Wikipedia talk:Manual of Style (dates and numbers)/Archive 106

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 100 Archive 104 Archive 105 Archive 106 Archive 107 Archive 108 Archive 110


Litre: l vs L

Do we have an official preference for the use of l vs L for litre? --Random832 (contribs) 21:27, 22 July 2008 (UTC)

As far as I know Wikipedia has no preference. "L" is the official symbol in the United States. The 16th CGPM decided to allow both symbols but "that in the future only one of these two symbols should be retained" and invited the CIPM to follow developments. --Gerry Ashton (talk) 21:52, 22 July 2008 (UTC)
(edit conflict) NIST SP-811 (1995) says "... although both l and L are internationally accepted symbols for the liter, to avoid [the risk of confusion between the letter l and the number 1] the symbol to be used in the United States is L". This sounds like good advice to me. Thunderbird2 (talk) 21:56, 22 July 2008 (UTC)
  • Agreed. I’ve long recognized that the lowercase l is hard to read and parse when using sanserif typefaces. I tend to use uppercase L for liter and lowercase l for the prefixed forms, such as ml for milliliter. I never had anyone correct or modify these. Greg L (talk) 00:02, 23 July 2008 (UTC)
I would not use "ml" and "L" in the same article, for fear that someone would expect consistency, and try to figure out some non-litre-related meaning for one of the symbols. (But then, I am accustomed to using a lower case letter followed by an upper case letter in SI symbols, especially mV and fF.) --Gerry Ashton (talk) 00:07, 23 July 2008 (UTC)
  • I suppose juxtaposing ml and L too close to each other would look odd. When possible, I don’t use “2 l bottle” as the typography sucks so bad you can hardly read it. And I’ve always found “5 ml” to look superior. I suppose it is easy to be “accustomed” to a lowercase letter followed by an uppercase letter because when you get to the derived units, all the single-letter unit symbols are uppercase. So, like you say, it is quite common to see something like mV, which is a lowercase prefix followed by uppercase unit symbol. However, the gram, meter, and second are all lowercase unit symbols and are used very, very often. And although it is rare to see these units in combination with the prefixes greater than 1000, you can get Mg, Gg, Tg (megagram, gigagram, teragram) etc. Greg L (talk) 00:19, 23 July 2008 (UTC)
How about "mL" and "L"? Gary King (talk) 00:31, 23 July 2008 (UTC)
  • I’ve done it that way too and don’t see any reason editors can’t do it that way. I don’t think this stuff needs to be prescribed on MOSNUM, do you? It seems the current international guidelines on usage are suitable for us too. Greg L (talk) 00:34, 23 July 2008 (UTC)
  • P.S. Although, I could see that writing the non-prefixed unit symbol L for liter might be prescribed or recommended on Wikipedia. I certainly never would write “2 l bottle” and would probably change any occurrences of it to “2 L” if I happened upon it. Greg L (talk) 00:40, 23 July 2008 (UTC)
  • We had a discussion about this at WikiProject Chemistry, I think, and decided to treat it as a regional variation per WP:ENGVAR, because there was some insistence that in the UK the symbol for liter (sorry, litre) is alway lowercase. I strongly prefer uppercase to avoid confusion, however. The perfect example is an old typewriter I used to have, which just didn't have a key for the number one--you had to use the lowercase el as a substitute! (In some fonts, however, the likely confusion is between lowercase el and uppercase i.) --Itub (talk) 12:12, 24 July 2008 (UTC)
  • For me this is not a regional/language issue. It's about clarity and (un)ambiguity. MOSNUM should make it clear that L is preferred. Thunderbird2 (talk) 14:39, 24 July 2008 (UTC)
  • Agreed. It should not be regarded as a regional/language issue. The rationale behind the U.S.’s adoption for uppercase L was clearly stated and is a good reason for Wikipedia to follow that convention. It’s simple enough: lowercase l, when viewed in sanserif typefaces is extraordinarily ambiguous looking and hard to read. How about this one: “1 l of Iletin I.” The drug name, I used in this example, just to juxtapose a numeral 1 and some uppercase I's and lowercase l's is Iletin I (I wrote the drug name in “code” typeface so it could be read). Is that about as clear as mud? The expression is somewhat easier to read using a serif typeface: “1 l of Iletin I.” Sanserif fonts are poor for distinguishing between I and l and 1 (again, in “code” face). Expressions with uppercase L, like “2 L of Pepsi” are infinitely easier to interpret in sanserif faces.

    Setting aside the issue of prefixed forms, which I don’t think needs any proscriptions or prescriptions (lawmaking bodies do best when the legislate the least) I agree that the symbol for liter should be uppercase L. Greg L (talk) 00:09, 25 July 2008 (UTC)

There is also the script small l (U+2113 ℓ) that was once used and still has currency in East Asian usages, but I could see using it on the English Wikipedia only in a rare case where both l and L might cause ambiguity. (The precomposed CJK characters, ㎕, ㎖, ㎗, and ㎘ use the script l form in most fonts.) Caerwine Caer’s whines 01:27, 25 July 2008 (UTC)

I agree entirely with Greg L and Gary King. "l" for "litre" is hopelessly hard to discern unless the context is very well established. Tony (talk) 01:40, 25 July 2008 (UTC)

Of course, Greg L has an obvious conflict of interest here, so it is no wonder that he favors LATIN CAPITAL LETTER L to be used. :) Caerwine Caer’s whines 05:31, 25 July 2008 (UTC) Indeed. Liter is my last name. Danger, is my middle name. Greg L (talk) 06:41, 25 July 2008 (UTC)

I agree that this isn't WP:ENGVAR. Use the capital version, L, and not the lowercase case, l, because lowercase l is indistinguisable from uppercase I. Which is why I usually hate non-serif fonts. Headbomb {ταλκWP Physics: PotW} 02:46, 25 July 2008 (UTC)

If Wikipedia's MOS expresses a preference for "L", we have made a decision to take some control over style in our publicaton. If we intend to take any control, we might just as well go ahead and express preference for "L" even when it is combined with a prefix, for the sake of consistency, to educate readers about the concept that any prefix may be combined with any unit, and to avoid having any readers wasting their time trying to figure out why inconsistent symbols such as "L" and "ml" appear in the same article. --Gerry Ashton (talk) 03:31, 25 July 2008 (UTC)
  • There are many disciplines where ml is routinely used. I would hate to prescribe something here on MOSNUM that starts out as an uphill battle. So I would advocate silence on the issue of mL. But, if we have to control every nuance of this issue, I would propose this, which I think balances the opposing needs of flexibility and clarity:

Since the lowercase L (l) is easily confused with an uppercase i (I) when using sans‑serif fonts, editors should write the unit symbols for the liter as follows:
  1. The unit symbol for the unprefixed form of liter on Wikipedia is uppercase L, e.g. “A 5.0 L engine” or “one gallon (3.78 L)”.
  2. The unit symbols for prefixed forms of the liter may be either the uppercase or lowercase forms of L, ml / mL and µl / µL, whichever is most common for that discipline. For instance, “typical porcine injection of glycopyrrolate is 2 mL per 100 kg” or “Add 50 ml of acetic acid…”
  3. In cases where prefixed forms of unit symbols of the liter are juxtaposed in close proximity to the nonprefixed unit symbol L, or are otherwise both used in the same article and in a manner or context where confusion might arise, editors should use the uppercase L in the prefixed forms, such as mL and µL.

This is my suggestion. Greg L (talk) 05:57, 25 July 2008 (UTC)
I think we should keep it simple, say
  • The symbol for litre is L (not l)
The prefixes are covered by the usual SI rule, so I don't see a need to mention them explicitly. Thunderbird2 (talk) 08:21, 25 July 2008 (UTC)
  • My proposal attempts to bow to reality that the SI prescribes a lowercase L for liter. Since prefixed forms of liter, like ml are easy enough to read and interpret and don’t suffer the ambiguity of appearance like the unprefixed “l” form, and since that is the style commonly used in many disciplines—particularly in Europe—the above three-part rule fixes ambiguity and allows editors to write with minimal intrusion by MOSNUM. And (hopefully) there won’t be edit wars over national styles. I think the more we stay away from hard, fast rules that are a big brush, the better off Wikipedia will be. Greg L (talk) 16:41, 25 July 2008 (UTC)
SI permits both upper and lower case. The current brochure includes the statement The alternative symbol, capital L, was adopted by the 16th CGPM (1979, Resolution 6) in order to avoid the risk of confusion between the letter l (el) and the numeral 1 (one). Thunderbird2 (talk) 16:49, 25 July 2008 (UTC)

The litre, and the symbol lower-case l, were adopted by the CIPM in 1879 (PV, 1879, 41). The alternative symbol, capital L, was adopted by the 16th CGPM (1979, Resolution 6) in order to avoid the risk of confusion between the letter l (el) and the numeral 1 (one).

Headbomb {ταλκWP Physics: PotW} 16:52, 25 July 2008 (UTC)
  • Yes, I used imprecise language that was incorrect. The BIPM allows—not prescribes—the use of the lowercase L. So I see no need to proscribe the use of lowercase L when it is paired with a prefix (50 ml), which are much less ambiguous than the non-prefixed lowercase l (50 l). I think we need to withstand the temptation to proscribe and prescribe so much here on MOSNUM in an attempt to gain fabulous consistency from article to article. Few readers notice such things but readers do notice when they encounter “mL” in a chemistry related article (for instance), when they customarily see “ml”. Further, ensuing editors’ battles, like the above-mentioned “regional variation” that Itub wrote of (12:12, 24 July 2008 post) wouldn’t be properly addressed at all with a blanket rule. Why go and piss off other editors at WikiProject Chemistry for minor reasons? Those editors over at WikiProject Chemistry probably don’t even know about this discussion here. We’re doing enough by prescribing uppercase L (2 L bottle) for non-prefixed instances of the symbol and further by suggesting that authors should also use uppercase even in the prefixed versions (like mL) if they are juxtaposed awkwardly next to a L symbol. We don’t need to go overboard and pretend as if we know what is best and tell everyone they have to change and do things a particular way now that they are on Wikipedia—particularly when the BIPM formally permits it with the SI, when some disciplines (like chemistry) commonly do it that way, and when some countries and dialects customarily use ml. What we’re trying to fix isn’t broken enough that it can’t be addressed just fine via the three components of above mentioned proposed guideline.

    If we want MOSNUM to be valuable and relevant, we’ve got to listen to what other editors are saying. You know, some of these editors at WikiProject Chemistry are highly educated chemists and don’t take kindly to being told what to do by novice editors who spend much of their time here on MOSNUM but who aren’t as informed as they ought to be in the practices observed in certain disciplines. Like I said above, legislative bodies govern best when they legislate the least. Tread lightly here, provide editing latitude, and fix only what’s really broken. If not that, then we really ought to solicit greater input from the WikiProject Chemistry editors (or move this discussion over there). Greg L (talk) 22:34, 25 July 2008 (UTC)


  • I therefore propose the following rule, which is SI-compliant, avoids all confusion as a practical matter, and affords editors and groups of editors the flexibility to write articles that reflect common practices in a given discipline:

Second draft of proposed typography for liter unit symbols


Unit symbols for liter
Since the lowercase L (l) is easily confused with an uppercase i (I) when using sans‑serif fonts, and since the BIPM endorsed the use of uppercase L for litre for use with the SI, editors should write the unit symbols for the liter and its decimal multiples and submultiples as follows:
  1. For in‑line text, the un‑prefixed unit symbol for liter (L) should be only the uppercase L, e.g. “a 5.0 L engine” or “one gallon (3.78 L)”. The script form of lowercase L, U+2113 (coded as ℓ) and which produces ℓ should not be used for in-line text.
  2. The unit symbols for decimal multiples and submultiples of the liter; that is, prefixed forms, may use either the upper- or lowercase L, e.g. ml / mL and µl / µL, whichever is most common within that discipline or is appropriate for the topic. For instance: “a typical porcine injection of glycopyrrolate is 2 mL per 100 kg” or “he then added 50 ml of acetic acid…”. The chosen prefixed style shall be consistent within an article.
  3. In the event that prefixed forms of unit symbols of the liter are juxtaposed awkwardly next to a L symbol, or where both appear in an article in a manner or context where confusion might arise, editors should use only the uppercase L in the prefixed forms, such as mL and µL.

The above rule-set means there would be no need to change articles like Cyclohexane, which contains only one instance of a unit symbol for a decimal submultiple of the liter and is clear as glass. There is no need to editwar over something which works well elsewhere throughout the chemistry world and is SI-compliant to boot. Greg L (talk) 18:41, 26 July 2008 (UTC)
Since the above proposed rule says that "L" should be used when there is no prefix, and also says "the style shall be consistent within an article", the logical conclusion is that any article that has an unprefixed "L" should use "L" throughout the article. All the stuff about "juxtaposed in close proximity" creates a contradiction; it implies that a prefixed "l" and a nonprefixed "L" may be used in the same article so long as they are separated by a lot of text.
I also fail to understand the significance of "in-line text". What difference does it make whether the text is in-line, in a table, in a picture caption, or in a citation (except that in non-scientific articles, the unit would probably be spelled out in in-line text)? --Gerry Ashton (talk) 19:00, 26 July 2008 (UTC)
  • Since the above proposed rule says that "L" should be used when there is no prefix, and also says "the style shall be consistent within an article", the logical conclusion is that any article that has an unprefixed "L" should use "L" throughout the article.
    No, you entirely misread what it says by taking two entirely different issues embodied in point #1 and point #2. It is clear enough. Where it says “the style shall be consistent within an article”, that is in point #2, which speaks only to the issue of prefixed forms. But to drive this home, I’ve added “chosen prefixed” to that last sentence of point #2.

    As to I also fail to understand the significance of "in-line text",
    I thought that would give editors the flexibility to use the script form in math-style text where the formula is set off from the body text and is not part of in-line text. Do you think that is a bad idea? What I’m trying to do is avoid writing a rule that would require wholesale changes in existing articles for no good reason.

    As regards All the stuff about "juxtaposed in close proximity" creates a contradiction; it implies that a prefixed "l" and a nonprefixed "L" may be used in the same article so long as they are separated by a lot of text
    No, it doesn’t “imply” anything, that’s exactly what it says and means and there is no contradiction. In total, it says that if the two aren’t used in a way that would cause confusion, then there is no problem. Do you think that if the Cyclohexane had some verbiage somewhere in the body text that mentioned “a 200 L chemical drum” that this is going to really confuse anyone because ml appears in the sidebar? Third-graders aren’t reading that article. And why are you so quick to condemn an SI-compliant style that is routinely used in chemistry if doing so—as the above rules mandate—wouldn’t cause confusion?

    A quick scan of your last 500 edits shows that your interest in chemistry-related articles is slim to none. As an R&D scientist who worked alongside a Ph.D. biochemist for many years on fuel cells (and am currently working on a medical device), I certainly know enough about chemistry to realize that there is simply no reason for non-chemists here on MOSNUM who like adding rules to style guides to be messing around in disciplines in which they have limited understanding. And why would you be so quick to ignore the wishes User:Itub? Let’s take a look at HIS contributions. As if he doesn’t know what he’s talking about in chemistry and should be ignored? Your being so quick to reject his wishes given his clear expertise in chemistry strikes me as a bit arrogant. The above three-rule set addresses the issue by proscribing the use of lowercase l for liter and also affords editors the flexibility to use an SI-compliant symbol so long as editors only exercise a little common sense in deciding if confusion might arise. Greg L (talk) 19:23, 26 July 2008 (UTC)

  • What Gerry Ashton seems to be saying, and I agree with him 100%, is that it is confusing to use L and ml in the same article. You don’t need to be a chemist to see that. Thunderbird2 (talk) 20:04, 26 July 2008 (UTC)
  • It’s not a blanket issue of “in the same article” and that’s why my proposal says “where confusion might arise”—so chemistry editors can use some common sense. As I mentioned above, writing “a 200 L chemical drum” in the body text of Cyclohexane isn’t going to be confusing just because “ml” appears in the sidebar. There’s no reason to go change that one instance of ml in the sidebar just because “L” is later added to the body text.

    You don’t trust chemistry editors to be able to use common sense in avoiding confusion? You think some non-chemists here on MOSNUM should tell them what they have to do because chemistry editors don’t understand their subject well enough to how to write chemistry symbols in their articles? I reject that notion. Utterly. I just don’t seem to have the stomach for writing restrictive guidelines that make blanket, radical changes in the practices of a discipline like chemistry unless we are fixing something that’s truly broken. Prescribing the U.S. convention (uppercase L) for the singular liter addresses an important issue and the BIPM recognized that. I also happen to think that we can point out that using ml and L in the same article certainly can be confusing and chemistry editors should look out for that and harmonize the style if necessary. I actually think chemistry editors can be trusted a bit to use their brains. Don’t you?

    Let’s see if we can get Itub to weigh in here; he’s actually a chemist. Greg L (talk) 20:11, 26 July 2008 (UTC)

I can see Greg L's point to a certain degree. Most advanced articles, in any discipline, will probably not create confusion by using "L" and "ml" in the same article, so long as they are not close to each other. Elementary articles could create confusion, but that could be addressed on an article-by-article basis. But what is the real custom in any discipline sufficiently well-organized to have journals and style manuals for those disciplines? Is there any style manual (or "instructions to authors" or other equivalent documents) for any discipline that advocates using different case for the same unit symbol, depending on whether it is prefixed or not, in the same article? If not, I suggest it is the practice of every discipline that has adopted a formal style that the case of a unit symbol be consistent throughout an article. --Gerry Ashton (talk) 20:23, 26 July 2008 (UTC)
  • Which is why the proposal says “whichever is most common for that discipline.” We don’t need to concern ourselves with the details of every discipline here on MOSNUM. It’s been my observation, having spent plenty of time dealing with chemistry in wet labs that “ml” is common as hell in chemistry. I can also tell that “mL” is often used in medicine. I don’t want to have to research all the individual practices within medicine and try to prescribe and proscribe here. There’s no reason in the world we can’t leave this up to the editors. If some editor who is wet behind the ears tries to edit a medical-related article incorrectly, then the more seasoned editors over there decide what is the proper practice. Greg L (talk) 20:41, 26 July 2008 (UTC)
Greg L's proposal is excessively complicated to allow for a situation that probably does not exist, that is, a dicipline that advocates changing the case of a unit symbol within an article, depending on whether the symbol is prefixed onr not. If such a case does exist, it could be recognized by remembering that the MOS is only a guideline and can be overridden if there is a good reason to do so. --Gerry Ashton (talk) 21:12, 26 July 2008 (UTC)
  • That’s absurd. Those three short points of the guideline aren’t at all complicated. The proposal beats the crap out of any of the alternatives I’ve seen here, which are
A) do nothing and leave instances of “2 l bottle” as they are…
B) prescribe L for liter but not address the issue of prefixed forms, like ml, if they are awkwardly juxtaposed in the same article (would it shock you, Gerry, to learn that lowercase L (l) for liter and ml in the same article are not all rare?)
C) attempt to “fix” problem B by requiring that ml always be mL even though chemistry often uses the former and is SI-compliant
Options B and C are worse than doing nothing at all. So how about we wait to see how some chemistry types feel about this? If there is going to be a guideline, it needs to give editors some flexibility, which doesn’t come with some ham-handed inflexible dictum that attempts to change the way the world works (where have I seen that before?). Or are the opinions of editors who actually understand chemistry no longer of any consequence here on MOSNUM and only the “regulars” who hang out in this venue have a say? We’ve seen what happens when regulars push an unwise guideline without properly considering the consequences and gathering greater input from the experts. Greg L (talk) 23:29, 26 July 2008 (UTC)
I'm a real life industrial chemist and I'm in favor of B or C. Actually, I favor C a little more. If you're gonna do it, might as well try to go all the way. I my experience, whether a chemist uses 3.78 l, 3.78 ℓ, or 3.78 L, depends on when that person was trained. —MJCdetroit (yak) 03:04, 27 July 2008 (UTC)

I'm not convinced that this is a matter that needs resolving in the MOS. Other than NIST, I'm not aware of any standards body mandating one form over the other and any attempt on Wikipedia's part to codify seems a bit of bureaucratic overzealousness that is adequately dealt with already by rule #1 on units of measurements in MOSNUM: "Unambiguousness: Do not write so you can be understood, write so you cannot be misunderstood." Caerwine Caer’s whines 04:43, 27 July 2008 (UTC)

Greg, thanks for all the praise towards me and the chemists, but I don't like your proposal. My favorite option here is doing nothing, followed by prescribing L. Your proposal allows inconsistency within articles, and consistency is the most important rule of style. House style guides exist not to ensure the "perfect" style but to help ensure a consistent style (if there were a perfect style then there would be only one style guide in the whole world!) Sometimes we decide that it is impossible to ensure consistency over all of Wikipedia (for example, with English variations and reference styles), and thus we have to settle for the more modest goal of consistency within an article.
If the legibility of "2 l bottle" is deemed a major concern needing a MOS fix, I would simplify the guideline to "avoid the use of the lowercase el for the unprefixed liter". Give the writer the flexibility of deciding how to fix it without inconsistencies. The options include converting to ml, or dm3, or using uppercase el consistently throughout the article. --Itub (talk) 05:56, 27 July 2008 (UTC)
  • Very well. We needed some chemists here (I’m a cadet chemist and couldn’t speak authortatively to the subject). I see, MCJdetroit, that you can live with B (simply prescribe uppercase L for the unprefixed symbol for liter). I simply don’t think trying to change ml to mL is realistic or wise; it is well entrenched in certain countries and disciplines, is SI-compliant according to the BIPM, and isn’t “broken” enough since it reads well enough. I think I can agree with Itub, that we either leave the issue entirely alone or, failing that, just go for option B. I would add though, that I thought I was doing that and a bit more with points 2 and 3 above, which effectively left ml and mL alone and gave editors the freedom to do what they want to do but further advised to look out for instances where ml was juxtaposed nasty-close to a simple L. Are you saying, Itub, that such details are really best left unspoken if we do prescribe uppercase L for liter? Greg L (talk) 06:31, 27 July 2008 (UTC)

I'm a chemist and I don't think I would agree that C is true. The upper case L is a quite common in chemistry literature. I don't like the lower case l, and I find that the curly l looks terrible. There was a fad at one point for using dm^3 in lieu of liter, and an even older fad for using c.c. (cm^3) in lieu of mL, but that seems to be non-mainstream now. --Rifleman 82 (talk) 08:13, 27 July 2008 (UTC)

  • No doubt mL is common in chemistry. My point was that ml is also common in chemistry. Since it is easy enough to read and since it is often used in chemistry, I see no reason to mandate mL. I see only the need to be consistent within an article. Greg L (talk) 22:51, 27 July 2008 (UTC)
Frankly I don't what being a chemist has to do with anything. Anyone can appreciate that l (lowercase l) can be confused with 1 (one) or I (capital I), and that L won't be. No need to be chemists. Physicists used litres too (and I see mL and ml equally in litterature and MSDS. So do biologists. So do cooks. So do mechanics. No one will ever be confused by "A 250 mL bottle" or "A 42.35 L fuel tank" unless they don't understand what a litre is, and nothing would be different here if it were written "A 250 ml bottle of water" or "A 42.35 l fuel tank". This isn't a WP:ENGVAR issue, nor is it a subjective preference of two equally clear styles. Using l for litre has clear disadvantages. We have an option, compatible with all the standards of technical writing out there, used in the real world, and free of disadvantages: all-across usage of L for litres.
And on top of the obvious advantages of all-across L usage, this is a simple rule. Using ml or mL depending if there's an unprefixed litre symbol other rules which either leads to inconsistency (2 L, but 250 ml), problems such as switching articles from ml to mL as soon as a quantity is given in the unprefixed form. And then switching mL back to ml if the unprefixed quantity is removed on the grounds that an article should follow the style used by the first major contributor. Headbomb {ταλκWP Physics: PotW} 08:46, 27 July 2008 (UTC)
"mL" looks really bizarre. I guess I'm not used to it at all, and am in a metrics country where it's not used in everyday life (I don't know what chemists do, and don't care all that much). However, I find "l" a real problem, as I said above. Is this proposal to mandate a usage, or to strongly recommend it? Is there any support for leaving a crack open where fields or editors want to choose? (This is not a loaded question: I need guidance as much as most editors on this point.) Tony (talk) 14:13, 27 July 2008 (UTC)
  • I agree Tony, mL looks bizarre to many editors. It is, however, common in U.S. medicine. That’s why I wouldn’t want to try to proscribe a restrictive rule here at MOSNUM. My proposed guideline, above, would only prescribe L for liter and would largely stay out of the issue of the prefixed versions, where it would leave the choice of ml v.s. mL up to editors of the respective articles; advising only that whichever version is chosen should be consistent within the article and, further, editors should watch for awkward or confusing placements of ml when the L is being used in the same article. Greg L (talk) 23:01, 27 July 2008 (UTC)
  • This sounds very good. Tony (talk) 01:22, 28 July 2008 (UTC)
  • Thanks. I can use the support. Others feel three points to the guideline are “complex.” but I feel those extra two points are just filling in the obvious blanks that naturally arise from the first point—which seems to have fairly wide support here as a minimal step to take. Greg L (talk) 02:09, 28 July 2008 (UTC)

