Misplaced Pages

Revision as of 05:19, 4 November 2008

Manual of Style

This page falls within the scope of the Misplaced Pages:Manual of Style, a collaborative effort focused on enhancing clarity, consistency, and cohesiveness across the Manual of Style (MoS) guidelines by addressing inconsistencies, refining language, and integrating guidance effectively.Manual of StyleWikipedia:WikiProject Manual of StyleTemplate:WikiProject Manual of StyleManual of Style This page falls under the contentious topics procedure and is given additional attention, as it closely associated to the English Misplaced Pages Manual of Style, and the article titles policy. Both areas are subjects of debate. Contributors are urged to review the awareness criteria carefully and exercise caution when editing. For information on Misplaced Pages's approach to the establishment of new policies and guidelines, refer to WP:PROPOSAL. Additionally, guidance on how to contribute to the development and revision of Misplaced Pages policies of Misplaced Pages's policy and guideline documents is available, offering valuable insights and recommendations.

Archives

General 1, 2, 3, 4, 5, vote, 6, 7, 8, 9 (AD/CE), 10, 11, 12, 13, 14, 14a, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 49a, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161 Binary prefixes B0, B1, B2, B3, B4, B5, B6, B7, B8, B9, B10, B11, B12, B13, B14, B15, B16, B17 Years and dates D1, D2, D3, D4, D5, D6, D7, D8, D9, D10, D11, D12, D13, D14 See also Misplaced Pages:WikiProject Dates Misplaced Pages talk:Date debate Misplaced Pages talk:Naming conventions (calendar dates) Misplaced Pages talk:Timeline standards Misplaced Pages:Measurements Debate Misplaced Pages talk:Manual of Style (dates and numbers)/Date autoformatting Complete rewrite of Units of Measurements (June 2008) Dec. 2008 RfC on Date linking/autoformatting (three proposals) Dec. 2008 RfC on Date linking/autoformatting (detailed)

This page has archives. Sections older than 5 days may be automatically archived by Lowercase sigmabot III.

RfC: Linking of dates of birth and death

Template:RFCstyle

Proposal: to add the words

These dates should normally be linked.

to the section WP:MOSDAB#Dates of birth and death, and to link the example dates, so the section would read

At the start of an article on an individual, his or her dates of birth and death are provided. These dates should normally be linked. For example: "Charles Darwin (12 February 1809 – 19 April 1882) was a British ..."

• For an individual still living: "Serena Williams (born September 26, 1981) ...", not "... (September 26, 1981 –) ..."
• When only the years are known: "Socrates (470–399 BC) was..."
• When the year of birth is completely unknown, it should be extrapolated from earliest known period of activity: "Offa of Mercia (before 734 – 26 July 796) ..."

...

Rationale There are some - most vocally perhaps Tony - who believe that pretty much no dates should be linked; and this seems to be what Lightbot was trying to achieve, too. But I don't believe that is the view of the majority. On the contrary, I think the balance of opinion, even amongst those who don't want to see pages becoming a "sea of blue", is that it is useful to have at least some date links on a page, to let people establish a broader context for the times in which a person lived, by clicking their way through the date hierarchy especially via pages like List of state leaders in xxxx or xxxx in the United Kingdom, etc. The proposal that at least the date of birth and date of death in a biographical article should be linked has been made independently in at least four different threads: by Scolaire in the section above #Dates are not linked unless; by Carcharoth in the section above #Concrete examples (year links); by Eleassar, relaying a question raised to him in talk, at WT:CONTEXT#Birth dates?; and by myself at User talk:Lightmouse#Date linking request (birth and death years). It therefore seems appropriate to put up this proposal specifically as a formal well-advertised RfC. Jheald (talk) 11:03, 29 September 2008 (UTC)

• Support. The date page hierarchy, and pages rapidly linked from it, provides a useful link to historical context for biographical articles. The biographical articles are stronger for such context; and the birth date and death date are the most obvious choice of dates to link. Jheald (talk) 11:03, 29 September 2008 (UTC)
• Oppose. (1) You're linking an anniversary day and month that is useless to our readers (please demonstrate some that are useful, and not just a magic carpet for discretionary browsers); and many editors will confuse this with the old autoformatting function. (2) Did you mean to "nowiki" the laborious constructions above that are concealed behind the piped linking ((] ] – ] ])? I'm sure this will go down very well with editors, who who will not only have to memorise how to do this, but will have to actually do it in every article. (3) You haven't demonstrated why it is worth forcing editors to make a link to a year page (birth/death): while it might be possible in a few rare instances to argue that the year of death page is vaguely useful (e.g., 1963 for the death of JF Kennedy, but even that example demonstrates how the fragmented facts about JFK in that year are better in the JFK article itself, or a daughter article on the assassination). (4) The "year in X" links are fine, except that concealing them behind what looks like a useless year-link is self-defeating, isn't it? Already, at least one WikiProject says not to use them. MOSLINK recommends the use of explicit wording to overcome the concealment. Tony (talk) 11:49, 29 September 2008 (UTC)
  • No, this has nothing to do with autoformatting. I'm proposing that such dates - the year, and the day - in the opening words of a bio article should be linked, end of story; something a number of other editors have also raised. The principal value being for the context that these links, and onward links from such pages, allow readers to click through to and explore.
    I'm not talking about "Year in X" articles, I'm talking about the bare year articles themselves. And I'm not intending to particularly mandate the   characters - they were there already, so I just left them. My proposal is very simple: as a rule, the days and years in those opening words should be linked. I want to see where the balance of the community rests on that question. Jheald (talk) 12:06, 29 September 2008 (UTC)
• Oppose, per many of Tony's comments, and just the fact that these year articles (much less day of the month ones!) don't provide useful historical context, they provide an often enormous list of trivial crap. If a large and well-organized WikiProject were capable of producing actually useful year articles that summarized the truly notable happenings in those years, I could maybe see the linking of years (only) for birth/death/establishment/disestablishment dates (only, for the most part). The problem with this though is that editors will see them linked in the lead sentence and then go around linking them all over the place, and we'd be pretty much back where we started. — SMcCandlish ‹(-¿-)› 11:55, 29 September 2008 (UTC)
• Support. I opposed delinking dates in the first place and I still do. -- Necrothesp (talk) 11:57, 29 September 2008 (UTC)
• Oppose, I've never seen much sense in date linking, and links to day-of-the-month articles result in triviality amost by definition. Fut.Perf. ☼ 12:41, 29 September 2008 (UTC)
• CommentSupport. I agree this is a good question to work out. My question is whether we should use what I think you are proposing, the well known and much disliked, "link to the day of year", "link to the year" (which is why people are asking about autoformatting), or if we should be suggesting {{Birth date|yyyy|mm|dd}} and {{Death date and age|yyyy|mm|dd|yyyy|mm|dd}} which provides protection against lightbot and allows for more flexibility in the future. As for those who oppose the "trivia dumping grounds", I suspect that if the links are to specific types of narrowly defined data (such as Births on January 15, 1900 or People who share a birthday on 15 January) most people would be fine with that. dm (talk) 12:43, 29 September 2008 (UTC)
  • As an aside, I changed this to a support. The templates I mention are real templates, and dont need any development. dm (talk) 03:47, 1 October 2008 (UTC)
• ;;partial Strong Support. In that case, the first example should be ""'''Charles Darwin''' {{DL|y=1809|m=February|d-12|mode=eng}} – {{DL|y=1882|m=April|d=19|mode=end}}) was a British ...", with the details of the template worked out later. (And yes, if the question is whether the dates be linked in the lead sentence, my answer is strong support.) Disagree with secret links to 1990 births or 15 January birthdays / January 15 birthdays (if, for no other reason, we'd need staff monitoring which of the latter is linked to.) — Arthur Rubin (talk) 13:22, 29 September 2008 (UTC)
  • Changed to neutral on the day of the year, even if the day of the year article is the one that links back to the person, and the year article does not, because of inadequate notability. It should, however, be pointed out, that ] ] would block autoformatting, and the only consensus we have is that autoformatting is bad, not the linking. — Arthur Rubin (talk) 01:58, 18 October 2008 (UTC)
• Support with comment There are many who do click and want to click on a date link to look at a reference in context, and whether that is trivial, banal or whatever is not my business, nor mine to judge. I can understand that someone may wish to click on a link to find out the context of a date of birth to the world around them at the time. Do I do it? No. Should it be allowable? Yes. For instance a child born during a battle in the local area, or being named Victoria, and that being the date of the coronation of Queen Victoria, or some other event that may have an effect on that person's environment. This information can be quite relevant. So the issue then becomes managing it, and making it useful. Is there 'overlinking' on dates, most definitely, and the information should be most specific, however, the request is specifically for Dates of Life. With regard to the comments about triviality ... for goodness sake, the difference between trivia and excellent knowledge is solely your own virtual framework and environment. If some people thrive on trivia, good luck to them, WP is here for all types. Not asking for extreme, let us find the median position. -- billinghurst (talk) 13:54, 29 September 2008 (UTC)
• Oppose. Linking some, but not all full dates in the article will be confusing. I don't see birth and death date-linking to be valuable at all. Most biographies do have categories for year of birth and death that would get your average browser to the year page anyway. Karanacs (talk) 14:21, 29 September 2008 (UTC)
  • Query to Billinghurst: Your assertion that many people click on and want to click on a date link seems unlikely—do you have sources for this? Tony (talk) 14:42, 29 September 2008 (UTC)
    • Thanks Tony. wikt:many Anecdotally from reading, especially the commentary when it was on User_talk:Lightmouse; some (light) discussions with genealogists, who are a little date focused. I too would love to see evidentiary information about date links and whether they are followed or not. If someone has the right wand to produce that data, it would be lovely. To Karanacs the proposal is just Dates of Life, not all dates. — Preceding unsigned comment added by Billinghurst (talk • contribs)
    • For tony to make such a statement that it is unlikely only proves that he is not paying attention to the comments being made against delinking of dates. I have stated on several occasions (as have others) that I do click on dates (sometime only to see if the article is associated to the date). As for evidence I recommend that someone does a query on the toolserver for all the date articles and see if the hits reduce over the next few months as more and more articles have the dates delinked. I believe we will find a marked reduction in the traffic to those date articles do to their delinking.--Kumioko (talk) 16:27, 29 September 2008 (UTC)
• Oppose. It adds complexity and I just don't see the value. Haukur (talk) 14:59, 29 September 2008 (UTC)
• Support I never agreed with Tony's 2 dimensional view that date linking is bad. Misplaced Pages is a 3 dimensional database of articles and is not bound by the 2 dimensional rules of a paper article. If we have an article in[REDACTED] that is linkable to an article then we should link to it (whether ir directly relates or not). That doesn't mean that it should be linked 4 or 5 times but it should be linked and the birth and death dates to me are reasonable. If we go along with this delinking of dates argument that tony presents then next we will be delinking the city and state of birth, military ranks, allegiances and any other link that is not directly related to an articles content. I think that this date argument sets a very ugly precedent. Additionally, given the volume of arguments for and against this venture it should be obvious to everyone (regardless of how they feel about whether dates should or should not be linked) that this does not meet consensus, regardless of how the vote previously came out.--Kumioko (talk) 15:06, 29 September 2008 (UTC)
• Strong but partial support I believe linking the year to the bare year articles for births and deaths in bio provides useful context information. I'm actually in favour of linking years (decades etc.) where ever the historical context is significant to the subject of the article, even if the subject itself is not significant to the period of time linked. However, I am not as convinced of the value of linking the month and day, especially since those links would not seem to add much context without the year. PaleAqua (talk) 15:53, 29 September 2008 (UTC)
  • Two simple points to PaleAqua: (1) Where was the consensus to link these items in the first place? (2) No one is suggesting a slippery slope to no wikilinking; rather, I sense that the motivation is the direct opposite: the encouragement of a stronger wikilinking system through the avoidance of extremely low-value dilutions. Tony (talk) 16:45, 29 September 2008 (UTC)
    • 1) Rhetorical statements are unhelpful. Where is so much of the information and documentation of templates, convention, etc. Wikis evolve, we are talking about a controlled evolution. 2) No, you are correct, no slippery slope suggested, it was Dates of Life only. Low value to you, statements to the contrary by others that dates of linking are not of low value seem to be ignored or derided as of low value. :-( billinghurst (talk) 17:12, 29 September 2008 (UTC)
• Support. It certainly does no harm. Also usefull for lovers of trivia. Let readers decide what they want to read. G-Man 19:55, 29 September 2008 (UTC)
• Oppose. One of the dates in the example, 26 July 796, would be displayed to those who have selected the "2001-01-15T16:12:34" date format preference as 0796-07-26. The unique format in the preference menu clearly defines this date as an ISO 8601 date, even though that term does not appear on the menu. Also, the discussion leading to the implementation of date autoformatting makes it clear this format was intended to be ISO 8601. ISO 8601 requires dates to be in the Gregorian calendar, and requires mutual consent before information exchange partners exchange any date before the year 1583. Since the date 26 July 796 is in the Julian calendar, both requirements are violated. --Gerry Ashton (talk) 20:15, 29 September 2008 (UTC)
  • Comment: this discussion is not intended to be about autoformatting. This is about hard-linking of the dates, rendered as written, which is how 99% of readers will see them. If there are bugs in autoformatting, then there are bugs in autoformatting. User beware. But we shouldn't let the tail wag the dog. The question is, regardless of autoformatting, should these dates be linked? Jheald (talk) 20:50, 29 September 2008 (UTC)
    • You're wrong. Every date linking discussion is always about date autoformatting until the date autoformatting cancer is excised and incinerated. --Gerry Ashton (talk) 21:38, 29 September 2008 (UTC)
• Oppose Just because an article exists on Misplaced Pages that can be linked to, doesn’t mean it should be linked to. Links should be topical and germane to the article and should properly anticipate what the readership will likely want to further explore. Linking of years (1982), isn’t germane most of the time and should be limited to intrinsically historical articles like French Revolution—in which case, the linked dates would be older, like 1794. What the bot is doing that I find really valuable is the de-linking of dates (October 21). If someone was born on that date in 1982, no one gives a damn if “On this date in] 1600 - Tokugawa Ieyasu defeats the leaders of rival Japanese clans in the Battle of Sekigahara, which marks the beginning of the Tokugawa shogunate, who in effect rule Japan until the mid-nineteenth century.” This isn’t not proper technical writing practices. Greg L (talk) 20:25, 29 September 2008 (UTC)
  • I just want to clarify that just because you don't "give a damn" doesn't mean knowone does. If knowone cared then there would be no need to have a On this day section in the main page.--Kumioko (talk) 20:39, 29 September 2008 (UTC)
    • Indeed. Maybe some people are interested in who else shared the same birthday, or that an English rugby union star was born on the feast-day of the patron saint of McDonalds. If WP has these pages, I think it's inappropriate to presume that because WP:YOUDONTLIKEIT, nobody else should be allowed to find them. Jheald (talk) 20:50, 29 September 2008 (UTC)
      • To Jheald: So you cite WP:YOUDONTLIKEIT. That’s sort of a “if it’s blue, it must be true” argument; if there was a WP:I REALLY REALLY LIKE IT AND IF AN ARTICLE EXITS ON WIKIPEDIA, IT SHOULD BE LINKED TO essay, I might “prove” my point. To Kumioko: I have no problem with the “On this day…” on the main page because all readers know what they will be taken to if they click on a link; they aren’t Easter eggs. And to both of you: This isn’t an issue of right or wrong; it’s a grey area centered around the issue of not desensitizing readers to our blue links through excessive linking. These are links to trivia. Too few readers, after they’ve stepped on these date land mines, want to bother with them any more. Greg L (talk) 21:24, 29 September 2008 (UTC)
        • I agree that date and year links can be land mines and unhelpful to readers. Let's make that clear first. However, I seriously doubt that links that are clearly birth and death years will mislead readers in your "land mine" sense. Take this example: "Charles Darwin (12 February 1809 – 19 April 1882) was a British ..." In my view, people may wonder what the links are, but when they click on them will realise "ah, an article on the year, that makes sense". They will then know this when they see it on future articles, and either click through as desired, or ignore them. What they won't do, in my opinion, is click on the link and think "oh, I was expecting an article on this person's birth" or "oh, I was expecting an article on this person's death". i.e. when clearly linked in a specified and limited context (birth and death years), year links are not Easter egg "land mines", and they are not excessive linking (two links per biographical article). Carcharoth (talk) 04:17, 30 September 2008 (UTC)
          • I have observed that articles tend to be more heavily linked in the lead section; birth and death dates also tend to be the first to appear after the subject's name. Linking to these date articles would strongly contribute to the strong sea of blue in the opening paragraphs. While death dates may be consequential in certain cases, the only possible exception birth dates being generally a non-event is Jesus Christ, and nobody knows JC's exact birth date or year anyway, so I think this is a red herring of a debate. 219.78.19.154 (talk) 15:47, 2 October 2008 (UTC)
            • Well, Alfonso XIII's birth was probably an event... Anyway, your overlinking argument is a good one. If we are to link some dates in a biographical article, then it would make sense to link birth and death dates, but doing it in the lead is not very good. If we say "do it only in an infobox", plus get rid of the autoformatting, then I like it better. -- Jao (talk) 15:57, 2 October 2008 (UTC)
              • Actually, Darwin's birth did involve a minor event; it was the same day as Abraham Lincoln's. I should prefer to have this trivium availabe behind a link to restarting the proverbially WP:LAME edit war about whether it should be in the lead...
                More seriously, the year of birth does provide context, and would provice more if the year articles were better. On medieval articles, it is often of some interest on what saint's day a given person is born; and so on. Septentrionalis PMAnderson 02:59, 3 October 2008 (UTC)
• Support - While I believe that most dates should not be linked, I believe that, in biographical articles, dates of birth and death would serve as helpful links. We link to the biographical articles of persons born on a particular date on that date's article, so why not link back to the date from the biography? – PeeJay 20:45, 29 September 2008 (UTC)
• Oppose - I don't see the value of linking the dates of birth and death. The previous objections to all date linking still seem to apply. Day-of-the-month linking is still trivial even when the date is someone's birth date EdJohnston (talk) 21:17, 29 September 2008 (UTC)
• Oppose - extremely low value links. If someone is interested in the "context" of who else was born on September 12, they can type those few characters into the search box themselves. These are trivial connections that clutter articles needlessly. Ground Zero | t 21:42, 29 September 2008 (UTC)

*Oppose - I have yet to see any argument that comes close to convincing me that these date links provide any sort of relevant context. Yes, they provide context, but the context is so general that it seems useless to me. And yes, I have heard the argument that "just because it seems useless to you, doesn't mean it's useless to everyone." This is a valid argument, but only to a point. Linking every word in every sentence to Wiktionary would probably be more useful than this, in my view. And I don't think that one would get any massive rash of support, either.--Aervanath lives in the Orphanage 22:31, 29 September 2008 (UTC) (Changed !vote: see below)

