Misplaced Pages

Revision as of 15:17, 23 October 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 6,943,852 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)
• 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)

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)

What is going on here w/r/t linked dates?

Came here from the autoformatting subsection, where I left some comments, only to find that the apparent "consensus" behind altering the date/year linking policy is only Tony's cherry-picked talk page (in short, it strawmans his opposition; downplays the opposition clear on the subsection page and here; and blames date linking for errors caused by autoformatting, which are far more efficiently solved by removing preference autoformatting if it's a legitimate problem.) Has the only vote so far been about British v American date formatting?

What gives? Is there really no consensus? And if so, why is the policy changing and why are bots being developed to auto"correct" existing pages? -LlywelynII (talk) 22:50, 2 October 2008 (UTC)

I second this, and have argued with Lightbot's owner about this before. Linking dates allows readers to use "what links here" on dates to find out what occurred on that date, and lets readers quickly see concurrent events worldwide for a given article's scope. This bot shouldn't be running until there's consensus. If the changes aren't noticed right away, it can be a real pain to undo it's efforts. -- Kendrick7 18:32, 5 October 2008 (UTC)

Autoformatting of dates is a pile of crap. It has been extensively debated for months. If Tony has a talk page that you don't think is convincing, that is a straw-man argument. Whether Tony's talk page is convinding or not, the concensus to not autoformat dates exists. Just read the talk page archives for this guideline. --Gerry Ashton (talk) 18:46, 5 October 2008 (UTC)

I do not agree with not linking the dates either and I would like to point out that the consensus was reached only after the 3rd or forth time of being no consensus.--Kumioko (talk) 19:28, 5 October 2008 (UTC)

About using "What links here" to find out what happened on a date: With autoformatting, "What links here" will pick up every reference published or accessed on that date, making it impossible to use autoformatting to find articles related to specific dates. —Remember the dot 19:45, 5 October 2008 (UTC)

That really only applies to dates from the 20th century on, maybe the 19th. There's no reason for this bot to be running around delinking dates from the fourth century, etc. -- Kendrick7 22:02, 5 October 2008 (UTC)

"What links here" on 5 October 427 can only show you links to "5 October" or "427", it can't show you links to "5 October 427" specifically. All the centuries are mixed together, making it very difficult to use linking to find events that happened on a specific date. —Remember the dot 22:17, 5 October 2008 (UTC)

You can get pages mentioning 5 October 427 from the intersection of sets of "what links here". Gimmetrow 03:46, 6 October 2008 (UTC)

From 5 October 419 through 7 November 427 .... — Arthur Rubin (talk) 03:58, 6 October 2008 (UTC)

Your point? You could get every article linking to any date in any range by getting every article lining to every date in the range. Gimmetrow 04:11, 6 October 2008 (UTC)

Arthur's point is that if you take the intersection of what links to October 5 and what links to 427, you'll include an article containing that text: "From 5 October 419 through 7 November 427", although it has absolutely nothing to do with 5 October 427. -- Jao (talk) 18:06, 6 October 2008 (UTC)

The intersection may include other articles, but it will include every article you want. Gimmetrow 18:20, 6 October 2008 (UTC)

Yeah, I know that, Remember. Still: a reader might want to know what else was occurring in 427 so they can get a wider historical context to the article they are reading, and linking it let's them do that in one click. Why is that such a terrible thing? -- Kendrick7 20:04, 6 October 2008 (UTC)

I have no problem with depreciating autoformatting. The question is w/r/t linked dates, particularly years. That seems very much not approved by consensus. Tony's arguments regarding "high value links" are rather silly. People may only click a few links upon visiting a page, but they don't click any of them by accident. If they click through the date, it's because they want context. More often, no one will click the dates, but it's useful information for those improving or examining year pages.

It boils down to reducing Wiki's information and functionality for aesthetics; personally, I'm against that. -LlywelynII (talk) 13:45, 6 October 2008 (UTC)

The old version of this manual used to encourage everyone to wikilink every single date. Your attempt to edit the manual can easily be interpreted as "go ahead and go back to the old policy of editing every single date, if that is your preference". This approach has been clearly rejected and your edit should not be allowed to stand. --Gerry Ashton (talk) 22:21, 5 October 2008 (UTC)

Plus the old policy was to only wikilink all full dates, which was done for the sake of autoformatting. For sole years, or month-year, the policy has always been to wikilink only when called for by WP:CONTEXT; thus, there has never been any consensus to wikilink all dates for the sake of linking. I'm not saying consensus can't change, just pointing out in what direction it would have to change, as many seem to be unaware of that. -- Jao (talk) 22:26, 5 October 2008 (UTC)

There's still no consensus that all year links should be removed, so Lightbot should be decertified as a bot, and those who unlink all dates using AWB or other automated systems, without checking each link for applicability, should be decertified for use of automated tools (after a warning). — Arthur Rubin (talk) 03:40, 6 October 2008 (UTC)

I second that removal of years has not been shown to be approved by concensus and decertifying Lightbot is an excellent idea, although I don't know where to go about saying so. Feel free to link to my support from the appropriate page. -LlywelynII (talk) 13:47, 6 October 2008 (UTC)

Agree that Lightbot should be decertified and anyone de-linking dates en masse should stop. --UC_Bill (talk) 14:55, 6 October 2008 (UTC)

And another agree here. Under the current discussion, Lightbot is well out of order. - fchd (talk) 16:00, 6 October 2008 (UTC)

• First, this argument that year links are necessary for the editors of year pages to orient themselves—to find leads to appropriate information to include on these pages, is utterly bogus. Has anyone heard of the search box? If you need to rely on WP itself rather than outside sources for your stimulus, just type in a year. Second, can someone point to the consensus for linking years in the first place? Tony (talk) 16:10, 6 October 2008 (UTC)

• There is no requirement to find consensus to allow for the linking of years and dates. Consensus would be required to enforce either always linking them (which no one is suggesting) or never linking them (which is what Lightbot is enforcing). There is no demonstrable consensus that these links must be removed, and given the concerns that continue to be raised it's continued use to remove all linked years is disruptive. Shereth 16:15, 6 October 2008 (UTC)
• Granted I may be paranoid that my keyboard will break but I don't like the idea that a subset of articles should be reachable only via the search box. — CharlotteWebb 16:16, 6 October 2008 (UTC)
• I realize that you think that these links should not exist. However, I think it is clear that there is no consensus for these mass edits. Please see Misplaced Pages:Requests_for_arbitration/Episodes_and_characters_2#Fait_accompli for why you should not being making these sorts of edits on a large scale without forming a consensus first. It feels like you're just trying to wear all of the opposition down by refusing to acknolwedge it and simply persist in making the edits until it's the status quo. -Chunky Rice (talk) 16:54, 6 October 2008 (UTC)