Guidance is redundant if there is no clear and present or significant problem. It is redundant if editors will not change what they do anyway. I searched Wikipedia and see no significant problem to solve. Evidence of a clear and present problem please. Lightmouse (talk) 14:30, 27 July 2008 (UTC)

Three articles in need of attention are Cubic metre, Netilmicin and Radiocontrast. It didn't take me long to find them so I imagine there are more. Thunderbird2 (talk) 15:01, 27 July 2008 (UTC)
Also look at Nestlé Pure Life. Headbomb {ταλκWP Physics: PotW} 15:09, 27 July 2008 (UTC)
  • Current guidelines give both of you plenty of justification to go and repair those articles that use both ml and mL without the necessity of creating any new guidelines here. Greg L (talk) 21:01, 27 July 2008 (UTC)

Three front runners

If I’m reading this right, there are 3 front runners. In order of increasing complexity, these are:

  1. Do nothing (maintain existing silence on the matter)
  2. Permit only uppercase L (not lower case l or its curly cousin ℓ)
  3. Permit L or l (not ℓ), consistently within an article

My own preference is #2, followed by #3. (The default is #1). What do others think? Thunderbird2 (talk) 10:37, 27 July 2008 (UTC)

  • No, I don’t think you’re reading it right. Greg L (talk) 20:39, 27 July 2008 (UTC)

I vote for #1. Do we need to clutter the page with this? Can we not let editors use common sense? My second preference is #3 (except if directly quoting a curly cousin). Both upper and lower case are valid according to SI. JIMp talk·cont 17:18, 27 July 2008 (UTC)

Yes, SI does permit both, but I don't suppose that approval extends to both styles in the same article. And there’s no reason why MOSNUM can’t prefer one SI-approved symbol over another, if we agree there’s a need to (it would not be the first time). I agree that common sense is the default option, but before you make up your mind, take a look at these:
Thunderbird2 (talk) 17:35, 27 July 2008 (UTC)

No, we should not have both in the one article (with the obvious exceptions). Yes, we are free to decide for ourselves. JIMp talk·cont 17:49, 27 July 2008 (UTC)

  • And current guidelines give you plenty of justification to go and repair those articles that use both ml and mL without creating any new guidelines here, T-bird. Greg L (talk) 20:58, 27 July 2008 (UTC)
  • T-bird misstated the options. Option #2, above (permit only uppercase L) is (at least) ambiguously written and can be interpreted to mean to apply to prefixed forms (like ml) as well as the singular form (L). This also happens to be a solution he personally prefers. The full set of options (omitting the script ℓ for simplicity), as I see it are as follows:
  1. Do nothing. This permits “2 l bottle” (awfully ambiguous and hard to read), “2 L bottle”, “2 mL injection”, and “2 ml of acid.”
  2. Prescribe uppercase L for the non-prefixed unit symbol of liter. This results in “2 L bottle”, and permits “2 mL injection”, and “2 ml of acid.” This option would not touch upon the issue of prefixed forms (like ml), it does not prescribe the common-sense principle that usage of ml or mL be consistent within an article, nor would it recommend that editors watch for awkward juxtapositions of ml next to L.
  3. Prescribe uppercase L for un-prefixed and prefixed forms. This results in “2 L bottle”, “2 mL injection”, and “2 mL of acid.” As Itub pointed out above, he prefers this option personally but notes that there are UK authors who are used to it this way. And I note that ml is very common in chemistry whereas medicine in the U.S.—and perhaps elsewhere—has mostly standardized on mL.
  4. Prescribe only that uppercase L be used for the non-prefixed unit symbol of liter, explicitly point out (rather than implicitly imply via silence) that ml and mL are per editors’ preference as they see fit, but require that editors be consistent with their usage within a single article, and further advise that editors watch out for awkward juxtapositions of ml next to L.
The start of this thread was intended to address the issue of “2 l” (lowercase el) and quickly expanded beyond that into proposals to address the prefixed forms, which are not as ambiguous to read. The U.S. convention (embodied in law) of using uppercase L for liter (also recognized by the BIPM as an acceptable alternative to lowercase el for use with the SI) is a good one.
I note that the single editor specializing in chemistry articles who is weighing in here personally prefers option #3. However, I am loath to write any guideline on MOSNUM without a clear consensus of a number of editors with a diversity of points of views and who all fully appreciate the feelings of other editors and understand the consequences of the changes being considered. Ill considered MOSNUM guidelines are like cancer: bad stuff and hard to eradicate once started. Without greater input from others, this guideline would be the product largely of editors who specialize only in writing guidelines. Accordingly, it seems best that we do nothing here. If someone still wants to do something, I will volunteer to post notices on a variety of Wikipedia article talk pages inviting chemistry and medical editors to weigh in here. Greg L (talk) 20:38, 27 July 2008 (UTC)
  • P.S. Jimp, when you wrote above “No, we should not have both in the one article (with the obvious exceptions). Yes, we are free to decide for ourselves” (in regard to articles like Beer bottle, which use both ml and mL in the same article), isn’t your position fully supported in my above three-point proposal? Greg L (talk) 20:53, 27 July 2008 (UTC)
  • P.P.S. To all: my proposed guideline, above, would only prescribe L for the non-prefixed liter and would largely stay out of the issue of the prefixed versions, where it would leave the choice of ml v.s. mL up to editors of the respective articles; advising only that whichever version is chosen should be consistent within the article and, further, that editors should watch for awkward or confusing placements of ml when “L” is being used in the same article. This addresses problems with articles like Beer bottle that T-bird and Headbomb pointed out, fixes the problem of lowercase el for “2 l bottle”, and other than that, entirely avoids the can of worms inherent in attempting to achieve an across-the-board style for the prefixed versions by prohibiting one of the conventions. Greg L (talk) 23:15, 27 July 2008 (UTC)

Greg, yes, we're free to decide for ourselves, if we decide to go with your proposal (which is pretty sound), we can. And, yes, it does address the issue of consistency within articles. Of course, Beer bottle etc. have other problems besides "l" vs "L" but your proposal would address this. JIMp talk·cont 03:19, 28 July 2008 (UTC)


One more alternative

It seems to me that it would be possible to express the intent of Greg L's proposals above using simpler language. How about something like this:

"Both the uppercase L and lowercase l are approved SI symbols for the litre [1] and may be used on Wikipedia. However, due to its visual similarity with the number 1 and the uppercase letter I in some fonts, use of the lowercase symbol without a prefix is discouraged. (That is, 100 ml is fine, but 0.1 l should be avoided.) To avoid confusion, mixing the upper- and lowercase symbols should be avoided. The Unicode "script ell" character or the precomposed characters , , and should not be used on Wikipedia."

Note that I've deliberately left any modifiers like "in close proximity" out of the sentence advising against mixing the two symbols: that way editors are free to decide, on a case-by-case basis, the appropriate scope (one sentence, one section, one article, a cluster of related articles, one WikiProject...) within which consistency should be maintained. —Ilmari Karonen (talk) 05:28, 28 July 2008 (UTC)

I can agree with this wording. --Itub (talk) 09:13, 28 July 2008 (UTC)
Your simplified wording is fine by me. You did a great job of compacting the essential elements of my proposal into a short and simple paragraph. Greg L (talk) 13:36, 28 July 2008 (UTC)
I support Ilmari Karonen's proposal. Thunderbird2 (talk) 18:10, 28 July 2008 (UTC)
  • Seeing no opposition, I posted Ilmari Karonen’s proposal. Greg L (talk) 21:08, 29 July 2008 (UTC)

I oppose. I'd rather have all-across L.

  1. The l in ml can still be misinterpreted for a 1 or a capital I. This presupposes that the reader knows that we're speaking of milliliters. This supposition is probably alright in most of cases, but as milliliters are a widespread units, but when speaking of centiliters (cl), deciliters (dl), kiloliters (kl), hectoliters (hl), one might easily get confused. Using the capital L exclusively would remove any possible confusion with Is and 1s (cL, dL, kL, hL,...).
  2. Allowing both makes things context dependents in weird ways and prone to edit warring. Articles which uses ml consistent will have to change as soon as a L is introduced. If the L is removed, do all the mL the article return to their former ml form? Do the mL stay as mL in anticipation that someone might use an unprefixed L? All-across Ls prevent this sort of stuff from happening, and saves people the trouble of looking for unprefixed L before deciding wheter or not to use ml or mL.
  3. There is nothing wrong with deprecating the use of l for liters. We've got a superior alternative that's just as familiar, recommended by NIST, allowed by BIPM, and which is problem-free.

Headbomb {ταλκWP Physics: PotW} 02:28, 30 July 2008 (UTC)

  • OK. But we have a general consensus to adopt at least part of what you desire: Write “2 L bottle” and be consistent and also don’t mix “add 2 ml to the 2 L flask”. Both ml and mL are currently on Wikipedia and this new guideline doesn’t change that. The guideline doesn’t permit anything “new”; it does, however, get rid of the most glaring problems that stand out like a sore thumb. Some of the articles you and T-bird cited above are problems with A) mixing ml and mL in the same article, and B) mixing ml next to L. This new guideline fixes both those shortcomings. Further, it gets rid of the glaring problem that was the starting point for this whole thread: “2 l bottle”.

    What you are advocating would certainly be “pretty” but will piss others off for reasons I stated above—including the fact that it is an SI-compliant style and is commonly done that way. That’s why we see the use of ml here. Now articles that feature ml and µl will at least be using them consistently with no mixing. We’ll just have to look at that and accept it. Other than us editors here flailing our arms over this, I expect that few readers will notice that one article does it one way than another. I suspect medical articles will naturally gravitate towards mL and that’s the way it ought to be in my opinion. We can let editors and groups of editors decide what’s better for their articles. If what we have now creates more problems than it solves—and I don’t see how that can be since the guideline only deprecates the worst practices that common sense should have fixed long ago—it can always be repealed or revised. Greg L (talk) 04:59, 30 July 2008 (UTC)

  • I like Ilmari's proposed text. Tony (talk) 10:49, 30 July 2008 (UTC)
  • As a US-centric editor, and a chemist/physicist to be (finishing my bachelors in December), I personally prefer 0.1 L and 100 mL to be used in all instances. The capital L places the emphasis on the L as being the principal unit with the m as a modifier. However, I've seen ml used before and am fully capable of reading it. Its my experience it mostly appears in non-authoritative sources, I can't recall seeing it in a textbook or peer-reviewed journal. Thats not to say its not there, its just not sticking out in my head. But its not something I look for either. However, 'l' without a modifier and the script L are unacceptable in my opinion. The first has the above stated problems. In the US at least, script L implies liquid phase or a quantum number, which I feel is an unnecessary ambiguity. I support Ilmari's proposed text as a matter of pragmatism. If I were to write my own proposal, it would say "The preferred method for abbreviating the liter unit on Wikipedia is 'L'. However, articles which are already written using 'l' are acceptable as long as all instances of 'l' are prefixed (100 ml is ok, 0.1 l is not). Over time these articles should be transitioned to 'L' but as long as the readability of the text is maintained, edits for the sole purpose of compliance with this directive are not necessary." EagleFalconn (talk) 17:11, 31 July 2008 (UTC)
    For the record, I completely agree with EagleFalconn's position. This has nothing to do with WP:ENGVAR and no more to do with chemistry than with any other discipline that happens to use the litre as a unit of volume. I see no good reason not to standardise on L, but can live with Ilmari Karonen's compromise, which is why I supported it. Thunderbird2 (talk) 17:31, 31 July 2008 (UTC)
  • Headbomb, the present version on MOSNUM says as follows:

  • The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since "l" can be easily confused for "I" (uppercase "i" ) and the numeral 1 when using sans-serif typefaces, editors should use only the uppercase L for the unprefixed liter (e.g., editors should write "A 10 L tank" instead of "A 10 l tank"). For prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter), either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer" are satisfactory) but the chosen style should be consistent so as to avoid the awkward mixing of styles.
  • Do not use the unicode "script ell" character and its variants (, , and ).
  • Articles should not mix the uppercase "L" for the unprefixed liter with lowercase prefixed forms like ml. Do not write "This soft drink is availible in both 250 ml and 2 L bottles", but rather "This soft drink is availible in both 250 mL and 2 L bottles".

And when you revert the text and leave off the underlined portion, you leave illogical edit summaries like “ ‘Articles should use the uppercase "L" and the lowercase "l" consistently;’ covers all of this.” For God’s sake; read what you are reverting it to (click on the provided link). The admonition you quote is followed immediately by only an example of how ml shouldn’t be mixed with L. If you want to claim that verbiage like “Articles should use the uppercase "L" and the lowercase "l" “consistently” applies to mixing of ml and mL, then add an explicit example. I don’t give a damn how it’s done but stop trying to tell me that it clearly covers the issue when it clearly does not. If you want to say that the following…

  • The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since "l" can be easily confused for "I" (uppercase "i" ) and the numeral 1 when using sans-serif typefaces, editors should use only the uppercase L for the unprefixed liter (e.g., editors should write "A 10 L tank" instead of "A 10 l tank"). For prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter), either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer").
  • Do not use the unicode "script ell" character and its variants (, , and ).
  • Articles should use the uppercase "L" and the lowercase "l" consistently; i.e., do not write "This soft drink is availible in both 250 ml and 2 L bottles", but rather "This soft drink is availible in both 250 mL and 2 L bottles".

…is clear that ml and mL should not be mixed in the same article, please explain how that is so damned clear given the example used in the last bullet. Greg L (talk) 02:01, 2 August 2008 (UTC)

The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since "l" can be easily confused for "I" (uppercase "i" ) and the numeral 1 when using sans-serif typefaces, editors should use only the uppercase L for the unprefixed liter (e.g., editors should write "A 10 L tank" instead of "A 10 l tank"). For prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter), either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer").

  • Do not use the unicode "script ell" character and its variants (, , and ).
  • Articles should use the uppercase "L" and the lowercase "l" consistently; i.e., do not write "This soft drink is availible in both 250 ml and 2 L bottles", but rather "This soft drink is availible in both 250 mL and 2 L bottles".
It could hardly be clearer. Headbomb {ταλκWP Physics: PotW} 05:45, 2 August 2008 (UTC)
I think the failure to communicate here is that Greg L seems to think of mL and L as different units, like oz and qt. If there were a rule that said the symbol for the quart may be either Qt or qt, that wouldn't tell you anything about whether to capitalize oz. Headbomb seems to think like me. Litre is a unit of measure. Milli- is a prefix, that is an internationally understandable way of saying "thousandth of a". Once one has decided to capitalize L, it is decided, and prefixes have no effect on the decision. Doing otherwise would be like writing "Qt" and "1/32 qt" in the same document. I will never think like Greg L does, but I suppose there may be other people out there who do. Greg L's insistence that the consistency only be for unit symbols that are awkwardly close to each other and not necessarily throughout an article prevents me of thinking of a concise way to explain this. --Gerry Ashton (talk) 15:33, 2 August 2008 (UTC)
  • I think the problem to communicate here is that you, Gerry, don’t realize that ml and L are different units of measure—that’s why there are 1000 mL per liter. Indeed, though they have the same base measure, they have different prefixes and are consequently different units. Talk:MOSNUM isn’t some sort or retarded forum where editors here are supposed to try to show off how stinking smart they are about units of measure. And I don’t think you did yourself much of a favor by blowing out "Qt" and "1/32 qt". Nice try though. Anyone with smarts God gave a goose can see what the problem is with the guideline’s wording that Headbomb is so enamored with. The greater trouble with what Headbomb is stubbornly refusing to see is that if the guideline says to use uppercase and lowercase consistently, and then uses only an example of mixing ml and L in the same sentence, then many a reasonable editor would think that is the scope of that bullet point. Will someone else please weigh in here other than these two guys? My problem is that the wording that Headbomb keeps restoring is does not give, via either explicit admonition, nor example that speaks directly to the point, that even though either ml and mL are OK, they can not be mixed in the same article. For Headbomb to gut out that simple wording on the pretense that what was there before is “unambiguous and clear” (and that the added specificity is unnecessary) is without foundation is beginning to look to me that he is beginning to fancy himself as the mayer of MOSNUM now. The consensus here was that ml and mL are to be used consistently in articles and I insist that the guideline be clear on that point. Greg L (talk) 18:20, 2 August 2008 (UTC)
Greg L wrote "I think the problem to communicate here is that you, Gerry, don’t realize that ml and L are different units of measure...." However, NIST's special publication 330, the official interpretation of SI for the United States, consistently refers to "units" and "submultiples" or "multiples" of units. It never refers to a prefixed unit as just a unit. So I reject Greg L's interpretation. I nevertheless believe that others will share his fallacious interpretation, and think the guideline should be clear about it. However, adopting a rule that says "L" or "l" must be used consistently throughout an article when prefixed, and used consistently when awkwardly close together when one instance is prefixed and the other is not, makes the rule complex to explain. A simple rule that the symbol for litre must use the same case, throughout an article, whether it is prefixed or not, would be much easier to explain. This is particularly so because people who are accustomed to careful writing are accustomed to the idea that once a style is chosen for an article, it should be used throughout the article. Allowing "ml" and "L" in the same article, so long as they are not awkwardly close to each other, will be a foreign concept to careful writers, and it will have to be clearly explained, because those writers will have a hard time believing we have adopted such a weird rule. --Gerry Ashton (talk) 19:00, 2 August 2008 (UTC)
  • When are you going to figure out what this is all about? Guidelines here aren’t for the “careful” writers you wrote about. If everyone here was a careful writer, we wouldn’t need most of the guidelines here. This is a list of the articles Headbomb and Thunderbird2 added above:
  1. Beer bottle
  2. Cubic metre
  3. Marskin ryyppy
  4. Nestlé Pure Life
  5. Netilmicin
All of these either had problems with ml and mL in the same article, or with ml and L being in the same article. One problem on Wikipedia is that different authors prefer different styles and sometimes the compromise solution is leave both in one article. So a guideline that says as follows:
either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer").
…and leaves it at only that isn’t clear enough because it can be interpreted as allowing both—which it does, just not in the same article. And why the hell did you write “Allowing "ml" and "L" in the same article, so long as they are not awkwardly close to each other, will be a foreign concept to careful writers, and it will have to be clearly explained, because those writers will have a hard time believing we have adopted such a weird rule” ??? Do you even know what text is in dispute? I’m not talking about my original proposal and if you had read what was actually being written here, you would have understood that. The text I want is as follows:
For prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter), either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer" are satisfactory) but the chosen style should be consistent so as to avoid the awkward mixing of styles.
So what’s wrong with that? Nothing. And it, or something else like it is going in. That was the consensus: that ml and mL shall not be used in the same article. The guideline Headbomb keeps reverting to does not properly reflect that consensus and stop telling me that it does because there is no truth to that at all. The added wording makes it clear that articles should be consistent with either ml or mL and nothing more. Greg L (talk) 19:42, 2 August 2008 (UTC)
If the rule were "Never use an unprefixed 'l'. Never mix prefixed forms of 'l' and "L" in the same article. Consequently, any article that uses an unprefixed 'L' must also use a capital 'L' in prefixed forms" then the rule would be easy to explain. But Greg L would not agree to that, he wants to allow "L" and prefixed lowercase forms, so long as there are no capitalized prefixed forms, and none of the 'L' symbols are awkwardly close to a prefixed form. Hence a compromise version reached consensus among the paricipants. But the compromise is hard to explain. --Gerry Ashton (talk) 19:51, 2 August 2008 (UTC)
  • No it’s not hard to explain. Read the below. Greg L (talk) 20:00, 2 August 2008 (UTC)

Edit warring against the consensus agreement

Headbomb and Gerry Ashton: There was a widespread consensus that ml and mL should not be used in the same article. There are a number of articles on Wikipedia that mix both. A guideline must be clear that though either style is OK, only one style should be used in an article. Here is the wording I'm proposing:

  • The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since "l" can be easily confused for "I" (uppercase "i" ) and the numeral 1 when using sans-serif typefaces, editors should use only the uppercase L for the unprefixed liter (e.g., editors should write "A 10 L tank" instead of "A 10 l tank"). For prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter), either lowercase or uppercase are acceptable (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer" are satisfactory) but the chosen style should be consistent so as to avoid the awkward mixing of styles.
  • Do not use the unicode "script ell" character and its variants (, , and ).
  • Articles should not mix the uppercase "L" for the unprefixed liter with lowercase prefixed forms like ml. Do not write "This soft drink is availible in both 250 ml and 2 L bottles", but rather "This soft drink is availible in both 250 mL and 2 L bottles

It seems clear that the last bullet point does not sufficiently speak to the issue of mixing ml and mL. In accordance with the consensus view here, I insist that any guideline wording make this clear. Don’t bother reverting it to a version of the text that doesn’t accomplish this agreed-upon principle or I will simply delete the whole damn thing from MOSNUM until we can vote on wording here. I don’t know what you are trying to accomplish here Headbomb. You earlier were pressing for requiring only that mL be allowed on MOSNUM and your efforts here almost look like you are trying to cripple the guideline so that editors continue to mix styles in the article. The underlined sentence is simple and straightforward. Stop deleting it. Greg L (talk) 20:00, 2 August 2008 (UTC)

  • All: Since Headbomb and I seemed to have gotten ourselves in a damned rut of reverting each other (so far successfully limiting ourselves at around two reverts per day), I hope all will agree that the following is clear as glass and is entirely in accordance with the consensus view:
  • The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since lowercase L ("l") can be easily confused with uppercase i ("I") and the numeral 1 when using sans-serif typefaces, editors should write the unit symbols for the liter and its decimal multiples and submultiples as follows:
  1. The un‑prefixed unit symbol for liter (L) should be only the uppercase L, (e.g., editors should write "A 10 L tank" instead of "A 10 l tank").
  2. Except as provided in point #3 below, prefixed versions of the liter (the decimal multiples and submultiples such as the milliliter and microliter) may be written with either the lowercase or uppercase L (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer" is satisfactory), but the chosen style should be consistent so as to avoid the awkward mixing of styles.
  3. Articles should not mix the uppercase "L" for the unprefixed liter with lowercase prefixed forms like ml. Do not write "This soft drink is availible in both 250 ml and 2 L bottles", but rather "This soft drink is availible in both 250 mL and 2 L bottles".
  4. Do not use the unicode "script ell" character and its variants (, , and ).
Greg L (talk) 20:30, 2 August 2008 (UTC)
I agree with Headbomb and EagleFalconn that there is a clear advantage in preferring L, but the consensus seems to be to permit both l and L. On that basis I see no need to depart from Ilmari Karonen's text. Thunderbird2 (talk) 22:47, 2 August 2008 (UTC)
  • There is absolutely no “consensus” for allowing lowercase ell for liter; there is precisely the opposite. Where the hell did you get that idea from the above? What Ilmari Karonen wrote above is clear as glass: “0.1 l should be avoided.” What were you trying to accomplish with your above post? You’ve been pushing for using only uppercase L for all forms (L, µL, mL). And then you switch gears and say that there is a consensus to permit “2 l bottle” when no one here thinks that is a good idea? Why would you write such a thing in an open forum? Greg L (talk) 00:44, 3 August 2008 (UTC)


