Wikipedia talk:Manual of Style (dates and numbers)/Archive 49

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 45 Archive 47 Archive 48 Archive 49 Archive 50 Archive 51 Archive 55

Small deregulation of 'units of measurement' section.

dave souza wrote:

Interesting to note the guidance "Spell out source units in text." I look forward to seeing articles where every reference to acceleration is spelled out as metres per second per second, to give a simple example.

As I understand it, that sentence was not present in the original guidance. The purpose of the guidance is to recommend succinct formats when inside parentheses. That first sentence was an afterthought by somebody. My guidance-about-guidance is that we should only give guidance if:

  • the issue is clear and present in Wikipedia articles
  • the issue is frequent enough to worry about
  • the guidance will result in a change in editor action

The essence of what dave souza says is that it is that it fails the third point. I am happy to propose a deregulation: Can we delete the sentence: Spell out source units in text. ?bobblewik 21:47, 13 May 2006 (UTC)

Prescribing always symbols in parentheses is just as illogical—and nearly as often ignored in actual editing. Gene Nygaard 14:35, 15 May 2006 (UTC)
What about something like: Spell out source units in text unless there is an agreed justification not to—e.g. m/(s·s), not metres per second per second.?—MJCdetroit 15:17, 15 May 2006 (UTC)
Response to Gene: If I understand you correctly, you support the proposal and are also making a proposal to delete the second sentence. That is fine by me.
Response to MJCdetroit: interesting thought but I don't think the added complication will make it pass the guidance-about-guidance. bobblewik 17:28, 15 May 2006 (UTC)
In general, though, this guideline ought to be followed, the text should say "kilometre" not "km", etc. This still ought to be recommended. - Centrx 18:16, 17 May 2006 (UTC)
'In general' is fine. Removal of the guidance does not mean that we endorse the opposite. We are merely deregulating because it fails the 'guidance-about-guidance' criteria above. dave souza pointed out one failing, Gene Nygaard pointed out a second. I don't see what problem the guidance is there to solve. If the guidance stays, we have to accept that the following expression is banned:
bobblewik 09:00, 20 May 2006 (UTC)
I made this proposal to deregulate 16 days ago. There have been no objections and it fails the guidance about guidance tests above. I will remove the following text within two days (unless somebody thinks we should vote about it):
  • Spell out source units in text.
  • Use digits and unit symbols for values in parentheses and for measurements in tables. For example, "a pipe 100 millimetres (4 in) in diameter and 10 miles (16 km) long".
bobblewik 21:11, 29 May 2006 (UTC)
Objection. I for one, like it the way that it is—regulated to say spell out units in text. I just think it looks better (in most cases). —MJCdetroit 03:38, 30 May 2006 (UTC)
That means that the following is banned:
Are you sure that is what you want to ban? bobblewik 18:41, 1 June 2006 (UTC)
I think that there should be a compromise where certain units should always be spelled out in text—kilometers, miles, feet, meters, millimeters, et ceteria. While other units could have exceptions to the rule. Miles-per-hour could be abbrivatied in text to mph and kilometers per hour to km/h or meters per second per second could be abbreviated in text. If deregulated, the following sentence would be acceptable: "a pipe 100 mm (4 in) in diameter and 10 mi (16 km) long". To my eye it is poorly written. It's better to be regulated with exceptions than not regulated at all.
In your examples above, my POV is that the Tokaido should read 589.5 kilometres (366 mi) but in the same right I see nothing wrong with 185 mph (295 km/h).
It's only my POV but I am basing it on ease and clarity of the read. The units that would have the exception would be units that if spelled out would easily "trip-up" a reader. I am sure we could come up with few. What do you think? —MJCdetroit 16:47, 2 June 2006 (UTC)

There should still be a recommendation that units be spelled out in text (though not in alternative-unit parenthesis) in all but exceptional cases like acceleration with "metres per second per second". Certainly, "metres" should be spelled out in text, and I think even "metres per second" should be spelled out in text, but the more complex the units become, the less appropriate it is to spell them out. However, with more complex units there is usually some alternative name, like "newton" to replace "kilogram-metre per second per second". -- Centrx 17:38, 2 June 2006 (UTC)

How about "Consider spelling out source units in text." Rich Farmbrough 08:36 9 June 2006 (UTC).
I would still prefer to declutter the Manual of Style and remove the bullet entirely because it fails my guidance-about-guidance (see above in this section). Symbolic forms in text are reasonable in many cases. However, your (Rich's) suggestion would have much the same effect as removal, so I could accept that. bobblewik 15:00, 10 June 2006 (UTC)
Agree with bobblewik. ..dave souza, talk 19:14, 10 June 2006 (UTC)
What is the purpose of this? All non-complex units should be spelled out. The styleguide ought to recommend style, not that the editor stroke his chin, look up into the sky and ruminate. Regarding "guidance-about-guidance", how would this guidance not be effective? Why could an editor considering whether to use "m" or "metres" not look at this and choose "metres". Under what conditions should they not? -- Centrx 04:33, 12 June 2006 (UTC)

Currency convention and Template:USD

Please see my comment at Template talk:USD. Thanks.