• This, I think, is what has me worried about this entire approach, and the Fait accompli from that ARbCom is exactly right. I've been watching the debate, and I completely understand and agree with the point of delinking dates and all that. However, this last step, completely depreciating date linking, was brought to the community (across many boards, appropriately), but only a 7 day period elapsed with maybe.. 20-odd editors responding during that time, and suddenly it was "consensus". I am not saying the consensus isn't there for this change, but clearly there needs to be more discussion of the issue. The matter should have been brought up via an RFC or a watchlist-details notice or some other means to invite a much larger discussion; this might have prompted different solutions (maybe the MediaWiki devs would have been kicked into gear to give us a usable autoformatting solution, but there have been other practical solutions such as templates as well after this change was made that seemed to have support) The end result would have likely been the same, but personally a result I would be more comfortable with it once a much larger discussion was made given the wide impact date linking has on WP. --MASEM 17:13, 6 October 2008 (UTC)

• Yes. Watchlist notification was discussed on 10 September but nothing was done. RFC was suggested on 13 September but it all seemed to get confused (Greg opposed RFC for a reason that seems to be of a personal nature, although he himself always said he wanted a larger input, did I get that right...?) and nothing was done there. No idea why, really. Of course, Tony's arguments have been visible in quite a few places, not only on MOSNUM, but still most people must have missed it (which would have been the case after an RFC or VP announcement as well, I'm sure; watchlist notification would reach more people). -- Jao (talk) 18:33, 6 October 2008 (UTC)

• I can't fault Tony on trying to spam (in a good way) as much as possible to get the word out, but the spamming was never really to a point of requesting input in a typical RFC fashion; I know when it was posted to WP:VG, it was more confusion on the point as opposed to any discussion. --MASEM 18:45, 6 October 2008 (UTC)

Well, Tony, if you only want year articles to be accessible via the search box, and not through links, the natural conclusion is that that should be the default for all links. - fchd (talk) 17:54, 6 October 2008 (UTC)

There is a difference between links that pass WP:CONTEXT and links that don't. We write "A demo version was released via download on May 1, 1999" and "the game received positive reviews from gaming websites" (examples from today's featured article), and nobody complains that the reader who suddenly feels an urge to read more about downloads, reviews or websites has to type those words in the search box. Why would a reader be more likely to wish to visit May 1 or 1999 than any of those three? -- Jao (talk) 18:33, 6 October 2008 (UTC)

For me it comes down to whether the year is "within living memory" or not. 1999 is within living memory, whereas an article dealing with events 70+ years ago is not, because a reader is increasingly unlikely already know the historical context of the article's subject or to have learned it from elders over the course of their lifetimes, and increasingly likely to need to know to really understand the article the further back in time we go. -- Kendrick7

I'd like to point out a fundamental difference between linking years and linking other terms. After a short time, everyone gets to know that we have articles on most years. This is not so with other terms, such as dummy load. So linking some terms serves to alert readers that an article is available on a topic, when it isn't obvious this is the case.

I advocate using infoboxes or templates for significant dates, and I don't mind if the years are linked within those infoboxes or templates. I also don't mind having the first instance of a year linked in an article, if the year is significant. Obviously years that are present in the reference list are seldom significant.

As for the degree of scrutiny needed before using a semiautomatic tool to delink the dates in an article, I believe a person should skim the article and get a sense of the state of the dates in the article. The use of a semiautomatic tool is justified when there are a number of inconsistent date formats in the article, or when nearly all the dates are linked. The use of a semiautomatic tool is not justified when the dates are in a consistent format and only a few dates are linked. --Gerry Ashton (talk) 22:02, 6 October 2008 (UTC)

User:Lightbot paused

I have paused User:Lightbot (at least if it behaves according to the instructions) per the above concerns. I would like to see some sort of consensus here that the task of de-linking dates has any kind of consensus prior to resuming the bot's work. Thanks, Shereth 16:07, 6 October 2008 (UTC)

• I can’t profess to be unbiased on this issue. I think linking to material that is unrelated to an article is unwise. I’ve written an essay on the matter (here) and expanded on that essay here on my talk page.

  Evidence for a consensus is unclear at this point. The above poll and discussion showed opinion was about evenly split (17 to 15) to no longer link the dates of births. I also believe there has been a developing consensus lately that the linking of calendar days (like March 12) is worse than linking years. Linked years is more of a grey area since there are more circumstances (like history-related articles) where the judicious linking of years is thought by many to be appropriate.

  There also seems to be an intertwining of issues. By de-linking calendar days/years, the bot was also removing autoformating. Autoformatting, which produced *prettier* results only for A) registered editors, who B) set their user preferences, was deemed as unwise by a consensus and has been deprecated.

  The complexity now, is that linking of dates is part of autoformatting and this won’t change until the developers disable the autoformatting function of the links. As a necessary consequence of delinking, Lightbot was replacing them with fixed-text dates in a specific format (Euro/International, or US). This aspect alone brings out passions and opinion is all over the map on how to choose date formatting in articles. A guideline that would key the date format to what is most appropriate to the subject matter failed and the current guideline is weighted towards defaulting to what the first major editor used. So formatting of dates after Lightbot has visited is intertwined with the issue of delinking dates.