• Oppose That would just makes everything more complex. Besides, I have yet to read a convincing argument on why date-of-birth and date-of-death links are necessary to aid the reader's understanding of the article's subject matter. Dabomb87 (talk) 02:20, 30 September 2008 (UTC)
• Support linking of birth and death years once at the appropriate place in an article (with the second option being a formal written support in the manual of style for using the birth and death year categories). Oppose linking of dates as these are, in my opinion, trivial links. This was my position in an earlier thread quoted above, though I may not have made it clear enough. I obviously disagree with those who think birth and death year links are trivial in biographical articles - it is my opinion that birth and death years are integral metadata information for biographical articles. Currently, such information is found either as: (a) plain text in the lead sentence, with some articles still having the dates linked; (b) birth and death date categories; (c) entries in the infobox; (d) entries in the Misplaced Pages:Persondata metadata information. Until the Manual of Style specifically mandates that the information for birth and death years needs to be in a form that can be analysed by computers (ie. metadata - and yes, linking is a form of metadata when used correctly), then delinking birth and death years without checking for the existence of the other metadata is a destructive process. I support reduction of overlinking, and avoiding a sea of blue links, but also support the retention of some form of clickable links to take the reader from biographical articles to our chronology categories and articles. Carcharoth (talk) 03:54, 30 September 2008 (UTC)
• I'd only support this as a reversion to the policy of all date linking, in other words linking dates of birth and death are no more or less valuable than any other date links. Either the standard should be to link all or to link none. - fchd (talk) 05:37, 30 September 2008 (UTC)
  • How so? Can you provide an example? I think the article with wikilink dates, eg. "He was promoted to Captain on 1 March xxxx ..." shows that THE date has only has relevance within the article itself, not to the world events at the time.
    At the moment, the issue with much of the discussion is the value judgments rather than relevance or usefulness. Many say it is of low-value where it means it is of low value to them. Whereas many of those supporting, say they find it useful, and they find it is of relevance for their research. I understand my biases, I would like the nay sayers to consider that it this is about relevance and perspective, not their values. --billinghurst (talk) 08:54, 30 September 2008 (UTC)
• Oppose Firstly, we already have consensus that wikilinking of dates is deprecated, so having this as part of the guideline would be a seriously retrograde step, and make a mockery of it, IMHO. Secondly, I would would be somewhat horrified at extensive wikilinking of birth and death dates: the vast majority of biographies I have come across have had these dates linked, and I just feel that these links add nothing to any of the articles. What I am talking about includes EIIR, where the only date I would probably retain is the date of coronation; I might also consider linking the dates of death of Mao Zedong and John F. Kennedy and other leaders who died in office, or other world figures who died at the height of their influence - for example John Lennon. However, we already have articles on the Coronation of the British monarch, Assassination of John F. Kennedy, and Death of John Lennon, which renders the linking unnecessary in the examples given, also proving Tony's point. I would say that even Albert Einstein's birth and death dates are but biographical facts which add little significance to the world if linked to date and year articles. If somebody really wants to look up 18 April 1955 for a context surrounding Einstein's death, they can just as easily type it in the search box or the address bar. It seems to be rather bureaucratic to oblige editors to add wikilinks to these whilst removing all the other wikilinked dates, when there is so much to do here on WP. Ohconfucius (talk) 10:29, 30 September 2008 (UTC)
  • Consensus was only reached after being repeatedly opposed. Tony simply kept resumbitting it until it reached consensus. I have been editing for a couple years on WP and I have never seen any change that has been so hotly contested as this. Your right though in that consensus was reached, now it is up to all of us to refine the details of the decision so that it best supports the project overall. I can live with the decision that dates should not be linked (although I don't agree with it per se) but I do think that certain key dates such as birth and death should be allowed.--Kumioko (talk) 18:27, 30 September 2008 (UTC)
• Support I use birth and death links all the time , as well as links in other key dates to get an historical context to what I am reading. Misplaced Pages year articles give a continuous timeline of what else was going on in the world at the time an event happend. They provide useful context and background and allow the reader to get immersed into a particular historic point in time. They are an invaluable resource unique to Misplaced Pages. Removal is a retrograde step. Lumos3 (talk) 12:43, 30 September 2008 (UTC)
• Weak support. Clearly some readers do find these useful, and the wide support for doing it can be seen in the fact that it has been so widely done. (If it had been introduced by bot, of course, this would not follow, but I see no sign that it has been.) We encourage multiple ways of linking articles together; categories and nav templates and links; this is merely another. I would much more firmly support weaker wording; but it is already established that normally means most people do, but you don't have to even for FA and GA, which should be weak enough. Septentrionalis PMAnderson 14:58, 30 September 2008 (UTC)
• Support, in the cases described by Carcharoth above. A less obvious way I have found links useful is to use them to see what is linked to a given article & the birth/death dates are one important way this works. Further, until this latest push to delink all dates, no one ever raised the issue that linking birth/death dates was unnecessary. I believe it deserves an exception -- & the spirit of ignore all rules more than justifies us to make an exception to any rule when the exception improves the encyclopedia. -- llywrch (talk) 18:50, 30 September 2008 (UTC)
• Qualified oppose. I don't have a strong opinion on the autoformatting question. Personally, I've always thought that our readers were smart enough to correctly read a date whether it was presented as 19 Jan 2008, Jan 19, 2008 or 2008 Jan 19. But I know that others disagree and I don't feel strongly enough about it to argue.
  On the more important question of whether the links are useful as links, I think they should pretty much all be removed. Linking a birth or death day to a page about that day of the month is invariably trivia. While many books publish such trivia, I do not consider that to be a proper function for an encyclopedia. There is nothing encyclopedic about the subject of the biography that the reader can learn by following the link to a page of other trivia that happened on all the other 19 Jans in time.
  The argument for linking years is better but still not strong enough in my opinion. The general argument for it (repeated by several people above) is that it provides historical context and can provide a path to the events which influenced the subject of the biography. I consider this a weak argument because the degree to which a newborn can be influenced by events outside his/her immediate family is trivially low. Child-development specialists will tell you that influences in the first 5-8 years are almost entirely domestic or, at best, highly local. The appropriate link for developmental context would be to the appropriate decate article covering the ages somewhere between 10 and 30. Likewise, a link to a death year tells almost nothing about the person's life except in the rare case where the death itself was a cause for notability.
  My opinion is also influenced by the observation that the "year" pages are massively overlinked. The odds of finding anything useful either on the page itself or by following "what links here" is miniscule. I've never yet followed one of those links and learned anything useful. Rossami (talk) 19:25, 30 September 2008 (UTC)
  • There are at least two uses for these links on birth/death dates. The first is an example of data management -- to maintain the Categories "X births" & "Y deaths". Not everyone who creates or improves an article remembers to include biography articles in these kinds of categories. The second is an example of user friendliness -- it helps end users to determine who was born or died on specific days. There are a lot of people out there who want to know who was born -- or died -- on a given day, & these links help them to research this information. While the Persondata information could offer the same information, so far Persondata is manually created & not yet present in all biographical articles. -- llywrch (talk) 03:13, 1 October 2008 (UTC)
• Support - It's clear that not everyone finds these links useful, but it's equally clear that some do find them useful to a degree, myself included. Jheald's proposal seems like a fair compromise. I'm confident that linking a date or two in the lead won't turn the rest of the article into an indecipherable sea of blue. --Bongwarrior (talk) 19:38, 30 September 2008 (UTC)
  • Bongwarrior: OK, “some” find the links useful. Is that the test you think should be used here: (“some”)? Or do you think it is more than just some, and that the body of readers who would actually want to read through lists of trivia in “year” article are sufficiently numerous to merit yet more blue links in our articles? Greg L (talk) 22:14, 30 September 2008 (UTC)
    • My impression from the above is that those who find the year links useful are coming more from the metadata side of things, rather than the trivia side of things. It would also be nice to have some acknowledgement that birth and death years are less trivial (though some people do clearly see them as still trivial) than a random mention of a year in a random article. And also that linking birth and death years does not contribute to a "sea of blue (links)", but is actually limited to a specific place (at the start of the article) and to two specific links. To expand on the metadata side of things, I'd be happy if a sustained effort were made to bring biographical articles into compliance with some standard style, ensuring that all the articles had Misplaced Pages:Persondata (currently woefully limited in its application - to respond to Kaldari's point below), that all biographical articles had birth and death year categories (or the 'unknown' equivalents) and the "biography of living people" tag (where applicable) and that all biographical articles had {{DEFAULTSORT}} correctly applied (to aid the generation of a master-index, as well as categorisation). If half as much effort went into that as into whether to link birth of death dates or not, then some progress might be being made. As it is, biographical articles account for around 1 in 5 of Misplaced Pages's articles (and, I suspect, a significant fraction of newly created articles), but only a small fraction use Persondata, thousands and thousands of biographical articles are not sorted correctly in the index categories, and many lack birth and death year categories. Many biographical articles also lack the {{WPBiography}} tag on their talk pages. This is one reason why I feel as strongly as I do about not just removing birth and death year links until a proper audit of the biographical articles has been carried out (you can, if you like, think of it as the "date audit" clashing with plans for a similar "biographical audit" and the "date audit" removing metadata links that might have been parsed by the "biographical audit"). To take that one step further, I wonder if the contributions log of Lightbot can be analysed to reveal how many birth and death years were delinked on biographical articles where no birth and death year categories were present? I presume such an analysis would be possible? Carcharoth (talk) 23:27, 30 September 2008 (UTC) I asked Lightmouse here if he can help.
      • I suspect that the actual readers click on links much less than we think they do. There's no evidence for their popularity. The concept of wikilinking is great, but needs to be rationed carefully. No studies have been conducted on readers' attitudes or behaviour in relation to them (for example whether readers tend to read through as much of an article as they're ever going to and then consider hitting a link, or whether they divert on the spot), but common sense tells me that the utility is fragile. Tony (talk) 02:30, 1 October 2008 (UTC)
        • You are probably right. I would say it would depend on a combination of factors: (1) Whether the reader understands the term or knows about the object/event/person linked {information/definition); (2) Whether the article contains sufficient context to explain things and avoid the need for a reader to click away to another article (insufficient article context); (3) Whether the reader is bored by the article they are reading and whether any particular link looks more interesting (diversionary browsing); (4) Whether the reader (after reading the whole article) wants to read up further on a particular topic (discretionary browsing). It depends on the reader to a large extent. What we, as editors, can do, is ensure articles have sufficient context to reduce the need to link, keep articles interesting, keep metadata separate from linking, and try to ensure high-quality linking (linking to good articles and to the correct articles) and to avoid overlinking. If there was ever a push for levels of linking, then one good metric would be "if a fact in article A is mentioned in article B and vice-versa, then that is a primary link", with other links being "background" or "definition" links. Trouble is, there is such a spectrum of reasons for linking, that levels of linking just allows for edit warring. If some software thing like "there is a reciprocal link" could be enabled to turn a link a different colour, that might work, but then too many different colours makes things silly as well. Maybe a preference to only have reciprocal links display? Carcharoth (talk) 03:00, 1 October 2008 (UTC)
          • Wow. Tony, are you saying we should unlink everytihng, not just dates or a few countries, but everything? Links aren't popular? We need to ration them? This certainly explains some of your underlying motivations. dm (talk) 03:47, 1 October 2008 (UTC)
            • It’s not too complex Dmadeo. Links should be judiciously used. They should be highly topical and germane to the subject matter. They should invite exploration and learning for the intended audience. Linking to electron is perfectly fine for the Atom article but would be boring and desensitizing to readers reading up on Planck units; the majority of the visitors reading that article already know what an electron is. The litmus test shouldn’t be whether or not some readers will find it interesting, but whether a good number of the target readership would find it interesting enough to click on. For too long, too many links have been added to Misplaced Pages’s articles because an article existed and could be linked to. But with 6,943,594 articles on en.Misplaced Pages, hundreds of them nothing but date-related trivia, plus even more on Wiktionary, the number of articles to link to is now astronomical and our articles have become excessively linked, effectively turning them into giant, boring, blue turds. Tony is right. We don’t need links to mind-numbing list of randomly-generated trivia nor to common countries. Nor to Manhole cover in the street out in front of Greg L’s house (it’s at a latitude of 47° 39′ 9.1″ for those who would actually be interested in that). It’s not that nobody is interested in clicking on all these links; it’s just that not enough readers are interested in clicking on them. IMO, the reaction to often strive for in readers when we provide links should be “Oh, WOW. I didn’t know they’d have an article on that too!”. Greg L (talk) 04:31, 1 October 2008 (UTC)
              • Precisely. As an aside, you should subst that 'number of articles' template, otherwise in a year's time it will show the number of articles at the time someone reads the archives, not when you wrote this - what do you mean, "no-one reads the archives"? :-) Though there could be a useful distinction, I think, between levels of information on an article and what to link to. Not everyone reading the Planck units article will know what an electron is - that is why you could link it once at the first appearance, and then not link it again (which is normal practice anyway). Consider the reader who wants to click "electron" but can't. They will either edit the article and add a link, or they will look "electron" up by searching for it. But they will be thinking as they do so "why didn't they give me a link to click on?!". But even relevant links are uninteresting to some. The first link on Planck units is units of measurement. I have no interest in clicking on that, but because it is relevant, it stays. So relevance is probably more important than whether a link is interesting. As for links to common countries, there are exceptions to every rule. If you have a list of countries, sometimes it makes sense to link all of them, rather than just some of them. Your "oh wow" point is one viewpoint (and something I agree with). The other is the semantic web - see WP:BUILD. Going too far one way or the other (overlinking and underlinking) could be very damaging. How would you propose to avoid underlinking? Carcharoth (talk) 05:02, 1 October 2008 (UTC)
                • I get the overlink problem, and it seems like a theoretical problem, but not really one in practice. I think it's a lot better to deal with a particular problem article with a simple MOS guideline and involved editors actually editing the articles. Trying to prescribe exactly how to do this in the MOS devolves into lists of what's acceptable and what's not (ie: unlink the United States, but not Australia). I've seen others describe this as overinstruction or instruction creep and I'm starting to feel that there's a small number of vocal people who really like the idea. I find it offputting. dm (talk) 08:19, 1 October 2008 (UTC)
          • This is exactly what I was afraid would happen, in Tony's opinion above[REDACTED] should be nothing more than a publically updated encyclopedia britanica with a few links sprinkled in the article for certain key events. Tony, THIS IS NOT A 2 DIMENSIONAL DATABASE, stop trying to force your narrow views on everyone else. I agree that many articles are overlinked and I understand what you are saying, but having the links is useful and they generate trafic to other articles perpetuating the cycle of publically updated information. If we start stripping off links then one of the primary selling points of[REDACTED] is lost and we might as well buy the paper set when the salesman comes to the door.--Kumioko (talk) 15:12, 1 October 2008 (UTC)
• I have to respond when my views are being misrepresented. I'm sure people aren't deliberately making things up, so I wish they'd check their facts first. (1) I see little value to the readers, and much unnecessary blue in prominent positions, in the linking of common country names, especially English-speaking countries. Just why every single popular culture article should have a link to "British", "UK", "American", "United States", "Australian", "Australia"—I've counted seven to one country in a single article—is quite beyond me. This includes such little-known entities as "India", "China", "Russia", and some European countries. If it's a world map our readers require, they should be made well aware of its existence on the main page, since these country articles swamp the linking reader with huge amounts of information, most of it unrelated to an article topic. (2) It's easy to accuse me, in an exaggerated and frankly quite unfair way, of wanting to strip away all or most links; but in reality, I'm pro-wikilink; I believe people who complain about the notion of a more selective approach to linking are, without their realising it, working against the wikilinking system by diluting the valuable links to such an extent that they are ignored by most readers. It's a great way to kill of a great system. I'm trying to make it more effective. Tony (talk) 15:40, 1 October 2008 (UTC)
  • …there! Another link to mindless trivia. Why? I link, therefore I am. Greg L (talk) 20:47, 1 October 2008 (UTC)
  • P.S. I agree completely when Tony wrote “I believe people who complain about the notion of a more selective approach to linking are, without their realising it, working against the wikilinking system by diluting the valuable links to such an extent that they are ignored by most readers.”  Well said. Greg L (talk) 23:16, 1 October 2008 (UTC)
    • It may surprise you, but I agree with what Tony said as well. We just draw the line at different points. People will always have different ideas about what to link and what not to link. If you want to successfully persuade more people to reduce overlinking, it might be worth expanding WP:CONTEXT to explain things in more detail. I also think part of the problem is that editors often think "do we have an article on this?", and then try a wikilink to find out (using preview). When it turns out to be blue, they check it (hopefully) and then leave the link there because they are pleased that we have an article on whatever. The pleasure at seeing a wikilink work is such that it can be very hard to consciously remove it. By the way, thanks for the essay (I'm sure I've seen a similar essay somewhere before). It makes some interesting points, even if I think putting vomit in the "see also" section is a bit over the top and faintly insulting, as is linking to insanity, but it's your essay. I would add some footnotes to the essay, giving examples of "fascinating" trivia from the October 16 article (I didn't read all of it, but I did skim it), but that might not be appreciated. Seriously, have you ever thought of putting articles like October 16 up for deletion? You sound like you would be happy if they were all deleted. Finally, thanks for the photo of a sewer manhole cover. I've placed this photo in the sanitary sewer article - might as well use the picture to improve an article as well (did you know some people actually collect pictures of manhole covers? See here. There is also some interesting history behind some manhole covers. But then if you are recoiling in horror at the thought of this, then I guess you wouldn't appreciate things like Station Jim either. Carcharoth (talk) 23:56, 1 October 2008 (UTC)
      • Carcharoth: When you write “Seriously, have you ever thought of putting articles like October 16 up for deletion? You sound like you would be happy if they were all deleted.”. Perhaps I might come across that way but, no, I wouldn’t want them deleted. Just de-link them.

        There are just too few people who are reading up on, for instance, Hugh Beaumont (actor), who are really going to read more than the first two entries after they click on a date link. I’d bet that 99.9% of the time, the typical reaction is “Hmmm… that’s what these links do” and then they click their browser’s ‘back’ button. Even with my challenge in the essay, it will be interesting if anyone can ante up and actually read only two of those trivia articles.

        By better anticipating what readers to a given article will be interested in further exploring, we increase the value of the remaining links. If someone is in a mood for long lists of historical trivia, it’s easy enough to type them into the search field.

        And I agree 110% with you when you write about the litmus test many editors use in deciding whether to link or not: if it can be linked to, then link to it. Greg L (talk) 00:35, 2 October 2008 (UTC)

        • Thanks for the reply. I must admit that when I wikilink an article without wikilinks (sometimes a badly written one - the lesson there is that it is better to rewrite the article before wikilinking), I have tended to add links to find out if we have articles on certain things, and only then winnowed the links down to those that are most relevant (and sometimes not even that). I will, in future, be trying consciously to increase the quality and 'impact factor' of any wikilinking I do. I still think that wikilinking tries to do too much - acting as (among other things): a dictionary/glossary; a 'related topics' section; and a further reading section. Carcharoth (talk) 02:08, 2 October 2008 (UTC)
• Support, I think the above comments provide a variety of compelling reasons why editors might want to link dates. What I would actually prefer is for editors to be given explicit discretion in whether to link these dates on any given article. Within the context of the rest of the MOS I think the proposed language is closer to that ideal than the existing text. Christopher Parham (talk) 02:01, 2 October 2008 (UTC)
  • I have five articles in mind already for an insertion of a link to Greg's essay on the sewer cover outside his house. Seriously. Link as much as you can, wherever there's a tiny opening to do so; after all, in today's world, everything can be related to everything else by one, two or three steps. it won't hurt the valuable links.Tony (talk) 02:27, 2 October 2008 (UTC)
    • The October 16 article has nigh on 12,000 incoming links. The same article has some 260 lines/events listed. The 364 links to other date articles created by {{months}} hardly dents the total. There is a serious imbalance here. 'October 16' is only one of 366 such articles with a very similar problematic. I am not saying that all articles should be back-linked from the date page, or that the majority are related to biographical d-o-b or d-o-d, but I would contend it is one valid perspective on the rather pandemic overlinking to date articles. Ohconfucius (talk) 04:22, 2 October 2008 (UTC)
      • If one looks only at links from article space and further exlude the 1199 lists, 1403 titles of the sort "2008 in medicine", 366 days, and 12 months, the count drops to 7425. Still high, but less outrageous. Looking closer at, say, XACML we see it is only linked by the date on a cited reference. I see no reason for linking citation data that is already well-structured, as in this date= field of a cite tag. On the wild assumption that only 2/3 of those are date= or accessdate= instances, that gets the number into a reasonable range.LeadSongDog (talk) 20:21, 3 October 2008 (UTC)
        • I suspect there may be some truth in that assertion, but unless and until all those citation templates are de-linked, we have no way of knowing. Ohconfucius (talk) 08:59, 9 October 2008 (UTC)
• Oppose. For the same reasons as other opposition. Lightmouse (talk) 13:36, 2 October 2008 (UTC)
• Come on… At least a couple of you “Support” editors ought to be taking me up on my challenge. If you can actually read four whole date and year articles, you can be the first recipient of your very own Sewer Cover Barnstar. Are there no takers? Greg L (talk) 03:03, 3 October 2008 (UTC)
  • I looked at your four date articles. I'm sure it's not going to convince you, but they didnt seem that bad. Someone had gone through and organized them enough to make them interesting. They arent going to be everyone's cup of tea, but I'm not sure why you're so offended by them either. I suppose suggesting you just don't look at them won't help either. dm (talk) 23:43, 3 October 2008 (UTC)
    • Dmadeo, when we, as editors, are deciding on whether or not to link a word or topic in an article we are writing, I would suggest setting the bar a bit higher than, “that didn’t seem so bad.” I might even be so bold as to suggest that we set the bar a bit higher so that in many cases, the reader’s reaction to seeing a blue link would be “Way cool… I didn’t expect they’d have an article on that too!” Greg L (talk) 00:14, 4 October 2008 (UTC)
      • With respect, I'd suggest setting the bar at whatever level makes you feel like contributing to articles. That level will be different for me and for anyone else, but thats fine. I encourage you to link however many words you'd like, as long as you dont mind when I do as well. dm (talk) 01:31, 4 October 2008 (UTC)
• Oppose. As illogical, fussy and confusing as when we decided after prolonged discussion not to autoformat, just a short while ago. 86.44.28.60 (talk) 18:36, 3 October 2008 (UTC)
• Oppose as confusing, since the policy is now *not* to link dates without particularly compelling reasons. "saving some curious readers the trouble of typing a year/date into the 'seach' gizmo" just doesn't seem sufficiently compelling. Sssoul (talk) 17:22, 5 October 2008 (UTC)
• Support linking years at least once. It is a powerful way to update and expand the year pages to use the 'what links here' button and see what pages refer to a particular year. Jcwf (talk) 00:23, 6 October 2008 (UTC)
• I don't know if this has been mentioned above, but there is a very relevant CFD discussion at Misplaced Pages:Categories_for_discussion/Log/2008_September_30#Category:Deaths_by_age. Some people, who seem to be in the majority, want to create a series of categories, automatically generated, of Category:Deaths at age 28, Category:Deaths at age 29, and so on. Whether they need the links being discussed here I don't know. Johnbod (talk) 00:59, 6 October 2008 (UTC)
  • I don't see the relevance. — Arthur Rubin (talk) 03:35, 6 October 2008 (UTC)
• Oppose Not needed, per very many above. Johnbod (talk) 00:59, 6 October 2008 (UTC)
• Support (changed from oppose): I've changed my stance here because, while I frankly still can't see how linking of dates is useful, it is clear to me that there is a significant minority of editors who do find it useful. If it's useful enough for even a few editors, then it is something which we should be linking.--Aervanath lives in the Orphanage 04:26, 7 October 2008 (UTC)

• Comment strange recursive reason for changing an opinion. You now support because you have seen a "significant minority" of other editors support? Do your previous convictions not amount to anything? Ohconfucius (talk) 11:18, 25 October 2008 (UTC)

• Support the proposal. I find it useful. Deb (talk) 13:55, 8 October 2008 (UTC)
• Oppose I have been very pleased to see a plethora recently of edit summaries reading "date audit per MOS:NUM" and I strongly feel that no wikilinks should be used for ornamental purposes which is what these links are. The next step would be linking full stops. (<- wikilinked) People are used to these links but they should go. People will get used to not having them, and if in one instance out of approximately 163 times reading the number 2008 they actually wanted to check out that page, they will not be annoyed at having to type it into the search box. __meco (talk) 17:32, 12 October 2008 (UTC)

• Oppose I think it's a horrible idea and will encourage even more pointless wikilinking. I agree with Meco and most of the others who oppose this proposal (sorry, no new reason I can think of for opposing, it's more or less all been said.). Doug Weller (talk) 09:55, 25 October 2008 (UTC)

• Oppose any naked linking of dates and years serves no purpose. So having it at the start of articles is a particular bad idea. --HJensen, talk 21:34, 26 October 2008 (UTC)

• Oppose linking of birth and death dates is trivia and anyone interested can easily type in the dates. Dates are incredibley overlinked as it is. RainbowOfLight 09:07, 27 October 2008 (UTC)

• Oppose-It's unproductive (hardly anyone actually uses these links) and they devalue other important blue links in the article lead--Shahab (talk) 04:46, 1 November 2008 (UTC)

• Oppose- this has been debated to death -- linking these dates will confuse the issue for many editors, who will take the date links in the opening para as licence to link all dates. Let's keep it simple and clean: no dates linked unless htere is a compelling reason. Ground Zero | t 18:53, 2 November 2008 (UTC)

Infobox templates

• BTW: up to this point it's 14 support and 13 oppose (if I counted right and ignoring any weak/partial distinctions). Sounds to me like there's no consensus either for or against this particular point. But it does point out that there is a large contigent of people who do want limited date linking, especially for something such as birthdates. As far as I know, lightbot is not unlinking the {{Birth date|yyyy|mm|dd}} and {{Death date and age|yyyy|mm|dd|yyyy|mm|dd}} templates, so perhaps we can say "In biographical articles, limited use of {{Birth date|yyyy|mm|dd}} and {{Death date and age|yyyy|mm|dd|yyyy|mm|dd}} may be helpful" dm (talk) 08:27, 1 October 2008 (UTC)
  • Strong comment: What this tells me most clearly of all is that we have lots of !votes from incoming parties totally unaware of the rest of debate (over three years worth) and thus largely-to-totally unaware of the negative aspects of date autoformatting. As just one example among many, I doubt that more than a handful of them have considered the fact that around 40% of surveyed articles had inconsistent date formats in them. This is largely because editors assume that the autoformatting just "handles it", and forget that 99.99% of Misplaced Pages's users are IP address readers, not editors, with no date preferences to set, who are all seeing "3 July 1982" in one sentence and "August 7, 1983" in the next – all because autoformatting ensures that most editors themselves simply don't notice the difference. This is happening in nearly half of our articles. That alone is enough to end this debate right now. — SMcCandlish ‹(-¿-)› 05:00, 4 October 2008 (UTC)
  • Do these templates render the dates in bright blue and have all of the disadvantages of the date autformatting system? Tony (talk) 08:33, 1 October 2008 (UTC)
    • Don't you mean the *advantagees* of date autoformatting? - fchd (talk) 08:40, 1 October 2008 (UTC)
      • Um, no, he meant disadvantages. — SMcCandlish ‹(-¿-)› 05:00, 4 October 2008 (UTC)
  • One or other of the two "birth date" templates MUST be used in infoboxes, if the birth-date is to be included in the emitted hCard microformat. Whether or not they link those dates does not affect this; and can be set according to whatever is the final community consensus. One or other of the two "death date" templates will be needed, when the hCard spec is updated to include "death date".Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 10:15, 1 October 2008 (UTC)
    • I don't have the expertise to understand this. What I can tell you is that it's great that many of the infobox templates have recently been modified so they don't augoformat the dates. Tony (talk) 10:48, 1 October 2008 (UTC)
      • In short: the templates are needed for technical purposes (related to metadata). It doesn't matter (for those purposes) whether they link the dates, or not. But people shouldn't be discouraged from using them, because of formatting, as not doing so will break one of the functions of the infoboxes in which they're used. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 11:12, 1 October 2008 (UTC)
        • For the record, these templates are currently not emitting links (since 1 September). Jheald (talk) 11:29, 1 October 2008 (UTC)
          • The birth and death date template age calculation may be wrong for a person who was born under the Julian calendar and died under the Gregorian calendar. They also provide no way to indicate what calendar was used for the dates. --Gerry Ashton (talk) 14:42, 1 October 2008 (UTC)
            • Those are valid concerns (and are being discussed elsewhere, I believe) but are unconnected to the issue of linking; also, such cases seem to be vastly in the minority. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 15:50, 1 October 2008 (UTC)
              • Not being a template programmer, I don't know if the concern can be fixed. I am reluctant to recommend a template that cannot fulfil its intended purpose, and might not be repairable. --Gerry Ashton (talk) 16:07, 1 October 2008 (UTC)
                • Forgive me for jumping in here, but can you explain something in simple terms to me? What is the purpose of the metadata, and the parsing thereof? I have seen countless mentions on this talk page that if dates were linked, such as birth and death, the collection of metadata would be made easier (am I right here - even if this can be achieved through plain text). This maybe the case, and several editors above wish it to be so, but I don't understand why. Maybe this issue isn't relavent here, but could somebody humour me. Dates should/would/could/may (whatever) be linked to allow for the easy collection of metadata. But why? (I'm not criticising metadata, or those who use it - I just don't understand it's purpose.) In anycase, for birth/death dates, is that not what {{persondata}} is for?–MDCollins (talk) 21:47, 1 October 2008 (UTC)
                  • It should be possible to generate a list of every biographical article on Misplaced Pages, along with the biographical data (where known). To do that, you generally need mature and comprehensive metadata coverage. Unfortunately, the maintenance of metadata on Misplaced Pages (en-Misplaced Pages at any rate) lags severely behind the rate of article creation (persondata, as you say, is one of the places where metadata should be placed, but as there are other places as well, such as the hcard format Andy mentioned above, and since persondata is used in only a small fraction of articles, there are problems). Wikilinks are sometimes analysed as a form of metadata, and certainly a mature and well-developed system of date markup would allow for applications. Geographical co-ordinates are given in a standard way - maybe dates should be as well. It is possible to go too far with this, though, since Misplaced Pages is primarily an encyclopedia, not a database (yes, I know the underlying software uses database tables, but I'm talking about the content here). It's a question of getting the balance right. I'm perfectly happy for dates and years to be mostly delinked (with a few exceptions), but the metadata concerns also need to be addressed. Carcharoth (talk) 23:34, 1 October 2008 (UTC)
    • No one was talking about removing those templates, anyway. — SMcCandlish ‹(-¿-)› 05:00, 4 October 2008 (UTC)

a compromise/interim proposal?

this whole date-link thing is like a Passover situation: since the linking of dates for autoformatting is now discouraged, it's pretty clear that bots and scripts that can assist with the huge task of unlinking that kind of date-link are really needed (otherwise it'll take years, and meanwhile editors who imitate what they see will keep adding them). the question is how to designate certain linked dates as *intentional* so that the bots, scripts and/or editors unlinking dates manually will leave those alone. one proposal (discussed here) was to designate such "intentional" date-links by putting them in the "see also" section with nondate words in them to let the bots and scripts know they shouldn't unlink them, for example:

]

]

