    • Proposal 3
      • "[N]ational preference" in relation to what?
      • Instead of "strong ties," I suggest "overriding association".
      • Instead of "Initially used," I suggest "prevailing usage"
    • Generally
      • There should be consistency within an article
      • A reference to a periodical's publication date should retain the format chosen by the publisher. (I don't mean a mention in the article, but in the formal cite/reference.)
      • Intelligent exceptions don't need to be mentioned, as they're covered by the WP:IAR pillar. Let's target an intelligent audience with reasonable judgment. I think this guideline will be best if aimed at consistency instead of at replacing judgment.
  • I prefer Proposal 3 with these suggested changes. --Danorton (talk) 15:15, 6 September 2008 (UTC)

Dan, I don't think there's a guideline saying that you have to format a reference the way the publication does. As long as the information is there, you're actually expected to change "Vol. 2, No. 1" to "2(1)" if that's the way you're formatting your whole reference list in a journal article or book. What one musn't touch is the spelling in the title of a paper/chapter and journal/book/conference proceeding (because it affects the searchability). But upper/lower case can and should be changed to conform to your house style. What I'm getting at is that the date format should be consistent within a ref list on WP, where possible, and it doesn't matter if it's changed to that which prevails in the main text. Tony (talk) 15:37, 6 September 2008 (UTC)

Okay, I think that such format should be preserved, but perhaps this isn't the proper forum for that discussion --Danorton (talk) 15:41, 6 September 2008 (UTC)
Both the American Psychological Association Style Guide and the Chicago Manual of Style have substantial sections about the publicaton date of various kinds of publications, and neither mention paying the least bit of attention to the date format in the work being cite. Following the format of the source is an undesireable departure from the normal practice of writing. --Gerry Ashton (talk) 15:43, 6 September 2008 (UTC)

How are we going, consensus-wise?

Leaving aside date autoformatting, which is a whole different kettle of worms, I think we're making solid progress on at least agreeing on some basics.

Points of agreement

  • ISO (year-month-day) format should not be used in written text: " Wikipedia was formally launched on 2001-01-15, as a single English-language edition at www.wikipedia.com." Although unambiguous, is unusual in written English.
  • We should use either 15 January 2001 (International format) or January 15, 2001 (U.S. format) for written text.
  • We should use an appropriate format for an article.
  • Where an article concerns an English-speaking nation, we should use the format commonly used in that nation. Except for Canada, which officially uses both formats, so either format is acceptable, if used consistently.

Points of disagreement

Where an article concerns an non-English-speaking nation

  1. Use International format
  2. Use whatever format matches the variety of English used in the article
  3. Use the format used in that country
  4. Use the format used by the first editor that added a date to the article (this line added, but not supported, by Teemu Leisti)

Changes from existing wording

The wording that stood for nine months was:

  • Articles on topics with strong ties to a particular English-speaking country should generally use the more common date format for that nation; articles related to Canada may use either format consistently.
  • Articles related to other countries that commonly use one of the two acceptable formats above should use that format.

The full text may be found at WP:DATE. Preceding sections deal with Wikipedia's preferred format (which equates to either 15 January 2001 or January 15, 2001) and subsequent sections deal with retaining date format unless there is a strong national tie. I think that the preceding and subsequent sections have strong consensus.

The long-standing wording has been challenged by one editor, and we are now engaged in a struggle to find a replacement wording, with various different proposals made, roughly equating to the points of disagreement above. We have had arguments for and against each point, and support for each point seems roughly even, with no clear consensus. --Pete (talk) 01:51, 7 September 2008 (UTC)

Comments

My own preference is for "3. Use the format used in the country."

Using International format throughout Wikipedia would lead to disruption unless strongly enforced. Likewise using U.S. format throughout. Tying date format to the variety of English used for English-speaking nations merely removes the decision one step beyond the real indicator: what format is actually used in that nation. The U.S. uses U.S. format and the U.K. uses International format and we don't need to look at the variety of English to see that this is true. --Pete (talk) 01:51, 7 September 2008 (UTC)

An example of Venezuela and Croatia was provided earlier in the discussion. I know what date format is used in Croatia: this is because I am Croatian. However, I don't know about Venezuela, just as I'm sure many don't know about Croatia - and I don't see why anyone should. Are we required to know what date format is used in 200+ countries of the world? Should there be a crib sheet for forgetful editors? GregorB (talk) 02:57, 7 September 2008 (UTC)
Looking at Calendar date, we find a list of nations sorted by format. Most nations use International format, so it's really a matter of remembering the exceptions, such as the U.S. and China. --Pete (talk) 06:44, 7 September 2008 (UTC)
  • That's a good summary of the state of the discussion. Except that there was a fourth opinion on the format to use for articles concerning non-English-speaking countries, which I added to the list.
We seem to have a consensus on most of the points, with only that one remaining. I support choice 1, "use international format", but I will not stand in the way if others clearly prefer some other choice. That said, I'll recap my reasoning: using the international format is (a) simple to remember and administer, and (b) international, as Wikipedia is. Teemu Leisti (talk) 07:38, 7 September 2008 (UTC)
We avoid decreeing international styles in other matters when the English-speaking world differs among itself. I don't see why date formats should be an exception. (It may be that Teemu Leisti thinks we should discard ENGVAR and construct an International English. I will oppose that, but this is not the place to discuss it.) As for the first point, it's all a matter of habit; I doubt American style is inherently harder, since those who use it find it easier. Septentrionalis PMAnderson 19:57, 7 September 2008 (UTC)

A point that is not agreed upon, but which I advocate: Although all-numeric dates in the format YYYY-MM-DD are allowed in citations, tables, and infoboxes, they may be interpreted as being governed by ISO 8601 so they should never be used in an article for years less than 1583, and should only be used to represent dates in the Gregorian calendar. --Gerry Ashton (talk) 21:35, 7 September 2008 (UTC)

I'd agree with this - in fact, there is a case for deprecating ISO 8601 dates altogether. As an editor, I like ISO 8601 (even if only in tables) - as a reader, less so (with DA effectively deprecated). But this is a separate issue, of course. GregorB (talk) 22:17, 7 September 2008 (UTC)
Concur with both above, after a little thought. If using ISO 8601 dates can lead to confusion, then we shouldn't use them at all, lest we mislead our readers. But of course this is a separate issue. --Pete (talk) 23:34, 7 September 2008 (UTC)
  • I agree with the “Points of agreement”. I don’t think it needs to be any more complex than that. I also don’t think that just because a UK editor (for instance) might be the first major contributor on an article on a national U.S. park and used the word “colour” a few times in the article means that the dates in that same article should be international format; the English-dialect spelling is a separate issue that shouldn’t darken the doorstep of what should be a straightforward rule-set on date formats. For those who might object to this notion, consider the opposite case; this same rule would apply to the reverse situation. As a major contributor to Kilogram, I used American English (“color,” “realize,” etc.). Does that fact mean I should have used U.S.-style dates for an article that doesn’t have a strong association with the U.S. because I used American English? Of course not. If anything, the kilogram has more to do with France than anything. So I used international date formating in Kilogram—but American spelling. While one newbie of a British editor not long ago tried to editwar over UK spelling, the article has been quite stable and no one has ever raised a stink over the international date formats in Kilogram; slight differences in date formats where the month is properly spelled out confuses no one. I think we really should just keep it simple with date formating; there is just no compelling reason to make this issue any more complex. Greg L (talk) 05:35, 8 September 2008 (UTC)

Decision on date format for non-English-speaking countries

I don't see the point of discussing this much longer, particularly since it's largely a matter of personal taste. The clear majority above support proposals which don't make the format used dependent on non-English date formats. Can we declare at least this issue settled (and thus stop the current edit warring over what version is the "existing" one?)--Kotniski (talk) 09:39, 7 September 2008 (UTC)

What's a "non-English" date format? I don't think anyone is proposing to use non-English date names. Nor is there any "clear majority" on wording, though I think that Proposal 2B is looking good as a compromise. --Pete (talk) 09:50, 7 September 2008 (UTC)
Well, you seem to want to use non-English date formats as a criterion for deciding which English date format to use. At least, that's how I understand your preferred wording.--Kotniski (talk) 09:56, 7 September 2008 (UTC)
I'm puzzled. We have International day-month-year, American month-day-year, and the ISO year-month-day. Which one do you think is "non-English"? --Pete (talk) 10:02, 7 September 2008 (UTC)
I think Kotniski is referring to "non-English-language" date formats. Tony (talk) 10:22, 7 September 2008 (UTC)
Well, what's a "non-English-language" date format then? DMY, MDY or YMD? I think I can see what you're getting at, but if the French use day-month-year, calling it a non-English format and ruling it out of the English-language Wikipedia is a bit of a stretch. I support using the format actually used in the country - which works regardless of whether they speak English or not - in exactly the same way that we use the temperature scale used in a country. The French use Celsius, but using your argument, each article would use whatever scale went with the variant of English used in that article, which I guess would be okay if you wanted to impose Fahrenheit on countries that use Celsius. Or impose month-day-year on countries that use day-month-year. --Pete (talk) 10:30, 7 September 2008 (UTC)
I think we've been through this already. Units are a different case altogether - choice of unit affects understanding, not just formatting style. The majority seem to support the view that date format should be treated like we treat spelling - just as we don't use the French spelling "couleur" as a guide to how we spell "colo(u)r" when writing about France, so we don't use the French "1 janvier" as a guide to whether we write "1 January" or "January 1". --Kotniski (talk) 10:39, 7 September 2008 (UTC)
I don't see this at all. And we've been through this before - NOBODY is advocating using non-English month names. Let me put it another way. When the French or Germans or Polish write a date, they write it as (say) 7.12.1941, and of course that means 7 December 1941, not 12 July.[6] We don't want to use a French or German or Polish month name, because this is the English-language Wikipedia, but we do want to use the same date format they use, in exactly the same way we want to use the same date format people use in the United Kingdom for articles about the United Kingdom. We're an international site, after all. The language is immaterial, because we're never going to use anything but English here. And no, I don't see a majority of people here linking date formats to spelling. Not unless there's a new way of counting. --Pete (talk) 11:30, 7 September 2008 (UTC)
I meant most people support proposals 2 or 2x, rather than proposal 3 which is the only one that concurs with your view. So I misunderstood your interpretation of the "current" wording. You now imply not that "'10 janvier 1999' gives '10 January 1999'", but "'10.1.1999' gives '10 January 1999'". That seems to me even less logical, since we're talking about quite different formats. Observe that in Britain we always write day before month when writing just figures (except in ISO), but it is not uncommon to see month before day when the month is written out. In other words, English speakers don't expect the written-out format necessarily to correspond to the numerical format.--Kotniski (talk) 12:00, 7 September 2008 (UTC)
Looking at how many people support tying the date format to English variety, it's nowhere near a majority. I'm guessing that you've counted me into your side because I've said that Proposal 2B is one that I can live with. We could work through each editor's views if you like. In any case, it's certainly not consensus. A poll is the starting point for discussion, not the end, as WP:CONSENSUS notes. I'm confident that we are making progress, but we aren't there yet. I'm sorry that you keep misunderstanding me. Maybe there's another way I can put it:
  • There are three elements in a date:
    • Day
    • Month
    • Year
  • There are three common ways of arranging these elements:
    • Ascending order: day-month-year International format
    • Descending order: year-month-day ISO format
    • Mixed order: month-year-day American format
  • By and large, each country will use one order in common use (say, on official forms). 1-2-3 means:
    • 1 February 2003 in the UK
    • 2 January 2003 in the U.S.
    • 3 February 2001 in China
  • Yes, there are always alternate formats in wide use in a country. The U.S. military uses day-month-year, English-language newspapers (even in England) use month-day-year, and July 4th and the Fourth of July are interchangeable all over. But, in general, each country has a preferred way of arranging dates, and although exceptions can always be found, trying to pretend that the UK commonly uses American format or vice versa is not likely to convince anyone with a working brain.
  • In Wikipedia we use the date format commonly in use in English-speaking countries in articles strongly tied to that country. Apart from Canada, where both formats are used officially.
  • The variety of English does not dictate the date format -as Canada demonstrates.
    • On the contrary, this demonstrates (if true) that the Canadian variety of English does not prefer a format. The others do not dictate, without exception, but they do strongly prefer. (It's not impossible to find, for example, honour in American either; but articles in American really should not use it, all the same; the exceptions here are on the same level.)
  • The names of the months in non-English-speaking countries are irrelevant. We're not going to use them in this English-language Wikipedia.
    • Yes, indeed, they are. That Skyring and he alone dragged them into this discussion shows him a sophist. Septentrionalis PMAnderson 20:23, 7 September 2008 (UTC)
  • I say that, like the English-speaking countries, we use the same rule for non-English-speaking countries.
  • In fact, this is the rule we've been using for nine months.
  • I cannot see any consensus for changing this to link article date format with variety of English.
  • What I can see are a few chauvinistic editors who feel every change of American format to International format as a blow to their vital organs and want to use every possible tool in their belt to stop this happening. This is about as helpful to our International encyclopaedia as declaring officially that we support one religion over another or that the Yankees are superior to the Red Sox.
This is an explicit appeal to construct an International English. Wikipedia is not an experiment in language reform; our purpose is to build an encyclopedia, in the English that now exists. Even if we had not agreed, long ago, not to prefer any dialect of it, Pete is wasting our time, and that of the editors whose work he complicated, in order to accomplish what he alone considers the Most Important Thing. Septentrionalis PMAnderson 20:23, 7 September 2008 (UTC)
No. I quite reject this. I'm certainly not advocating that Wikipedia use an International English - whatever that might be. We reflect the world - we don't try to change it, except insofar as we provide an encyclopaedia for everyone, which may change the world in a small and positive fashion. We use the varieties of English found in English-speaking nations in articles about those nations. Full stop. --Pete (talk) 23:26, 7 September 2008 (UTC)
In short, I say that articles strongly tied to a country should use the date format in common use in that country.
  • Apart from ISO year-month-day, which looks awkward in English-language prose. --Pete (talk) 17:21, 7 September 2008 (UTC)
    • As 7 September 2008 looks awkward in American, and September 7, 2008, does in British and Australian. Septentrionalis PMAnderson 20:28, 7 September 2008 (UTC)

Since Skyring claims not to understand what Kotniski is saying, let me explain. Only a week ago, Skyring was arguing that If we are writing articles about Venezuela, we should use the format common in that country, so long as it can be expressed in English. Kotniski is answering that contention, which would, if pushed, lead to writing in Spanglish; as indeed all too many of our articles are. I fully concur that the dating format used in Venezualan Spanish (which is an instance of what Kotniski means by "non-English date formats") should not determine our usage.

Skyring's attempt above, to claim that "non-English date formats" somehow means the Spanish names of the months is frivolous. It convinces me that Skyring has ceased to argue in good faith; if there is another reversion, I will therefore consider which venue of dispute resolution to employ. Septentrionalis PMAnderson 20:14, 7 September 2008 (UTC)

That's a rather inflammatory accusation, Pmanderson/Septentrionalis, and one I do not see supported by the facts. For instance, see section "How are we doing on consensus?" above, which is a good, productive summary by Skyring/Pete of the state of the discussion so far.
This whole enterprise is rather getting to be like arguing the number of angels that can dance on the head of a pin. However, precisely because it raises so much emotion, and because it concerns text elements that are used in nearly all articles, we should aim to arrive at a MoS guide that is easy to understand and to apply, and acceptable to a great majority of reasonable people.
Can the participants please read the section "How are we doing on consensus?" above, and say if they agree with the presented points of agreement and disagreement? If we are agreed on that much so far, then we can go at the final points of disagreement, hopefully in a spirit of mutual respect, cooperation, and Assuming Good Faith. Teemu Leisti (talk) 20:50, 7 September 2008 (UTC)
Thanks, Teemu! Pmanderson seems to have quite misunderstood what I've said. --Pete (talk) 23:18, 7 September 2008 (UTC)
Lets examine Proposal 2B. The American date format will be used when an article specifically applies to an American subject, otherwise the International format of Day-Month-Year will be used for all other articles(Oops, sorry lets not forget that Canadians can use whatever little ole format they care to use when it involves their countries ties). I think that's what, 90% or more of all articles will now be International format? So now Americans need to just throw out their taught format and get with the Day-Month-Year format, or stick to the 10% or less of the English Wikipedia that applies to American formatting if they don't like it. Of course for editors like Pete, who come from a country that uses Day-Month-Year format, maybe thinks, so what's the bother, huh? Could it be the fact that we Americans like to sit down at our computers and contribute to this English Wikipedia and feel comfortable whether we're writing about the rainfall statistics for Montana (US), Queretaro (MX), Wales (UK) or any other country, state or city upon this planet, without having to second think the dating convention most of us have used our entire lives? We are writing it in English. This is an English Wikipedia and I assume the entire English speaking inhabitants of this planet knows that 12 January 2008 and January 12, 2008 mean the same date. Why can't Americans sit down and write a new article on any subject matter we like, reference the content, add images, tables, or whatever manner of wiki-markup we like, using the dating format we are most comfortable with? What distracts most from editing and reading a Wikipedia article, is the variant of usage and non-continuity of dating format within an article, not which is used continuously throughout. If an article was begun using Month-Day-Year, or if it began using Day-Month-Year, then that is the format. This isn't algebra, calculus, chemistry or any other complicated formula that must be written in exacting form to derive a meaningful result. This is simply the presentation of a date, in English, where BOTH forms are understood throughout the English speaking populations of the world. It's nonsense. Stick to what came first with consistency throughout an article and we'll all deal with a little less policing, disruption and edit warring, and create an encyclopedia. Are we here to get bogged down in the bureaucracy of debating something that is understood and simply applied?--«JavierMC»|Talk 02:59, 8 September 2008 (UTC)
When people get emotional about date formats, it's hard to see their comments in an objective fashion. OK. Looking at the logic employed here, where you say that 10% of articles would be in U.S. format and consistency is a good thing, then that looks to be a powerful argument to make all date formats International. As for people editing articles, I don't think that too many worry overmuch about what date format they are using. Most would just put down their dates however they like regardless (or, more likely, ignorant) of what the Manual of Style says. Getting good content into Wikipedia is what really matters, and those who care about presentation and formats and styles can go through and tidy up later on. As has been repeatedly pointed out, nobody is going to get confused over the date format, whether it's 12 March 2004 or March 12, 2004. I think that, just as we say July 4, 1776 for an article about the U.S. then so too should we say 31 August 1997 for an article about England, 26 January 1788 for an article about Australia and 14 July 1789 for an article about France. In each case, that is the format actually used in the country. We're an international effort, you agree? --Pete (talk) 03:21, 8 September 2008 (UTC)
Yes, but an English-language international effort. The point I wanted to settle concerns just the last of your examples - 14 July 1789 for France. Of course it's not 14 July 1789 in France, it's (I guess) 14 juillet 1789. Is this French format of any relevance to us, writing in English? You say it is; the clear majority in the foregoing discussions support proposals that imply that it isn't. Can't you just accept that you're outnumbered on this one and let us focus on the other remaining points of disagreement?--Kotniski (talk) 07:28, 8 September 2008 (UTC)
I wouldn't insult the intelligence of your fellow editors if I were you. Let me ask you two questions:
1. Is 14 July 1789
a) day month year format or
b) month day year format?
2. Is 14 Juillet 1789
a) day month year format or
b) month day year format?
No need to answer - no matter what you say, you aren't going to be able to convince anyone that the date format used in France is anything other than day month year. --Pete (talk) 11:54, 8 September 2008 (UTC)

←Oh, no one will disagree with that. But you miss Kotniski's point entirely: it's of absolutely no consequence to the writing of text in English. I and, I'm sure, many other contributors here are tiring of this roundabout and your slightly narky tone, probably brought on by your sneaking suspicion that you're fighting a losing battle. I ask you to adopt a more reasonable stance and to help us move the debate towards a solution: what appears to be your anti-American drive will not gain traction (if it's not anti-American in intent, it certainly is in effect; please remember that about 70% of native anglophones are American). Oh, and one last thought: do foreign-language Wikipedias use the prevailing date-format in the country that an article is related to? Um ... no, they most certainly don't. Check for yourself. Tony (talk) 12:10, 8 September 2008 (UTC)

The simple fact is that here in the English-language Wikipedia, we use the date format appropriate to the country. We use month day year for the U.S. and we use day month year for the U.K. And we use the appropriate format for other nations. Have done so for years. Here's how the wording developed:
2004: It's generally preferable to use the format used by local English speakers at the location of the event. For events within Europe and Oceania, that is usually 11 February 2004 (no comma). For the United States it's usually February 11, 2004 (with comma).[7]
2005: It is usually preferable to use the format preferred in the variety of English that is closest to the topic. For topics concerning Europe, Australia, Oceania and Africa, the formatting is usually 17 February 1958 (no comma and no "th"). In the United States and Canada, February 17, 1958, (with two commas—the year in this format is a parenthetical phrase) is correct, and in Canada, 17 February 1958 is common..[8]
2006: If the topic itself concerns a specific country, editors may choose to use the date format used in that country. This is useful even if the dates are linked, because new users and users without a Wikipedia account do not have any date preferences set, and so they see whatever format was typed. For topics concerning Ireland, all member states of the Commonwealth of Nations except Canada, and most international organizations such as the United Nations, the formatting is usually 17 February 1958 (no comma and no "th"). In the United States, it is most commonly February 17, 1958. Elsewhere, either format is acceptable.[9]
Early 2007: If the topic itself concerns a specific country, editors may choose to use the date format used in that country. This is useful even if the dates are linked, because new users and users without a Wikipedia account do not have any date preferences set, and so they see whatever format was typed.[10]
Late 2007: Articles on topics with strong ties to a particular English-speaking nation should generally use the more common date format for that nation.[11]
2008:Articles on topics with strong ties to a particular English-speaking country should generally use the more common date format for that nation; articles related to Canada may use either format consistently. Articles related to other countries that commonly use one of the two acceptable formats above should use that format.[12]
The attempt to link date formats to the variety of English dates from a couple of weeks ago and has no consensus. --Pete (talk) 13:28, 8 September 2008 (UTC)
  • “International?” This is en.Wikipedia. Whereas there are a great deal of English-speaking Chinese, they do after all, have their own version of Wikipedia in their native language. I really don’t see a compelling case for using inappropriate date formats in articles closely tied to a particular country just because an article was authored by a particular editor who used a particular dialect of English with particular spelling. Just because a particular editor resides in central Africa and speaks with tongue clicks shouldn’t influence the date format used in September 11 attacks (which currently uses the international date format, which I think is asinine).

    And I know Tony disagrees with me on this point, but I don’t see the need for using international date formats in an article on a U.S.-related subject (like a U.S. national forest), just because a UK author was the first major contributor and used the spelling “colour” in the article. Autoformatting (and its linking to mind-dumbing trivia) is history. That’s good. Now we authors have to look at what we are making everyone else see. As for the formatting of dates, far too much turf war-mentality has pervaded the editorial community here. It should be one, simple rule:

There’s no legitimate need for any more complexity than that. If we are to start keying the date formats used in articles to the dialect of English the editor happened to use while writing the article, all the articles I’ve been writing (science-related topics) would have U.S.-style dates because I use U.S.-style spelling. That makes no sense. Keep it simple. Dragging English-language dialect complexities into this discussion is totally unnecessary. Greg L (talk) 19:14, 8 September 2008 (UTC)
Looking at the September 11 attacks article, it seems to have U.S. format throughout, which is entirely proper. But that's by the by. I still can't see any consensus for any particular way of doing things, and certainly not for any major change to the long-standing text, and the even more long-standing way of doing things, which is to have date formats tied to the nation's preferred format. If those who want date format tied to the variant of English chosen by the first editor are going to edit-war over this without first gaining consensus, then it looks like we're going to have to get some loftier eyes looking over it. I'd like to think that we can come to an agreement without doing this - after all, as I've shown, we kept more or less consistent practise for several years without refighting the War of 1812 over trivial issues - but progress on agreement seems to have stalled, with personal attacks becoming bolder. I'm also annoyed that this attempt at change is consuming time that could be spent more productively by all editors. The removal of date autoformatting is a thorny enough issue - why does it have to be made even more complex by changing a long-agreed way of doing things? I'm going to restore the wording back to the way it was, and if edit-warring breaks out again before we have a consensus for the change, then we can take it further and higher and make an admin or two earn his/her keep. --Pete (talk) 00:23, 9 September 2008 (UTC)
  • I was wondering why the hell you saw the dates differently on the September 11 attacks article. Blame the damned autoformatting. Though I happen to be an American, I had my user prefs set to international-style dates. Thus, because I am part of that 0.1% *privileged elite* that seemingly deserves to have a special view of the page no one else sees, I was seeing what regular readers don’t see. I just turned my date prefs off and I encourage all other editors to do the same so we always see what IP users see. In this particular case (the Sept. 11 article), the default date that most readers see is the appropriate one for that particular article. Greg L (talk) 03:18, 9 September 2008 (UTC)

Arbitrary section

No idea where in the above morass I should have put this, so apologies for starting a new section. I was directed here after having date links removed, and am utterly incredulous that autoformatting is being abandoned. I strongly oppose such a move. Granted, there are problems with linking all those dates to irrelevant articles, but that just means that a new improved autoformatting system should be developed and then phased in. The current system is not perfect. But not having any autoformatting at all is much worse. Furthermore, why was no-one ever informed of this until people started removing the links? The sheer volume of debate this has generated would appear to indicate that it is a controversial change that should have been better thought out. Modest Genius talk 22:46, 8 September 2008 (UTC)

I copy the responses to a similar statement above:
This has been debated for years, Necrothes. And it is finally coming to an end. The “linking” of dates is unwise and the autoformatting of dates was a terribly unwise decision made two years ago to resolve editwarring. Because of autoformatting of dates, 99.9% of Wikipeida’s readership (regular IP users) were often seeing unsuitable date formats in articles. Further aggravating the problem is we editors couldn’t even see the problems because our preferences settings were blinding us to this fact. The first step in fixing these articles is to unformat (delink) the dates. Then we simply write out the date in a method that is most suitable for the topic. The current MOSNUM wording…

Articles clearly written in a national variety of English should generally use the more common date format for that nation; articles in Canadian English may use either format consistently. Articles on topics with strong ties to a particular English-speaking country will therefore normally use the corresponding date format.