  • Since I'm uninvolved with this topic would you like me to take a look and decide "what has consensus"? Fnagaton 23:27, 2 August 2008 (UTC)
  • Thank you very much for the offer Fnagaton; good timing. I see that Headbomb did a pretty good job of compacting my four-point version into something that I believe covers all the bases. Other than maybe rounding off some milliliter values and a few other tweaks, I’m happy with what he’ done. Greg L (talk) 23:42, 2 August 2008 (UTC)
  • Cool stuff. :) Fnagaton 23:43, 2 August 2008 (UTC)
  • Damn. Help Fnagaton! I just took a real close look at what he wrote and realized it had logical lapses. I think, Fnagaton, that what we need is not a review of what the consensus is (that much doesn’t seem to be in dispute), but how best to communicate the concepts. Headbomb tried to condense my above greebox into this:
  • Two symbols are recognized for the liter: L and l (lowercase L). Either forms are acceptable, but only one of them should be used in a given article (e.g., do not write This soft drinks comes in 355 mL cans and 591 ml bottles but rather This soft drinks comes in 355 ml cans and 591 ml bottles or This soft drink comes in 355 mL cans and 591 mL bottles).
  • However, when there is an unprefixed liter symbol, use the uppercase form thorough the article (e.g., do not write The soft drink comes in 2 l plastic bottles, 355 ml metal cans, and 241 ml glass bottles but rather The soft drink comes in 2 L plastic bottles, 355 mL metal cans, and 241 mL glass bottles). This is to avoid confusion with the uppercase i (I) or the numeral (1), especially when using sans-serif typeface.
  • Do not use the unicode "script ell" character and its variants (, , and ).
  1. First, the first bullet point implies that lowercase L is OK for liter. The consensus clearly was that it was not. This was fully and unambiguously addressed in my point #1 above.
  2. Second, the second bullet point shows an example of “2 l plastic bottles”, which only reinforces what seems to be flat-out declared as being OK in the first. This shouldn’t have been an option in the first place. This was fully and unambiguously addressed in my point #1 above.
  3. Thirdly, Ilmari Karonen wrote (05:28, 28 July 2008 post) of keeping what is consistent ambiguous so “that way editors are free to decide, on a case-by-case basis, the appropriate scope (one sentence, one section, one article, a cluster of related articles, one WikiProject...)” This was addressed in my point #2 above, where it speaks of “but the chosen style should be consistent” and left it up to editors and groups of editors to decide upon the scope of what should be consistent. Headbomb’s first two bullet points make explicit mention of “articles” and this breaks Ilmari Karonen’s good suggestion. Both Tony and Jimp expressed that they thought that versions I wrote that were closer to my above greenbox. I don’t understand Headbomb’s obsession in reducing what I have down to the point that he breaks the intent.
No one should have to put up with so much aggravation. If someone who knows how to properly and completely compact my four-rule set so the permissible practices are perfectly clear, with no logical lapses, and so the scope of what it applies to is left broad (per Ilmari’s suggestion), then be my guest. But if they can't do it right, don't screw with it please.
Headbomb, your dicking this all up. Run your “compactions” by others here for review and vote beforehand please.
Fnagaton: And yes, I just now noticed T-bird’s little post above (22:47, 2 August 2008) that his interpretation of all the above is that there is a “consensus” to allow lowercase ell for liter (as in “2 l bottle”. He has to know that isn’t true and I suspect him of trying to bait me here. His clearly-stated position all along has been to permit only uppercase in all forms (L and mL and µL). So I really don’t know what he’s trying to accomplish here but by no stretch of any reasonable imagination, could any human being who is truly trying to be helpful possibly think there is a consensus for that. He writings baffle me and don’t appear to have been intended to be helpful in the least. So please advise, Fnagaton, as to your read on what is the consensus view based on what has transpired to date. Greg L (talk) 00:44, 3 August 2008 (UTC)
I have to agree with Greg L on one point. His bullet that begins "The internationally accepted unit symbols for the liter are both the uppercase L and the lowercase form (l)" is better than "Both the uppercase L and lowercase l are approved SI symbols for the litre" because our guideline only partially accepts "l", so we should make sure we make it clear that it is the international community that fully accepts "l", not us. Greg L is quite wrong to get upset about Thunderbird's posts; Thunderbird's post make perfect sense if you ignore Greg L's notion that "L" and "ml" are different units. Since they are actually a unit and a submultiple of a unit, allowing "ml" means we do indeed allow the use of "l" for "litre", just so long as it is prefixed. --Gerry Ashton (talk) 01:13, 3 August 2008 (UTC)
The agreed wording includes the verbatim text
  • Both the uppercase L and lowercase l are approved SI symbols for the litre [2] and may be used on Wikipedia.
You really need to familiarise yourself with WP:AGF and WP:CIVIL. Thunderbird2 (talk) 01:02, 3 August 2008 (UTC)
  • Yes, I knew you were going to come back with your accusations of my being uncivil and all that T-bird. Your MO is clear. But you are simply wrong on each and every count above. Every one. The “agreed wording” was NOT agreed upon by a consensus of any sort; it was simply one of a number of proposals that were thrown out. And it said (says), in full, as follows:
"Both the uppercase L and lowercase l are approved SI symbols for the litre [3] and may be used on Wikipedia. However, due to its visual similarity with the number 1 and the uppercase letter I in some fonts, use of the lowercase symbol without a prefix is discouraged. (That is, 100 ml is fine, but 0.1 l should be avoided.) To avoid confusion, mixing the upper- and lowercase symbols should be avoided. The Unicode "script ell" character or the precomposed characters , , and should not be used on Wikipedia."
So you are being impermissibly misleading and clearly unproductive when you make an allegation with the simple conclusion that the “consensus seems to be to permit both l and L”. The issue is much more complex than that and you full well know it (or at least should know it). You also know full well that there was no “agreed wording” of any sort and that what you cite was just one of many proposals. So stop playing all coy and clueless here. Greg L (talk) 01:49, 3 August 2008 (UTC)
Ilmari Karonen's words were supported by EagleFalconn, Greg_L, Itub, Thunderbird2 and Tony. I naively assumed that is what you had posted on 29 July. I see now that you posted your own interpretation, and yet you have the nerve to accuse me of being misleading. The present wording needs to be replaced with Ilmari's text and any changes to it agreed here. Thunderbird2 (talk) 07:59, 3 August 2008 (UTC)

Possible compromise form?

While there seems to be a consensus to employ uppercase ell (L) for liter, it appears to me that there is no consensus on capitalization of the ell when it is combined with a prefix. Accordingly, I would like to offer a modification of Ilmari Karonen’s proposal that may capture the preferences of most of those involved in this debate:

"Both the uppercase L and lowercase l are approved SI symbols for the litre [4] and may be used on Wikipedia. However, due to its visual similarity with the number 1 and the uppercase letter I in some fonts, use of the lowercase symbol without a prefix is discouraged. (That is, 100 ml is fine, but 0.1 l should be avoided.) When a prefix is used, either the uppercase L or the lowercase l may be used, whichever is most common for the given discipline, but the form selected should be used consistently throughout the article. (E.g., 10 ml [or mL] of iodine and 50 ml [or mL] of water, but not 10 ml of iodine and 50 mL of water.)

"The Unicode "script ell" character or the precomposed characters , , and should not be used on Wikipedia."

Hopefully this may prove a less contentious formulation. Askari Mark (Talk) 01:35, 3 August 2008 (UTC)

  • Pretty good Askari Mark. I like what you have. Note though what Ilmari Karonen wrote (05:28, 28 July 2008 post) of being a bit ambiguous as to what editors should be keeping consistent so “that way editors are free to decide, on a case-by-case basis, the appropriate scope (one sentence, one section, one article, a cluster of related articles, one WikiProject...)” This was addressed in my point #2 above, and I think it is an important point. So I would strike “throughout the article.” With a few other tweaks, I would arrive at…

With the SI, the accepted unit symbols for the liter are both the uppercase L and the lowercase form (l). However, since lowercase L (l) can be easily confused with uppercase i (I) and the numeral 1 when using sans-serif typefaces, the un‑prefixed unit symbol for liter should be only the uppercase L, (e.g., editors should write "A 10 L tank" instead of "A 10 l tank"). Prefixed versions of the unit symbol (the decimal multiples and submultiples such as the milliliter and microliter) may be written with either the lowercase or uppercase L (e.g., either "A 200 ml bottle" or "A 500 mL glass of beer" is satisfactory), but the chosen style should be used consistently. Articles should not mix the uppercase L for the un‑prefixed liter with lowercase prefixed forms like ml; that is, do not write "this soft drink is availible in both 250 ml and 2 L bottles", but rather "this soft drink is availible in both 250 mL and 2 L bottles". Do not use the unicode "script ell" character and its variants (, , and ).

Greg L (talk) 02:27, 3 August 2008 (UTC)
I'm all for reducing verbiage in MoS; there's enough to wade through already. Can this topic not be dealt with in a few shortish sentences, something like: The only abbreviation to be used for liter, when unprefixed, is "L". When prefixed, it may alternatively be written with a lower-case "l" (e.g. "ml"), provided that this is done consistently in an article and provided that the unprefixed abbreviation is not used in the same article. --Kotniski (talk) 05:19, 3 August 2008 (UTC)
Things would certainly be simpler if we stuck with the common sense answer (exclusive use of L). We have to dance around and have a paragraph because of the allowing of the lowercase l, but only on the conditions that there's no unprefixed version of it around. Using Greg L's last version as the basis and chopping/editing stuff, I came to this. What say you? Headbomb {ταλκWP Physics: PotW} 06:40, 3 August 2008 (UTC)
  • The symbol for the liter in a given article may be chosen to be either exclusively L (uppercase L) or exclusively l (lowercase L), as both are internationally recognized. Do not use the unicode "script ell" character and its variants (, , and ). If an unprefixed liter symbol is present, use exclusively the uppercase symbol (L) across the article, as the unprefixed lowercase symbol (l) can easily be confused with the uppercase i (I) or the numeral 1 when using sans-serif typeface.
  • Do not write A 591 ml bottle and a 355 mL can but rather A 591 mL bottle and a 355 mL can or A 591 ml bottle and a 355 ml can.
  • Do not write either A 2 L bottle and a 355 ml can or A 2 l bottle and a 355 ml can but rather A 2 L bottle and a 355 mL can.
I don't like that; the first sentence (which is all many editors are likely to read) gives entirely the wrong impression. I'm sticking by my even more concise version above; anyway, I'm off on a week's vacation, so I'll leave y'all to resolve this vital issue:)--Kotniski (talk) 06:55, 3 August 2008 (UTC)
1) What exactly is that "wrong impression"? 2) Let's keep things simple, but let's not oversimplify them or sacrifice presentation for sake of simplicity. The community has chosen a complicated solution, so this will require a certain level of explanation. I don't like allowing both prefixes one bit, nor do I see any reason for doing so, but that's what was chosen. Headbomb {ταλκWP Physics: PotW} 07:17, 3 August 2008 (UTC)
With one exception, I have no problem with Greg L’s reformulation of my proposal (other than a few minor cleanup tweaks); I was simply trying to keep as close as possible to something that most editors appeared to like. I disagree with the striking out of “throughout the article”, which I believe is much preferable stylistically to the mixed use of ml and mL within an article. I appreciate, obviously, leaving editors with the maximum flexibility, but consistency (as in date formats or ENGVAR) within an article almost always improves its appearance. I have no problem with a cluster of articles, under a WikiProject or not, using the same formulation, but any single article should IMHO employ a consistent style.
I would also like to note with respect to “excess verbiage” that it’s inevitable if the (non-silent) option is to retain the flexibility to use ml or mL; the briefest formulation is only available if the mL form is mandated. Otherwise, the length is driven by the need to explain the rules and by the length and number of examples provided. Askari Mark (Talk) 00:46, 4 August 2008 (UTC)

Clarity on auto-formatting dates?

User:Tony1 just removed the auto-formatting of all the dates in Military history of Canada‎ on the grounds that it was no longer encouraged at WP:MOSNUM Now, personally, I really couldn't care less either way, but I'd like to know what consensus is on this. I didn't see anything in WP:MOSNUM that suggested a major reversal of guidelines, or anything suggesting a large scale removal of such formatting was in order. If I somehow missed it, I'm sorry. If such a consensus has been made, it should be more clearly displayed on the guideline page. If not, I'm not sure what to do about Military history of Canada‎. Thanks, TheMightyQuill (talk) 15:15, 17 July 2008 (UTC)

This is an area of significant debate and there isn't a definitive consensus: there are advantages and disadvantages of date wikilinking. TONY falls on the side of 'mostly disadvantages', so sometimes removes all date wikilinking form the main body of articles (FACs I believe). The MOS says date wikilinking is optional (it used to be required, so 'no longer encouraged' means 'can be done, but doesn't have to be') and can't be definitive due to these advantages and disadvantages. Rjwilmsi 16:05, 17 July 2008 (UTC)

Wouldn't "In the mean time, err on the side of leaving things the way they are" be a standard intermediate policy? Whatever, this really doesn't matter.- TheMightyQuill (talk) 17:24, 17 July 2008 (UTC)

I agree. I just noticed a bunch of fully formatted dates were being unlinked with an edit summary stating that they are no longer encouraged. So I came here, read the page, and didn't see language supporting this. So I came to this talk and see that it is being discussed. I don't see any consensus that supports these mass changes. I happen to really like the date preference formatting feature. I'd be fine with it going away it that's what was decided, but I don't see that it has been decided. --Elliskev 19:54, 17 July 2008 (UTC)
Since I learn't about this (I support it, I think the linking is/was ugly) I have starting telling people to remove them at peer reviews before GA assessment. Should I slow down on that? — Realist2 (Speak) 20:09, 17 July 2008 (UTC)
It seems to me that, until we get some clarity, it is now a stylistic issue. I don't think I would hold up a GA or an FA or any other A over a preference of style (assuming that there is consistency). --Elliskev 20:17, 17 July 2008 (UTC)
The script-assisted edits by User:Tony1 left the articles Ludwig van Beethoven and Wolfgang Amadeus Mozart with inconsistent date formatting (Beethoven's birthday is given as "December 16, 1770" and "16 December, 1770"); this is clearly in violation of MOS:NUM. This inconsistency is eaxctly what auto-linking avoids. Until we get a Template:Date which covers all dates (at least A.D.), auto-linking is the lesser evil. Furthermore, as other have pointed out several times here, there is no consensus or even a change in the guidelines which would justify the whole-sale change of articles. Note: I whole-heartedly agree with the delinking of partial dates. Michael Bednarek (talk) 13:23, 18 July 2008 (UTC)
I've gone through the Beethoven article and rationalised all dates to the correct format for German topics. My experience has been that any significant article on a non-U.S. subject will generally have a mish-mash of date formats, which is visible to the vast majority of readers, except for registered editors. I've had my date prefs turned off for years so I can see dates in the format most users see them. --Pete (talk) 18:04, 18 July 2008 (UTC)

I would like to see links to the consensus discussions to remove autoformating and in fact links to the discussion to make autoformating optional. I am afraid this has become a personal crusade of Tony1 and hope there is a clear consensus somewhere to support this. Rmhermen (talk) 13:49, 18 July 2008 (UTC)

Tony's responses:

I apologise if I've upset anyone by my removal of date autoformatting from articles. This has been part of a trial to uncover any problems in a script. I have to say that people have reacted positively to the reduction in the amount of bright-blue words and underlining in their articles, with two exceptions out of many instances. When the issues were put to both editors, they changed their minds, which is telling: date autoformatting is a complicated issue that many WPians do not understand and just go along with because everyone else has been doing it for ages without question. One of these two users was Tony the Tiger, who's generally a maximalist when it comes to linking, and not an easy to convince. When he got to the bottom of how date autoformatting is fundamentally different from plain linking, he wrote to me: "In response to your post on my userspace, yes I concur linking dates is among the most useless type of linkage." I could quote a lot of positive comments here, but won't unless you request it.

Elliskev and MightyQuill: MOSNUM does no longer encourage the autoformatting of dates. It used to be mandatory, then late last year, I think, this was weakened from "are autoformatted" to "can be autoformatted". This change was not made by me, although more recently I've clarified the language after debates on this page. Optionalisation is not a sudden change, but has evolved over at least two years; you can find numerous debates in the archives here during that period. I think there was significant support for compulsion in the earlier debates, but in the past year or so, and particularly more recently, this has not been the case—rather, there has been continual grumbling on this talk page about:

  • the inadequacies of the system;
  • MediaWiki's failure in the face of a number of requests to address the shortcomings of date autoformatting, since early 2006 I think it was; and
  • the more recent realisation that it's an in-house system only, which conceals from us the inconsistencies and glitches in raw date formatting that millions of our readers out there must put up with.

The in-house-only aspect was not part of debate until recently (I myself didn't realised it until a few months ago). This fact has considerably strengthened support for optionalisation. It's easy to say that there's no consensus, Rmhermen, but I haven't seen significant opposition at MOSNUM to optionalisation for some time.

As for the Beethoven and Mozart articles, it was my fault not to have checked through more thoroughly (I normally do); but I have to ask why you're complaining now? All of your readers out there have been seeing those inconsistencies ever since they were entered; at least removing the autoformatting has made what they see obvious to you, and I hope will lead to their being fixed. If you'd posted on my talk page, I'd have done so immediately. I see that Elliskev has reinstated auto to Schubert, and Michael Bednarek to Beethoven. Who is going to fix the inconsistencies uncovered by the removal of the autoformatting? I'm happy to do the Mozart right now, since that hasn't been reverted.

Elliskev: no one has suggested that a FA or an FA would be held up because of a date autoformatting issue? You weren't implying that, surely. User Realist may well have decided to encourage people to remove autoformatting at GA, and I strongly support him/her, but in the end, it's optional: those who believe the removal is a significant improvement can only put their case to those who are uncomfortable with what has become part of the furniture over the years.

I hope this response has cleared the air. MightyQuill, I wonder whether you're willing to trial the non-autoformatting in the MilHist article for a week or so, perhaps returning to it a few times to judge whether it displays your high-value links a little better, and reduces the colour-clutter slightly on the page. Same or Mozart. I have no wish to pick fights with people who firmly object.

I haven't put most of the arguments for not using autoformatting that have occupied debate here, but we can regurgitate them if you like. I'm interested in your opinions, and hope that you might be willing to support this issue. I believe someone else is setting them out fully in an essay soon to be posted. TONY (talk) 17:46, 18 July 2008 (UTC)


I'm with you, Rmhermen. Where is the discussion and consensus to remove autoformatting? When I last re-read MOS:NUM, it said nothing about autoformatting being deprecated or to be removed. Right now I view date formatting the same I way I view American vs. British/Commonwealth English: both have their appropriate place and shouldn't be changed because WP:IDONTLIKEIT. Honestly, in the debate, I can see the reasoning of both sides in the date issue, but, personally, I like having the dates formatted. On the other hand when the dates are all replaced as "January(ampersand)nbsp;14" or "14(ampersand)nbsp;January" (which ever the preferred order) in the edit window, I think that may be a little intimidating to novice editors. — Bellhalla (talk) 16:00, 18 July 2008 (UTC)
Are you aware that autoformatting only works for the small percentage of Wikipedia readers who are both registered users and who have set their user preferences to a specific choice? In other words, the vast majority of readers, who are not registered, do not benefit from autoformatting and just see an unhelpful "sea of blue". (Most people have learned not to click on date links as they rarely go to an informative article). One of the arguments against autoformatting is that it caters to the Wikipedia elite at the expense of the general public. Also, once autoformatting is removed, the mistakes in date formatting become obvious and can more easily be fixed. Mistakes are hidden from wiki editors by the autoformatting. —Mattisse (Talk) 16:19, 18 July 2008 (UTC)
I've corrected the three date glitches that have been uncovered in the main text of the Mozart article. TONY (talk) 18:13, 18 July 2008 (UTC)

Personally, I couldn't care less about formatting, but if some people do, why not find a solution that satisfies everyone. User:Michael Bednarek mentioned the Template:date which I see doesn't work. What about producing a template like {{ymd|2008-07-18}} that would format dates properly both for editors and for anonymous readers, and use a bot to convert them all (and simultaneously unlink them), instead of just unlinking them on a page by page basis? - TheMightyQuill (talk) 18:16, 18 July 2008 (UTC)

Quill, believe me, every available option has been pursued without success. Bellhalla's point makes a false analogy, IMO, between US vs Br date formatting, and date autoformatting versus no date autoformatting. The logical analogy is between US vs Br date formatting, and US vs Br varieties of English. MOSNUM provides a counterpart to our engvar rules on the use of the date formats in an article just as engvar does for the use of the varieties of English in an article.
Can I say, the differences are yet more trivial between the two standard date formats than between the transatlantic varieties of English. Month/day or day/month? Who cares? TONY (talk) 18:39, 18 July 2008 (UTC)
Perhaps my wording didn't clearly express what I was trying to say. I was not trying to tie the implementation or not of autoformatted dates to the use of American or British English. What I meant was that removing (or installing, for that matter) autoformatted dates in an article that already is using one style should be treated the same way one would approach the switching of an article from American to British English: through discussion and consensus. In my reading of the MOS, as it stands now, there is no wording that requires or even suggests the removal of autoformatting.
And, yes, I'm aware that date preferences in practice are only used by a fraction of registered users, and all of the other reasons on both sides. I personally lean towards linking (because even though an American, I prefer the d-m-y format), but will support whatever the consensus about autoformatting is. And right now, I haven't seen a Wikipedia-wide consensus to remove autoformatting from articles. — Bellhalla (talk) 11:39, 19 July 2008 (UTC)

Tony, please stop trying to convince me, because I really couldn't care either way about the date format. My problem is that you're taking action (in the form of mass removals) when there is no consensus to do so. There is a very big difference between MOSNUM not encouraging the formatting of dates, and MOSNUM discouraging the formatting of dates. Only the latter would legitimize mass removals. Whether or not I would support such a change in guidelines is a totally different issue than whether I would support mass removals before such a change has occurred (ie. before consensus has been reached). - TheMightyQuill (talk) 19:30, 18 July 2008 (UTC)

It isn't U.S. vs British date formating. It's U.S. vs everyone else. Most of the world uses d-m-y. If wikipedia si an international project, then appropriate internationalisation procedures should be in place. --Pete (talk) 18:47, 18 July 2008 (UTC)
Extreme care should be taken in using bots to format dates. In particular, there should be no particular default format, as this invariably leads to dates being presented in the wrong format. An example is the birth date and age template, which defaults to m-d-y, and thus generates inconsistencies when used for articles that should be d-m-y. Another difficulty is with dates in direct quotes, which can be in archaic forms, or using the writer or speaker's preferred format, rather than what Wikipedia might later decide.
I've often wondered whether there could be some sort of metadata template at the head of articles, laying down preferred formats for dates, varieties of English, units of measurement etc., which could be in bot-accessible format. --Pete (talk) 18:47, 18 July 2008 (UTC)

(reset indent)Response to Tony I hope I hit all the points addressed to me.

MOSNUM no longer encourages autoformatting, but does it discourage it?

Where auto-formatting exists, shouldn't removal be discussed on an article-by-article basis?

Inconsistencies uncovered by removal of auto-formatting should be fixed, just as inconsistencies uncovered by editing should be fixed.

I wasn't suggesting that FAs or GAs are being held up, or that they should be held up. Poor way of expressing it on my end. Hypothetically, if an article was held up because it included auto-formatted dates, that would be a bad thing. However, suggestions in a GA or FA review are (I'm guessing here) less likely to be ignored on the basis of optionality. People are going to do as suggested in hopes of a Support !vote --Elliskev 20:00, 18 July 2008 (UTC)

Thanks for clarifying this, Elliskev; it certainly would be a bad thing. TONY (talk) 00:47, 19 July 2008 (UTC)
I have not observed this happening at FAC. —Mattisse (Talk) 22:49, 20 July 2008 (UTC)
For the record, after juggling this issue in my mind for the last week or so (and turning off date preference settings), I am in favor of delinking all dates. --Elliskev 16:59, 1 August 2008 (UTC)

When definitely not to link dates

It may be a toss-up in the case of most articles but I've just recently decreased the average frequency of photons emitted from a couple of pages. Certain articles contain date ranges of the form day1–day2 month year (albeit month day1–day2, year). Blue lemon can't deal with this. For autoformatting to work this has to be written day1 month–day2 month year. We're trying to write English here: we shouldn't have to twist it to fit some ill-conceived formatting feature. I propose that whenever such unautoformattable dates appear in an article, for consistency no dates should be autoformatted. JIMp talk·cont 02:27, 18 July 2008 (UTC)

In the interests of consistency, we should therefore dump all date autoformatting. The only way to deal with such autoformatted date ranges so that they appear correctly is to show both dates in full. This looks very awkward, and leads to novice editors "fixing" them for readability, and getting into a tangle of wikidates. --Pete (talk) 17:58, 18 July 2008 (UTC)
Pete, I agree entirely, and I know others do. Date ranges (January 3–9, 1998) and slashes (the air-raids of the night of May 21/22) are just two examples in which the autoformatting will not allow us to do what is stylistically neatest and standard practice; the same applies to expressions such as "February–April 2006". They represent an advantage in not using the system and a disadvantage in using it; editors need to weigh this up when choosing whether or not to use it. Yes, autoformatting in the main text of an article shouldn't be applied here and not there. None of this, of course, changes the optional nature of the system.
On that note, may I point out that I'm perplexed to see your changing of all of the raw formatting from US to the international format some six hours ago in the Beethoven article. Critically, this change will be seen by our readers, and it is probably only because the raw format is now again concealed from the WP "elite" by the autoformatting that no one has raised the issue. The section in MOSNUM that governs this is largely based on the highly successful WP:ENGVAR policy. These policies are in place to promote stability and minimise edit warring. I know it was a lot of work, but I wonder whether you'd like to reconsider this change, although there may have been a good reason for this that I'm unaware of. TONY (talk) 00:43, 19 July 2008 (UTC)
Regarding Beethoven, the use of day-month-year is the correct format. If Germany has moved away from International Dating format, then I am unaware of it. --Pete (talk) 01:10, 19 July 2008 (UTC)
Regardless of date format chosen, I don't understand why you autoformatted the links in your article Ludwig van Beethoven, given your statement about about dumping date autoformatting above. Since autoformatting covers up inconsistencies and screw ups from the eyes of wiki editors, it leaves the vast majority of unregistered readers, plus wiki editors without Preferences set, to see the true result, while providing everyone with a multitude of meaningless links to dates. —Mattisse (Talk) 23:28, 20 July 2008 (UTC)
You caught me at a moment of transition. The arguments against wikilinking dates are compelling and if I change a date, I now also remove the brackets. --Pete (talk) 23:41, 20 July 2008 (UTC)

A rank-and-file perspective

I've been editing on a varying scale for five years and with dozens of identities. Until recently, I'd never hewed to the MoS, simply because I felt some of the prescriptions were silly. Date-linking was probably my main complaint, and on its account alone I found it hard to take the Manual seriously. Instead, I'd pull up FAs, go through them, follow the precedents I liked, and ignore those I didn't. The MoS was to be disregarded; I never tracked changes or updates.