]

some editors feel that solution is "too much trouble", but it's nowhere near as much trouble as leaving the massive job of unlinking now-deprecated links to be done by hand just because a bot might undo a link someone cherishes - and even if people are doing the unlinking job manually, they still won't know just by looking at them that this link here is cherished by someone, but that link there is free to go.

the editors who want birth/death dates linked at the start of biographies are proposing that that positioning should designate those dates as "cherished/untouchable". the trouble is that bots can't be taught to recognize position in an article; for a bot to understand that it should leave a link untouched, the link needs to include a non-date word. it's not easy to think of a non-date word to insert in birth/death-date links without making the sentences awkward - especially considering that the formatting breaks whole dates up into two parts: ] ].

but: would the people who want to keep birth/death dates linked in the first lines of biographies be satisfied if the years remained linked - at least until we think up a better solution - and the calendar dates were unlinked? i understand that some editors feel the birth/death year links are potentially of interest for the context they might provide, but the calendar-date links don't provide context - they provide what amounts to historical trivia, which (i posit) *is* very adequately relegated to the "see also" section where the few readers who want to know what mishmash of events went on on that date throughout history can easily find it, marked explicity as "].

if that would satisfy the people who cherish these birth/death-date links in bios, then the gallant bots that are waiting to assist with the necessary task of unlinking meaningless/now-deprecated date links could at least proceed with the unlinking of calendar dates. in good Passover fashion they would leave alone any calendar date link that includes a non-date word, like "]. so if some calendar dates that someone cherishes do get unlinked, that can be repaired by adding a non-date word or phrase to the link to shield it from the next "pass" of the bot and moving it to the "see also" section so that sentence flow is not encumbered.

and meanwhile we could think some more about how to designate linked years for "Passover purposes".

i hope someone sees what tree i'm trying to bark up here. Sssoul (talk) 12:52, 13 October 2008 (UTC)

The major problem with this idea is that users should never have to alter their behavior to accommodate the bots; in fact the situation is always the reverse, in that bot behavior should be modified to accommodate the editors. If a bot cannot be written in a way that does not require human editors to modify the way they are editing to make the bot's task possible then the bot should not operate. This also raises the issue of existing articles - requiring users to go back and "mark" existing dates in order to prevent the bot from removing links is placing far too large a burden on the editors. In short : bots are created to make editors' tasks easier, not to force them to do more work. As such I find this proposed compromise - while a valiant attempt at finding a middle grounds - to be unacceptable. The central question that must be answered (and has not yet been significantly explored) is whether or not a retroactive removal of now deprecated links is acceptable to the community. Shereth 17:25, 13 October 2008 (UTC)

thanks first off for recognizing that i'm trying to be constructive. maybe i haven't made idea clear enough, though: i don't mean that any editors would *have* to mark any dates in any special way; everyone would be free just to leave the calendar-date links as they are and let them be unlinked. perhaps that wouldn't be a problem for anyone, since the argument that some date links provide historical context doesn't apply to calendar-date links. it seems plain that the vast majority of calendar-date links were created purely for autoformatting, and since linking for autoformatting is now discouraged/depracated, i don't understand (at all) what the resistance to removing those links is based on.

but meanwhile in case someone really feels it's valuable, encyclopedic, etc, to offer readers a link to ], they *could* create something like that if they wanted to, knowing it wouldn't get unlinked by a bot. (maybe by another human, but that's a different question!)

also, policy/guideline changes very often *do* entail people changing the ways we edit. in this case the policy deprecating links for autoformatting is the source of the changes. the bot is just meant to assist with making them. Sssoul (talk) 17:59, 13 October 2008 (UTC)

If I might just co-opt something you said, which I believe to be the very root of the problem : "The policy deprecating links for autoformatting is the source of the change." This whole mess gets fixed when one very simple question is answered, that question being, "Does the deprecation of certain types of year/date links necessitate the subsequent (and retroactive) removal thereof?" If the answer is yes, then Lightbot should resume, and the above idea can be implemented to allow editors to retain some links with some changes. If the answer is no, then the bot should not be resumed and the fate of now deprecated links left to the editors of the articles in which they are found. That question has yet to receive any community-consensus based answer, and I believe answering it would fundamentally solve the problem. Shereth 18:27, 13 October 2008 (UTC)

... i'm really struggling to fathom how the policy deprecating linking for autoformatting could possibly be interpreted as implying "but let's keep all the existing date links that have no purpose except for autoformatting" - but if that indeed needs clarification, how do you suggest seeking clarification - another RfC? mediation? or ...? Sssoul (talk) 19:07, 13 October 2008 (UTC)

ps: having slept on this ... Shereth, when you wrote "the major problem with this idea is ...", it seems to me that what you actually meant is "my main objection to this idea is ..." your view is as important as anyone else's, of course, but it seems to me that it would be fair to hear some other people's reactions to the idea before labelling one aspect of it a "major problem". Sssoul (talk) 08:53, 14 October 2008 (UTC)

• There is a fundamental flaw with Shereth's argument and objection. Humans are endowed with intelligence, while bots are not. In order for things to be automated/automatable, things have to be done according to a certain logic. Economists realised that long time ago, and Taylor invented the concept of the production line; the vacuum cleaner was invented so we no longer clean floors the same way as before, the same applies to almost every labour-saving device you can think of. To say that humans should go about and be humans in exactly the same way as before is, with all due respects, bollocks. Ohconfucius (talk) 09:23, 14 October 2008 (UTC)

• Re Ohconfucius: You have completely missed my point. See the response by Lightmouse below which catches my sentiment exactly. Software should simplify the tasks undertaken by humans and not complicate it. I do not mean to imply that we shouldn't adapt the way we do things to accommodate new technological advances. If that's what you got out of my statement then you completely misunderstood what I was saying.
• Re Sssoul: Fundamentally I'm not opposed to removing date links. Fundamentally I really don't give a crap one way or the other. What's got me so animated here is that as an administrator I see (and deal with) complaints from sundry editors who have got their knickers in a bunch because of Lightbot removing some links and they do not see any consensus to remove links. I'm somewhat tired of having to deal with/respond to the situation. The reason I am so adamant about getting consensus is so that next time I see someone pitch a fit on a noticeboard about Lightbot removing date links, I can just point to the consensus and be done with it, rather than having this debate re-ignited for the umpteenth time. Shereth 13:51, 14 October 2008 (UTC)

thank you Shereth - so how do you propose seeking the consensus you consider necessary to clarify whether the deprecation of date-linking for autoformatting includes the premise that existing date-links that serve no purpose anymore should be undone? Sssoul (talk) 15:35, 14 October 2008 (UTC)

An RFC, coupled with a clear and unambiguous statement (such as "Does the deprecation of bare year links mean they should be systematically removed via bot") would establish a sufficiently strong consensus on the matter. Shereth 15:41, 14 October 2008 (UTC)

well ... why just "bare year links"? the now-depracated autoformatting also involved calendar-date links. maybe both questions had better be asked at once, to avoid someone saying yet *another* RfC is needed. Sssoul (talk) 15:59, 14 October 2008 (UTC)

The wording was just an example suggestion, naturally it would make sense to cover all of the bases. It might also make sense to keep the individual points as discrete questions to keep folks from getting confused about what they are or are not supporting, but it definitely would make sense to cover all of the issues in a single RFC and get it done with. Shereth 16:05, 14 October 2008 (UTC)

thanks for clarifying. would it be worth making the "by bot" question a distinct point from the other two? as in (and i hasten to add that i don't mean this is exactly how they should be worded):

• 1] does the depracation of autoformatting mean existing calendar-date links that served no purpose but autoformatting need to be systematically undone?
• 2] does the depracation of bare-year links mean the ones that exist need to be systematically undone?
• 3] if yes to either: is it desirable to enable a bot to do the systematic unlinking, or should it be done only manually? (i feel like it would be fair to point out right away that there are ways to "earmark" both kinds of date-link so that a bot would not undo them, but that that part *would* need to be done manually - a link to a brief description of the suggested ways to "earmark" the links could be included.) Sssoul (talk) 17:04, 14 October 2008 (UTC)

I like this approach. Shereth 17:27, 14 October 2008 (UTC)

You make a fair point, Shereth. Software should eliminate, reduce or simplify human tasks. If only such reasoning had prevailed when auto date formatting was proposed ("its easy, all you have to do is link some but not all dates and in this exact way"). Now we have to find a way of clearing up mess attributed to autoformatting. We can do it with or without software assistance. Lightmouse (talk) 17:35, 13 October 2008 (UTC)

I won't disagree with you. I just have to repeat my insistence that community-wide consensus regarding the fate of said deprecated links be cemented prior to taking any sitewide actions. Shereth 18:27, 13 October 2008 (UTC)

Sssoul proposed bot delinking of calendar dates such as ] and described some other features of his proposal. You are making yourself very clear that you think it requires the expressed opinion of many people. That is clear. You are one of the many people and your opinion is valid. Would you personally accept Sssoul's proposal? Lightmouse (talk) 18:51, 13 October 2008 (UTC)

I do not oppose the de-linking of calendar dates, no. This does not mean I personally accept Sssoul's proposal in its entirety Personally I do not see the need for such a compromise if only a demonstrable consensus could be reached one way or another. For all I care you can turn Lightbot loose on calendar dates, as I believe the primary objection that keeps coming up is regarding years, not the calendar dates. Shereth 14:01, 14 October 2008 (UTC)

I do object. Linking calendar dates should be even rarer than linking years, but there are occasions where it adds value. The link from Pope Sylvester I to December 31, St. Sylvester's Day, should stay. Septentrionalis PMAnderson 14:15, 14 October 2008 (UTC)

Pmanderson, do i understand that you see no possibility of designating that particular calendar date as valuable/meaningful in that article by inserting a nondate word in the link - for example ]? Sssoul (talk) 15:35, 14 October 2008 (UTC)

Thanks. That's clear, as usual. I agree that the ideal solution would not be a compromise. I also agree that the heat of the debate is about years. I hope we can both agree that it has been about 'solitary years' (blah blah ] blah blah) rather than the year that was linked to enable autoformatted of a full date (blah blah ] ] blah blah). Thus I would summarise Sssoul's proposal to turn into a bot specification such as:

• leave solitary years alone and delink all other date components/compounds unless they contain a non-date term.

Many articles must be exempted as a whole e.g. date related articles themselves. Lightmouse (talk) 14:51, 14 October 2008 (UTC)

for the record, i personally would also prefer a non-compromise solution - i'm just seeking to help editors who are saying they want to keep particular dates linked as exceptions to the current policy deprecating date-linking. what i'm trying to do is point out how they can "protect" the links they want to keep, if they want to. and i focussed this latest suggestion on calendar-date links because i think there's less distress about/resistance to unlinking those. and i still hope a satisfactory resolution can be found for dealing with year links. Sssoul (talk) 15:35, 14 October 2008 (UTC)

It's already "marked as valuable/meaningful"; it appears three times in the infobox: once as his death, once as the end of his papacy, and once as his feast; also in the sentence: In the West, the liturgical feast of Saint Sylvester is on 31 December, the day of his burial in the Catacomb of Priscilla. At least the last two should be linked; they should not be convoluted to satisfy some piece of MOScruft. Both mean, and should say and link to, 31 December. When MOS ordains bad writing, as it does all too often, it is malfunctioning. Septentrionalis PMAnderson 17:27, 14 October 2008 (UTC)

thanks Pmanderson - i understand what you're saying. i just want to note in passing that no one is suggesting eliminating calendar dates from any articles - only asking whether all how many of them need to be linked to lists of miscellaneous events that happened on the same date throughout history. i understand that you feel this one *does* need to remain linked to such a list for the article to be understood, but that stating explicity in the link what it's a link to would not be okay with you in your opinion be detrimental to the article. Sssoul (talk) 17:40, 14 October 2008 (UTC)

I don't support linking all of them - and have said so at least twice; the effect of this proposal, however, is to link none of them, and substitute (for some of them) see also links which many readers will not know exist. Septentrionalis PMAnderson 18:05, 14 October 2008 (UTC)

okay, i've amended my statement above. i don't understand why "many readers will not know" that there are links in the "see also" section, but please note that the bot will leave intact any date-link that has a non-date word in it, wherever it appears; such links could appear in the body of the text, for example:

• For fuller understanding, see the list of miscellaneous events on 31 December throughout history.

i feel that making date-links that you consider important explicit that way (and eliminating the masses of meaningless date-links) will have the advantage of clarifying for readers that your link really *is* worth following. Sssoul (talk) 06:40, 15 October 2008 (UTC)

• I would still tend to believe that, even in the case of Sylvester, '31 December' is a low value link. It is infinitely more informative to link to the appropriate Name day article, because that's where the true meaning is. Ohconfucius (talk) 07:57, 15 October 2008 (UTC)

i agree, Ohconfucius, but unless i'm misunderstanding Pmanderson, he/she takes a different view. i'm just pointing out that the proposal *does* accommodate date links that an editor truly feels are essential to understanding an article. Sssoul (talk) 08:09, 15 October 2008 (UTC)

• Pmanderson's just pointing out the exceptions which should prove the rule. There will always be exceptions, and we just need to agree on a way of treating them. His opposition to delinking by bot appears to be a bit Luddite to me. Is there an agenda? Ohconfucius (talk) 08:39, 15 October 2008 (UTC)
  • Then I suggest Ohconfucious look up Luddism: opposition to machinery on other grounds than whether it will work better. (Sometimes those grounds are also valid; harmony between editors is a good, and one which bot reversions tend to corrode.) But in this case, the bot will work worse: bots should not be used where there are exceptions, because they will not notice them, and (if reversed) they will come back and edit war for them. (He might also look up exception proves the rule; that saying has two senses: the legal one is irrelevant here, and the other is destructive testing. Enough exceptions blow up the rule.) Septentrionalis PMAnderson 17:42, 15 October 2008 (UTC)
    • Incorrect. The original legal definition is the exact relevant one. If the rule is to remove all square brackets which surround 'mmdd', 'ddmm' or yy, then it is evident that all others are deliberate and are to be kept per the intention of one or more editor. It is a rule which humans and bots alike would have few problems in policing. Ohconfucius (talk) 08:51, 17 October 2008 (UTC)

"bots should not be used where there are exceptions, because they will not notice them" - again, the point of the proposal is that bots and human editors will notice exceptional date-links that are "earmarked" as exceptional. without earmarking them somehow, neither bots nor human editors have any way to recognize that the links ] or ] should remain linked in one particular article. the proposal is pointing out a way to earmark the links that some editor values highly and wants to keep. if a link isn't worth the trouble of earmarking it, one might well ask whether it's really a high-value link. Sssoul (talk) 07:31, 17 October 2008 (UTC)

Quite correct. (2 messages up). The bot should not be run until a convention for "earmarking" dates is established and published, preferably for at least one month. You're claiming your argument is to the contrary? — Arthur Rubin (talk) 21:54, 17 October 2008 (UTC)

• If you accept that deliberately linked dates must be formulated differently to just placing square brackets around calendar dates and calendar years (and you do, don't you?), then it is the very next logical extension that we can resumed delinking all those which are ], ], ], ] or ], delinking by bot or by script regardless of where they are placed. I just fail to understand what further objection there could possibly be to restarting the delinking by bot? Ohconfucius (talk) 23:34, 17 October 2008 (UTC)

(outdent) The root question one should be asking is whether or not the benefit of having superfluous/deprecated links removed is worth the cost of having editors manage the exceptions to the rule, a question I sincerely hope to see answered by a broader audience. Shereth 17:53, 15 October 2008 (UTC)

I think most of us (but probably not all) would be perfectly okay with mass catch-too-much delinkings if we were somehow certain that the relinkings we made to the article after this bot-edit would not be reverted again by bots. Sssoul's suggestion is one way of ensuring that, but I'm sure there are others. -- Jao (talk) 18:04, 15 October 2008 (UTC)

• I agree with the questions/statements by both the above. I see no real issue to un-doing (whether by bot or by script) the date links which currently exist. Those date links which are 'deliberate' need to be re-made in another way which is obvious to bots other editors alike. Ohconfucius (talk) 08:46, 17 October 2008 (UTC)

Anderson and others, if it means so much to you to link a certain anniversary day or year (and I haven't yet seen one that is useful, frankly), then simply make an explicit piped link in the "See also" section. It is as simple as that, and everyone is happy. Tony (talk) 08:38, 17 October 2008 (UTC)

• The 'why should we put in extra effort to put "special" date links in? argument' is exactly as Tony says. If it's worth putting in, it should be worth the effort. And if it's worth the effort, there should be no complaints about needing to put the work in ;-) Ohconfucius (talk) 08:46, 17 October 2008 (UTC)

putting "highly-valued date links" in the "see also" section should perhaps be regarded as a "strongly encouraged option", since some people have expressed anxieties that a reader might somehow miss that section. the main point (as i understand it) is that making "highly-valued date links" explicit by including a non-date word in them is what will designate them (for both human editors and bots) as "this link is highly valued by some editor, not merely a remnant of autoformatting or overlinking". and that "earmarking" will serve its purpose no matter where in the article the link appears. if some editors really strongly prefer to add For more historical context see the list of ] or For whatever reason, see the list of ] to a paragraph instead of to the "see also" section, it's no skin off my nose. bots will leave those links linked thanks to the non-date words they include; human editors might debate whether or not the links are really useful and/or what the best position for them is, but they'll know someone feels they are valuable links. Sssoul (talk) 09:04, 17 October 2008 (UTC)

The (weak) majority that the year of birth and death in biographical articles should be linked seems as strong as the expressed argument (not consensus) that all years should be unlinked. Any bot needs to take into account appropriate exceptions. — Arthur Rubin (talk) 21:59, 17 October 2008 (UTC)

the bot - like human editors - will recognize exceptions that are designated as exceptions. the proposal is to designate "exceptional date links" by adding a non-date word to them. most people in the RfC above were talking about keeping the birth/death-years linked; all they need to do is add something like For more historical context see the list of ]. calendar-date links don't appear to be very high-value to most of the people in the RfC above, but if they are of value to someone, they too can be designated as exceptional. it's not that hard. Sssoul (talk) 23:11, 17 October 2008 (UTC)

That is clearly unreasonable. I would consider "(born July 4, ])" the maximal acceptable tagging to be required of editors if the RfC fails to reach a consensus for exclusion. (The status quo is inclusion.) — Arthur Rubin (talk) 23:29, 17 October 2008 (UTC)

I suppose another alternative I would consider acceptable would to be to include:

• {{for|other events occuring in the year of birth|1943}}
• {{for|other events occuring in the year of death|2008}}

before the lead. You might consider that less acceptable than the status quo, as would I, but I would consider that an acceptable alternative. — Arthur Rubin (talk) 23:33, 17 October 2008 (UTC)

• There are two problems: (1) these examples are piped in a way that conceals from the readers their destination. (2) Positioning this type of link before the lead is far, far too prominent for what is almost certainly a pathway to a sea of irrelevant material. I strongly disagree with this suggestion on both grounds. "See also", when spelt out, is both unintrusive and more likely to be clicked on by readers than a concealed solitary year link in the running prose. I have no idea why you find objection in this solution. Tony (talk) 01:12, 18 October 2008 (UTC)
  • The ] approach violates the guideline for the #See also section (so it needs to be discussed in the talk page of that guideline), and conceals the name of the article. {{for}} in the #See also section seems appropriate for some cases, but the status quo that birth and death years are linked requires a consensus to overturn, as no consensus has been established that year links are always inappropriate. {{for}} in the lead seems an appropriate option, if you insist that the bot should be allowed to run amuck without consensus. {{for}} at the start of any section would be allowed by my proposed modification to the proposal here, and I see no reason why the link to the birth year should be moved out of the lead. I'd accept , as an alternative, the infobox templates emitting the year link, although you seem opposed to that, possibly because you think the year link is misleading, even though it's the actual name of article. If that is your reasoning, I can't understand why you think it's misleading, unless you want to propose moving all the year articles (and handling the templates which link to them). (I've got some idea how many templates are in question, considering the 1900s to 1900–1909 moves.) — Arthur Rubin (talk) 01:43, 18 October 2008 (UTC)

seeking common ground again: it seems that Arthur Rubin's concerns are with year links, not calendar-date links. it appears that so far only two people in this discussion object to unlinking calendar-date links that no one has "earmarked" as exceptional/high-value links. common ground is good.

as for year links (which were not meant to be the focus of this "compromise/interim proposal" - but so be it), they are currently misleading: autoformatting plus overlinking mean year links currently appear to be meaningless. a major part of the point is that when a year link *does* have meaning/value, "earmarking" it will make it explicit what the meaning/value is, which will increase the likelihood that readers might actually make use of the link.

if the main question is where exactly in an article the explicit/earmarked year links should appear, maybe suggesting recommended options would be sufficient: the "see also" section is one possibility - the info-box is another - adding a footnote would be another - a sentence added to the paragraph where the date appears would be another. Sssoul (talk) 12:42, 18 October 2008 (UTC)

Question for Shereth: You say "What's got me so animated here is that as an administrator I see (and deal with) complaints from sundry editors who have got their knickers in a bunch because of Lightbot removing some links and they do not see any consensus to remove links. I'm somewhat tired of having to deal with/respond to the situation." You later talk of editors who "pitch a fit" about the issue at a noticeboard. I have not yet asked, but need to now, whether this is still the case. How many editors have "pitched a fit" in your experience (it's strong language, so we're not talking of just queries and requests for where the practice is mandated, such as I've seen at Lighmouse's talk page). I'd be pleased if you placed evidence before us so we can judge the extent of the problem in numbers, intensity and timeframe. I note that new practices and policies, especially those that change long-established practices, are indeed the subject of emotional reactions by editors who may have spent considerable time and energy in inserting square brackets. But that is not what should concern us here, since editors have now been spared that manual labour in their creation of new text, and are greatly assisted by automated and semi-automated means WRT existing text. Lightmouse and others, including myself, have had considerable success in engaging with editors who query, or even complain of, the removal of the links in question.

Please be more explicit in laying out the evidence so that we can discuss it in informed terms. Tony (talk) 15:16, 18 October 2008 (UTC)

Oppose the linking of days of the year and years. All the proponents of these links should first earn their Sewer Cover Barnstar by honestly and truly accepting the challenge before coming here to run variations of this theme up the flagpole to test the winds. Absolutely no one actually *enjoys* reading these lists of trivia; the nearest I’ve seen an editor get to earning their barnstar was half the full challenge. And the opinion of that editor after that exercise was this: “That wasn’t so bad.” Well… that reaction comes up quite short of a ringing endorsement for linking to these God-awful articles. Step right up, you advocates of date linking; be the first to actually be able to stomach reading four entire trivia articles that the links take readers to. Then come back here and report to the others if your experience was…

1. Worse than having a stick poked into your eye.
2. Worse than getting on a bus and having to sit for five minutes next to a bum who smells like butt crack
3. A thoroughly boring experience and you don’t really expect any reader to actually read more than 2% of what’s there before hitting the “back” arrow on their browser.

After one of you has actually earned the barnstar, then we can talk about why you really think linking to these articles is such a thoroughly marvy idea. Greg L (talk) 01:37, 20 October 2008 (UTC)

Strong support Dates of births and deaths are highly important and linking them will help people find things that are relevent --Hamster2.0 (talk) 18:34, 2 November 2008 (UTC)

Slam-dunk oppose. Articles are overburdened with blue and red links already.--Goodmorningworld (talk) 13:47, 3 November 2008 (UTC)

People should stop opposing delinking. It is minor and inconsequential.

People should stop opposing delinking. It is minor and inconsequential. Lets move on. Lightmouse (talk) 16:16, 25 October 2008 (UTC)

• Indeed. Embracing just a 10% dose of Misplaced Pages:Don't-give-a-f*ckism helps me better deal with the goings-on here. That article is a reminder to observe the two golden rules of life: 1) Don’t sweat the small shit. 2) Everything is small shit. Greg L (talk) 18:09, 25 October 2008 (UTC)

Says you. I won't go in to why (there's already plenty of that elsewhere in the archives here and on this very page), but I disagree completely with date unlinking like this. To whomever had this idea: it was terrible and I fail to see the point in it. —Locke Cole • t • c 19:27, 25 October 2008 (UTC)

• Because sound technical writing practices dictates that links should be topical and germane to the article they’re in. They should anticipate what the readership visiting any given article would be interested in further exploring. The date articles that are linked to are nothing more than lists of randomly chosen trivia that rarely has a thing to do with what subject matter of an article. Too few readers actually bother to read them. If you don’t yet understand this point, try reading these dozen trivia articles, top to bottom, with no skimming: January 1, January 2, July 15, July 16, September 22, September 23, 1925, 1933, 1955, 1965, and 1987. No, seriously. Read all twelve. All of them. Tough, ain’t it? Come on… you didn’t read them. Go ahead… try.