  I would propose that we all get onto the same page as to whether there is any meaningful difference between linking of dates and autoformatting of dates (for simple years, like 1987, there isn’t), and try to progress forward from there. Trying to arrive at a consensus is made more complex by the fact that many editors arrive here late to the discussion after articles have been affected by Lightbot; we have to start from square-one with these editors. Greg L (talk) 19:41, 6 October 2008 (UTC)

  • We can wikilink dates without invoking autoformatting: ], ] will give March 12, 2008. Mind you, I strongly suggest a template form for this instead of hand-writing it. --MASEM 20:26, 6 October 2008 (UTC)

Greg, you wrote:

• the bot was also removing autoformating and Lightbot was replacing them with fixed-text dates in a specific format

Lets be clear about one thing. Lightbot does not delink autoformatted dates. Many people would be delighted if it did, but it does not. Lightmouse (talk) 21:37, 6 October 2008 (UTC)

• • I have to say, it would be great if you would write up a description of what Lightbot does on it's user page because right now, there's no way to tell, as far as I can see. -Chunky Rice (talk) 21:45, 6 October 2008 (UTC)

• Indeed. Let’s get it from the horse’s mouth. I’ve struck the contested text. What are the true facts here Lightmouse? What is your bot doing that has editors’ nickers so in a bunch? I’ve clicked on some of your activity as assisted by some AWB software and the result was the deletion of double brackets around dates. Of course, I completely agreed with what you were doing there and think it improves Misplaced Pages. And I think you properly read the general consensus when you made your move with AWB. But now I’m confused. Are there two kinds of computer-assisted activities going on here? Note further that by taking away the brackets, the dates get locked into their raw way they were coded. For editors who were looking at the world through their damned date preference setting, many would think AWB was changing the date format. The effect of AWB is confusing to some and this is aggravated by the thoroughly moronic action of autoformatting, which gives only some editors a special, rose-colored view of editorial content that no regular user sees.

  There is no point revisiting the issue of what date format to use in articles; that was thoroughly hashed through, starting here in Archive 110, via two run-off-style polls. It hasn’t even been a month since then, so it is unlikely the mood has changed.

  So task at hand is to push for a clear consensus on the circumstances under which it is appropriate to employ links to calendar days and years. Not too many editors disagree with the premiss that links should be sufficiently topical and germane to any given subject to merit being linked to; the issue is where to draw the line and how to memorialize the nuances in an easy-to-follow, clear guideline. Any bot activity should narrowly limit itself to whatever that guideline calls for. Greg L (talk) 21:56, 6 October 2008 (UTC)

• Whatever comes of this (or any related) discussion, I'd also like to see the question of whether or not bot or script-assisted removal of wikilinks to dates/years is appropriate finally put to rest. Shereth 22:02, 6 October 2008 (UTC)
• (EC)Thanks to an idea in your sig, we can nowrap the "faked" date to prevent it moving about. (see User:Masem/datetest for an example). Again, this needs to be simplified via a template, but its doable. Just that the template needs to know what format to pump out. --MASEM 22:05, 6 October 2008 (UTC)

I am always a little surprised when I come across the assertion that Lightbot delinks autoformatted dates. It delinks any date except autoformattable dates. That is it. A solitary year is not autoformattable. I personally like the phrase 'date fragments' but some people didn't like that. The issue was extensively discussed in the bot approval. The bot user page (User:Lightbot) provides links to its three separate approvals, look at the bullet points in the one called 'Lightbot 3'. I wrote it in bullet point form in an attempt to make it clearer. If you are still uncertain about what a date that isn't autoformattable means, come back to me. Lightmouse (talk) 22:19, 6 October 2008 (UTC)

• Maybe we are talking about two different bots. What about this example, Lightmouse? Let’s agree on the simple facts here. Greg L (talk) 22:24, 6 October 2008 (UTC)

← And we come to the crux of the issue, Masem. There are nice ways to link fixed-format dates that circumvent autoformatting and gives all visitors to Misplaced Pages the same date format. On Sept. 16, we settled on the date format editors should use in articles. So now the issue to settle is the circumstances under which dates should be linked. Here’s my take:

• The issue is not whether or not these lists have any socially redeeming value whatsoever; it is whether or not they are sufficiently topical and germane to any given subject to merit being linked to; that’s all.

  The nearest thing to a completely random list that has been successful is the Guinness Book of World Records. But, given the nature of what’s in that book, and the fact that is is organized into classifications (natural disasters, human feats, etc.), it can actually be read rather linearly with some measure of enjoyment. Misplaced Pages’s random lists of who-knows-what come up quite short of “compelling reading.” I don’t buy into the implicit argument that ‘since nearly everything is in date articles, they are suitable links to put into any article.’ To rebut that attitude, I submit How to Bore People in Five Simple Steps.

  Links to years in truly historical contexts are appropriate: in an article on the Great Depression, judicious use of links like 1929 make sense and do a good job of exploiting the promise of hyperlinking, as first envisioned by Paul Otlet in his 1934 book, Traité de documentation (Treatise on Documentation) as interestingly covered here on YouTube.

  But for general-purpose uses like birth years? I don’t think so; if visitors are reading a Misplaced Pages’s article on, for instance, Frank Gehry, they are most likely there because they are interested in famous architects and beautiful architecture. Accordingly, we add value to the Frank Gehry article and encourage learning and exploration by providing a link to Falling Water, not by linking to 1929 (the year Mr. Gehry was born). But if there was an article on Notable architectural events of 1974 (the year of his first major design), then by all means, let’s provide a year link to that article.

  As for specific calendar days, like like March 12, so few readers would be interested in wading through any of these lists, we would only diminish the value of links and desensitize readers to them were we to link to them.

  I also think Misplaced Pages’s Fairness In Advertising policy ought to be better applied. For specific calendar days (which ought to be quite rare) links would work as follows:

Pearl Harbor was attacked December 7, 1941 (list of random events throughout history on Dec. 7).

There’d be far fewer of date links being clicked on after that. In all seriousness, I suggest that year links be aliased so they better disclose to the reader what they will be taken to. I suggest as follows:

The Great Depression followed “Black Thursday” which occurred on October 24, 1929 (other notable events of 1929).

Greg L (talk) 22:24, 6 October 2008 (UTC)

Hmmm, a date that will live in infamy, but apparently not be linked. Maybe we should just say that Pearl Harbor, by amazing coincidence, occurred on Pearl Harbor Day. I think birthyears should be linked, because the world a person is born into tells you a lot about their life, and the year articles exactly provide that. -- Kendrick7 02:50, 7 October 2008 (UTC)

• I can see what it was approved to do - which appears to be whatever it wants. What I want to know is what it actually does. Please write a short summary. -Chunky Rice (talk) 22:34, 6 October 2008 (UTC)

To answer Greg's question... you are not referring to Lightbot. You are referring to Lightmouse. Lightbot can run when I am asleep, Lightmouse can't. The Lightmouse contributions often involve a script and my fingers pressing 'Save page'. I find it difficult to answer the request by Chunky Rice because it does a lot, there are hundreds of lines of code. Think of the list of all things that might be called a 'date', then think of a list of all things that might be called a 'valid autoformatted date', then subtract the latter list from the former list and you will have a list of all the things it might delink. For example, in its last edit, it removed one link to '1961' and one link to '1968'. You can see from its recent contributions that it is mostly solitary years because that is what most non-autoformattable dates are. Lightmouse (talk) 22:48, 6 October 2008 (UTC)