After all, what was the big idea? How did date-linking follow usual linking guidelines, wherein links are used to improve depth of understanding, explain references, and lead to related subjects? What does April 7 have to do with Henry Ford? He probably never gave a whit of thought to April 7 until he happened to croak on it. How much worth do random events occurring in 1828 have for readers of Francisco Goya? How relevant is it that Andrew Taylor Still, "the father of osteopathy", happened to be born in the year Goya happened to die? And why do I keep saying "happen"? Because dates are happenstance by nature, irrelevant to almost every topic. If we're just linking dates for diversity, to lead readers in new browsing directions, why should dates in particular get this special privilege? And if we just like to turn as much text as possible to that lovely blue hue, why not just make it the default text color on Wikipedia?

So anyway, after much avoidance and intermittent wikibreaks, I returned to editing on a small scale. I decided to play things by the book this time, and the book is what I turned to when I got to cleaning up our inchoate Eleanor McGovern article. Did I really have to link these stupid dates? Looking through the MoS and expecting the bad news, I didn't find it. I went to the Village Pump, asked about changes in datelinking policy, and got no answer. Eventually I wound up at Tony's talk page, and now I wind up here.

The policy needs to change. Optionalization is a fine start, and an absolute minimum, but the MoS should come out clearly and unambiguously against date-linking. It's not an IDONTLIKEIT concern; it's a matter of getting guidance from our guidelines. The MoS should not create ex juris enclaves of special linking policy where common consensus about overlinking does not apply. Sure, there is a tension between WP:OVERLINK and WP:BUILD, but even the more link-happy of these two makes clear that links should only be made to relevant pages, and April 7 has precious little relevance to anything except April 8. If we are honest to our general guidelines, the MoS must establish a strong recommendation against linking dates in all but the most unusual circumstances.

And one more word, on the "American" date style. I'm an American, though raised partly in the UK, and I don't think date styles are an issue at all, let alone analogous to differences in spellings or measurement systems. Just a few weeks ago I made sure that Pan-Slavic colours was moved back to its original American spelling, and if WP ever switched to an all-metric system, I'd go into conniptions. And yet, I wouldn't give a hoot if we switched over entirely to "7 April" over "April 7", and neither would most Yanks. Ask the average American on the street which style belongs to which place, and we wouldn't be able to tell you. The same American who flinches to hear "kilometres" would find nothing dubious or foreign about "7 April". Further, the "international" style is preferable simply because it lends itself less naturally to ordinal suffixes and other such frippery. Though I doubt it will ever happen, date formats could be standardized without loss or upset. But really, my main issue is date-linking, and we can spare the other debate for another time.

We can act now to establish a single clear and sensible date guideline which conforms to the spirit of our wider policies, and we should. Mr. IP (talk) 22:41, 20 July 2008 (UTC)

I was reading about the space program. What has this got to do about it?
  • Very well written Mr. IP. There are a number of us who agree with you. I for one, agree with everything you wrote above. I think MOSNUM does itself a disservice by even giving as much “air time” as it does to the whole issue of autoformatting of dates. Note how if we were to code [[2005-08-06]], the vast majority of readers see only  2005-08-06. Is that June 8, 2005 or August 6, 2005? It’s not intuitively clear until the 13th day of the month. Only we registered editors—a small, privileged minority—would see something attractive and unambiguous such as August 6, 2005 or 6 August 2005. And I too am an American but in general-interest articles, I use the Euro format as it is well recognized by Americans and is logical and “metropolitan” to boot. You’ll be pleased to hear that the authoring community on MOSNUM is waking up to this realization that user preference settings that benefit only registered editors are not proper solutions to anything. We are also getting better in resisting the efforts of those who add trivia to dates, like October 16, and labor to force the “advertisement” of their work with links in articles to that trivia. If I’m reading an article on the space program, I don’t need to click on an Easter egg link to find that on this day (October 16) in 1925, Angela Lansbury was born. In fact, there is now a bot removing the majority of these links from Wikipedia. Greg L (talk) 00:27, 21 July 2008 (UTC)
  • What bot is that? I am not sure this is accurate; bot's shouldn't be making optional style changes at any rate (for one reason since a bot going the other direction would be equally valid, and competing bots is not a good idea). Christopher Parham (talk) 02:04, 21 July 2008 (UTC)
  • Naturally enough, I agree wholeheartedly with IP and Greg L. Can I say that the mandatory removal of autoformatting would be heaven-sent, but is unwise until the community has a chance to realise that the sky doesn't fall in when you remove auto. Quite the opposite. Here's a capped list of the disadvantages of date-autoformatting, which people may wish to paste onto the talk page of an article that could benefit from the removal of autoformatting, along with a flagging of the intention to free the dates soon. Once the issues are explained, I think there'll be little or no objection.

____________________

Title: "Proposal to remove date-autoformatting"

Dear fellow contributors

MOSNUM no longer encourages date autoformatting, having evolved over the past year or so from the mandatory to the optional after much discussion there and elsewhere of the disadvantages of the system. Related to this, MOSNUM prescribes rules for the raw formatting, irrespective of whether a date is autoformatted or not). MOSLINK and CONTEXT are consistent with this.

There are at least six disadvantages in using date-autoformatting, which I've capped here:

Removal has generally been met with positive responses by editors. Does anyone object if I remove it from the main text in a few days on a trial basis? The original input formatting would be seen by all WPians, not just our millions of readers; it would be plain, unobtrusive text, which would give greater prominence to the high-value links.

____________________

One final matter: Parham, it is pure contrarian fantasy that someone would invent a script for imposing date-autoformatting in this contet. I'm unsure what motivates you this time to apparently object to a highly popular move; is it based on "I don't like it"? Your support would be greatly appreciated. Tony (talk) 02:30, 21 July 2008 (UTC)

I have no objection to the changes made so far. In general I always support changes that allow the editors actually building articles to use their best judgment about what is appropriate in a specific case. I don't think the benefits you've identified above are especially substantial, but the benefits of autoformatting are also limited and fairly trivial. I would oppose a return to mandating any particular style, however. As I also mentioned above, no bots should be making these changes at the moment, and to address your first point, it's quite possible the script already exists: at one point autoformatting was mandated, and I assure you that people will write scripts to address even the most blindingly trivial MOS mandates. Christopher Parham (talk) 02:47, 21 July 2008 (UTC)
That said, I think FA reviews in particular should be careful not to cast these changes as anything but optional, or to suggest that the MOS encourages them at this time. Christopher Parham (talk) 02:57, 21 July 2008 (UTC)
No one is proposing mandatory anything. If contributors like the idea of freeing the dates in an article, there is nothing wrong with using a script to avoid a copious manual task, sine the effect is identical to that of the equivalent manual task. Given the wish by most contributors not to dilute their high-value links and to have as smooth and clean a copy as possible, it's not surprising that the move is generally popular. Your comments might be more relevant if the move was not so popular. I agree that FA nominators should be made aware that the use or non-use of autoformatting is their choice, and that the review process excludes reference to that choice. Tony (talk) 03:09, 21 July 2008 (UTC)
The topic originator just recommended that the MOS discourage autoformatting and the next post was to agree with him, so obviously some people are leaning that direction. Perhaps any form of "mandate" is incorrect here, but I would prefer the choice remain with editors. I am fine with a script being used to remove autoformatting, so long as this isn't emploed to systematically remove autoformatting from large swaths of articles. Christopher Parham (talk) 03:23, 21 July 2008 (UTC)


Autoformatting dates and references

I might be missing this, but how does the current suggestion of removing autoformatting of dates within the article body affect dates for references? Are these encouraged to be not linked (which means that the accessdate parameter in most templates will need to be changed) with the same date formatting as the article, or are these considered outside the bounds of the article, thus linking as normal? --MASEM 14:00, 21 July 2008 (UTC)

Personally, I think that unambiguous dates such as "July 1, 2008" do not really need to be wikilinked, but the ISO-format "2008-07-01" dates used by the citation templates should still be wikilinked for clarity. —Remember the dot (talk) 01:46, 25 July 2008 (UTC)

Here are three articles, with dates completely delinked and correctly and consistently formatted, in three different styles, as samples:

  1. Using the cite xxx family of templates (cite web, cite book, cite news, etc.): Ima Hogg
  2. Using the {{citation}} template and international date style: Samuel Johnson
  3. Using manual citation method: Tourette syndrome

SandyGeorgia (Talk) 05:00, 25 July 2008 (UTC)


Wow, after reading and threading through the various discussions, I can see why the recent campaign to establish a new practice regarding wikidating has run into so many "bumps on the road." One of the problems was that the project groups have not been fully involved in either the deliberations or in the notification of the proposed changes. All of a sudden, a number of editors who have championed this change in the MoS direction have appeared, often without any previous history in editing an article and have summarily revised the text to reflect the new wikilink date style often to the consternation of the editors previously developing the article. I have now started to introduce the topic for a discussion at the various project groups in which I interact. FWiW Bzuk (talk) 16:23, 1 August 2008 (UTC).

I don't care much what decision is reached on the autoformatting issue, but one thing is perfectly clear: anyone mass removing these links without an established concensus should cease and desist, and there certainly shouldn't be any bots already doing this (as was indicated above). Was there a prior discussion/concensus regarding the change in the MOS? If so, can someone provide a link? Personally I like the idea of having all dates formatted consistantly to my preference. I had assumed that dates were autoformatted to a single default for non-registered users. Given what has been said above, I take it this is not the case? I assume this avenue has already been explored? PC78 (talk) 17:30, 1 August 2008 (UTC)

Have other option to deal with date autoformatting even been discuseed? I can think of a few:
  1. Education - the argumant that only small percentage of readers/editors used date autoformatting, and therefore we should not use them at all is SILLY. What if people had taken that attitude to cars, electricity, indoor plumbing, etc.? Why wasn't a program to educate users of the advantages to registering and setting preferences considered first?
  2. Change the date color - Why not make date autoformatting links a different color than blue? Green seems like a good option, and certian users here seem to like the color!
  3. Disabling date autoformatting - If there is no good side to date autoformatting, and there use is as bad as people are making it out to be, then why not propose disabling the date autoformatting feature in the Wiki software altogether? End of issue, as the date formats simply would not work. What sense is it to have preferences if no one uses them anyway, or the option is removed at the page level?
I'm sure there are more options available than just these. Why can't we discuss others ways to handle the whole issue? - BillCJ (talk) 18:46, 1 August 2008 (UTC)
  • PC78, establishing when a “consensus” has been reached on Wikipedia is darned hard to prove when there are some editors who vehemently oppose the “consensus view.” But it seems quite clear to me that the newly developed consensus is that the autoformatting of dates is—at least—not worth the fuss, and (probably), introduces more problems than it fixes. For one thing, the only people who really benefit from the “autoformatting” are registered editors who’ve taken the trouble to adjust their user preferences. In fact, some of the autoformatting options result in particularly crappy looking dates for most readers. The notion that Wikipedia is doing the world any good whatsoever by providing tools that the vast majority of readers (un-registered I.P. users) don’t benefit from has long been discredited. There is an increasing awareness in the Wikipedia authoring community that “features” that don’t benefit unregistered I.P. users aren’t proper solutions to anything and only make a privileged elite feel good about their handiwork. Further, autoformatting results in the linking of dates, which over-links articles because the links only take readers to random lists of trivia which have next to nothing to do with the subject matter of the article, and thus, the disadvantages of autoformatting outweighs its benefits.

    And when you wrote that “I had assumed that dates were autoformatted to a single default for non-registered users”, you assumed a bit too much with that assumption. Depending on the autoformat chosen, there are different defaults and some are particularly crappy looking—beyond worthless really—for most readers. Note how if we were to code [[2005-08-06]], the vast majority of readers see only  2005-08-06. Well, is that June 8, 2005 or August 6, 2005? It’s not intuitively clear until the 13th day of the month. Only we registered editors—a small, privileged minority—would see something attractive and unambiguous such as August 6, 2005 or 6 August 2005. But we registered editors are insulated and oblivious to this ugliness and can’t fix what we can’ see. The simple solution is to stop thinking of tools that only work for us as being of any value whatsoever.

    And BillCJ, your argument that that the consensus view (solutions that only benefit a small minority and actually messes things up for the vast majority) is “silly,” just proves the point that some registered editors on Wikipedia are only here to impress themselves and an elite club of registered editors. The rest of us simply reject that attitude; we know what Wikipedia’s proper mission is: to benefit the general reader just as much as our precious little selves. Greg L (talk) 19:05, 1 August 2008 (UTC)

(Edit conflict) Now I'm an "elitist" because I think everyone should be able to use the same tools I do? Wow! Thanks for your thoughtful considered response on the substance of my comments. Keep up the good work! - BillCJ (talk) 19:19, 1 August 2008 (UTC)
  • It was hard for me to get past the part where you wrote “Education - the argumant that only small percentage of readers/editors used date autoformatting, and therefore we should not use them at all is SILLY.” Greg L (talk) 19:25, 1 August 2008 (UTC)
    • Yeah, sorry, I won't use such big words next time, ok? - BillCJ (talk) 19:31, 1 August 2008 (UTC)
Greg L, I agree. I came into this a little over a week ago 100% in favor of linked dates and 100% in opposition to any active delinking. I figured It's a feature, not a bug. However, after reading the argument that you have restated—linking doesn't offer any benefit to the majority of reader—I decided to test it. I turned off date preferences. After a week or so of random article strolls, I have decided that that argument is not only sound, but compelling. The majority of readers are treated to articles with multiple date formats (yyyy-mm-dd being the most obnoxious). What's worse is that they inconsistency is glaringly obvious, since all the dates are bright blue. I hope a consensus is developed to scrap the practice. --Elliskev 19:17, 1 August 2008 (UTC)
  • Elliskev, the consensus is to scrap it. It’s just that it’s been discussed for a long time and some of the threads have been archived. So new editors, wondering “where the &%*!@ my autoformatting go?” are coming here all brand new to the discussion. Like you, they should stay unregistered for a week or two; that”ll open their eyes to how it was a brain-damaged notion to have even considered making tools in the first place that would only benefit registered editors. Greg L (talk) 19:25, 1 August 2008 (UTC)
Just to be clear, the consensus is that the autodate format is "optional" not to scrap it or have I misread the MoS? FWiW Bzuk (talk) 19:36, 1 August 2008 (UTC).
That was my understanding, as well. How likely does it seem that we could develop a consensus that actually takes a stand, one way or the other? --Elliskev 19:39, 1 August 2008 (UTC)
Agreed - there is no consensus to scrap it, despite certain opinions expressed here. Further, the whole "elitist" argument is inappropriate (and somewhat offensive), as there is nothing "elite" about registering for Wikipedia. Suggesting otherwise only serves to distract from the issue at hand. --Ckatzchatspy 19:43, 1 August 2008 (UTC)
No, Bzuk, you misunderstood the meaning of "optional". Only the "brain-damaged" or "elitist" would think that "optional" means we actually have a choice! Why are we wasting are time here, BillZ? - BillCJ (talk) 19:44, 1 August 2008 (UTC)

(ec) My understanding is that it is finally an option to scrap autoformatting in an article I am writing if I want to, which I do. That is, autoformatting is no longer forced on me. If I want to consider the vast majority of readers who are the unregistered public, I am now allowed to do so. —Mattisse (Talk) 19:52, 1 August 2008 (UTC)

(ec)If it's to remain optional, there needs to be some leeway given to editors who make a good-faith effort to fix some of the articles that are inconsistent in date format. If I come across an article that has a combination of inked and unlinked dates of various formats, I'd be inclined to choose a format and unlink everything. I'm pretty sure that eventually somebody would accuse me of unilaterally pushing my preference. --Elliskev 19:54, 1 August 2008 (UTC)
The opposite behavior, i.e. demanding autoformatting, ideally would be considered unilaterally pushing a preference also, in a perfect world. —Mattisse (Talk) 20:00, 1 August 2008 (UTC)
Just to be clear, see the top of WP:MOSNUM. You'll note what it says about it being a guideline (as opposed to a policy). The means it's advisory, and so is optional. Ultimately you can do the sensible thing and mostly other editors will go along for the ride. If you do something silly, they'll outnumber and outlast you and the sensible thing still eventually happens.LeadSongDog (talk) 20:15, 1 August 2008 (UTC)
Yes, but. Guidelines do carry a lot of weight. I may be outnumbered, but if I'm backed up by a guideline, then my !vote counts for more - usually. --Elliskev 20:37, 1 August 2008 (UTC)

GregL: What this issue really needs is a centralized discussion, where a binding concensus can (hopefully) be forged. It's all too easy for you to see a concensus where you want to, but it won't necessarily be so obvious to everyone else. What I don't want to see is a repeat of the debacle with placeholder images in biography articles, where a group of editors plastered them all over the place and a concensus was later formed against their use.

Is there any reason why a single default can't be imposed, as opposed to the current "no preference"? PC78 (talk) 21:26, 1 August 2008 (UTC)

Yes. Currently 99% of the readers of Wikipedia are not registered. There is no way to "impose" a preference on them. For those few readers who are both registered and who have set a preference, you would have to mandate a date preference that seems to have political implications on Wikipedia: what date format would you mandate?
What an editor finds out, if that editor removes Preferences and therefore the "formatting" of autoformatting, is that any given article can be a hodge-podge of date formats. That is what 99% of readers of Wikipedia see. —Mattisse (Talk) 21:40, 1 August 2008 (UTC)
Your figures on type of readership may well be correct but begins to muddy the discussion here. I believe that the consensus has not yet been reached other than an acceptance of the optional format for wikidate linking. One of the major concerns expressed was that there were already a number of editors who acted as if policy had been determined and made large-scale changes to existing articles, adding much to the confusion and uncertainty that resulted. FWiW Bzuk (talk) 22:11, 1 August 2008 (UTC).
I'm rather puzzled that it apparently requires another elusive wikipedia "consensus" to fix a problem that readers have always seen. Autoformatting was always a dreadful idea, at best a half-baked solution to a problem only half-understood. --Malleus Fatuorum (talk) 22:19, 1 August 2008 (UTC)
Great, issue solved. "I don't have a dog in the race" but if there was an agreement on a consensus which some other editors have also alluded to, point the way and let's move on. FWiW Bzuk (talk) 22:24, 1 August 2008 (UTC).
Common sense ought to trump consensus any day of the week, don't you think? Does autoformatting cause a jarring reading experience or not? It's not rocket science, and neither does it warrant all this pallaver to fix what was was obviously a poor choice of work-around made in the past. --Malleus Fatuorum (talk) 22:32, 1 August 2008 (UTC)
(2nd edit conflict) One man's "common sense" is not necessarily anothers. When you want to change something that is long established and will affect a significant proportion of articles, then yes, you really do need to establish a concensus first. I feel it necessary to reiterate what I said above; people shouldn't be mass removing these links when such action is not supported by concensus, guidelines or policy. PC78 (talk) 22:47, 1 August 2008 (UTC)
Who are "these people" who are "mass removing these links"? Why the hyperbole? Why not look at what's actually been said and done, instead of what the less well-developed parts of your brain are making you feel has been said and done? --Malleus Fatuorum (talk) 22:51, 1 August 2008 (UTC)
Hey, try and avoid the needless insults, OK? It does not contribute to a proper discussion. As for your question, why even ask? A simple look at Special:Contributions/Tony1 demonstrates the point. --Ckatzchatspy 22:54, 1 August 2008 (UTC)
Exactly. Apparently it's not me who needs to "look at what's actually been said and done". PC78 (talk) 23:07, 1 August 2008 (UTC)
Regardless of all of this "pallaver", there is no consensus to eliminate the format of wikiautodating, there is however, a perceived consensus to have the date format as an optional practise, unless I have misunderstood the direction of this "string." FWiW Bzuk (talk) 22:38, 1 August 2008 (UTC).
I believe that you have missed the important point. Commonsense trumps consensus. Sadly though I should probably have said "uncommonsense", given the deep misunderstanding that appears to surround this issue. --Malleus Fatuorum (talk) 22:45, 1 August 2008 (UTC)
From where do you get this notion? As I indicated above, "common sense" is not universal. Nor is it a guideline or policy. Concensus, on the other hand, is. PC78 (talk) 22:49, 1 August 2008 (UTC)
I will leave this discussion now, as I'm reminded that there is no greater fool than one who argues with a fool. --Malleus Fatuorum (talk) 22:54, 1 August 2008 (UTC)
My sentiments exactly. Oh, and welcome to the real world, where not everybody shares your opinions. PC78 (talk) 23:07, 1 August 2008 (UTC)

My personal view is that as long as the dates are displayed correctly for the region to which the article relates or for which a consensus exists within the article or topic, then there's no problem. It does create less of a "blue sea" effect. I've always been totally against wikilinking years (other than those that are part of dates) as it serves no purpose at all, but with the dates the issue had always been what I perceive as a US-centrism that sometimes operates and spreads when left unchecked. (i.e. if people in the US do something one way and people elsewhere do it differently, the latter is seen as "wrong".) Agreed with PC78 that the best way to effect change management is to involve people rather than force it from above. Orderinchaos 19:38, 2 August 2008 (UTC)

I'm convinced. I just turned off date preferences in my preferences. I've been removing bare year links on sight for a long time now, and I now intend to removed date autoformatting when I see it. -- Donald Albury 20:47, 2 August 2008 (UTC)

  • Orderinchaos, and that’s the trouble with the current implementation of autoformatting of dates. Rather than take a look at the requesting reader’s I.P. address and figuring out what country they hail from (something that is routinely done on Web servers), Wikipedia did an unwise foray into making formatting tools that only benefit registered editors and actually often screw things up for the majority of readers (I.P. users). How Wikipedia thought that was a good idea is beyond me. Editors have to be intellectually fair here and get beyond the fact that they like what they see and really make the effort to see just how junked up Wikipedia becomes for the vast majority of readers. Greg L (talk) 20:52, 2 August 2008 (UTC)
  • P.S. What would solve all of this is if Wikipedia developers made a parser function that looked to the requesting reader’s I.P. address so it knew what country they are from. This is routinely done on Web servers of all sorts so they can spoon-feed custom content to the reader and collect reader statistics. There could be certain parser function classifications and groupings of these classifications, such as “US/other” or “US/Commonwealth”, “NorthAmerica/Other”, “US/UK”. Then we could have templates like {{US/Commonwealth|trunk|boot}} so text would read “…he put the bomb in the trunk of the car…” for US readers and “…he put the bomb in the boot of the car…” for readers from the UK and Australia. Similarly, editors could write {{US/Commonwealth|color|colour}}. For dates, (although I personally don’t have a problem looking at 2 August 2008), one could write {{US/other|July 4, 1776|4 July 1776}}.

    But we don’t have these tools, which would be designed to improve the reading experience for the vast majority of Wikipedia’s readers: the non-registered I.P. user. The autoformatting of dates was sooooo unwise because they allowed us to start using crappy-looking code for regular users to look at just so we privileged few could benefit. 21:20, 2 August 2008 (UTC)

  • If we had some sort of parser for date formats/spellings/units of measurement, then we could present articles in a more internalionalised manner, more accessible to people, regardless of where they live. I suspect that Wikipedia users in (say) China, would regard the whole thing as a Yankee wank.