…is just fine; there should be absolutely no editwarring over the issue. See my above, expanded response (with the fascinating picture of Tokugawa Ieyasu) for more info on just why autoformatting was—and remains to this day—a piss-poor set of tools that never should have been made in the first place. Greg L (talk) 17:11, 5 September 2008 (UTC)
Necrothesp, I appreciate that you were unaware of any discussion, but discussion on depreciating autoformatting of dates, already longstanding, drew in a large number of editors on this page through July/early August as it moved towards consensus. It was then posted to the Village pump and to WP:Centralized discussions through August.[13] Additionally, several Wikiprojects were individually made aware of it, as were a large number of individual article writers. Thus, autoformatting was depreciated. 86.44.27.255 (talk) 19:37, 5 September 2008 (UTC)
Not to mention the fact that de-linking was even the subject of a "successful" petition to the developers that the developers simply ignored, in 2006. This is by no means some sudden change. — SMcCandlish [talk] [cont] ‹(-¿-)› 21:40, 6 September 2008 (UTC)
Teemu Leisti (talk) 03:27, 9 September 2008 (UTC)
None of which explains why autoformatting should be removed, or why dates should be unlinked before a new system is in place to handle them. Furthermore, dozens of items are posted to VP every day, how was anyone supposed to spot a single far-ranging policy change amongst all the rest? Modest Genius talk 20:24, 9 September 2008 (UTC)

Proposal to stop mass unlinking dates for 30 days and no other changes

After taking a few days away from this topic, I've come back in the hope of helping move this group forward. In that light, I propose a 30 day hiatus from mass unlinking of dates. This proposal in and of itself does not address any of the other concerns, justifications, proposals, etc. Offering support for this does not mean support for anything else, other than you want to start helping this group of people getting larger every day start rebuilding a working consensus here. I personally believe that there is a middle ground here although we have not found it yet. Please keep replies to one line with support/oppose/neutral. As an aside, and hopefully in good humor, please feel free to put the date you've replied in whichever format you prefer. dm (talk) 01:19, 9 September 2008 (UTC)

DA design process was fatally flawed; rip it out. User:Gerry Ashton JD 2454718.564595

I can't disagree with Gerry's comment: all is said in nine words. What I don't understand is the passion behind a few editors who are shouting from the rooftops about the moves by a number of editors to improve WP thus. They have gone as far as to mount an RFC against me, although as yet uncertified. There, the demand was for "a few days"; now I see the demand has blown out to "30 days"; next, they'll up it to what? Tony (talk) 02:01, 9 September 2008 (UTC)
I think that hanging would be too merciful in your case. One must always consider the possibility that a good drawing and quartering may be beneficial, pour encourager les autres. ;-) b--Malleus Fatuorum (talk) 02:16, 9 September 2008 (UTC)
Support a 30-day moratorium. None of this is personal, it's about what is the best thing for Wikipedia's readers. BTW, I don't have a choice of date format ;-) the ~~~~ signature uses dd month year. If I had a choice, I would use yyyymmdd (all numeric). Truthanado (talk) 02:22, 9 September 2008 (UTC)
No, it's about what's best for wikipedias's readers. Anyone remember them? --Malleus Fatuorum (talk) 02:31, 9 September 2008 (UTC)
Sure do. I have amended my comment. Truthanado (talk) 02:39, 9 September 2008 (UTC)
Sorry, Truthanade, just adding "readers" provides no substantive reason for a moratorium, so we're left wondering whether this is entirely to preserve the "preferenced" display of dates for a few WPians who can't stand "the other guys' month-day order". The readers should come first, not editors' whims, and your addition of the word "readers" is hollow without a cogent argument. Tony (talk) 02:43, 9 September 2008 (UTC)
It has been pointed out to me that my comment above may be open to misinterpretation. Just to be clear, my view is that this is another attempt at procrastination, another demonstration of the unwillingness of some to face the self-evident facts of the matter. I see that my attempt at what I thought was humour was misplaced, for which I apologise. Let me now say loudly and clearly that date autoformatting was a poor work-around to fix a non-problem that was only half understood, and that it ought to be eliminated asap, for the benefit of our readers. No need for more pointlesss delays. --Malleus Fatuorum (talk) 02:57, 9 September 2008 (UTC)
  • Autoformatting of dates is like dumping garbage from New Jersey into the ocean. Just because the practice had been done for a long time and there’s lots of garbage at the bottom of the ocean, doesn’t mean it was ever a good idea. It’s time to stop. Greg L (talk) 03:10, 9 September 2008 (UTC)
  • I wasnt aware of Wikipedia:Requests for comment/Tony1 until I went to Tony's talk page to ask about it. Bad timing I guess, I would not have suggested that approach. Please WP:AGF when considering this proposal. My main concern is the loss of the structured data, which I think does have value. As I said on Tony1's talk page, I agree or am neutral on many of the arguments here, but the tone of voice in this group is very unhelpful. Sarcasm, analogies to garbage, and snide remarks don't help you convince people, but they do start to convince people they're better off staying away. I really do believe there is a middle ground here that addresses the concerns people have raised. I just wish people were willing to work them out without trying to argue louder and longert. Ironically, the more people who find out about this due to the mass date unlinking, the more times that argument will take place. dm (talk) 03:28, 9 September 2008 (UTC)
  • Strong support — this was the goal of the RFC anyway. --UC_Bill (talk) 15:17, 9 September 2008 (UTC)
  • Support — there is a nearly identical proposal (by UC_Bill actually) that also includes a test where we turn off DA for some time, to see what the reaction from editors is positive or negative. --Sapphic (talk) 01:58, 10 September 2008 (UTC)
  • Would accept ten days' moratorium: I find it puzzling why this discussion is getting so nasty and personal. If we haven't sorted this out in 10 days, what makes anyone think it can be resolved in 30?? — there is nothing like a looming deadline to concentrate the mind. Ohconfucius (talk) 02:12, 10 September 2008 (UTC)

Pointless procrastination, be it 30 days or 10 days long. If this hasn't been sorted out in several years, what makes you think it will be now? The faultiness of DA has been thoroughly explained. Benefit of the readers? I don't see how. --Eleassar my talk 08:32, 10 September 2008 (UTC)

  • Support. At the very least, the linked dates should not be ripped out, but changed to a template, so we can put back the ones where Wikipedia:WikiProject Time subproject guidelines suggest they should be there. — Arthur Rubin (talk) 20:44, 10 September 2008 (UTC)
  • Oppose Agree with Eleassar. No compelling need to prolong this another 30 days. Professor marginalia (talk) 21:50, 10 September 2008 (UTC)
  • Comment If I count this through after two days, it seems to be 6 to 5 1/2 (Im calling the 10 day moratorium a 1/2). This obviously isnt a vote, and its clear the separation between the points of view is wide. I'd suggest that this does not yet represent consensus, so lets see what a few more days brings. dm (talk) 05:16, 11 September 2008 (UTC)
  • Support. There must be a way to make autoformatting work - and removing the markers for dates will make it very difficult to re-institute in the future. Has anyone suggested that there be a space for a marker/header for each article, and that if the user has NOT chosen preferences, that such header would determine which format is used? Headers might even be able to be added to US articles by a bot, based on categories. The default could even be dD MMMM YYYY unless the header has been put there. Another advantage would be that users who wanted the ISO format would still be able to see it --JimWae (talk) 05:26, 11 September 2008 (UTC)

Edit-warring on national ties.

I think it's about time the page was protected until consensus is agreed on the wording. I can't see anyone's proposal having support enough to warrant a change to the status quo. Changing the wording to your preferred version and hoping nobody notices a controversial change is no solution. Neither is edit-warring until one side drops from exhaustion. That's not the way we do things here. At least it's not the way things should be. --Pete (talk) 23:59, 28 August 2008 (UTC)

(Later) I've made a request for protection. It is obvious that a dispute exists and we should sort it out ourselves, rather than force some poor admin to trawl through the mess to make a decision via AN/I. --Pete (talk) 00:21, 29 August 2008 (UTC)

It is obvious that one editor, perhaps two, disagrees with the majority. That is not usually cause for page protection. It is also obvious that Skyring reverted three times and then called for protection; that's not done. Septentrionalis PMAnderson 02:38, 29 August 2008 (UTC)
You have repeatedly deleted text without consensus when there are several alternate wordings being discussed. Looking at the history of the wording, inserted by Tony in December last year, you let it ride for nine months, despite being a frequent contributor to this page. Why is it suddenly something you needed to edit-war over? I suggest it is because you noticed that I was changing articles on European subjects to European date formats, and you couldn't bear the thought. You edit-warred over my changes, which were in concurrence with the MoS, and then you edit-warred over the Manual of Style. --Pete (talk) 04:27, 29 August 2008 (UTC)
You have repeatedly declined to recognize that WP:Consensus can change. There is considerable evidence that it has, most clearly the #Straw poll above, but also the discussion at ANI; there is no evidence the old consensus, whoever it was, remains. Changing established formats is disruptive; you successfully disrupted the article on the South Ossetian war for some days. Please stop. Septentrionalis PMAnderson 13:00, 29 August 2008 (UTC)
I disagree with your assertions above, I don't see consensus support for any option, but we now have an opportunity to find a solution that doesn't involve edit warring. Hopefully one that both of us, all of us, will feel part of. --Pete (talk) 00:05, 30 August 2008 (UTC)

I've restored this section from archive to give some context. I think that we are making good progress on agreeing on points in dispute and that it is not helpful to edit war over the wording without consensus. I'm happy to see the long-standing wording changed to something that reflects discussion and consensus, but when an editor keeps changing the wording to suit his preferred POV, then this is not helpful in keeping discussion calm and civil. I ask that we keep the long-standing wording until we find consensus, whatever that may be.

Alternately, we could all edit-war our hearts out until one side or the other gives up in disgust. --Pete (talk) 00:01, 8 September 2008 (UTC)

Pete, I inserted the sentence only when Sandy put a robust case about Venezuala's date formats (international) vs the prevailing engvar used in that country. I did not intend that it have the implications it has turned out to. I agree with what Anderson says above; I wish you'd accept that you're outnumbered vastly in what seems to be an anti-US push, and that you're behaving aggressively. Tony (talk) 11:21, 8 September 2008 (UTC)
I don't see this at all, actually. Support for tying date format to varieties of English looks pretty soft. I see a lot of different views. As for aggressive behaviour, no, I think you should look elsewhere. What's your preference - edit-war over the wording or discuss it? --Pete (talk) 12:25, 8 September 2008 (UTC)

Misconception vis a vis date formats

Apologies for only having skimmed the above discussions, but I don't have hours upon hours to peruse it in detail. As such, I'm only going to comment on a recurring conception I noticed that I don't think is entirely accurate. Specifically, there seems to be an idea that MMMMM dd, YYYY, formatted dates are "the American way" of writing things. While I don't dispute that the format is dominant in the United States, it's not universal. I myself use dd MMMM YYYY dates and have since high school. Admittedly, it's uncommon, but the differences in date formats among nations are not nearly as consistent as the differences in spelling. Powers T 13:14, 8 September 2008 (UTC)

I don't know about that; the differences in spelling aren't absolute either. Honor isn't universal in American, although I would only consider honour in a tone more appropriate to wedding invitations than an encyclopedia; -ize is found in British English, largely on the grounds that it correctly represents the Greek, and so on. Septentrionalis PMAnderson 13:42, 8 September 2008 (UTC)
Stupid suggestion, but would it make sense to attempt to harmonize the alternate spellings and the date issue so that there's one set of rules for both. I know they're closely aligned presently, but it might help to make sure that the exact same distinction is made for both. --MASEM 14:08, 8 September 2008 (UTC)
You mean US spelling means US date-format in an article? It's one of the options above. Sometimes hard to detect the eng.var, though I don't think that's a great concern. The rules are really needed for where there's a sniff of a dispute in an article. Then it's worth hunting through to determine the eng.var. Tony (talk) 14:16, 8 September 2008 (UTC)
But that's my point: that calling mdy formats the "US date-format" is inaccurate at best. I don't think copying the spelling rule -- American spellings for American topics -- will work with dates. Powers T 17:11, 8 September 2008 (UTC)
I didn't say the differences in spelling were absolute. =) Powers T 17:11, 8 September 2008 (UTC)

There seems to be a lot of fiction about national date formats being bandied about, so I decided to do a Google sample of the Web to see who's using what.

  • When you type "international date format" into Google, the vast majority of sites talk about the ISO date format. Apparently, most believe that the international date format is the ISO format (2000-12-31).
  • A survey of British news sources indicates that, while most use European format (31 December 2000), quite a few use American format (December 31, 2000), and significant number switch back and forth, sometimes on the same page.
  • A survey of American news sources indicates that most of them use American format, but a significant number of institutions use European format, and a few institutions use ISO format.
  • A survey of Canadian news sources shows that most use American format, some use European format, and the ISO format is more common than the European format. Commonly, they use American alphanumeric format (December 31, 2000) with ISO numeric format (2000-12-31). The variant ISO alphanumeric format (2000 December 31) is also in use.
  • A survey of Australian and New Zealand news sources indicates that they use both European and American formats, and like the British they often switch back and forth.
  • As far as spelling is concerned, the Americans use American spelling most of the time, the British use British spelling most of the time, the Canadians usually use British spelling (often with the American date format), and the Australians and New Zealanders use both spellings, often inconsistently (the labour policies of the Labor Party)... (But I've also seen the U.S. government Report on the Sulfur Emissions of the Sulphur Creek Sulfur Plant)

So, notwithstanding all the debate, I don't detect a great deal of consistency in the national use of date formats, nor do I see much linkage between dates and spelling. It seems to be largely a matter of personal or institutional preference.RockyMtnGuy (talk) 15:38, 8 September 2008 (UTC)

I did a little further query experimentation on that first point. Compare +"international date format" +"8601" with +"international date format" -"8601" and we get different results than RMG found above. Hit counts were 1270 versus 5350. Google just ranked the +"8601" links higher (for whatever reason). I do note however that the W3C standard for Dates in HTML5 is quite unambiguous, as is RFC 3339. LeadSongDog (talk) 18:50, 8 September 2008 (UTC)
Even in the latter ("-8601") case, most sites are still talking about the ISO format without saying "8601". Most of the rest are talking about random things like the proper case to use for the month name in Greek...RockyMtnGuy (talk) 22:55, 8 September 2008 (UTC)
You can narrow it down further with a -"ISO", but I take your point. There are some interesting exceptions though., such as jbase, where "international" is used in contrast to "domestic" (US) in reference to the dd/mm/yy vs. mm/dd/yy distinction. My real point was that googlediving is far removed from solid research even if it is a useful indicator. LeadSongDog (talk) 17:10, 9 September 2008 (UTC)
You seem to be confused. Australians and New Zealanders use Australian and New Zealand spelling consistently. This is very similar to British spelling in most areas and is generally fairly consitent in the modern era. The Labor Party in Australia as a historical exception, explained in the article, and does not indicate an inconsistency (it's a name for starters) Nil Einne (talk) 04:48, 9 September 2008 (UTC)
Nil is correct. And Canadians do not "usually use British spelling". "Sulphur Creek" is a name, too, based on colonial spelling. Tony (talk) 07:59, 9 September 2008 (UTC)
Okay, try this Canadian web site http://www.chem.ucalgary.ca/asr/ for variant spellings. It belongs to the "Alberta Sulphur Research Ltd., Center for Applied Catalysis and Industrial Sulfur Chemistry, University Research Centre." It informs you that "Alberta Sulphur Research Ltd. was incorporated as a non-profit sulfur research organization...in 1964". Later, you find that "Dr. Clark is the Technical Manager of Alberta Sulphur Research Ltd. and Professor of Chemistry in sulfur chemistry at the University of Calgary". However, I take it that Australians and New Zealanders are a bit more consistent in their spelling.RockyMtnGuy (talk) 19:33, 9 September 2008 (UTC)
IUPAC has determined that "sulfur" is the spelling to be used by chemists worldwide. Tony (talk) 16:09, 12 September 2008 (UTC)

Rationalising MOSNUM and MoS (main)

It's high time we addressed that fact that both pages duplicate each other in many respects. I am to blame for this; it resulted from my major rewriting of MOSNUM ?two years ago, when I boldly added the main sections of the new MOSNUM to MoS (main); before then, MoS (main) was silent on all of those areas and did indeed just link to MOSNUM, although in a less well-organised way than we can do it now. In retrospect, transclusion would have been much better, and I can't think why this wasn't done. Anderson sets out the options at Wikipedia_talk:Manual_of_Style#Transclusions. Tony (talk) 14:13, 8 September 2008 (UTC)

nbsp and dates

As the delinking of dates rumbles on, I have to ask: is there any consensus on the use of non-breaking spaces in dates? Various archived discussions seem to have suggested their use both between day and month, and between month or day and year. On the other hand, some of the discussion suggests that its use is purely discretionary. I'd like to overhaul my portfolio of articles to comply with the new MOSDATE, and I'd appreciate advice as to the extent to which I should use non-breaking spaces. Choess (talk) 17:47, 8 September 2008 (UTC)

  • It’s a good idea Choess, but I don’t know if one can mandate such nuances; perhaps we could have wording on MOSNUM that encourages that as a “best practice.” It takes a long time editing on Wikipedia to learn about such things as non-breaking hyphens and non-breaking spaces. Personally, I use a non-breaking space between the date and the name of the month so I don’t have a orphaned numeral hanging at the end of a sentence (in the case of Euro-style dates), nor a line beginning with a numeral (in the case of American-style dates). I think it is better to not mandate so much in a collaborative writing environment: just let new editors simply write out dates, and let more experienced editors do cleanup when they feel the need. Greg L (talk) 18:35, 8 September 2008 (UTC)
Isn't that what MoS is for, though - telling the experienced editors how to clean up (and resolving conflicts between them as they do so)? New editors are probably only hazily aware of MoS, so they will do the same whatever we write. --Kotniski (talk) 18:42, 8 September 2008 (UTC)
It's a good practice as far as typography goes, but it further erodes the readability of the markup, even for experienced editors. Powers T 19:46, 8 September 2008 (UTC)
  • Well, there are more disruptive markups being routinely used on Wikipedia than Sept.&nbsp;11, 2001. How much worse is that non-breaking space that {{nowrap|Sept. 11, 2001}}? (Yes, I know these two example markups don’t do the exact same thing; it’s illustrative). So I don’t think it can reasonably be called an issue of experienced editors having a problem with this sort of simple stuff (the ability to deal with that would pretty much define “experienced editor”). The issue is just a subtlety: do we start mandating advanced markup techniques in style guides. Maybe. But I don’t think that is best. Greg L (talk) 20:36, 8 September 2008 (UTC)
    • I don't know that either is being used routinely, but the latter is certainly more readable. I'm not suggesting "experienced editors" can't "deal with" a non-breaking space; I'm suggesting that the readability is decreased. Which it is. And the benefit of using the non-breaking space has to be weighed against that loss of readability. Powers T 23:04, 8 September 2008 (UTC)

We prefer dates with the month spelled in full. I think any markup that joins the day, month, and year into one non-breaking unit is too long, and the jaggedness of the resulting rendered text is a greater problem than wrapping the date. If there is to be any non-break space, I'd put it only between the day and month. --Gerry Ashton (talk) 20:46, 8 September 2008 (UTC)

  • I agree with you Gerry. It should be “10 October 2001”” or “October 11, 2001” (spelled out fully). Still there shouldn’t be “October
    11, 2001”, a non-breaking space improves it immeasurably. I entirely agree with Choes that this is best typography practices. The issue, in my mind, is whether or not one requires the non-breaking space. I would say that if we wanted to mandate the use of a non-breaking space (always a good idea), one might be doing good if MOSNUM provides guidance to new editors as to how exactly one makes a non-breaking space and clearly explains why one should do so. Greg L (talk) 20:51, 8 September 2008 (UTC)
Whatever happened to the recent mantra that registered editors should see the same thing the reader sees? Is it suddenly inconvenient in this context?LeadSongDog (talk) 21:16, 8 September 2008 (UTC)
What exactly is being seen differently in this example? Powers T 23:04, 8 September 2008 (UTC)
  • Gerry was objecting to abbreviating the name of the month. And I agree. I don’t abbreviate month names anyway and was just being sloppy with making an example of using non-breaking spaces. Greg L (talk) 03:06, 9 September 2008 (UTC)

A non-breaking space could easily be provided when the date is autoformatted... Gimmetrow 03:22, 9 September 2008 (UTC)

  • Which will be wonderful and will rank right up there just under antibiotics, Gimmetrow—after the feature is made to work for everyone instead of just we editors (which just glosses over editorial problems with date formatting and makes 99.9% of our readership look at junky, inappropriately written text). But don’t hold your breath for those tools; it would take a lot of tinkering under Wikipedia’s hood to make that happen. Greg L (talk) 04:02, 9 September 2008 (UTC)
    • The autoformatting code in MediaWiki already does similar formatting things for everyone, right now. It would not take much tinkering to have [[9 September]] produce 9&nbsp;September. Gimmetrow 13:29, 9 September 2008 (UTC)

Won't ALT+0160 (with Windows) produce an nbsp that looks like a regular space in the edit, e.g. September 9, 2008? Or, is this what is meant by a literal hard space? If so, what kinds of problems are caused by using literal hard spaces? --Elliskev 12:42, 9 September 2008 (UTC)

It sounds like things are coalescing around nbsp between month and day, at least for U.S.-style dates. I would think that the same would apply for international-style dates: 11
October 2001 also seems like poor line breakage. I agree that trying to glue the entire date together with nbsp is overkill. While it might be nice to see dates explicitly mentioned in WP:NBSP, I think it's best to recommend, rather than mandate, them for the time being, because of the legibility issues. Unfortunately, I don't think there's an easy fix for those. The problem with using an ALT-160 space is that it is very difficult for an editor reading an article to know whether or not a given space is a literal hard space. If I'm trying to fix dash problems in an article, for instance, I can always scroll up to the text of an article and see whether a given dash is a hyphen-minus, en-dash, or em-dash. But it's impossible to tell by casual inspection to what extent ordinary whitespace in an article has been replaced by non-breaking space. Maybe if a WYSWIG editing tool were ever built, but for now, I'd say we have to put up with the written entities. Choess (talk) 14:36, 9 September 2008 (UTC)
  • Tony1's comments: Let's first remember that the DA mechanism has never prevented wrapping; as well, the dates after our signatures wrap, which hasn't seemed to bother anyone. SandyGeorgia, the FA delegate, is hugely in favour of adding the hard-space between month and day (= day and month). Lightmouse, who wrote the DA removal script, is doesn't like it. So I'm caught between two friends who have opposite takes on this. The effect of the hard-space is nice, except in a few places adjacent to an image or infobox where there's a bit of text-stretching results from it; but that's no big deal. I do worry that it looks bad in the edit-mode and is a pain to key in; I desperately wish that Noetica's push six months ago for a short-cut like the double comma hadn't failed (like most proposals put to WikiMedia, it seems). I'm on the fence. Tony (talk) 14:52, 9 September 2008 (UTC)
    • I should prefer not. FAC does not need to require more things that are not intuitive and are hard to key in; it spends too much time on that now. Septentrionalis PMAnderson 16:03, 9 September 2008 (UTC)
    • A non-breaking space between the month and day in February 14 may prevent a line beginning "14, 1958". Such spaces should be considered; they may not be worth the trouble and the less readable code, and they may cause other bad line breaks, so they are entirely a matter of editorial discretion; since they are not visible, they need not be included uniformly within an article. is what I would put if I were Jimbo, but it may not be worth the space; the first sentence, may imply the whole thing. Septentrionalis PMAnderson 16:16, 9 September 2008 (UTC)
Without the link, allowing a line break between month and day looks bad. But having to add &nbsp;s to every date would be a pain, and hard to read while editing. If autoformatting was able to convert a linked date in accordance with a user's preference, could it be made to produce an unlinked, hard-spaced date? If not, what about creating a set of templates to do that? So {{1 January}} would produce "1&nbsp;January" by default, or "January&nbsp;1" for users with that preference. And the other way around for {{January 1}}.
—WWoods (talk) 17:22, 11 September 2008 (UTC)

Birth dates?