  Furthermore, there were some stupid autoformatting going on where only registered editors could see *pretty* dates like 6 July 2005—the vast  majority of our readership (I.P. users) would see the raw date formats such as 2005-07-06. In other words, it was unwise to have done autoformatting and its associated linking in the first place. Greg L (talk) 20:14, 25 October 2008 (UTC)

  So instead of consistently formatted dates, now we have articles with a mish-mash of date formats (often times in the same paragraph and sometimes even the same sentence). How is this an improvement? Wouldn't it have made more sense to get the date format for IP users changed to something everyone could agree upon? Or better yet, if linking is really your primary concern, come up with a different syntax for date handling that allows dates to be autoformatted and yet remain unlinked. Instead a few people here have seemingly decided their way is the best way and made this a crusade. I really hope this reaches the ArbCom so those involved can be told that making wide sweeping changes without vast community consensus (especially changes which will be visible to our readership) should require the highest involvement of the community. —Locke Cole • t • c 23:15, 25 October 2008 (UTC)

People here have tried for years to get the developers to implement non-linking date autoformatting, to no avail. The devs are either uninterested, or actively resistant. It is not going to happen. Setting a default date format for users who haven't set the preference would fix half of the problem, but would still leave ugly and stupid links scattered through all our articles. Better to rip out date links for now. If unlinked date formatting with a suitable default is implemented later, it will have a new syntax anyway so there is little to be gained by keeping the current broken date links in the meantime.--Srleffler (talk) 05:19, 26 October 2008 (UTC)

Um, no. Nothing that impacts large numbers of articles on Misplaced Pages is “minor and inconsequential” – least of all when it lacks consensus. The simple fact is that date autoformatting was deprecated, and with little opposition once the reasons for it were explained. Since DA is embedded in the link coding and editors cannot remedy that, the only way to remove DAs is to delink dates. The problem is that there has never been any consensus for “de-linking on sight” (or by bot) of linked dates in general, much less “years in XXXX”.

“Deprecation”, by the way, means “no longer using, but leaving the extant coding in place” – not “removal with extreme prejudice”. I hate to say, “I told you so,” but as I pointed out in August, mass removal (by whatever means) of existing DAs would prove highly disruptive. Those who preferred to be rid of them ASAP wikiwide have pursued their agenda and reaped their due reward, so I have no sympathy for them regarding the tremendous amount of time they’re having to waste defending their position. I will note, though, that I proposed in that posting a far less disruptive approach, which remains a practical option, although it won’t achieve the “ASAP” removal that some here so dearly desire.

As for the “historical years”, again, there is no consensus for their removal at all; this isn’t a function of DA. Their removal is due entirely to the prejudice against this type of linking by some of the same folks, who feel that the product is unencyclopedic and useless. Whether one agrees or disagrees with this position, it is the sole reason for their being delinked at this time. The solution to the uproar over delinking DAs – and, separately, historical year linking – is, of course, to find a consensual path forward … not telling others to “just get over it”. After all, without consensus-seeking, it’s just as fair to proclaim, “People should stop delinking existing DAs. It is minor and inconsequential.”. Askari Mark (Talk) 20:31, 25 October 2008 (UTC)

• Well said Askari Mark. So what do we do to clearly identify a consensus? If there was ever anything that is the subject of dispute, it is when one tries to identify when a proper and true “consensus” has been arrived at. I guarantee you, that some editor who is willing to don orange robes and set himself alight over this issue, will set the bar much higher than does the middle-of-the-road editors.

  Trying to settle upon any course of action on Misplaced Pages is damned tough because it has no elected representatives. Further, even if 50 editors weighed in here, any decision that lets a bot loose is guaranteed to result in editors coming here wondering why they weren’t consulted on the matter. As a result of this phenomenon, all decision making on Misplaced Pages tends to be neutered down. For example, it took three years to get Misplaced Pages to stop being all alone in its use of “256 kibibytes” of memory when everyone else inhabiting this pale blue dot uses and understands “256 kilobytes”. Three years. That was a Misplaced Pages:Use common sense sorta issue but change doesn’t come about easily here. Maybe it sometimes takes being WP:BOLD to get anything done. I truly don’t know what is the best tact to take here. Greg L (talk) 21:01, 25 October 2008 (UTC)

  IMHO, a change like this is worthy of notice in Mediawiki:Sitenotice for at least a week (preferably a month). You need lots of input, not just MoS regulars, for something like this. —Locke Cole • t • c 23:19, 25 October 2008 (UTC)

• Locke Cole, in response to your 23:15, 25 October post: Yes, it would have been better to have a simple set of rules governing the choice of date format to use in articles. I had proposed a simple one: Use Euro-style dates (6 September 2008) for most articles but use U.S.-style dates if the article is about or is closely associated with the U.S. (and a handful of U.S. territories that would have been listed). But I lost.

  As for a date format that works for I.P. users, we don’t have those parser functions available for use in templates. At least not yet.

  Note that I am an American. But I am also an R&D engineer and I have noooo problem looking at dates in either format. I’m all about writing to produce the least confusion for the greatest portion of our readership. So an article on Chicago should say “January 20, 1985” (which it does) whereas an article on International System of Units should say “1 August 1793” (which it doesn’t;  it uses American-style formatting and they’re all linked!). Don’t blame me. Greg L (talk) 00:13, 26 October 2008 (UTC)

  P.S. I have my date preferences turned to "none". Every editor should set theirs the same way, otherwise, we aren’t seeing what I.P. users see. Greg L (talk) 00:16, 26 October 2008 (UTC)

• What makes me gobsmacked is why the order of month–day or day–month should matter to anyone. In all English-speaking countries, both are used (the US military, for example, uses day–month, Canada uses both, many newspapers elsewhere use US format despite the predominant use of international). Why do you care so much about it when you readily accept "colour/color" and "travelling/traveling"? It's a mystery. Those who moan on and on about the dispensing with date autoformatting seem to believe that it's quite OK for our readers to be exposed to both types of this binary system (just as for what is essentially our binary spelling system). Please take the emotion out and give that in-house indulgence a rest. We have much more important business to get on with. I ask all responsible editors here to assist in the task of removing date autoformatting. As for the linking of trivial date fragments, far better solutions to the provision of gateways into year pages et al. have been provided on this page. They should be embraced and the year-page crowd should spend its efforts improving woefully bad year-pages, not shoring up millions of unused pathways to those pages. Tony (talk) 01:26, 26 October 2008 (UTC)

• 'In all English-speaking countries' except England! Nimbus (Cumulus nimbus floats by) 01:47, 26 October 2008 (UTC)

• Do you read English newspapers? Um ... please have a look. Tony (talk) 01:54, 26 October 2008 (UTC)

• No, I don't read the papers, in most cases they are a waste of wood but I will take a special trip to the newsagent to see. If I want to know the date I go to the clock in the bottom right of my screen where Microsoft tells me it is currently 26 October 2008. I guess their 'regional' setting option is there for a reason. Nimbus (Cumulus nimbus floats by) 02:12, 26 October 2008 (UTC)

• Actually only irresponsible editors would continue to delink dates while there's an obvious lack of consensus. Such editors should be taken to task for their behavior IMO, up to and including blocking for obvious and continued disruption and violations of WP:POINT. Continuing to remove date links only makes the situation all the more urgent for those opposed to such delinking and creates entirely unnecessary drama (especially considering that such date links are, essentially, harmless; the wiki has somehow managed to avoid collapse while having date links all these years, why the sudden urgency to remove them as fast as possible?). —Locke Cole • t • c 09:02, 26 October 2008 (UTC)

• And your sideline sniping at people who actively work to improve WP is not irresponsible? Let me think about that for a moment. You should be ashamed of yourself, calling for the blocking of those who shoulder the work (unlike you); your view of the situation appears to be based on utter ignorance of the long long debate that has gone on here and ended in consensus in August. You're a Johnny-come-lately, perhaps, who's upset at having missed the debate—sorry, but that's just too bad. Maybe you'd like to bone up on the four information and consensus pages linked to at the bottom of this page. Tony (talk) 00:42, 27 October 2008 (UTC)

Greg, I believe we do have consensus on deprecating DA – or perhaps more accurately, further use of DA (not its removal). Rarely have I seen as many editors come to MOSNUM (or any other policy/guideline talk page) originally in opposition and change their mind once the issue was clearly explained to them. Normally, opinions are locked in and loaded, the trenches grow fuller, and the bombardments more intensive. It’s quite understandable that editors would like to be able to read articles in their preferred format, but the argument that it leaves a mess for non-editors – those whom we write here for – was pretty cinching. I think most editors who were involved in the debates appreciate (or came to appreciate) that it would be preferable for something to be written by the developers to enable this without the use of links. There were several threads posted about this issue in the VPP and I know the members of the more active WPs were aware of it, so I think it’s fair to say that there is consensus for the deprecation of DA.

The problem now being encountered is the one I foretold would happen. First, DA is intimately tied to wikilinking. It’s all well and good to say wikilink only where necessary and useful, but there’s a broad range of opinion on what “necessary and useful” really means; wanton – and particularly semi-/fully automatic removal of links – tramples on these viewpoints indiscriminately and most particularly affronts those who didn’t participate in the discussions because, well, they rarely or never do until something impacts them directly. Each one then demands an accounting and explanation/cajoling/convincing, which even the most ardent advocate of delinking dates has no doubt wearied of by now. In effect, these “delinking activists” have decided to “be bold”, forgetting that that approach works best on the small scale of an individual article, but not on the larger scale of a wikicommunity.

To get past this brouhaha, I think what should be done is something along the lines of what I proposed in August. To wit, pursue the broadest consensus on what to do about delinking DAs; I think Locke Cole’s suggestion that it be advertised via Mediawiki:Sitenotice is excellent advice since it wraps it all up in one intense episode and then we can all get back to using a larger portion of our wikitime in constructive editing. (However, here I would insist that “year in XXXX” be explicitly excluded, for the reasons I outlined to Nimbus above.) This proposal should first describe the consensus for deprecating continued use of DAs and why it was justified (telling the story one more time instead of hundreds of times), and offer two options for community comment approval: a) active removal of existing date links by manual, semi- or fully automated means; or b) permit the editors to decide on the fate of these links on the articles they are active in on the respective talk pages. (Of course, if one of these editors wants to be bold on that particular article, they can, but if protested must then abide by the talk page result.) Is this latter option quick and efficient? No, not at all. But what it does take into account is the variable range of tolerance for change among people in general, and the distaste by the majority for diktats by “outside experts” that meddle in their workspace. I think you’ll find that most of these “involved editors” will opt for removal, albeit not as fast as others would like; the advantage is that these editors will then “own” the decision as much as those who have already wrestled it into existence here at MOSNUM.

As for Tony’s wistful cry for a standard date format, … something about “If wishes were horses…” goes here. Choosing one over the other means selecting a winner and a loser … and, naturally, as in locking pages being edit-warred over, the winner will be the “wrong one”. As for myself, I’d settle for all the formats being consistent within each article. That would be major progress over what we have now. Askari Mark (Talk) 03:41, 26 October 2008 (UTC)

Which "Tony" is this? I think you'd better strike out your accusation, which is in total opposition to my view. "wanton – and particularly semi-/fully automatic removal of links – tramples on these viewpoints indiscriminately and most particularly affronts those who didn’t participate in the discussions"—you make it sound like having lewd, animalistic sex on a busy street; but this is more like hard work, requiring the exercise of fine judgements in some cases as to which format is most appropriate. The simple rebuttal is that from many thousands of date audits, only a handle of editors have complained, and most of them have been easily won over when it's explained. There's one of those on my talk page from a few days ago. You appear to be overstating the opposition, which, to my eyes, is coming mostly from IT-trained editors, who tend not to balance functionality over disadvantage in the way that normal people do. That accounts for the concentration of loud complainants and the silent agreement/support by the vast majority of users (let alone readers). Tony (talk) 08:58, 26 October 2008 (UTC)

Tony, I’m sorry you mistakenly took my “wistful” note as an insult; actually, I was agreeing with you. I’m completely comfortable with the international format and have always preferred it. However, on Misplaced Pages, it’s not what I want – or what you want – but rather what the community wants. Askari Mark (Talk) 22:11, 26 October 2008 (UTC)

Hear hear. If more participants here took such a reasonable approach, I would be more willing to discuss the issue. What I will note is that a blue-linked date doesn't really detract from the article, since there is no chance of the reader being fooled into thinking that the link is relevant to anything but the date itself. I'd also note that article-consistent date formats are double-plus-good, regardless of the status of DA itself, and year linking should be subject to editorial discretion, especially for historical years. Franamax (talk) 04:21, 26 October 2008 (UTC)

Franamax, if you'd engaged with the long debates about the matter, you'd be better informed as to the disadvantages of DA. Tony (talk) 08:58, 26 October 2008 (UTC)

• Thank you Askari Mark, for your careful analysis of the current situation we find ourselves in. I think your above observations best sum up the views “middle-of-the-road” editors who have seen all the evidence. Maybe you and Tony can suggest at a way to move forward. Greg L (talk) 04:55, 26 October 2008 (UTC)

• I think Askari already provided a reasonable suggestion of the way forward: advertise on the site notice before continuing with date delinking. Get widespread community support, then go to work on this (massive) undertaking. Specifically, we should decide on a few options, create a poll, and advertise that on the site notice (or alternately the watchlist notice). I see the options as really being: 1) Deprecate linked dates (with the understanding that removal will be gradual, done during the course of regular editing, not via bots). 2) Eliminate/forbid linked dates (this is (1), but with approval for bots to remove the links immediately). 3) Deprecate linked dates and request that the developers provide a method for formatting dates using some alternate wiki markup/syntax (presumably if this got enough support from the community the devs might feel compelled to make this a priority). 4) Maintain the original status quo, allow linked dates, etc. If there's additional options, feel free to list them (or your modifications). Once we have a list of options we can move forward with a wider community poll/discussion. —Locke Cole • t • c 09:02, 26 October 2008 (UTC)

• A key thing to remember is that there is a trick or two that we can link to full dates without invoking date autoformatting - it would require a bot assistance. Just as the DA implemenetation by mediawiki is convolutes linking and date autoformatting, the arguments on this issue are also convoluted, and sometimes the fact that we can separate the two issues (the dissadvangates of non-reg/non-pref users for DA, and the overlinking due to dates) needs to be stated as well as another option. --MASEM 09:12, 26 October 2008 (UTC)

Um, the reluctance of people to go down that path just might have to do with their feeling that the issue has been discussed to death on this page over the past two years, and resolved. You're flogging a dead horse. Tony (talk) 09:09, 26 October 2008 (UTC)

I don't see anyone else who read the discussion saying it was resolved. And I don't see an explicit consensus, except for date autoformating. — Arthur Rubin (talk) 09:46, 26 October 2008 (UTC)

Date delinking is by definition, minor and inconsequential, it makes no sense to object to such edits. Lightmouse (talk) 10:28, 26 October 2008 (UTC)

OK, I'll bite. If date delinking is, indeed, "minor and inconsequential", it makes no sense to object to such links. — Bellhalla (talk) 12:49, 26 October 2008 (UTC)

er ... Lightmouse is quoting an administrator who recently berated him/her for making "extremely minor, inconsequential edits" (ie date-unlinking edits). Sssoul (talk) 12:56, 26 October 2008 (UTC)

Yup, and it seems like an abuse of admin tools, frankly. The reasons provided are thin and contradictory, and the users involved have seemed to pop up out of nowhere. It's really odd. Tony (talk) 15:02, 26 October 2008 (UTC)

I don't know why some editors who are still discussing this issue, and doing so as if it were a major step back. This debate is not only going round in circles, but I object to editors who are starting 'name-calling'. Specifically, I take issue with "only irresponsible editors would continue to delink dates while there's an obvious lack of consensus. Such editors should be taken to task" – are delinking editors really being irresponsible or just being bold? I defy anyone to to defend why the tons of date links in each article in this category should not be de-linked forthwith. It would be blindingly obvious to all, except those not possessed of it among us, that it's just applying a tiny dose of common sense. Before DA was deprecated, the MOS already said that dates should be linked to only if it enhances the understanding of the article. So even before deprecation, there was little reason for these links to exist. Now that DA nonsense is gone, I say good riddance to these links. Ohconfucius (talk) 14:18, 26 October 2008 (UTC)

Yes, the following editors say that date delinking is minor and inconsequential

• Martinp23
• Tombomp
• xeno

Lightmouse (talk) 17:08, 26 October 2008 (UTC)

I don't really have a solid position on whether or not it should be done - what I know is that delinking dates is minor and inconsequential and should not be the sole purpose of your AWB edits. Don't conflate an AWB misuse issue with a MOSNUM debate. The shaky consensus of delinking dates is another matter altogether, and a can of worms I have no desire to eat from. –xeno (talk) 17:58, 26 October 2008 (UTC)

• Agree, by Greg L. Minor an inconsequential. I think (my opinion; don’t shoot me) that Wikipedians’ reaction has more to do with the perceived insult of “trespassing” where uninvited. De-linking improves Misplaced Pages for regular I.P. readers and increases the value of the remaining links. Greg L (talk) 18:01, 26 October 2008 (UTC)

  Not really.

  • Delinking makes very little difference to IP readers, since they don't run into the problems that autoformatting produces; it would be a relatively minor benefit to them to make date formats uniform within an article, but this can be done without delinking, and delinking will not produce uniformity - it will merely expose inconsistency.
  • One thing that this long discussion has made clear is that there is no consensus on which date links have value. Tony thinks none do; I value a minority; some editors think almost every year link has value as context. In the absence of consensus, we should, as often, leave well enough alone.

  How about it? Septentrionalis PMAnderson 18:15, 26 October 2008 (UTC)

Lightmouse should stop insisting on delinking. It is minor and inconsequential.

Delinking dates is not the sole purpose of my edits. Lets all be clear, linking of dates was implemented without consensus and was a big mistake. It is only now that delinking is accepted that we are cleaning up the mess on Misplaced Pages. Not one pro-linking editor has told me that they run a bot to clean up the many easily fixable but common date errors such as ] and ] ]. Not one. All talk about how great date linking is, no action to make it work. Lightmouse (talk) 18:22, 26 October 2008 (UTC)

We are not interested in motive; we are not ArbCom. But this is an unfortunate combination of falsehoods: linking of dates and autoformatting was indeed accepted by consensus as a technical fix to a behavioral problem. It was, I agree, the wrong solution; it would have been better to have treated it as we have since treated spelling differences: by agreeing to differ.

The road forward now, however, is not another technical fix, imposed while ignoring the functions of linking in individual articles. It is a behavioral solution: again, leave well enough alone, and talk to editors on talk pages, and see if you can persuade them to delink. We have the arguments for that; if WP:Autoformatting does not persuade editors to delink, we should again agree to differ.

It is bad practice to treat everything as a nail because you happen to have a hammer. It is the same bad practice to run a bot because you can. Septentrionalis PMAnderson 18:34, 26 October 2008 (UTC)

And it's good practice to hammer in nails when it makes a product. I'm sure Lightmouse has no intention of pursuing your ridulous demand that editors at each article be "persuaded" to delink: they don't own the article (nor does Lightmouse, but he at least is improving a major fault in our articles). Tony (talk) 00:38, 27 October 2008 (UTC)

Tony may well be right that Lightmouse has no intention of accomplishing his ends by gathering consensus at the various pages. This would be regrettable, and would suggest recourse to dispute resolution.

But this is not a major flaw; indeed, in many articles linked dates are harmless: they have none of the flaws of WP:Autoformatting lists. All that needs is an article with one or two dates, after 1752, in positions (like the birth-death list in the lead) where the inconsistency between British and American formats on final punctuation does not arise. Septentrionalis PMAnderson 15:05, 27 October 2008 (UTC)

Please give a reference to the consensus for linking. Lightmouse (talk) 18:37, 26 October 2008 (UTC)

It was proposed, and immediately accepted, by Tim Starling in 2003; see archive 6 of this page. It was still consensus in April 2007 (sufficiently so that no-one, in a long discussion on dates and how they should be linked, suggested that it might be changed); see Archive D4. That's more consensus than any poll is ever likely to be. Consensus has changed since. Septentrionalis PMAnderson 19:13, 26 October 2008 (UTC)

O come now, that's the least convincing case I've heard for this C-word. Tony (talk) 00:22, 27 October 2008 (UTC)

Why thank you; that is to say that a rule which has been on a guideline page for four years and was demonstrably non-controversial during that period was never consensus. Will you hold to that postion the next time some of the hare-brained notions that infest MOS are challenged by a mere speaker of English? Septentrionalis PMAnderson 14:52, 27 October 2008 (UTC)

Just to add a voice to the debate: Because of the serious problems with linked/autoformatted dates, I feel that full linked dates should be delinked on sight. The best option is to do so with a bot that is set up to make all dates in an article use a consistent format. Linked years before 1900 should probably be reviewed by a human. The vast majority of them are linked in error, but in rare cases it may be beneficial to put an article's topic into historical context.--Srleffler (talk) 03:01, 27 October 2008 (UTC)

What serious problem are you speaking of? And why is unlinking the dates and removing the autoformatting as it currently exists the better choice when there are other ways forward? —Locke Cole • t • c 05:26, 27 October 2008 (UTC)

Please make a proposal; we could use one. Septentrionalis PMAnderson 14:53, 27 October 2008 (UTC)

Proposal made below, please chime in if you have anything else. =) —Locke Cole • t • c 02:26, 30 October 2008 (UTC)

The problems with linked dates have been well discussed here and elsewhere. If you're participating in this discussion you should really already know what they are. The worst problem in my opinion is that they have caused date formatting for readers to be inconsistent, by obscuring the true date format from the editors who would otherwise be likely to fix format problems. Delinking dates is a first step to beginning to clean up the mess. Editors won't fix what they can't see. Almost as bad is the fact that linked dates clutter articles with worthless links such as March 20 and 2008. Links to dates without years, and to year articles are almost never appropriate, especially for years in the 20th century and later. If you want to propose another way forward, great, but be aware that ways forward that involve getting the developers to change the software have not proven successful.--Srleffler (talk) 16:40, 27 October 2008 (UTC)

Wouldn't a better fix be to make date formatting work for non-logged in users (IP editors)? I'd agree that the links themselves aren't quite so important (though I prefer to have SOME dates linked, just not ALL dates linked), but the formatting is something we should work on getting fixed and kept. —Locke Cole • t • c 02:26, 30 October 2008 (UTC)

Proposal to strengthen the watery proscription of year–month–day numerical dates

A user has brought to my attention this little mess, in which the numerical dates (1) jostle with the numerical publication numbers, (2) are probably not comprehensible by most of our readers (is "1943-09-08" September 8 or August 9? It's made harder by the non-linear US medium–small–large date format (month–day–year) as a reference point for many readers.), and (3) are inconsistent with the US format used elsewhere in the main text.

Currently, MOSNUM says this:

YYYY-MM-DD style dates (1976-05-31) are uncommon in English prose, and are generally not used in Misplaced Pages. However, they may be useful in long lists and tables for conciseness and ease of comparison. Because some perceive dates in that style to be in conformance with the current ISO 8601 standard, that format should never be used for a date that is not in the (proleptic) Gregorian calendar, nor for any year outside the range 1583 through 9999.

What the hell is "ease of comparison"? No one has ever explained that bit.

I propose that this be strengthened thus:

YYYY-MM-DD style dates Wholly numerical dates such as "1976-05-31" are uncommon in English prose, and are generally not used in Misplaced Pages the main text of Misplaced Pages articles. However, they may be useful in long lists and tables for conciseness and ease of comparison. Because some perceive dates in that style to be in conformance with the current ISO 8601 standard, that format should never be used for a date that is not in the (proleptic) Gregorian calendar, nor for any year outside the range 1583 through 9999. In tables in which there is a shortage of space, three-character abbreviations of the months may be used in the order that is consistent with the prevailing date format in the main text (e.g., "Aug. 5, 1961" or "5 Aug. 1961").

I'd rather ban them in all tables, but that might be going too far. Tony (talk) 02:17, 27 October 2008 (UTC)

I see no reason to use them in tables when it seems to be ok to use "12 Nov 2008" or "Nov. 12, 2008" which is exactly one or two characters longer than "2008-11-12" but reads much easier. ISO should only be used in references and that until we work out how to replace DA. --MASEM 02:24, 27 October 2008 (UTC)

• Agree with Tony. Greg L (talk) 02:29, 27 October 2008 (UTC)

Your rewrite of the ISO 8601 part, in my view, changes the meaning. Before, the date 1582-10-14 was totally unacceptable for any purpose and could be changed on sight, because it is in the YYYY-MM-DD format and the year was less than 1583. The new statement says the ISO 8601 standard should never be used for years outside the 1583-9999, so it might be ok to use the format so long as it was not accompanied by a statement that it conforms to the ISO 8601 standard. --Gerry Ashton (talk) 02:29, 27 October 2008 (UTC)

The "ease of comparison" bit is straightforward and should be retained. Basically, the yyyy-mm-dd sequence is hierarchical, inspectable and sortable. It's more easy to see that 04 comes before 05 than to see that Apr comes before May. Franamax (talk) 03:53, 27 October 2008 (UTC)

The "sort" template deals with this. eg sort|2008-11-12|Nov. 12, 2008. --MASEM 04:07, 27 October 2008 (UTC)

The documentation for the sort template is horrible, and in my book, a template is no better than its documentation. --Gerry Ashton (talk) 04:14, 27 October 2008 (UTC)

It's pretty straight forward: two parameters, first is the sort key, second is the displayed text. It basically expands to a bit of display:none CSS for the sort key. The template doc can be fixed but I've used the sort template with iso dates as the sortkey but normal dates for songs (see, for example List of songs in Guitar Hero III: Legends of Rock) --MASEM 04:18, 27 October 2008 (UTC)

Masem, I've checked that article and I find no instances of {{sort}} - perhaps you subst'd the instances? I do see the sort technique though (rather clunky) and I see another template now {{dts}}, so that's good to know.