    However, making complex editting templates within the text is not going to make joining in easier or attractive to new editors. Putting each date, each unit of measurement, each boot/trunk/hood/bonnet, each or/our or ise/ize ending inside brackets and slashes is a prospect to fill me with despair. --Pete (talk) 00:22, 3 August 2008 (UTC)

  • Yeah, but the beauty is, no editor would *have* to use these templates. I’ve been the “first major contributor” to a number of articles and I use “color” (not colour), “liter” (not litre), and “kilogram” (not kilogramme). I know full well that this drives UK readers nucking futs (and the reverse for me at times). We editors (and new editors) would be free to edit without the use of the templates simply by following all the current MOS and MOSNUM guidelines. It would primarily be those editors who are running around currently trying to change “kilogram” to “kilogramme” who would avail themselves of these tools. These editors would do all the work for us and would be pleased to do it too! Cool, huh? Greg L (talk) 02:51, 3 August 2008 (UTC)
I don’t mind dropping the wikilinking of dates at all, since it’s a lot of work for the benefit of very few readers (see Greg L’s note to Orderinchaos a few posts above this) and it cuts down on the “sea of blue.” It’s better to have all the dates in a common format in any given article. What I would prefer to see eschewed is the use of purely numerical dates (e.g., 2008-08-02) where one cannot discern whether the American or European approach is being used, unless the date happens to be later than the twelfth of the month. Askari Mark (Talk) 00:43, 3 August 2008 (UTC)
Now that you have brought up the ISO dating, it is a complete mystery to many users as to what the day-month-year connection is. Is it August 2, 2008 or 8 Februry 2008? If you are advocating a standard, then use a written out date, regardless of the country of origin. All en-wiki users would understand August 2, 2008 or 2 August 2008. As to "poking sticks" into editor's eyes by firmly refusing to use country-specific spelling conventions? What a revelation, no wonder some editors are perplexed when a US-centric approach like the one above is being advocated. FWiW Bzuk (talk) 04:11, 3 August 2008 (UTC).
  • Agreed, we need to spell out the name of the month for greatest clarity. If it’s an article on a general science issue that isn’t attached to any specific country, I just use 2 September 2007. If it’s an article, on a subject about or of particular interest to readers in the US, I use September 2, 2007. And in neither case, do I link to silly lists of trivia that have absolutely nothing whatsoever to do with the subject. I’ve been fairly successful with this approach, only rarely having other editors modify what I wrote—and even then—all they’ve done is add the dumb links. Since I subscribe to the theory that articles that look like they’ve been attacked with a spray can of blue paint numbs the reader to links and inhibits exploration and learning, I just de-link the damned things. Greg L (talk) 05:14, 3 August 2008 (UTC)
I should note every single time I've ever keyed in a date, even though I've autolinked, I've always used the most appropriate for the region or the article (if there's a lot of dates already there I stay consistent). So anything I've ever written will look exactly as it should if it's unlinked - I have no doubt the same would be true for most experienced editors. Orderinchaos 06:14, 3 August 2008 (UTC)
IP based geolocation and rendering is clearly the right solution for the unregistered reader. Unambiguous, maintainable wikitext is clearly needed, whether that be ISO, fully spelled, or some other date form in the wikitext. What is rendered should be what the reader will understand, not what the editor likes. And variation of rendering between articles is a lame excuse for a standard. So rather than waste all this energy on perpetual, cyclical debate over work-arounds, why not focus attention on the real question: How do we pursuade the developers to fix the implementation of date rendering at some time in the proximate future? LeadSongDog (talk) 06:25, 3 August 2008 (UTC)
  • LeadSongDog, someone has to step up to the plate and write a buzilla. But not much will happen, I don’t think, unless one is socially active on the live chat relay whetever-they-call-it. But I could see geolocating as being damned popular for editors who hate looking at non-local UK/US spelling variations. Greg L (talk) 06:05, 4 August 2008 (UTC)
  • There is one already open (I can't access it from here to give you the number) with a very long comment history. There is reluctance amongst the developers to use IP-based mechanisms for some reason having to do with the server performance cost of maintaining cache coherency on an enormous table of IP addresses. I don't know if this is really a tough nut to crack, but it seems to me that they think so. I would have thought that they only need current session IP addresses, or perhaps even a session cookie could enable it, but I'm far from current in my programming skills. LeadSongDog (talk) 15:50, 5 August 2008 (UTC)
Please not ISO dating, that just opens up another issue. FWiW Bzuk (talk) 20:09, 3 August 2008 (UTC).
You may be missing the point. The reader should not see the format in the wikitext. It should always be appropriately rendered in a reader-friendly format. Any unambiguous wikitext should work correctly. Why should this scare anyone?LeadSongDog (talk) 02:48, 4 August 2008 (UTC)

On a pedantic note, Greg said that the spelling 'kilogram' instead of 'kilogramme' annoys UK people. The spelling 'kilogramme' may exist in the UK but I think the spelling 'kilogram' is the current dominant spelling in British English. It is the spelling used in British law such as The Units of Measurement Regulations. This pedantry does not detract from his point. Regards Lightmouse (talk) 21:19, 3 August 2008 (UTC)

I was reading about the speed of light. What has this got to do with it?
  • Orderinchaos, regarding your 06:14, 3 August 2008 post, that would be the way I would do it too (if I actually wanted to link to dates—which I don’t). But Wikipedia’s date formating options includes a total abortion that should never have been dreamed of, let alone included as a tool for editors to use. Note how if we were to code [[2005-08-06]], we (the minority privileged registered editors) would see either August 6, 2005 or 6 August 2005. But the vast majority of readers see only  2005-08-06. How ambiguous is that? Is that June 8, 2005 or August 6, 2005? It’s not intuitively clear until the 13th day of the month. The proper thing to do is simply stop using date formatting tools that don’t benefit everyone equally.

    And 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…

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 October 21st, they can type it into the search field. When we over‑link, we just turn articles into a giant blue turd. Greg L (talk) 05:57, 4 August 2008 (UTC)
I'm not going to defend date autoformatting, but there is no ambiguity in the date 2005-08-06. It means 6 August 2005. That is the beauty of ISO dates - they have only one possible interpretation.Thunderbird2 (talk) 16:15, 5 August 2008 (UTC)
ISO dating does not have a universal understanding despite your ardent affirmations in it. At WP:AVIATION PROJECT, the issue was introduced years ago and a Polish editor indicated that the ISO dating was problematic. After checking, it was confirmed that not only did many editors use a garbled version of ISO dating but that foreign visitors often were completely mystified as to the date. If we are not going to use autoformatting for dates, the recommendation made in the project group and should apply to all other situations is that a clearly written out date is preferable. FWiW Bzuk (talk) 16:22, 5 August 2008 (UTC).
Yes, I agree that confusion can arise through lack of familiarity. For that reason ISO dates, where used at all, and despite the clear benefit of unambiguity, should be treated with caution. Thunderbird2 (talk) 16:31, 5 August 2008 (UTC)

ISO dates are a standard format on Wikipedia. Just look at the main article on the main page: Ann Arbor, Michigan or any other with lots of references. The autoformatting mess is because well-meaning people combined two things (create more links, create autoformatting). Let's not combine a migration away from autoformatting with a prohibition on ISO dates - just think of the effect that will have on delinking thousand of articles like Ann Arbor, Michigan i.e. you will have to examine the article for its dominant region, and edit all the ISO dates to become US dates, only then will you be able to remove all the links. Lightmouse (talk) 17:10, 5 August 2008 (UTC)

No one is suggesting removing ISO dating but in starting a new article, a recommendation would be to have dates given in a logical, readable format that is universally understood. ISO dating IMHO, doesn't do that. FWiW, the first templates that were established "fixed" ISO dating. Those editors that used "scratch" cataloging, eschewed the format. Bzuk (talk) 17:26, 5 August 2008 (UTC).
In an ideal implementation, of course editors should be able to write in whatever syntax they are comfortable with. Bots should transform the wikitext to a single, consistent, unambiguous format (extended ISO 8601 or full international or full US format would all work). The consistent format in the wikitext is then readily rendered to html in accordance with user prefs or (for anon readers) IP geo-coding default prefs. LeadSongDog (talk) 17:57, 5 August 2008 (UTC)
Bots, like all computers and computer programs, are not yet sufficiently advanced to be called stupid. There will always be situations where a shorter format is better, and a full format is better. There will always be situations where the ISO format will be better because it is sortable with readily available software, such as spreadsheets. There are situations where text appears to be a date but is not, and could be incorrectly altered by a bot. ("March 12 miles" ordered the captain.) --Gerry Ashton (talk) 18:52, 5 August 2008 (UTC)
Surely you can find a better counter example. MOSNUM calls for "March twelve miles", doesn't it? In any case "March twelve" is not a date, it's a date fragment or day-of-the-year. But I take your point. The thing is, the only way to avoid that confusion entirely is to use some sort of markup either for dates or for date-like-strings-that-are-not-dates. Once you choose to do that, all options are on the table.LeadSongDog (talk) 19:10, 5 August 2008 (UTC)
(ec)This may need its own centralized discussion... Why in the world would we want to endorse keeping ISO dating? Under what circumstances is ISO dating compatible with English prose? --Elliskev 18:00, 5 August 2008 (UTC)
(ec)The obvious reason for keeping ISO dates in the wikitext is that it is the same in all languages. Remember that most wikipedia articles today are in languages other than English. In citations, especially, this is a huge labour saver. That says absolutely nothing about what is rendered into English prose. That should be whateveer is selected in user prefs (or the equivalent default for AnonIP readers).LeadSongDog (talk) 18:57, 5 August 2008 (UTC)
That makes sense. Do you foresee mandating ISO dates in wikitext? If so, do you foresee it as being enforceable? I expect that most editors will never adopt ISO date formatting. --Elliskev 19:12, 5 August 2008 (UTC)
Enforceable? On WP? Excuse me while I pick myself up off the floor! No, it's just a tool. But if the tools were to work correctly I think we would in fact see majority adoption. After all, who wants to type "19 Sēremōnaþ 2008" (anglo-saxon) instead of "2008-06-19"? LeadSongDog (talk) 19:39, 5 August 2008 (UTC)
OK. I could have picked a better word than "enforceable"... You know what I meant. I'm just asking if planning some fancy tool (that may or may not ever become available) around the expectation that people will adopt ISO format dating isn't going to be a big waste of time. It seems like it would be a lot easier to just recommend that dates in a given article be formatted consistently in a format that makes sense for the article. As for the labor-intensity, Wikipedia doesn't have a shortage of editors. --Elliskev 19:56, 5 August 2008 (UTC)
"I, the Great Karnack, foresee an argument in this place until the future becomes the past." - but frankly, as long as ther is a choice to be made, people will fight over which choice. Correct implementation moots the question, making all unambiguous formats equally valid for editor entries, while giving the readers the format they preferred. Most major software products have already gone down this road. See your operating system's "Control Panel" for instance. Simply put, it's wrong for implementations to push a choice on the user. Implementations should adapt to the user instead.LeadSongDog (talk) 21:40, 5 August 2008 (UTC)
I agree. With everything you said. I still think the idea that looking to ISO format dating as a solution is goofy. If the world was populated entirely of computer programmers and data miners, I might think otherwise, but it's not. Anyway... I call mootness on the whole question. Once we have an ass-kicking, mind-reading, date-formatting tool that can crawl through the encyclopedia transforming every occurrence of Christmas 2008 to the much more understandable 2008-12-25, I might come around. Until then, I'd be content to see articles with consistent date formatting (while reserving my right to leave my date preference set to "No preference"). --Elliskev 22:26, 5 August 2008 (UTC)
  • T-bird, regarding your above 6:15, 5 August 2008 post: “That is the beauty of ISO dates - they have only one possible interpretation”. Wrong. There is only one way they are “supposed” to be interpreted (according to the standard), but there are two ways to interpret them. The problem with the all‑numeric ISO date format (like 2005-08-06) is the specification defines the proper parsing order (yyyy‑mm‑dd) and says it is not yyyy‑dd‑mm.

    Do you think most readers are familiar with the ISO specification? I can point to a hundred people I know who’ve never heard of the ISO, let alone read their specification. Or do you hope that if Wikipedia were to start using the ISO‑style, all‑numeric date, the rest of the world will go “Ah HAAA (those smart Wikipedia editors), I see” and the world will soon intuitively understand the style and instantly recognize and parse such dates? The ISO specification was valuable for establishing a standardized way for the storage and retrieval of computerized, numeric-only dates in databases and spreadsheets and what not. But the result is still slow to read and takes more mental energy to parse than simply writing 2 February 2008 or February 2, 2008. There clearly was never any need for some standards organization somewhere on this planet to “define” how dates with the month written out (like 2 February 2008) should be parsed and interpreted because there is no alternative way to interpret them. What’s good for machines isn’t necessarily good for humans.

    So, to paraphrase your fallacious statement: “That is the beauty of dates with the month written out - they have only one possible interpretation”. Just because there is a “standard” of some sort for the computerized storage and retrieval of data and similar purposes, doesn’t mean it’s a good convention to use when writing dates in the body text of encyclopedic articles. That much is just too obvious. Greg L (talk) 22:58, 5 August 2008 (UTC)

There are further problems with ISO dates. Because ISO publications are not freely licensed and substantial charges must be paid to access them, many otherwise careful writers will not access the publications to resolve subtle issues, such as
  • If a time is given with an ISO date/time, is there any presumption about the time scale (UT1, UTC, EDT)?
  • If a date is in the period when the Gregorian calendar was adopted in some countries but not others, is there any presumption about which calendar was in force?
  • Are years BC allowed? If so, how are they formatted? --Gerry Ashton (talk) 23:37, 5 August 2008 (UTC)
Looking at the history of this talk page shows me the timestamp on Gerry's post as "2008-08-05T19:37:50" because I'm in Eastern Daylight Time and I've told the WP servers that in my prefs. The general 8601 representation of that would be "2008-08-05T19:37:50(-4) but MediaWiki doesn't bother to tell me where I am every time I look at a time. Instead it tells me elsewhere that it's doing something other than the basic implementation, technically remaining conformant by saying so. If I had not set a time zone, MediaWiki would have used UTC, or "2008-08-05T19:37:50Z". The Gregorian calendar is used unless otherwise specified, even for dates before it was introduced. Where specified, extra digits can be prefixed to represent dates further off, using a + or - prefix, as in "-011500" for the year from 11501 to 11500 BCE. But of course none of this matters to editors or readers of articles. They would continue to work in familiar representation, perhaps editing with new markup, i.e. {{date|7|August|2008}} LeadSongDog (talk) 04:46, 6 August 2008 (UTC)
Absolutely support Tony1. Turn off your auto-date prefs and and you'll see very soon that this is a no brainer. Date-linking was always an illogical and insular practice. Undo, wiki-wide, at last.
On the subject of ISO dating, my not-very-considered and perhaps ignorant view is that it should be avoided. It plays on the ambiguity non-US-date-convention-using people such as myself feel whenever they see a xx-xx-xxxx date format on the web, then further confuses by reversing that format. If you don't know ISOs, you're fucked. Also it's ugly, and when used, necessitates being linked for clarity, which does nothing for its looks and reintroduces confusion re dates usage and linkage.
My main support above all else is for the de-linking and internal consistency of dates in article text proper, which should virtually never be ISO in the first place. 86.44.31.35 (talk) 12:14, 6 August 2008 (UTC)
  • Isn’t there a clearly developing consensus here for the following guideline:
Since links to lists of historical trivia are rarely germane to the subject matter of the main article, the linking of dates and years (e.g. April 2, 1978) is discouraged unless the article is especially historical in nature. So long as the autoformating of dates results in linked dates, autoformated dates are discouraged. The all-numeric, autoformat option that is coded as follows [[2005-08-06]] and which displays as 2005-08-06 for non-registered I.P. users (the vast majority of readers) should not be used.
How say ye? Greg L (talk) 02:53, 8 August 2008 (UTC)
Now how do we change cite templates to read written out dates? FWiW Bzuk (talk) 02:56, 8 August 2008 (UTC).


Autoformatting dates and references (2)

  • Tony1's response to Bzuk's comment above: It's not a quick fix. We asked for date-autoformatting (DA) to be made optional (on/off) feature at the cite-web template, but despite the goodwill of the regular admins, attempts to tweak the program have run into technical and programming issues, and are still being worked through (see the previous link). Yes, it would give me frissons of unbridled pleasure to see DA in a huge number of citation and infobox templates suddenly deprecated and immediately addressed; but I'm concerned about the sudden deprecation at MOSNUM of an issue that is technically complicated, unlike the removal of DA from the main text, which is simple and easy to achieve. (The evolution of optional DA in the main text—as opposed to addressing its built-in compulsion in templates—has been a different matter: it has required no mandatory "retrofitting" and has merely given more power to users to do what they think best in the main text.)
  • Responses to Greg L's suggested text: Greg, I applaud the bold move, but I have a few issues in the proposed text:
  • Linking vs. autoformatting: Let's not confuse these concepts, which could be head-spinning for newbies and many other WPians.
  • Historical articles: In an article on the Roman Empire, the link to the month and day in the DA, "September 29, 256", is utterly trivial, but if folk really insist on keeping a link (not a DA) to an early year, it would be better for them to be straightforward about it: "September 29, 256". Link to an early year if you must, don't DA—it's as simple as that; why complicate matters by even mentioning the historical issue? (I grudgingly accept that early-year pages might occasionally be slightly relevant simply because they contain little information and thus might provide a bird's-eye view of the world at hand; I wince, though, at the narrow cultural base of those pages.)
  • "So long as DA results in linked dates, autoformated dates are discouraged.": This appears to cast the move away from date-autoformatting in a weaker, temporary light, and focuses only on the ugly, disruptive visual appearance of the system. However, the entanglement with linking is only one of six reasons I can think of to drop DA. Above all, it's the utter triviality of the difference between month–day and day–month that I believe should not be "addressed" because it's not a problem in the first place (like "colour/or").
  • ISO dates: My personal preference is to deprecate ISO dates, but I've picked up during this month-long debate that some people have strong reasons for using them, particularly in citation templates. I lack the expertise to make a definitive judgement in that respect.
  • Summary: Therefore, if there were consensus for moving from optional to deprecatory at this (early) stage, I'd be more inclined to go for something simple like this:

The autoformatting of dates is generally discouraged in the main text of articles. In particular, the all-numeric, autoformat option ([[2005-08-06]], which displays as 2005-08-06 for the vast majority of readers, i.e., non-registered IP users) should not be used.

  • But I think we first need to gain greater acceptance of the freedom not to DA in the main text ("hey, it's nicer and the sky doesn't fall in"). I suggest that DA continue to be optional for the time being, and that the proposal above to allow plain dates in the main text alone be acted on in MOSNUM soon, without getting tangled up with the complexities of the templates issue. Allowing main-text-only consistency in raw date formatting in an article is a critical enabling factor along the way to persuading the community that templates should be technically updated along these lines. That way, templates can be encouraged to evolve along with the changed approach to main-text DA. To bite off everything at once would be an impossible task. Tony (talk) 02:59, 9 August 2008 (UTC)
  • While I like Tony’s simple rendering, I lean more to Greg L’s proposed approach because it explains the “why” of the rule, and thus appears less like a diktat from the “MOS cabal”. After all, I think we most all agree that most editors tend to support discouragement of autodate linking once their eyes are opened to the problems that it poses. I would like, though, to propose a more concise rendering that combines the best of both proposals:
The autoformatting of dates is generally discouraged in the main text of articles because the great majority of users cannot take advantage of it. This includes all non-registered users, as well as those registered editors who do not select a date preference. In particular, the all-numeric, autoformat option ([[2005-08-06]]) should not be used, because it displays as 2005-08-06 which leaves the identity of the month and day unclear. (Note, though, that certain templates do require this ISO yyyy-mm-dd format for proper functioning.)
Is this closer to what we’re looking for? Askari Mark (Talk) 01:05, 10 August 2008 (UTC)

I think there is, at least by implicaton, a loss of information when old dates are wikilinked and the original date is written in YYYY-MM-DD format. People often associate this format with ISO-8601, and this MoS (dates and numbers) says "ISO 8601 dates (1976-05-31) are uncommon in English prose" thus establishing that in the absence of information to the contrary, when a date in this format is encountered in Wikipedia, it is an ISO-8601 format. As such, it is in the Gregorian calender (proleptic if need be). When a user preference transforms the date into some other style, information about which calender is used is lost. Granted, it would be unwise for the editor writing an article to depend on the mere use of ISO-8601 dating to communicate that the proleptic Gregorian calendar is in use, but nevertheless information can be lost. --Gerry Ashton (talk) 01:30, 10 August 2008 (UTC)

Further question(s)

  1. Where was this discussed? I see several people above saying it came out of the blue, and I don't see any other discussion ora pointer to where one is. Help?
  2. Surely someone has pointed out that the blue link under dates can be changed with a simple css class?
  3. Autoformatting may only help wpians who are logged in and have set their prefs, but it doesn't hurt any readers (other than the blue links - see previous point)

I'm just waay confused where this came from, where was the discussion, and obviously there hasn't been consensus. The autocratic change to the MOS seems totally out of line with "the way we work". -- SatyrTN (talk / contribs) 02:15, 9 August 2008 (UTC)

I've corresponded with Satyr, who says s/he is fine with optional DA, although personally prefers it. Tony (talk) 11:57, 9 August 2008 (UTC)
I'd like to add my "response" to Tony1's disadvantage list (which can be found in the "hidden" grey bar on Wikipedia talk:Manual of Style (dates and numbers)#A rank-and-file perspective):
  1. Auto-formatting does no harm, even if it seems to only do "good" for a few.
  2. This is incorrect - it's definitely not trivial. Date formatting is an issue that's been around for years and years - and obviously there are proponents and opponents of each format. If it were trivial, something would have been worked out one way or another - the fact that it hasn't speaks loudly.
  3. This issue can be solved with a simple css formatting - get the code guys to add a <div> around auto-formatted dates, and anyone can get rid of the blue underline.
  4. "Errors" in wikilinking is an issue for all wiki-markup - tables are an example that is *tons* more difficult than dates, and somehow we survive. Yes, there are errors. There are errors in applying categories, but we still do it.
  5. "Edit mode clutter" is a specious argument. Citations can be far more cluttering in edit mode than adding brackets around a couple pieces of text.
  6. These "limited applications" may indeed be a valid issue.


Customary units and gallons

I have a pair of ideas for resolving the U.S. v. US (or should that be U.S. v US) issue for the page. Caerwine Caer’s whines 03:32, 28 July 2008 (UTC)

I've always been in the "U.S." camp, but if the page is consistently written with "US", we should leave it be. —MJCdetroit (yak) 03:51, 28 July 2008 (UTC)
This seems a solution to a non-problem ... and a solution that'll only work if there are no other "US"/"U.S."s on the page. JIMp talk·cont 04:10, 28 July 2008 (UTC)
Reducing the number of points of potential disagreement, even if it fails to be perfect is a good thing. The page (as I found it when I noticed the problem) wasn't consistent. Caerwine Caer’s whines 05:07, 28 July 2008 (UTC)

Are we not, though, adding another point of potential disagreement with respect to US-centricism? The US customary system is certianly not my customary system. JIMp talk·cont 05:26, 28 July 2008 (UTC)

Caerwine, there wasn't a problem until you lurched in and made it a problem. I don't understand the utter inflexibility in your campaign to disallow "US", which is increasingly on the go in the US itself. Optionalisation here was pursued with rigour by SMcCandlish, himself an American. This is senseless. Apart from the stand-alone issues, the dots look very awkward next to "UK" and "USAF" and the countless other "US ..." acronyms that are officially not dotted. But no one is about to de-dot articles that have the dots in "US"; that is frowned on unless consensus is first gathered. We have more important things on our plate, don't we? Tony (talk) 05:57, 28 July 2008 (UTC)
The "U.S." is an initialism, not an acronym. That "US" is sometimes used as a symbol or to build symbols or acronyms is neither here nor there. We are writing encyclopedia articles here, not headlines or text messages where extra characters are inherently bad. I realize that a number of people are lazy have a style preference, and so they want to eliminate periods from abbreviations. However, others such as myself, still find them desirable, when abbreviations are needed. When there is no need to abbreviate, then we shouldn't and thus avoid problems with both sides of the period controversy. Of course, what we really need are separate Wikis for the American and British languages, but the powers that be think that is a bad thing for some reason. Caerwine Caer’s whines 07:02, 28 July 2008 (UTC)
  • I agree with everything MJCdetroit, Jimp, and Tony have said here. There is no problem. The only problems that can occur is if we attempt to upset the apple cart here and change editors’ practices. Wikipedia is not like some magazine or professional print encyclopedia where there is one chief editor at the top who sets a manual of style and all the copy editors toe the line. And as a result, Wikipedia will never have the project-wide consistency those publications enjoy. What we can do is establish guidelines to address practices that unnecessarily cause confusion. Both “U.S. gallon” and “US gallon” confuse no one. Greg L (talk) 14:24, 30 July 2008 (UTC)

It's nothing to do with laziness; that argument was run a year ago and laughed at. Typing dots is no problem for me: it's the reading of them that I don't like, and which makes the initialism look clunky. I hope that this campaign you've been conducting will cease. Tony (talk) 16:01, 30 July 2008 (UTC)

That's one problem with the written language, it just doesn't convey humor (which I thought my use of a strike through would adequately imply) as well as the spoken does. I was in no way implying laziness as the sole or even primary motivation of the style preference against using periods in abbreviations and alphabetisms. Nor is laziness necessarily bad, as it is one of the reasons why mathematical notation is used instead of writing out formulae in words. However, while not as much of a problem for the U.S. as it is a well known alphabetism, with alphabetisms that may not already be known to the reader, not using periods leaves the impression that it is supposed to be pronounced as if it were a word. There should be a distinction between the two, and periods are the only unambiguous one at present, though other potential solutions such as all caps being reserved for alphabetisms, so that one would write that "the US is a member of Nato" but never "the US is a member of NATO" would also be a workable style. The problem with that alternative is that not enough people use it, and indeed, most people would consider the use of "Nato" instead of "NATO" to be incorrect rather than a style variation. Caerwine Caer’s whines 01:36, 3 August 2008 (UTC)

Customary units

Since the United States is the only country with significant usage of a set of customary units, how about saying instead of U.S. customary units each time they are mentioned, use United States customary units the first time, and customary units each succeeding time. Caerwine Caer’s whines 03:32, 28 July 2008 (UTC)