Should birth dates be linked? Please leave any comment at Wikipedia talk:Only make links that are relevant to the context#Birth dates?. --Eleassar my talk 19:09, 8 September 2008 (UTC)

  • Linked? Like, to trivia? So instead of clicking on an Easter-egg-like date in hopes of reading more on the birth of the person, they are taken to a long, tedious list of mind-numbing random trivia that has nothing whatsoever to do with the person the article is about? I hope not. Greg L (talk) 20:43, 8 September 2008 (UTC)
For the reasoning see the linked discussion. --Eleassar my talk 21:02, 8 September 2008 (UTC)
  • P.S. Caveat: With one exception. I would suggest that one could link years (never dates) in intrinsically historical articles, such as French Revolution. In such an article, if one wrote as follows:
…then I wouldn’t have a problem. It would give readers interested in history an opportunity to find out what else was going on during those years. Greg L (talk) 21:04, 8 September 2008 (UTC)

Ruminations on date linking

I just removed [14] the text which allegedly presents the arguments for delinking deprecation. I don't have time to follow every detail of this page, but I questioned these a while back, and the arguments haven't improved. The reasons were faulty because

  • autoformatting does something for non-registered and non-logged-in users. Not only does it involve some formatting, but everyone benefits from what was, I think, was the main reason for the date linking, the reduction of date-format wars.
  • rendering the linked dates without the links would solve this (and see later)
  • the iso format implying a Gregorian date is pretty much irrelevant. The guideline already says don't use iso-style dates in the text, and publication dates in refs should match text. Accessdates for online material will presumably be late enough that this isn't a concern.
  • idioms should work in the text as written.

Datelinking may have some disadvantages, but the disadvantages presented so far are weak. Hardly seems worth throwing the whole system out if these are the best you can come up with after this much time. These linked dates represent a substantial investment in metadata, and it has uses. Rendering linked dates without links keeps the metadata investment while addressing the "overlinking" issue.

I see the faulty reasons have already been restored. If you cannot defend the arguments for linking after this much time, then those reasons will be removed. Gimmetrow 02:55, 9 September 2008 (UTC)

  • Well, I undid that edit but had an edit conflict with Gerry, who already did it seconds beforehand. And my edit summary comprised “We’ve discussed this long enough.” Greg L (talk) 03:04, 9 September 2008 (UTC)
    • After "long enough" if you can't come up with better reasons than those four, they you have nothing. Gimmetrow 03:07, 9 September 2008 (UTC)
  • We’re going in circles with you and going nowhere with you Gimmetrow. Please don’t be disruptive to MOSNUM. Your “not getting it” is beginning to look like an intentional mental block. I’m baffled why you could possibly not understand the concept that giving editors a special view of Wikipedia’s pages causes nothing but problems and only sweeps editorial problems under the rug. So you can understand, here is some sample text. Take a look at it while logged in and while logged out.
Do you see now? To save other editors the effort of having to log out, here is how the above was coded:

During the French Revolution, the Estates-General convened in Versailles on [[May 5]], [[1789]] and lengthy speeches by Necker and Lamoignon… and then on [[2005-05-05]] a memorial was built on the site…

And here is what virtually every single user of Wikipedia sees (because they aren’t *special* like we registered editors):

During the French Revolution, the Estates-General convened in Versailles on May 5, 1789 and lengthy speeches by Necker and Lamoignon… and then on 2005-05-05 a memorial was built on the site…

But for a registered editor (one in a thousand users) logged in with “Euro-style” preferences, here’s what they see:

During the French Revolution, the Estates-General convened in Versailles on 5 May 1789 and lengthy speeches by Necker and Lamoignon… and then on 5 May 2005 a memorial was built on the site…

It should be quite clear that both date formats 1) are inappropriate for use in an article on the French Revolution, and 2) they appear lousy or awkward for the vast majority of Wikipedia’s readership, and 3) editors who are logged in and have their user pref settings adjusted to their suiting can’t even see where these problems exist! As editors who are largely responsible for this stuff, we’ve been marching along all fat, dumb & happy, totally oblivious to what we’ve been making the vast majority of our readership look at. As I stated above, we editors should never use tools that allow us to look at regular editorial content in a way that is at all different from what everyone else has to look at. The only preferences settings that affect what we see should be date offsets for when edits are made—that sort of thing.
Now, if this goes anywhere like it has with other editors arguing the same point you are (“autoformatting is wonderful”), this is the point where you reply how we ought to simply then make autoformatting work for all users, including IP users. Been there; done that. Please get point and stop being disruptive. We’ve gone all over this stuff; I can’t help it if you don’t get it. Slapping {disputed} tags all over MOSNUM because *Gimmetrow isn’t pleased* isn’t going to get you your way.

And I’d say you pretty much burned your bridges when you began your argument here with “ I don't have time to follow every detail of this page.” You might as well have said “I didn’t participate in all the discussions—and don’t feel like getting up-to-speed anyway—but don’t like the consensus that developed so I deleted what has been here for a week since it offended my sensibilities and then slapped a {disputed} on the text after I was reverted.” That was so not a smooth move on your part. Greg L (talk) 03:40, 9 September 2008 (UTC)

If you cannot justify the deprecation, then you need to remove the false reasons. If you cannot bother to do your homework, then there's nothing further for me to say. Whatever consensus might have been for date delinking, it's gone, derailing any hope of date delinking for now. Gimmetrow 03:58, 9 September 2008 (UTC)

  • I just did justify it. Read the above post. If you still don’t “get it,” you’re apparently not trying to understand the issue anymore and I can’t help you. Greg L (talk) 04:04, 9 September 2008 (UTC)
    • Greg, you have inappropriately removed a disputed tag without resolving the dispute. You have not justified anything. You are formally asked to either remove the disputed "reasons" or restore the notice. If you do not restore the notice or fully resolve the dispute immediately, I will consider your actions a user-conduct violation. Gimmetrow 04:08, 9 September 2008 (UTC)
  • The consensus for deprecating autoformatting doesn’t disappear just because Gimmetrow fervently wishes that to be so and solemnly pronounces that there is no consensus. I’ve seen these very tactics with a couple of “won’t let go” editors over the IEC prefixes. Doesn’t fly. Just because an editor places a {disputed} tag on text doesn’t mean that action was justified and legitimate. You already admitted that I don't have time to follow every detail of this page and then you had the gall to come back and slap the {disputed} tag on MOSNUM instead of having stayed here and discussed and debated the entire time. And all of that was after you deleted text and another editor reverted that stunt of yours. If you want your voice to remain relevant, stay in the discussions and don’t make pronouncements that you don’t even have time to get yourself up to speed on the issue. What you did is basically vandalism by someone who didn’t want to devote the time that others have invested into this debate. Greg L (talk) 04:19, 9 September 2008 (UTC)
  • I patiently waited some weeks, watching this page for any defense of those "reasons". If you refuse to defend them, then I will consider them indefensible, and remove them. Gimmetrow 04:32, 9 September 2008 (UTC)

The links are of marginal value as metadata because they fail to capture which calendar the date is in, and whether the date is AD or BC. --Gerry Ashton (talk) 04:11, 9 September 2008 (UTC)

  • I am quite done further soiling myself with this debate tonight. Someone else will have to deal with Gimmetrow for the next 48 hours. I’m fed up. Goodbye. Greg L (talk) 04:21, 9 September 2008 (UTC)
  • Gimmetrow, you amaze me. Let's take two central statements you've just made above:

"autoformatting does something for non-registered and non-logged-in users.... the reduction of date-format wars." "rendering the linked dates without the links would solve this (and see later)"

Um ... can you link to any edit war over the removal of DA? Pete's little wars in a few places (of which I disapprove) over the selection of the date format in articles related to foreign-language-countries started well before the deprecation of DA; such selection has nothing to do with DA—they concern what our readers see. As Greg L has pointed out countless times, we don't want editors to have blinkers on. DA doesn't stop edit wars, it stops us from properly managing the dates our readers see. Have a look at the pie graph of the appalling state of our date management. If you did some date auditing yourself, you'd be confronted by this reality. Your notion of retaining autoformatting, but "unlinked" is dangerous: we won't be able to tell what is and isn't autoformatted, and we will continue to be blind to what our readers see; thus, the poor management of DFs will continue. Tony (talk) 04:45, 9 September 2008 (UTC)

    • How can we be blind if, as is claimed, only a tiny fraction of registered users have a pref setting, and therefore most editors see exactly what readers see? And the date format wars were quieted back in 2003 with dynamic dates. For the past five years, anyone set on one style could set a preference. Gimmetrow 04:50, 9 September 2008 (UTC)
I don't know who's claiming that, but I think it's been mangled since—only a tiny fraction of readers, not logged-in registered WPians, have the autoformatting blinkers on. Tony (talk) 07:56, 9 September 2008 (UTC)
Doesn't that beg some questions? Do registered WPians "have autoformatting blinkers on"? Gimmetrow 01:14, 10 September 2008 (UTC)
Yes, I think we generally do. I took mine off some time ago and urge others to do the same. JIMp talk·cont 17:15, 10 September 2008 (UTC)
So, even if we were to agree date prefs were "blinkers", you don't have them. Gimmetrow 02:36, 12 September 2008 (UTC)

The conclusion here seems to be that date linking for autoformatting is deprecated, but the actual reasons behind that deprecation are weak and disputed, and the section still presents none of the reasons against date delinking. Certain editors refuse to recognize that there is any dispute at all, and refuse to engage other editors. What can we do about this? Gimmetrow 02:36, 12 September 2008 (UTC)

My view is the reasons for deprication are quite strong, and the only reason for it that has any merit is it helps to find dates that a person thought should be autoformatted. If, in the future, the software is improved to do an adequate job of autoformatting, the new mechanism could either operate on the dates that are already linked, or the linked dates could be transformed by a bot into a new XML tag or template that the new system could process. I personally have done all the waiting for vapourware as I intend to in this lifetime, but others may see it differently. --Gerry Ashton (talk) 04:03, 12 September 2008 (UTC)
What does "vapourware" have to do with this? Multiple programmers have written or offered to write code and work to resolve problems, but nobody here seems willing to support them. But that's not central to the holes in the four "reasons" currently presented in the guideline page. Just to pick the last "reason", currently written as
The syntax of dates differs from format to format: the American format should be followed by a comma when the syntax of the sentence as a whole would not otherwise require punctuation; the International format should not be. This requires either awkward and unintuitive input requirements or an editing system capable of evaluating English syntax.
This assumes the text needs to follow these punctuation rules 1) not only as written, but for all possible date preference settings, and 2) that this is something we must avoid. Neither is an obvious given. We follow other restrictions designed to make pages accessible to readers with special needs. Even if it that doesn't apply here, perhaps that "small minority of Wikipedia’s readership" can accept that, if they choose a preferred date format, some punctuation may read incorrectly. Gimmetrow 04:28, 12 September 2008 (UTC)
Software that someone has offered to write is certainly vapourware. I also view any software that has not been through a successful beta test as vapourware, but I suppose some people might not be that stringent. To me, vapourware is that which people delay action in anticipation of, when there is little assurance the software will go into production (no matter the reason). --Gerry Ashton (talk) 04:34, 12 September 2008 (UTC)
Again, irrelevant tangent. Gimmetrow 04:49, 12 September 2008 (UTC)

Date format resolution attempt

Once again, this time in the (perhaps overly optimistic) hope of finally reaching agreement. The principal alternatives are summarized in four options—A, B, C and D. Please note two points:

  • Common ground: For all four options: there is no change to the current situation where there are ties to English-speaking countries, except that the rule for Canada-related articles is clarified: "Use the date format chosen by the first major contributor".
  • Headings: Since the only distinguishing feature of each option is how articles without ties to English-speaking countries are treated, the headings encapsulate only this feature.
(A) Engvar. Date format follows language variety; month-day format is considered a feature of U.S. English, just as the spelling "color" is. Where both ENGVAR is unclear and there is no strong tie to an English-speaking country, the "first major contributor" rule applies.
(B) First major contributor. Where there is a strong tie to an English-speaking country, use the most common date format used in that country. Where there is no tie to an English-speaking country, the "first major contributor" rule applies.
(C) Day-month default. Use day-month format unless strongly tied to an English-speaking country that uses month-day.
(D) Country-driven. Use whichever of the two acceptable formats most resembles the date format used in the country in question, even if non-English-speaking. If this cannot be determined, the "first major contributor" rule applies.

I suggest a poll on these four alternatives, with concise arguments (if any) since we've probably been through them all at length anyway. Or maybe I've missed something out. Anyway, once we know what principle we're adopting, we can decide on the precise wording.--Kotniski (talk) 10:49, 9 September 2008 (UTC)

May I suggest that Greg L set up one of his tabular polls such as he's done for other difficult issues? It would be be useful if everyone could mark each of columns A, B, C and D with a number drawn from this list (i.e., any of these numbers for each of Options A, B, C and D):

1 = I like this a lot.
2 = I like this mildly.
3 = I don't favour this, but I could live with it.
4 = I object to this.
5 = I strongly object to this.
0 = Complete opposition
1 = Could be much better
2 = Ambivalence
3 = Could be improved, but I support this
4 = Complete support

A blank in a square would represent a neutral attitude to that option.

Perhaps the process of adding the signature could be simplified from the previous tabular pools—just write your unsername in the name column and sign underneath the table. No one would bother falsifying, since the edit history reveals all. I'll ask Greg at his talk. Tony (talk) 11:14, 9 September 2008 (UTC)

This sounds good. I'll go alert the previous participants once the table is set up. I've added Option D, which is pretty much the long-standing status quo, as there are a handful of nations such as the Philippines and Pelau which use U.S. format. --Pete (talk) 11:49, 9 September 2008 (UTC)
I've struck out the original options and replaced them with the (completely opposite) options actually provided with the table. Having larger numbers to indicate higher support seems more intuitive. --Pete (talk) 03:07, 10 September 2008 (UTC)
I've clarified the options and removed my precipitous vote and the queries which necessitated the clarifications, to avoid clutter. Votes for option D will need to indicate a default to A, B, or C for cases where no country is involved. Awaiting the table (but might be best to hang on a bit longer in case of further tweaks to the list of options).--Kotniski (talk) 12:12, 9 September 2008 (UTC)
I agree with Kotniski about tweaking. Three things.
(removed two resolved - hope no-one minds this anti-clutter comment-removal policy - Kot)
  • Option A will need to default to "first major contributor" if ENGVAR can't be determined. Tony (talk) 12:19, 9 September 2008 (UTC)
I think that follows from ENGVAR, but try to clarify if you like.--Kotniski (talk) 12:30, 9 September 2008 (UTC)

←OK, I've rewritten them with the intention of preserving the meanings, except for filling in a few minor holes (I think Pete's Option D in particular, needed this). Kotniski, please revert or modify them as you see fit. Could others please offer suggestions here to forestall chaos, if you can bear to keep your fingers off them—two editors directly tampering is more than enough, and comments/suggestions from others are very welcome. Tony (talk) 14:24, 9 September 2008 (UTC)

Well, I'm happy to relinquish that role to someone else, particularly as Tony and I might be felt to be on the same side of the debate (and I've got work to do...) I will just make one final tweak to D though. --Kotniski (talk) 14:31, 9 September 2008 (UTC)
And we need option (E) Engvar. Established usage. Date format follows language variety; month-day format is considered a feature of U.S. English, just as the spelling "color" is. Where the variety of English is unclear, stay with an established format. Where both ENGVAR is unclear and there is no established format, the "first major contributor" rule applies.
I think this is just a more detailed wording of (A); it doesn't need to be listed as a separate option.--Kotniski (talk) 16:15, 9 September 2008 (UTC)
I also would prefer strong ties in (A) as it is in WP:ENGVAR; else we will have bafflegab about the ties of Adrianople to England. Septentrionalis PMAnderson 15:59, 9 September 2008 (UTC)
I agree on the need for strong ties. Tangential involvement isn't really enough to disregard the warnings (originating at ArbCom level) about changing styles. --Pete (talk) 03:07, 10 September 2008 (UTC)

All four options assume a single date format is used in each location. I don't think that's accurate, as I brought up above. Powers T 14:02, 9 September 2008 (UTC)

I think it's understood that, where it says "used", it means "most commonly used". If not even that can be determined, then we have to make an exception (as for Canada). But it's the general principle we need to decide first, before we address the detailed wording and possible exceptions.--Kotniski (talk) 14:24, 9 September 2008 (UTC)
Anderson: I've added "strong", but you do realise that this narrows the applicability of the tie when in the negative, don't you? Can you check through from that perspective? As for Option E, do you mean "there is no established date format in the article. If I'm right, I think I'd rather write it into A. Thus, there's a distinct chain of decisions. Does anyone object? We'll have to retain "and there is no strong tie to an English-speaking country, otherwise a US-related article without a distinctly US-spelt word is in trouble if in day-month format already. Just thinking through the worst-case scenario so that we're covered. Is this what you mean?: "Date format follows language variety (month-day format is considered a feature of U.S. English, just as the spelling "color" is). Where the variety of English is unclear in an article without strong ties to an English-speaking country, retain the established format in the article. Where both ENGVAR and the existing date format are unclear, the "first major contributor" rule applies." I'm off to bed. Tony (talk) 16:34, 9 September 2008 (UTC)
Even "most commonly used" is misleading. Frankly, I don't see the difference among options A, B, and C; they all require mdy format in articles "strongly tied" to the United States, which I think is unecessary. D has the same problem. Where's the option for "always use the first contributor's choice"? Powers T 22:50, 9 September 2008 (UTC)
  • I object to (C) with or without strong, so I don't see much difference there.Septentrionalis PMAnderson 17:37, 9 September 2008 (UTC)
  • I thought there might be a real disagreement about established, by which, yes, I mean in the article. I'd rather say it, so that we can change without somebody digging up a stub from 2002 and claiming that means he gets his way; and I see Kotniski thinks it's implied anyway. Septentrionalis PMAnderson 17:37, 9 September 2008 (UTC)
Common sense would hopefully apply, as well as common co-operation. We're working together, not every editor out to battle and win to write Wikipedia in his preferred polemic. --Pete (talk) 03:07, 10 September 2008 (UTC)
  • All: With regard to Tony’s 11:14, 9 September 2008 proposal, I’d be pleased to set up a vote table here. I’m tied up for the work-part of the day (west coast U.S.), but will be pleased to do this tonight. By the way, so we don’t confuse experienced editors here, I think we’ll set up the “degree of support” values just as before: greater support equals higher value. Greg L (talk) 18:15, 9 September 2008 (UTC)

Poll on guideline for writing fixed-text dates

Which guideline do you support for determining the most suitable fixed-text date format (2 February 2009 or February 2, 2008) editors should use in Wikipedia’s articles? If a clear general consensus develops for one particular option, we will go with that option. Otherwise, we will conduct a run-off vote on the two options that have the greatest number of points. The current options are as shown below:

(A) Engvar (date format follows dialect of English most suitable for the article, default to first major contributor—no list of countries): Date format follows language variety; month-day format is considered a feature of U.S. English, just as the spelling "color" is. Where both ENGVAR is unclear and there is no strong tie to an English-speaking country, the "first major contributor" rule applies.

(B) Go with first major contributor unless there is a strong tie to an English-speaking country—no list of countries: For articles on, or strongly associated with a particular English-speaking country, use the most common date format used in that country. Where there is no tie to an English-speaking country, the "first major contributor" rule applies.

(C) Default to international unless U.S. and its territories—listed countries for editors’ convenience: For articles on, or strongly associated with, the U.S. or its territories (or countries listed in this guideline that use U.S.-style dates: Micronesia and Palau), editors should use the U.S.-style date format (“February 2, 2008”), otherwise, editors should use the international date format (“2 February 2008”) in articles. The dialect of English used for the article (“colour” v.s. “color”, etc.) shall have no bearing on the chosen date format. New articles on or strongly associated with Canada should use the international format but, for existing articles related to Canada, whichever format was used by the first major contributor shall be retained.

(D) Country-based, even if non-English-speaking—no list of countries: Use whichever of the two acceptable formats most resembles the date format used in the country in question, even if non-English-speaking. If this cannot be determined, the "first major contributor" rule applies.


Note that the below chart does not allow ~~~~ signatures to be used. You must copy/paste or hand-edit your signature. For assistance in writing your signature, you may copy the time below in red from the preview window after clicking on “Show preview” while in edit mode:

04:04, 28 December 2014 (UTC)


0 = Complete opposition
1 = Could be much better
2 = Ambivalence
3 = Could be improved, but I support this
4 = Complete support
DEGREE OF SUPPORT FOR OPTION 
Editor   A     B     C     D  
Greg L 0[1] 2 4[2] 2
JavierMC 2 4 0 0
Teemu Leisti[3] 1 0 4 1
Truthanado 0 0 4 0
Sapphic[4] 0 0 0 0
Powers[5] 0 0 1 0
Pete 0 0 4 4
Kotniski 3 4 2 1
Jimp 4 3 0[6]
Anomie[7] 3[8] 0[9] 0[10] 0[11]
Donald Albury 0 0 2 0[12]
GregorB[13] 4 3 2 1
Septentrionalis 3.5 3[14] 0[15] 0
Alexf 2 0 4 0
Aervanath 4 4 2 1
SharkD 0 0 3[16] 2
NerdyNSK 0 0 3 4
Remember the dot 3 4[17] 1 0
Jeandré du Toit 0 0 1 0
MJBurrage 0 1 4[18] 1
Occuli 0 1 4 1
Deb 4 1 1 4
Arthur Rubin 3 1 0 4
Danorton[19] 0 0 2 0
Woodstone 2 1 4 3
SMcCandlish 4 0 0 0[20]
Professor marginalia 4 3 2 1
Twas Now 1.5 4 3 0
Mdcollins1984 3[21] 3[21] 1[22] 2[23]
Edison 4 1 0 1
Christopher Parham 4[24] 4 0 0
Askari Mark[25] 4 4 3 2
Ohconfucius[26] 1 0 4 1
Necrothesp 0 0 4 0
Orderinchaos 0 0 4 4
Hiding 0[27] 2 4[28] 2
ChrisDHDR 0 0 3 3
OwenBlacker 3 2 3[29] 1
Total 66.5 56 86 46
Average 1.75 1.47 2.26 1.21

The average scores, as of the 0-0-3-3 vote by ChrisDHDR, are as follows:

Option A = 1.76
Option B = 1.50

These are invalid statistics, since A and B are actually the same thing, and even averaging them will produce incorrect numbers, since some respondents rated one higher than the other, not realizing they amounted to the same concept in different wording. Discarding the lower values of the two, the real stat is inserted below. — SMcCandlish [talk] [cont] ‹(-¿-)› 03:41, 12 September 2008 (UTC)
Interesting logic. How do you account for editors giving different levels of support for options A and B? It seems that people perceived them as different. (Later) In fact, looking at your own vote, you have "Complete support" for Option A and "Complete opposition" to Option B. --Pete (talk)
Option A/B = 2.01

Option C = 2.24
Option D = 1.22
See Run-off poll below.

Poll comments
  1. ^ My opposition to option A (Engvar, or “English variation”), is that I am an American and use American English in all my contributions. I also routinely use international-formated dates in articles that have no strong ties to the U.S., such as Kilogram. Note that the French are more closely tied to the development of the kilogram than any other country and the kilogram clearly has no strong national tie to the U.S. So it makes no sense in my book to require that dates in the Kilogram article follow the U.S. convention (“February 2, 1799”) just because the article was written using American English. Greg L (talk) 18:30, 10 September 2008 (UTC)
  2. ^ I don’t see the need for using international date formats in an article on a U.S.-related subject (like a U.S. national forest), just because a UK author was the first major contributor and used the spelling “colour” in the article. IMO, it should be one, simple rule: There’s no legitimate need for any more complexity than that. If we are to start keying the date formats used in articles to the dialect of English the editor happened to use while writing the article, all the articles I’ve been writing (science-related topics, such as Kilogram) would have U.S.-style dates because I used U.S.-style spelling in those articles. That makes no sense. Keep it simple. Dragging English-language dialect complexities into this discussion seems totally unnecessary.