• Maybe the simplest way to convey what Lightbot does is to provide four links here that illustrate its typical activity with dates. Greg L (talk) 22:51, 6 October 2008 (UTC)

I am confused by that request. There are over 380,000 examples. You can pick any one of them just by going to the contributions. Why are we doing this? Lightmouse (talk) 22:58, 6 October 2008 (UTC)

• The problem I have is that the approval includes "other edits," which aren't specified, to edit dates/numbers etc. as "part of general MOS guidance", which is also very vague. I don't have a good idea what this bot does. Just a sentence saying, "This bot unlinks non-autoformatted dates." would be helpful. Right now, there's no way to tell what it's doing. -Chunky Rice (talk) 23:00, 6 October 2008 (UTC)

Lightbot unlinks non-autoformatted dates. Lightmouse (talk) 23:01, 6 October 2008 (UTC)

And that's the entirety of what it does? -Chunky Rice (talk) 23:02, 6 October 2008 (UTC)

It is also approved to edit units of measure in a variety of forms. Note that approval might not translate into activity. Lightmouse (talk) 23:09, 6 October 2008 (UTC)

If you want to know what it is approved to do, see:

• Misplaced Pages:Bots/Requests for approval/Lightbot 3 and the two other approvals (links on that page)

If you want to know what it actually does, see:

• Special:Contributions/Lightbot

it is currently focussed on delinking solitary years because people believed such links as inferior to autoformatting links. There seems to have been a flip flop in that belief. Lightmouse (talk) 23:15, 6 October 2008 (UTC)

I think there is not so much a flip-flop in belief, as a change in which belief is being discussed. Towards the beginning of the discussion it was brought up that many editors saw that full dates and calendar dates were linked to enable autoformatting, so they just linked every year in incorrect imitation of what they saw. It think that's true, that's what really happened. Now the discussion has shifted to "what about the years that were linked deliberately?" Of course, Lightbot can't tell the difference. Perhaps if Lightbot could search an article for unlinked years, and not operate on any article that contains an unlinked year, that would reduce the problem. After all, if some years are unlinked, that would imply that the ones that are linked were deliberate. --Gerry Ashton (talk) 23:28, 6 October 2008 (UTC)

Lightbot can't do that, it can only work with a few characters and sometimes a whole line. Even if it could, that would mean that the four useless links to '2003' in The Escape Engine would not be unlinked because the year '2002' is not linked. Or the useless links to '2009' in Upcoming Telenovelas could not be unlinked because there is an unlinked year '2008' (that article is definitely overlinked because it also contains linked solitary months). Lightmouse (talk) 23:52, 6 October 2008 (UTC)

Sorry, but that sounds like an even stronger case for Lightbot to be dropped. While I generally support delinking of individual years as I believe the links have limited value (and I've delinked a number manually when making other edits to articles), it is clear from the above debate that a significant proportion of the editors here do find value in them. Also, when and if Lightbot is re-started, and all it is doing is delinking standalone year links, perhaps a more informative edit summary than "(Date links per wp:mosnum/Other)" might be in order? - fchd (talk) 05:27, 7 October 2008 (UTC)

... in case a newcomer's perspective is of any value: as Gerry Ashton wrote: "many editors saw that full dates and calendar dates were linked to enable autoformatting, so they just linked every year in incorrect imitation of what they saw." having gone through that exact phase not very long ago, i really welcomed the new policy deprecating date-linking because of its beautiful clarity. as you can see from my edit history i was unlinking/reformatting dates manually for a while, then tried a script for a day or two; and i'm deeply dismayed to learn that the policy is still so controversial. but reading the arguments being presented here ... it seems people agree that the autoformatting needs to be either abandoned or changed to template form; it seems people agree that not every date should be linked; it seems people agree that some dates (mainly years) do deserve to be linked. the trick is to formulate a rule that's clear (including to newcomers).

it's simplistic but: what about putting links to the date pages that people consider important/valuable in "see also" sections, rather than making them "in-line" links? Greg L's suggestion that such links should be identified as (for example) {(])}} would work very nicely in the "see also" sections, as would "1978 in music"-type links. and it seems like it would be clear enough (even to people who haven't read the policies) that not every date mentioned in an article needs to be listed there. Sssoul (talk) 09:40, 7 October 2008 (UTC)

If anyone hasn't noticed, Lightbot has re-started again. - fchd (talk) 11:03, 7 October 2008 (UTC)

Apologies, that appears to be Lightmouse the user rather than Ligthbot the bot. Either way, the end effect is about the same. - fchd (talk) 11:24, 7 October 2008 (UTC)

Sssoul's idea sounds like an excellent move. There would be nothing more disconcerting to readers than to see some years bright blue and some black. Consistency in the main will be preserved, and the few occasions on which year pages might be deemed vaguely relevant to a topic may be convered in the "See also" section. Tony (talk) 12:01, 7 October 2008 (UTC)

Putting ] in 'See also' rather than in the main body sounds good to me. Lightmouse (talk) 14:35, 7 October 2008 (UTC)

Excellent suggestion, Sssoul! That sounds like a beautifully phrased compromise. I wholeheartedly agree.--Aervanath lives in the Orphanage 16:11, 7 October 2008 (UTC)

Of course the article year X is going to link to notable events in year X, it's WP:COMMON sense. Readers looking for temporal context shouldn't have to scroll all the way down to the see also section; that's a terribly WP:BURO-cratic solution. We wouldn't do something like that for articles providing geographical context (i.e. a link to Azerbaijan), we just use an inline link and everyone is happy. -- Kendrick7 16:18, 7 October 2008 (UTC)

Mostly going to have to agree with the above. The compromise sounds nice at first, but I don't really see it as a solution. "See also" is the appropriate place for related topics that can't be linked in the main body of the text; inline links are always superior, if for no reason other than the fact that readers interested in context should not be expected to scroll to the bottom of the article. If a link to a year (or date) is appropriate to the context of the article, it is appropriate as an inline link. Shereth 16:23, 7 October 2008 (UTC)