Customary units is long-winded and could still lead to ambiguity in spite of your prior mention of the full form. JIMp talk·cont 04:10, 28 July 2008 (UTC)
I fail to see how customary units is more long-winded than U.S. customary units. Also some of the proscriptions, such as when "sq" and "cu" may be used would seem to me to be equally applicable to other customary units in thise rare instances when they are used. Caerwine Caer’s whines 05:07, 28 July 2008 (UTC)
I fail to see that too. What I meant is that it's more long-winded than "US" or "U.S.", which is generally sufficent. JIMp talk·cont 05:18, 28 July 2008 (UTC)
  • On Wikipedia, we’re supposed to be using U.S. customary units as the primary unit only in articles that pertain to the U.S. and any other countries/topics that still use those measurements (some Canadian disciplines?). So an article referring to the fuel efficiency of an American-made vehicle only needs to say “miles per gallon” because the audience that is accustomed to these units aren’t confused in the least. The article doesn’t need to say “miles (U.S. customary - statute) per gallon (U.S. customary - liquid)”; not as a primary measure nor as a parenthetical conversion. Were we to do as you are suggesting, the text would become unnecessarily verbose, cumbersome, and awkward looking. And Wikipedia would look like we were preparing for the inevitable invasion by an alien space race (“we just wanted you to get up to speed as soon as you landed”). If a reader needs extra clarity, they can click on the link; that’s what they’re there for. Greg L (talk) 14:15, 30 July 2008 (UTC)
(and Detroit wonders why they have trouble exporting cars....)LeadSongDog (talk) 14:21, 30 July 2008 (UTC)
  • If they made an appealing car that I liked better than my Miata and my wife’s Honda, I’d buy it. Greg L (talk) 14:28, 30 July 2008 (UTC)
Trouble Exporting??? Hell trouble just selling 'em in North America. I live it everyday. Trust me it's a combination of the morons at the top and the unions that have wrecked the big three and it's pretty damn scary for the rest of us stuck in the middle. Thank God my company also does business outside of the automotive realm. But this discussion now has little to do with the mosnum, so we all agree that it's not practical to do what Caer suggested. —MJCdetroit (yak) 15:03, 30 July 2008 (UTC)
  • Yup. Agreed. More than impractical. Undesirable. Greg L (talk) 16:49, 30 July 2008 (UTC)

Gallon

This one I don't support as much, since unless we want to change a whole bunch of pages, we'll need to mention U.S. gallon (or US gallon or both) at least once, so it doesn't solve the issue, but perhaps changing as may mentions as would be unawkward to liquid gallon, leaving the U.S. implied instead of the liquid would work. Caerwine Caer’s whines 03:32, 28 July 2008 (UTC)

I'm not sure what you mean but liquid gallon is unclear. JIMp talk·cont 04:10, 28 July 2008 (UTC)
There are two U.S. gallons (liquid and dry), tho the dry gallon is seldom used, unlike the dry quart and dry pint which are used fairly often. (Indeed, I regularly buy a dry pint of grape tomatoes when I go grocery shopping.) Caerwine Caer’s whines 05:07, 28 July 2008 (UTC)

Yes, there are two US gallons. However, you can generally guess which one is meant by that which is being measured. Also the imperial gallon is usually used for liquids. JIMp talk·cont 05:24, 28 July 2008 (UTC)

The relevant U.S. statute, NIST's Special Publication 811 (page 68) and the 3rd edition of the American Heritage Dictionary (unabridged, see Measurements Table) all fail to list a U.S. dry gallon. It is not at all clear that such a thing exists today. --Gerry Ashton (talk) 06:58, 28 July 2008 (UTC)
  • It would seem that “U.S. gallon” (or US gallon) confuses no one unless it were to be used in an article on a discipline that uses U.S. customary dry volume, such as U.S. wheat production or wheat prices on commodities markets; there are 8 dry gallons per bushel. And even then, such an article would be using ‘bushels’ and not gallons. It the U.S. it is generally assumed that unless specified otherwise, “gallon” refers to the liquid gallon (3.78 L) so the extra specificity of “U.S. liquid gallon” is unnecessarily awkward and cumbersome. I would think that any article that refers to the U.S. dry gallon (4.405 L), should say so: “U.S. dry gallon”, link the first occurrence, and say “dry gallon” thereafter in the same article. I would think this would be a rare occurrence indeed; I know of no use of “dry gallon” on Wikipedia other than to discuss the unit itself. I would add though, that I haven’t a clue what would happen if I drove two miles to a nearby feed store that sells bulk grain for livestock and asked for “two gallons” of something. But, again, this is the only context (“bushel” confusion within the discipline of agricultural grain sales) where this ambiguity could possibly arise in real life. Greg L (talk) 13:52, 30 July 2008 (UTC)
  • I did just about what Greg L describes; I went to a US store that sells farm supplies and asked for enough buckwheat to plant 3000 square feet. They sold me 5 pounds of seed. --Gerry Ashton (talk) 18:08, 30 July 2008 (UTC)
  • Load up the ol’ buckboard. Yee HAA!. Or elsewhere, that would be “Gerry Ashton, while visiting from California, walked into a European store that sells farm supplies and asked for enough buckwheat to plant 250 square meters. They sold him 2 kg of seed.” IMO, you used a piss-poor example to point out the shortcomings of the U.S. customary system; I can come up with much better examples to accomplish that. Greg L (talk) 22:29, 30 July 2008 (UTC)
  • P.S. I’m a big, big advocate of the metric system. I'm also a big fan of writing clearly. Whereas “Gallon (U.S. customary - liquid {statute of 5th of Queen Anne})” is highly specific (and helps to highlight the shortcomings of U.S. customary system), such unnecessary specificity usually doesn’t help Wikipedia’s articles read better—just the opposite. That’s the only thing this is about: communicating clearly. We’re not here to change the way the world works. Greg L (talk) 22:37, 30 July 2008 (UTC)


  • P.S. Caerwine, all this extra specificity you are proposing in your various proposals above doesn’t add clarity nor minimize confusion. The resulting verbosity would be clunky and awkward and wouldn’t enhance understanding whatsoever for the target audience that needs these units of measure. “Gallon (U.S. customary - liquid {statute of 5th of Queen Anne})” isn’t needed unless you are making a chart comparing different units of measure. If your objective is to point out how stupid, confusing, and unhelpful the U.S. customary system of measurement is, your proposals are a splendid vehicle to accomplish that end. But that doesn’t seem to be your objective here; you seem to want to clear up confusion. The trouble is, there is no confusion to clear up. Greg L (talk) 14:46, 30 July 2008 (UTC)

Miles

miles (U.S. customary - statute) per gallon (U.S. customary - liquid) - you're not serious, right? We're ALMOST NEVER going to have to specify the US "statute"/survey mile vs the one based on the international foot - the difference is literally only two parts per million; a quantity would have to be given to six or seven significant figures before it matters, and sources often don't specify what mile is used ("statute" mile is often used simply to disambiguate from nautical miles). 30 mile/US gallon is 1.27543×107 m-2 is (reciprocally) 7.8405 L/100km no matter which mile is used. --Random832 (contribs) 15:09, 30 July 2008 (UTC)

  • Read what I wrote. No, I'm not serious. I was pointing out, via excess example, that Caerwine’s suggestions amounted to excess specificity that did nothing to enhance understanding or reduce ambiguity. Greg L (talk) 16:19, 30 July 2008 (UTC)
    • Perhaps you should read as well. My suggestions were not about adding extra specificity, but about changing to an alternate specificity so as to hopefully avoid the U.S v. US styling issue as much as possible. Since the Imperial gallon is both a liquid and a dry measure, I was hopeful that adding the specifier liquid instead of U.S., might prove acceptable as an alternate means of specifying which gallon was intended. I was doubtful that it would be, but not so much I was willing to assume failure without testing the proposition. Caerwine Caer’s whines 23:03, 30 July 2008 (UTC)
  • Understood: why not float the proposal. And, yes, the proposal didn’t fly; about as well as a wet noodle. “US gallon” (or “U.S. gallon”—I don’t care) is well recognized, sufficiently clear, and causes no confusion (except perhaps in all but the very, very rarest of articles). Greg L (talk) 23:22, 30 July 2008 (UTC)
  • Caerwine, I'll be very disappointed to see any more of this campaign to go dotting all over the place, and alternatively to make clutter-compromises such as spelling out the well-known names "United Kingdom"/"United States" where initialisms do very well now. This change you propose is just poiny and pointless. Tony (talk) 00:00, 31 July 2008 (UTC)

Nautical miles per Canadian gallon

This summer, at one particularly remote harbour (or harbor using US spelling), I managed to fill my yacht up with 14.62 Canadian gallons of diesel fuel (they hadn't changed their pumps for a few decades), and I managed to get about 25 nautical miles (we don't use your wimpy landlubber miles at sea) per Canadian gallon. (Of course, it's a sailboat.) So let me offer the following suggestion: If you don't use SI (metric) units, specify which country or which organization defined the standards you are basing your units on. Are you using American imperial gallons, or British imperial gallons, or Canadian imperial gallons. And, are your miles statute miles, or nautical miles, or Norwegian miles. (I add the last one because I drove in Norway, and I speak a bit of Norwegian, so I know they often specify their mileage in litres per Norwegian mile - the Norwegian mile is 10 kilometres or about 6.2 English miles). So, if your are using non-SI units, don't assume that everyone uses the same gallons or miles as you and specify who's standards you are using. RockyMtnGuy (talk) 22:04, 4 August 2008 (UTC)

  • It is not so complex; common sense always prevails. I would say it depends on the article. If it were an article that was particularly germane and specific to a particular practice (a certain sail boat race) in a particular place (from one island off Norway to another), then one should look towards current literature on that subject. In that specific case, I would advocate using whatever the customary units of measure are that the participants in that particular race are familiar with and would use parenthetical conversions to whatever units would commonly be used internationally in sail boating. If it is an general-interest article on sail boating, then you would use the units of measure for sail boating that are typically used internationally, and would add parenthetical conversions to one or two other common conventions as practical sense would dictate. To address the specifics of your set-up question and do so literally, it doesn’t matter what you did at that pump; it only matters what subject matter is and what are the customary practices for those who have experience in the art. Greg L (talk) 22:49, 4 August 2008 (UTC)
I'm somewhat sensitive to this topic because I once was involved in a major metric conversion project for a giant multinational company. While the SI system is straightforward, trying to figure out what "traditional" units people might be using can be a challenge. People often don't know. For instance, the Americans fueling up at that pump halfway between Seattle and Anchorage were probably unaware that they were buying Canadian gallons of diesel fuel. All they were aware of was that the price was rather high (the Canadian gallon is 20% bigger than the American gallon.) And, when the pilot of an airplane announces, "We are 150 miles from Salt Lake City", you probably think he means statute miles, when he probably means nautical miles (which are 15% larger) because aircraft navigation is generally done in nautical miles. When you order a pint of beer in Britain, you get 20% more beer than in the U.S. When you order a pint of beer in Canada, you aren't sure what you are getting, because they might serve you a British pint, or an American pint, or possibly 500 mL of beer. In international trade, when they say "barrels", do they mean "US oil barrels", or "British imperial barrels" and when they say "tons", do they mean "long tons", "short tons", or "metric tons". If it's loaded aboard a British-owned ship flying a Liberian flag with a Greek captain and a Phillipino crew, whose standards are they using? In other words, in an international environment, you can't be sure that the "traditional" units someone else is using are the same as the "traditional" units you are used to. So, if it's not in SI units, you should specify whose national standard you are using. RockyMtnGuy (talk) 02:35, 5 August 2008 (UTC)

Of course a Canadian gallon is a British gallon is an Australian gallon. They are all the imperial gallon as opposed to the only two other gallons in current use: the US liquid & dry gallons. So, keeping things simple, we need only specify that it was an imperial gallon. As for Canadian pints: in my limited experience in Canadian pubs the word pint refers to any largish beer glass, not necessarily 20 imp fl oz, but this doesn't mean that a Canadian pint is anything different to an imperial pint, all it means is that bar owners are good at playing on people's confusion. JIMp talk·cont 03:24, 5 August 2008 (UTC)

There are few ways to anger voters faster than messing with the size of their beer. A pint of draught is always an imperial pint (20 imperial fluid ounces). If a bar is going to cheat you on draught, they'll find a way to water it or underfill your glass. They'll rarely take the chance of using illegal glasses because it's tangible evidence of intent. Bottles and cans vary, but most are 355ml and all are expressly labled with their capacity. Most establishments will simply do it right, as it's worth their license to get caught.LeadSongDog (talk) 15:30, 5 August 2008 (UTC)
I have some bad news for you. Since Canada went off the British Imperial system, the pint has no legal standing in the country. As a result, a "pint" is whatever the seller says it is. So, if you go into a Canadian bar and order a "pint" of beer, they will serve you whatever they want to. From a legal perspective, the authorities will only object if it is less than 8 fl oz (which is considered a "cup" of beer.) So a pint could be anything from 8 fl oz to 20 fl oz. As a result, you will almost never get 20 oz pint (568.2 mL) of beer, but you will often get 500 mL (17.6 oz) - if you are lucky. If not, you might get a 16 oz American pint and what's worse, it might actually be American beer. RockyMtnGuy (talk) 06:28, 8 August 2008 (UTC)
See The Weights and Measures Act SCHEDULE II (Section 4) for the legal definition in Canada. While there, see SCHEDULE I, wherein there are three symbols for the Customary Unit litre which are "L, l or l" defined as "1/1 000 cubic metre" LeadSongDog (talk) 13:44, 8 August 2008 (UTC)
The Imperial pint was grandfathered into the system as nominally equal to the British Imperial pint (20 oz). However, the key phrase in the law is "subject to prescribed limits of error". Measurement Canada doesn't seem to take the pint seriously, so the prescribed limits of error seem to be +/- 12 oz. That's broad enough to include the UK pint, the US pint, the so-called metric pint (500 mL), and some obscure French pints. However, now bars seem to be serving "sleeves" of beer rather than pints, which neatly dodges the issue since a "sleeve" is not a unit of measure. RockyMtnGuy (talk) 03:32, 9 August 2008 (UTC)
Hi RockyMtnGuy, I hope it would be very unlikely that a Canadian establishment would be serving American-style beer. The legend/rumour about "watered down" beer is that during World War II, 10% watering down occurred for the sake of economy and war contingencies, but in postwar years, the standard remained. It does sound like an urban myth, but there may be some truth to it. FWiW Bzuk (talk) 15:14, 8 August 2008 (UTC).
No, it's sad but true. As a result of Free Trade, many Canadian drinking establishments now serve American-style beer, and many Canadians actually drink the stuff. I personally think the difference arose as a consequence of prohibition, because after prohibition was repealed, many states still objected to alcohol consumption and set the legal alcohol content of beer as low as possible. On the other hand, during prohibition Canadian breweries sold large amounts of illegal beer to consumers in the US, and they brewed it with a lot more alcohol. Traditions persist, particularly in Quebec where some beers reach 11% alcohol. RockyMtnGuy (talk) 03:32, 9 August 2008 (UTC)

Superscripts

Headbomb recently added the following.

Do not use the unicode characters ² and ³. They are harder to read on small display, and are not aligned with supercript characters (see x1x²x³x4 vs. x1x2x3x4).

What's the general feeling here? I came out against this rule the last time it was added to the MoS. It was since removed. I'm not calling for its removal again but I would like to see what kind of ground we're standing on with respect to consensus for this. JIMp talk·cont 05:15, 28 July 2008 (UTC)

  • There might be differences in readability depending on computer platform and browser so we all might be getting different data with which to form opinions. I used to be a fan of the special unicode superscript-like characters because Wikipedia’s rendering engine use to add extra leading (line spacing) when you coded superscript (e.g. x<sup>2</sup>). That made short paragraphs hard to read because you could hardly tell when one paragraph ended and another started. However, on my Mac (OS X and Safari), the Unicode characters look damn small. Now that Wikipedia has fixed leading, I now prefer a coded (x<sup>2</sup>) superscript. And certainly, the two superscript methods should not be mixed in the same formula! Greg L (talk) 13:29, 28 July 2008 (UTC)
    • These are two competing problems, occurring in different browsers. We should explain both and let editors take their choice. Septentrionalis PMAnderson 22:04, 28 July 2008 (UTC)
      • For reference, here's what I see ([5]) (which makes me realize that they are harder to read on any display. I use FF3 and 1280x800 resolution) Headbomb {ταλκWP Physics: PotW} 00:54, 29 July 2008 (UTC)
  • Support doing away with the unicode chars. --Rifleman 82 (talk) 03:33, 29 July 2008 (UTC)
  • Support doing away with the unicode chars. Greg L (talk) 13:32, 30 July 2008 (UTC)
  • My preference in the past has always been to exclusively use the unicode characters for powers of units - since superscript tags mess up line spacing. Hmm, Greg L now seems to be saying that's fixed... testing x2 testing - ok, Support. --Random832 (contribs) 14:55, 30 July 2008 (UTC)
  • (From Greg L (talk) 21:51, 30 July 2008 (UTC)) Testing…
Lorem ipsum dolor sit amet, maecenas eligendi tincidunt aenean, sit et hac hendrerit massa, morbi maecenas nec vel auctor. Aliquam sit, tincidunt justo arcu neque eu mi fames. Vitae tellus suspendisse sed sit, dapibus ante purus erat non dui vivamus, dolor ultricies maecenas lacus luctus nunc, integer cursus tellus, anim a sem. → 592 mm3 ← fusce non, hendrerit etiam turpis vivamus hac, eget magna laoreet. Ipsum class risus, vitae leo lacinia rutrum cursus mauris nunc, purus tincidunt quisquam est blandit sed, auctor auctor. Feugiat pede metus sed ut integer duis, quis nec purus, ac ad in ac convallis. Odio morbi pellentesque facilisis. Praesent sed tempus phasellus turpis nec elementum, justo eu volutpat tincidunt perferendis, mauris enim nullam et pellentesque sociosqu sodales, eros nulla sociosqu nascetur mauris euismod. Libero urna morbi lacus, quisque varius massa dapibus egestas aliquam vulputate.
    • Yup. Exaggerated space for the superscript, whereas the next space is too narrow. Integer actually overlaps the dot in risus. Septentrionalis PMAnderson 21:56, 30 July 2008 (UTC)
  • Pmanderson, you still didn’t specify which browser/OS [I see, from Microshaft. I can guess.] nor did you specify the amount of leading change. And indeed, Wikipedia’s new rendering engine does still mess with leading. I was wrong; it is not zero (at least on my platform) But it is nothing remotely like it used to be. Using Mac OS X 10.5.4 (the very latest) and Safari 3.1.2 (the very latest for PowerPC on a Mac) and using Firefox 3.0.1 (the very latest), the shift is precisely 1 pixel for both browsers. I did a screen capture and made my measurements at 2X magnification. Having noticed the huge difference a month or so ago (for me anyway) in the new rendering engine and how superscripts no longer make a paragraph look like a whole new paragraph wherever superscripts are used, I assumed it was zero change. It isn’t zero. But it is clearly as small as possible and isn’t objectionable on my browsers. Pmanderson, I don’t know what your shift is (do tell) but it is so small now for me as to not be disruptive to the paragraph and its readability advantages, compared to squint², are well worth it IMO. I also see that whereas they still have the special “m²” and “m³” symbols one can choose from the insert table, the developers saw fit to remove the individual ² and ³ symbols; you have to copy/paste or hand-code them as unicode. I assume they did this when the adjusted the rendering engine. I take that as a hint for what they expected to occur here. Greg L (talk) 22:15, 30 July 2008 (UTC)
  • P.S. I would be interested in exactly what version of Explorer you are using. If it is so old that only a small proportion of readers are using it, it no longer needs to be supported for an optimum user experience. For instance, Web sites like CNN are now typically laid out so users must scroll horizontally unless their monitors have at least 1024 pixels of horizontal resolution. If a user has a 800 × 600 or smaller monitor, they have a less-then-optimum experience. Greg L (talk) 23:30, 30 July 2008 (UTC)
Note that in the Wiki markup hints at the bottom of each edit page theres javascript to achieve m² and m³. Oh, and I have Firefox 2.0.0.16 and i have changed line spacing. -- Eiland (talk) 22:35, 2 August 2008 (UTC)
  • Support per above and also because the unicode chars are more likely to get mangled when copypasting. Sept, how bad is the spacing for you? - Dan Dank55 (talk)(mistakes) 03:04, 5 August 2008 (UTC)

Placement of "Dates of birth and death" in the presence of infoboxes

Why do we insist on placing the dates after the name in the intro when it causes clutter and can be tucked away in infoboxes? What is so special about the dates anyway? I suggest we change the MOS to suggest that the DOB may be removed from the intro if an appropriate infobox exists. --Adoniscik(t, c) 16:19, 30 July 2008 (UTC)

No harm in having it in both places. Some editors remove infoboxes on principle, and it would be a shame to misplace the information. Infoboxes shouldn't contain anything that isn't included and sourced in the text anyway. Septentrionalis PMAnderson 18:35, 30 July 2008 (UTC)

What basis is there for removing something that neatly presents all the facts? It's much easier to glance at an infobox, which has a standard format, than to fish through text. Also, why shouldn't the infobox contain anything that isn't in the text? Just provide a citation. --Adoniscik(t, c) 19:07, 30 July 2008 (UTC)

Talk to those who don't like them; but I've seen "loud", "ugly", "redundant" and "stupid". I don't care either way myself; but I object to either side of this debate being enforced here. Septentrionalis PMAnderson 20:03, 30 July 2008 (UTC)
Placement of birth/death dates in biographical articles, including infoboxes, is a long-standing and accepted practice. WP:IDONTLIKEIT is not a reason to abandon that. Dl2000 (talk) 00:06, 31 July 2008 (UTC)
There's every reason not to clutter the top with duplicated information. I've heard complaints by people who'd rather infoboxes contained new information, or less, rather than shoving dates of birth and death in the reader's face twice in two prominent positions. It's a reason I objected to the blanket use of infoblots in the first place. Tony (talk) 00:06, 31 July 2008 (UTC)
Nor is "I do like it". And the fact that an urchin is widespread and long-standing, like poor grammar, is no reason to object to arguments for improvement. Wasting readers' time with duplication in almost the same place degrades the project. Tony (talk) 00:11, 31 July 2008 (UTC)
As far as I understood, it was standard encyclopedic practice because it allows the viewer a quick and easy understanding of the era that person lived in and whether or not there is something significant about their physical lifespan. For example, Abd-Allah ibn Ibadh, who was missing this information, was immediately placed in Category:Possibly living people, even though a two second check would have pointed out that he started a religion almost 1500 years ago. Is this about Kenan Evren? Cheers, CP 03:48, 1 August 2008 (UTC)
The infobox exists, I think, to concisely present the most important information about a subject. For countries, that's the population, capital, area, etc. For politicians, that's date of birth and death, their most notable offices, whom they served with, preceded, succeeded, etc. This information must also exist in the article text; I find infoboxes useful to quickly find this information. --Golbez (talk) 04:59, 1 August 2008 (UTC)
Why must the information also exist in the article text? I contend the opposite; we are duplicating information that is better presented in the infobox. The standard formatting also makes it possible for search engines to mine the data. --Adoniscik(t, c) 15:04, 3 August 2008 (UTC)
This hits the nail: I don't care whether the information is in the opening para of the lead or just to the right in the infobox: there should be more emphasis on avoiding straight repetitions of information. Tony (talk) 03:06, 9 August 2008 (UTC)

Banning metric units

From time to time, I come across ways in which people ban metric units. Sometimes this is because they do not like them, sometimes it is because they fail to appreciate the effects of things they do. One of the common ways relates to quotations. See: Crotalus mitchellii angelensis. In this case the text is written as:

  • the locality given is "about 4 miles southeast of Refugio Bay, at 1500 feet elevation, ..."

An edit to add metric units was reverted. I agree that conversions in quotes is an issue in many cases. However, this instance is somewhat trivial and the text should be recast so that the units are not part of a quote. Regional styleguides do not have to address this problem but I think an international publication does. I think this is similar to sex biased text 'a doctor should know his patients' i.e. the bias is initially not noticed, then people notice it but worry about how to fix it, then they simply stop thinking in such ways and write non-biased text in the first place. In a similar way, I would like to be able to point at guidance about units inside quotes. It does not have to be dogmatic, merely encouraging thought of alternatives. Lightmouse (talk) 10:56, 1 August 2008 (UTC)

The article talk page is a better place for this. Fnagaton 14:11, 1 August 2008 (UTC)
Fnagton is right. The MOSNUM already specifically deals with conversions inside quotes; either square brackets:
  • "about 4 miles [6 km] southeast of Refugio Bay, at 1500 feet [460 m] elevation"
or annotate (especially with very obscure units where dual conversions maybe needed):
  • "the ark was 300 cubits long by 50 cubits wide..."[1]

  1. ^ 450 x 75 ft or 135 x 22.5 m
MJCdetroit (yak) 15:21, 1 August 2008 (UTC)


I can see that I was not very clear. The issue is not about the 'how', it is about the 'whether'. This is an issue the crops up frequently. I have added a bullet that I think encapsulates what I mean:

  • If an unconverted regional unit is used within a quote, consider rewording to place the unit outside the quote.

If anyone has any better wording, feel free to change it. Perhaps the correspondence will give context.