More worryingly, even though I figured there was a way and I'm somewhat used to the wiki-hunt, how was I supposed to know about those templates? I naturally checked at Help:Table, which leads to Help:Sorting and both are Meta pages and seem to contain no links to the en:wiki sorting templates. So not only is the template docs not so good, I can't even find a way to get where I'm supposed to be confused! Asking the average editor to figure all this stuff out whilst forbidding the simple yyyy-mm-dd format so they can get their table to work at all just seems a bit much. I dunno, was there some obvious clue I should have picked up on? I'm not that swift sometimes... Franamax (talk) 05:30, 27 October 2008 (UTC)

That's a notable, but fixable problem, if we decide to completely rid ourselves of ISO-type dates in the article body; both sort (which is general purpose for any sortable key, so those lists use for "the"-less sorting, the dates as given are effective what it would expand out to) and dts help achieve this for the only place where ISO dates (outside of the ISO date article itself) are left being used. These do need to be discused in at least the help for sortable tables, but that's easily fixed. --MASEM 05:43, 27 October 2008 (UTC)

It's time to call this hyperbole that's been repeated here way too many times. The assertion that YYYY-MM-DD formated dates such as 1943-09-08 "are probably not comprehensible by most of our readers" is unreferenced, weasel worded, and just plain insulting to the reader's intelligence. Can anyone furnish reliable sources that show that "most" of any group similar to the users of the English WP: (Internet-using, English-literate, and globally distributed) have problems understanding that numbers are normally written most significant digit first? LeadSongDog (talk) 04:43, 27 October 2008 (UTC)

I personally have no problem at all, when I see a leading year, I switch over to hierarchical mode, so I would read the above as Sep. 8, 1943. However, I'm Canadian, so I'm used to seeing multiple date formats, maybe that's not a universal experience? Certainly 02/03/04 can be ambiguous, but to me the leading four-digit year in dashed YMD notation is not confusing. (And *ahem*, that's where linked day articles can make things very clear - but I digress :) Franamax (talk) 05:02, 27 October 2008 (UTC)

• It’s not that they can’t be understood. They’re just not normally written that way so they are less intuitive and takes more effort to read. People read by looking at words. That’s why “July 4, 2001” is much easier for Americans than parsing out 2001-07-04. The latter is ugly to boot. What Tony proposed is simple common sense. Greg L (talk) 04:54, 27 October 2008 (UTC)

• I think we need to sort out date autoformatting before we start banning certain date formats entirely as seems to be the proposal here. Strongest possible oppose to this proposal. —Locke Cole • t • c 04:59, 27 October 2008 (UTC)

• Now that we're down the road of more than one acceptable date format (Int and Amer), I'm going to start pushing for yyyy-mm-dd format for all China-related articles. No more Mr Nice Guy ;-) Ohconfucius (talk) 07:46, 27 October 2008 (UTC)

I don't know how you could get anything but 8 June 2004 out of 2004-06-08: for those used to month-day-year, the year is just moved in front and the rest is as usual; for those used to the logic of little to big, this is just as logical only in reverse. But I can't speak for everyone on this. The point is though that this format is uncommon in English ... how they do it in Chinese doesn't matter since this is the English WP. ISO dates serve no purpose here on WP. There are templates to deal with sorting (if they are unfindable, make them findable) and there are three-letter abbreviations to save space. I strongly support Tony's proposal—in fact I'd be happy to make it an all-out ban. JIMp _talk·cont 09:08, 27 October 2008 (UTC)

Rejoinders from Tony:

• Gerry, I was going to ask you to suggest a rewording of the last sentence in the proposal, given your note above; but other suggestions here may have superseded that. Are you in agreement with the reworded proposal?
• I'm an experienced reader/editor, but I still have to dwell on these numerical dates to mouth out in words the month and the day. I guess it's confusing that throughout the English-speaking world we use both month–day and day–month order, which is just fine when the month is spelled out. But as Americans who travel to the UK or Australia soon discover, 2/5/08 means May 2, 2008, not February 5, 2008. It's just asking for trouble on an international site not to spell out the month. We need dates to be simple and explicit on WP. This numerical gobbledygook should be flushed down the pan for the sake of our readers.
• I see much to commend Masem's point that a three-character abbreviation of month is just fine for saving space in tables. I'm going to be bold and change the proposal to this effect.
• Franamax, I don't see why "04" and "05" has any advantage over "Apr" and "May", since every English-speaker is adept with the sequence of months (as with the days of the week). The problem is readily identifying which of the digits is the month and not the day (emphasis on readily—we can all stare at it like a chess puzzle and work it out, but I suspect many of our readers simply glaze over when they see dates that look like telephone numbers.
• Locke Cole, this is quite a separate issue from DA, of course. Tony (talk) 10:24, 27 October 2008 (UTC).

Please compare these three examples of lists starting with dates, from the "mess" I linked to at the top of this section. Feel free to comment below. Tony (talk) 10:47, 27 October 2008 (UTC)

Example A: numerical dates:

• 1943-12-17 — Magnuson Act (Chinese Exclusion Repeal Act of 1943), Sess. 1, ch. 344, 57 Stat. 600
• 1944-02-03 — Mustering-out Payment Act, Sess. 2, Pub. L. 78–225, 58 Stat. 8
• 1944-06-22 — Servicemen's Readjustment Act of 1944 (G.I. Bill), Sess. 2, ch. 268, Pub. L. 78–345, 58 Stat. 284
• 1944-06-27 — Veterans' Preference Act, Sess. 2, ch. 287, Pub. L. 78–359, 58 Stat. 387
• 1944-07-01 — Public Health Service Act, Sess. 2, ch. 373, 58 Stat. 682
• 1944-12-22 — Pick-Sloan Flood Control Act, Sess. 2, ch. 665, Pub. L. 78–534, 58 Stat. 887

Example B: dates with abbreviated months:

• Dec. 17, 1943 — Magnuson Act (Chinese Exclusion Repeal Act of 1943), Sess. 1, ch. 344, 57 Stat. 600
• Feb. 3, 1944 — Mustering-out Payment Act, Sess. 2, Pub. L. 78–225, 58 Stat. 8
• Jun. 22, 1944 — Servicemen's Readjustment Act of 1944 (G.I. Bill), Sess. 2, ch. 268, Pub. L. 78–345, 58 Stat. 284
• Jun. 27, 1944 — Veterans' Preference Act, Sess. 2, ch. 287, Pub. L. 78–359, 58 Stat. 387
• Jul. 1, 1944 — Public Health Service Act, Sess. 2, ch. 373, 58 Stat. 682
• Dec. 22, 1944 — Pick-Sloan Flood Control Act, Sess. 2, ch. 665, Pub. L. 78–534, 58 Stat. 887

Example C: full dates in a standard WP format:

• December 17, 1943 — Magnuson Act (Chinese Exclusion Repeal Act of 1943), Sess. 1, ch. 344, 57 Stat. 600
• February 3, 1944 — Mustering-out Payment Act, Sess. 2, Pub. L. 78–225, 58 Stat. 8
• June 22, 1944 — Servicemen's Readjustment Act of 1944 (G.I. Bill), Sess. 2, ch. 268, Pub. L. 78–345, 58 Stat. 284
• June 27, 1944 — Veterans' Preference Act, Sess. 2, ch. 287, Pub. L. 78–359, 58 Stat. 387
• July 1, 1944 — Public Health Service Act, Sess. 2, ch. 373, 58 Stat. 682
• December 22, 1944 — Pick-Sloan Flood Control Act, Sess. 2, ch. 665, Pub. L. 78–534, 58 Stat. 887

Comment—Yes, the numericals in A are the neatest visually and are compact, but the abbreviated version is almost as neat and compact. These examples do show the full dates in C in the poorest possible light, and I think they work best in most contexts. But a bit of deft formatting can improve their look—I'd be inclined to put the full dates last, in parentheses, like this:

• Magnuson Act (Chinese Exclusion Repeal Act of 1943), Sess. 1, ch. 344, 57 Stat. 600 (December 17, 1943)
• Mustering-out Payment Act, Sess. 2, Pub. L. 78–225, 58 Stat. 8 (February 3, 1944)
• Servicemen's Readjustment Act of 1944 (G.I. Bill), Sess. 2, ch. 268, Pub. L. 78–345, 58 Stat. 284 (June 22, 1944)
• Veterans' Preference Act, Sess. 2, ch. 287, Pub. L. 78–359, 58 Stat. 387 (June 27, 1944)
• Public Health Service Act, Sess. 2, ch. 373, 58 Stat. 682 (July 1, 1944)
• Pick-Sloan Flood Control Act, Sess. 2, ch. 665, Pub. L. 78–534, 58 Stat. 887 (December 22, 1944)

After all, the names of the Acts are what should come first—they're a much better point of departure than the exact date of passage. Tony (talk) 10:47, 27 October 2008 (UTC)

Tony, I don't necessarily disagree with any of your above, except the bit about the chess puzzle. YYYY-MM-DD is really not all that puzzling. In fact, here in Canada at least, where we're headed to one-day financial clearing, cheques now have that same format (sans dashes) Image:CanadianChequeSample.png. Now it's true that Americans still measure in cubits, but I don't see a huge stretch to accomodate three rather than two date formats, namely "dd Mon yyyy", "Mon dd, yyyy" and "yyyy-mm-dd" (and expansion of "Mon" to "Month", since someone will always want September 11, 1918 rather than 11 Sep 1918). The YMD usage should be discouraged in articles, but not forbidden in tables where it can be used to good effect and by it's own context in the column solve the chess puzzle anyway. Franamax (talk) 10:58, 27 October 2008 (UTC)

And my pick for conciseness and readability, unfortunately, is your first set of examples. :( Franamax (talk) 10:58, 27 October 2008 (UTC)

I'm not sure how many e/c's I've fallen behind now - anyway, look at your Example A: there's a good temporal layout which is intuitive and obvious for any reader, by the very nature of the progression of dates. Of course, this would only be applicable when the editor wished to demonstrate a temporal progression of events, but I think example A makes that progression most easily clear. Franamax (talk) 11:06, 27 October 2008 (UTC)

Not intuitive for every reader, experto crede; and B and C are less clear only for those readers who have trouble with the months of the year. This is not the Simple English Misplaced Pages. Septentrionalis PMAnderson 16:11, 27 October 2008 (UTC)

Of course it's not a separate issue from date auto formatting, because if DA worked as expected for all users (logged in or not) then the matter of what date format to allow/disallow within articles would be a moot point. So please get back to helping resolve DA rather than plowing ahead with other changes. —Locke Cole • t • c 13:25, 27 October 2008 (UTC)

I oppose a ban on ISO compliance within Wikipdedia. It seems to me that Misplaced Pages and international standards should be friends by default, not enemies. I will support a ban on using ISO format for non-ISO dates (the Gregorian problem referred to by Gerry Ashton). As much as I respect Tony on many issues, I think he is wrong to ban them. I wonder what the citation people think. I think MOS guidance should have a 'clear and present problem' test and this would fail such a test, I think. Lightmouse (talk) 13:19, 27 October 2008 (UTC)

• I like option B. It’s very human-friendly, and it’s compact enough for this backwater use. To me, it is a superior trade off. I haven’t thoroughly looked over the ISO standard, but I’m quite sure they weren’t suggesting that all-numeric dates should be the preferred way for dates to be written out for humans to read in body text. I’m confident the point of the standard was to bring order (standardization) to those instances where all-numeric dates are normally used. You see all-numeric dates in computer databases, on blue prints (technical drawings), tables of data, etc. I’ve seen some editors get awfully enamored with ISO dates and defend them for use in body-text like “after the 1941-12-07 attack on Pearl Harbor”. That’s not what they’re for. Greg L (talk) 16:01, 27 October 2008 (UTC)

• I'm indifferent whether all-numeric dates are declared contrary to this guideline, and our "blessing" is transferred to three-letter abbreviations. If three-letter abbreviations are adopted, we should decide whether or not a period is to be used after the abbreviation. As an aside, I was watching the World Series, and I noticed an advertisement displayed on the wall behind the batter (it was actually added in the television production process; the real wall is a green screen. The ad was for some product or movie due out in December, and the date was formatted something like "12.15.2008". (I'm certain periods were used as separators, I'm not sure which date in December the thing is becomming available). So apparently even mass-marketers are willing to believe Americans have some flexibility reading date formats. --Gerry Ashton (talk) 18:16, 27 October 2008 (UTC)

Interesting, Gerry. I think it's foolish of the advertising company: I'd ditch the year (it's obvious) and say "out December 15", or if desperate for space, "out Dec. 15". From a communications standpoint, the format they used was a big mistake. Tony (talk) 01:16, 28 October 2008 (UTC)

• OK, I have a suggestion on whether the abbreviated months should have periods or not. I note that my local paper, which probably follows one of the major U.S. newspaper style guides (NY or Chicago), handles it this way: When someone writes a letter to the editor, they will insert a parenthetical pointing to which issue the reader is referring to. So you see text like this:

I see from your Sunday article (“Americans shit their pants when they see metric stuff,” Sept. 28), that you left no stone unturned…

Again, note that this is being used in a parenthetical, which simulates the back-water nature of publication citations. Note that the paper also uses flex, three-to-four-letter abbreviations and periods. This allows months like June to be written out. Thus, I believe the progression goes as follows: Jan. Feb. Mar. April May June Jul. Aug. Sept. Oct. Nov. Dec. It abandons the rigid, three-letter-only notion in order to make it more human-friendly. Greg L (talk) 19:53, 27 October 2008 (UTC)

If we write April in full, would we not also write March in full? –xeno (talk) 20:16, 27 October 2008 (UTC)

• I’m not sure about the exact practices of the newspaper. I’ve written a number of letters to the editor (go figure) and have them archived. So I know for a fact about some of them. They might indeed spell out March. I am assuming that they follow an industry-standard practice and I would recommend that we simply adopt that practice since I can see what the goal is and it makes sense. I happen to have the phone number of the gal at the paper who calls to confirm whether it was really me who sent in the letters. I think I’ll call her and ask her how the progression goes. Greg L (talk) 21:31, 27 October 2008 (UTC)

If the only concern is reducing the amount of space taken up by months, we could achieve the goal nearly as well by allowing short months (e.g. June) to be written in full. However, if all abbreviations are exactly three letters, tables can be formatted with fixed-width fonts to force all the elements of the date to align vertically. I suppose exactly how the editor envisions the reader will use the table will determine whether the improved alignment is worth tolerating the ugly fixed-width fonts. --Gerry Ashton (talk) 20:38, 27 October 2008 (UTC)

• Addressing Xeno and Gerry: For tables, where exact fit and alignment are important, I don’t see any reason at all why we can’t have three-letter abbreviations. For sidebars and citations, and similar non-body-text uses where tight fit isn’t a consideration, I would suggest flexible abbreviations like newspapers use for parenthetical use when citing a specific edition. I just got off the phone with the editor of our local paper. They use the Associated Press manual of style. He said all U.S. newspapers follow the AP guideline so this abbreviation method is consistently used across the country. It is as follows:

• Jan.
• Feb.
• March
• April
• May
• June
• July
• Aug.
• Sept.
• Oct.
• Nov.
• Dec.

Greg L (talk) 21:46, 27 October 2008 (UTC)

Thus, we have…

Example D: Associated Press-style abbreviated months:

• Sept. 17, 1943 — Magnuson Act (Chinese Exclusion Repeal Act of 1943), Sess. 1, ch. 344, 57 Stat. 600
• Feb. 3, 1944 — Mustering-out Payment Act, Sess. 2, Pub. L. 78–225, 58 Stat. 8
• June 22, 1944 — Servicemen's Readjustment Act of 1944 (G.I. Bill), Sess. 2, ch. 268, Pub. L. 78–345, 58 Stat. 284
• July 27, 1944 — Veterans' Preference Act, Sess. 2, ch. 287, Pub. L. 78–359, 58 Stat. 387
• April 1, 1944 — Public Health Service Act, Sess. 2, ch. 373, 58 Stat. 682
• Dec. 22, 1944 — Pick-Sloan Flood Control Act, Sess. 2, ch. 665, Pub. L. 78–534, 58 Stat. 887

Greg L (talk) 21:52, 27 October 2008 (UTC)

• Now that we have the above AP-style abbreviated month progression captured above, wouldn’t it be good to capture this information in MOSNUM? I’d offer to do it, but I’m a hundred times more familiar with Talk:MOSNUM than MOSNUM itself; I’d probably put it in the wrong spot. Greg L (talk) 00:11, 28 October 2008 (UTC)
  • Does it serve any purpose? Month abbreviations are well known, and anybody who wants can reconstruct them as easily as we constructed them. Septentrionalis PMAnderson 00:56, 28 October 2008 (UTC)

• Maybe they are well known to you, but I doubt that one editor in ten here can properly write down the AP version of the month abbreviations without looking. Greg L (talk) 01:06, 28 October 2008 (UTC)
  • Is there any profit to the encyclopedia to insisting on March and April over Mar. and Apr.? Would there be any to insisting on the other way around? If not, let us abstain. Septentrionalis PMAnderson 01:09, 28 October 2008 (UTC)

• It’s so abbreviations here don’t somehow look awkward. I’ve always wondered why my casual attempts at monthly abbreviations looked odd. I’d try Jan. Feb. Mar. Apr. Jul. Sep. and they’d look *funny*. Then I’d try four-character abbreviations and they’d look odd. It’s because newspapers (and probably many magazines too) observe a specific practice where they have variable number of characters. The benefit to Misplaced Pages is a style that looks natural and reads fluidly. I would simply propose that when there are tight space considerations and a particular need for columns to line up, three-letter abbreviations are fine. If it is in main body text, spell out the name of the month. If it is side-bars, citations, and other such uses, follow the AP’s flexible-character abbreviations observed by every subscribing English-language newspaper in the world. Why is there any controversy to this?

  Maybe the AP-style monthly abbreviations are well known to you, but I doubt that one editor in ten here can properly write down all twelve months without looking at the above list and without error. After all, the editor of a major city newspaper had to go to his AP handbook to provide them to me without error. Besides, MOSNUM isn’t only for über-experienced editors like you; it is here to help editors with minimal editing experience (which is why many would come here in the first place). I’m curious why you would even question the value of memorializing the abbreviations here. Misplaced Pages is authored largely by novices. The Associated Press is populated by writers, all of whom have journalism degrees and they saw fit to put it into their manual of style. Greg L (talk) 01:24, 28 October 2008 (UTC)

  • The AP stylebook is intended for reporters: writers who are in a hurry, and whose stories will not be revised after publication, usually in a matter of hours. We have more lesiure.
  • We also have a Manual of Style which is now infinitely less comprehensible than the AP 's, because we have crammed more into it than our organization can make clear. The solution to that is to be more hesitant about cramming even more into it.
  • Our MoS was never intended to cover all the topics of any stylebook; it was intended to say what was necessary for our purposes: stuff they omitted, matters related to wiki format, content rules needed because anybody can edit us (like ENGVAR), decisions where stylebooks disagree (including agreements that we may differ from article to article), and those purely stylistic points which are most important to us and so worth spending space and complexity on. . Septentrionalis PMAnderson 03:05, 28 October 2008 (UTC)

Common Use doesn't make it right It's common-use in US English prose to call Eastern Time: EST even though that's not correct (in other words NY time gets called EST even in the summer). Therefore the argument that common-use in US English prose should determine Misplaced Pages convention is a weak argument in comparison to the unamibugauce ISO standard (even for non-traveled US citizens). The idea that anyone could get Aug 9th out of 1943-09-08 is unlikelyhood piled on top of unlikelyhood. I agree with that statement that it's "insulting to the reader's intelligence" to think they can't understand 1943-09-08. —Preceding unsigned comment added by Zaurus (talk • contribs) 02:20, 28 October 2008 (UTC)

People who proclaim that "no-one could possibly misunderstand what I meant" should at least learn how to spell first. Septentrionalis PMAnderson 03:05, 28 October 2008 (UTC)

Agreed, as far as I know, no dating system in the world has ever used YYYY-DD-MM. Short of spelling out the month, it's the most disambiguous way to give a date. –xeno (talk) 14:18, 29 October 2008 (UTC)

• The harm, in posting a list of AP-style month abbreviations here is… what? Greg L (talk) 03:25, 28 October 2008 (UTC)
  • Every additional sentence makes it that much more difficult to find stuff, and that much easier for eyes to glaze over. Small harm, but so is the benefit. It's laudable of you to dig this up; but think of it from the next reader's point of view, when this is just another bit of clutter. Septentrionalis PMAnderson 03:34, 28 October 2008 (UTC)

• Maybe I'm wrong, but I think newspapers hardly ever publish tables containing month names (or numbers) where it would be nice to have them align vertically. Misplaced Pages is more likely to do so. So while the newspaper could confine itself to two sets of month names (full and abbreviated), we would need three (full, AP abbreviations, and three-letter abbreviations). I would prefer to stay with just two sets, full and three-letter abbreviations, even if the latter don't read quite as smoothly. --Gerry Ashton (talk) 03:34, 28 October 2008 (UTC)
  • Not that much more smoothly; most users will be seeing proportional font. Septentrionalis PMAnderson 03:36, 28 October 2008 (UTC)

• Another point against ISO-like formats is : Won't someone thing of the children?? Seriously, if we're trying to write a work that is accessable across as much regional variances and ages as possible, I will tell you that the ISO-like format is not something that a grade schooler will be able to recognize, but I know in the US that the months of the year are pre-school if not earlier (and I presume in the EU and other parts of the world, taught in DD MMM YYYY format). I have no problem if ISO is used in wikicode, but presented work should avoid this format save for discussions of what the ISO format is. --MASEM 14:30, 29 October 2008 (UTC)
  • Masem, I think your heart's in the right place thinking about the kids. But think of the favor we'd be doing them to teach them the more efficient standard, so they wouldn't have to guess the meaning of 11-9-2001. I would add that my two kids don't have a problem with yyyy-mm-dd. And I think if Misplaced Pages were to adopt this formatting, it would expose it more to kids, giving them a head-start with the more efficient standard. It should especially be easy for US kids, cause they already use mm/dd, so it's just switching the year to the front. As for the EU, I think they often use dd-mm-yyyy, as mmm is language specific.--Zaurus (talk) 04:59, 30 October 2008 (UTC)

Exception

Any such re-writing needs to allow for the fact that templates like {{Start date}} and {{Birth date}} have parenthetical YYYY-MM-DD format dates, hidden by CSS (and thus visible to people with CSS unavailable or disabled as, say, 31 May 1975 (1975-05-31)) in order to output the ISO-format dates required in microformats. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 11:27, 27 October 2008 (UTC)

Note that the {{TL:Birth date}} provides a date in the format required by HCard. That relies on vCard, RFC 2426, which is intended for directory, "white-pages" information, as would be found on a business card. Thus it is only intended for living people, or people who have died recently. Using it to describe people who have been dead for millenia can be expected to cause trouble, and sure enough, it does. The vCard spec relies on RFC 2426 to define some of the terms it uses, and in RFC 2426 we find " 'date', 'time', and 'date-time': Each of these value types is based on a subset of the definitions in ISO 8601 standard." This means that it is wrong to use {{Birth date}} (and most likely, any of its cousins} for any person who's birth date is written in any calendar other than the Gregorian calendar. --Gerry Ashton (talk) 23:17, 27 October 2008 (UTC)

hCard is based on vCard, but is used for more than just "white pages" info, and hCards are legitimately used for more than living (or still-warm) people. The issue of pre-Gregorian hCard dates (which I raised on 19 April 2007) remains to be resolved (the microformat community is notoriously and lamentably slow in such matters). While your latter point may arguably be valid, it has no bearing on the point I made. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:29, 27 October 2008 (UTC)

If there is to be an exception, my post has bearing on how the exception is worded. The exception might or might not use the phrase "ISO 8601". The exception might recklessly claim that the format YYYY-MM-DD is permitted for invisible machine-readable purposes, but Misplaced Pages specifically rejects any implication that ISO 8601 applies, and makes no statement about what calendar the date is written in. Or the exception might say these microformats are OK so long as the year is greater than or equal to 1583, or 1753, but is forbidden for earlier years. --Gerry Ashton (talk) 23:42, 27 October 2008 (UTC)

The use of abbreviated month names in tables has to my mind some limited merit, but considerable associated complication. Because most editors will not take the trouble to memorize the list of approved abbreviations and their punctuation a (semi-)automated spelling repair would have to be implemented to catch all the variations and bring them into conformance. Should US/international variants also apply, then that automation will need to be sensitive to the choice of variant that is adopted for each article in question. The usual argument for the US style "Jan 1, 2000" (that it avoids starting sentences with a numeral) does not apply in tables. I see no merit in using the abbreviations in citation/cite template inputs or outputs. These should be done with outputs fully spelled out (should that be the ongoing consensus) and with inputs in any unambiguous form that parses correctly. I agree with Gerry that the misapplication of the term ISO 8601 to describe simple YYYY-MM-DD form is problematic, particularly for historic dates and for timezone sensitive articles, such as those dealing with aviation events. We should explicitly distinguish the form YYYY-MM-DD from the ISO 8601 standard.LeadSongDog (talk) 17:54, 29 October 2008 (UTC)

• Rant time: I’m not saying that only abbreviated months should be used and that all-numeric dates should not be used; each has its place. I’m just saying that abbreviated months have their uses, and when they are used, they should be properly written. There is far too much tendency from certain elements here to semi-invent special formats for use here on Misplaced Pages, or to adopt some cockamamie standard (like all-numeric dates) just because it has been anointed by ISO, BIPM, or whomever. The ISO only standardized how numeric dates should be formatted; it didn’t suggest they should be used in normal prose. Newspapers do too use abbreviated months. And that’s why the Associated Press, whose practices are observed by all newspapers throughout the English-speaking world, has a guideline governing how to write them. It’s a logical system, is what we are used to reading, and is what we subliminally recognize as looking right. Short of some use like cramming abbreviations into a super-tight chart or some other unusual situation like that, you never abbreviate all the months names into either fixed three or four-letter groups. The proper abbreviations are as follows:

• Jan.
• Feb.
• March
• April
• May
• June
• July
• Aug.
• Sept.
• Oct.
• Nov.
• Dec.