Samsara (talkcontribs) 15:24, 19 May 2006 (UTC)

Gibi/Giga crap

I argue that the gibi-, etc, prefixes for byte measurement are completely useless and only serve to introduce more confusion, and only validate deceptive advertising practices.


1) Computers are base 2 machines, and therefore it is completely inappropriate to use base-10 labelling. 2) There is no SI unit for "byte", and bytes are therefore not part of the SI system, and not subject to its rules. 3) We have 50+ years of using kilobyte as 1024 bytes, megabyte as 1024 kilobytes, etc. Changing this now will only serve to confuse everything. 4) The ONLY businesses that actually use the base-10 values for bytes are hard drive companies that are trying to decieve their customers by artificially inflating their size. By adopting the xxbi- system, we only further validate their deception. 5) It plain just sounds stupid. —Preceding unsigned comment added by MithrandirBooga (talkcontribs) 09:53, 19 May 2006

I strongly agree with you. However, the consensus seems to be against us, which is unfortunate (which is why I stopped bothering to argue--it was like talking to a brick wall). I will also note that the IEC units are confusing people--look at the recent question posted on Talk:Xeon for a perfect example of this. jgp 15:06, 19 May 2006 (UTC)
Also, just a little note--I share your frustration, but calling people "retarded" (as you did in your edit summary) isn't helping anyone. jgp 15:35, 19 May 2006 (UTC)
See these two prior exhaustive discussions. There is really nothing more to add (on either side) that hasn't already been said:
Response to your lines of reasoning:
1. Computers are only base 2, indeed even digital, machines for a very limited definition of "computer." See Computer.
2. The prefixes were contrived by the IEC, not SI, though SI, IEEE, and to a lesser extent NIST have all accepted their usage.
3. The prefixes change nothing, they only provide a specific, unambiguous notation for binary powers.
4. Call it deception if you like, but they are using SI prefixes in the correct manner, regardless of whether or not you agree with the ethics of doing so.
5. You could always come up with your own prefixes, publish them, and convince IEC, SI, and IEEE to use them in favor of the current ones. Failing that, I move that we keep current policy since the prefixes have been and are being adopted by major standards organizations. -- uberpenguin @ 2006-05-19 16:18Z

In looking through the discussion, I see that the arguments for using the new promulgated standards are overwhelmingly erroneous. For example, comparing this to the use of metre would only be valid if we were currently in the French revolution, when the standard was new. For example, in practice, the new standards actually seem to increase confusion, and in no situation would cause problematic confusion as they are used only in the context of computing and the difference is only a small one anyway. I think this should be brought up again. Wikipedia should not be supporting or enforcing the introduction of novel, heavily opposed terms that will be neither helpful nor clear to most readers. These are invented words created nearly out of the thin air, their use is not historically or mathematically necessary. - Centrx 19:22, 19 May 2006 (UTC)