    The whole point of making it country-based is to make the article read as smoothly as possible for the likely audience. If it is an article on a U.S. National Park (likely a readership dominated by U.S. citizens), then use U.S.-style date formats to make the article as natural for that audience. For everything else, use the international date format. This should all be about making Wikipedia better for our readers; we needn’t be so concerned about hurting the feelings of editors. Greg L (talk) 01:08, 10 September 2008 (UTC) (UTC)

  3. ^ I completely agree with Greg L's comment. Plus, as bad as basing the choice upon the variety of English used in the article is, I dislike even more having to go through the article edit history to find out which format the first editor used. The choice should impinge depend only on the subject of the article, not on historical accidents of that sort.

    And as for the proposed "default" format being day-month-year: take a look at the dates produced by ~~~~ signatures. I haven't noticed objections to this. Teemu Leisti (talk) 01:09, 10 September 2008 (UTC)

  4. ^ Dates should never be plain text. Autoformatting makes this entire poll irrelevant.
  5. ^ I would prefer options that don't require articles with strong ties to the United States to use MONTH DAY YEAR format, because while common, that format is far from ubiquitous in the U.S. -- Powers T 02:02, 10 September 2008 (UTC)
  6. ^ How they write dates in another language has no relevance on en.WP. International date format juxtaposed with US spelling is no great shock—the US military does it. Similarly international spelling juxtaposed with US date format is not shock either—Canadians often do it ... some non-Canadians do so too. The distinction between new & old Canadian articles will be too hard to maintain ... indeed the idea that Canadian articles should be in international format is wrong-headed (much as I prefer international format): Canadian do use both. User:Jimp 10 Sep 08
  7. ^ !votes were solicited at WP:VPR, so here is mine. I will not watch this page for replies, as I am uninterested in arguing over this issue. IMO, too much of this months-long debate has been an exercise in WP:POINT for various points. I also do not look forward to the day when the date warriors decide all references must not use YMD format, as I greatly prefer that short format in the references list. Anomie 11:43, 10 September 2008 (UTC)
  8. ^ Option A is the only consistent option. The rest are even more encouraging people to fight and editwar over the format. WP:ENGVAR might need updating to indicate which countries use which formats. Anomie 11:43, 10 September 2008 (UTC)
  9. ^ Same a Option A, except it allows someone to come along to a random article with an established spelling but no dates and force a conflicting date format (for WP:POINT or WP:POV reasons). Anomie 11:43, 10 September 2008 (UTC)
  10. ^ Frankly, Option C seems to be pushing some sort of "DMY is the best" POV, and invites insane levels of speculation over who this "likely readership" really is—most cases where the "likely readership" can really be determined can probably be more easily settled by applying the "strong national ties" rule. Anomie 11:43, 10 September 2008 (UTC)
  11. ^ Option D: How they write dates in another language has no relevance on the English-language Wikipedia. Anomie 11:43, 10 September 2008 (UTC)
  12. ^ As an American, I want to say that I think this issue has been overblown. I'm quite comfortable with using the 'international' date format everywhere, it is perfectly clear as long as we spell out the months instead of using numbers. User:Donald Albury 11:52, 10 September 2008 (UTC)
  13. ^ A), as an extension of WP:ENGVAR, appears to be the most neutral and well-established solution. B) is effectively very close, as the first major contributor will likely use "matching" date format and variant of English. D) is barely passable: it is almost as if it prescribes using rhymes when writing about poetry. GregorB 12:24, 10 September 2008 (UTC)
  14. ^ Needs a note about established styles --Septentrionalis
  15. ^ although I oppose both of them wholeheartedly, C is worse than D, because it is unclear what countries it applies to. (If the argument is that the Philippines is effectively an English-speaking country, the way to deal with this is to use A, and argue the point when it arises. But I doubt it is, and we do have a Tagalog Wikipedia for this reason; and a Spanish one too. -- Septentrionalis
  16. ^ Actually, I would prefer a fifth option: use the international format in all articles, about all countries, in all of English Wikipedia. -- SharkD
  17. ^ I like proposal B's flexibility when handling Canada-related articles, and think that proposals C and D are just going to make lots of people mad. I'd also like to point out that articles such as Acid2 use U.S. format for the prose, but international format in the timeline because international format is more logical in tables. -- Remember the dot
  18. ^ This is the best option presented, but to be honest I am an Americain who would prefer that all of Wikipedia just adopt "day month year" as its standard. All countries that use English use this format (including America), even in those few where it is not the most common it is easy to understand in prose etc. For tables the standard should be ISO (YYYY-MM-DD) for both space and sorting reasons. --MJBurrage
  19. ^ They're all profoundly vague and arbitrary, so we'll no doubt revisit this soon. You can't get much more arbitrary than "first major contributor," although I have to admit that it isn't vague. They're all vague as to how "country" relates to the issue. (What country does "Ocean liner" relate to?) For (A) and (B), "strong tie" is vague because most articles have "strong ties" to multiple countries. (Both the US and Australia have strong ties to Nicole Kidman). For (C), the TV soap "Dallas" is "related to the U.S.," but it's more popular outside the US than within. Few here would mind if you put the dates in Chinese. (And I live in Texas!). Similarly, for (D), what the heck does "country in question" mean, anyway? What's the question? -- Danorton (talk) 21:53, 10 September 2008 (UTC)
  20. ^ Options A and B are identical, if you bother to actually read WP:ENGVAR! I have zeroed B because we already have ENGVAR and it has served us well; whatever MOSNUM says on this should also be repeated (not necessarily at full length) in ENGVAR, and the general advice at ENGVAR should not be contradicted in any way by MOSNUM without an incredibly good reason (KISS principle - editors will not follow guidelines if they become too convoluted). C is unnecessarily complicated, and is already covered by A/B except for the two countries that use US format dates but do not have strong ties to the US (they can simply be specifically mentioned if anyone feels this is really necessary). D mistakes the map for the territory, the menu for the meal. Whatever version prevails, ENGVAR should be updated to account for it, not just MOSNUM. PS: This complicated table is a really poor way to conduct a simple poll. SMcCandlish 22:50, 10 September 2008 (UTC)
  21. ^ a b I'd like a combination of A and B really. I would prefer a three-tier system: "Use the style most appropriate if an article is strongly related (for example The White House for US, Tony Blair for UK). If, like kilogram, a country is not related, or if two countries boast an equal claim or a conflict will arise (Nicole Kidman was mentioned earlier), use the same format that ENGVAR implies. If ENGVAR hasn't been applied in the article, or isn't consistent, the status quo or first major contributor rule applies." –MDCollins (talk) 22:53, 10 September 2008 (UTC)
  22. ^ Default to international will cause a lot of problems in my opinion. –MDCollins (talk) 22:53, 10 September 2008 (UTC)
  23. ^ Can't see much difference to option B. –MDCollins (talk) 22:53, 10 September 2008 (UTC) [This is in reference to D.]
  24. ^ Option A is the simplest and is likely to produce the best reading prose in most situations. I think B is pretty much the same in most situations. - C. Parham
  25. ^ I see no essential difference between Options A & B; they just seem to be saying the same thing in different ways. - Askari Mark 01:48, 11 September 2008 (UTC)
  26. ^ I believe we need International as a default. To keep falling back on WP:ENGVAR for the date format is a fallacious argument, as has already been evoked - the two are not mutually exclusive and could go together. Firstly, let's be clear about our concepts: the 'first major contributor' is not a default, but a variable; secondly, the situation exists already, and the proposal would strengthen people's belief that they owned any given article just because they created it, and as such I am totally opposed to proposal. The only important thing is that all date formats in the text of any article should be consistent throughout, the next is how we aboid edit wars over which takes precedence. The 'first major contributor' option is a recipe for further conflict. Ohconfucius (talk) 02:08, 11 September 2008 (UTC)
  27. ^ Per Greg. I agree completely with his position. The article topic should guide to the date format used. Hiding T 09:58, 11 September 2008 (UTC)
  28. ^ Per Greg. I agree completely with his position. The article topic should guide to the date format used. Hiding T 09:58, 11 September 2008 (UTC)
  29. ^ C is my favoured option — OwenBlacker, 22:07, 11 September 2008 (UTC)
Comments on "Poll comments"

This section is intended to provide a forum for rebutting and commenting on the above ref-based vote statements. Greg L (talk) 22:16, 9 September 2008 (UTC)

  • Question: How do options B and D differ from each other? I largely copied & pasted that stuff. It seems these can be consolidated into a single option. Greg L (talk) 22:29, 9 September 2008 (UTC)
I've been struggling to work out the exact differences between all of the options all day - it all seems very subtle.–MDCollins (talk) 22:35, 9 September 2008 (UTC)
  • It appears the only difference between B and D relates to non-English speaking countries. Not a big difference. Greg L (talk) 00:27, 10 September 2008 (UTC
I don't think there needs to be any difference at all. English-speaking or not, there's only so many ways you can arrange day, month and year. Six. Only three of the six are in widespread use, and we only support two of these in written English. --Pete (talk) 02:53, 10 September 2008 (UTC)
  • Clarify point C: I believe (from Calendar date that Micronesia, Palau and the Philippines also use "U.S. style dates". Canada's position should probably be clarified again for any who come to vote without reading the mass of discussion above.–MDCollins (talk) 22:39, 9 September 2008 (UTC)
  • Thank you MDCollins. I revised option C per your observation. Greg L (talk) 01:17, 10 September 2008 (UTC)
  • I can support fixed text dates for the expected reader's benefit (C) so long as there is the (future) possibility that ALL dates will be displayed according to user date preferences if they are defined. Truthanado (talk) 01:29, 10 September 2008 (UTC)
  • Truthanado, in the future, when/if U.S. readers are accustomed to international date formatting and that is a thoroughly natural and fluid way of reading dates for that readership, the Wikipedia should go with the flow. It’s all about making text flow as naturally as possible for the likely audience. For that matter, we can loose “feet” and “pounds” in all our articles after the U.S. converts over to the SI. Again, go with the flow. Greg L (talk) 01:37, 10 September 2008 (UTC)
    • I think you'd be hard pressed to find an American reader confused by the "European" date format. (So long as we're talking long dates and not short dates like mm/dd/yy!) That's why I'd like an option for "always use dd mmmm yyyy" or "always use the first contributor's choice". All four of these options require a different format for U.S. topics. Powers T 02:02, 10 September 2008 (UTC)
  • The "common ground" point about Canada is not there. Now, I guess it applies to A, B and D. Perhaps a note above the table? Tony (talk) 01:44, 10 September 2008 (UTC)
  • Sapphic. Yes, your vote statement (“Autoformatting makes this entire poll irrelevant”) is true—but only for some editors (0.1% of Wikipedia’s readership). Autoformatting doesn’t work for 99.9% of Wikipedia’s readership (I.P. users and even registered editors if they haven’t set their user preferences). In those cases, writing [[February 2]], [[2008]] in an article on a European subject results in February 2, 2008 for virtually everyone who reads the article. Since the article is on a European subject, it ought to use European date formatting. Also, this abomination: [[2008-02-02]] produces 2008-02-02 for 99.9% of our readership instead of the pretty-looking, spelled-out dates you’ve been thinking it’s been producing. Greg L (talk) 01:50, 10 September 2008 (UTC)
  • Tony: I don’t understand your point. Please explain your point in detail and what you would propose. What is it about Canada that isn’t resolved? Greg L (talk) 01:50, 10 September 2008 (UTC)
  • Tony: Pondering your 01:44, 10 September 2008 post some more, I would think the best way to handle Canada is to discuss here, what the unresolved issues are with it and try to identify what the common ground is, and then to tweak each option individually as is required to properly address Canada. Making guideline wording internally logical and consistent isn’t easy and I’m not at all confident that it will be practical to make a fits-all, global statement (a “note above the table”) about Canada that will properly drop into each rule. Greg L (talk) 01:57, 10 September 2008 (UTC)
  • Powers: The actual implementation of (A) or (B) would probably resemble the present wording, using generally, so variations in usage and in editorial preference could change from the predominant national format. Septentrionalis PMAnderson 14:38, 10 September 2008 (UTC)
  • Anomie⚔: did you read option C beyond its title??? You wrote that it “invites insane levels of speculation over who this ‘likely readership’ really is—most cases where the ‘likely readership’ can really be determined can probably be more easily settled by applying the ‘strong national ties’ rule.” In actuality, it is a “strong national ties” guideline. The guideline boils down to “If it’s an article related to the U.S., use U.S.-style dates, otherwise use international-style dates.” Greg L (talk) 17:44, 10 September 2008 (UTC)
  • What about tables where RFC 3339 (ISO 8601) dates make the info comparable, sortable, doesn't cause line breaks, nor weird middle alignment in other cells? How about RFC 3339 dates in references? What if the first editor of an article was a scientist who used the RFC 3339 format? Why unlink and force people to read middle-endian dates when they've already set their preferences for small-endian of RFC 3339? -- Jeandré, 2008-09-10t19:10z
    • RFC 3339 contains these provisions:
      • "There are many ways in which date and time values might appear in Internet protocols: this document focuses on just one common usage, viz. timestamps for Internet protocol events." Since our articles rarely describe Internet protocol events, we should never describe whatever we adopt as RFC 3339, even if we end up adopting something that looks similar (which I don't advocate).
      • "All times expressed have a stated relationship (offset) to Coordinated Universal Time (UTC)." We often give dates without knowing or caring about the time zone; this is another reason this RFC does not suit us. --Gerry Ashton (talk) 19:35, 10 September 2008 (UTC)
    • ISO 8601 requires the Gregorian calendar, and agreement from the reader before using any year before 1583. We have lots of old dates, so this does not suit us very well. --Gerry Ashton (talk) 19:35, 10 September 2008 (UTC)
      • ISO 8601 says it uses the Gregorian calendar, but this is the same as anything else we use. I don't understand the problem with old dates that you're proposing. If it's necessary to express a Julian calendar date, it's still possible to use the conventions of YMD order, hyphen separator, etc.; this is no worse than any other date format. --JWB (talk) 20:34, 10 September 2008 (UTC) 2008-09-10T20:34Z (UTC)
        • You certainly can use the "YYYY-MM-DD" format in a way that violates ISO 8601, if you are prepared to invent and gain consensus for a new standard, and then communicate that standard to editors and readers. Just be sure you don't call it ISO 8601. --Gerry Ashton (talk) 20:44, 10 September 2008 (UTC)
Comments on poll itself

Polls rarely resolve anything here. This one was unnecessarily geeky; plain-text polls work just fine (to the extent that polls work at all). It should not have been phrased in terms of "votes". Adding a "the vote so far" line in the middle of a poll is biasing and divisive. A and B are the same thing, and obvious other options are missing, so the poll's data set is simply invalid. There is no snowball in favor of any particular option, so all this will result in is a finding of "no consensus". This should have simply continued as a normal consensus-building discussion. — SMcCandlish [talk] [cont] ‹(-¿-)› 22:51, 10 September 2008 (UTC)

Polls are extremely useful for provoking debate and are invaluable for finding out what editors’ opinions on a subject are. Much thought is exchanged in these forums. As for a finding of no consensus, as the lead-in intro speaks of, if there is no clear consensus (and one doesn’t seem to be developing here), we can hold a run-off poll with the two highest scoring options and see if a general consensus develops. If there is no status quo, we can just leave MOSNUM guidelines as they currently are, or editwar on MOSNUM. Greg L (talk) 23:06, 10 September 2008 (UTC)
Right... I guess that's why m:Polling is evil was written, along with WP:NOT#DEMO and WP:POLL, all of which have way, way more buy-in than WP:Voting is not evil (if you don't believe me, check the "What links here" pages; the latter has under 50 links, while the page that WP:POLL resolves to has over 2000, and that's not counting all the references to the Meta page). You can assert all you like how "extremely useful" polls supposedly are here, but a huge percentage of editors disagree with you. A run-off poll? You have to be kidding. I don't know if you were around for the WP:ATT debacle, but it was a debacle pretty much precisely because it turned into a divisive poll, and the hundreds of involved editors factionalized, leading to a breakdown in normal consensus-building processes. I can already see that happening here on a smaller scale. I'm not sure what you mean by "if there is no status quo"; "status quo" means "the current situation", and of course there is one. I'm not sure why you want to polarize this so much (into a poll to begin with, into a two-option runoff if you don't get what you want, into a false dichotomy of the binary choices "do nothing" and "editwar", etc.) This is not Middle-earth, and this is not a war between the Forces of Light and Darkness. — SMcCandlish [talk] [cont] ‹(-¿-)› 23:20, 10 September 2008 (UTC)

For a topic that supposedly found consensus this is sure a hot potato. I think that it is obvious at this point that regardless of how many Suport and oppose votes it got this topic does not meet consensus.--Kumioko (talk) 22:27, 10 September 2008 (UTC)

Don't confuse the idea of deprecating date autoformatting because of all of its problems, which has been building consensus for years, with a squabble over how to fix a problem of the aftermath of that decision. They are different conversations. — SMcCandlish [talk] [cont] ‹(-¿-)› 23:20, 10 September 2008 (UTC)
Kumioko: Not yet anyway. We’ll see if one develops with a run-off vote. More contentious issues (like the IEC prefixes) were sorted through with polls. This date issue isn’t nearly as contentious. Greg L (talk) 23:19, 10 September 2008 (UTC)
There is no point in a "run off vote". Wikipedia operates on consensus (WP:CONSENSUS is a policy, not a guideline), not voting (WP:NOT, including WP:NOT#DEM is a policy, not a guideline). Even WP:POLL makes it clear that polls should only be used to help build consensus, not as an actual voting mechanism. The only outright votes on WP are WP:RFA and related (bureaucrats, board members, ArbCom, etc.), since they are elections of human "officials". — SMcCandlish [talk] [cont] ‹(-¿-)› 23:35, 10 September 2008 (UTC)
  • SMcCandlish: Your position with your 4-0-0-0 vote is clear. You participated in a poll, and then you declared that you think all polls to be evil. You can’t have it both ways. Clearly, a lot of editors want to participate in this because very many have voted on this issue and also engaged in well behaved debate and discussion. You are out of step with the rest of us. So stop vandalizing this page by deleting polls and being all-around disruptive. We’re trying to see if we can develop a consensus here and your obstructionist moves get in the way of that objective. Greg L (talk) 00:17, 11 September 2008 (UTC)
I've addressed all of this at WP:ANI#Disruption of Wikipedia talk:Manual of Style (dates and numbers), since it isn't really on-topic here. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:16, 11 September 2008 (UTC)

I think you are being a little previous with holding a run-off poll. I advised as many interested editors as I could find about the poll and only a few hours later you're unilaterally moving on. Give the sun at least a day to travel around the globe, please! Our editors are scattered around the world. In fact, while I think a run-off poll is a good idea, I'd like to see the original poll run a week from the start date, and then after discussion we come up with some options which summarise the two "winners". The wording of the four original options has changed over the course of the poll, too, and I'd like to see some of the fat trimmed off before going further. We have consensus on a lot of stuff, such as defaulting to first major contributor, and that Canada uses both formats. I disagree with some of the wording changes as well - the first option reads a little too much like a campaign ad for my taste, and option three seems to have become very bloated and confusing. If we do this thing fairly and openly, then the process makes any post-poll criticism difficult. And why not include the longstanding wording as an option? I'm very dubious that there is any real reason for change.

Who wrote that? Please sign your posts. :-)
I'll also note that the changes to option three now makes options three and four almost identical in effect, as all nations but a handful use International format.
But having said that, It's good to have a poll produce an easily-accessible set of results. If nothing else, it is plain that no consensus yet exists, despite a few strident contributors repeatedly claiming that their side had "a clear consensus". But I think that we are getting there. --Pete (talk) 01:29, 11 September 2008 (UTC)
I agree: let the poll run for a while. As a matter of fact, the poll period should have been stated at the start. How about adding the close date to the poll description? I suggest a week from the start. Teemu Leisti (talk) 02:11, 11 September 2008 (UTC)
Concur with all of the above. NB: The edits that have closed the poll will need to be undone, including the removal of the blank entries and the summary of the results (which is bogus anyway; since options A and B are actually the same thing, with B simply being a strangely-worded version of A, the numbers are badly skewed in favor of C; it's simply e-gerrymandering). I would also suggest removal of the pre-emptive second poll, or at least a big note that it is not presently open, or editors will be confused as to which poll to participate in and/or will have their participation in the first one colored or discouraged by the claim in the second one that the first has already produced only a pair of choices. I'm not sure what anyone really expect to get out of these polls with regard to useful data, since the artificial forking of A/B has hopelessly muddied the water. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:16, 11 September 2008 (UTC)

Why do we even need this poll? I've been editing here for quite a while, and I never noticed a problem. Yes, it would be nice if all the dates in an article are consistently formatted (and I think they usually are), but the level of detail this poll goes into is crazy. First, as much as I try to do the right thing around here, I'm probably going to write any date in the "American style" (January 1, 2009) out of habit, having used it consistently for -- let's just say over 30 years. If I notice other nearby dates are in "reversed order" (international style), then I will probably use that style, too, without much thought. I'd guess that a lot of Americans (a large percentage of the editors here) are like me. I wouldn't say we are lazy, but it's such a non-issue that we don't pay any attention to it. Likewise, if I read an article on U.S. National Parks and all the dates were in international format, I probably wouldn't even notice. The only time I have an issue is when the month and day are ambiguous (8/1/09 is August 1, 2009 to me, but could also be 8 January 2009). As long as the month is spelled out, I really couldn't care less as to the order of the two. Any guideline that requires me to go back and figure out the style that the first primary contribution used is just plain dumb. I'd wouldn't bother doing it, so I'd probably break it and let someone else (who cares) figure it out. I'd rather see a policy that ALL dates be in international format than have to refer back, potentially through hundreds or thousands of edits, to figure out which way I should write a date. Therefore I reject every proposal here as being overly specific and unworkable in the reality that is en.wiki. --Willscrlt (Talk) 10:01, 11 September 2008 (UTC)

Run-off poll on guideline for writing fixed-text dates

Let’s see if we can develop a general consensus now. Which guideline do you support for determining the most suitable fixed-text date format (2 February 2009 or February 2, 2008) editors should use in Wikipedia’s articles?

The average scores in the above four-option poll, as of the 0-0-3-3 vote by ChrisDHDR, are as follows:

Option A = 1.76
Option B = 1.50

These are invalid statistics, since A and B are actually the same thing, and even averaging them will produce incorrect numbers, since some respondents rated one higher than the other, not realizing they amounted to the same concept in different wording. Discarding the lower values of the two, the real stat is inserted below. — SMcCandlish [talk] [cont] ‹(-¿-)› 03:41, 12 September 2008 (UTC)
Option A/B = 2.01

Option C = 2.24
Option D = 1.22


  • SMcCandlish, You’re arguments make no sense and took a lot of gall. Note how many editors are voting differently on A and B. The two options are clearly different, otherwise so many editors wouldn’t be voting differently on them. Isn’t that sort of a “Well… Duuhhh”  fact? Either your mental faculties operate on a higher plane that the rest of us can not even approach—or even fathom (“we’re not worthy”)—or your extreme bias on this issue (you voted 4-0-0-0 on the above four-option vote) blinds you to the reality of the situation and you are unable to get the WP:POINT. Whereas I can’t rule out the former possibility, it is more likely that latter. Please allow the rest of us to participate in this poll in peace, as well as its accompanying vote comments and discussion and debate; we want to see if we can arrive at a consensus and your interventions are not helping. You’ve already declared that “all votes are evil”. And the outcome? Editors keep voting in both this and the above poll. Your response to that? You then you came up to this introductory preamble, donned orange robes, poured gasoline on yourself, and set yourself alight to now make the point what we’ve all participated in is illogical and therefore invalid. You’ve made your point. Please stop being disruptive.

    And, by the way: if A and B are truly “the same thing,” whey then, didn’t you vote 4-4-0-0 instead of 4-0-0-0? If you’re going to have a message point, please try to be consistent. I see you didn’t even notice the consequences of the logical shortcomings of this new tact of yours and go up and revise your vote in the four-option poll. I’ve seen this sort of mentality before and you will not be permitted to hijack WT:MOSNUM and Wikipedia by exploiting every tool available (like this ANI) to halt progress towards something just because you strenuously object to it.

    Finally, note that I did not revert your disruptive edits here. This wasn’t the proper venue to make your point, but I’m not going to revert you and get into an edit war with you as to where you may attempt to make your points. But from hereon, I’ll see you down below in the discussion area; let’s not clutter this introductory preamble any more, shall we? I’ll allow you several more hours of ranting and then let’s solicit suggestions on how best to try to make progress towards a consensus. Greg L (talk) 18:21, 12 September 2008 (UTC)

    P.S. I see that Septentrionalis insists on having special rules for himself and wants to vote with a “-1” vote. I twice set his vote to a “0”, thinking his vote set a bad example for others and encouraged extreme voting in hopes of gaming the system to unduly influence the outcome. To all: Septentrionalis is clearly not above the rules that everyone else here is abiding by; in the final tally, his vote will be counted as a zero. I encourage everyone, to not spoil the mutual assumption of good faith here and to vote per your conscience—but within the rules—in order that we might work in good faith towards a consensus.