It is not Apr., Jun., and Sep. Nor is it “Sept. Octo. and Nove” (only “Sept.”). Proper practice is to use variable-length abbreviations. Further, I utterly reject any argument that there is “too little room on MOSNUM for this list,” or that it “lengthens MOSNUM to the point where readers can no longer find what they need,” or that the “AP has professional writers with journalism degrees who are so damn busy that they need such a guideline but Wikipedians have more time and they don’t need such a guideline”; those arguments are totally absurd. Beyond absurd.

What we’ve got here are some editors who are far too damn pushy and are blocking common-sense adoptions of simple guidelines that enjoy world-wide recognition. Some editors here behave as if MOSNUM and WT:MOSNUM are some sort of private reserve for them to make their personal and indelible imprint upon. What I really think is that some editors here have been doing things a certain (wrong) way, that they A) are completely convinced their way is logical and best, and B) don’t want to see their work eventually changed in any way. It’s not that I’m suggesting some editors here lack bad faith; I just think they are so biased on this issue, that they are desperately resorting to arguments that are utterly fallacious. That’s my opinion… so shoot me. We are not going to be inventing new home-grown conventions for abbreviating months in citations, sidebars, and parentheticals. What’s with this tendency on MOSNUM that it enables editors to come here and push through practices (like “256 kibibytes”) that no one else in the real world observes on this pale blue dot? It’s time to stop pushing through guidelines—home-grown in some cases—just because someone thinks they’ve figured out a *better way to a brighter and more logical future* or are misinterpreting what some standards organization is really trying to accomplish.

After I get this simple damned list of how to abbreviate the months on MOSNUM, I’m going to stop wasting oodles of time fighting every common-sense minor point. Instead, I intend to focus on fixing the very underpinnings of how MOSNUM functions, arrives at a consensus (or not), identifies when a consensus has truly been reached, resolves conflict (which is pretty much all that occurs here), and avoids making half-baked decisions. Right now, this place is broken beyond all recognition, requires that editors invest far too much time arguing, and it gives far too much weight to the editor who shouts loudest, longest, and is most willing to push the boundaries of edit warring. Greg L (talk) 18:20, 29 October 2008 (UTC)

Time for decaf;/? I've said before, this is the wrong venue for discussion of what constitutes consensus. That's Misplaced Pages talk:Consensus (which desperately needs fixing). If fixed there, it'll spill over here. Some samples from todays news here, here, here, here and here make it pretty clear there is zero consistency of date style in the real world press, whatever the AP stylebook says. Articles even use different styles in their RSS dateline from the displayed dateline ensuring that aggregators like google news get something different. 18:48, 29 October 2008 (UTC)LeadSongDog (talk)

• Disagree: A prohibition against YYYY-MM-DD is ridiculous considering how fractured actual real-world usage is. Trust me, I check expiry dates on Canadian pharmaceuticals, and on a regular basis manufacturers don't even use the same format from one lot to another! Also, if people stop fighting to delink dates (why delink first use of a date??), there won't be a problem with ambiguity (even ambiguity with a sensible format to it), because explanation will be a click away! Why do we have pages for those dates anyway if we aren't going to link to them? ps: There seriously needs to be less bitterness on this page too. It was a terrible read. ;_; - BalthCat (talk) 22:58, 29 October 2008 (UTC)

• I don't follow your reasoning. How can the pharmaceutical situation be an argument for allowing numerical date formats? It just goes to show that using month names is much better. As for using explanatory links, that poses major WP:1.0 problems. -- Jao (talk) 23:16, 29 October 2008 (UTC)

• Who the hell is proposing to completely abolish YYYY-MM-DD? It has its place. All I had proposed is to adopt the nice, smooth AP method of abbreviating months when one needs abbreviated months. This isn’t rocket science here. Greg L (talk) 01:40, 30 October 2008 (UTC)

• This section, "Exception", is a subheading of Proposal to strengthen the watery proscription of year–month–day numerical dates. Although the proposal has been edited, it says that all-numeric dates are not used "not used in Misplaced Pages the main text of Misplaced Pages articles." Tony proposed that. And you agreed with him (or so it seems; with the edits, maybe not is all as it appears). --Gerry Ashton (talk) 02:33, 30 October 2008 (UTC)

• LeadSongDog. I looked at your list of examples in your 18:48, 29 October 2008 post, above. They didn’t demonstrate that there are different ways the world press abbreviates the names of months. One spelled it out in full. One used all-numeric, and three spelled it “Oct.” (although Reuters left off the period). I am not taking a position on whether dates should be spelled out or in all-numeric or abbreviated. I have a very narrow, focused objective: for those uses where dates should be abbreviated, and where they don’t have to fit into ridiculously tight spaces like in tabular charts, then follow the A.P. method. That’s all. I’m waiting for a better argument than “a list of twelve months takes up too much room on MOSNUM.” Cow pancakes. And, your comment about decaf was amazing prescient; I happen to have accidentally made a double-strong cup of coffee only a half hour before that post. Greg L (talk) 03:11, 30 October 2008 (UTC)

Guess I was too vague. this AP story uses "Aug. 28" in the text and "Tuesday, Oct. 28, 2008" in a photo caption. this Bloomberg story uses "Oct. 29" in the dateline and "October 29, 2008 13:29 EDT " in the article's Last updated, while showing "Updated: New York, Oct 30 18:26" on the page banner. this Guardian story shows "Monday October 27 2008 " on the dateline while the page has "Oct 28 2008" in the Related information list. this Radio Netherlands story uses "29-10-2008" in the dateline and "28 August" in the Imported violence box. This Reuters Africa story uses "Wed 29 Oct 2008, 21:02 GMT" on the upper dateline, "Oct 29" on the lower dateline, and "Thu 30 Oct 08 | 22:24 GMT" at the masthead. Among them we have inversions of sequence, variant punctuation, variation of whether to state the time zone,, and variation of whether to state the day. Clearly they are not all following one style. Hope you enjoyed the buzz;/) LeadSongDog (talk) 22:44, 30 October 2008 (UTC)

Help with terminology

Can anyone explain to me whether something can be simultaneously:

• minor and inconsequential
• controversial
• too important to build into 'general fixes' of AWB

It seems to me that people are mangling the english language by saying that date delinking fulfils all three bullets. Lightmouse (talk) 12:46, 27 October 2008 (UTC)

As I've said to you above, you need to stop conflating an MOS debate with an AWB misuse issue. Suggestions on how to properly work in these minor/inconsequential delinking edits have been given on your talk page - simply ensure that you enable RegEx typo fixing and skip when no typo is fixed. This also means you actually have to check every edit, which provides the added benefit of lowering your false positive rate. As to why delinking can't be built into general fixes: because there is no way for AWB to tell when a date link is provided for context vs. when it was provided for autoformatting. –xeno (talk) 12:51, 27 October 2008 (UTC)

If I understand much of the previous commentary, the explanation you seek lies within context and scope. Few would argue that the removal of date links in any given article is not "minor and inconsequential", while the act of doing so on a broader scope can very well be "important and controversial". By way of an illustrative example : cutting down any given tree in a forest is minor and inconsequential, but cutting down every tree in the forest is important and controversial. I cannot speak to the AWB issue directly (I have never used AWB and have no experience with it) but it is fairly clear that even the most minor sort of edit becomes a big deal when it is performed many thousands of times. Shereth 13:56, 27 October 2008 (UTC)

• Are you implying that DA and date-linking is like the trees in a forest? If so, I believe you and I are on completely different planets, so th speak. For me, one date-linked page is like one plastic bottle left to litter a beautiful beach. We should pick up and dispose of all the plastic bottles, and restore the beauty of our beach. Cleaning up is important and should not be controversial at all. Ohconfucius (talk) 15:21, 27 October 2008 (UTC)

• LM asked for an example where something could be considered "inconsequential" yet "important" at the same time; I provided one. Your example serves equally well. I am neither advocating nor opposing the delinking of dates. I am merely interested in seeing a solution develop out of community consensus rather than a few interested editors, given the scope of the work being proposed. Cleaning up is important and should not be controversial, but it has generated controversy nonetheless and the question of whether or not this is beneficial cleanup of litter or harmful razing of trees is one the community should answer. Shereth 15:26, 27 October 2008 (UTC)

• Thanks for the clarification. Ohconfucius (talk) 16:08, 27 October 2008 (UTC)

No. The whole purpose of AWB is that it allows many small edits to be made. The 'minor and inconsequential' rule serves to avoid server burden for edits that are hardly noticeable, such as removing double spaces. It was not put in place to permit those with coercive power to censor pro-MOS edits that they don't want done. You suggest that autoformatting links are trees in a forest, well, the MOS says those trees need chopping down. You can't have an MOS that can't be implemented. Lightmouse (talk) 14:19, 27 October 2008 (UTC)

the MOS says those trees need chopping down. No, it doesn't; this is not The Most Important Thing. The marginal benefits involved in removing many (not all) date-links do not justify thoughtless and uncivil mass editing; the equivalent of a forest-fire. If someone is editing an article anyway, I would encourage xim to trim overlinking while xe are about it, and that is what is this guideline is supposed to produce. No more; that's why it's a guideline. Septentrionalis PMAnderson 14:47, 27 October 2008 (UTC)

• Then how about thoughtful and civil mass delinking? By my reckoning, lists are responsible for more than 50% of the date links within wikipedia. By selectively culling date links in articles like I have done in Category:Lists of television series episodes, I believe I am are carefully, selectively, thoughtfully clearing the dead wood so that trees may be seen and appreciated for their true beauty. BTW, these hundreds of edits I have now performed have caused no more than 2 editors to complain, and only about the wrong destination date format - pretty trivial, if you ask me. Ohconfucius (talk) 02:17, 29 October 2008 (UTC)

Date links themselves are of marginal and rare benefit. At least some classes of date links are suitable for mass-removal with little human oversight. Links such as March 20, and links to years since 1900 are almost entirely inappropriate. Cases where those links would be justified are so exceedingly rare that I doubt they justify the great increase in effort required to do manual link removal rather than automating it.--Srleffler (talk) 17:22, 28 October 2008 (UTC)

• Real guidelines express what any random dozen Wikipedians may be expected to say on a given issue. They exist because the majority of the dozen will need to explain themselves to the dissentient minority and the newbies, and they are tired of repeating themselves; so the majority bungs down what they agree on in Misplaced Pages space, and work on making it clearer over time. Therefore they don't need "enforcement", much less an enforcer; the real rule rests in the general consensus, and another random dozen Wikipedians will support it when necessary. The phrasing in Misplaced Pages space is an approximation of this, stuck together with Scotch tape and piano wire; just as WP:PRO says.
• What rule about date linking has that degree of support is unclear; the stream of complaints out of the wordwork casts doubt that any statement short of not too many links would. Septentrionalis PMAnderson 01:05, 28 October 2008 (UTC)

• And how long, exactly, is a piece of string? Your prescription is far too vague to be of use in a guideline. If the above approach is adopted, we enter into a very subjective realm with little or no clarity. I don't think anybody is suggesting banning all links to date/year articles any more. I believe most people seem to accept that some small number of date/year links may be valid in certain biographical and historical articles, and it has been actively proposed that these dates should be piped or displayed differently. We need not get into circuitous debate if we can just agree on that. The reason no-body has come up with a 'how', yet appears to be because most who favour retaining some sort of linking are resistant to anything but a direct wiki-link to the date and year article, without piping and without the use of templates. This is not about scotch tape any more... Ohconfucius (talk) 02:37, 28 October 2008 (UTC)
  • If there is no consensus on a clear line, we must be vague - or, what is often preferable, say nothing. Doing otherwise involves both bullying and scotch tape; some editors may prefer that, but I don't. Septentrionalis PMAnderson 02:52, 28 October 2008 (UTC)

• • Vague is as good as useless. You seem to be implying that the existence of rules equates to bullying. I see a consensus, but with isolated pockets of resistance. Having clear rules in such a context cannot be considered bullying, surely. Ohconfucius (talk) 02:56, 28 October 2008 (UTC)
    • Vague is as good as useless. Not among civilized editors, who can take a hint. But that is the case for silence. Septentrionalis PMAnderson 03:15, 28 October 2008 (UTC)

If there is no consensus - and in the sense I have outlined there is none here - then any definite rule will be opposed by many editors. A rule so opposed cannot be imposed without bullying. I have said no more. Septentrionalis PMAnderson 03:10, 28 October 2008 (UTC)

Val and delimitnum.

I've improved {{delimitnum}} a bit. It now supports the most numbers it can and a lot more leading 0s. See for details. It is still subject to the rounding problems and ending 0s however.Headbomb {ταλκ – WP Physics: PotW} 18:16, 29 October 2008 (UTC)

• Thank you very much Headbomb. I’ve got a stupefyingly huge proof page here at User:Greg L/Delimitnum sandbox. Due to the maddening math-based technique we have to resort to, this sandbox throws a variety of mathematical challenges at {{delimitnum}}.

  The long-term solution is to induce a charitable developer to give us a half-hour of his time and give us a character-counting parser function. Then people like you and SkyLined could go ape-shit and make these templates bug-free with a quarter of the effort. Do you know any such developer whom I can throw myself at his feet as he sits on his programmer’s throne in judgement of my non-programmer’s soul? Greg L (talk) 01:18, 30 October 2008 (UTC)

Greg, if I knew where to go, this thing would've been settled a long while ago. Well a long while ago being somewhere between now and March 2008 'cause I didn't know anything about template and template writing back in March. Well I still don't know that much about them, but you don't need to know a lot to write templates. Headbomb {_{κοντριβς} – WP Physics} 04:35, 30 October 2008 (UTC)

• All I have to do is write some Excel macros now and then to remind myself that my strong suite isn’t programming. Greg L (talk) 06:45, 30 October 2008 (UTC)

Date linking redux

After skimming the archives linked to from WP:MOSNUM and noting that the "overwhelming consensus" was achieved via a straw poll consisting of about a dozen unique editors, I really think we need to step back, stop unlinking dates en mass, and really work towards something that will be more agreeable to all involved.

I've asked a dev how long it might take to fix date autoformatting so it applies to anonymous/IP editors. I'm also prepared to work on this myself (I'm somewhat familiar with MediaWiki and have done PHP work before). But we need to agree on what exactly it is we want as an alternative to linked dates. Do we want a new syntax for dates? Do we want the dates to be formatted to one specific format for all IP visitors? Or maybe we should have a way to specify a format on each page (with a default being used on pages that don't explicitly state which format they want)? Are there other concerns we should consider?

This is a much better solution than simply unlinking dates and being generally disruptive. The only other alternative I think is to open a straw poll on whether or not to unlink dates and post a notice in MediaWiki:Watchlist-details. —Locke Cole • t • c 01:08, 30 October 2008 (UTC)

• (*Must be too much caffeine in me still.*)  What is clear here is that there can be no such thing as a consensus on Misplaced Pages if there is one single editor who disagrees with the decision. Decision making doesn’t work here on MOSNUM—and probably anywhere else on Misplaced Pages. I asked a developer to make a magic word to delimit long, scientific notation numbers six months ago. See the above thread. Has the magic word been made yet? Noooo. Has the simple parser function it relies on been made yet. Noooooo.  Has a developer bothered to write “Boo” in response to our “knock knocks”? Nooooo. Has a developer even acknowledged whether I exist and exhale carbon dioxide? Hell no. Don’t hold your breath for a developer to descend on feathered wings to the accompaniment of heavenly voices who is going to solve this for us anytime soon. Just loose the links; they screw up formatting for I.P. users and not a damned editor here is willing to read even a dozen of those articles they link to. Greg L (talk) 01:34, 30 October 2008 (UTC)
  • If it were just me opposing this you'd have a point. But it's not. There are a number of people opposed to this (on this very page, not to mention the archives where random editors seem to have turned up (and subsequently left, likely due to the attitudes of certain proponents of this change)), and it should be obvious that something like this is going to require slightly more than twelve unique editors to go around changing this on the entire wiki. Or maybe that's not obvious, in which case, my apologies. At any rate, I've offered to work on this on my own time. The current devs of MediaWiki also do so on their own time (and do so without compensation, well, except for Brion anyways). Now what needs to happen is a sane sensible discussion about what is needed (besides massive unlinking which is highly inappropriate given the amount of backlash about this). I'm generally only concerned with the autoformatting of dates (though I believe some date links would also be appropriate; I oppose removing all date links entirely, but I also don't think every single date should be linked (but they should be formatted)). So a helpful reply here might be going over what you'd like to see done, development-wise, to address date formatting. —Locke Cole • t • c 01:58, 30 October 2008 (UTC)

• Please don’t attempt to buttress your message by pointing out how many others feel as you. Do I really have to cite some world history where lots of people were solidly behind some thoroughly dreadful ideas? Stupidity in numbers is not a new concept to humanity and Misplaced Pages is far from immune to this phenomenon. I think most of what is going on here is editors A) are used to blue dates and don’t like change, and B) don’t realize how thoroughly lousy of an idea it was to have made autoformatting in the first place. One editor was so use to the “blue” of date linking, he proposed simply using color as a way to highlight dates again. Talk about not being able to handle change…

  As to the rest of your objective, I rather agree with you and would like to have parser functions that knew what country the requesting reader was from. But I don’t see the imperative with trying to get autoformatting working. Were it me, I could think of much better uses for such a parser function. In fact, I see this perceived imperative to get date autoformatting working as emblematic of how decision making here on MOSNUM is brain-damaged beyond all recognition. I don’t mind looking at U.S.-formatted dates nor Euro-formatted dates one twit. Neither style bothers me. But I also know that not all readers of Misplaced Pages (readers, not editors), feel this way. Americans are a rather sheltered lot (big ponds of water on both sides, you see), and they’ve not been exposed to a variety of date formats. I simply think that we should have Euro-formatted dates for most articles and have American-style dates in articles closely associated with America. And be done with it.  Much too much argument and calories burned feeding pissed-off neurons has been devoted here to this trivial issue. Greg L (talk) 02:49, 30 October 2008 (UTC)

• P.S. Question: Just how long do you think it will reasonably take (really, really) to make a parser function and template (or magic word), that would know from what country a requesting reader hailed from and which could look to a lookup chart and categorize them into camps, like America/Europe or America/Commonwealth? Greg L (talk) 02:59, 30 October 2008 (UTC)

• While you may disagree with the various people who support date formatting/linking, to ignore them out of hand simply because you think they're "wrong" isn't how things work here. Anyways.. to your question: unfortunately I don't think we'll be getting a function can change the page based on the incoming readers nationality. MediaWiki (and Misplaced Pages in particular) uses a cache for pages to minimize server strain. Pages are rendered once then stored to be quickly fed to readers as needed. Only when a page is edited is the page re-rendered (or when a template used on the page is changed, etc). So changing it for each reader would break that (we could use Javascript perhaps, but that's messy and some people have an aversion to pages using Javascript). My suggestion would be to agree on a default date format, then provide a function that allows that to be overridden on a page by page basis. The priority for settings would go something like this: 1) logged in user preference would override page preference, 2) page preference would override default, 3) default. The last two would apply to logged out/anon/IP editors, the first would apply to people like you and I. Would this work as an alternative to detecting the readers date format? —Locke Cole • t • c 03:15, 30 October 2008 (UTC)

• I don't know who or where the "number of people opposed to this" are. There's a lot of hot air here. I've delinked probably a thousand articles, and how many complaints have I had? Well, I am pleased to inform you I have had exactly two. Two individuals who asked me to pay greater attention to the date format. Ohconfucius (talk) 04:58, 30 October 2008 (UTC)

• Personal complaints directed at your talk page? Yes, two seems like a believable number. Then you look at this page and the archives and note that the number of people complaining exceeds the number of people who supported this proposal when it was taken to a straw poll (that magic number appears to have been twelve). I think it's clear in that case that there's no consensus for this and we should work on something that will be more agreeable (fixing the issues presented by those who support this change, mainly that date autoformatting doesn't work for logged out users). —Locke Cole • t • c 13:30, 30 October 2008 (UTC)

• And this is because links to trivia that no one reads is unwise, and autoformatting that only benefits registered editors but screws it up for everyone else is unwise. Solution? Get rid of the autoformatting. And as long as the autoformatting is removed properly,  it’s a job well done. If there is anyone who still thinks linking to these trivia articles is a good thing, they should first read Misplaced Pages:Why dates should not be linked, and then go earn themselves a Sewer Cover Barnstar here. Greg L (talk) 06:40, 30 October 2008 (UTC)

• So you know for a fact that nobody reads these "trivia" pages? You've seen the server logs and/or have some method of determining the kind of traffic these pages get? And you think there's no possible technical solution to the dates presented to unregistered/IP users? No way to turn on autoformatting for them so dates appear consistent throughout the article? Can you be any more arrogant in making all these assumptions? —Locke Cole • t • c 13:30, 30 October 2008 (UTC)

Having just moved the IEC trolling discussion to a separate page, I'm tempted to do the same with the date linking discussions. Not that I want to censor them, but I don't see why this talk page has to endure so much traffic over what is a comparatively minor issue out of all the many it covers. Anyway linking is relevant to other guideline pages, not just this one, so it makes sense to have a separate page for discussion of that topic. Any objections?--Kotniski (talk) 10:49, 30 October 2008 (UTC)

So you want to shuffle these discussions from one obscure page to yet another obscure page that nobody watches and get even less attention (when the goal should be the opposite)? Are you serious? —Locke Cole • t • c 13:30, 30 October 2008 (UTC)

So you asked if there were any objections, and receiving one objection, you went ahead and moved the discussion anyway? Reverted. –xeno (talk) 14:06, 30 October 2008 (UTC)

I think it's better off here, for now. Hopefully, it can soon be archived. --Dweller (talk) 14:12, 30 October 2008 (UTC)

• “You've seen the server logs and/or have some method of determining the kind of traffic these pages get?” Locke Cole: To the extent that these trivia articles get any traffic at all is because so many dates link to those articles and we continually get new readers who are clueless as to what the links take them to. Few readers actually read these things. How do I know? Let me know when you’ve actually read the twelve on WP:Why dates should not be linked. Be honest though, I don’t think it is humanly possible to actually read twelve whole such articles. And even if you manage to actually succeed at that endeavor (“I did it and survived! That wasn’t so bad.”), that fact would come up a tad short of demonstrating that dates ought to be routinely linking to; these trivia articles—particularly the day-of-the-year ones like October 30—are rarely any more germane and topical to the subject matter than some horoscope. Greg L (talk) 20:56, 30 October 2008 (UTC)

• Werdna, a dev on MediaWiki responded here to some questions I had about date autoformatting and the possibility of an extension to add an XML style tag to format dates. As to your questions/response: Again, how do you know these things? How do you know they only get traffic because of just the date links? Why do you think new readers would be at all confused about clicking on a date link and being taken to an article with information on that specific date when every other article link functions the same way (taking you to something relevant to the linked text)? You point to your very unscientific (and unconvincing) personal evidence that nobody reads them, but do you stop to think that they're less intended to be literal articles (read top to bottom) and more to be an interesting collection of events on a specific date? Anyone who seriously reads them (like a book and not like a list) has other issues to deal with. So, now that we have a dev stating that the changes people want would be easy enough to implement, can we please stop with the delinking until we've sorted out how to move forward? —Locke Cole • t • c 23:11, 31 October 2008 (UTC)

Shuffling to WT:Date linking

OK, let's combine this with my suggestion above and take this topic to a separate page. Further discussion will be at WT:Date linking. (Just let me set this up...) —Preceding unsigned comment added by Kotniski (talk • contribs) 13:55, 30 October 2008 (UTC)

Reverted per WP:BRD, please gather consensus before shuffling the discussion (I disagree, it'll only serve to stifle it as people won't have the new page on their watchlist - might as well keep the battleground here so that it can be archived properly into the archives of this page). –xeno (talk) 14:09, 30 October 2008 (UTC)

Sigh. Whatever. But I"m sure many people have taken this page off their watchlists due to the absurd amounts of traffic generated by one or two marginal issues like this one. People who really consider this an important issue could watch the other page, and that would allow people who don't care about date linking to come back here and talk about other, perhaps even more important things.--Kotniski (talk) 14:15, 30 October 2008 (UTC)

I don't think it's worthwhile to have the discussions scattered over many different pages. Let's keep it here, archive it here, and future generations will know where to look for the great Misplaced Pages Date Delinking Debate. –xeno (talk) 14:17, 30 October 2008 (UTC)

Footnote following "Dates ... should not be linked, unless there is a reason to do so."