In-line links and citations are always better. Perhaps we should reflect on why and when an editor should link to a date, rather than how. A bullet-point list of criteria in the style guide should suffice; and perhaps linking should be the exception rather than norm. Millstream3 (talk) 16:26, 7 October 2008 (UTC)

I still like "within living memory" as a good rule of thumb, which would make most everyone happy, at least from the examples most people are providing against linking, which involve years from 1990 onwards. Links to years even octogenarians can't remember anything about which provide temporal context to the article are OK, links to years less than 70 sols ago are generally to be avoided. If I'm writing an article that involves the year 1058, I insist that this year should be linked, and I'm not going spend the rest of my life reverting LightBot and script-kids every few days. -- Kendrick7 17:25, 7 October 2008 (UTC)

plainly there's a whole range of different views, but a "within living memory rule of thumb" is way too arbitrary to address the main problem i see with the date-linking - which is that unclear/inconsistent policies give too many people the mistaken impression that *all* dates should be linked. and the fact that it can be difficult to decide which geographical place names to link isn't (to me) an argument in favour of leaving excessive masses of dates linked for no reason - which is the current situation.

everyone in this discussion so far seems to agree that currently there *are* too many date links, mainly due to the now-deprecated (?) autoformatting, and to editors who think that since some dates are linked then *every* date should be linked. the bots/scripts were developed to assist in undoing some of that excess. i understand the objection to the bots/scripts - in the course of undoing masses of useless/ill-conceived date links, they've also undone some date links that someone felt were useful. so the point is to find some way to eliminate the excess date links and the confusing principles that mislead people into excessive date linking without doing away with date links that some people consider valuable.

some people who want to keep certain specific date links feel that scrolling down to the "see also" section is too much trouble. but leaving some dates linked creates an ongoing need to undo overzealous date linking - which is *also* too much trouble. so what other compromises do people propose for a clear and consistent policy on date linking? Sssoul (talk) 18:06, 7 October 2008 (UTC)

ps: Kendrick7 wrote: "Readers looking for temporal context shouldn't have to scroll all the way down to the see also section", and Shereth wrote: "readers interested in context should not be expected to scroll to the bottom of the article." i don't think i understand why not - if someone is interested in the temporal context, skipping to the bottom doesn't seem particularly difficult.

but if that's really too much to ask of interested readers, maybe a template could be created to add a box of "links to dates mentioned in the article" to the "contents" box on articles where there are editors who feel strongly about making it ultra-simple for readers to jump to date pages. Sssoul (talk) 20:13, 7 October 2008 (UTC)

I don't think a rule of thumb for articles relating to 70 year old+ events is any more arbitrary than the argument that we can't link to the year 472 because too many articles link to the year 2005. In my opinion, you're alternatives fail WP:CREEP; we can put that in the rules, but no one is ever going to go to this much trouble. -- Kendrick7 20:38, 7 October 2008 (UTC)

um ... i haven't raised any arguments related to the number of articles that link to 2005. the "70+" idea seems to me pretty arbitrary in its choice of "cut-off date"; but the main problem i see with it is that it will *look* arbitrary - for example in biographies of people whose lives/careers "straddle" the cut-off date. policies that look arbitrary won't be very helpful in alleviating confusion over what dates to link.

as for WP:CREEP, i don't think my proposals would require elaborate instructions. "don't link dates in articles; links to important dates can be added to the 'see also' section" seems pretty straightforward. (yes, a template attached to the "contents" box would call for a few more instructions - that's one reason i prefer the "see also" proposal.)

"no one is ever going to go to this much trouble" ... well, everything is "too much trouble" if no one feels strongly enough about it. i thought the whole point was that some editors feel strongly about making it ultra-easy for interested readers to link to some year pages. if that's not the case, let's go back to the "see also" idea.

anyway to reiterate: the proposals so far seem to be:

• link all years prior to 1939 and unlink all other dates - is that right? (i don't know anything about bots/scripts so someone will have to chime in about whether a date-unlinking bot/script could be taught to do that. i feel this policy wouldn't do much to alleviate the confusion about what dates should/shouldn't be linked, but ... the confused will always be with us, i guess.)
• unlink all dates in articles, and put date links someone considers important in a separate section - either the "see also" section or a box that could be appended to the "contents" box on articles where someone wants it. (i hope date-unlinking bots/scripts could be taught to leave sections like that alone. maybe this is "too much trouble", or maybe it sounds promising.)
• unlink all dates. (bots/scripts exist that can assist with this, but some people protest that certain valuable date links are being or may be unlinked.)

any other ideas? Sssoul (talk) 21:38, 7 October 2008 (UTC)

I think the best bet is for editors to link dates when they believe that the date provides valuable context, and not link them when they do not. I don't expect editors to have a problem exercising this type of editorial judgment. Christopher Parham (talk) 21:57, 7 October 2008 (UTC)

... but the current situation is that there are masses of date links that are *not* based on editorial judgement - they're based on the now-defunct autoformatting policy and on misunderstandings of it and/or of other policies. the masses of ill-conceived links need to be eliminated; the question is how to designate date links that someone feels are genuinely valuable for understanding the article so that those don't get eliminated along with the useless/ill-conceived date links. Sssoul (talk) 07:51, 8 October 2008 (UTC)

There are also masses of dates that have been unlinked not based on editorial judgment, by this bot - the best way to ensure that date linking reflects the judgment of editors is to leave the decision for editors to make an a case-by-case basis. Christopher Parham (talk) 02:32, 9 October 2008 (UTC)

What we were to only allow year dates to link to "YYYY in field" pages, each of those having a separate table for other "YYYY in field" pages? That is, say I've got an article on a politician (and only a politician); then date links from that page would link to "1999 in politics" (and possibly "1999 in United States politics" if the field is considered too large). If the topic was a crossover, the editors would have to select the best appropriate links, so a politician that may have been a professional athlete before would have both "in politics" and "in sports" year links. In other words, this is sort of a category structure (which it what sounds like people want but keeping it inline). Now, and I would say this is critical, this works under the assumption that we normal avoid surprise links (eg linking to "YYYY in field" but only displaying "YYYY" with no additional context), but if we made this universal across pages, this would no longer be a surprise.