Did you read through the linked previous discussions? Calling the prefixes 'novel' and 'heavily opposed' seem to indicate you haven't. Allow me to rehash some major points here. The IEC prefixes have been adopted or endorsed by IEC, SI, and IEEE, three of the largest standards organizations dealing with fields where these prefixes are most useful. I hardly think that qualifies them as novel. Furthermore, this issue has come up several times in the past (I linked only two), and every time support has been overwhelmingly for continuing IEC binary prefix usage on Wikipedia. Bringing it up for discussion again would very likely result in the same conclusion and a lot more of the same dialog and arguments and counter-arguments that have been brought up before.
Contrary to your impression, the new standards are designed to reduce confusion, since there is already significant confusion when using prefixes like 'mega' in a computing sense. The hard drive situation was already mentioned as the most obvious point of confusion in the situation of misapplying SI prefixes. Most hard drives are marketed using the correct definitions of SI prefixes, which causes confusion when software reports the quantity with the common but non-SI definition. In concrete terms, does a hard disk capacity of "30 GB" mean 30×109 bytes or 30×230 bytes? The definition of 'giga' is ambiguous in this area. Though someone with some experience likely knows that in hard disk contexts, the SI definition (109) is likely, the lay man may have no idea. 'Gibi', on the other hand, is never ambiguous and is therefore desirable to use in appropriate contexts. 1 GiB RAM can only mean 230 bytes of solid state memory. In practice, most authors link the first or more occurrences of IEC binary prefixes where they occur in articles. If the usage of the prefix does cause some confusion for the reader, they can simply click the wikilink and learn what the prefix means.
Everything I have said was more or less already stated in the two links I gave above... Please read through those if you want to continue this dialog, since it's really a waste of both our time to retype the same lines of reasoning used several times past. -- uberpenguin @ 2006-05-19 20:15Z
Also, you may want to scan through Talk:Megabyte#Regarding "Adoption by the NIST" (Oct 2005) and Talk:PlayStation Portable/Archive02#Mebibyte vs Megabyte. -- uberpenguin @ 2006-05-19 20:27Z
I had read the discussion before, and I have read it again, and even before reading it I knew quite well that these were IEC-promulgated standards, though contrary to your statement, there is no evidence that the BIPM has "adopted or endorsed" the IEC prefixes, and the NIST website explicitly states that they are not part of SI. Bytes aren't even SI units. The hard drive manufacturers who apply the SI prefixes with their normal use are not doing so out of some concern over the accurate implementation of international standards, but in the interest of deceiving the customer for profit. While a use of "gibi" is not ambiguous, it is for the vast majority of persons who would misinterpret the ambiguity of "giga" in computing contexts, an unknown word that would lead to confusion. That is, a person who does not understand the computing use of the prefixes, would in encountering the new binary prefixes, find his uncertainty not cleared up but confused further, even if he cared in the first place. One of the discussants states that is, in fact, alright because the reader would then "learn something", as though, first it was the purpose of Wikipedia to distract the reader from what he actually is here to learn, and second, as though finding out a minor textual difference promulgated by a standards body was actual real learning or as worthwhile as learning a conceptual truth or an actually important fact. Nothing said here has discredited the fact that these are ill-conceived and unnecessary pseudo-words recently created by committee that are used only in official documents and some software programs, and not by anyone in everyday usage. Of the paltry 26,000 results (excluding Wikipedia) for "gibibyte" on Google, the first five pages are either definitions from dictionary-like sites or are people talking about the term as a term (and depreciatingly so), not using it as part of the writing. -- Centrx 18:26, 2 June 2006 (UTC)

I feel it's more important to promote precision as a guideline for encylopedic writing than to be overly worried about potentially forcing a reader to learn something (and perhaps even come to understand the entire reason the IEC prefixes exist in the first place). Using SI prefixes with "byte" leads to ugly ambiguity, and the IEC prefixes simply avoid that. I'm not on a mission to change all the articles on Wikipedia, but at least as far as the ones I'm heavily involved in, I see no utility whatsoever in perpetuating a historical accident that has led to somewhat confusing misuse of SI prefixes with non-SI units. I'm convinced that the IEC units are a reasonable solution to this problem and therefore use them in my technical writing. I have no further arguments to make that have not already been made in previous discussions. If you want to move to start yet another vote on this, that is your prerogative, but in absence of any new information or arguments, I doubt the outcome will be different. -- uberpenguin @ 2006-06-03 19:39Z

Oh, as to your comment about BIPM, I realized that the links I gave don't show their mention of the IEC prefixes. My appologies. In the eighth edition of the SI brochure (French, English), there is a side bar in section 3.1 which states in part, "Although (the IEC binary) prefixes are not part of the SI, they should be used in the field of information technology to avoid the incorrect usage of the SI prefixes." (this is the report on the CCU at which this text was adopted; section 4.5) I did not mean to suggest that BIPM has adopted the IEC binary prefixes as part of SI. However, if the quoted statement doesn't qualify as an endorsement, I think we're talking across terms. I continue to hold that the adoption or suggestion of usage by major organizations like IEC, IEEE, ANSI (they simply affirmed the IEEE's adoption), NIST, and even BIPM is more than enough reason to consider the prefixes suitable for technical writing and this encyclopedia. -- uberpenguin @ 2006-06-03 19:53Z

While I can accept that because of this many editors will use the binary prefixes, and that it could possibly be an appropriate stylistic difference (more in the vein of American vs. British spelling), that does not mean that it should be so explicitly recommended as it is currently in the MoS, saying that any edit of binary 'kilo-' 'kibi-' "should be accepted". The MoS should not enforce new standards, the only reason for their acceptance being an appeal to authority promulgated less than a decade ago. Unless and until these new standards become commonplace and commonly understood, using them does not resolve ambiguity, rather it introduces it.

Two other points about that text: First, saying that the SI standard meaning for byte prefixes is the "popular use" is incorrect. The most common use is with the original binary meaning, and it is nearly all only commercial manufacturers who use the other meaning. Second, the MoS even says that confusion may result from using the binary prefixes, recommending that they be piped link to avoid confusion, as though the user should be forced to (after laughing at the name "mebibyte") to go to some other article that is not necessary to finding the information they want in the first article, simply because the article has uncommon usage. -- Centrx 20:36, 3 June 2006 (UTC)