(A) Engvar (date format follows dialect of English most suitable for the article, default to first major contributor—no list of countries): Date format follows language variety; month-day format is considered a feature of U.S. English, just as the spelling "color" is. Where both ENGVAR is unclear and there is no strong tie to an English-speaking country, the "first major contributor" rule applies.

(C) Default to international unless U.S. and its territories—listed countries for editors’ convenience: For articles on, or strongly associated with, the U.S. or its territories (or countries listed in this guideline that use U.S.-style dates: Micronesia and Palau), editors should use the U.S.-style date format (“February 2, 2008”), otherwise, editors should use the international date format (“2 February 2008”) in articles. The dialect of English used for the article (“colour” v.s. “color”, etc.) shall have no bearing on the chosen date format. New articles on or strongly associated with Canada should use the international format but, for existing articles related to Canada, whichever format was used by the first major contributor shall be retained.

(e) None of the above: In which case please indicate what wording you would like.


Note that the below chart does not allow ~~~~ signatures to be used in <ref> notes. You must copy/paste or hand-edit your signature. For assistance in writing your signature, you may copy the time below in red from the preview window after clicking on “Show preview” while in edit mode:

04:04, 28 December 2014 (UTC)


0 = Complete opposition
1 = Could be much better
2 = Ambivalence
3 = Could be improved, but I support this
4 = Complete support
DEGREE OF SUPPORT FOR OPTION 
Editor   A     C     e  
Greg L 0[1] 4[2] 2[3]
Jeandré du Toit 0 1 4
Orderinchaos 0 4[4]
Woodstone 0 4 0
Willscrlt 0 1 4[5]
Lankiveil 2 4 2
Hiding 0 4[6] 3[7]
AliceJMarkham 1 1 4[8]
Teemu Leisti 0 4[9] 0
Septentrionalis 3.5 -1[10] 4[11]
Powers 0 1 4[12]
Professor marginalia 4[13] 2 0
Kotniski 4[14] 2 0
OwenBlacker 3 3 [15] 0
JavierMC 4 0[16] 0
Danorton 0 2 4[17]
SMcCandlish 4[18] 0 0
Ohconfucius 0[19] 4
JRG 0 4[20]
Tony1 05:04, 13 September 2008 (UTC) 2 3
Pete 0 4
Blank
Blank
Run-off poll comments
  1. ^ My opposition to option A (Engvar, or “English variation”), is that I am an American and use American English in all my contributions. I also routinely use international-formated dates in articles that have no strong ties to the U.S., such as Kilogram. Note that the French are more closely tied to the development of the kilogram than any other country and the kilogram clearly has no strong national tie to the U.S. So it makes no sense in my book to require that dates in the Kilogram article follow the U.S. convention (“February 2, 1799”) just because the article was written using American English. Greg L (talk) 23:33, 10 September 2008 (UTC)
  2. ^ I don’t see the need for using international date formats in an article on a U.S.-related subject (like a U.S. national forest), just because a UK author was the first major contributor and used the spelling “colour” in the article. IMO, it should be one, simple rule: There’s no legitimate need for any more complexity than that. If we are to start keying the date formats used in articles to the dialect of English the editor happened to use while writing the article, all the articles I’ve been writing (science-related topics, such as Kilogram) would have U.S.-style dates because I used U.S.-style spelling in those articles. That makes no sense. Keep it simple. Dragging English-language dialect complexities into this discussion seems totally unnecessary.

    The whole point of making it country-based is to make the article read as smoothly as possible for the likely audience. If it is an article on a U.S. National Park (likely a readership dominated by U.S. citizens), then use U.S.-style date formats to make the article as natural for that audience. For everything else, use the international date format. This should all be about making Wikipedia better for our readers; we needn’t be so concerned about hurting the feelings of editors. Greg L (talk) 23:33, 10 September 2008 (UTC)

  3. ^ To all: Please note the title wording of option “e”, it is for something else you propose. If you don’t want to suggest something else, keep your vote blank (or maybe a zero). But if you have a 1–4 vote, accompany your vote with a reference-based comment detailing your proposal. I didn’t put this column in this table and don’t quite no what to do with it. But I assume that some editors here have been treating it like it means “do nothing”. In my specific case (my “2” vote), that is what I am suggesting as my alternative for e: do nothing. Greg L (talk) 19:47, 12 September 2008 (UTC)
  4. ^ Please read in conjunction with Greg L's, as I think he has expressed my view pretty clearly. As an Australian I have no problem in editing US, Canadian etc articles with American date format as they are more immediately comprehensible to the intended audience. Orderinchaos 10:00, 11 September 2008 (UTC)
  5. ^ This whole thing is too labor and research intensive. It hasn't been a noticeable problem and shouldn't be such a big deal. If I had to choose, I'd say either make ALL dates be in international format or allow non-scientific articles about the U.S. and its territories and people be in U.S.-style format. Articles about people probably should be in the person's own preferred format (M D, Y in my case) regardless of their country (I could move to France and still M D, Y would probably be my preference for many years). The only thing that's really important is that we do not allow ambiguous numerical dates (e.g. 8/1/08 could be August 1 or 8 January). As long as we don't allow those, the text format concerns are just silly. --Willscrlt (Talk) 10:13, 11 September 2008 (UTC)
  6. ^ If you get rid of auto-formatting, this is my choice. Hiding (T 10:54, 11 September 2008 (UTC)
  7. ^ I am not sold on getting rid of auto-formatting. I appreciate the reasons why, but surely the devs must be able to implement the feature another way? Hiding (T 10:54, 11 September 2008 (UTC)
  8. ^ As suggested below, simply adopting the international standard as the universal format for enwiki would be my preference given the options suggested to date. --AliceJMarkham (talk) 11:29, 11 September 2008 (UTC)
  9. ^ Like I commented earlier, I think that running a runoff is premature, but I'll vote for my preference anyway. Teemu Leisti 14:05, 11 September 2008 (UTC)
  10. ^ The wording is unclear; therefore it requires improvement in order to rank as 0
  11. ^ Silence; which would restore our normal default practices, which should amount roughly to B, without the confusion between "established" format and "first user"
  12. ^ I prefer either: "All dates should be formatted as DD MMMM YYYY, except on specific articles where the consensus of editors dictates a different format," or "All date formats should follow the 'first significant contributor's' date format choice, except where a consensus exists to override that choice." -- Powers 15:37, 11 September 2008 (UTC)
  13. ^ Consistent with MOS ENVAR, time to settle this and move on, not endlessly explore more options Professor marginalia
  14. ^ Concur with the Professor's sentiments. Have no personal objection to C, but foresee turbulence trying to implement it, as it clearly favours one of what have up to now been considered two equally acceptable alternatives.
  15. ^ Whilst I agree with Professor marginalia and Kotniski, I would prefer Wikipedia to have more of the more-widely-used international format. I could quite happy live with the engvar equivalent, though. — OwenBlacker 22:14, 11 September 2008 (UTC)
  16. ^ Telling American editors that they can use M-D-Y format only when it pertains specifically to their country seems overly biased to me.
  17. ^ Use the worldwide convention of day followed by month (e.g. “2 February”) unless the article topic has overriding cultural ties to the alternate convention of month followed by day (e.g. “February 2”).Danorton (talk) 03:03, 12 September 2008 (UTC)
  18. ^ This poll's data is hopeless invalid for reasons I've already explained elsewhere. Furthermore, option A is misstated: If there is anything unclear about ENGVAR, then ENGVAR should be clarified. The idea put forth in the poll that ENGVAR is necessarily unclear frequently enough that its alleged inclarity has to be mentioned in the poll itself is an incredibly blantant (even if unintentional) bias against A inherent in the poll. C is not viable, because it conflicts with ENGVAR (an established guideline that there is no consensus to undo), and it will cause incredible amounts of strife and ill-will that ENGVAR has kept at bay for a long time. If ENGVAR needs modification, bring that up at WP:MOS (ENGVAR is a section of MOS), and WP:MOSNUM can be updated to do with dates whatever revised-ENGVAR demands. Noo ption E is needed, since revising ENGVAR and MOSNUM is necessary anyway, no matter what the consensus results of this debate are, while a major change (like banning US-style dates in non-US articles, one of the "E" options suggested) is something that is going to have to gain widespread consensus through normal consensus-building processes, not "voting". SMcCandlish 03:28, 12 September 2008 (UTC)
  19. ^ There is no formal or linguistic tie of English variant spelling with date formats Ohconfucius 13 September 2008 (UTC)
  20. ^ We need to make dates that are important to be notable in the article. I suggest we have the option to add bold text to some of our dates to show them up on the page like linking does at the moment. That way they show up but don't link to irrelevant pages. JRG 13 September 2008 (UTC)
Run-off poll Discussion
  • To sort out some confusion that seems to have crept in, this poll is not about whether you support autoformatting or its removal. Whether or not a date is linked for autoformatting makes no difference to its format (except for a minuscule percentage of users). All dates - linked or otherwise - have a format in which they are displayed to the world. What we are trying to decide now is what that format should be. The "removal of autoformatting" has concentrated minds on this issue, but the need for clear guidance on this matter is independent of that decision.--Kotniski (talk) 15:23, 11 September 2008 (UTC)
  • Second poll serves no purpose, as the first one's results are invalid because the wording of the options was changed again and again while the poll was running, and it was then pre-emptively closed when still active, not to mention that options A and B are actually the same thing, making the numbers skewed, as addressed above. Furthermore, the divisive and polarizing false dichotomy presented by the second poll is not helpful to consensus-building, as option C is really nothing but a slight modification of A/B to be more specific (that's a simple editorial discussion about the wording, not actually a completely different option with regard to what to do about date formatting!). — SMcCandlish [talk] [cont] ‹(-¿-)› 04:03, 11 September 2008 (UTC)
I agree that the run-off poll may have been a bit premature, but your claim that all the options are really the same thing is not true at all (and doesn't seem consistent with your vote of 4 for one of them and 0 for the others). The difference between A, B and C is quite clear. A links date format to spelling; B gives editorial freedom to choose date format independent of spelling; C forces day-month date format (Anglophone-country-related articles excepted in each case). If we are to have a run-off between A and C then we can do it in the form of a simple question: in non-U.S.(...)-related articles written in American English, should dates be day-month or month-day? --Kotniski (talk) 07:54, 11 September 2008 (UTC)
Please do not misstate what I said. I said that A and B are the same, and noted that C is simply a variant of them (by contrast with D, which is off in left field). Why would I redundantly "vote" for something else when I've already selected the option that makes sense to me? Option A does not link date formatting to "spelling"; it links date formatting to the parameters laid out in ENGVAR (strong national ties to a topic, if none then first major contributor, and that's precisely what B says in different wording). C does not do what you say it does. Please actually read it closely. It modifies the ENGVAR/A/B position by saying use US format when there are strong national US ties specifically, otherwise use international. As I said, it is a variation. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:01, 12 September 2008 (UTC)
Or do you mean: "Should they normally be day-month by default, or still subject to the selection of either format based on a rule?" Tony (talk) 08:13, 11 September 2008 (UTC)
Noo, I don't think so; option A would imply month-day in such cases.--Kotniski (talk) 14:41, 11 September 2008 (UTC)
I'm not going to vote on a "run-off poll" because I don't think the first poll has run it's course. In any case, it serves no purpose unless the options are reworded to encompass the views in the first poll. To encourage further thoughts, and allow the score of 3 (with improvement) necessitates improvements can/must be made before taking it further. That said, if there is a clear loser in the first poll, thoughts can be narrowed down.–MDCollins (talk) 09:02, 11 September 2008 (UTC)
  • I'd also say that the second poll serves no purpose, because the original poll - being based on range voting rather than preference voting - does not really require a run-off. Option C won with a very slight edge over option A. (Effectively, it could be called a tie.) [See below] If the same editors vote again in the same way - and there's no reason to vote differently, unless one has changed his or her mind - then the result of the second poll will be exactly the same as the first one; narrowing down the choices does not really change anything. GregorB (talk) 09:20, 11 September 2008 (UTC)
    • 2.2 to 1.7 is hardly a tie, but I do think polls are unproductive anyway. Orderinchaos 12:03, 11 September 2008 (UTC)
  • For WP:CONSENSUS purposes, it is absoletely a tie (i.e., "no consensus"), since consensus is a measure of communal agreement, not vote counting. The no-consensus sitatution is exactly what I predicted the result would be, despite the gross bias in favor of C inherent in the poll itself on at least two levels, and it uses the wrong voting format for a two-item poll anyway. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:01, 12 September 2008 (UTC)
Polls are a good indicator of community feeling and a springboard for consensus. One thing that is quite clear is that there is no consensus for any specific change to the longstanding wording, which makes Anderson's repeated changing of the wording without consensus and despite repeated warnings both here and on his talk page a matter of concern. We're working through discussion, with ongoing polls, so there's really no excuse for edit-warring to push a personal POV. --Pete (talk) 12:16, 11 September 2008 (UTC)
Orderinchaos, you're quite correct about the winning option - I somehow managed to misread the numbers. (I've corrected my comment.) Not having a tie actually makes the run-off poll even less useful... GregorB (talk) 12:25, 11 September 2008 (UTC)
It's actually much closer than it looks; see my stats corrections above. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:01, 12 September 2008 (UTC)
True. Very true. I've "voted" in it anyway just to try to balance out the inherent anti-A language in the poll's wording a tiny bit, lest a strongly-C but bogus result later be bandied about after a lot of people have forgotten that the results were bogus. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:01, 12 September 2008 (UTC)
  • I object to this poll. Polling and polling again and again until one happens to come up right is an elementary form of fraud, one of the first reasons m:Polling is evil. Septentrionalis PMAnderson 15:02, 11 September 2008 (UTC)
    • I wasn't going to comment on the run-off poll's regularity but still: one can't have a two-horse range voting, because it is easy to "tamper" with the results. If you want the option you prefer to win, you simply award it 4 points, and give 0 points to the other option, regardless of what you actually think about them; anything else would reduce your option's winning chances. Makes no sense, and should be considered void. GregorB (talk) 19:05, 11 September 2008 (UTC)
      • Range voting on a runoff is contrary to the term runoff. A runoff in which there are only two choices (like this one) should be by preference voting only. That is, voters vote for either one, not both. Forcing the voters to make a choice increases the probability that the result of the vote is a true representation of the will of the voters. Having said all that, this vote (or any vote for that matter) is meaningless. Wikipedia is governed by consensus which involves discussion, counter-discussion and obtaining an overall consensus of the will of the Wikipedia community as a whole. Truthanado (talk) 01:56, 12 September 2008 (UTC)
        • And these are not votes; see m:polling is evil. One of the evils is that claims of victory based on averaging (as above) are inherently a misunderstanding of the system. !Votes are not intended to be tactical, and there should be no scope for them to be so. These are merely abbreviated statements of our various positions; thus, for example, I do mean to evaluate C at -1. That's what I think of it; it's a vague statement of an abominable position, so it would need improvement to reach a 0. Septentrionalis PMAnderson 04:40, 12 September 2008 (UTC)
          • Good point. If the scale if incomplete, or merely non-equidistant, then averaging the votes makes no sense. The poll shows only two things: 1) there is no consensus (not even a rough one), 2) some solutions do not have enough support and, consequently, stand very little chance of being accepted. In short: back to the drawing board... GregorB (talk) 08:38, 12 September 2008 (UTC)
  • All: m:polling is evil is an essay; it is not a guideline, rule, or policy of Wikipedia. And the whole point of that essay is that polling—in and of itself—is not helpful if done in the absence of discussion and debate. In this case, where we have vote comments and then a well used debate section to discuss the vote comments, the system—as a whole—is a tool to find out where others stand on the issue, share thoughts and try to influence opinions towards an effort to arrive at a consensus view. No one said democracy is easy, that’s why it doesn’t work in all areas of the world—some cultures just aren’t ready for it. And certainly, on Wikipedia, where we have a collaborative writing environment, it is essential that we try to arrive at a consensus view rather than let extremists manipulate the system (the “-1” vote of Septentrionalis and the ANI by SMcCandlish) in a vain effort to get their way.

    I find it noteworthy that those who seem to find fault in this entire endeavor A) participate in it (in full), and B) tend to have highly polarized votes (lots of zeros or a “-1”). I applaud the rest of you here, who are voting with nuanced votes and discussing and debating in good faith. Greg L (talk) 19:15, 12 September 2008 (UTC)

Please stick to the rules (moved)
I see that Septentrionalis insists on having special rules for himself and wants to vote with a “-1” vote. I twice set his vote to a “0”, thinking his vote set a bad example for others and encouraged extreme voting in hopes of gaming the system to unduly influence the outcome. To all: Septentrionalis is clearly not above the rules that everyone else here is abiding by; in the final tally, his vote will be counted as a zero. I encourage everyone, to not spoil the mutual assumption of good faith here and to vote per your conscience—but within the rules—in order that we might work in good faith towards a consensus.

And please, no running average statistics. It makes a visible target to aim for and encourages extreme voting in an effort to game the outcome (like Septentrionalis’s effort). Just vote honestly and assume good faith in others. Don’t worry; cheating won’t be rewarded in the end. Greg L (talk) 18:53, 12 September 2008 (UTC)

  • I see that Greg L has added personal attacks to this unnecessary, ill-constructed and misleading poll. This is regrettable; especially since his personal attacks are false, and if he read his own talk page, he would know them to be false.
  • First of all, my !vote on A remains what it was before: 3.5: I am perfectly content with it, but a rephrased B would be more to my taste - so would silence. As for others: as Gregor commented some time ago: changing votes to 4/0 because one wishes maximum effect in choosing between the two option is the predictable result of incompetent poll design. (Several, however, hold 4/0 as their opinions of the two options.)
  • We agree at least that there should be no running averages. Indeed, there should be no averages at all; these are not votes - they are the best effort to express complex opinions in a comparable fashion. They are not numerical data; at best they are rank data, but !votes of 2 on the same option can mean very different things.
  • In particular, -1 is the best I can contrive to express my opinion in this shorthand: C is badly phrased, and needs fixing; but even if it were fixed, it would be atrocious guidance, so it is worse than 0. This is not an attempt to game an average; there should be no averages.
  • Those who invent rules out of their head and attempt to impose them on their fellow editors without consultation may expect to be ignored when the rules do not fit what their fellows wish to say. Septentrionalis PMAnderson 19:38, 12 September 2008 (UTC)
  • Septentrionalis: You’re not abiding by the rules that everyone else here is abiding by. Stop trying to justify your actions and stop trying to hide behind the apron stings of “personal attacks” when someone calls you out for cheating. The rules that this is a 0–4 vote is in the preamble and is imbedded as an editors’ note in each and every vote row; don’t say you didn’t notice. And stop saying you are somehow more *special* than the rest of us here and don’t perceive the need to vote within the constraints everyone else has seen fit to vote within. The proper remedy for someone who twice restored his vote to -1 after being told on his talk page why he shouldn’t do so is to have your entire vote disqualified. Rather, in the end, your “-1” vote will be treated like a zero. We can’t be encouraging others to do what you did. Greg L (talk) 19:50, 12 September 2008 (UTC)
    • Greg L is at least consistent. He sponsors continual polls, which Wikipedia discourages. He speaks of them as votes, which is the chief reason we discourage them. He invents a system which encourages quantitative treatment of opinions, rather than looking at what editors actually think, which is the second reason we discourage them. He imposes rules without consultation, and expects everyone else to abide by them; ignoring the clear case that they are inappropriate for the question being considered. He is having a fit about a purely procedural question; the procedure happens to be inappropriate for most polls, and doubly inappropriate for Wikipedia. He complains, personalizes, and polarizes the discussion when he gets opinions he doesn't like, which is the chief reason we don't have polls. (And he doesn't understand the purpose of a column which encourages third alternatives, the best way of getting out of no-consensus binds.)
    • It should be no surprise that he !voted 4/0 for a wording that would attempt to impose an inconvenient and obnoxious rule on all the editors who aren't here and don't know about this poll. Septentrionalis PMAnderson 20:29, 12 September 2008 (UTC)
  • You are perfectly welcome to start you own poll—one that has no rules and any editor can do whatever they want. But you haven’t done so and have chosen to vote here, where there are rules. If you try to game the system to get your way, your vote might well be fully tossed. You’ve made your feelings abundantly clear. Don’t be a dick in the process. Greg L (talk) 23:07, 12 September 2008 (UTC)
    • You know, that's the way most polls in Wikipedia work: they're places for discussion, without an owner or any rules but custom and civility. Your solitary insistence on "rules" you have invented interferes with discussion, if only by wasting time. If I thought we needed more polls, I would start one accordingly; but we need, if anything, about two less.
    But let me quote actual policy: Wikipedia is not a moot court, and rules are not the purpose of the community. Instruction creep should be avoided. A perceived procedural error made in posting anything, such as an idea or nomination, is not grounds for invalidating that post. Follow the spirit, not the letter, of any rules, policies and guidelines if you feel they conflict. If the rules prevent you from improving the encyclopedia, you should ignore them. Disagreements should be resolved through consensus-based discussion, rather than through tightly sticking to rules and procedures. Septentrionalis PMAnderson 01:29, 13 September 2008 (UTC)
    • Now, do you have anything to say on the substance of matter at issue? Septentrionalis PMAnderson 01:31, 13 September 2008 (UTC)
So one person can distort a whole poll process by "voting" –40, yes? It's gaming the system to use anything outside the given range. Tony (talk) 01:44, 13 September 2008 (UTC)
Only if someone takes the average in the first place. That's theoretically unsound, and contrary to Wikipedia practice. All my !vote means, or can mean, is that I would be perfectly content with silence, happy with A, although it could be tweaked, and find C worse than useless; the only advantage of numbers is the easy comparability with other opinions, which is in part specious. The chief lesson of statistics is learning what not to do with them. Septentrionalis PMAnderson 01:52, 13 September 2008 (UTC)

deindented: I'd say his attempt at gaming the system was at least transparent, and not totally blatant. By giving one option a 3.5 and then giving 4 to 'none of the above' is really what gets me. Clearly, if one option is acceptable enough to gain 3 12 marks, I don't see how one can seriously put hand on heart and say none are acceptable. If one person, quite contrary to the instructions given at the start) can put negative marks, I'm sure that others among us may want to re-evaluate the marks we gave too! Then it will all degenerate into the sort of playground brawl that is of no benefit to anybody. Ohconfucius (talk) 02:28, 13 September 2008 (UTC)

    • I didn't say none of the proposals are acceptable. Either A or B is; saying nothing (often the best guidance, in a Taoist spirit) would also be acceptable, and may be necessary. If there is no consensus, proclaiming anything as if it were consensus would be a falsehood. Septentrionalis PMAnderson 14:16, 13 September 2008 (UTC)
  • We’re trying to see if we can develop a general consensus. We can see past Pmanderson’s horseplay. And I for one, refuse to get pulled any further into his drama on this issue. It is childish. We’ve made our point here that we don’t want others to try the same stunt as he did so we can find out what people really think, pursuade others, and find common ground. And, fortunately, no one seems to have perceived the need to follow suite and game the system. The rest want to participate in good faith as we all (mostly all) had initially set out to do. Greg L (talk) 03:11, 13 September 2008 (UTC)

Non-vote-based resolution

Instead of a poll that does not actually encapsulate all of the reasonable views on this topic, and duplicates one of them as two differently-worded "options", I think it would be a lot more productive to go back to what the goals or values are, and work from there. I would propose the following:

  • Value: Wikipedia guidelines should not conflict. Ideally, they simply agree in a synergistic fashion, although when necessary simply make well-explained and scope-limited exceptions of the one for the other. We already have a guideline at WP:ENGVAR. It may not be perfect but it should not be ignored. Any solution arrived at here should agree with it as much as possible (if not entirely), since it already represents consensus by virtue of being a long-standing guideline. Ergo, WP:MOSNUM and WP:ENGVAR, regardless of the outcome, should be made to agree with each other.
  • Value: Reading and editing Wikipedia should be pleasant, not a chore, annoyance, fued or exercise in nitpicking.
  • Goal: The primary goal of having any date formatting recommendation at all is reader usability (including readability, parsability, lack of ambiguity, simplicity, etc.).
  • Goal: The secondary goal is article consistency (really a subset of reader usability, but distinct and important enough to be considered on its own).
  • Goal: The tertiary goal is editability – editors should not have to return to the same guideline again and again to figure out what to do because the situation has been made unnecessarily complex, nor be led into revertwars because of ambiguous or conflicting guidelines.
  • Goal: A fourth goal is ability and ease of automation by templates and bots. While a derivative of consistency, it is worth keeping in mind as a discrete issue.
  • Goal: A fifth goal may be non-interference with developer ability to come up with another means of date autoformatting that does not conflict with editorial needs. Such interference isn't likely anyway, but let's just say it up front.

