Wikipedia talk:Manual of Style (dates and numbers)

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive1

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive2

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive3

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive4

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive5

Archive at Wikipedia talk:Manual of Style (dates and numbers)/archive6

See also: Wikipedia talk:Naming conventions (calendar dates)

See also: Wikipedia talk:Timeline standards

Vote at Wikipedia talk:Manual of Style (dates and numbers)/vote

Mav proposed a required margin of 75% for any option to win. While I sympathize with the general idea, we need to be clear about what this means. This vote was started to determine whether the current MoS policy should be changed. If we set a margin of 75% now, that means that the current MoS recommendation will not be changed if none of the options (including the so-called compromise options) wins by such a margin. This is of course very much in my interest, but Mav will probably argue like this: If none of the options wins, we need to change the MoS to one of the inconsistent options, i.e. "everyone can use their preferred style."

I want to make clear that this is not acceptable nor democratic, since it would basically say "This option, which was never reached through an open consensus or a democratic process, is going to win, unless any of those other options reaches more than the 75% margin." If we can agree that any of the options that propose to change the MoS needs to reach a 75% winning margin, that's fine with me. --Eloquence 04:43 24 Jun 2003 (UTC)

James, nobody said that you should get "75% of votes" to win, Mav proposed that the vote margin should be 75%, so if the option "Month Day" had 20 votes, option "Day Month" would require 25 votes to win (20=75% of 25), at least that's how I understand him. I can certainly accept whatever option receives the most votes, though. --Eloquence 05:48 24 Jun 2003 (UTC)

Ok. I must have mis-understood. I have removed my mis-understanding, least there be any . . um . . misundertanding. :-) wikipeace. FearÉIREANN 05:53 24 Jun 2003 (UTC)

Eloquence: For any option to win there needs to be wide support for it. If, as JT says, most users want to only use their style then let them in the articles they work on when and where it makes sense. And the use of the word "policy" is absurd in this whole matter; never, not once, has the MoS ever been policy; It's only function is to serve as a guidebook on the preferred house style for editing (and then it is not enforced on authors but rather is something that interested copyeditors generally follow). It is a loose agreement, soft conventions if you will, on how Wikipedians as a whole think looks and works best. Therefore these style guidelines are useless if they do not have a great deal of support.

Oh and the argument that the previous MoS guideline is the default fallback (or previous "policy" as you stated) is is logically flawed for this simple reason (pardon for the shouting); WHEN THAT GUIDELINE WAS WRITTEN THERE WERE NO REDIRECTS FROM THE [[DAY MONTH]] STYLE TO THE [[MONTH DAY]] STYLE. So, ''of course'' it was the ''only'' guideline we had at that time because only [[MONTH DAY]] links actually worked! But after the initial part of this discussion started 4 months ago a few users went ahead and made all 366 redirects. The vote went nowhere and became such a confusing mess and joke that people on nearly all sides of the argument were able to logically argue that their choice was the actual winner. Now since the redirects were there (thus finally permitting [[DAY MONTH]] links) then the natural fallback was to allow both and prefer neither. The aborted vote effectively killed the old guideline since it became very clear that a sizable group wanted to link dates in the International style. But another sizable group wanted to link using the American style. So we have both. I'm going to put the margin back in now. If you take it out then I'll take out your date since I don't recall you being given the authority to choose such a date (or even call for a vote, for that matter). ---mav 06:04 24 Jun 2003 (UTC)

Sorry, but shouting does not change the facts: The only policy that has ever been formulated in consensus is the one we have now -- to recommend the Month Day format. After all, you only changed the MoS text after this discussion started. All the mucking around with redirects is certainly helpful to adopt another option, but these options can be voted on. If people want the laissez faire solution to win, they can vote for it. It's there. To say that this is going to be the default if no other options get so-and-so many votes is simply an undemocratic way to get what you want. Let's imagine for a second that only 1 person wants that option to win and all other options have a high, but very similar number of votes. According to your logic, the 1 vote option should win anyway because it's now technically feasible to use it. This is obviously not acceptable. --Eloquence 06:17 24 Jun 2003 (UTC)