Hello.

I spotted this clause with its footnote that unhelpfully does not point to any particular section on this page and I'm breaking out in a rash at the thought of reading all the gumpf up there.

The current wording is unhelpful and the footnote even less so.

I suggest the unhelpful footnote is discarded, which will allow users to argue "reason to do so" at the talk page of the individual article if they believe there's a specific need for a specific instance.

How does that sound? --Dweller (talk) 13:14, 30 October 2008 (UTC)

I also like someone's suggestion above that "the first date in an article should be linked" - this also helps to alleviate the worry that date articles will eventually be orphaned. –xeno (talk) 13:36, 30 October 2008 (UTC)

Sounds a bit artificial, confusing to editors who don't know the rule - and orphaning is a non-problem since date articles are themselves massively interlinked.--Kotniski (talk) 13:47, 30 October 2008 (UTC)

See WP:Walled garden and the definition of isolated article at WP:O – we don't just want articles not to be orphaned, we want them to be reachable by a finite number of clicks from the Main Page. Personally I'd say "by a finite number of clicks from any other article", but maybe that's just me. -- Army1987 (t — c) 10:08, 1 November 2008 (UTC)

(ec) The footnote could certainly go AFAIC. And after these eons of discussions we must be in a position to say at least something more specific than "unless there is a reason". I propose adding (after the quoted clause) a sentence like "Date items are normally linked within articles on date- and year- related subjects." (wording to be improved). --Kotniski (talk) 13:42, 30 October 2008 (UTC)

This should also resolve the orphaning issue.--Kotniski (talk) 13:44, 30 October 2008 (UTC)

Before this becomes complicated, can I suggest that in this thread, we stick to the question: Can we just drop the footnote? I'll create a new section below for the wording of the main text. --Dweller (talk) 13:47, 30 October 2008 (UTC)

Agreed - drop the footnote. It's unhelpful and strange. –xeno (talk) 13:52, 30 October 2008 (UTC)

I agree: the time has come to remove this troublesome footnote, which is out of keeping with the register of a style guide (points to localised squabbles and reduces the authority of MOSNUM). Yes, we definitely need to write in an exception for chronological pages themselves; and any fear of orphanage should be allayed by the opportunity to insert into the "See also" section of an article dates that are deemed by a contributor to be especially important, explicated by a piped phrase: that is a gateway to all sibline year articles. Editors are much less likely to question this than the formulaic blueing of chronological items in the body of the prose.

There is nothing special about "the first date in an article", and allowing it to be blued will open the floodgates for visitors and newbies to blue all dates/years. We do not want a return to that. 13:52, 30 October 2008 (UTC) —Preceding unsigned comment added by Tony1 (talk • contribs)

OK, footnote's gone. I don't think it was that controversial - more tricky is agreeing any change to the text that preceded it, and I've left that. See below. --Dweller (talk) 13:57, 30 October 2008 (UTC)

Excellent! --HJensen, talk 22:32, 30 October 2008 (UTC)

Text of: "Dates ... should not be linked, unless there is a reason to do so."

Kotnitski has suggested in the previous section that this text be reworded. Can we discuss this here, to leave the previous section for discussing the footnote? --Dweller (talk) 13:50, 30 October 2008 (UTC)

Dweller, just be bold and put the silly footnote out of its misery now. Tony (talk) 13:53, 30 October 2008 (UTC)

Gone. Let's discuss the remaining text now.

I think that the existing text is slightly vague. I'd be happier with its tone if the word "specific" was in it, clearly showing that exceptions are not the rule.

How about:

should not be linked unless there are specific reasons to do so.

Thoughts? --Dweller (talk) 14:25, 30 October 2008 (UTC)

Nice and succinct, but it would be nice to add a couple examples of a specific reason... –xeno (talk) 14:48, 30 October 2008 (UTC)

Try:

...for example, dates with articles, like September 11th 2001. Even then, linking of irrelevancies to the article is inappropriate, for example, linking a sporting event to the notable date that it happened to take place on.

? --Dweller (talk) 15:00, 30 October 2008 (UTC)

Don't understand - the date in your example doesn't have an article, the article is about the attacks. Or do you mean in contexts like: "...he noted that Al-Qaeda was responsible for September 11 and...", where the date has come to signify other events?--

An example of a date worth linking to is February 30, 1712. Note the colon, which defeats date autoformatting. --Gerry Ashton (talk) 16:31, 30 October 2008 (UTC)

<ec> To be honest, I really struggled to think of an example of when it would be justified to link a date. Please do suggest an improvement! If it was down to me, I'd have left it as per my first suggestion, without an example. Following the ec, I see the suggestion of Gerry, which is nice and quirky. --Dweller (talk) 16:35, 30 October 2008 (UTC)

I see in fact that this matter is already addressed quite satisfactorily at WP:CONTEXT (there is also a section at WP:MOSLINK, but that doesn't say any more than this page does). So I suggest wording like the following: "Dates (years,....) are not normally linked, except in articles on specifically date-related topics. For more information about circumstances in which it is appropriate to link dates, see WP:CONTEXT#Dates." Then at least any further debate on this topic could be kept in one place (i.e. there).--Kotniski (talk) 17:01, 30 October 2008 (UTC)

Not a bad idea, but I don't really see consensus for the recent changes there, either.... — Arthur Rubin (talk) 19:56, 30 October 2008 (UTC)

Sorry to break in, but I actually think the wording as it is does a good job. It tells that dates should not be linked, and if they should, editors should agree on a good rationale for linking. I think the current formulation conveys that idea fine. I don't think examples are needed. We don't provide examples for when a policy guideline should not be followed (but just says it shouldn't when common sense and the occasional exception will improve the article). --HJensen, talk 22:40, 30 October 2008 (UTC)

OK, but if the same issue is addressed in two (or even more) different guidelines, then they should either say the same thing or (preferably) have all the information in one place and summaries with links to it in the other places. --Kotniski (talk) 07:07, 31 October 2008 (UTC)

maybe just a sidelight here (sorry) but this part of WP:CONTEXT#Dates is pretty confusing/misleading:

"If an explicit link is provided, preferably in the lead, it alone can be the gateway for the reader to access the available sibling articles for other years (e.g., ]) ..."

why is it suggesting providing "explicit" year-in-X links and then giving an "aliased" year-in-X link as an example? to me "explicit" means "not aliased". Sssoul (talk) 07:35, 31 October 2008 (UTC)

I think this matter would be best discussed over there. I'm going to add a link to that section from this MoS page.--Kotniski (talk) 08:40, 31 October 2008 (UTC)

• Misplaced Pages:Manual_of_Style_(links)#Piped_links says: "Avoid piping links from "year" to "year something" or "something year" (e.g., ]) in the main prose of an article in most cases. Use an explicit cross-reference, e.g., ''(see ])'', if it is appropriate to link a year to such an article at all. However, piped links may be useful:
• in places where compact presentation is important (some tables, infoboxes and lists); and
• in the main prose of articles in which such links are used heavily, as is often the case with sports biographies that link to numerous season articles."

MOSLINK also says: "It is possible to "pipe-link" words that are not exactly the same as the linked article title—for example, ]. However, ensure that it is still clear what the link refers to without having to follow the link."

I believe that the way around this is to insert into the "See also" section at the bottom of the main text an explicit link to one or two of the most most important year-in-X pages for the topic; this is ideal for providing a gateway to all of their sibling year-in-X pages. The horrid likelihood we must all face is that readers almost certainly ignore links that look as though they lead to the same trivial, irrelevant solitary year facts that they've learnt to ignore over the years. I would like to remove the third bullet from the first quote above. Tony (talk) 12:10, 1 November 2008 (UTC)

yeah, that third bullet contradicts what it says at WP:CONTEXT#Dates about "multiple links throughout the article unnecessary"; it ought to be removed. adding something about piping plain-year links to highlight their relevance - Other notable events of 1955, for example - seems like it would be a lot more pertinent. Sssoul (talk) 12:28, 1 November 2008 (UTC)

Links in "See also" sections should never be piped, as they are supposed to just suggest the reader that other related articles exist, so they should use the real title. If it is not obvious what the article at the other end of the link 1955 is about, the fault is of the article's title, not of the unpiped link. Personally, I'd move an "article" like 1955 to List of 1955 events, and make 1955 a proper article, like 1345 – or redirect it to the decade (e.g. 1950s) in the case of a year for which writing such an article is unfeasible. -- Army1987 (t — c) 13:11, 1 November 2008 (UTC)

Yes, I thought this a while ago until someone (Sssoul, was it?) suggested the "See also" solution with pipe. I think the example was * ], which might beckon the readers in more effectively than just the blue year under "See also". Same for year-in-X links in "See also": ]. It doesn't have to be piped, but allowing piping for such links leaves open the option of more reader-friendly items, where the editors can think of them. There was a good deal of angst to the overall solution when raised above (or is it archived now?), but I suspect most of this came from those who were letting off steam about the removal of date autoformatting and chronological links in general. I'm keen for selective, explicit gateways to chrnological pages, bolstered by my belief that its' really hard to get readers to click on anything, especially 1998 and the like. Tony (talk) 13:21, 1 November 2008 (UTC)

IMO, the title of an article should be the most likely caption for a link to it, plus optionally a disambiguator in parentheses. If you think that a link like 1998 is not useful, why keep that article there, rather than moving it to List of 1998 events, where it belongs? -- Army1987 (t — c) 13:32, 1 November 2008 (UTC)

renaming pages "like 1955 to List of 1955 events" is a great idea; and if/when that's done, links to those pages should be kept explicit, just like year-in-X links should be kept explicit. meanwhile, my main point was that that third bulleted item at Misplaced Pages:Manual_of_Style_(links)#Piped_links that Tony1 pointed out above really needs to go. the suggestion about piping bare-year links to make their relevance explicit was just a sidelight; and even if that's against some immutable rule about "see also" sections, it would still make sense to pipe them in the body of the article, footnotes, etc: "For more context see ]." unless/until all those Events-in-Year pages are renamed, that is. Sssoul (talk) 14:05, 1 November 2008 (UTC)

Excellent and most logical suggestion. Maybe I'll start first? Ohconfucius (talk) 14:30, 1 November 2008 (UTC)

But deciding whether to link or not an article should be based on the potential of the article, not on its current status. Otherwise, there would be bots going around removing all red links they encounter (as red links are of no use whatsoever to readers who aren't also editors), and there would be no suggestions against the avoidance of linking redirects with possibilities (see the sentence immediately after the bulleted list in MOS:LINK#Piped links, and elsewhere).

If someone makes 1955 a redirect to 1950s (or an article like 1345), links such as 1955 will immediately stop being useless. (I'm not advocating linking years everywhere, only when they are particularly relevant to the context of the article.) So there is no hurry to change ] to ], IMO. (If you do need something more explicit for "See also" sections, I'd suggest "* 1955, for other notable events of that years", which is the style I'd normally suggest when the relevance of any link in the list isn't obvious. But, IMO, it's better to fix year articles than links to them.) -- Army1987 (t — c) 16:25, 1 November 2008 (UTC)

overlinking and remnants of autoformatting means bare-date links currently appear to be meaningless. that's the reason i see for making the relevance of date links explicit, when they are indeed relevant. that could be done by renaming the articles, as you suggest, or by piping the links with an explicit statement of their relevance.

i don't get why redirecting 1955 to 1950s would be desireable; if a link to 1950s is appropriate, an explicit link to 1950s should be used.

meanwhile, though, weren't we discussing Misplaced Pages:Manual_of_Style_(links)#Piped_links? Tony1 is right that that third bullet should be removed. Sssoul (talk) 16:41, 1 November 2008 (UTC)

What would be appropriate for 1955 would be an article, like the one for 1345; but, until such an article is created, a redirect to 1950s is a much better placeholder than the current 1955 article. Read both articles, and decide which one would give more historic context to someone reading them. -- Army1987 (t — c) 16:47, 1 November 2008 (UTC)

yes, a link like "For more context see: The 1950s" is probably better for most purposes than "For more context see: Other 1955 events". but i don't agree with the idea of making a link to 1950s look like a link to 1955, either through redirects or any other form of disguise. if a link to 1955 is of no value in a given article, it should be unlinked; if a link to 1950s would be valuable, it should be an explicit link. Sssoul (talk) 16:59, 1 November 2008 (UTC)

But I'm not suggesting either ] or a redirect as a definitive solution. I'm suggesting a redirect from 1955 to 1950s as a temporary placeholder until a more specific article, like 1345, is written. -- Army1987 (t — c) 18:23, 1 November 2008 (UTC)

i understand - and disagree. my view is that if a link to 1955 isn't useful in a given article it should be unlinked (or be replaced by an explicit useful link), not "temporarily redirected" to another page. Sssoul (talk) 18:35, 1 November 2008 (UTC)

← So, we should remove all red links, all links to redirects with possibilities, and possibly even all links to very short stubs or very bad articles, until a decent article is written at the link target. Or is there any reason why years should be treated differently? -- Army1987 (t — c) 20:50, 1 November 2008 (UTC)

smile: 1955 isn't a red link or a "redirect with possibilities", and if it isn't useful/relevant to an article it should be unlinked just like other links that aren't useful/relevant. WP:MOS_(links)#Overlinking_and_underlinking seems quite clear about that, and it treats low-value year links the same as other low-value links. and if a link *is* useful/relevant, it should be explicit, not "disguised" as in your ] example. by "temporary redirects".

meanwhile, can we get back to the subject of how to phrase "Dates ... should not be linked, unless there is a reason to do so"? thanks Sssoul (talk) 21:43, 1 November 2008 (UTC)

I'll quit this discussion as it's obvious nobody agrees with me, but I just wanted to point out that I didn't actually argue for ]. -- Army1987 (t — c) 02:12, 2 November 2008 (UTC)

Oh Great and Merciful Developer… hear our wishes

Would some developer please checkout Bugzilla 13025 as well as 15677? We beseech thee to hear our wishes. Why have you forsaken us for six long months? We ask not that thee create an entire magic word. We understand now that our earlier requests were… immodest. We humbly request only that you bless us with a character-counting parser function. Without such a parser function, the fields of our current number delimiting templates are bare. Our eyes thirst for nice functioning templates that have no bugs. Our midwives weep. We beseech thee to give us but a half-hour of your time to give us the tools so that we can fish the templates and feed our articles. Greg L (talk) 23:55, 30 October 2008 (UTC)

Ramen. Headbomb {_{κοντριβς} – WP Physics} 04:24, 31 October 2008 (UTC)

Date links and date pages

Since several editors are out unlinking dates, the topic of the date pages (January 1, etc.) needs to be addressed. The date pages have a particular formatting layout that should be maintained regardless of whatever consensus is reached on unlinking dates. I have seen a few (not many) instances of unlinking (usually script-assisted) of years in the date articles. These articles should be explicitly excluded from this practice - if not, a real mess could be made of things. If there is widespread feeling that these items should be unlinked, it should be brought up and discussed at WP:DAYS because it might require an overall layout change to the date pages in order that the pages 1.) are updated across the board and 2.) are kept in a format that maintains the utility of the pages. -- Mufka 01:39, 1 November 2008 (UTC)

Can you be explicit about which sets of articles? Timelines, for example, could do with a guideline about the formatting of non-linked month–day items when they head each item in a sequential list (I'd go for bold). Tony (talk) 04:16, 1 November 2008 (UTC)

My mistake, I thought saying January 1, etc. was sufficient. I mean January 1 through December 31 only. The days of the year. -- Mufka 13:46, 1 November 2008 (UTC)

I'm certainly in favour of writing in an exemption for the linking of chronological items within articles on explicitly chronological topics, such as January 22, 1970s, 19th century, 2002; but not, of course, for the date autoformatting/linking of full dmy/mdy items (which confuses the issue for the readers). The fact that dm/md items are autoformatted for logged in, preferenced Wikipedians is irrelevant: the reason for square-bracketing them is to link to sibling articles. Does anyone else agree? If there's a general feeling that this might be acceptable, we could raise it at MOSLINK. Tony (talk) 13:57, 1 November 2008 (UTC)

Developer help with dates

I got a response from a dev with SVN access to MediaWiki to answer a couple of questions here. As this appeared to be the primary reason for unlinking dates and removing auto formatting, can we please stop the mass unlinking that's going on? —Locke Cole • t • c 23:00, 31 October 2008 (UTC)

The current date preferences choice is unacceptable. By offering the format choice "2001-01-15T16:12:34" it creates an overwhelming implication that date autoformatting obeys ISO 8601, because that very specific format is strongly associated with that standard. However, data autoformatting does not actually conform to the standard. Since the means to invoke date autoformatting is unacceptable, date autoformatting is also unacceptable. --Gerry Ashton (talk) 03:46, 1 November 2008 (UTC)

An implication is not a contract. The preferences do not say it is ISO 8601, it only outputs something resembling that. Similarly, AFAIK there was no contract saying ISO 8601 was valid as input for wikitext, just that it would convert linked dates to whichever format the user had chosen. I read your essay (linked from your user page) and I agree that it could be improved. So instead of dismissing auto formatting out of hand as you have done, why not propose some changes or outline how you believe it should work? —Locke Cole • t • c 05:09, 1 November 2008 (UTC)

• Why is there this dogged insistence from one person that there should be DA? While there may be dispute to de-linking dates, there is an undisputed consensus that it is deprecated, so its not an issue any more. It's dead. So what, even if it were browser-configured, automatic and totally transparent to the user. Why should we be bothered to put in the effort to discuss it, let alone have to get developers involved? It's not as if we have a policy to use ISO dates which is not prose. We have date formats perfectly comprehensible to all, so lets 'fuhgeddit'! Ohconfucius (talk) 05:29, 1 November 2008 (UTC)

• Er... it is disputed, in fact. Or did you miss the various people who have turned up on this page (and in the archives)? And I fail to see the problem with date autoformatting: it allows editors to use any date format they feel comfortable with without making articles inconsistent in date display. That's a good thing, last time I checked, and is something worth working for. —Locke Cole • t • c 05:54, 1 November 2008 (UTC)

• Let’s not count our chickens before they hatch. All the developer (Werdna) said is that accomplishing what is desired wouldn’t “be too difficult.” That comes up quite a bit short of anyone committing to do something about it (or finding someone who gives a damn) and is a light-year short of actually receiving a finished product. Now that we actually have specific knowledge of a developer (someone who has been programming since he was 14 years old), I suggest we try to further develop this relationship. It would sure be nice now, to be able to present a unified front as far as our requests go. A big turn off for any programmer would be bickering and changing our minds. Greg L (talk) 06:12, 1 November 2008 (UTC)

• • You are absolutely correct. He hasn't committed to doing it, nor is it done. But we also haven't come up with a specific plan or design (which I've noted further up on this page) in order to proceed. Once we have that we can go back with specifics and ask (or as I've offered, I can work on developing it). FWIW, I've been programming since I was 16 or so (circa 1991), but I'm also older than Werdna. :P At any rate I think any dev would be reluctant to work on something without a clear decision on what should be done and how it should work. —Locke Cole • t • c 06:35, 1 November 2008 (UTC)

Locke Cole, how would the ISP mechanism deal with Canada, which uses both formats? Or India, which uses both? That is just one issue that will stuff it all up. And someone will need to make a choice about an awful lot of countries; it's by no means a hard-and-fast rule as to which use which: have you had a look at this article? Enough to send shivers down the spine of any programmer. Tony (talk) 12:16, 1 November 2008 (UTC)

Locke Cole asked "why not propose some changes or outline how you believe it should work?" My preference is to read dates that follow the same variety of English as the rest of the article, so I would not use date autoformatting if it were offered. So I'm not going to help design date autoformatting, but I will point out problems that may result in date corruption, or false claims that we conform to a standard. --Gerry Ashton (talk) 14:45, 1 November 2008 (UTC)

Tony, I don't see anything in the questions to Werdna asking for IP geolocation. Trying to implement that would be a Bad Idea, but I think Locke Cole simply wants a single default (at least, a single default for each article) for all preference-less readers. While I don't see this kind of autoformatting as particularly necessary, I also don't see it as particularly harmful, if it can be safely done without requiring editors to mark up all dates in some way, especially not as links. ("Safely" here refers to problems like detecting if dates are within quotes or detecting whether extra commas are needed, etc. It still doesn't sound like a perfectly simple task to me.) -- Jao (talk) 15:28, 1 November 2008 (UTC)

Regrettably some type of markup would be necessary I suspect because there's bound to be false positives in any function that tries to detect and format dates automatically on the server side. But is that really such a bad thing? We already have all kinds of markup in the form of templates and parser functions, a simple XML-style tag to surround dates doesn't seem like that big of a deal, really. —Locke Cole • t • c 00:16, 2 November 2008 (UTC)

On second thought, your arguments are starting to get to me here. Not as to its usefulness, I still think it's unnecessary, but that's just me (well, and a lot of others). But I think I could find this type of DA acceptable. Say we accidentally have "<date>January 20, 1961</date> – November 22, 1963" in John F. Kennedy. To almost everyone, this inconsistency will be hidden, so they will not know it should be fixed. But some will come there logged in with non-US date preference, and they will see "20 January 1961 – November 22, 1963"; eventually, in all probability, one of them will fix it. I still don't really see how it's better than no DA, but it sure beats the old DA. Even if this will be implemented in the (near or far) future though, I don't see a case for not unlinking dates. I understand that a bot could convert old link-markup to new tag-markup, but it would still not catch all dates, because all dates are not linked as it is. -- Jao (talk) 01:30, 2 November 2008 (UTC)

Although the comma issue (also see Tony's comment in the section below for that) would still have to be solved, of course. -- Jao (talk) 12:10, 2 November 2008 (UTC)

• It’s not complex. Why is so much effort being devoted to this? A simple guideline for fixed-text dates for editors to use in writing new articles is all we need:

1. For articles on, or strongly associated with, the U.S. or its territories (or countries listed in this guideline that use U.S.-style dates: Micronesia and Palau), editors should use the U.S.-style date format (“February 2, 2008”), otherwise, editors should use the international date format (“2 February 2008”) in articles.
2. New articles on or strongly associated with Canada should use the international format but, for existing articles related to Canada, whichever format was used by the first major contributor shall be retained.

Nothing more complex that this is needed. We don’t need to shelter our readership from the occasional *shock* of seeing dates in a less-than-customary format. Whereas such dates may be unused by certain readers, dates in either format confuses absolutely no one. Greg L (talk) 18:35, 1 November 2008 (UTC)

So much effort is being devoted to it because this isn't a print encyclopedia, there are opportunities here for better presentation than simple unchanging static text. —Locke Cole • t • c 00:16, 2 November 2008 (UTC)

• I agree Locke. I truly look forward to country-sensitive parser functions. Templates that use these parser functions will be extremely useful tools. They would allow us to create templates that group countries into classifications. Then, text that truly interrupts the flow of thought (and is sometimes truly confusing) can be addressed. For instance, we could code {{dialect|Commonwealth|US|The solicitor put the suspect’s colour-coded files it the boot.|The attorney put the suspect’s color-coded files in the trunk.}}.

  My concern is that just because someone left a message on the talk page of a developer and received a response (*collective audience gasp*) mustn’t be construed as a reason for delaying the deprecation of formatted dates. Bugzilla requests have historically taken a l-o-n-g  time to get any action. Many Bugzillas never get any action whatsoever—not even a response. So, while it might be *pretty* to think this will be acted on soon, it just isn’t realistic to assume as much. Nor is it wise, IMO, to postpone meaningful action in the mean time. This autoformatting is really junking things up for regular I.P. readers. And… (I might catch flack for thinking this and having the chutzpa to actually express it here) I also think the trivia articles these links take readers to are simply appreciated far too rarely by the typical reader. I think the best thing to do is keep on bot-delinking so long as the bot makes the articles read better for I.P. readers than what is currently there. Greg L (talk) 01:06, 2 November 2008 (UTC)

• Of course it could take time for this to happen, while the issue of overlinking and inconsistent dates should be one solved now, but this going to the point that the auto-correcting change that people have done on dates should leave some computer-understandable bits around that do nothing to the markup to the reader so that when this happens (now that there's progressive talk on the devlist), reverting dates to the new approach will be trivial. Otherwise, editors, with the dates completely stripped of such codes, will have to manually process articles. There's a number of ways to do this, all that are still compatible with scripting tools, so its not like the breadcrumbs can't be left. --MASEM 12:26, 2 November 2008 (UTC)

I really, really don't like that Canada wording. Suppose User:JoeCanuck, a WP newbie, writes an impressive article on Waterfowl in the Gulf of Saint Lawrence, using the U.S.-style format. Someone changes all the dates to international. JoeCanuck asks about this on the talk page, getting the answer "for articles on Canada, you should use the international format". "Ah," he answers, goes to Gulf of Saint Lawrence and spots a U.S.-style date there, using his newfound knowledge about WP to "fix" it. Someone reverts him; new talk page answer: "That rule doesn't apply here, because this article was created before November 2008." I wouldn't be surprised if we never heard from JoeCanuck again after this. All right, so maybe I'm exaggerating, but I think that a policy that distinguishes between old and new articles is never a good idea. Apart from that, I still think this proposal (as well as the one which does not favour a specific format by default, only taking a stance for articles connected to certain English-speaking countries) is very clear and simple to follow. -- Jao (talk) 01:30, 2 November 2008 (UTC)

Just my two cents. I would propose something more-or-less like:

• Every article should consistently use one date format, except within direct quotations (and in tables etc. using YYYY-MM-DD);
• If a significant majority of the full dates in the article refer to events happening in places where the month-day-year format is commonly used in English, use that format; likewise for the day-month-year format.
• If all dates refer to events happening in places with no significant English-speaking population, or in places where both formats are in common usage in English (e.g. Canada), or if roughly the same number of dates refer to events in places using each format in English, just choose either format. The format choosen should preferably be consistent with the variety of English used in the article, i.e., use 2 November 2008 for articles written in British English, and November 2, 2008, for articles written in American English.