Feel free to identify others, of course.

We can also identify some non-values and non-goals, such as advancing Commonwealth English over US English or vice-versa, irritating users by using US dates in British articles or otherwise PoV-pushing mostly-American date formatting against the preference of most of the rest of the world, or insisting on the use of geeky formats like 2008-09-10, which for many readers will be ambiguous and hard to parse, even if it is a "standard" in some other context.

SMcCandlish [talk] [cont] ‹(-¿-)› 03:57, 11 September 2008 (UTC)

Comments on non-vote-based resolution

  • My take is that poll options A and B are nothing but differently worded equivalents, since B is A in a nutshell, and nothing in the more specific details of A precludes B. C is a minor variant of this, with limitations. D is simply not relevant. The real question at hand is whether the A/B (strict ENGVAR) or C (modified ENGVAR) solution is the better one. I lean toward the former, since ENGVAR has been show to work well, and I don't see a particularly strong justification for the additional specificity of C. What I'll call "option E", someone's proposal to use nothing but the international format, isn't viable; it is will meet with a lot of resistance from most US editors (and some non-US ones), and would simply be ignored by enormous numbers of editor, probably the majority of them, in fact.

    While I'm in favor of A and its simplicity, some modification of it (various poll comments indicated that certain "voters" saw room for adjustment) could be in order. So, I would propose that ENGVAR (A/B) be taken as the base from which to work (since it is, after all, an extant guideline), and that those who have issues with it (C issues, or other ones) raise those concerns, explain them well, and let's see what we can agree on. I would further suggest that actual draft guideline wording to play with would be useful – it could quiet a lot of the debating and get us to collectively focus a little better instead of arguing so much about abstractions.
Here is a stab at language to work on:

The formatting of dates, as 23 February 2008 versus February 23, 2008, should be consistent throughout an article, and follow one of those two patterns. Which format to use is dependent upon the dialect of English used in the article, which should be that most suitable for the article's topic, or that used by the first major contributor to it if there are no strong national or dialectal ties to the subject. US and strongly US-tied topics automatically call for February 23, 2008 formatting, while UK (and usually British Commonwealth) subject matter necessitates the 23 February 2008 version. [other guidance after this, e.g. on ISO dates and malformed dates that mix and match both styles, e.g. comma in intl. style, no comma in US style]

Clean and simple, A/B and even aspects of C merged into plain English. I'm sure it could be improved of course, but I'm trying to get at the ideas, principles and goals, not arrive at perfect wording on the first pass. By including "dialectal ties" it accounts for situations like Palau that are not strongly US-tied (unlike, say Guam) but still use the US style. "Usually British Commonwealth" accounts for the fact that both formats are used by large numbers of Canadians.
That's just A reworded, isn't it? The difference between A and B (if no-one's explained it yet) is that B does not link date format with dialect (if the first major contributor used both colour and June 1, 1999, it would stay that way, whereas under A one of them - the date, persumably - would have to be changed).--Kotniski (talk) 07:40, 11 September 2008 (UTC)
So, what exactly is worth arguing about now? I have to strongly point out that arguing for always using international format, or always doing it except for US topics, or other more blanket measures is going to conflict with WP:ENGVAR very sharply. We have no reason to do that, and frankly I don't think we have the consensus authority to do that without also gaining consensus at ENGVAR to change it. Let's just keep it simple.
SMcCandlish [talk] [cont] ‹(-¿-)› 04:45, 11 September 2008 (UTC)
  • It appears to me that there is a discussion underway here that effectively proposes changes to ENGVAR, but until today no mention of this fact had been made on the talk page relating to ENGVAR. Not everybody follows the D&N sub-page, so any proposal to alter ENGVAR must involve people who are actively involved in that page, not just people here.
My preferred solution would be a feature request to have a country code available to allow pages to be customised to the user's location. This would use, in decreasing order of preference, a specified user preference for logged in users, then the browser preference (if set), then finally default to the country that the IP address is located in. Such a feature would allow for delivering content in the "correct" format for the reader's location, with a major re-write of both this section and ENGVAR. It seems likely to me that these arguments, which have been repeated ad nauseum already, will continue indefinitely until such a feature is implemented. --AliceJMarkham (talk) 06:31, 11 September 2008 (UTC)
Alice, we've taken the decision to move away from computerised solutions of any type, since they hamper the efficient management of date formats. WYKIWIG (What You Key in Is What You Get) is the obvious solution.
SMcCandlish, I don't care whether engvar or first major contributor is uppermost in deciding the format for articles that are unrelated to an anglophone country, just as long as we have a cascading chain of decision-making for cases where there's uncertainty) to forestall disputes. If there's to be a country-driven rule (D), a list of which format is used in which country would be necessary; I find that idea quite unsatisfactory, as though we should be using the local name for cities (Wien rather than Vienna). And if the Austrian date format is to count, perhaps we should use their word for the month: "15 Marz 2003. On Canada and South Africa, both of which use both formats, the ease with which the most recent query was solved by clicking on "Earliest" in the edit history indicates to me that this way of resolving engvar issues is good for date formats, too, for anglophone countries that use both. India also has a sizeable English-language population, some of it native-speaking; they use both formats. Tony (talk) 07:36, 11 September 2008 (UTC)
No offence, Tony, but that's just plain silly. Nobody has seriously suggested using non-English date names. It's something made up as a strawman to frighten people off. Nor do we really need a list: apart from a handful of nations, everyone uses International date format. A few use ISO year-month-date, but we're not going to use ISO in plain text. Now, I'll grant you that none of the other language Wikipedias try to tie date format to location, but neither do they tie it to variety of language. You know why? They all use International format. Day-month-year.[15][16]. Why don't we, as an international project, follow suit? --Pete (talk) 08:32, 11 September 2008 (UTC)
Because in written German (and other foreign language projects), there only is one format. In written English both formats are common, hence this discussion.–MDCollins (talk) 09:09, 11 September 2008 (UTC)
"No offence, Tony, but that's just plain silly." - I think Tony was employing sarcasm here, not advocating the use of foreign month names! Anyway, as I said in the vote comments above, I would favour something like "Use the style most appropriate if an article is strongly related to a country (for example The White House for US, Tony Blair for UK). If, like kilogram, a country is not related, or if two countries boast an equal claim or a conflict will arise (Nicole Kidman was mentioned earlier), use the same format that ENGVAR implies. If ENGVAR hasn't been applied in the article, or isn't consistent, the status quo or first major contributor rule applies."
If this suggestion actually is ENGVAR - please tell me. In any case, a guideline such as that makes it perfectly clear how to work out the required format, without needing in depth knowledge of ENGVAR - providing that it complies with it as mentioned above.
MDCollins (talk) 09:07, 11 September 2008 (UTC)
"Wien" rather than "Vienna" is perfectly logical if you're going to be influenced by text as Austrians write it. May as well use their words for the months. Tony (talk) 09:19, 11 September 2008 (UTC)
I haven't seen anybody seriously suggest this. Putting it forward as some sort of real possibility is quite irresponsible. However, on a more relevant matter, I've seen editors advocating use of International format throughout, and that's a serious suggestion which is becoming increasingly attractive to me. Nobody is likely to get confused if the date is 1 February 2003 or February 1, 2003. One date format for all wikipedias, no more petty arguments, a genuine commitment to internationalism. Every year at Wikimania we hear Jimbo and other wikilights talking about how Wikipedia is a global effort, and yet I see some editors here striving their hearts out to push American formats, American units of measurement and so on in a charming example of cybercolonism. --Pete (talk) 09:50, 11 September 2008 (UTC)
  • I'm using this example to illustrate—logically—why the country-driven proposal is unacceptable. Tony (talk) 10:59, 11 September 2008 (UTC)
Why is it unacceptable now, when it's what we've been doing for years? I've posted above a brief history of the development of the wording, and it's pretty much always been, "use the format of the location". Wikipedia hasn't fallen apart over this in the past, and I doubt it will do so in the future. I've also been looking at WP:ENGVAR, and you know what? It's never linked date format to variety of English. The reason why it is used is a reference is that EngVar is a example of linking usage to location and minimising conflict over personal preferences. This notion that date format is somehow tied to variety of English is a complete furphy. --Pete (talk) 12:02, 11 September 2008 (UTC)
If we're going to deliberately avoid using end-user preferences or system information to format dates, and considering that almost all english speaking countries use it, I'd agree with Pete's suggestion that we simply adopt the international format as being the best available solution. The above !vote doesn't offer that as an option, though. --AliceJMarkham (talk) 11:20, 11 September 2008 (UTC)
That would be my preference, too, but apparently that would be a step too far, as it would be opposed by many American editors, who make up about half the contributors to en.wikipedia.org. So, using day-month-year in all articles except those related to the US—i.e., choice C in the polls—looks like the the best compromise that is realistically attainable. Teemu Leisti (talk) 13:46, 11 September 2008 (UTC)
It is not a compromise; the mere suggestion demonstrates refusal to discuss in good faith. Septentrionalis PMAnderson 18:58, 11 September 2008 (UTC)
Participants in this discussion have different opinions. In order to reach a consensus, one side either has to be persuaded by the arguments of the other side; or, failing that, people have to negotiate, and give up a part of what they wish for in return for getting another part. This is known as compromising, and it's often done in real life, too. I don't see how you understand this as demonstrating "refusal to discuss in good faith".
If I really had my way, everyone would use the ISO 8601 notation, yyyy-mm-dd, but I realise I would get very little support for it, so I don't even propose it. We can't always get what we want, which is something to be kept in mind by everyone. Teemu Leisti (talk) 23:34, 11 September 2008 (UTC)
Ah, I see. This is a "compromise" with a Never-Never Land, in which not only do all of us here lose our senses and decree that only ISO is to be used for inputting dates, but the rest of Wikipedia tugs their forelocks and complies.
There are severe ethical objections to such a system, starting with the purpose of Wikipedia: we are the encyclopedia that anyone can edit, including those who dislike ISO and the much larger number who have never heard of it. But what gives you the idea that such an idea would work? Editors drift in, and add to the encyclopedia in English; they are not dragged through every subpage in MOS - and therefore they're not going to add ISO dates. If such an effort were started, they would begin by emending such dates into English, as now they begin by correcting "teh".
This is an exaggeration, like Tony's reductio ad absurdum of using German names, of the flaws of option C. Teemu's dreamland would require all editors to use a date format unidiomatic to almost all of them on all articles. Option C requires the same of only half our editors, on all articles not flying the Stars and Stripes; the mischief is the same, the scope wider. Septentrionalis PMAnderson 00:47, 12 September 2008 (UTC)
I'm not sure this subject deserves quite that much righteous indignation. Teemu Leisti (talk) 04:21, 12 September 2008 (UTC)
Righteousness? The only moral claim I have made is that anyone should be able to edit Wikipedia. If you disagree, talk to Jimbo. I discussed only the practicality of Teemu's utopianism; whether a small group of editors should propose to bully all Wikipedia into the format of their choice is another question entirely, which we cannot profitably discuss. It is plain that we have no common ground of principle. Septentrionalis PMAnderson 04:56, 12 September 2008 (UTC)
I suggest you step back for a bit and reread your own comments, pretending someone else were addressing them to you. I outlined to you a bit of my thoughts during this process, and you reacted by accusing me of refusing to discuss matters in good faith. I assure you that that is exactly what I am doing, and, it seems to me, what most of the other contributors here are doing.
No one is suggesting that the MoS is a straightjacket that all contributors must immediately follow—in case of date formatting, or in case of any of the other rules enumerated therein. It never has been, and never will be. It is a guideline for best practice, to be followed by editors interested in cleaning up articles, and to be referred to in case of disputes. The discussions and arguments undergone in pursuit of formulating the MoS will hopefully forestall many similar article-specific arguments in the future.
The subject of how to display dates is not a complicated one, and there are not that many choices available. I, and others here, have been attempting to lay out these choices, discuss them, agree on common ground, and most of us are now participating in a straw poll to find out the general lay of the land. I attempted to explain the reasoning that goes into attempts at compromise. Reacting to this by such high-handedness ("dreamland", "utopianism", "talk to Jimbo", etc.) does not really contribute that much to resolving the issues.
You wrote: "But what gives you the idea that such an idea would work? Editors drift in, and add to the encyclopedia in English; they are not dragged through every subpage in MOS - and therefore they're not going to add ISO dates." Nothing gives me the idea that such an idea would work. That's why I have not suggested it, as I noted in the paragraph where I explained my reasoning. I was merely attempting to make the point that in order to make progress, we all have to make compromises, sometimes even self-censoring those of our preferences we know are too advanced for the masses to follow immediately. (That was a joke, I hasten to point out.)
I know the euphoria one can get from righteous indignation on the Internet, crushing one's deluded opponents with high-flying rhetoric. Then I grew up a bit, and nowadays attempt to speak to people as if I were face-to-face with them. I try to moderate my rhetoric. (I don't always succeed.) It makes for fewer hurt feelings; in case of Wikipedia, faster progress; and I don't get that empty feeling afterwards. Rather than spending one's flaming energies on arcane Wikipedia formatting rule debates, it's preferrable to save it for things that really matter.
That said, despite the triviality of the debate's subject, this has been a fascinating process to go through, and I'm interested in finding out how it all turns out, if it ever does. Teemu Leisti (talk) 06:24, 12 September 2008 (UTC)
Which American editors are opposed to it? Are you sure there's "many"? Powers T 21:16, 11 September 2008 (UTC)
OK, I don't know exactly how many there are, but I'm sure that were international notation to be imposed on all articles, the protests would be loud, clear, and numerous. Which is why I'm not even insisting on my second-preferred choice, dd MMMM yyyy being imposed on all articles. I'm prepared to compromise, but at least one editor sees this as "refusal to discuss in good faith". Go figure. Teemu Leisti (talk) 23:34, 11 September 2008 (UTC)
It is not what Wikipedia has been doing; it's what Pete's been doing all by himself. Like much of MOS's subpages, the text on which he relies is a relic of some obscure discussion, which survived until somebody else actually noticed it. Septentrionalis PMAnderson 19:02, 11 September 2008 (UTC)
A couple of comments: 1) This is not a proposal to modify ENGVAR (though I have notified ENGVAR/MOS about this discussion, as ENGVAR is a part of it). It is a proposal to a) lay out wording for MOSNUM that does not conflict with ENGVAR, or b) lay out wording that does conflict with ENGVAR if there is consensus to do so, and then to take that proposal to MOS for alteration of ENGVAR, so the guidelines agree. Obvious, I would prefer option a over option b. 2) The idea that date formatting is not tied to variety of English is absurd. 3) Yes, using the format expected by people from an anglophone country (US style in US, "international" style in UK/AU/NZ, etc.) when writing about them and things related to their country, because... 4) Yes, date formatting is obviously implied by ENGVAR, since how we write dates is just as much a matter of our variety of the language as is the spelling of colo[u]r and whether that car part is a trunk or boot (or something else - I don't know what South Africans call it). — SMcCandlish [talk] [cont] ‹(-¿-)› 04:17, 12 September 2008 (UTC)
That would seem to offer a straightforward and restrained solution to the problem. Septentrionalis PMAnderson 04:56, 12 September 2008 (UTC)

Changing date formats

Hello. A recent argument has come up on the Autumn Phillips article. I created the article and since then it has always used the date-month-year date format, which according to the MoS is perfectly acceptable for Canadian-related articles. My understanding has always been that date formats should not be changed unless for good nationalistic regions (i.e. on an article about a British person, it would be acceptable to change to the UK format from the US format). I have understand that changing date formats otherwise is disruptive and unnecessary. A user, G2bambino, has insisted on changing to the month-date-year format for Phillips' article saying that most Canada articles use that. I have explained that the MoS says either can be used and advises against changing date formats unless necessary. But he does not believe and continues to change it. Please can someone shed light on what the MoS says - should date formats be changed if the current one is perfectly acceptable for that article and has always been used.--UpDown (talk) 11:30, 9 September 2008 (UTC)

Both formats are acceptable for Canada and therefore changing date formats is disruptive. See the comments at the top of WP:DATE. This applies regardless of which way the changes are made. --Pete (talk) 11:35, 9 September 2008 (UTC)
This is exactly what I thought, and thank you for the clarification.--UpDown (talk) 11:37, 9 September 2008 (UTC)
This would have been an issue with or without the recent debates on (1) the removal of DA, and (2) the selection of date format for articles related to non-anglophone countries, since the Canadian situation is unchanged. I do think it needs to be spelt out in the MOSNUM text, though: clearly, the existing format wins out, and if your antagonist wants a duel to the death, go back in the edit histories to determine the first choice, as for our highly sucessful WP:ENGVAR procedure. I'll go to that article now. Tony (talk) 11:40, 9 September 2008 (UTC)
The first date format used for the article was day-month-year (I created the article), and this has been used ever since and I don't believe anyone has ever tried to change it before.--UpDown (talk) 11:43, 9 September 2008 (UTC)
Yep, 10 seconds at the edit-history page confirmed that, and I've left a note to this effect at the talk page. Easy-peasy. Tony (talk) 11:44, 9 September 2008 (UTC)
And is that the new guideline, then? The first date format used stays indefinitely? If that's to be the case, and it applies across the board, then fine. I'll adhere to that and keep it in mind for future. But, I want to be clear that that is indeed the way things will be set from now on across all Canadian articles. --G2bambino (talk) 15:09, 9 September 2008 (UTC)
With two qualifications: The format may be changed by consensus. For that matter, I don't think that most of us who discuss the matter are Canadians: if we misrepresent what Canadians find natural, the provision here should change.Septentrionalis PMAnderson 15:52, 9 September 2008 (UTC)
This was just one person's personal preference. There is no consensus on date formats in Canada, given the country's close ties to both Britain and the U.S. and the rather heterogeneous origins of its population, so if an article starts out with one date format it should generally stay that way. Additionally, Phillips is related to the British family, so maybe the British date format is preferable, and she was born and grew up in Quebec, which uses the French date format (3 Mai 1978).RockyMtnGuy (talk) 16:47, 9 September 2008 (UTC)
As I said to you at Talk:Autumn Phillips: thanks for your input, but you're unfortunately wrong on a number of counts. The facts so far are 1) either format is acceptable, 2) most Canadian subject articles use the [month] [day], [year] form. I'll further elaborate that this therefore isn't a matter of "personal preference," unless you're specifically alluding to the desire to draw the formats of articles with a common subject nationality into some kind of consistency; I'm not sure how many times I have to repeat myself about what the majority of articles with Canadian subjects use, regardless of whether or not that reality emerged purely by coincidence. (I imagine it evolved as such because that is typically how english Canadians write the date; never as [day] [month] [year].) If, however, there is some policy (whether existing, or soon to be) that says something along the lines of: for articles with Canadian subject matter, either date format is acceptable, however that which is first used shall remain unaltered and be the benchmark format which all subsequently added dates shall follow, then that would trump consistency across articles, and guide future editing. --G2bambino (talk) 20:12, 9 September 2008 (UTC)
  • Yep, that's the advantage of being the first major contributor to many articles: you get to choose those optional styles. I don't see the matters you raise as being enforceable reasons to change the existing formats—they are too prone to debate and argument as to their importance—an invitation to disputes, which is exactly what we want to prevent. In any case, does it really matter? The differences are trivial. This is a case of leave well what is there. Tony (talk) 17:02, 9 September 2008 (UTC)
I'm not exactly sure that trying to attain consistency is really all that trivial, per say. But, trust we Canadians to say "whichever makes you happy!" and therefore blow away much ability to have consistency. I'm just curious, though: will the first come, permanently stay idea become a guideline for all Canadian articles? From people's words above, I get the impression that it's currently not, and can't find evidence to the contrary. --G2bambino (talk) 19:51, 9 September 2008 (UTC)
  • I think we may need something even more cut and dried than 'first major contributor'. The first major contributor's efforts may have been subsequently eclipsed by another major editor. Will the person who then comes along to perform the audit function go to the page history to ascertain who is the first major contributor? Unlikely scenario, methinks. Ohconfucius (talk) 03:20, 10 September 2008 (UTC)
The words opening WP:DATE should be guidance enough: In June 2005, the Arbitration Committee ruled that when either of two styles such as 14 February or February 14 is acceptable, it is inappropriate for an editor to change an article from one style to another unless there is a substantial reason to do so. For an article on Canada, either style is acceptable, and the reasons given above do not seem to me to be substantial enough reason to change, especially where this results in disruption. --Pete (talk) 03:24, 10 September 2008 (UTC)
This is a bit that I find a little unclear: are strong ties to a country considered "substantial reason" in ArbCom's ruling? What does the full text of the ruling actually say? GregorB (talk) 19:14, 11 September 2008 (UTC)
Read the ruling yourself; I'm not sure who made this apply to this form of date warring, but the case deals with the AD/CE variety - and the ruling is clearly general. Septentrionalis PMAnderson 19:21, 11 September 2008 (UTC)
I've read it and it is interesting, because it could be reasonably argued that applying options C or D from the poll would actually contravene the 2005 ruling. Moreover, it appears to me that ArbCom ruling plus ENGVAR yield exactly the A option from the poll and nothing else. GregorB (talk) 20:18, 11 September 2008 (UTC)

Proposal to temporarily disable date autoformatting

While I personally like the date autoformatting feature (including the date links) I realize that many people do not. So, rather than argue about this, I propose we simply perform an experiment to gauge the reaction from editors.

PROPOSAL

  1. Temporarily stop unlinking (or linking) of dates — except as consensus on each article's talk page dictates — for some specified period of time.
  2. Have the Wikipedia system administrators temporarily disable the date autoformatting feature in the MediaWiki configuration for the English Wikipedia.