I appreciate your point of view, but must point out that there's little point of even having a manual of style if it doesn't make any firm recommendations. Look back at the early history of the text and you'll realize that it was added because nearly every time some anonymous editor decided to change IEC binary prefixes to SI prefixes in computing contexts, a revert war or argument would erupt. Eventually the discussion moved to the village pump and then here and here the current text was adopted. The current phrasing largely exists to prevent the same conflict with the same arguments on both side being brought up again and again. In other words, I think this issue NEEDS firm wording because it's tiresome to argue this on a case-by-case basis.
It's similar, though not as fundamental, to SI units versus imperial units. It would be horribly unproductive to go through that discussion in every article where the HP could be used in lieu of the Watt, the ft. lb. in place of the Newton. Therefore our MOS states in no uncertain terms that SI units are to be used in scientific articles. Similarly, we should either firmly support or firmly oppose usage of IEC prefixes in computer and informational contexts. Skirting around the issue only makes its mention here an utter waste of space since the MOS would then be useless for preventing the same argument from playing out on talk pages over and over. -- uberpenguin @ 2006-06-03 21:02Z
An additional side note... Can we PLEASE avoid the "their names sound funny" comment? Personally, I think bosons, fermions, phonons, magnons, quarks, and leptons all sound ridiculous in everyday conversation, but one rarely sees marauding bands of physicists hell bent on the discontinuation of the names on grounds of their phonetic curiosity. I'm not so sure that the word "kilobyte" wouldn't have sounded strange to the vernacular English speaker a hundred years ago. -- uberpenguin @ 2006-06-03 21:14Z
Words sound funny not only because they are unfamiliar, but because they have odd sound combinations and have been invented nearly out of thin air without regard for how they sound or look, or where they come from. All of those physical particles are derived from people's names, maybe used for hundreds or thousands of years, and older words. A hundred years ago, an English speaker might mistake a "kilobyte" for a thousand-fold bite, but nevertheless the words are groupings that are derived from older words. Instead, the IEC has engineered a word that satisfies its criteria of non-ambiguity but results in a Frankenstein of a word, patching letter combinations ("bi" with "ki", "me", etc.), without considering it as a speaker or a writer would (at least the physicists are creative). The result is an unusual sound combination in English, with i/eh - b - i/ee - b. In the front, where with mega, m and g are much different sounds created with different parts of the mouth, as are p and t in peta, we end up with m and p with b separated by a short vowel (this is the reason why they say it's pronounced "mebeebyte", because b-short-b is no good, but the same problem occurs here). These physics words are simple, common formations: bose-on, phone-on, magna-on, quark (quirk, quack), leapt-on. They only sound funny because they are unfamiliar.
The policy for American vs. British spelling seems to work fine, where it is established that the status quo style in use for an article is to remain, possibly except in a total rewrite. Any revert war on that matter would contradict the manual of style, just as any revert war would in this matter.
The difference between this case and Imperial units is that basic metric units have been around for more than 200 years, the metric system has been in predominant use by many major countries for more than 100 years, has been used uniformly in science for at least 50 years, and is presently the overwhelmingly dominant system of measurement throughout the world. The words "mebi" and "gibi" didn't even exist 10 years ago and their use is overwhelmingly less common than the other prefixes. Unless and until this usage becomes commonplace, the binary prefixes are totally unlike Imperial vs. metric units. Its only claim is that it is promulgated by a major authority, but it may very well turn out like the decimal calendar (French Republican Calendar). -- Centrx 22:06, 3 June 2006 (UTC)
*sigh* If you truly want to argue that your definition of "funny sounding" is more correct than mine, then I have no idea what I'm doing on Wikipedia. The IEC prefixes are shortenings of phrases like "kilo binary", "mega binary", etc. Complain about the imperfection of their construction all you like; I've spoken my peace. My only suggestion now is to propose any changes you want to this policy and call for a vote. -- uberpenguin @ 2006-06-03 22:23Z

Why link dates?

I know this is probably somewhere is the piles of archives for this page, but can someone tell me why all dates are linked? The articles about days in the year never seem relevant to the article in which they are linked, and the date preferences don’t affect most readers, they are only useful to us editors with accounts. --Arctic Gnome 15:32, 13 May 2006 (UTC)