There you go again with the word "policy" -- MoS is not a set of policies. Please re-read what I said above about what the MoS is. It can only be enforced through the good will and agreement of a clear majority of users -- not a simple majority!. If and when there is no such supermajority for any one thing then we cannot have a guideline on that. You are trying to elevate a loose guideline to a policy that you can then hit people over the head with. IMO that isn't very wiki. I question the validity of this entire vote and have already pointed out its major flaws. We have no process on when to decide to have a vote (expect in the one case were Jimbo said we could do so) and no process on choosing what the particular choices are. This makes this vote invalid and void. And the SHOUTING was a response to your argument that the default fallback is the current "policy." I've already explained why that is a load of rubbish. And on Wikipedia guidelines and policies are, very often updated to reflect current usage. It has already been proven that the current usage is to tolerate both and prefer neither. --mav

If a user consistently unbolded intro text, or changed the birth date format, and reverted to that version even though a copyeditor corrected it, they would be reprimanded and quite possibly banned. We may not call the style guide policy because we accept that users do not consistently follow it, but it effectively is policy, because copyeditors implement it and will certainly get very annoyed if users repeatedly revert to a non-MoS version.

I did not start this vote. If I did, I would go entirely different about it, like I did with m:Article count reform. I do not like the arrangement of options and I do not like that preferential voting does not allow me to express my disapproval of any particular option. I did, however, use this existing vote as a basis because many people have already participated in this discussion.

They did participate in this discussion because there was an increasing desire by some users to use a different date style than was the current MoS recommendation at the time. Currently, most users favor changing the date style. Some users favor allowing both. Because the vote was not organized properly, with no deadline set and no outcome announced, some users have started to simply use their preferred option: Some have used whatever is in an article, others (like me) have used the existing MoS, others have used Day Month. Now the inevitable happened, and two people with different standards (me and Awrel) collided. You now suggest that the "use whatever you want" option is the effective "status quo" and that only through a massive majority (75%) could this ever be changed. This is ridiculous because the policy was never officially changed (through an open process) in the first place, and because it is quite clear and always was from the vote results that a large number of users do not favor the solution of allowing both styles, because, like me, they dislike inconsistency across different wiki pages. You care so much about keeping users -- then please also care about keeping users like me, instead of scaring them away by undemocratically changing our policies to allow more inconsistency in style.

You can question the vote results. You can delete the whole vote page for all I care. I won't set up a new one. If you do not accept the vote, I will insist that the MoS not be changed until a new vote is held, and a new result announced, and I will not accept any other process.

If I was you, I would start campaigning for the inconsistency option. It doesn't seem to be winning. --Eloquence 06:49 24 Jun 2003 (UTC)

Neither was Ralph Nader in the 2002 election. Yet enough people voted for him to make the razor thin margin between Republican Bush and Democratic Gore to result in a tally that fell in Bush's favor (since the great majority of Nader voters were Democrats). This vote is fundamentally flawed because, as you said, there is no way for people to register their disapproval (and the degree of their disapproval) for each choice. Therefore, even though it is approval voting, people tend to go for the style they know and love instead of voting for something that all parties can at least live with (that is why registering disapproval is so important). It has been months since this very poorly-devised poll was set up. I say we bin the whole thing and have you set up a real nice poll (like you did for the article count reform). I hate this poll with a passion -- sorry if that has spilled over. We also should let Jimbo set the ground rules on the vote (margin and timeframe) since he is fair and is the only person with authority around here. --mav

Fair enough -- I'd like to wait with that until mid-July though, because setting up a good vote takes some time. I'll also work with Martin on this because he's usually good at arriving at compromises. Can we agree to leave the MoS page as it is in the meantime? --Eloquence 07:11 24 Jun 2003 (UTC)

The page needs to explain that this vote is being scrapped and the framework for another better vote is being worked on. In the meantime there should be no convention in this regard. --mav

OK, we disagree about the "no convention" part. How about allowing copyeditors to adapt articles to the current MoS, but not to revert the reversion of such a style change? And no, this would not be the official MoS text, only the solution for the transitional period. --Eloquence 07:27 24 Jun 2003 (UTC)

This is all becoming rather a moot point. I was quite serious when I said I would code automatic conversion. A basic demostration is now available at http://piclab.com/wiki/wiki.phtml?title=Astronomy_and_astrophysics . The user preferences aren't done yet, so at the moment it just acts as if the user has requested MD,Y format. I think I was a bit ambitious with my spec: I still haven't done yearless conversion (e.g. DM -> MD with no Y around), and I don't capitalise months yet. Do people think these features are necessary?

Some other points for discussion:

• Should a style be imposed based on what country the user's IP address comes from? Or what their browser setting are?
• Should "session preferences" be implemented, allowing ordinary readers to change their date formats, etc.? What about long-term cookies, to save these preferences?
• Are there any thoughts on whether or not article titles should get the same dynamic treatment?