That should allow us to see the response from editors, with minimal changes (one line in a configuration file, vs. hundreds of thousands of edits to articles to remove links) and in such a way as to be easily reversible should the need arise. While this proposal does nothing to address the issue of overlinking, it does address the more general issue of whether to keep date autoformatting — and if it turns out that the most people prefer that date autoformatting be turned off, then the unlinking of dates can be performed in an officially-sanctioned manner by a bot, which is far better than the scripted-but-still-manual process taking place right now. --UC_Bill (talk) 21:12, 9 September 2008 (UTC)

Support

  • UC Bill
  • Daytrivia
  • Sapphic — I prefer we keep DA, but at least this test would help decide things one way or the other.
  • Weak support, but this decision should not be made here. This problem arose precisely because many editors with opinions on date formatting did not know that MOSNUM was discussing the problem. This affects many users, and many articles. WP:VPP is the minimum, and a banner announcement would be better. Septentrionalis PMAnderson 14:43, 10 September 2008 (UTC)
  • Weak support, Kumioko — I don't know if this is the right way to test the problem or the solution but I agree with UC Bill that it is easily undoable if the general public doesn't like it. I too like the date links and although some would say that users don't do it, I have been known to click on a date link to see what else happened on that day. In the end though it seems like no matter what anyone says about this change (unless they are in support our arguments are written off as trite and ignored.--Kumioko (talk) 18:08, 11 September 2008 (UTC)
  • Support first option but don't turn off autoformatting. Dates should not be unlinked till this process is sorted out. I'm annoyed that no one was told about this until Tony's bot started delinking pages I had put lots of work into. JRG (talk) 04:10, 13 September 2008 (UTC)

Oppose

  • Oppose in form suggested --Gerry Ashton (talk) 21:43, 9 September 2008 (UTC)
  • Strong oppose mostly because we should never live-test a feature that the bulk of the users aren't aware of. Or basically, implementing a change to "shock" the environment (in order to get editor response) will be a terrible backlash. --MASEM 22:26, 9 September 2008 (UTC)
  • Oppose, for same reasons as Masem. Teemu Leisti (talk) 01:22, 10 September 2008 (UTC)
  • Oppose - Personally, I don't really care, because I do not use it. However, it appears to me that autoformatting is not the cause of our problems, so we should look in a more appropriate place. Disabling a function which a sizeable minority uses is just going to be a huge source of complaint that we haven't got a handle or the real problem. Ohconfucius (talk) 03:12, 10 September 2008 (UTC)
  • Strong oppose. “Shock therapy” is disruptive, and this is not the proper forum for raising the issue. Askari Mark (Talk) 01:57, 11 September 2008 (UTC)
  • Oppose—this would not be an accurate way to test community reaction. Tony (talk) 03:38, 11 September 2008 (UTC)
  • Oppose - for those of us managing content areas, it means we can't easily see where date formatting needs to be managed. I think it would also provoke unnecessary drama. Orderinchaos 10:15, 11 September 2008 (UTC)
  • Oppose Seeing as how this was started and an RfC was started on the issue, I feel that those who are continuing this anti-Tony crusade are starting to become disruptive. I would ask that they stop and keep it to one forum at a time. Ottava Rima (talk) 13:47, 11 September 2008 (UTC)
  • Oppose. Lightmouse (talk) 13:58, 11 September 2008 (UTC)
  • Oppose—this will not give users a taste of what we're calling for. JIMp talk·cont 14:09, 11 September 2008 (UTC)

Comments

  • I'm opposed in principle to disabling the autoformat feature, as it would mean tossing out a good function because aspects of it are not working. That aside, if this test proceeds, it must be separate from the other functions of the date feature (such as the ability to control how dates and times display in a user's watchlists, histories, and so on.) The "preferences" selection also controls that display, but there is no reason to remove that functionality. --Ckatzchatspy 21:30, 9 September 2008 (UTC)
While the same user preference settings are used in both cases, only the in-article date autoformatting is enabled/disabled with the $wgUseDynamicDates configuration setting. I also disagree with disabling the autoformat feature, and am hoping that by temporarily disabling it, enough of an outcry will come from upset editors that it becomes clear to everybody involved that there is more support for keeping it than disabling it. But I'm also open to the possibility that I'm wrong, and that only a small percentage of editors will be upset, or that even more will be quite happy with having autoformatting disabled. The majority here seem to dislike autoformatting, but I don't believe the regular editors of this page are anything like a representative sample of the concerned editors, and so I think this experiment of temporarily disabling autoformatting is the best, most visible way to attract attention to the issue. --UC_Bill (talk) 21:40, 9 September 2008 (UTC)
I believe if the trial is to be conducted, the articles should look as proponents of autoformatting removal ultimately want them to. That is, dates that are autolinked without a pipe, directly to a month, day article, and/or to a year article, should have the link removed, as well as leave the date in the same format as in the wikitext. Otherwise the test is meaningless. --Gerry Ashton (talk) 21:47, 9 September 2008 (UTC)
That defeats the whole point of the test, which is to see what the reaction will be to the removal of autoformatting, without having to change hundreds of thousands of articles. If (as I expect to happen, but maybe I'm wrong) a lot of editors complain after discovering that date autoformatting has been disabled, then it would be better to have to only restore one configuration setting, rather than fix hundreds of thousands of articles. On the other hand, if the temporary disabling of autoformatting turns out to be popular, it can be left off permanently, and a bot can handle the unlinking in a matter of days. --UC_Bill (talk) 22:05, 9 September 2008 (UTC)


As far as I can see, most of the editors here don't object to the autoformatting per se, merely that the vast, vast majority of wikipedia readers don't see the autoformatted text, and instead see a whole jumble of dates in inconsistent formats. The removal of the links is related to, but not just a consequence of the autoformatting; they are being removed because they link to meaningless trivia unrelated to the subject in question. The removal of the "displayed links" but keeping the autoformatting by nowiki tags solves the latter problem, but not the former, in terms of what readers actually see. If the autoformat can be disabled temporarily, it might make editors more aware of what "readers" actually see. This of course can be tried on a local level by turning off your preferences, even if only for a short while. Personally I would prefer to see consistently formatted, unlinked dates; I am perfectly happy reading dates that aren't in my (former-) preference/default style.–MDCollins (talk) 22:10, 9 September 2008 (UTC)
I object to autoformatting per se as it is implemented, and I object to all the proposals to fix it that I have seen so far. I feel the minimum requirements for autoformatting are (1)accept for input a date that has been marked up with a means to determine if it is AD or BC, and a calendar (Julian or Gregorian) (reasonable defaults could apply) and (2) output the result in a form that is not ambiguous, using an acceptable default for readers who are not logged in or who have not selected a preference. The lack of ambiguity requirement rules out all-numeric format. --Gerry Ashton (talk) 22:44, 9 September 2008 (UTC)
  • Autoformatting introduces far more trouble than it’s worth. Everything about it sucks: it makes *pretty* date formats (*sound of audience in enthusiastic applause and cheering*)—but only for registered editors who’ve set their user preferences (*sound of audience doing a disappointed “awe” in unison*); it automatically creates links to other articles (*sound of audience in enthusiastic applause and cheering*)—but the links are to mindless lists of random trivia that don’t have jack to do with the article 99% of the time (*sound of audience doing a disappointed “awe” in unison*). Good riddance. Greg L (talk) 01:31, 10 September 2008 (UTC)
Then you support a trial period of having it disabled? --Sapphic (talk) 01:52, 10 September 2008 (UTC)
  • Sorry, no Sapphic. Autoformatting could be a wonderful tool if Wikipedia got under the hood of how its servers worked and made it so the requesting reader’s I.P. address could be mapped to a country. Then we could make autoformatting work for everyone. But such a radical change to the way Wikipedia works behind the scenes isn’t in the offing anytime soon. It was simply wrong of the developers to have made a tool that allows editors to write what is essentially custom content just for us editors to enjoy but which often totally mucks things up for regular readers. Wikipedia isn’t about us; it’s for making articles that read as smoothly as possible for regular readers. The best option—the one I support—is to no longer use these tools and to deprecate their existence from current articles. Greg L (talk) 02:10, 10 September 2008 (UTC)
  • DA and wikilinking of dates ought to be completely independent, and the discussion should also be independent. It will never be resolved until we decide to address it in those terms. I care immensely about [not having] date-wikilinking because it exaggerates the blue sea of mess, but I really couldn't care less about the debate here if an article should carry international or American date formatting (except that it should be one or the other). DA is fine if it didn't have to rely on wikilinked dates. Then each WP reader has a choice whether to engage DA or not. Ohconfucius (talk) 03:07, 10 September 2008 (UTC)
There is a patch submitted to do exactly what you suggest. You can vote for it at that page, if you like. There are actually two patches for that bug, so make sure to add a comment explaining that you prefer the one that turns off linking of DA-formatted dates, as opposed to the one that autoformats non-linked dates. --UC_Bill (talk) 17:34, 10 September 2008 (UTC)

←Let me get this right: disabling would render all dates as their raw format (without square brackets), and would display them in plain black text such as we see the post-signature dates, yes? Would it mess up any template-generated dates? Is is possible for WP alone to disable it, not WikiMedia (this is the assumption)? Tony (talk) 03:27, 10 September 2008 (UTC)

No, but if that's preferred, it could be implemented with a patch (even simpler than the last one, since it would just be removing the links and leaving the format untouched.) So far as I know (and judging from bug requests to have it turned on in other languages) date autoformatting is only implemented on the English wikipedia anyway, so WP == WikiMedia in this regard. I'm not sure what you mean exactly about template-generated dates, but if the template outputs dates with non-piped, non-prefixed linked dates then it would affect them the same as any other similarly linked dates.
Without a patch, date autoformatting can be disabled (with a one-line configuration change) but linked dates will still be linked; they just won't be reformatted according to user preferences. So disabling DA that way would be the same as switching all editors to the "No Preference" date formatting option. --UC_Bill (talk) 17:16, 10 September 2008 (UTC)
I have submitted a new patch that will effectively nullify date autoformatting. It causes non-piped, non-prefixed linked dates to be displayed without a link and in the same format as they are in the page code. If you think that would be more useful for testing, let us know. --UC_Bill (talk) 23:25, 10 September 2008 (UTC)

Why are we voting on this? WP Admins don't have the power to do this. We'd have to make a request to the developers and when it comes to autoformatting I think we've shown that to be akin to talking to a brick wall ... in a vacuum ... JIMp talk·cont 09:02, 10 September 2008 (UTC)

Please stop being obnoxious and insulting to the developers. There was response from them to the date formatting issues — several years ago — and the developers at the time have since been blamed for implementing what was really just a poorly thought-out process. Of course they want nothing to do with it now, unless a coherent plan can be arrived at that takes into account all positions (aka a "consensus.") We're voting on this particular proposal because it's about as minimalist as it can get (to help with building consensus.. start simple) and it has the very concrete effects of limiting the friction (by stopping mass unlinking) and gathering useful data (by means of an experimental disabling of DA to gauge editor opinion.) After the test, we would have a very good indication of the percentage of editors that are happy with having DA disabled, vs. those that are upset about it. If the proposal gets enough support then it can be brought to a wider audience, and then (assuming it continues to meet with support) to the developers. Then we can conduct a test, and then make permanent plans to either fix DA or disable it for good, based on accurate information about editor preference. Make sense? --UC_Bill (talk) 17:28, 10 September 2008 (UTC)
People make mistakes, sure, but what I so obnoxiously & insultingly was referring to were the several coherent partitions to the developers to fix the mess. Anyhow, it's an interesting idea if it is possible—do you have a plan for implementing it and doing so so as not to surprise users unaware that it's an experiment we're conducting? JIMp talk·cont 17:49, 10 September 2008 (UTC)
Yes, the plan was described in my previous comment. First we build consensus here for the two-part proposal (stop mass unlinking, disable DA in the config.. or possibly apply a patch to disable it in a different way, described in a reply to Tony1 above) then we present it to a wider audience (presumably through VP or someplace) and then — assuming it meets with widespread approval — we present the plan to the developers. Since it doesn't involve any permanent changes to the system and is easily reversible, and since it would provide useful, concrete data, I believe the developers would be amenable to the proposal, if it has support from enough of the community. Presenting it at VP should limit the amount of "surprise" and in any event, getting feedback from editors is pretty much the whole point of the test, so the "surprise" factor isn't really a huge issue. --UC_Bill (talk) 18:26, 10 September 2008 (UTC)
Well it is an interesting idea. It might be a bit tricky to implement. I still think it would involve a fair bit of surprise, though, in spite of any mention at VP. Another thing that is bothering me is that this is not what we advocates of date linking depreciation have in mind. We are calling for consistently and appropriately formatted dates in raw text. Simply turning the autoformatting off will generally produce the opposite. I'd rather conduct a different experiment: let the dates on all daily FAs for a month be delinked (if only temporarily) and gauge the reaction. JIMp talk·cont 14:09, 11 September 2008 (UTC)

That does have the advantage of being easy to implement. Lightmouse (talk) 14:24, 11 September 2008 (UTC)

As long as the mass random (Tony1 insists on not using lists to manage/document the unlinking) delinking stops while we're gathering data on editor reaction, I could agree to this. --UC_Bill (talk) 16:54, 11 September 2008 (UTC)
Jimp wrote: "I'd rather conduct a different experiment: let the dates on all daily FAs for a month be delinked (if only temporarily) and gauge the reaction." This sounds like an experiment that would be worth doing, and would not be too disruptive. I support the idea. Teemu Leisti (talk) 22:48, 11 September 2008 (UTC)
Bill, thanks for your response to my queries above. I strongly oppose any change that simply switches everyone to "no preference", but leaves all of the dates in bright-blue. On your reaction to Jimp's frustration with WikiMedia: please try to empathise with those of us who put a lot of work at Bugzilla to try to get them to change things. However, I'm kind of glad they were like a brick wall, because in those days we didn't realise the full extent of the disadvantages of DA. I must say that I find the sudden activity in writing patches so that square brackets might be retained to be rather like putting a raincoat over your clothes before jumping in the shower, rather than taking off your clothes. And I still don't see why the cleansing of DA markup from articles should affect your stats, since identifying dates is easy with a script. But maybe that's not your methodology. What is of greater concern is that, like User:Colonies Chris, I'm highly suspicious of the scope of the sample, given that the breakdown of the stats bears absolutely no relation to our experience of auditing dates in general articles. Where are all of these unlinked dates coming from, to give such high proportions to them in your tabular representations? How are global averages (that's what stats are) going to inform an organised, strategic campaign to run scripts to remove DA? Tony (talk) 09:03, 12 September 2008 (UTC)
The unlinked dates appear mostly in references and templates (or both, in the case of {{cite}}) and thus might end up as linked anyway because of additional processing (in templates) or escape notice (in the case of bare dates in references.) That's why both Sapphic and I were trying to eliminate dates from within references from the statistics. Sapphic's current processing (still underway I believe) is keeping track of dates separately depending on whether they're in a template, a reference, or in the article text (at least, that's the assumption based on what she's written on the project talk page) which should make it very easy to get a breakdown any way we like. De-linking dates does make identifying them more difficult (because the link brackets serve as easily identifiable boundary markers) but not impossible — however it does make it impossible (for a script) to know which dates are capable of being reformatted, as others have pointed out. At this point I don't expect anybody to listen, but I'll once again make an appeal to stop the manual/script-assisted mass unlinking of dates; it's not helping anything, no matter what your position on date autoformatting is. And it's undeniably being a DICK to proceed with the delinking while other editors have given you plenty of reasons to hold off, and even offered to develop ways to do it in a more efficient manner. --UC_Bill (talk) 15:50, 12 September 2008 (UTC)
It's uncivil to issue orders to anyone in your edit summaries once, let alone twice in a row, THE SECOND ONE SHOUTED. That's likely to make anyone ignore what you're saying. In any case, you still haven't explained satisfactorily why these stats are valid or useful for any strategy in the future. I don't buy the methodology. I don't buy the approach. I'm not convinced by the belated, frenzied attempts to come up with yet more techno solutions when getting rid of techno has been agreed to and is widely popular. People don't want it. It does indeed look as though you're trying to stall reform, and I have no intention of taking the slightest notice of your threats. You've tried to personalise the issue by launching an RFC against my "behaviour", which is pretty nasty. Please don't continue that kind of intimidation; I don't buckle easily. Tony (talk) 16:07, 12 September 2008 (UTC)

Poll being conducted

  • For those who might miss it, a poll on a proposed guideline for writing fixed-text dates is shown above. Click here  to go to it. Greg L (talk) 22:23, 9 September 2008 (UTC)
    • Can I suggest the poll is moved to a subpage to make it easier to find? Do we need to encourage other editors to come and give their views? If so, is it possible to post a notice at the top of the people's watchlist requesting participation? I'm sure this is possible, but haven't got a clue how it works.–MDCollins (talk) 22:33, 9 September 2008 (UTC)
      • My personal preference is to encourage the maximum number of editors possible to participate in the vote. I rather dislike conducting things in remote backwater venues. Let’s see what Tony thinks would be the best forum since he proposed that I place this here. I really would like to clearly resolve what exactly we will be voting on before we get too many more votes. It seems to me that options B and D are identical, other than the nature of the wording. Greg L (talk) 00:17, 10 September 2008 (UTC)
        • D requires you find out what the non-English speaking country, with no obvious tie to an English speaking country, uses as their English equivalent format, and apply that throughout the article. Right?--«JavierMC»|Talk 00:22, 10 September 2008 (UTC)
          • Ugh. Too much research and fussing. This is en.Wikipedia; editors shouldn’t have to research what date format some central African country uses when writing about how Gorilla habitat is disappearing in that country. What format a non-English-speaking country uses has no more bearing on making the article read smoothly for English-speaking readers than does whether the people in the central African country use tongue clicks when they speak. We really ought to make this simple: Is the article about the U.S. or closely tied to the U.S.? Then use U.S. date format. Otherwise, just use the international format. If we keep trying to make this so complex, by the time we’re done, we can get a Nobel Peace Prize for developing a hybrid of chaos, game, and quantum theories just for figuring out how to write out a God-damned date. Greg L (talk) 01:01, 10 September 2008 (UTC)
            • You write the paper and I'll sign it :) --«JavierMC»|Talk 01:11, 10 September 2008 (UTC)
              • LOL. Thanks for your support. Greg L (talk) 01:24, 10 September 2008 (UTC)
              • Why do we even need the exception? Isn't the easiest way to say "always use day-month-year"? Powers T 02:10, 10 September 2008 (UTC)
                • Because articles such as Boston, Massachusetts and Spokane River Centennial Trail will have a much higher proportion of Americans reading it than other articles not closely associated with the U.S. To keep such articles as natural and smoothly reading as possible and with few (!) brain-process interrupts, such articles should use the date format that is most natural for that audience. Wikipedia should not be used as a vehicle to promote change in how the world works. That’s why articles on an American muscle car will have the displacement of the V8 engine in cubic inches first, and as a parenthetical conversion to liters. We’re here to make easy-reading, informative articles that are written appropriately for a given target audience. We are not here to promote the adoption of the SI, the adoption of IEC prefixes like “mebibyte” (symbol MiB), dates formats, or anything else. We go with the flow in order to produce good articles. This mission is made a bit more complex because of the enormous breadth of the readership Wikipedia enjoys, the enormous breadth of the topics it covers, and the nature of the collaborative writing environment. Greg L (talk) 02:28, 10 September 2008 (UTC)
                  • Most contributors aren't going to read MoS before adding content. You pretty much have to have been an editor for a while before you are a regular user of MoS. I'd rather have good referenced content than drive editors away by hounding them over style. There's always bots or wikignomes to come along and shuffle everything into shape. I don't know what dating rituals mountain gorillas use, but in Africa it's all international format. --Pete (talk) 03:30, 10 September 2008 (UTC)
                    • Indeed, most contributors aren’t aware MOSNUM even exists. But it’s still a good thing having rules here or chaos would rule supreme. I think that, in most cases, contributors become aware of MOSNUM’s rules when edit conflict arrise and the more well informed editor cites the rule here. I can pretty much guarantee you that if MOSNUM adopts shitty rules, conflict arrises and Wikipedia’s articles suffer. Greg L (talk) 04:15, 10 September 2008 (UTC)
                  • My point, Greg, is that the conception that American readers are going to be somehow horribly jolted by the appearance of day-month-year date formats is grossly mistaken. It's a very minor issue, far different from spelling or measurements. Everyone seems to be assuming American brains will freeze when encountering these dates, without providing any evidence of that. I submit that it's not worth instituting an exception to the date formatting rule (and inviting the corresponding a) arguments over whether an article is strongly tied to the U.S. and b) inconsistency in our MOS standards, making it harder for contributors to contribute) to avoid whatever extremely minor "hiccup" in reading might result from a slightly unfamiliar date format. At the very least, a nation-neutral option ought to be available in the poll. Powers T 16:06, 10 September 2008 (UTC)

I am not going to take part in this poll. I have stated my preferences already, feel free to read and repeat them. I hardly noticed date inconsistencies until we started discussing it. Date formatting is less important to me than having raw text appear as WYSIWG without blue links. So I will go along with any of the options proposed by Tony. Lightmouse (talk) 09:05, 10 September 2008 (UTC)

  • Powers, I agree. As an American, I don’t have a problem with either format. But some editors clearly have strong feelings about it and we need to avoid editwarring and have a guideline that makes the most sense. Greg L (talk) 18:45, 10 September 2008 (UTC)
  • I haven't seen any serious opposition (to deprecating autoformatting, yes, but not to using dd mmmm yyyy format dates in American topics), but I admittedly haven't read all of the megabytes of discussion above. Powers T 22:41, 10 September 2008 (UTC)
    • Yes, there is; not all American editors - and fewer American readers - are accustomed to military idioms. The most recent discussion of this very point has been archived here. Septentrionalis PMAnderson 22:45, 11 September 2008 (UTC)

History of date autoformatting and ISO-8601

Since the recent debate concerning date autoformatting sometimes touched on how the software was originally intended to work, I have gone through old versions of this page and prepared an essay, History of ISO 8601 and date autoformatting in Wikipedia. If it seems too long, skip to the "Summary" section.—Preceding unsigned comment added by Gerry Ashton (talkcontribs) 10:35, 10 September 2008

Run-off poll

A run-off poll, on the two highest scoring options (A and C) has been posted here in hopes of developing a general consensus on this issue. Greg L (talk) 00:21, 11 September 2008 (UTC)

Notice of ANI

All:SMcCandlish posted this ANI over my posting and moderating of the above polls. Note that I did this at Tony’s request here on my talk page. The desires of SMcCandlish are clear with his 4-0-0-0 vote. He participated in a poll, and then he declared that he thinks all polls to be evil. Clearly, a lot of editors want to participate in this because so very many voted in the poll and also engaged in well behaved debate and discussion. Then SMcCandlish, uhm… *modified* this page by deleting the run-off poll. I’ve responded to SMcCandlish’s fallacious charges at the ANI but am too pissed off now to participate here at the moment. Someone else (Tony?) will have to deal with that character right now. Greg L (talk) 01:23, 11 September 2008 (UTC)

I never said a lot of editors did not want to participate in the poll. My main issue is your unilateral closing of it while in-progress (i.e. preventing a lot of editors from participating!), and my second issue is with the enormous bias introduced by falsely forking option A into options A and B which really are the same thing, which makes it look like C has far more proportional support than it actually does. Also, I never said polls were evil, despite your claim (the fifth one, I believe) that I did. I said they usually don't help much, and that many editors oppose their use. — SMcCandlish [talk] [cont] ‹(-¿-)› 05:58, 11 September 2008 (UTC)
S McCandlish did cite the very well-known essay m:polling is evil. Its opinions on the bad side of polls, and that they must not be majority votes, are very widely held; the conclusion, which is the title, is often felt to be a bit strong (including in the essay itself) - but that is what it's called. Septentrionalis PMAnderson 05:27, 12 September 2008 (UTC)
Stanton, you're both Wikifriends, and I urge you to put this behind you—forgive and forget. I'd close the RFC, Stanton, and I'd bring back the original poll, Greg. Run-off might be possible later, but some of the options involve quite major change. It would be wise to get more community involvement before starting a run-off. Tony (talk) 03:50, 11 September 2008 (UTC)
It's not an RFC, and it's not even really a personal dispute (other than I've noted some NPA transgressions). It's a procedural matter about how the poll has been conducted, and the revert force being used to ensure that it gets conducted one person's way, and about how the results are invalid and misleading. I consider it a moot issue at this point, because several others have also objected to the premature closure of the first poll, so there is clearly no consensus for it, and I think I've made the case clearly enough that option A and option B are actually the same thing and that the numbers have to be adjusted as a result. Forgive and forget: I don't harbor grudges to begin with, and none of this has any effect on my ability and willingness to work collaboratively with Greg L, with whom I am most often in agreement. I simply cannot abide polls that are (even if unintentionally) skewed, especially if their intent is to shape the direction of consensus-building. It just doesn't work that way. And if my attempts to resolve such problems get reverted (with free bonus personal attacks), I will go through proper channels to resolve the issue, not continue to edit war. WP:ANI, WP:RFPP, etc., exist for a reason. I've never filed an RFC against anyone, nor a block request against anyone other than vandals and sockpuppets, because I simply don't get angry enough with anyone to go there. It's not about winning, it's about consensus working properly. — SMcCandlish [talk] [cont] ‹(-¿-)› 05:58, 11 September 2008 (UTC)
A and B are in the same spirit, and many of us could accept either of them, but they are distinct. A would have an article written in American, except for international dates, adapt the dates to the rest of the language - and conversely for an article in any Commonwealth dialect (save Canadian). B would leave both cases alone; but it is true that both A and B would repudiate Pete's crusade to impose the format he is accustomed to on Wikipedia. This result, however, is an absurdity: for example, if I were Jimbo, I could have improved on the wording of both A and B; that does not amount (as averaging would have it) to partial acceptance of the unclearly worded and nationalist C. Septentrionalis PMAnderson 15:12, 11 September 2008 (UTC)

Citation templates

Resolved: Just a pointer to another discussion.

Linking of dates is no longer encouraged in MOSNUM. However, these links are at present institutionalised within the above templates. I am referring specifically to the 'accessdate' field, which links the ISO-formatted dates to the relevant date/year articles - no other date fields are linked. I am proposing that all the dates in the above templates, where linked to year or date articles, be de-linked. The discussion is taking place here. Ohconfucius (talk) 04:30, 11 September 2008 (UTC)

Agreed - that's a problem I've hit on a number of articles in my watchlist. Orderinchaos 12:09, 11 September 2008 (UTC)
Just use accessmonthday or accessdaymonth and accessyear instead; they aren't linked. SandyGeorgia (Talk) 04:36, 12 September 2008 (UTC)

Canada

I think what to do about Canada should be a separate issue from the main issue. There are less permutations if Canada is dealt with separately. I also see NO support in the discussion for "all new Canadian articles will use international format (dD MMMM YYYY)". Every Canadian newspaper at http://digital.vancouversun.com/epaper/viewer.aspx uses MMMM dD, YYYY format. What basis is there for the suggestion that International format (dD MMMM YYYY) be used for ALL new Canadian articles? For me (and for many Canadian newspaper readers) reading such an article would be like reading a British article about Canada. Having that specification for Canada within a proposal seriously detracts from any support I would give it.--JimWae (talk) 06:18, 11 September 2008 (UTC)

In my auditing of article dates, I've come across a few Canadian articles with dmy, although my impression is that more than three-quarters use mdy. Canadian articles a much more likely to have dmy inconsistencies where mdy appears to have been the global choice. Quebecois writing in English here would be highly likely to use dmy, I'd say, and a minority of native anglophones prefer dmy. The solution is not to force either format on the whole lot, but for Canadians to work out whether the "first major contributor" rule should apply when there are rumblings, as in the recently contested article on this page. If not, do they have another surefire method? There is only one other possiblity I can think of: go with the current choice (IF that is clear), and if unclear in its jumble of both, then go back to the first contributor's choice. Tony (talk) 08:25, 11 September 2008 (UTC)
Should be noted too that almost all Australian newspapers use American date format for some incomprehensible reason (and in many cases, have done for well over 100 years), even though nobody uses them here in general or formal use. So the media are a very poor barometer. I do agree though that in my broad observation from my time in Canada though, outside Quebec where European dates are institutional in every sense, Canadians generally use MMM dd, YYYY. Orderinchaos 12:12, 11 September 2008 (UTC)
Concur on Canada (I've lived there too, recently). — SMcCandlish [talk] [cont] ‹(-¿-)› 05:03, 12 September 2008 (UTC)

I think choice C is unnecessarily presumptive in choosing THE format for Canada by choosing THE format for ALL future Canadian articles--JimWae (talk) 19:40, 12 September 2008 (UTC)

Fractionating the survey is the only way

The only way to move this whole thing forward is to break up the issues into smaller bits (one of which is Canada, one is articles related to non-anglophone countries (History of Russia), and one to articles without strong country ties (Mars, Elephant) and to ask for people's preference range for each.