The Great Date Format Debate was one of the earliest and most widely discussed issues of Wikipedia (ok, so there were a lot fewer contributors then - so a higher percentage participated in the discussions). In the end it was decided to stop the battling by rewriting the very code of Wikipedia so everyone would see only what they wanted to see - and never be exposed to the awful format preferred by the "others". This led to repeated calls to rewrite the code to hide the differences in British and American English, and then to hide all English language differences. However, with the much larger community and the multiplying of rules and guidelines, these calls never gained much traction. Rmhermen 15:50, 13 May 2006 (UTC)
"rewriting the very code of Wikipedia"... Shades of E.E. Doc Smith! Rich Farmbrough 17:49 13 May 2006 (UTC).
Full dates, with day, month, and year are linked so that the Wikipedia software, when displaying the text, converts those dates into the format specified by the registered user in his "Date and time" preferences. Otherwise, I see no reason why dates, such as those which simply have years, should be linked (categorically). This should be changed in the MOS. - Centrx 19:13, 13 May 2006 (UTC)
There are at least 5 pages worth of archives (see "Note on Archives:" at top), from the most recent discussion alone (over the last 2 months), of this issue! But basically, there is a bugfix being worked on (go vote for it), so that dateprefs work without making the dates into links, at which point the guideline will be re-examined.
Consensus on delinking solitary year links is currently unreachable. We're currently suggesting deferring to WP:CONTEXT on a case-by-case basis (see linking years above). -Quiddity 20:55, 13 May 2006 (UTC)
Last time I checked, the guideline was: always link full dates (so that the software can convert the format —oh, I'll go vote for the bugfix, absolutely), link incomplete dates only if the answer to the question: "Does the link provide more info about this article's topic?" (rephrased in my own words) is "Yes, it does". I think that's reasonable and easy. --Gennaro Prota(talk) 23:56, 13 May 2006 (UTC)

At the moment there is no guidance at all on this MoS page about when and whether to link dates. The whole discussion has been removed, not even leaving behind a comment that the question is being debated. There is, however, in the first line about dates, an instruction to "See sections, which follow, regarding when such linking is appropriate", although there are no longer any such sections. Could we please put something back in here to reflect the current consensus or lack thereof, to guide us mere humble wikipedians who are just trying to improve pages? Current usage is inconsistent but includes heavy date linking including isolated years, so the question does need a guideline, otherwise people are likely to understand the apparent rule to be to always link all dates of any kind, which makes for really annoying reading. My own vote would be to not routinely link any dates unless the link is specifically relevant to the article, and instead to use separate notation or software parsing to convert dates to each user's preferred format without requiring links, as discussed in the bugfix. - Mglg 19:22, 17 May 2006 (UTC)

So, is there any reason to categorically recommend the linking of all dates? What are the problems with this revision that need to be resolved? - Centrx 01:32, 18 May 2006 (UTC)

It's extremely disappointing to see the lengths to which people will go to encourage the useless bluing of chronological items that are irrelevant to the autoformat function, e.g., 2004. I have no idea why the MoS has now been altered to eliminate the advice against doing this; it should be reverted forthwith. I note other recent changes that have weakened the quality of prose. Tony 01:49, 18 May 2006 (UTC)

And if you do that, it will be immediately reverted. Things here are decided by consensus, not by the immortal word of Tony. Two things came out of that long discussion - firstly, that there was absolutely no consensus as to what to do about links (thus no consensus to have a "kill them all" guideline), and secondly, that there seemed to be some consensus not to mass-remove them ala Bobblewik. If you choose not to participate in a discussion, don't bitch later when it doesn't go your way. Rebecca 01:53, 18 May 2006 (UTC)
The effect of those changes was not to recommend that all links should be deleted. First of all, what is the reason for linking to the example dates under "Years, decades, and centures" when, as with all words linked according to context, most usages of them will not be linked. As was agreed and implemented in section "Partial dates", "editors should take into account the usual considerations about links, including the recommendations of Wikipedia:Only make links that are relevant to the context." It is not a part of the recommendation of the styleguide that all dates be linked, which is in fact what the Examples would indicate. So, why should they be kept? - Centrx 06:05, 18 May 2006 (UTC)

About links to years

The project page currently says

There is less agreement about links to years. Some editors believe that links to years are generally useful to establish context for the article. Others believe that links to years are rarely useful to the reader. Some advocate linking to a more specific article about that year, for example 2006.

Could I suggest that we add a bit more guidance here? (Translation: I would like a bit more guidance.) As a starting point for discussion, here's the kind of additional advice we could provide.

Linking to recent years is generally not useful to readers. Possible exceptions:
  1. When an event listed in our article for that year is relevant.
  2. "As of now" constructs used to report the current situation.
She published one more novel, Pride and More Prejudice, in 2001.
After his Royalist sympathies became known in 1649, he moved to France.
The magazine had nearly 5,000 subscribers as of 2006.

(I seem to remember the "As of" bit being mentioned in either this MoS page or WP:CONTEXT.)