The unfortunate drawback is that this cannot be bot assisted, at least easily. A bot might be able to determine the page's primary field by looking for the first WikiProject on the talk page, but this is going to fail on crossover articles, and there's potential for hit and miss. Individual editors would be needed to standardize this approach Wiki-wide. --MASEM 22:04, 7 October 2008 (UTC)

well ... does every year that some editor feels is important to link to have associated "YYYY in field" pages? i kinda doubt it. Sssoul (talk) 07:51, 8 October 2008 (UTC)

A technical response to Sssoul's questions about bot capabilities:

• a bot can delink all dates after a 'threshold date' such as 1939
• a bot can delink all dates except those that contain a non-date word such as ]. But it can't distinguish between ] in one section and ] in another because a bot doesn't know about sections.
• a bot can delink all dates (we already knew this)

My other idea: full date linking (autoformatting) is the disease, overlinking of partial dates is merely a nasty symptom that has got out of control and keeps coming back. We could try for consensus for bots to treat the disease rather than the symptom. I am sure many of the pro-delinking people would support that. Lightmouse (talk) 22:08, 7 October 2008 (UTC)

For the record, I didn't suggest linking to years mentioned prior to 70 years ago would be mandatory! While I generally agree with Christopher above, if LightBot could be taught the difference between 1939 and 1939 BCE/1939 BC (well, those articles don't exist yet, but you get the idea), I would have no objection to it making a one time pass to de-link all years and decades after 1939. I would guess that would cover 90% of all year links, given Misplaced Pages's tendency towards WP:RECENTism. -- Kendrick7 22:17, 7 October 2008 (UTC)

Yes, it is easy for a bot to distinguish between solitary years such as ], ] and ]. I notice that there is increasing acceptance that full autoformatted dates should also be delinked. That could be done at the same time. Lightmouse (talk) 22:44, 7 October 2008 (UTC)

From the RfC above, there is no consensus at all to unlink dates of birth and dates of death at the top of bio articles, whether full or not. And in the absence of a clear RfC that can be linked to, I'd suggest there's not much evidence of consensus to delink any other dates either. Jheald (talk) 23:09, 7 October 2008 (UTC)

While I think dates should be unlinked, I have to agree with Jheald: there's no consensus for a mass unlinking of anything quite yet.--Aervanath lives in the Orphanage 04:25, 8 October 2008 (UTC)

I'd agree that dates are overlinked (unless we can get the MediaWiki dev's to incorporate geo-presence for formal date formatting), but I'm most unhappy with the current mass unlinking. I'd suggest it stop for now, except by strictly manual methods.

I'm still intrigued by LightMouse's comment on LightBot's method: "I find it difficult to answer ... because it does a lot, there are hundreds of lines of code". Interesting that, bot approval is just a matter of confusing up the code 'til no-one can understand it? Changes in guidelines are immediately enforced with spaghetti code? Trust me, it really does work, honest. Hmmm. Franamax (talk) 08:40, 8 October 2008 (UTC)

Response to Aervanath and Franamax: so it's a case of WP:IDONTLIKEIT, is it? By that I mean, you just appear to be unsettled by the kind of prompt adaptation of which wikis were built for. The longer the cancer of overlinking and the dysfunctional date autoformatting is left, the harder it is to fix. Every new editor comes to WP and copies the practices they see. It is not practical to make such an important change in slow motion. Were you thinking of a decade-long program? Tony (talk) 10:17, 8 October 2008 (UTC)

Well, no, I try not to frame my arguments in blue. I do generally accept the overlinking rationale (pending geolocation auto-preference, wherein date-links would make perfect sense). What I'm not comfortable with is the pace and scale, in particular when I see bot-op and script-assisted edits. I worry about what gets left in the dust behind the vehicle. In particular, I'm not clear on when exactly date-linking is appropriate. Did we arrive at a consensus somewhere that it shall never ever occur? Franamax (talk) 10:32, 8 October 2008 (UTC)

a few people have expressed this greater confidence in date-unlinking that's done manually - which puzzles me some. as long as links are not designated as "this is a link someone thought about and wants to keep", doing the unlinking manually just means it takes longer than doing it with the help of a well-designed script or bot. i don't see the point of slowing down a process if there's agreement that it needs to be carried out. if someone doesn't support the process then i don't suppose they want it carried out slowly *or* rapidly.

moving well-founded links to the "see also" section and "piping" them when necessary - for example ] - would be a way of designating them as well-founded, considered, intentional, etc. maybe there are other workable ways to designate them, but that's one suggestion on the table at the moment. Sssoul (talk) 12:08, 8 October 2008 (UTC)

Comment. I'm glad Lightbot was paused. While I agree that irrelevant years should be de-linked per WP:CONTEXT, it appears that some people believe this means "articles about years should be orphaned". And I don't think a bot can be able to understand whether a link is relevant or not. I did remove the link in http://en.wikipedia.org/search/?title=Quintic_equation&diff=244401457&oldid=239497393, as the fact that the theorem was published in 1824, rather than in 1624 or in 1924, is totally irrelevant to the point being made (that there is no formula for general quintic equations over the rationals in terms of radicals); on the other hand, linking the year when somebody was born, or a historic event happened, or a book was published, in the article about the person/event/book itself, provides the historical context in which the person lived, etc. -- Army1987 (t — c) 16:59, 10 October 2008 (UTC)

Avoiding linebreaks in dates

If dates aren't linked, allowing a line break between month and day looks bad. But having to add &nbsp;s to every date would be a pain, and hard to read while editing. If autoformatting was able to convert a linked date in accordance with a user's preference, could it be made to produce an unlinked, hard-spaced date? If not, what about creating a set of templates to do that? So {{1 January}} would produce "1&nbsp;January" by default, or "January&nbsp;1" for users with that preference. And the other way around for {{January 1}}.
—WWoods (talk) 19:32, 15 October 2008 (UTC)

Developers do not put improvements into production on a timetable that allows interaction with the editing community. That makes consensus-building with developers just about impossible. Without developer assistance, there is no access to preferences, and even if there were, few readers have preferences set. So WWoods proposal amounts to always using the format "15&nbsp;October 2008" and there is no consensus to always use that format. --Gerry Ashton (talk) 20:29, 15 October 2008 (UTC)

'Avyear' template

We have a template {{avyear}} in the aviation project (as in Template:Avyear) which links to years in aviation, it is used only in the infobox next to an aircraft's first flight date. First flights are listed in the relevant aviation year section, I strongly believe that this linking is totally in context and was very disappointed to see it delinked by Lightbot, there are now 100+ 'year in aviation' articles that are not easily accessible. Notwithstanding previous discussion about 'years in aviation, music', easter eggs etc I think it is (was?) a fantastic feature. I was demotivated after seeing many hours work undone and stopped editing for a week or two (not a protest, just very cheesed off). Sure, don't link to individual days or years if they are not in context but please leave an easy way of linking to very relevant 'year' articles. Nimbus (talk) 23:56, 20 October 2008 (UTC)