Moved from Lightmouse talk page:begin

Here's another problem: I've noticed that Lightbot has recently been making edits to certain articles on snakes, applying the convert template in particular. While I have nothing against this per se, what I do not approve of is that it has also been making edits to type locality statements. That's not good, because these are supposed to be literal quotes. You could fix this by programming Lightbot to ignore all text between an instance of "type locality" and the end of the paragraph it occurs in. --Jwinius (talk) 10:51, 1 August 2008 (UTC)

Thanks. I noticed that too. I had a look and would like to fix it, perhaps on the lines you suggest. Have you considered recasting the text so that it is not dependent on units inside quotes e.g. 'it was said that the length was '? Lightmouse (talk) 10:58, 1 August 2008 (UTC)

That would seem like a simple solution, but once again, it's supposed to be a literal quote, so I can't change it either. Type locality statements usually come from the first ever descriptions of a species or subspecies: they may be short or long, although never more than a single sentence in my experience, and may even be wildly wrong. That's why they're "between quotation marks." --Jwinius (talk) 11:39, 1 August 2008 (UTC)

Moved from Lightmouse talk page:end

Other edits that are relevant are:

I believe that the editor is pro-conversion. So perhaps I have mistitled this section, but if you look at the edits you may see the generic point. I would be intereted to know what people think. Lightmouse (talk) 15:39, 1 August 2008 (UTC)

Ordinarily, an editor may introduce any comment within a quotation by putting the comment within square brackets. The only exceptions I can think of are mathematical expressions and quotations that already contain square brackets. I think if Jwinius wants type locality statements to be exempt from a practice followed by most proficient English writers, he should seek to put such an exemption in the MoS. --Gerry Ashton (talk) 16:53, 1 August 2008 (UTC)

Lightmouse, is there a reason why you can't use square brackets in the manner suggested by MJCdetroit and Gerry Ashton? Thunderbird2 (talk) 14:18, 3 August 2008 (UTC)

It was tried but reverted. Regards Lightmouse (talk) 14:35, 3 August 2008 (UTC)

I see, but how about more generally. Let's say a fisherman encounters a "30 foot sea monster". Can your bot be taught to convert that to "30 ft [10 m] sea monster"? Thunderbird2 (talk) 15:23, 3 August 2008 (UTC)

Yes, it could be coded to handle "30 foot sea monster". But you would also want it to handle:
  • the crew described it as a "incredible sight to see, a 30 foot sea monster" but 20 foot was its actual size
There are several problems that I have not been able to solve:
  • there is no limit to the number of characters between the quote mark and the digits
  • there is no difference between a begin quote mark and an end quote mark. In the example that I give, the quote '30 foot' and the non-quote '20 foot' look the same to the code.
  • the convert template does not do square brackets. So with my limited skills, the first integer feet values up to 100 feet would need 100 lines of code: one line for '30 foot' and another line for '31 feet' etc.
Lightmouse (talk) 19:23, 3 August 2008 (UTC)

What is the interface to the bot like, that is, does it go through the text change-by-change and ask the operator for permission to do each change? Or maybe it does the whole article and just asks the operator if the entire set of changes to the whole article is ok? Or maybe it just changes the article and it is up to the operator to read the article before hand to see if there is anything that wouldn't work? --Gerry Ashton (talk) 20:03, 3 August 2008 (UTC)

Answer to your second sentence is not simple. There are two forms of automatation: monobook, and AWB. They both produce what people call a 'diff' page i.e. what you see when you click the 'Show changes' button. As you know, 'Show changes' allows you to edit again, save, or cancel. AWB also has a bot mode. In bot mode it will proceed to save the page without human intervention although you can opt to intervene if you want. Quality control for the bot mode involves pre-edit selection and post-edit checking (at least it does in my case). You can see for yourself if you have AWB or you install the monobook script.
Answer to your first sentence: 'No'. Neither form of automation seeks consent for each change within an article. Consent is for the entire set of edits as shown in 'Show changes' mode.
Lightmouse (talk) 20:57, 3 August 2008 (UTC)
Actually, AWB allows you to cancel edits (or lines, I'm not sure which) in diff mode for most edits (disambig, and possibly some AWB plugins, have different options). — Arthur Rubin (talk) 20:26, 5 August 2008 (UTC)

Question on #1 vs number one

I notice that many feature articles on musicians use #1 or #3 etc. to indicate chart placement of a single or album. Some editors think this is fine, others say the the number should be written out as in "the record reached number one on the pop charts" and "the album rose to number three on the hot 100. There is no mention in the WP:MoS. Any guidance for me? Hopefully, —Mattisse (Talk) 22:26, 4 August 2008 (UTC)

  • There is also the option that our local newspaper uses: The record reached No. 1 on the charts.” Greg L (talk) 22:37, 4 August 2008 (UTC)
    • WP:Record charts, which is in Category:Wikipedia style guidelines, says for instance "Oz Music Charts has a list of #1 hits..." (So if that doesn't look right to you, please tell them.) - Dan Dank55 (talk)(mistakes) 02:46, 5 August 2008 (UTC)
      • We can't punt this question to that page, which is a wreck and should be dealt with. I just deleted a large list of not-all-reliable sources from the bottom of the page, and it's supposed to be a MoS page! Many music articles appear at FAC, so we need a real guideline. And that page shows why MoS needs to be fixed and coordinated; it's not very helpful, and obviously has had little oversight. What do other style guides say about this music issue? Epbr123 has mentioned that the usage might differ depending on whether it's in text or a table. SandyGeorgia (Talk) 02:55, 5 August 2008 (UTC)
        • This is one of many issues where we need to leave well enough alone. We do not need to legislate for legislation's sake. There are several ways to do this; all of them are sound English. (To my ear, "No. 1" is British English, but if so, WP:ENGVAR says all we need to on the matter.) Septentrionalis PMAnderson 20:34, 5 August 2008 (UTC)
          • But allowing "No. 1" goes against the current guideline of having to spell out numbers under 10. Should this be made an exception to that rule? Epbr123 (talk) 08:49, 6 August 2008 (UTC)
            • Yes, there should be an exception for idiom in general. I had not realized that Kotniski's reformulation had removed it. Septentrionalis PMAnderson 17:03, 6 August 2008 (UTC)
So, then would #1 be against the guideline per the same rule? Currently at least one half FA articles on musicians (by random survey) contain this format, including recent ones. —Mattisse (Talk) 15:53, 6 August 2008 (UTC)
Tony is skiing; perhaps we'll get a discussion of the issue once he returns. SandyGeorgia (Talk) 15:56, 6 August 2008 (UTC)
Comparing Numero sign, Number sign and Pound sign sheds some light on origins. On typewriter keyboards in Britain, the Pound sign (₤) took the place which on US keyboards was occupied by the Number sign (#), so typists instead rendered the Numero sign (№) as (No.) instead. (Indeed teletype operators even called # "the pound sign"). So while top tens in other countries would use #1, the British lists would use No.1 as a typographical convenience. LeadSongDog (talk) 18:21, 7 August 2008 (UTC)
  • And I know that I recently had a letter-to-the-editor published in my local U.S.-based newspaper and they replaced my “#1” with “No. 1”. I suppose the main reason for doing so—and this is in a country the # sign is easy enough to type—is because “No. 1” looks better for main body text. I suspect # works better if its use is limited to tables and whatnot. Greg L (talk) 02:22, 8 August 2008 (UTC)
    • U.S. newspapers generally follow AP Stylebook (or claim to), which says "No. Use as the abbreviation for number in conjunction with a figure to indicate position or rank: No. 1 man, No. 3 choice." - Dan Dank55 (talk)(mistakes) 03:56, 8 August 2008 (UTC)
Outside tables, I dislike the use of the hash as an abbreviation for "Number", on the grounds of its clunky visual appearance and negative effect on the reading process. It's possible that a significant proportion of readers don't at first comprehend the meaning of #, especially those who are unfamiliar with the types of text in which # is thus used; they probably do get it after a while, but only after unnecessary bumps in their reading process. It doesn't seem to be an engvar issue. In tables, I'm less concerned, since in that context the hash is less visually intrusive (rather, its concision can be an advantage) and more readily identifiable in the presence of many other hashes as "Number".
So I'd support deprecation in running text: "The use of the hash (#) as an abbreviation for Number is acceptable in tables, but it should be avoided in running prose and line-by-line lists in favor of No., unless there is a particular reason to use the hash." Tony (talk) 00:16, 9 August 2008 (UTC)

ISO and "year month day" dates

Under the Dates section, would people oppose examples showing ISO or "year month day" formats?

X mark.svg Incorrect February 14th; 14th February; the 14th of February
Yes check.svg Correct 14 February; February 14
X mark.svg Incorrect October, 1976; 1976, October
Yes check.svg Correct October 1976; 1976 October
For example, the Olympics began on 2008-08-08 or on 2008 August 8. Placed in double square brackets, these will auto-format correctly anyways.

The reason I bring this up is that under Full date formatting subtopic Strong national ties to a topic, it says editors should use a date format familiar to the nation. East Asian countries such as China and Japan use "year month day" dates. In addition to that, their month names are numbered 1 through 12, making the overall date look very similar to the ISO format (ISO: 2008-08-08, Chinese: 2008年8月8号, Japanese: 2008年8月8日). Some articles on computer data formats may also use ISO.

According to the Dates section of the article, ISO dates are rare in prose. As it does look odd, I'd be okay with not promoting ISO dates in articles, but I think that the "year month day" format is plausible. Of course, any changes to this article would also be made on the main Manual of Style article as well.

Wikky Horse (talk) 16:58, 10 August 2008 (UTC)

Non-anglophone countries: "it says editors should use a date format familiar to the nation"—This is an incorrect paraphrasing of the guideline. See the following guidelines:

"Articles related to other countries that commonly use one of the two acceptable guidelines above should use that format." And "ISO 8601 dates (1976-05-31) are uncommon in English prose, and are generally not used in Wikipedia. However, they may be useful in long lists and tables for conciseness and ease of comparison."

Tony (talk) 00:52, 11 August 2008 (UTC)

I have never seen English-language prose that placed the year before a spelled-out month. As far as I can tell this is a neologism invented by Wikky Horse. --Gerry Ashton (talk) 01:27, 11 August 2008 (UTC)
If we had a mechanism that would format ISO dates according to user or IP location prefs, that would be fine, but most users don't have accounts and so if they turn to the China article they will see ISO dates in the middle of sentences, which looks really odd for most users of this English-language wiki. ISO dates in tables aren't too bad, but putting numbers in text is something to set the eyes grating. --Pete (talk) 02:07, 11 August 2008 (UTC)
Wikipedia does have a mechanism that will format ISO dates. Wikipedia just doesn't use that mechanism except for logged-in users who have specified a preference. This whole issue would lose most of its importance if Wikipedia were to simply assign a date format for non-logged-in users to see. (sdsds - talk) 02:31, 11 August 2008 (UTC)
There are problems with defaults:
  1. Various editors and readers have preferences as to what format they want to see their dates in.
  2. Various articles have appropriate date formats.
Finding a way to provide a reader with an appropriate date format for the topic is going to be difficult. --Pete (talk) 02:42, 11 August 2008 (UTC)
Following up, ISO dates are rarely used in prose statements. FWiW Bzuk (talk) 20:10, 11 August 2008 (UTC).
Hi Bzuk. While it may be true that use of ISO dates in English prose is common only in a few contexts, no one is suggesting widespread use of unwikilinked ISO dates. What has been suggested is use of wikilinked ISO dates which get autoformatted to some other format for presentation to readers. (Unless the reader has specified a preference for ISO dates, of course! ;-) (sdsds - talk) 01:46, 12 August 2008 (UTC)
Now this is the real reasoning behind my comments, see: Discussion on Autoformatting dates. The use of autoformatted dates in any format even the $(^)%& ISO dates (sorry I let my feelings about ISO dates slip) are being deprecated, and an editor is now free to use a simple, written out system. FWiW Bzuk (talk) 02:02, 12 August 2008 (UTC).

Changes to the text - date autoformatting

User Ckatz, who has previously taken me to task WRT the move away from date-autoformatting, is making changes to the wording of the date autoformatting section, with what I perceive to be vague or unconvincing edit-summary justifications. I've suggested that instead of edit-warring, the matter be discussed here. I do believe that the existing wording expresses fact rather than opinion, which is Ckatz's issue. Tony (talk) 10:38, 11 August 2008 (UTC)

My concern (as well as that of others) is with your moves to remove autoformatting without consensus, as opposed to the "optional" nature of the guideline at present. However, with regard to the separate matter of the tweaks to the guideline, the old version read as follows:

"Careful consideration of the advantages and disadvantages of the autoformatting mechanism should be made before applying it: the mechanism does not work for the vast majority of readers, such as unregistered users and registered users who have not made a setting, and can affect readability and appearance if there are already numerous high-value links in the text.."

while my changes were to have it read:

"Careful consideration of the advantages and disadvantages of the autoformatting mechanism should be made before applying it; autoformatting has no effect for unregistered users or registered users who have not made a setting. "

If discussion is desired, great - but I do think the changes are pretty self-evident. Saying it "does not work" really pushes the "don't use it" POV, as does the unnecessary (and unverified) "vast majority of readers" - hence my more specific language. The second part of the text, which I removed, is not suitable for a guideline as it clearly reflects an opinion about the effect of autoformatting. I was pretty clear about this in the edit summaries. --Ckatzchatspy 19:41, 11 August 2008 (UTC)
I think that a better change in the direction you suggest would be to highlight the fact that users who do make a setting may enjoy improved readability by seeing dates in the most familiar format. It's false to say that autoformatting has no effect on unregistered users; autoformatted/non-autoformatted dates appear differently to them because the former are linked and the latter are not. Christopher Parham (talk) 23:19, 11 August 2008 (UTC)

Tony's response: I agree with C Parham's second statement: it's false to assert, as Ckatz's change did, that "autoformatting has no effect on unregistered users": how many outsiders have said to me, when I raise the topic of the blue dates, "yes, I wondered why earth: it's odd", or similar queries? My disclosure that it's so Wikipedians can see those blue dates in a preferred format receives quizzical looks. Of course, they have to travel over the disruptive blue displays without the preferred formatting. Therefore, Ckatz's proposed change in the text ("no effect") is starting to look POV.

"The second part of the text,... is not suitable for a guideline as it clearly reflects an opinion about the effect of autoformatting"

This risks being "opinion" itself. There's wide acceptance that (1) some links are more valuable than others, and (2) text can be overlinked (the "sea of blue" that is referred to on this page and in the archives—by Greg L, as one of many, in less flattering words; put "turd" into your finder). A significant motivating factor for me is to make the linking system work better by removing low-value links. I'm unsure why this is not uppermost in your minds, too. So C Parham, your suggested insertion of "users who do make settings may enjoy improved readability" is not as straightforward as it might at first seem. We need to take into account the entire utility of the text and the linking system.

"Saying it "does not work" really pushes the "don't use it" POV, as does the unnecessary (and unverified) "vast majority of readers""

The first one is a fact; the second is a conjecture in the vein of "The sun will come up tomorrow morning"—unverifiable in the strictest terms, but not worth wasting people's time over the null hypothesis. WP is the ?seventh-most-visited site in the world, I think: many millions of hits a day. Compare that with the few thousand regular users, and the few tens of thousands who make occasional edits: how many of the visitors are registered, preference-set and logged in? And nothing like all edits—even by regulars—are done by registered, preference-set and logged-in users. I'm surprised to be having this discussion about the existence of the "vast majority", frankly. In any case, some people advise that we should set "no preference", so that we see what outsiders see, to make it easier to pick up the raw-format inconsistencies and broken formatting that plague some articles.

The realisation that disciplined linking is important to the project is well-established at WP:MOSLINK, WP:CONTEXT and, indeed, MoS main. Such phrases as:

"Do not make too many links" (MOSLINK), and "Links should add to the user's experience; they should not detract from it by making the article harder to read. A high density of links can draw attention away from the high-value links that you would like your readers to follow up. Redundant links clutter the page and make future maintenance harder." (MoS main)

have been there for ages, undisputed. The proposal to remove from MOSNUM wording that is consonant with these statements is going against a large-scale, long-term trend in WP.

I do hope to convince both of you to support the three main issues: (1) DA is generally undesirable, while not disputing editors' right to use it where they wish; (2) the move a while ago, in effect, to give editors at each article the power to use or not use DA was worth supporting; and (3) raising the issue on individual article talk pages is an entirely legitimate part of the way WP evolves—through reasoned debate. Tony (talk) 01:24, 12 August 2008 (UTC)

Tony, as you know I was initially sceptical about the change, but I can see the reasoning behind it and I have instituted the new approach to articles in which I presently am working, to the deafening sound of no one complaining. A suggestion that may have been made before but bears repeating is to involve the project groups in the evolving discussion about autoformatting dates. FWiW Bzuk (talk) 01:30, 12 August 2008 (UTC).
the deafening sound of no one complaining. Heh. That's been my experience too. I've removed wikilinks from dates when I've changed them for consistency or appropriateness and nobody's jumped on me.
However, unless I unlink all dates in an article, my actions in ensuring consistency have the paradoxical effect of making it appear to editors with date prefs set that there is now inconsistency, and they sometimes undo my changes to match the format displayed by the autoformatting (which of course varies according to the viewer's prefs).
I'd like to get rid of wikidates entirely, unless there is a bloody good reason to link the date. September 11 or the Fourth of July springs to mind, but in these cases there are specific articles. --Pete (talk) 03:34, 12 August 2008 (UTC)
Thanks for your comments, Bzuk and Skyring. My experience is largely the same, except that in my survey of some 100 FAs, in which I posted a note asking permission to remove the DA and provided the capped arguments for doing so, I found small clusters of resistance at: some Canada Wikiproject articles (although notably some users supported and one or two converted to support after a while); a few MilHist articles (Victoria-Cross-related, etc.), and IT articles (opposition apparently driven by Ckatz). I'm not denigrating this resistance at all, except that it was rather fulsome in one or two places. People now have the power to choose, and I can only put in a persuasive, reasoned pitch, then go away. Apart from these clusters, there were a few other isolated objections by single users. Some people were enthusiastic about the move; others supported, varying from strongly to mildly to a few "don't care much, do it if you wish". Many talk-pages have no comments, including those with subsequent sections (during a two-to-three-week period). The survey is not yet finished. Tony (talk) 05:03, 12 August 2008 (UTC)

Date-format consistency: let's bring MOSNUM into line with reality

This arises from the discussion above, initiated by Lightmouse (although not with a title I'd have chosen). The debate is prompted by MOSNUM's statement: "The same format should be used in the main text, footnotes and references of each article". In that discussion, inter alia, Lighmouse says:

I started this debate [on the scope of within-article consistency in date formatting] because there is a mismatch between guidance (demands consistency) and the reality (tolerates inconsistency). The debates about consistency have always been about the main text only. Date formatting must be one of the most talked about and most badly handled issues on Wikipedia. Yet three formats are widely seen on the same page without significant comment. Few editors care enough about date consistency to do much about it. It is not a big deal. For now, please, just cut the scope of guidance down from whole-article to main-text.

This is a view that I find compelling, except that I take it Lightmouse also intended that date formats in citation-driven references should be internally consistent where they are different from date formats in the main text.

SandyGeorgia, the FA delegate, has been concerned about this "mismatch" between rules and practice for some time; she finds it most uncomfortable to pass FAs that are, technically speaking, in breach of the MoS guidelines. This occurs, for example, in all articles that use the popular cite web template for the references section, which displays ISO dates in contempt of MOSMUM's statement:

"ISO 8601 dates (1976-05-31) are uncommon in English prose, and are generally not used in Wikipedia.

Note, however, that there's a little loophole that might be construed as allowing ISO dates in citation lists:

However, they may be useful in long lists and tables for conciseness and ease of comparison."

even though this is at odds with the quote at the top of this section <clears throat loudly>.

Sandy commented in the discussion above:

However the wording change is done, please make sure that readers understand that:

[the main] text should use one, consistent style for date formatting and linking, and
references and footnotes should use one, consistent style ...
The aim is that article text and footnotes/references may differ in style because of citation template programming, but within the footnotes and references, we should still find consistency in both linking/delinking and style of dates used. That is, if ISO dates are used, they should be used consistently [in references].... we're not letting footnotes and references off the consistency hook; we're just recognizing that current template implementations on Wiki make it very hard to achieve consistenty between article text and citations. We can still achieve consistency within each. I suppose the "consistency dividing line" would be consistency above the See also section, and consistency below the See also section, in terms of WP:LAYOUT.... In other words, I agree with the proposal, but disagree with the section heading here.

I believe that herding together the templates into a rationale, coordinated part of the project is impossible to do in the short term, and that we should go along with Lightmouse's and Sandy's call. In effect, the simplest way of acknowledging reality is to be strict about date formatting consistency within (1) the manually entered dates in an article, and separately (2) the citation-generated dates that are largely out of editors' control. We are aided in this proposal by the fact that the citations are neatly corralled at the bottom of articles, and the main text, with a few rare exceptions, contains manually entered dates.

At the moment, MOSNUM can say what it likes about whole-article consistency until it's blue in the face, but the developers of the citation templates have taken no notice, and we tag along. Tony (talk) 05:14, 12 August 2008 (UTC)

Tony, the discussion over date formats proceeded through WP:AVIATION PROJECT and the general consensus was to have an unambiguous, easy-to-understand, written out system. ISO Dating was specifically rejected as an overall project format. The problem remains that the citation templates are written in an arcane and confusing American Psychiatric Association convention, specifying the use of the ISO dating. New editors are directed to these templates although the APA guide is a university-based system that I remember being instituted because professors were tired of having to explain the then, more common Modern Language Association style guide that was the usual format for research papers. The APA guide is a simplified but yet less flexible system that ties date of publication not to the publication but the author because it was more closely associated with a Harvard citation style (APA also eliminated the place of publication for the same reason of simplifying the system). Any attempts to have the citation templates (which were provided as a means to assist writers with no formal research background or cataloging experience in providing verifiable sources of information to back up entries) reworked or even to provide an alternative in the form of the MLA system have been summarily rejected, often by the expedient of only one dissenting voice. After a number of efforts that were rebuffed, I resorted to "scratch cataloging" to introduce the MLA format to articles in which I was devoting some time. FWiW Bzuk (talk) 15:50, 12 August 2008 (UTC).
Bzuk: I agree entirely with what you say. I just loathe APA style; and I believe they make lots of money by selling their hard-copy manual, too. They change a few nuts and bolts every so often and bring out a new edition that everyone has to buy. Of course, it's not available online, is it ....
Dealing with the citation template mess is something the community might want to consider at some later stage. In the meantime, I encourage all editors to avoid citation templates and to retain proper control over their formatting by manually keying in their references. It's not rocket science, and produces a much better result than those templates. At the moment, the templates are quite a separate issue from weening ourselves off manually keyed-in DA. Tony (talk) 02:30, 13 August 2008 (UTC)

I do not understand this discussion, or what people still believe are the limitations in any of the various citation templates. There is currently a bug at {{citation}} where it won't accept delinked US-style date formats (it converts them to international style), but to my knowledge, this meme that x, y or z can't be done with a given template is simply wrong. With the exception of that bug, all of the templates can handle all combinations (linked, not linked, US-style or international-style), and that one bug can be dealt with manually by moving the date out of the template. I suspect that many editors are repeating these memes without having experimented with different methods. Again, there are three completely delinked samples in:

SandyGeorgia (Talk) 02:35, 13 August 2008 (UTC)

Thank God for a rational voice. There really is no problem that can't either be easily worked around, or hopefully quickly sorted out. --Malleus Fatuorum (talk) 02:41, 13 August 2008 (UTC)
Perhaps the problem is that, because different templates handle dates differently, one actually has to take the time to figure each one out. But it's not rocket science; even I can do it. SandyGeorgia (Talk) 02:44, 13 August 2008 (UTC)
That's a problem, and one that hopefully will soon be fixed. As a user of the {{citation}} template though, I feel somewhat insulted by the suggestion that it's "to assist writers with no formal research background or cataloging experience". I have a "formal research background", and I have no reason to doubt that others who choose to use the template do so as well. --Malleus Fatuorum (talk) 02:53, 13 August 2008 (UTC)

Date-format in regions

In American English the date is written as mm/dd/yy
In British English AKA European English the date is written as dd/mm/yy.
I suggest that on all American related articles we should use the American format and in European related articles, we use the European format. This makes sense. Also the same goes with articles related to South Africa, Australia ect. I believe this should be added to this wiki-page Ijanderson977 (talk) 18:24, 12 August 2008 (UTC)

A couple of judgement calls. Where would Grand Theft Auto IV fall? Created and designed in Britain owned by a US company and set in fictional American locations. Or Cary Grant born in UK but made his name in US. Laurel and Hardy will be a fun one as well. - X201 (talk) 10:31, 13 August 2008 (UTC)
I'm suprised to see that neither Wikipedia:Manual of Style nor [[Wikipedia:Manual of Style (dates and numbers) explicitly forbid the formants mentioned by Ijanderson977 (although they are not among the acceptable options given). In any case, there is no chance whatsoever that the community would tolarate an endorsement of either format because they are so confusing. Who is to say if an article is American or European oriented? How do we know if the editor who wrote the date was obeying the Manual of Style? --Gerry Ashton (talk) 19:25, 12 August 2008 (UTC)

Written out dates, 12 August 2008 and August 12, 2008 are confusing? What system are you advocating? FWiW Bzuk (talk) 19:45, 12 August 2008 (UTC).

Ijanderson977 is advocating that if an article relates to Europe, and needs to mention today's date, it would be 12/08/08, but if the article relates to America, it be written 08/12/08. That's what confusing; dates where the date is written in letters are not confusing. --Gerry Ashton (talk) 21:15, 12 August 2008 (UTC)

That's not what mm/dd/yy or dd/mm/yy advocates unless it is in ISO dating format. When I see dd/mm/yy, it is a code for writing out 12 August 2008 and mm/dd/yy means August 1, 2008 to me; that's why I was confused, as I didn't see an argument for using numerical equivalents, merely for the arrangement of the day or month sequence. FWiW Bzuk (talk) 21:30, 12 August 2008 (UTC).