Cheers, CWC(talk) 20:36, 19 May 2006 (UTC)

You've gone to the link for 1649 and, to present an argument for the use of such links, have chosen a context-sentence that just happens to be relevant to the points at the top of the linked page. This is extremely rare in practice. As for linking 2006, save me from it, please.
I agree that the further back in history you go, the more likely it is that year-pages will contain only a small amount of information, and that it might be slightly more useful in some contexts—particularly with respect to BCE items. However, providing a blanket imprimatur to link simple years without auto-formatting function will open a hornet's nest. The reasons for not linking these items have been covered ad nauseum and, IMV, are valid and important to the readability and appearance of WP text.
Nothing stops a reader from typing a year into the search box.
That the WP techs chose to accompany the auto-format function with a link function is highly regrettable, and has caused no end of tension, in this room and elsewhere. It would be simpler to remove that link so that the auto-format function is limited to just that: autoformatting. Tony 01:39, 20 May 2006 (UTC)
If the information is already mentioned in the article for that year, why link to the year article? It may limit the number of links, but it is not a reasonable recommendation. Aside from it being a rare circumstance that is not very useful, it is reasonable that there will be places where it would not be appropriate to link to the year, even if the event is mentioned in the year-article, and then in other cases it may be reasonable that the year be linked even if the event is not mentioned in the year-article; the proposed policy does not correspond to the actual reasonable usage of year links.
As for "As of year" links, these are usually linked like "As of 2006", not "As of 2006", at least until recently, and as I recall the only reason for this for Wikipedia meta-organization, that is, to help in the maintenance of articles that have time-dependent information that should be updated. Other than this, which is not linking years themselves, what would be the reason for these links? - Centrx 01:59, 20 May 2006 (UTC)
  1. I just found Wikipedia:As of (do-oh!). We probably should link to that essay/article/How-to in this MoS page.
  2. I agree with Centrx that there will be cases where linkifying the year number will be helpful to readers, and also cases where it is not helpful. What I'm asking for here is advice (not Policy, not even a Guideline) to help editors recognise these cases.
  3. I can't come up with a good example of when it clearly would be useful to link to a year. Could someone else provide one?
Cheers, CWC(talk)
There is no consensus on this matter. We've debated it again and again at length, and that is the one thing that has become very obvious. Can we please just drop the "but I want it my way!" comments every fortnight, stop making an issue of it (don't go out of your way to delink, don't go out of your way to link) and get on with building an encyclopedia? Rebecca 05:04, 20 May 2006 (UTC)
The Manual of Style currently suggests a uniformity of linking that implicity contradicts other parts of this same styleguide. You have yet to answer my question regarding this, under section Wikipedia talk:Manual of Style (dates and numbers)#Why link dates?, after reverting an edit. I assume, then, that part of the edit was acceptable? - Centrx 05:32, 20 May 2006 (UTC)

(Back to margin) I apologize for starting this discussion. I had not realised that the divisions were so deep. (I decided to read the recent discussion, but when I saw "171 kilobytes long" I quickly changed my mind.)

(I was expecting/hoping that there would be consensus for a statement such as "most editors agree that linking years is (1) helpful in <some rare situation> and (2) not a good idea in <some other rare situation> but other cases are up to individual editors". But we're a long way from even that much consensus, aren't we?)

Two final questions:

  1. Is there any consensus as to whether an article should be uniform with respect to this? (That is, either all year linked or none linked.)
  2. Is there any consensus about whether it is proper to linkify previously unlinked years, or delinkify years, when editing an article?

These questions are, of course, closely related.

With trepidation, CWC(talk) 18:20, 20 May 2006 (UTC)

Hehe. I'll weakly suggest that the consensus was: 1) Uniformity is not required. Preferably only link things that actually give additional context (which is subject to your own subjectivity). 2) De/linking years is fine as long as it's not automated mass-edits. And post-1970 year links are understood by all to be the biggest over-linked problem that keeps starting all this rucuss, so (in less notable entries (eg not G.W. Bush)) one can generally delink at will. (add more weasel words if that isnt a hesitant enough explanation ;) -Quiddity 03:33, 21 May 2006 (UTC)
There is clearly never going to be unanimity. But there is pretty much consensus (among those who have considered the current situation) that very few date links are useful as links (as opposed to wikifiying mechanisms). The definition of "very few" will vary. Unfortunately there is a folk memory that all dates must be linked, until and unless that ceases, it will not be possible to assume that every linked date has been carefully thought about. Rich Farmbrough 17:10 10 June 2006 (GMT).

"1700s" is not a century, but a decade

1700s is far more often used to refer to a century than to a decade, isn't it? I'm not sure I've ever seen the latter usage, actually. We certainly shouldn't use it to refer to a decade—it would be horribly confusing. We should use "first decade of the 18th century" instead, if necessary. Is it common for style guides to prohibit using such things to refer to entire centuries? —Simetrical (talk • contribs) 03:45, 21 May 2006 (UTC)

I read it not as a prohibition but a note that the article 1700s is about the decade, not the century. Rmhermen 20:04, 21 May 2006 (UTC)

Canadian currency abbreviation

  1. The article says C$ is the usual abbrev.
  2. i've always seen it as CA$ though
  3. and the ISO currency code is CAD (as we're using for pounds sterling in the example)

which should we be using? i glanced through, but could gather no sense of consensus from prior discussion: -/archive29, -WikiProject_Numismatics/Style(-talk), -WikiProject Numismatics/Archive 2#Naming_conventions. -Quiddity 04:48, 21 May 2006 (UTC)