It is (or ought to be) a general principle in hypertext that the text of links should make clear where the link takes. If someone sees a printed version of that page, would them be able to figure out what article the 1962 refers to? Remember, not everyone has the ability of hovering on links to know where they go in advance. -- Army1987 (t — c) 09:02, 22 October 2008 (UTC)

Surely none of the links on a printed version of an article will work?! It seems discussion on this date linking subject is not over yet, watching with interest. In the meantime I will neither link or unlink 'aviation years' to stay safe. Nimbus (talk) 15:34, 22 October 2008 (UTC)

In printed encyclopedias, including some predating the WWW, I've seen marks in an article suggesting to read some other article, such as underlining its title or preceding it with a picture of a finger. Now imagine one has a complete printed copy of Misplaced Pages. If someone sees "atomic nuclei", they can realize that the editor was suggesting to read the Atomic nucleus article, despite the irregular plural; writing "atomic nuclei (see Atomic nucleus)" would be overkill. But which would be clearer, "1962" or "1962 (see 1962 in aviation)"? -- Army1987 (t — c) 11:34, 23 October 2008 (UTC)

Intention to implement Sssoul's solution

The discussion above has a bearing on a number of style guide pages; I have no problem that it's being discussed here alone at the moment. Later today, I'll insert links at those pages to this section.

Pursuant to Sssoul's excellent suggestion, I intend to add this suggestion (above) to style guide pages that are relevant (MOSLINK and, until it's merged into MOSLINK, CONTEXT), unless there are good reasons not to. I can't see how anyone could object; there are at least four compelling reasons to support this:

• Inline solitary year links are very unlikely to be clicked on (they're not explicit, and readers soon learn that they lead to unfocused information).
• Related to this, being able to spell out "]" is a huge advantage and is likely to attract many more clicks.
• The assumption that inline is superior to "See also" is very doubtful. The opposite argument could easily be run, that readers are more likely to branch out to linked articles after they've read an article, rather than aimlessly interrupting their reading to go elsewhere at important places in the main text. We should not assume that readers have a marijuana bong next to their computer.
• It is long established that the undisciplined linking of every year is undesirable; linking selected years will encourage editors to start linking all of them, which would be a serious backwards step to the move towards selective linking to build the web more strongly.

In summary, it solves the issue that some editors may wish occasionally to privilege a particular year by linking it (1963 in the JF Kennedy article), and provides explicit gateways into the WikiProject Years articles. Since the use of "concealed" year-in-X links are already deprecated, this is an ideal opportunity to address that issue as well.

Accordingly:

Where there is strong reason to link to a year-article, editors are encouraged to insert a piped link into the "See also" section (]) rather than linking the item in the main text (]). Similarly, a "concealed" year-in-X link (]) should be avoided in the main text in favor of an explicit link in "See also" (]). This recommendation does not apply to articles on years, other chronological items such as decades and centuries, and year-in-X articles.

Tony (talk) 03:40, 8 October 2008 (UTC)

• Wow! Is there now a consensus for this? This looks great. May I suggest that the guideline explicitly state that if a year gets linked in the body text, that it also be piped? And may I also suggest that we standardize on terminology? I would propose that “date” shall refer to either a “calendar day” (May 12) or to a “year.” There has been confusion during our debates because of the dual meaning of ambiguity of the terminology. If someone else has more suitable terminology that is already well embraced elsewhere on Misplaced Pages, I’m all for it. Greg L (talk) 04:06, 8 October 2008 (UTC)

• Unfortunately, I think this is jumping the gun a little. There has NOT been a consensus reached above on Sssoul's solution (no matter how much I like it), and therefore this proposal is premature. We should hold off a bit until the discussions above have reached a consensus.--Aervanath lives in the Orphanage 04:27, 8 October 2008 (UTC)

likewise - i don't feel enough opinions have been expressed yet. it's my understanding that the current MOS policy about the deprecation of linking dates for autoformatting purposes *is* based on a consensus broader than just a few people - i personally wasn't aware of any discussion of that, but since it's been adopted in the MOS i sure hope it doesn't need to be hashed out again. that consensus seems to me to justify unlinking *most* full dates, as well as excessive/ill-conceived linking of partial dates, but there's no consensus yet on what to do with date links that some editor feels are genuinely valuable to understanding an article, and i feel a consensus on that is important before proceeding.

although it seems premature to implement this proposal, i may as well note, for the record, that i don't think "the same applies to concealed links ])" is very clear/communicative. Sssoul (talk) 06:34, 8 October 2008 (UTC)

That's funny, I thought piped links were deprecated in the "see also" section. — CharlotteWebb 10:24, 8 October 2008 (UTC)

(1) This is simply a more formal and quite transparent attempt to generate the consensus that Aervanath feels hasn't yet been generated. Several encouraging remarks were made in support of Sssoul's notion above, and there appeared to be no opposition thus far. I had assumed that people would write "support", "comment" and "oppose", so I'll start the ball rolling by writing "Support" below.

(2) Concealed links have been deprecated for I don't know how long in MOSLINK, and are strongly discouraged by at least one major WikiProject. Unfortunately, they're widespread in lists and in articles on sports, film, and certain other topics. I suspect that readers just ignore them, which is a pity.