Where on earth would dd/mm/yy mean 12 August 2008 format, as opposed to 12/08/08 ?? Perhaps you meant dd Month yyyy.LeadSongDog (talk) 05:43, 13 August 2008 (UTC)

Yep, numbers alone are dangerous—I have to think carefully to interpret them, and sometimes still can't determine which system. The advantage of spelling out the month is that (1) it instantly classifies all three components of a full date, and (2) it's more accessible to most readers. Tony (talk) 05:32, 13 August 2008 (UTC)

Guidance is only needed if there is a significant problem to solve *and* it would not be solved by the wiki. The format 'xx/yy' is not present in any significant amount and is quickly changed by the wiki. Thus guidance is not required. Lightmouse (talk) 10:01, 13 August 2008 (UTC)
I was just saying isn't it best to use the American style date format on American articles and European format on European articles. I thought that would be easy enough to understand and it makes sense. Ijanderson977 (talk) 23:26, 13 August 2008 (UTC)
We say that: 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.
That Ijanderson977 appears to have missed it while considering the subject should be a hint on how widely read, and hence how useful, our guidance is.</irony> Septentrionalis PMAnderson 21:14, 14 August 2008 (UTC)
"...articles related to Canada may use either format consistently" - There are at least four different date formats in common use in Canada, plus some uncommon ones, so you can't reasonably expect to decode dates unless you have some more information than a person's location. Since this is the English-language Wikipedia, you can at least avoid the French date formats (which Canadian residents cannot), leaving only the British, American, and ISO date formats to contend with (and the ISO format is in common use in Canada). However, if you see something like "10/11/12" in a Canadian article, you really don't what it is. To avoid confusion with mixed date formats you must:
  1. Always use a four-digit year
  2. Always spell the month
That makes it unambiguous, regardless of whether the author means:
  • 10 November 2012,
  • October 11, 2012, or
  • 2010 November 12.
Otherwise you can't avoid ambiguity. RockyMtnGuy (talk) 00:10, 15 August 2008 (UTC)

Autoformat quick-question

Can someone tell me why we don't auto-format dates for non-registered users? -- SatyrTN (talk / contribs) 02:10, 13 August 2008 (UTC)

Your question seems to embody a misunderstanding. Autoformatting was a quick and very dirty fix that nobody in their right mind ought to have agreed to. Non-registered users see the dates as they were written. If, on the other hand, your question is why didn't the developers come up with with a system that did what many were deceived into thinking that autoformatting did, then your guess is as good as mine. Laziness? Incompetence? Ignorance? --Malleus Fatuorum (talk) 02:19, 13 August 2008 (UTC)

No, I meant what I asked. Why is it that a visitor to our site that doesn't log in to any account will see 2008-08-23 when a logged in user will see it auto-formatted? Why don't we auto-format dates for users that aren't logged in? -- SatyrTN (talk / contribs) 03:00, 13 August 2008 (UTC)

I don't think there is universal agreement that we should present information in the same format that the reader is most accustomed to. Few other publications see any need to do that; the only series I can think of is the Harry Potter books. --Gerry Ashton (talk) 03:08, 13 August 2008 (UTC)

Wikipedia should end its addiction to tinkering with date formats and mandating date format consistency within articles. Ordinary people do not care much. Wikipedia clearly does not care much about what ordinary readers see. As an issue, date format is less of a concern than regional spelling (color vs colour). Spelling can be wrong for the region but unambiguous date formats are not wrong, merely less common. Lightmouse (talk) 09:47, 13 August 2008 (UTC)

I bet the reason why we don't auto-format dates for non-registered users is because we couldn't reach consensus on which format to convert them into! Basing the choice on something arbitrary and essentially random would be the only "fair" way to do it. (sdsds - talk) 16:15, 13 August 2008 (UTC)
Or perhaps basing it on the country of origin, as determined by their IP, which would at least be right most of the time - that would be much more fair, IMO. -- SatyrTN (talk / contribs) 16:25, 13 August 2008 (UTC)
  • SatyrTN: Regarding your 03:00, 13 August 2008 post (“Why is it that a visitor to our site that doesn't log in to any account will see 2008-08-23 when a logged in user will see it auto-formatted?”), exactly. Somewhere along the lines, some developers had a total brain-abortion and developed template tools that only benefit readers if A) they sign up as a registered editor, and B) go to their user preferences and adjust their default setting. As a result, if editors code [[2005-08-06]], we (the minority privileged registered editors) will see either August 6, 2005 or 6 August 2005 (after setting our users preferences). But the vast majority of readers see only  2005-08-06. How ugly and ambiguous is that? Is that June 8, 2005 or August 6, 2005? Although this is an ISO-standard, all-numeric format for consistently storing and retrieving data in computers, it is certainly not human-friendly. It’s not intuitively clear what date it is until the 13th day of the month. The proper thing to do is simply stop using all templates and tools that don’t benefit everyone equally.

    What would solve all of this is if Wikipedia developers made a parser function that looked to the requesting reader’s I.P. address so it knew what country they are from. This is routinely done on Web servers of all sorts so they can spoon-feed custom content to the reader and collect reader statistics. There could be certain parser function classifications and groupings of these classifications, such as “US/other” or “US/Commonwealth”, “NorthAmerica/Other”, “US/UK”. Then we could have templates like {{US/Commonwealth|trunk|boot}} so text for US readers would be “…he put the bomb in the trunk of the car…” and UK and Australian readers would see “…he put the bomb in the boot of the car…”. Similarly, editors could write {{US/Commonwealth|color|colour}}. For dates, (although I personally don’t have a problem looking at 2 August 2008), one could write {{US/other|July 4, 1776|4 July 1776}}.'

    The simple solution is for editors to stop using the formatting tools. That will also stop linking dates to mindless lists of historical trivia. If one is writing an article on an intrinsically historical subject, such as Napoleon Bonaparte, then linking “1799” can be topical. But for most articles, the resultant links to seemingly random trivia are not topical and just junk up articles with too many blue links, which discourages learning and exploration. If someone is reading up on Planck units, they don’t need to read that “Max Planck first proposed them in 1899”, and click on the link in hopes of learning more about his proposal, only to be faced with stuff like “July 17: The French Bretonnet-Braun mission is destroyed in the battle of Togbao, in Chad, by the warlord Rabih az-Zubayr.” (*aack*) Greg L (talk) 17:21, 13 August 2008 (UTC)

  • OK, we’ve discussed this more than long enough. See the endless threads, above, on this issue. It keeps coming up again and again and again. I simply don’t understand why there has been so much paralysis on action regarding discouraging the use of [[2005-08-06]]. It is beyond worthless. I added a footnote (∆ here) to the table of date formats, here. Given that this particular format produces (to quote O.J. Simpson) “ugly-ass” text for the majority of readers, this seems a common-sense thing to do. I guess we’ll just sit back and watch for who reverts it and why; that will at least clarify for me why this paralysis has gone on for so damned long. Greg L (talk) 17:38, 13 August 2008 (UTC)
Your simple solution is to get hundreds of editors to change thousands of articles.
My simple solution is to get a few developers to change a relatively insignificant amount of code.
Which is simpler? -- SatyrTN (talk / contribs) 17:47, 13 August 2008 (UTC)
I applaud you being bold, but since consensus has not been reached on discouraging the use, I've reverted your change until we do have consensus. -- SatyrTN (talk / contribs) 17:49, 13 August 2008 (UTC)
  • Good luck with the developers. I’m not holding my breath. Greg L (talk) 17:52, 13 August 2008 (UTC)

I can't work out where to comment, but my comment would be that I'd support dropping date formatting as long as we adopt the compromise we have for spelling, first contributor unless article subject indicates a national tie and consistency. Hiding T 18:12, 13 August 2008 (UTC)

  • Hiding, you know where I stand on linking of dates. As for formatting, I couldn’t care about this issue and am utterly baffled that there is so much conflict over it. I’m an American, where “February 14, 2007” is common but when I write articles that aren’t closely tied to the U.S., I use the fluid and ‘metropolitan’ “14 February 2007”. The Euro style progressively builds from small to big and saves a comma. More importantly, either method confuses absolutely no one.

    As to spelling, there is already a perfectly workable solution on WP:MOS:

(My emphasis.) That guideline encourages real contributions from editors rather than encouraging them to run out and start a near-worthless little stub of an article. It also keeps authoring on Wikipedia a fun hobby for everyone by not discouraging editors who have invested a lot of time and energy into an article, only to see someone wade in later and change the spelling to someone else’s national convention. Is there some proposed change afoot being contemplated? To bad if it is; the current guideline seems to be one of the wisest policies I’ve seen on Wikipedia.

And finally, why do you perceive there is a link between the dropping of linked dates and spelling? Or is it that you are offering your support for one issue if you gain support for another? Greg L (talk) 20:05, 13 August 2008 (UTC)

  • You seem to be reading more into my comment than I wrote. I haven't mentioned linking of dates, but since you opened that can of worms, I prefer to link to dates when they are useful, and object to arbitrary removal of them, and support piped links of the type [[1938 in comics|1938]]. With regards formatting, I think we are in agreement, but I'm finding it hard to tell. Why do you think I perceive a link between the dropping of linked dates and spelling? To my best knowledge I'm not currently engaging in quid pro quo, but if it is on offer, I'll gladly take some. I just happened to follow a link on {{tl:Cent}}. Hiding T 23:11, 13 August 2008 (UTC)
  • I've asked this before and don't recall much of an answer. It would be fairly easy to have autoformatted/linked dates render without links. When links were desired they could still be created with pipes, since that circumvents autoformatting. This would require no changes to articles, would retain autoformatting for those who want it, and would remove a whole lot of unnecessary blue. What are the downsides to this solution to the datelink problem? Gimmetrow 18:50, 13 August 2008 (UTC)
  • Is there anyone here who thinks [[2005-08-06]] is a *good* thing? I’ve seen precious little support for it and thought there was a consensus to loose the thing. No editor would be required to wade back into their articles and deprecate all instances of it; deprecation can be left to those editors who give a darn. And if they do the task properly (not mix styles or use the opportunity to change the date style used in the article), it’s easy as pie. Greg L (talk) 20:10, 13 August 2008 (UTC)
  • Is this intended to be a response to me? Gimmetrow 20:15, 13 August 2008 (UTC)
  • Not necessarily. It’s directed to anyone who thinks [[2005-08-06]] is a *good* thing. If there are few (or even no) supporters, then it’s time to stop using a date format that messes things up for 99.9% of Wikipedia’s readership. P.S. I believe your question was provoked by how my post was seeming clueless to your position. Yes. I understand where you stand on the linking of dates. I would indeed be interested to know how you feel about that one particular date formatting tool. Greg L (talk) 20:19, 13 August 2008 (UTC)
  • Oh, Gimmetrow: In answer to your question as to whether there are any downsides; I think you have an outstanding idea. But I’ve read here that the formatting tools were the product of a developer who likes loves the links and doesn’t respond to requests to loose them. That’s about the only downside of your proposal that I can see. Greg L (talk) 20:26, 13 August 2008 (UTC)
  • Actually, GregL, I use 2008-08-13. Often. It's the quickest way to enter a date. Though, honestly, I usually do that in refs, not in the main body of the article. And Gimmetrow, I would totally support taking the links off of autoformatting. I'd also totally support auto-formatting for non-logged in readers. -- SatyrTN (talk / contribs) 21:06, 13 August 2008 (UTC)
  • I think a format like 2008-08-14 is by far the best way to render a clear and unambiguous date. If several dates are used, it is immediately obvious in which sequence they are. In a table they line up neatly. No confusion of meaning is possible. It is easily searchable. In my daily work in an international company we almost exclusively use that format. −Woodstone (talk) 09:32, 14 August 2008 (UTC)
  • Woodstone: It’s beside the point whether you use “2008-08-14” for such purposes. I’m talking about the use of “[[2008-08-14]]”, which editors are using for in-line prose and like the results because they don’t realize that it junks up in-line prose with ugly all-numeric dates for the vast majority of readers. Either that, or the do realize that it junks it up for most readers and don’t give a crap. It was unwise for the developers to have provided a tool that produces the intended effect only for registered editors. That decision was simply brain damaged. Greg L (talk) 15:19, 14 August 2008 (UTC)
  • I don't understand why it "junks up" the prose. It is just a notation. We don't write "August fourteen, twothousand eight" do we? People see a number and read words. After the first few times it becomes natural to just see a "date", not a "code". −Woodstone (talk) 19:46, 14 August 2008 (UTC)
  • I do write "August 14, 2008," routinely. Whenever I meet an ISO date from the first third of the month and context does not make the answer obvious, I have to stop and remember which one encodes the day and which the month. To us, ISO dates are obnoxious and confusing. That is sufficient reason to use them with extreme caution. Septentrionalis PMAnderson 21:27, 14 August 2008 (UTC)
  • I agree with Woodstone for the reasons he stated. Firstly, it is ironic that ISO8601 was created to solve the problem we are discussing, yet we want to ban it. Secondly (and perhaps more importantly to Greg) if we continue with the status quo libertarian view, we can delink dates throughout an article simply by removing the square brackets - if ISO is banned, we effectively ban delinking as a single action. Lightmouse (talk) 09:53, 14 August 2008 (UTC)
(ec)Yes, Greg, unambiguous dates in wikitext that do not need translation are a good thing because they are more often correct. No, Greg, dates wikilinked by default are not a good thing when they are badly rendered into ugly html which might be confusing to the few people reading wikipedia who have never used a computer before. There is no end run, we have got to get the developers to fix the code.LeadSongDog (talk) 21:14, 13 August 2008 (UTC) - addendum: Isn't it odd that the "search" function can render dates according to preference when logged in, or as "13 August 2008" when not logged in? Why is it so hard to accept the same thing elsewhere?LeadSongDog (talk) 21:39, 13 August 2008 (UTC)
  • I’d take advantage of a template that formated dates so they display as “13 August 2008” for I.P. users and as “August 13, 2008” for registered, logged-in editors, who set their user preferences setting to the U.S. style. That would be nice for readers. Except, all these tools currently link to trivia. Doing a bugzilla requires someone who is “tight” with the developer community and “hangs” in the relay chat rooms and all that. If anything is to be done, someone like this needs to step up to the plate. Greg L (talk) 22:40, 13 August 2008 (UTC)
Downside: Certain usage usually goes together. "Colour" goes with "13 August 2008". "Color" goes with "August 13, 2008". Writing styles typical of the U.S. military also go with "13 August 2008". If the date style does not match the style of the rest of the article, a certain dissonance is created, which may seem more disturbing than a slightly unfamiliar date style. --Gerry Ashton (talk) 22:56, 13 August 2008 (UTC)
The advocates of a clear, unambiguous "written out" date seem to have made some inroads here. ISO dating is not used in prose and its continued use as part of a citation template is the only reason for its continued use in Wikipedia although "dumping" it as a citation prerequisite should now be a primary concern. Despite a recent comment by an editor or two that claims to use ISO dating in normal correspondence, the problems still exist as to translation by a reader unfamiliar with the system. The "mix" of date styles or formats is especially jarring for a reader. If the majority of the information is in textual form, then a written out date, August 13, 2008 or 13 August 2008 should prevail. FWiW, the one benefit of Wikipedia editing is that it is a fluid and flexible medium that has evolved over time, and that changes such as the ones being proposed can eventually be incorporated. Bzuk (talk) 23:25, 13 August 2008 (UTC).

Some dates are part of reading flow (body text). Some dates are just reference data (citations). The obsession with date consistency got us into this autoformatting mess and readers really do not care that much. I am happy to read 'Fourth of July' in one paragraph and 'September the Eleventh' in the next and I am happy if citations have compact ISO formats. The formats that almost all readers see today are not the problem, just take all the square brackets away. Lightmouse (talk) 00:12, 14 August 2008 (UTC)

Responses from Tony1, from bottom upwards:

  • I go with Lightmouse's call to "just take all the square brackets away".
  • Bzuk: Whether to use ISO dates in cite web et al. is mostly a matter of current editor choice rather than a task for developers. However, because a few editors strongly prefer ISO, I'd be willing to merely recommend against using it in citation-generated dates, as well as the current deprecation of its use in the main text. Better still, I think users need to be educated that there are significant advantages in retaining local control over formats by dispensing with citation templates altogether. The priority, to me, is to ensure that where templates are used, editors have the ability to switch date autoformatting on or off—a simple "link on / link off" field; this includes commonly used templates in infoboxes.
  • Gerry's right on inbuilt dissonance. The worst example is the clutch of articles on September 11, 2001, which renders manually entered DA dates on my display as "11 September 2001", defying the iconic phrasing we all know, including the concomitant "9/11"; that's dissonance for you. This is an example of why we should give editors the power to choose one of the two standard date formats in relation to the topic, rather than delegating this to distant developers in cocoons.
  • I agree with everything Greg L says, except his penchant for the alternative of lobbying to spread the system to an IP-based one. I wonder why it's worth it? Who cares whether it's month–day or day–month, when the month is spelled out? It's soooo trivial and not worth the technical/editorial palahva. Simple and locally controlled is best, for us and for our readers.
  • LeadDogSong: "There is no end run, we have got to get the developers to fix the code" (also Gimmetrow "It would be fairly easy to have autoformatted/linked dates render without links"). I disagree on two grounds. First, the code is not worth fixing, for most of the reasons under the cap that has been promulgated here and elsewhere. Second, the developers at Bugzilla are notoriously resistant to calls for technical change. Witness the failed call for the decoupling of DA and linking at https://bugzilla.wikimedia.org/show_bug.cgi?id=4582 Bugzilla], including, after more than a year, an 88-person petition of WPians (Jan 07). Brion Vibber is on record at the start as dismissing calls for decoupling linking from autoformatting ("this was done in part to encourage linking of dates so they can be used as useful metadata."—12 Jan 06); hmmm ... shows insight—no wonder the attempt is still in abeyance two and a half years later. To take the developers' side, though, it doesn't help when there are multiple suggestions for syntax and technical solutions; as well, developers, especially volunteers, stand to lose if there are any problems, so they're cautious folk. We're also dealing with WikiMedia, not WP; do you know how many sites use WM software? It's not easy.

Again, I ask, why bother to go through so much angst in search of a solution to a non-problem? I can think of lots of better places to direct our programming and editorial talent. The "end run" LeadDogSong talks of is here now: drop it altogether. Simplicity wins, usually. Tony (talk) 00:53, 14 August 2008 (UTC)

Tony1, did I hear the sacrosanct citation templates could be done away with? My heart goes pitter-patter (LOL). In reality, a great number of new editors who have little or no experience with research writing still need to rely on a "system" although either the Modern Language Association or American Psychiatric Association or Chicago style guides can be taught/learned with ease. All of these style guides have a simple to follow format. FWiW Bzuk (talk) 01:56, 14 August 2008 (UTC).
<utters quietly:>Yep, those templates are problematic, and I don't think they save any typing, either (someone correct me if they do). They simply introduce different errors. All our new editors need are a few models to choose from. It's not rocket science.
To me, the citation templates operate in the worst tradition of "garbage in, garbage out" as the users who are unfamiliar with the premises and concepts of bibliographical referencing, simply stick in more errors. The templates were thought to be an aid, they now appear as a major hinderance. FWiW Bzuk (talk) 11:48, 14 August 2008 (UTC).
  • To all: Indeed, I am talking about using “2005-08-06” only in tables and other special-purposes uses where tabular data and tight spacing is important. But for regular prose, there should be no use of “[[2005-08-06]]” because it only benefits we editors, 99.9% of readers see computerese, which doesn’t look nice. Either editors should loose the brackets and write out the dates (“2 August 2001”, or “August 2, 2001”). Either that or the developers should provide us with a formatting tool that benefits all readers. It was unwise for the developers to have provided a tool that produces the intended effect only for registered editors. That decision was simply brain damaged. Greg L (talk) 15:19, 14 August 2008 (UTC)
  • P.S. What I am specifically saying is that even a non-linking formatting tool like {{formatdate|15 May 2005}}, which would yield “15 May 2005” for I.P. users, and “15 May 2005” for European editors, and “May 15, 2005” for U.S. editors, is no good either. All of us editors need to stop thinking Wikipedia is some sort of private preserve where we can employ formatting tools that benefit only us. Greg L (talk) 15:34, 14 August 2008 (UTC)
Agreed - I believe everyone recognizes a date no matter what the format. Do we really need to have "preferences" for date formatting? Are we that sensitive? —Mattisse (Talk) 17:19, 14 August 2008 (UTC)
Wait - explain that. Why isn't formatting the date there any good? Why does formatting the date benefit only us? -- SatyrTN (talk / contribs) 15:53, 14 August 2008 (UTC)
  • SatyrTN, Wikipedia’s date formating options includes a total abortion that should never have been dreamed of, let alone included as a tool for editors to use. Examine the below table and take note of who sees what.
What you type What logged-in registered users see (settings on first row) What others will see*
-- January 15, 2001 15 January 2001 2001 January 15 2001-01-15 No preference --
[[2005-05-15]] May 15, 2005 15 May 2005 2005 May 15 2005-05-15 2005-05-15 2005-05-15
* Non-registered users and registered users not logged in
Note how if we were to code [[2005-08-06]], we (the minority privileged registered editors) would see either August 6, 2005 or 6 August 2005. But the vast majority of readers see only  2005-08-06.

Most registered editors stay logged in. As soon as they don’t see their name up at the top (every 30 days now), we log in again. Very few of we editors ever peruse Wikipedia and look at articles as a regular I.P. user. But 99.9% of Wikipedia’s readers are I.P. users.

So editors *think* they’re doing something wonderful with this stupid format option and, in fact, all we give 99.9% of our readers is the worst of both worlds: they don’t get the written-out dates we privileged editors see, and they get the damned links to mindless trivia that has next to nothing to do with the subject the article is about.

So that is why I advocate that editors should not use this particular date *formatting* option. If editors want “2005-08-06” for a tabular or special use, they should simply write it out. If editors want proper dates with a month written out (which they should want to usually do), they should at least use one of the other date formatting options. Better yet, just choose a style (either “August 6, 2005” or “6 August 2005”) and don’t link it.

This is why I suggested adding a footnote (∆ here) to the table of date formats, to produce this. Greg L (talk) 17:22, 14 August 2008 (UTC)

I'm still confused why every argument on this page is between "leave it the way it is" and "get rid of auto-formatting". Does no one see the benefits of expanding auto-formatting? So that all readers benefit from it? -- SatyrTN (talk / contribs) 19:48, 14 August 2008 (UTC)
I have my preferences turned off, as I am beginning to understand many logged in readers do. I would very much resent "big brother" deciding what my format "should" be. —Mattisse (Talk) 20:36, 14 August 2008 (UTC)
  • SatyrTN: Like other editors here, autoformatting is getting more and more despised because its shortcomings are increasingly seen as overcoming its virtues. I don’t have a problem looking at “August 6, 2007” any more than “6 August 2007”. So the autoformatting tools—for me anyway—are fixing a problem that was trivial to begin with. I can only guess that there may have been an unreconcilable holy war here between a “UK camp” and a “US. camp” and that precipitated the date autoformatting tools. The “autoformatting” tools really do two things: they autoformat, and they link.

    On the subject of links: 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 we’re going to start linking to trivia, that suggests we might as well link to every other possible word; and attitude of “if Wikipedia has a damned article on it, LINK TO IT!!!!” When we over‑link—like with links to trivia—we just turn articles into a giant blue turd. A properly linked article invites exploration and learning by properly anticipating what the typical reader might be interested in further exploring.

    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…

Tokugawa Ieyasu and what the hell he did in 1600 is of no consequence to someone reading a science article. It certainly doesn’t have anything to do with speed of light. If readers want to read a mind‑numbing list of random trivia that occurred on October 21st, they can type it into the search field.

Now, I don’t care to try to impress this basic lesson on technical writing upon every damned 8th-grade editor who takes the time to register and become a Wikipedia editor™®©. They can knock themselves out with the power of wikilinking and link the living shit out of everything they touch their hands to. I don’t want linked dates in articles I’ve worked hard to expand and make professional. I’m happy as a clam as long as MOSNUM doesn’tencourage the use of date *formatting* or linking or permit other editors who like to link dates to wade into nice mature articles and junk them up with even more links.

Finally, I’m all for expanding the use of date formatting so it benefits all readers (I.P. readers too, even though they account for *only* 99.9% of Wikipedia’s readership). But such tools currently don’t exist and the developer who ‘rules’ on this likes his damned blue links to trivia and seems to be entirely pleased with what they do so long as he sees what he likes when he’s logged in. So I’m not holding my breath waiting for properly conceived tools that will truly benefit the vast majority readers. And I simply think it was profoundly unwise for a developer to have made a date format like [[2005-08-06]] where registered edtiors are deluded into thinking they’re making some pretty-looking dates like August 6, 2005 or 6 August 2005 but pretty much the entire rest of the planet sees 2005-08-06. Whoever thought that was a keeno idea should, IMO, be blocked for a week for stupidity. With the use of [[2005-08-06]], not only does the rest of the world see the all-numeric date in in-line body text, but if they have the misfortune of clicking on the blue link, they’re taken to the nothing but random trivia. Greg L (talk) 21:56, 14 August 2008 (UTC)