Well, I have no opinion. Whatever's used most. It only seems natural to me to use the full country code; it reduces confusion, I should think. —Simetrical (talk • contribs) 19:48, 21 May 2006 (UTC)
I agree that their doesn't seem to be any agreement. I would prefer the full country code. This would at least make it clear which currency was being referred to. It would also reduce confusion where short codes (e.g. $ or £) are used by more than one currency. There are two potential problems I can see (1) What do we do with historic currencies and (2) I suspect that most people don't know what the country codes. Solving (2) would mean we would need links on the country code. So for example we would have CAD$200. Personally I am attracted to the idea of a template to do this, we could then incorporate other information into the link later (e.g. exchange rates or "xxx in todays money"). However, if implemented this would mean these templates being used in a lot of places and I don't understand the implications for this on the servers. --MarkS (talk) 20:00, 22 May 2006 (UTC)

Linking in examples

Since no one has seen fit to explain why unlinking the dates in the examples is not a natural consequence of the agreed changes to the styleguide, which explicitly states that dates be linked according to the usual criteria of linking—which does not entail that all terms are linked, bar none—least of all those reverting the changes who must therefore clearly have some insight into the matter, I can only assume that it requires a separate discussion section so that those with dim vision can see it.

First, the styleguide, changed according to broad agreement, explicitly states that, "when considering whether such a date should be linked or not, editors should take into account the usual considerations about links, including the recommendations of Wikipedia:Only make links that are relevant to the context." Does this entail that all dates in an article should be linked? The examples indicate that it would.

It may be objected that the examples in the section "Years, dates, and centuries" are in fact examples only of how dates are linked to proper pages. While this is reasonable, an editor coming here for guidance on the formatting of dates, as dates and not as links, which is half the purpose of the page, will find examples indicating that all dates are linked, without exception. At the least, in this section, this should be corrected by a statement more explanatory than telling the reader to search through the whole page for some information about it, when the relevant section is no longer even named "Linking". This section, which opened declaring it describes linking, then proceeds to describe the formatting of dates, per se, not as links.

More strikingly, we find in the section "Dates of birth and death" that all but the "floruit" dates are linked, a reasonable exception but one not carried over to "circa" dates, which are likewise approximate and even less contextually relevant than dates where the event actually occurred in the same year as the others in the linked year-page. Are we to think that the birth of Mo-tzu in the same year as Socrates is contextually relevant to the life of Socrates, and that Visigoths defeating a Celtic invasion of Gaul has any relation to the birth of a Romanian monk, which may or may not have actually occurred in that year? - Centrx 06:44, 21 May 2006 (UTC)

Suggestion: the last extensive changes were agreed with the aid of a draft version comparing origininal and changes (User:Stephen Turner/Date Proposal). That ended a long and elaborated and detailed discussion, where finally there was consensus. So here's my suggestion, either ask Stephen whether you can use his draft to try out further improvements, either copy-paste Stephen's proposal to (e.g.) User:Centrx/Date Proposal, and inform us on this talk page if you have a version that you think meriting consensus.
Also, re. the whole section on dates, I'm sorry to say (would like it different, but it isn't) there are probably no "minor" changes that could be operated there. Words & presence of square brackets have been weighed with an outmost precision to come to a consensus. So any word changing that delicate balance of what is said, what is implied, what is outspoken (etc) would need a new consensus. Anyway if it's an improvement you're proposing I'd be behind you. But I don't see an improvement in your proposals yet. --Francis Schonken 07:17, 21 May 2006 (UTC)
The proposition is rather clear for section "Dates of birth and death": omit links for Socrates, Dionysius Exiguus, Robert Menli Lyon, and Rameses III. For the remainder there are a couple of possibilities that may warrant examples. - Centrx 07:25, 21 May 2006 (UTC)
Me not seeing that as an improvement yet is also clear, I suppose. So, unless you can convince me and/or some other people, a change to the project page in that sense is not appropriate afaik. --Francis Schonken 07:47, 21 May 2006 (UTC)
Why is not an improvement? Let's consider, for the moment, only circa dates. If the circa dates should be linked, when it is not even certain that the person was born or died in that year and the link proceeds to a year-page where the only other events took place in lands in which it would take a whole year even to travel to, how is that relevant to the context of the article about the person, and why should not the floruit dates also be linked? - Centrx 18:26, 21 May 2006 (UTC)
Centrx: I understand and agree with your statements about delinking the examples, as they insinuate/imply that all dates should be linked. But i also understand the neutral/pro-linkers perspective that these are merely technical examples of where to put the square brackets (eg [[1940s]] not [[1940]]s). So i'll concur with Francis, that any changes should be aimed at the introductory text, not the link examples.
I've changed the "Partial dates" heading to "When to link". Perhaps that could be changed further, or the section moved up, or just further clarification added to the introductory text. This is such a contentious/POV issue that draft suggestions should be made here before being implemented, otherwise revert battles will restart. The top section of /archive48 is worth reading through, and the "Stephen's proposal" comparison format is an excellent example to work from. -Quiddity 17:42, 21 May 2006 (UTC)
Note that other sections also contain "when to link" info (full dates; dates on disambig pages not linked;...) also the intro of the first section under the "Dates" header:

This section describes how to link to years, decades and centuries. See sections, which follow, regarding when such linking is appropriate. (my bolding)

That's how it was intended, and how we agreed (that is, until a new consensus would establish - but not before) --Francis Schonken 17:57, 21 May 2006 (UTC)
For the first section, "Years, decades, and centuries", why not replace the sentence "See sections, which follow, regarding when such linking is appropriate" with the sentence "when considering whether such a date should be linked or not, editors should take into account the usual considerations about links, including the recommendations of Wikipedia:Only make links that are relevant to the context" or something similar? - Centrx 18:26, 21 May 2006 (UTC)
Don't get it, why would you do that?...
  1. Would create internal contradictions (full dates are not usual considerations, but the unusual ones presented on this MoS page - the essence is that for full dates a completely different set of recommendations is described than for partial dates)
  2. From what I remember from previous discussions, this would not be something for which a consensus could easily be found. But again, I recommend that if you want to pursue this option that you use a temp page, with the differences explicited, and *with the full text of all the sections* so that other wikipedians at least can check whether you accidently created internal contradictions.
--Francis Schonken 07:53, 22 May 2006 (UTC)
Check the section. Section "Years, dates, and numbers" does not contain any full dates, only partial dates, for which there was consensus to add that line. An abridged version of it could be added to that section, the sections can be integrated and reordered. - Centrx 02:21, 24 May 2006 (UTC)

"spoken quote" example

Do we need this addition to Wikipedia:Manual of Style (dates and numbers)#Direct quotations?...

However a spoken quote on its own is amenable to some format editing, for example if Ken Livingstone says on the radio: "Congestion charges will start at 3 pm and stop at midnight", this can be written in line with the Manual of Style: "Congestion charges will start at 3 p.m. and stop at midnight", as long as it still reads the same.

Either the source is a printed one, per WP:RS, and then we quote it as it is printed (even if that source quotes yet another spoken or printed or whatever source, with or without copying errors), either the source is spoken only (TV or Radio interview or so - which has a bit more WP:V problems and generally would be avoidable but nonetheless...), if it is acceptable as Wikipedia material, then write it down to the best of your abilities, and conforming to MoS (e.g. don't use American spelling for Ken Livingstone - if you know what I mean).

If two separate sources use different spelling, well then use common sense: the most authentic format of the quote (aka, from the most reliable source) would best be used, etc... Jeez does all rulecruft that comes down to "use common sense" have to be spelled out for you guys? --Francis Schonken 15:14, 22 May 2006 (UTC)

Jeez, do you have to collectively insult everyone who didnt write that paragraph?
And yes, kinda. Common sense encoded as law/rules is the only way to prevent actual idiots from continually making the same mistake over and over and over again (and re-arguing it over and over...). See entire history of mankind.
What about WP:BEANS? - afaik that's the way wikipedia handles this, not by spelling out ever more detailed instructions. So, if Rich has encountered a problem in this sense somewhere on an encyclopedia page, let him spell it out here on talk, maybe we can help. Note also wikipedia:how to create policy#Guidelines for creating policies and guidelines, No. 4 "Avoid kneejerk reactions", saying that a one-off incident would in most cases not lead to a guideline/policy update. --Francis Schonken 18:51, 22 May 2006 (UTC)
I don't understand how WP:BEANS is relevant? There is no potential mischief that could result from that paragraph (other than the usual misinfo mischief). Possibly you meant Instruction creep, which No. 4 refers to? But yeah, finding a balance between anarchy and instruction creep is what this is all about. :) --Quiddity 21:27, 22 May 2006 (UTC)
Regarding the actual question, i agree, that is a completely superfluous addition, and should be removed. -Quiddity 18:17, 22 May 2006 (UTC)
Not a big deal, either way. Rich Farmbrough 07:54 9 June 2006 (UTC).


MoS says to prepend and wikilink $ sign with US$ but should I do something similar with a euro sign or just leave it as just an unlinked euro sign? RN 19:44, 25 May 2006 (UTC)

Following the same format here, I suppose it should be linked. This should be added to the MoS. -- Centrx 22:24, 26 May 2006 (UTC)
The € sign does not need an EUR appendage like the US dollar or Australian dollar. The Euro is used only in Europe and there aren't any alternative variants within each country that uses the currency. i.e. There isn't a French Euro or an Italian Euro. —The preceding unsigned comment was added by Rcandelori (talkcontribs) 04:24, 27 May 2006 (UTC)