Maybe the wording should be tweaked, but I think nothing more complex or arbitrary than that is needed. -- Army1987 (t — c) 02:35, 2 November 2008 (UTC)

Army1987's formulation does not fit well for articles that are not about people or places at all (for example, Standard deviation). Such an article could use either format, even if some of the people connected to the topic are from an English speaking country (for example, Francis Galton). Also, articles about creative works that are not set in any particular real location should use the format associated with the author, if the author spoke English or lived in an English-speaking country. --Gerry Ashton (talk) 03:34, 2 November 2008 (UTC)

Well, in Standard deviation there will usually little reason to write a full date (with the day of year), anyway; the one about fiction is a good point. I just fail to see the reason why an article written in American English should not use the November 2, 2008, just because it is not specifically about the US, even if maybe the only two or three dates in it refer to events which took place in the United States. -- Army1987 (t — c) 09:15, 2 November 2008 (UTC)

• Splendid Jao, a one-rule guideline for dates. I like simplicity:

1. For articles on, or strongly associated with, the U.S. or its territories (or countries listed in this guideline that use U.S.-style dates: Micronesia and Palau), editors should use the U.S.-style date format (“February 2, 2008”), otherwise, editors should use the international date format (“2 February 2008”) in articles.

Does that work for you? Greg L (talk) 05:06, 2 November 2008 (UTC)

For me personally, yes. I don't know if forcing the day-month format on Canadian topics is a wise decision, but speaking only for myself, I would have no problems with it. I also, of course, enjoy the simplicity. -- Jao (talk) 11:01, 2 November 2008 (UTC)

There would be a lot of complaints. Tony (talk) 11:13, 2 November 2008 (UTC)

Even simpler: "Each article should generally use the date format commonly used in the variety of English in which the article is written: for example, use 2 November 2008 for articles written in British English, and November 2, 2008, for articles written in American English. (For varieties of English in which several different formats are in widespread use, such as Canadian English, just choose any one of them and use it consistently; in case of doubt, use the format used by the first major contributor to the article.)" Together with WP:ENGVAR, it automatically requires British format for British topics and American format for American topics, and doesn't impose absurd restrictions such as that Gasoline (which is written in US English) should use day-month-year. -- Army1987 (t — c) 17:30, 2 November 2008 (UTC)

(unindent) Yep, this issue (like the linking one) was also discussed ad nauseam and with wide community participation, consensus was pretty clearly reached in the end. I don't see much point continually raising these same issues. We should be glad they're settled, even if the consensus isn't exactly what each of us personally would prefer. It's not as if they're a big deal or anything. I suggest people find some new and more significant things to discuss - there's still plenty out there.--Kotniski (talk) 12:37, 2 November 2008 (UTC)

I haven't quite finished reading all the new discussion, but there is a thing I'll say right now: Will people please stop pretending that Locke is alone in this? You have read the discussions, you were there, so you know he isn't, and pretending he is is very dishonest. Shinobu (talk) 09:06, 3 November 2008 (UTC)

(Okay I think I've mostly caught up.) Firstly, there is no consensus for turning dates into plain text, regardless of what some editors here say. I've waded through all the discussion a while back, and now again through the new discussion, and it's quite clear there is no consensus for that. Now, the time I can devote to dredging through MOSNUM talk is limited, so when I post a comment, it usually sinks in the archive and I'm forgotten as it were, while some other people apparently can post daily and leave much more of a vocal impression. I'm not saying people shouldn't be allowed to do that, au contraire. But when these people then start to pretend that we slowpokes don't exist and start to derive from that a kind of pretend consensus, or rather start implementing their loved solution by fiat without consensus, they're in the wrong. Good, now that we've got that over with, on to the proposed solution, namely a kind of date markup that doesn't link.

According to a developer it wouldn't be hard to implement. This aligns well with my own programming experience: these kind of things tend to be rather trivial to implement, especially since most of the work has essentially already been done (tag detection, date formatting). (Note further that we could achieve the same effect with template markup and JavaScript for people who want another format than the default. So if a wikimarkup solution isn't viable for whatever reason, that shouldn't stop us. But it looks like it's simple to implement.) However, the developers need a clear guide on what to implement, because otherwise you end up implementing it in ten different ways, getting whistled back to the drawing board after people say ‘no we really had a slightly different thing in mind’. So what exactly do we want?

Syntax: <date>12 March 456</date> looks like a reasonable proposal. This would then also determine the default display format. Alternatives in the same vain: <date value="12 March 456" /> or template inspired: {{date|12 March 456}} or perhaps as a parser function: {{#date:12 March 456}} Perhaps we should ask a developer what is easiest.

Functionality: Now we're adding this, we might as well think if there's anything else we need or want. Should a page be able to have a default format or is it enough to specify formats in the tags themselves? Should registered users be able to specify a ‘short’ format, for use in tables, etcetera? If so, how do we want this implemented? Automatic detection from text entered, possibly with override? Should users be able to turn all dates into links, or to turn all date links off?

I have done enough programming to know that this should all be relatively simple to implement and I think it would pave the way to a solution acceptable to all of us. And this is of course a benefit of such a technical solution: instead of forcing what one group of editors wants down all our throats, a specialized markup based solution can potentially make everyone happy which makes it clearly the best road forward. Shinobu (talk) 09:56, 3 November 2008 (UTC)

Explain how "everybody" is going to be made happy by being told that whenever they want to write a date in an article they're editing, they can't just write it, but they have to use some special syntax that apparently serves no useful purpose except to solve a problem that never was a problem.--Kotniski (talk) 14:19, 3 November 2008 (UTC)

A date autoformatting that works without overloading linking is desirable as long as it has all the features we've asked for default settings and the like. The currently implementation sucks and needs to be dropped, that is a given. The devs say that a fixable version is possible. Thus, instead of dropping all easily-discovered computer-readable dates, we should try to find a format that in the current is simply a passthru to the data within but can be easily modified (whether a single template or bot activity) to make it work for the new DA. This will require additional markup, yes. Just like learning how to reference material properly, provide interwiki links, create tables, and so forth. It is no more difficult than other basic wikiediting tasks. DA was never a problem, it was always a feature and its a feature that we want to have as long as it degrades gracefully for unregistered/no preference users. The devs say it can be done , it's just not going to be today, so it seems silly to rush to make reinstalling the new DA more difficult by wiping the metadata of dates (the link brackets) with another piece of metadata that at the present time does nothing but help prepare for that. It's thinking forwards compatibility, not backwards. --MASEM 14:42, 3 November 2008 (UTC)

You really would have to be cross-eyed to find difficulty in reading month–day or day–month, whereever you come from. No one has yet answered why our readers and our editors are SO uneducated or blinkered that this would be an issue spending more than 2 seconds thinking about. Why are people wasting time here, hmmmmm?? Tony (talk) 15:23, 3 November 2008 (UTC)

Of course everyone can probably figure out day-month vs month-day. Having to do that once or twice in an article isn't going to kill anyone. But when you get to history-based articles with a lot of dates, someone reading in the date format that they aren't use to will be spending more time trying to figure out and scan visually for dates that are in the format they aren't use to seeing. DA (when done right with proper defaults and no linkages) is a usability feature we should want to strive for. Otherwise, why do we even argue over the date format of US+select other country-oriented pages over that from the rest of the world? Why not just go all international day-month-year style and never worry about date formatting again? If we are going to admit that US-centric topics should use US-style dates, then it also makes complete sense to want and desire for a DA system that works properly. --MASEM 15:29, 3 November 2008 (UTC)

If the dates are in a consistent format on a page, I don't believe that any reader will have any significant difficulty in figuring them out. If you've learnt to read English and use the Internet, you're almost certainly used to seeing both types of date (and even if you haven't, it's a no-brainer to figure them out). Spelling probably has a far greater effect on individuals' comprehension - why is no-one interested in making editors write {{altspell|colour|color}} or the like? The only answer seems to be that we've seen so much discussion and effort put into the specific question of dates (generated originally, I guess, by a few people who got over-emotional about what date format they preferred to see), that people have been misled into thinking there must be a real problem here. THERE ISN'T. Or anyway, not one that's worth developers, editors (and even readers) wasting any significant amounts of time on. This effort should be spent doing some of the countless things that really would make Misplaced Pages better.--Kotniski (talk) 15:44, 3 November 2008 (UTC)

Indeed. Trebor (talk) 19:50, 3 November 2008 (UTC)

Tony, as you seem to think we're all "wasting our time" here, perhaps you should move along and find something else to work on as this really seems to be bothering you. I'm sorry for that, I really am, but your view is (IMO) the minority view and you need to stop minimizing these things simply because you don't appreciate them. —Locke Cole • t • c 02:17, 4 November 2008 (UTC)

Fake apologies don't really solve anything, to be honest. Can we agree that linking dates to day or year pages is generally pretty dumb? And that having the vast majority of our readers (the ones who aren't logged-in) seeing a mish-mash of date formats is pretty dumb? And that with regards to a new method of date autoformatting, it will take a long time to get consensus here and an unknown (but, based on past experience, probably long) time for the developers to implement it? If so, it's still correct at the moment to be unlinking and standardising dates in articles. If not, which bit do you take issue with? Trebor (talk) 02:57, 4 November 2008 (UTC)

It wasn't a "fake apology", he seems to believe this is a "waste of time" and wonders why we're all "wasting time" on something he's apparently sorted out already for all of us. And I'm sorry, but that's just no the case at all. To your questions: I'm on the fence about year and date links, I don't see the harm in linking to these articles. But yes, having our regular readers seeing inconsistent dates within the same article (and often within the same paragraph or even the same sentence, I'm sure) is dumb. So let's fix that: by making auto formatting of dates work for regular readers. I don't know how long it would take to get consensus on date autoformatting (as far as how to implement it, I don't think there's any real opposition to it in principle). I do know that once we have something we can go to a dev with (something they can act upon) it shouldn't take long at all so long as it isn't unnecessarily complex. No, it's not correct to unlink dates: somehow Misplaced Pages has survived with them for all these years, I fail to see the immediate urgency in removing them now. —Locke Cole • t • c 04:31, 4 November 2008 (UTC)

I would forgive Lock for appearing to be confusing 'cause' and 'effect'. The current movement to delink dates rests on a script which is capable of harmonising date formats within an article in one fell swoop (except for ISO dates). This, I believe is the correct approach to fixing the problem to which Lock was referring. To use some sort of theoretical DA for the purpose is akin to sweeping dust under the carpet than to sweeping it up. I hope I will have un-confused you enough to get you off that fence, and to hoover under that carpet. ;-) Ohconfucius (talk) 04:41, 4 November 2008 (UTC)

I havent been part of this conversation for a month or so as things have been pretty busy in real life and I found myself getting annoyed by some of the same tones that Locke is responding to. The civility of the group has improved somewhat from two months ago, with dismissiveness the major remaining sticking point. To give credit where credit is due, Tony, Greg and others have been extremely persistent and focused on their goals. What concerns me is that it's an all or nothing goal which goes beyond guidelines to impose and execute their vision. I'd be much more impressed if participants in this discussion were actually editing more articles and not just running bots (though I do recognize that we all contribute in different ways).

I'd also like to suggest people do their best to avoid dropping back to their old circular logic when responding. WP:CHILL comes to mind. Someone new comes to the group and says "why"? Depending on who responds first, the thread quickly winds up touching DA, ISO dates, trivia, IP vs registered editors, overlinking, context, wording, what does specific mean, markup vs links, etc. It winds up with annoyed editors who feel there's a cabal here because the only people responding are those who have the passion, anger or persistence to keep following this talk page. It's not fair to say that only a few people have an opposing point of view. It is fair to say that many of them run away from here quickly. I just wish there was some middle ground we could find without having to stake out far end of the spectrum positions. dm (talk) 05:19, 4 November 2008 (UTC)

'Forced' date format

I hate to wade into this, as I am only a hack editor. I tried reading most of the above, and hope this doesn't conflict. Most of my editing has dealt with TV show episode lists and such. I somewhat agree that the dates listed in the tables don't need to be linked to anything, but do think that some robot or other script would still want an easy way to tell a date from the rest of the text. Does or could there be a template that would forcibly format a date, somewhat like {{forcedateformat|USFull|2008-11-01}} would output an unlinked November 1, 2008, no matter what the user settings are. Andyross (talk) 16:51, 1 November 2008 (UTC)

Not a template as such, but the software does indeed contain such a feature. To use it, you type the following characters: November 1, 2008. Miraculously, this produces precisely your desired output. And this looks so much like a date that it can be recognized as one by bots and humans alike.--Kotniski (talk) 18:08, 1 November 2008 (UTC)

That's not quite fair. As I noted in another one of these interminable threads (with different specifics), on January 23, 1900 people were killed. — Arthur Rubin (talk) 18:12, 1 November 2008 (UTC)

Maybe it happened in Albany, New York's capital. But I don't see any calls for putting templates round all phrases containing commas just because they might be misinterpreted by some hypothetical bot.--Kotniski (talk) 18:50, 1 November 2008 (UTC)

Above, we have an editor who wants a sort of return to date autoformatting so that all users see the same date format displayed according to what they prefer. Here we have an editor who would like to be able to force all editors to see a date in the 'same' format, no matter what his preferences are. Funny, takes all sorts to make a world! Ohconfucius (talk) 00:52, 2 November 2008 (UTC)

In the American date format the year is a parenthetical phrase and should be followed by a comma, so, on January 23, 1900, people would have been killed, if they weren't one thousand and nine hundred people killed on January 23 of current year. -- Army1987 (t — c) 02:17, 2 November 2008 (UTC)

Many many US writers will disagree that a comma is required, except on rare occasions for disambiguation. This is one of the reasons that dates should be left for editors to control (according to our guidelines), and not handed over to another programmer's toy. Why is the order of month and day so important, anyway? Tony (talk) 10:10, 2 November 2008 (UTC)

I'm a native speaker, and I'm skeptical many many US writers will disagree that a comma is required. Please cite some. Septentrionalis PMAnderson 18:03, 2 November 2008 (UTC)

Part of my idea was to make it EASIER to adjust date formats as Wiki standards change, or if an article just needs to be updated. By wrapping dates, something as simple as a find/replace could change the date formats. Maybe even putting in a {{fdf|USLong}} would default everything after it to a particular format. And, one of the 'formats' could be AUTO, so it could cover all options. Andyross (talk) 14:26, 2 November 2008 (UTC)

And another nice reason for wrapping dates: By setting a format at the beginning of the article, you can keep the format consistent throughout the article. I know one of the issues discussed were articles with mixed date formats. That's so much of an issue when you have multiple editors and countries. By wrapping dates in a template, other editors will hopefully also wrap theirs, giving a consistent look to the end user. Andyross (talk) 21:43, 2 November 2008 (UTC)

By not wrapping dates in a template, other editors will hopefully not wrap theirs, and everything will be very VERY easy - you write what you want the reader to see, as most editors do already and will always continue to do. Honestly, I've never seen so much discussion of potential solutions to a total non-problem as I have over this dates issue. --Kotniski (talk) 08:45, 3 November 2008 (UTC)

Unnecessary vagueness

• Misplaced Pages:DATE#Unnecessary_vagueness

The last example in the table is an example of a misplaced modifier not unnecessary vagueness. (dogs with unbreakable parts vs. toys with unbreakable parts) Since the error was obvious, and I'm good at tables; I'll go ahead and remove it for now. If you have any objections, please change it back.--Hamster2.0 (talk) 18:10, 2 November 2008 (UTC)

We have an assortment of toys for physically active dogs with unbreakable parts.

We have an assortment of toys with unbreakable parts for physically active dogs.

PS: In case you want to put it back after a couple of edits without a "roll back," just paste this before the end of the table (aka. the |} symbol.)--Hamster2.0 (talk) 18:24, 2 November 2008 (UTC)

 |-
 | We have an assortment of toys for physically active dogs with unbreakable parts. || We have an assortment of toys with unbreakable parts for physically active dogs.

No consensus for mass unlinking of dates

I have been watching the debate for some time, not intervening because it seemed to be a pointless argument. But then I came across this edit , part of a mass unlinking to January 2. It seems to me that the so-called policy is now actively damaging the encyclopedia; whether or not there was a consensus to discourage further linking of dates is one thing, but it seems clear to me from the debates that there is no consensus to prohibit any links to dates or to remove all existing links. Breaking historical redirects without going to WP:RfD is just one example of the damage being done.--Rumping (talk) 18:59, 2 November 2008 (UTC)

• What's so special about January 2? The dates should all be delinked. ;-) Ohconfucius (talk) 01:41, 3 November 2008 (UTC)

I'm sorry, that edit was my fault. I was undoing the backlinks via Twinkle to get a list for AWB, and trying to not to touch any article that actually related to useful dates (the actual articles for September or July 5, for example). I must have missed this one. My sincerest apologies. NuclearWarfare _{My work} 22:04, 2 November 2008 (UTC)

Errors happen (I make them too—date auditing requires skill and judgement in some cases, and there are very occasionally slips on the wrong button); but that doesn't prove this specious and unsupported argument that cleansing this sorry function from WP is damaging it. Quite the opposite: DA and the linking of date fragments have both done significant damage for our readers, and need to be addressed as promptly as possible, not left to mar the project for years by some ultraconservative, anti-bold don't-rock-my-boat knee-jerk reaction. Please stop churning—it's not going to change anything and is wasting all our time. Tony (talk) 01:01, 3 November 2008 (UTC)

I agree. I wouldn't put it as forcefully as Tony does, but in my opinion (which I developed after reading around the issue a while ago), there is a good argument for mass delinking to avoid the mess seen by unregistered readers, and then rebuilding of any collateral damage. Though given the connotations of that phrase, "collateral damage" might not be the best phrase to use. The specific mistake pointed out above is just that - a mistake. Carcharoth (talk) 05:26, 3 November 2008 (UTC)

I still don't understand the "significant damage" though. Headings are bold, dates are blue. How does this impair the ability of our readers to navigate the articles? All wikilinks are optional - they're there so that if you want to explore a topic further, you can do it, presto. If you don't want to explore, don't click on it. In the case of year links, there are any number of reasons one may wish to click - in-line - to see what else happened in the same year as Apple introduced the Lisa, both in computing and in the world at large. The current structure of our year articles isn't particularly helpful here, focussing as they do on exhaustive lists of births and deaths - but that's a problem with our year articles, not the incoming links.

day-month linking is a little different, since apparently it is causing a problem with editors missing incompatible formats in-article (I wouldn't know, I purposely never set a date preference). Even here though, we're missing an opportunity. The top of each ] article could have a little message about how if you sign up for an account, you can have dates displayed in your preferred format. With one link to an explanatory page that would casually mention the other benefits of an account. Like having a watchlist, and how much we like helping everybody to edit here.

My main problem with this mass action though, is that it doesn't have to be done in a way that is solely destructive to embedded data. Why can't a neutral template be substituted instead? Let's say {{mmmdd}}, which would simply deliver its input? There's no expansion problem there, no expensive parser function, and it would be easy (and dev-independent) to add a class to make the wikilinks non-blue. And it would keep our ability to make future improvements, and the learn-by-see editors would find it easy to copy. There are lots of ways to go other than "I hate this and I want it gone now" - combined with "I'll use a script to do it, someone else can fix the mistakes". We've been down this path before, have we not? Franamax (talk) 07:09, 3 November 2008 (UTC)

It's the learn-by-see editors that are the problem - they will waste their time and intellectual energy using a template that serves no purpose. Honestly, WE DON'T NEED THIS - no-one would even be suggesting it if we hadn't previously had the autoformatting idiocy in place. Templates should be used to make things easier for editors by saving them from writing complex formatting or repeated text or whatever, not to complicate their lives just to preserve "embedded data" for some unspecified future purpose. If we wanted to mark such data, we wouldn't be starting with dates, which are 99.9% recognizable anyway.--Kotniski (talk) 08:54, 3 November 2008 (UTC)

That's glossing over a few of the points I just tried to convey. A simple template requires no great intellectual effort, thus none is wasted. Metadata that "serves no purpose" is only revealed to serve a purpose when someone figures out what purpose it can serve. And the template can begin neutral but can be made to serve a valuable purpose - it can be used for date autoformatting! Two very simple dev changes would be needed to make it happen: add a magic word for default formatting; and add cookie capability for IP users, so that the average reader could set a preference for their own PC. There are lots of ways to go, but "idiocy", "knee-jerk", "churning", "wasting all our time" - why don't we just MfD this talk page then? Doesn't seem to be much talking going on, though I've yet to see the consensus for mass de-linking. Franamax (talk) 09:12, 3 November 2008 (UTC)

Well, any template like this is an unnatural and time-wasting distraction for people who just want to write stuff. Date autoformatting is a totally valueless purpose, as has been pointed out many times. The only purpose I can see is that people might write scripts for extracting information from Misplaced Pages automatically (text mining and the like) - but (a) we don't do that with any other kinds of information, so why for dates? (b) dates are pretty much recognizable anyway, so why are people making such a fuss about marking them up and not asking for any other types of information to be marked up? The consensus is referred to via a footnote on this MoS page and anywhere else it's mentioned, so you don't have to look far. Stand back from this and you'll see what a pointless issue to keep raising and re-raising - I can only assume people are making the false deduction "there was a solution, therefore there must have been a problem". There is NO PROBLEM with just writing dates as ordinary text, just as there is no problem with writing anything else as ordinary text. So let's stop trying to solve the non-problem with schemes that would make editors' work harder, and stop continually disrupting the MoS and this talk page by raising issues that have been dealt with satisfactorily and require nothing further except implementation.--Kotniski (talk) 10:43, 3 November 2008 (UTC)

We certainly don't want to encourage newly registered editors to select a date preference: they need to leave the default option "No preference" untouched, and WPians who have already selected a format preference should seriously consider reverting to "No preference". Why? Because we are all responsible for cleaning up the mess that date autoformatting has led to. This includes spotting inconsistent formats in an article (easy to insert when the raw format our readers see in the remaining square-bracketed dates is concealed from us) and forcing the point on the global choice of format for an article, both through the display mode. These are essential to housecleaning, and are hampered unless "No preference" is chosen. Tony (talk) 13:36, 3 November 2008 (UTC)

Protected

I've seen too many reverts on this page popping up on my watchlist in the last day. Edit warring does not achieve anything, only discussion achieves progress. So I've protected the page for a week so people can discuss before editing. MBisanz 13:51, 3 November 2008 (UTC)

This may have been overkill. There are a number of different disputes about autoformatting, and they came up on the same day; there is no established war that I can see. I would suggest unprotection; I'm not sure this even amounts, at the moment, to a {{disputedtag}}, although the addition of this new issue on "specific" may boot it to that level. Septentrionalis PMAnderson 16:09, 3 November 2008 (UTC)

The sequence that hit me was , , , , , , , , , , . Does that not look like an edit war over a tag? MBisanz 16:14, 3 November 2008 (UTC)

Yes, that does; thank you for isolating it. (All except the first, which was the substantive edit that provoked the tag.) Septentrionalis PMAnderson 16:33, 3 November 2008 (UTC)

My opinions on the merits are

• there plainly is a dispute; going to the brink of 3RR to remove a tag is evidence of underlying dispute.
• I don't see the functional difference between reason and specific reason in this context, and I wish someone on either side of this nonsense would tell me. Some other wording should be possible. Septentrionalis PMAnderson 16:33, 3 November 2008 (UTC)

I tried addressing this issue in this section, above, but frustratingly the discussion left the point I raised and moved on to a whole new area (piped links). To answer your request, the nuance is that with a word like "specific", it demonstrates that there is no blanket "reason" for all uses of date linking, but that in specific circumstances, an individual date may be linked with unusual justification. --Dweller (talk) 16:52, 3 November 2008 (UTC)

If that's all, then it's redundant. Dates should not be linked, unless... already denies the existence of a general reason to link all dates, and the only one ever suggested (to serve autoformatting) is expressly denied at the end of the paragraph. Septentrionalis PMAnderson 02:51, 4 November 2008 (UTC)