The debate is complicated because the number of permutations is vast: there are five categories of article, four possible decision-making criteria, and the spectre that some of those criteria need to default to others when things are unclear. Believe me, things can be unclear—I've audited dates in a lot of articles. The prime objective is to prevent disputes, so every base needs to be covered. I have a concrete plan for this. It's here. Tony (talk) 12:30, 11 September 2008 (UTC)

I am skeptical that this kind of polling will be useful either. People will often pick simply whatever first appeals to them, without consideration of the ramifications of their selection (a major problem in the original poll, above). There is no new consensus that WP:ENGVAR is suddenly invalid, ergo most of these questions are auto-settled the instant they arise. ENGVAR's strong impliation, even as presently written, is that for Canada and S. Africa, where both date formats are commonly used, there is no strong national/cultural/linguistic tie to either of the two date formatting styles, and for non-Anglophone countries the question is even more moot, ergo it defaults to English variety (and thus date formatting) used by first major contributor. I've elsewhere tossed around a few other ideas on exceptions, examples, and how to deal with unusual cases, but I don't think I need to get into them again right this instant. I just want to make a major observation that a lot of the handwringing that is happening is being done by people who simply had no idea how inconsistent the date formatting has been, even inside the same article, and are now panicking about it, even though for 99.9% of WP readers it has always looked this way and the sky hasn't fallen down yet. This heightened state of "oh crap, we have to fix this, now, now now!" has resulted in proposals that are a) redundant, as the matter is already covered by extant guidelines (even if they need clarification), and b) against current consensus, because they would directly contradict extant guidelines. I wish everyone would calm down. To the extent there is a real problem, it can be addressed by tweaking ENGVAR to be clearer and adjusting the MOSNUM wording so that both guidelines are in step with each other. Those camps who want to take more stringent measures, like mandating DD Month YYYY format for all articles but US-centric ones regardless of which variety of English is used in the article, or banning Month DD, Year format totally, even for US articles, have a major new consensus to try to form with everyone at MOSNUM and MOS/ENGVAR, against a long-standing pre-existing one. Do not confuse these two issues, folks. 1) Adjust what has worked for a long time to work a little better. 2) Radically change it. Not equivalent! :-) — SMcCandlish [talk] [cont] ‹(-¿-)› 04:31, 12 September 2008 (UTC)
And therefore, if there is indeed no consensus, we can just be silent. ENGVAR and our header on established styles are enough - as long as there is no language to encourage date warring. This is why I have supported both A and B. Septentrionalis PMAnderson 04:45, 12 September 2008 (UTC)
Right. — SMcCandlish [talk] [cont] ‹(-¿-)› 05:01, 12 September 2008 (UTC)
If there's no consensus, I'd just remove the dispute tag and leave it as is: that will entrench Method C, and forestall any campaign to convert US to international in articles related to non-anglophone countries or no country at all (am I right?). Just why there's a dispute tag in the anglophone-related section I don't understand; I'd have thought the "Retain the existing format" was the disputed one. Tony (talk) 06:44, 12 September 2008 (UTC)

I think the tag was there in relation to the sentence Pete kept re-adding, which has now gone (hopefully for good, since the one thing to have come clearly out of the poll was a rejection of option D). So I've removed that tag. I agree with Tony that the present wording (without the option-D sentence) should remain (although I think it would entrench method B, not C). Silence would mean a lack of guidance which will encourage edit-warring and endless pointless discussion (effectively the many KB of debate we've just have could be duplicated on every article's talk page). While option C won the poll, and I don't personally have any problem with it, I think it's too radical to implement on the basis of this poll alone. If its supporters can promulgate it in such a way that the community at large cannot claim to have been unaware of it, and still gain consensus for it, then that's obviously fine. --Kotniski (talk) 07:21, 12 September 2008 (UTC)

I agree. Wikipedia is an international effort, that is why there is a Spanish, Russian, Indian, etc. wikipedia that applies to those non-English speaking internationals. This however is English Wikipedia and the dating style of Russia should have nothing to do with what version we use in English Wikipedia. They are not tied to an English speaking country, therefore whatever style came first and predominately throughout such an article should be the overriding dating format.--«JavierMC»|Talk 08:05, 12 September 2008 (UTC)
Excellent. Please see the section below and add your name. Tony (talk) 11:41, 12 September 2008 (UTC)

Date link removals

I've noticed that certain users are going around with bots removing ALL links to dates from articles, claiming that this page supports this, although I can't find any sign of a discussion or consensus about this. I have no objection to removing excessive date links, but I find removing ALL date links to be totally over the top. All I can find on this page is a vauge statement about not linking dates unless there is a reason to do so. This is surely entirely subjective, and not particularly helpful. Perhaps some guidance on what WOULD constitute an approporiate link to a date would be helpful, such as on infoboxes. And can we agree that removing all date links from articles is not a particularly sensible practice? G-Man ? 19:26, 11 September 2008 (UTC)

The guideline is unfortunately vague. It should be changed to reflect the current consensus (or at least I hope there is a consensus) of not unlinking dates until the discussion about the date autoformatting is finished. Unlinking dates before that time is potentially harmful. GregorB (talk) 19:33, 11 September 2008 (UTC)
I agree with GregorB. Tennis expert (talk) 20:29, 11 September 2008 (UTC)
  • Guidelines on have always encouraged restraint in linking. I accept that there may be cases where linked dates are helpful to the understanding of a given article, and thus should not be de-linked. However, I believe these are in a minority of articles where only very selective linking would be required. These are often overlooked when the task is automated. I know that some may regard this as a pendulum swing away from the near-obligatory linking of dates due to its use in autoformatting of dates which has been deprecated. I think that deprecation is for good reason: overlinked articles make it more difficult for readers to navigate and discern which links can potentially enhance understanding of a topic. The relevance test could be the existence of a legitimate back-link from the date article. However, I feel that it is on the whole preferable to de-link all dates first (by script or by bot) because it is so labour intensive, and some may be manually re-inserted later in a considered and selective manner in instances where the understanding of the article is enhanced. Ohconfucius (talk) 06:26, 12 September 2008 (UTC)
  • GregorB, the discussion about DA went on for at least two years, and thank the lord has been resolved. There is nothing vague about the guideline: "deprecated" is quite clear, and the change has wide support among WPians. Thanks for your concern; I'm happy to discuss it further on my talk page if you have specific questions. Tony (talk) 06:34, 12 September 2008 (UTC)
To me, it is unclear whether merely saying DA is deprecated gives a license for mass unlinking. Let me explain why I think mass unlinking is potentially harmful at this point. Unlinking will presumably be done with bots. There still is a chance for doing some sort of "smart" unlinking, where square brackets are removed and the underlying date format is changed. For example, a script could be made that looks at date formats in an article, determines which is the majority format, and brings the rest in line (only wikilinked dates, of course). This kind of cleanup will be impossible after the wikilinks have been removed. I've said already that the current DA syntax is a kind of semantic markup, however primitive it may be, and once you remove it, you lose information and cannot go back the same as you cannot unring the bell. This is not an argument against deprecation of DA (although it might be), but an argument for not doing it hastily. GregorB (talk) 08:29, 12 September 2008 (UTC)
I think it is useful to allow date conventions at the user level. I think the problem was that it was initally set up as a link. When I am reading an article and see a date, I want the date to be clear to me. I do NOT need a link to everything that happened on that date back to 5000BC. I think what would solve nearly ALL these issues would be a new date reference that did not cause a link, but did allow individual readers to use their preferred format Autkm (talk) 06:43, 12 September 2008 (UTC)
Gregor, it's interesting that you raise this issue of the potential loss of meta-data. Mark-up is not at all required for a bot or a script to identify and treat dates in an article. The current script does this already; among its most valuable functions (cleverly extended by Gimmetrow last month) is to iron out the inconsistencies in date formats in an article, whether they're linked or unlinked. In fact, it's a little too active in doing this, which requires careful human oversight, since it's yet to be upgraded to disregard date formats within quotations and image titles. This has led to a few justifiable complaints, and has been a lesson to those runnning the script against complacency. Given the intricacies of how date formats are chosen globally for an article, I can't see that a bot or a script could ever monitor each article for compliance with the guidelines. It's hard enough getting editors to do so. What is a bot to think of the articles that are 50/50 US/international formats? In mixed cases, and in the absence of ties to an anglophone country, I go for the least disruption by retaining (and making consistent) the existing format wherever possible, typically leaving a note to local editors to ping me if they decide on the other format. I'd be interested in your feedback from actually running the script. Please let me know if you're interested in doing so. Tony (talk) 08:47, 12 September 2008 (UTC)
Exactly: wikilinking not only says "this is a date" (this is trivial, except for ISO 8601, which could conceivably be a license plate number, a phone number, or the like), it also says "it is safe to change its format". Once wikilinks are gone, there is no way to be sure without human oversight, as you noted yourself. In the end, we might as well decide to do a "dumb" unlinking - at any rate, it will do no harm. But whatever we do or don't do, it is important to be aware of the consequences. (P.S. I'm interested in this script you mentioned - will contact you when I catch some time...) GregorB (talk) 09:38, 12 September 2008 (UTC)
We can replace wikilinking with a template (that explicitly calls out day, month, and year, along with a format parameter); the template would not produce a wikilink'd date, but one spelled out by the format. It needs to be as simple/stupid as possible, something like {{d|2008|9|10|int}} to make it easy to format. This can be done by a smart bot and leave dates in a machine format that, if smarter approach to DA is installed in MediaWiki, we can likely make one change in this template and all such dates will immediately be DA friendly. We can also create similar templates for date ranges as well. Yes, it is a PITA to write that instead of [[2008-09-10]] but it also much more forward-compatible. --MASEM 20:55, 12 September 2008 (UTC)

Dispute about factual accuracy: new tag

Gimmetrow has recently posted the tag in the autoformatting section. I can't begin to imagine which fact is open to interpretation. Could he please be specific? Tony (talk) 06:31, 12 September 2008 (UTC)

The four "reasons". Wikipedia_talk:Manual_of_Style_(dates_and_numbers)#Ruminations_on_date_linking. Gimmetrow 12:14, 12 September 2008 (UTC)

Here are your rebuttals of the reasons, pasted in from that link:

  1. autoformatting does something for non-registered and non-logged-in users. Not only does it involve some formatting, but everyone benefits from what was, I think, was the main reason for the date linking, the reduction of date-format wars.
  2. rendering the linked dates without the links would solve this (and see later)
  3. the iso format implying a Gregorian date is pretty much irrelevant. The guideline already says don't use iso-style dates in the text, and publication dates in refs should match text. Accessdates for online material will presumably be late enough that this isn't a concern.
  4. idioms should work in the text as written.

1. MOSNUM says "The feature can ony be seen by registered users, etc". You're interpreting the feature as the blue splotch, whereas clearly the meaning in this context is the formatting of the date elements itself.

2. MOSNUM says "The resulting links are normally to lists of historical trivia which often have little or nothing to do with the subject of the article. The use of these formatting tools therefore tends to produce overlinked articles." These are very good reasons to deprecate—the reasons are valid. You are proposing a remedy that retains a key disadvantage of DA: that editors don't see what their readers see, and that good date management is hampered by our failure to see the underlying formats. This remedy, in any case, is irrelevant to the reason stated.

3. I don't know.

4. I don't understand. Tony (talk) 12:47, 12 September 2008 (UTC)

Not sure why you responded down here rather than up in the other thread. But
  1. The "feature" includes some date formatting for everyone, and I don't just mean blue.
  2. If this is only about linking, then editors do see exactly what readers see, and they would under my idea. But on point: if dates were autoformatted but rendered without a bluelink, however dates contribute to the appearance of overlinking, that contribution would be eliminated.
  3. This argument is basically saying a Julian date written as [[1 January]] [[1500]] might autoformat to 1500-01-01 and be understood as a Gregorian date because it looks like ISO. Probably, in any article written so [[1 January]] [[1500]] is not ambiguous, 1500-01-01 won't be either, but assuming it's a problem, there is another solution. I think it's generally agreed that 1500-01-01 or 1900-01-01 is inappropriate in prose. That particular date format preference could be discouraged, or those choosing that preference could accept whatever problems it causes.
  4. We write text as: [[1 January]] [[2001]] was the start of the year. If someone has MDY as a pref, this happens to appear as: [[January 1]], [[2001]] was the start of the year. The text as written is (let's assume) correct, and the text as autoformatted MDY is (let's assume) incorrect because it should have a comma after 2001. Should we take this into account when writing the original text? Perhaps we should, but it's not obvious to me it's any great burden to use an idiom where the punctuation is the same either pref. But perhaps the "small minority of Wikipedia’s readership" who choose a date pref can accept that sometimes the punctuation may be incorrect.
Gimmetrow 13:41, 12 September 2008 (UTC)
  • Response to 1: This is news to me; what kind of formatting? When I come in as an anon, for example, I just see what I see with "no preferences" chosen as a logged-in reader. That's not "formatting" as I see it.
  • Response to 2: How can they see "exactly" what we see? They have no prefs set.
  • Response to 3: I'll leave that to others.
  • Response to 4: In DA syntax, the omission of the comma in US formatting is inconsequential for WPians, since the system adds it in display mode (although I'm unsure whether it does for IP readers). The script adds it when delinking. Tony (talk) 14:41, 12 September 2008 (UTC)
To 1: Formats are "cleaned up" to some extent for everyone, even readers not-logged-in. Thus [[January 1]] [[2001]] displays as [[January 1]], [[2001]] (with a comma) for everyone who doesn't set a pref.
To 2: "If this is only about linking". This "reason" refers ONLY to the appearance of blue, it has nothing to do with rearranging text. Everyone sees the same links.
To 4: Yes, the system makes [[January 1]] [[2001]] display as [[January 1]], [[2001]] for everyone without a preference, but that's not what this "reason" is about. It's saying that a raw text sentence
[[1 January]] [[2001]] was the start of the year.
ought to appear as
[[January 1]], [[2001]], was the start of the year.
in MDY format, with a comma after 2001. The system doesn't add that comma for anyone, so those with the MDY preference setting would see a sentence with incorrect punctuation. Is this a significant problem? Is this a "reason" for dumping the autoformatting system? Gimmetrow 16:34, 12 September 2008 (UTC)
I think Reason 4 is referring to date ranges, punctuated with an en dash ("August 13–15, 1991"), and slashed dates ("the night of 29/30 June"). It could be taken to refer to that frightful date-final comma we sometimes see, and I agree then that the wording should be more specific. On Reason 1, I suppose the focus of whoever drafted the final text was on the order of the month and day; that is what most folk will think of. You're right, strictly speaking, but I'm sorry that you and Greg L, and now you and I, have got into a tussle about it. These seem to involve refinements rather than being worthy of a dispute tag, don't you think? In other ways, Gimmetrow, I'm very grateful for your technical expertise and razor-sharp mind, I hope you know. Tony (talk) 16:54, 12 September 2008 (UTC)
  • {disputed} tags are being misused on MOSNUM. Those should be used only to dispute FACTS in articles. One should find these things (briefly) in articles like September 11 attacks or General relativity. They are too often being used to dispute value judgements of what constitutes best editorial practices in a style guide. Gimmetrow: stop abusing the system. The general consensus is in support of the current wording. It seems the more vehement the opposition to this stuff, the more likely editors are to resort to this sort of tool. It is disruptive. Greg L (talk) 18:37, 12 September 2008 (UTC)
The tag wanted is called {{disputedtag}}; for less than a whole page, it should be invoked with the argument "section=yes". Septentrionalis PMAnderson 18:46, 12 September 2008 (UTC)
  • On substance: these are indeed the reasons why those who dislike autoformatting dislike it. I would welcome reasons to like it, myself; I would find a MOS that gave reasons for alternatives much more useful. Septentrionalis PMAnderson 18:51, 12 September 2008 (UTC)
Well, it seems to me that last two reasons involve very narrow situations and admit of solutions other than dumping the entire DA system. In that sense I'm questioning the validity of the reasons. The second reason is a little more complicated, but I find the links of some use as metadata. For instance, I've checked for possibilities for Today's Featured Article by getting the intersection of links from WP:FA and the what-link-here for a specific date. It's possible for these links to exist as metadata without appearing blue. Gimmetrow 20:03, 12 September 2008 (UTC)
  • On 3, I agree; but Gerry Ashton expressly holds that the possibility of producing error by autoformatting to 1582-01-01 is a fatal objection. We have filed a bug requesting that autoformatting into all-numeric, which was originally presented as a concession to a single editor, be removed, but that hasn't happened either.
  • 4 means that, however autoformatting is done, a large class of sentences will be mispunctuated. This is not narrow.
  • The meta-data should be added as an advantage, although it seems marginal over searching for "September 13" and "13 September" in article space. Septentrionalis PMAnderson 20:39, 12 September 2008 (UTC)
  • 4 could be handled by a template either like {[ 1984,]} or (probably better:) {[Comma-maybe]} that either inserts a comma or does not depending on whether US format is presented (and other puctuation does not follow) --JimWae (talk) 22:13, 12 September 2008 (UTC)
    • We could indeed do that; but it would require that everybody entering a date use some such non-intuitive syntax - and it would require a editor typing in one format to know and think what the other format would look like and whether it would need a comma. 4 used to say something about this, but I see it has been oversimplified. Septentrionalis PMAnderson 22:32, 12 September 2008 (UTC)

Support or oppose on just one matter

OK, I'm sure everyone's getting tired of polls and drawn-out debate on how date formats should be chosen for articles that have no strong cultural ties to anglophone countries. Instead of a complicated choice of options, it would be very useful to know whether just one of these options can be ruled out or should still be in contention. Then we might be able to move forward.

Specifically, I want to measure your view of just one issue, the "country-driven" method. This is the method by which the date formats used in articles related to a foreign-language country should be chosen on the basis of the date formats closest to those used in that country. By this proposal, where US format is used in such articles, this would have to be changed to international format.

Here are examples of US formatted articles related to countries that use day-month-year format:

I ask you to bear in mind the complexity of the situation in many countries in the three categories of countries listed, here and below. Even though the information lacks citations, it's clear that more than one system is used in many countries, not just Canada. Tony (talk) 09:59, 12 September 2008 (UTC)

Support—use the date format closest to the one used in the non-English-language country


Oppose—the format used in the non-English-language country is irrelevant

  • Tony (talk) 09:59, 12 September 2008 (UTC)
  • it should be "non-english language country", please ;-) Ohconfucius (talk) 10:12, 12 September 2008 (UTC)
  • Non-English "first language" in a country is irrelevant and the first-major-contributor rule applies. For English-speaking first-language countries, the style appropriate to them is important (to prevent edit-warring) for articles "with strong ties". In which case For articles without strong ties, or equal claims, ENGVAR (leading down to first-major contributor) applies.–MDCollins (talk) 11:30, 12 September 2008 (UTC)

* Country is entirely irrelevant whether it's an English-language country or not. Powers T 12:56, 12 September 2008 (UTC)

  • Yes, we've already decided this in the first poll. Option D came a clear last. We don't need a new poll every day. In fact I think it's time to close all ongoing polls and decide what conclusion to draw from the results. As I said above, I think we should leave the current wording of the guideline in place for now, and try to establish whether there is really wider community consensus for option C (which won the poll but would be a radical and controversial departure from existing policy).--Kotniski (talk) 13:13, 12 September 2008 (UTC)

Neither point had got through to me. OK, poll closed. Seems after the fact, now. I wish I hadn't wasted time with this, and instead had asked the pertinent question for Option C. It will need to be the subject of a poll. Tony (talk) 14:42, 12 September 2008 (UTC)

Can someone archive some of this mess?

This whole talk page is huge and if someone, who knows how, can archive the dead sections, would be much appreciated. I would but don't want to mess up the "system." --«JavierMC»|Talk 23:07, 12 September 2008 (UTC)

Consider it done. Archiving the whole page after moving sections that should be elsewhere. Teemu Leisti (talk) 16:29, 13 September 2008 (UTC)


Archive 105 Archive 108 Archive 109 Archive 110 Archive 111 Archive 112 Archive 115

Date format

There has been much discussion recently about matters connected with dates. The main issues are as follows:

  • What to do about autoformatted dates. These are linked dates of the form January 19, 1988 or 19 January 1988, which have traditionally been used in Wikipedia articles, purely for the purpose of making the autoformatting tool work (this allows logged-in users to select a preference as to how such dates are displayed). This has now been deprecated (i.e. it is no longer the recommended style). The manual of style page itself currently states the reasons for this decision (basically it's because the benefits are minuscule, and it leads to massive overlinking) and also contains a reference to the discussions which led to the decision to deprecate. Please do not come here asking when or why this happened unless you have trouble understanding that information.
  • What format to use for dates. With the deprecation of the autoformatting links referred to above, people came to realize that we don't have very clear guidelines on what date format should be used in articles (basically whether - and in what situations - month should precede day, as in January 19, or day precede month as in 19 January). Discussions and polls on this matter can be found at /Archive 110.

Summary of date format polls

My proposal is that this polling (at /Archive 110) now be closed and we decide how to proceed next. This is how I personally would summarize the results of the polls (others may differ):

  • The idea that our date formats should depend on the date format used in non-English languages (option D in the poll) has been rejected.
  • The idea of making date format an element of English variety (so that - to simplify - June 1 goes with American spelling and 1 June goes with UK spelling) was well supported (option A), slightly more so than what is effectively the status quo (option B, which allows date formats to become established in articles independent of US/UK spelling).
  • The most popular (though not overwhelmingly so) approach was (option C) the idea of preferring the day-month style in all articles which do not have a specific connection to the U.S. (or other English-speaking countries which do it the other way). This will be considered a radical move in some quarters (which haven't necessarily been fully aware of the implications of this poll).

I would suggest the most logical way forward would be to establish first whether there really is community consensus for this last option. A well-publicized and precisely-worded proposal would do it. If such a proposal fails (unfortunately I suspect it will, though who knows...) then we could come back and look at option A again, which (being relatively uncontentious) might well gain consensus.--Kotniski (talk) 08:03, 13 September 2008 (UTC)

Looking at the polls, they are well-supported, and the tabular format worked well. There were some attempts to influence voting and interpret results, but there is a clear winner in Option C, which led with an average of 2.26 over the runner up (Option A) on 1.75. The run-off poll shows an even larger winning margin.
Accordingly, I think we can insert the strongly supported text into the MoS. However, the wording needs work, though not to the extent that it transforms into a less-popular option. --Pete (talk) 11:30, 13 September 2008 (UTC)
Of course you do; but this was not a vote. There is no consensus for your crusade, and only one of these perpetual polls even shows a slight majority preferring your chosen option over an alternative. Concentrating on it alone, with its flaws, is cherry-picking. Septentrionalis PMAnderson 14:23, 13 September 2008 (UTC)
Why don't we move the discussion to the subpage, as the text at the top says? Teemu Leisti (talk) 15:18, 13 September 2008 (UTC)
Because it doesn't. Kotniski's question is "what do we do now?" and should be answered here, if not somewhere even more public. This concerns a lot of articles, and should be done in the light. Septentrionalis PMAnderson 16:05, 13 September 2008 (UTC)
Fine. Also, see new section #Archived /Date format as /Archive 110 below. Teemu Leisti (talk) 17:11, 13 September 2008 (UTC)
  • Teemu Leisti: The best venue for this next, critical step, is to keep this poll in this high-profile venue so the maximum number of editors can participate. One of the shortcomings of Wikipedia’s procedures is how controversial discussions have in the past been moved to remote backwater venues where it tends to drop off editors’ radar. That’s not good. The more controversial the issue, the more we need to foster the greatest participation by the Wikipedia community to ensure we are getting a good measure of the community’s mood. So let’s keep the voting here, well out in the open where the maximum number of editors can voice their opinion and discuss this matter.
Very well, I see your point. Teemu Leisti (talk) 17:11, 13 September 2008 (UTC)