-- Tim Starling 10:03 24 Jun 2003 (UTC)

I don't think the year should be necessary, because otherwise you'll likely get a mixup of different formats on a page that has an extensive timeline. It will start with "January 10, 2003" and then only say "February 15". However, we need to make sure there are no false positives. It must be case sensitive because "may" is also a verb and "march" is a noun/verb. What is the Y,,,DM format for?

I have no opinion on whether the words/numbers should be links. Requiring them to be links reduces the likelihood of false positives, but some people prefer not to have all dates linked.

Auto-detection for anons would be neat, but "nice to have", i.e. not necessary. Anything beyond that is IMHO overkill. --Eloquence 11:03 24 Jun 2003 (UTC)

Okay, yearless dates will be implemented. The current code requires links, and I think it's easily worth the inconvenience, to reduce false positives. We don't want to get into parsing sentence structure. Plus I'm biased -- I like articles with lots of links. The "Y,,,DM" format simply demonstrates the fact that my regular expressions allow an arbitrary combination of spaces and commas between DM and Y components. It's a shortcut, but I can't think of any problem with it.

What about the number of supported date formats? 3 may be enough. 4 or 5 may be even better. :)

On an unrelated matter, does anyone know what happened to user:-- April? I last saw April 6 months ago. :) Excuse me while I go think of some more... -- Tim Starling 11:41 24 Jun 2003 (UTC)

Unrelated trivia: did you know that Ershi Huangdi[?] was the 2nd August Emperor of the Qin Dynasty?

If we were to standardise, the custom of Wikipedia seems to be to choose the "most international" choice, which is surely day-month-year:

• We have already standardised on "aluminium" and "metre" becuse the relevant international standards bodies use these non-US English spellings. So US-majority-rules is an argument that has been rejected.
• It's clear (from computer national date formats) that day-month-year is favoured by most nations.
• As an example, ISO uses day-month-year on its English language home page.
• Was there not an American film called "Born on the 4th of July"?

(BTW note the false logic at the top of this page: day-month-year is not "British style" it's "most-places-other-than-US" style and so by that argument ought to apply to all articles except those about the US.)

So if there was a standard, it ought to be day-month-year. But it's pointless micro-management to look for consistency in this. Now there are day-month redirect pages there is no problem. Let anarchy prevail.Andy G 18:20 24 Jun 2003 (UTC)

Hasn't this been discussed already? We have a vote in progress to decide this issue. It seems that at the moment, DM,Y is winning. Note that the format "the 4th of July" is not among the proposed alternatives. Does anyone think it should also be supported by the software? Perhaps in the format "the [[4th of July]], [[1776]]"? -- Tim Starling 00:07 25 Jun 2003 (UTC)

I think it'd be good to set up redirects from 4th of July[?] (and maybe July 4th[?] to July 4. But I think you can get away without supporting it in software, because 4th of July[?] vs 4 July probably doesn't excite the same level of passion... :) Martin 20:25 30 Jun 2003 (UTC)

The "turnout" seems rather low to justify any sort of mass changing of articles. Besides I can imagine certain types of writing (including discussion of date formats) where automated global changes to date formats would significantly damage articles. -- Paul (Hobgoblins for small minds)

There's not going to be a mass changing of articles, there's going to be a software feature allowing a user to choose their preferred date format. Dates will be converted on the fly. The default is to leave it as-typed. This is not a pipe dream, this is a piece of code which is mostly already written, and will likely be in CVS by the end of the week. If you think it's going to make articles unreadable, now is the time to say something. Do you think there should be a method of overriding the user preference on an article-by-article basis, say with a special command at the top? -- Tim Starling 00:56 1 Jul 2003 (UTC)

The date/time page is too long for me to edit, but I'd like to put in another plea that Julian/Gregorian and double dates be handled properly! -- Someone else 01:16 1 Jul 2003 (UTC)

Do you have a specific feature request in mind? For example, do you want "[[December 25]], [[1642]] o.s." to be converted to "[[January 4|December 25]], [[1643|1642]] o.s."? Or perhaps "[[December 25]], [[1642]] o.s. ([[January 4|n.s.]])? -- Tim Starling 02:52 1 Jul 2003 (UTC)

I had dates such as 22 Feb[?] 1615/6 and pages such as George Balanchine, Aleksey I of Russia, Peter I of Russia, George Washington, Vladimir Lenin in mind -- Someone else 03:15 1 Jul 2003 (UTC)