(3) Charlotte, thanks for pointing out the possibility that piped links are deprecated in "See also" sections; however, I can't find mention of this just where you'd expect to, at WP:SEEALSO; nor can I readily see the point of such a deprecation. Tony (talk) 10:35, 8 October 2008 (UTC)

Are they not intended to list related articles by name, similar to disambig pages (one blue link per line and all that happy). This seemed sensible enough I took for granted that it was a guideline already. In any case Misplaced Pages:Piped link#Intuitiveness should cover it well enough. As I understand it, other than for dissanbiguation purposes piped links should generally only be used when the constraints of the surrounding prose (that is, what does and doesn't flow well in a sentence) leave no other viable option. This is not an issue in appendices such as the "see also" section or Main article: ], and I as a reader would expect these to show the article's proper title, especially if I had the paperback edition. — CharlotteWebb 20:23, 8 October 2008 (UTC)

• Support—This opens explicit gateways to year pages; it's highly likely that inline links to solitary years are rarely clicked on by readers. It is perfectly consistent with the trend on WP towards more careful, "smart" linking to maximise the utility of the system. It's the type of content that our "See also" sections appear to have been designed for. Tony (talk) 10:35, 8 October 2008 (UTC)

Well geez, if solitary year-links are rarely clicked on, doesn't that invalidate the argument that they commonly lead the reader on a path to nowhere? Presumably, the readers clicking the links will be genuinely interested in the other things that happened in 1929, 1941, 1939, 1905, 1968, 1989, 1918 - I'll stop now. Those years are relevant in-line to the article flow - they set the context for the story. Franamax (talk) 11:03, 8 October 2008 (UTC)

I'm no enemy of year-pages; on the contrary, I'm keen that WikiProject Years be revitalised and that year-pages (and decade pages) be lifted out of their current moribund state. Clearly, the inline carpet bombing of our text with bright-blue years never worked (many readers would have wondered WTF they were). Making links explicit in the "See also" section is a much better way of promoting them as focused secondary articles, and nicely addresses the disadvantages of blue years scattered through the main text, which does not have community support. Tony (talk) 12:09, 8 October 2008 (UTC)

• Support. Lightmouse (talk) 11:16, 8 October 2008 (UTC)

• Support - it is, without question, an improvement on the current situation. Millstream3 (talk) 11:29, 8 October 2008 (UTC)

• Support. I still wouldn't mind date links in infoboxes, but if this proposition can keep them out of article prose, it's for the better. A link to 1963 in the See also section of John F. Kennedy assassination is harmless and might even be useful – there's Category:1963 in the United States and Category:1963 crimes, but categories aren't ordered chronologically and are harder to read. -- Jao (talk) 13:14, 8 October 2008 (UTC)

• Jao, note that the link will be piped and contain at least one non-date word. For example, ] rather than ]. It will be impossible to implement using automation otherwise. Lightmouse (talk) 13:30, 8 October 2008 (UTC)

• Yes, isn't that already part of the proposal? I'm sorry if I was vague, but I meant specifically "a link to the 1963 article", not "a link that looks like 1963". -- Jao (talk) 13:54, 8 October 2008 (UTC)

• I can see this being reasonable when talking about one or two year links, but what if we've got a famous person with events that occur pretty much every year of his or her professional life (say, 40 odd years); Inline, these would not be a problem, but now you've got a spam of them in the seealsos. This is not a easily viable solution to this for multi-year topics. --MASEM 13:17, 8 October 2008 (UTC)

• But very few years are relevant. Elizabeth II of the United Kingdom has done a lot of stuff, but there's certainly no point in linking to 1949 (the year she moved to Malta) or 1991 (the year she addressed the US congress). There has been a very stable consensus not to link these non-WP:CONTEXTual years at all, so these are not what anyone will (or at least, should) be putting in See also. -- Jao (talk) 13:54, 8 October 2008 (UTC)

• They may not be relevant to you but given that we talking about this approach in that someone will be interested in clicking a year link to find other events during that year, there is a likelihood that someone would be interested to see what other events happened in 1949, even if that event is not a significant facet of Queen Elizabeth II's history. Anytime you start talking revelance, it becomes very subjective and leads down the road of edit-warring to no end.
• The way I'm seeing this is that we want to replicate the usability of categories that allow users to jump to other related topics, but not using categories, which.. well, seems to be wasting an existing capability. It might be too grand a scheme now, but I'm thinking that if we plot out a good tree of "Year in XXXX" categories, we can make this all work via templates and categories and be more effective for end users. --MASEM 14:18, 8 October 2008 (UTC)

It's the dilution problem again. The more linked years, whether in the running prose or the "See also" section, the less likely the reader is to bother with any of them. The idea is to be highly selective; that is a much more effective drawcard to encourage reader interest in year articles per se. And remember that year articles, and year-in-X articles, all provide easy passage to their siblings, yes? Tony (talk) 15:03, 8 October 2008 (UTC)

I think before we get decide on this approach (again, I see the validity of it), we need to have a good working understand of what should be highly significant dates that are to be linked in this approach. ("linked" here could mean placed as seealsos or as current inline links). I realize this is not a simple task, and one that is likely better suited by giving a range of example cases which state that this is the case, and similar examples which are not, with cases otherwise not covered to be treated case-by-case. Once we know what the approximate volume of dates will be that we will want to link in this fashion, then a better assessment of which why is better can be made. --MASEM 15:25, 8 October 2008 (UTC)

• Support with a few minor changes. 1) I would change "chronological item" to "year" as the wording seems to address only years and not dates. For example, there is ongoing discussion above about whether birth/death dates should be linked. 2) I would add that it would also be appropriate to use the template:see also to add year links to a particular section. 3) I think year links on date articles (e.g. links on the March 12 article) should be exempt from this policy and link directly to the year inline. Queerudite (talk) 14:07, 8 October 2008 (UTC)

Queerudite, excellent points one and three; is the wording now clear? I'm a bit wobbly about your second point, since if taken to extremes it would lead to clutter. If there's some way of wording it to yield highly judicious usage of this possibility, it might win support here ("Occasionally, if a year page is of close relevance to a section, the template:see also may be used for this purpose.") I'd be surprised to find whole-year articles that were sufficiently relevant to just a section; it's hard enough to find relevant year articles for an entire article. Unsure; what do other people think? I'm tending to think that this is more trouble than it's worth. Tony (talk) 15:03, 8 October 2008 (UTC)

If it is contentious perhaps my second point should be left for a separate discussion. I do think it would helpful, for instance on the biography of a politician, a certain section might have a see also 2001 in politics and in a different section see also 2003 in politics; or an astronaut might have sections on 1969 in spaceflight and later 1973 in spaceflight. It seems the alternative would be to place these in the See also section on the bottom of the article, when it is no longer contextually relevant. Queerudite (talk) 15:17, 12 October 2008 (UTC)

• Oppose per WP:CREEP. You make all the rules you want, you'll never get 10,000 editors to start doing this. -- Kendrick7 14:22, 8 October 2008 (UTC)

Prescription for no policy or style guidance at all on WP, Kendrick? Tony (talk) 15:03, 8 October 2008 (UTC)

You'll never get people to remember to make a special exception for just year inlines as opposed to every other kind of inline. It'll simple lead to silly practices, like having to do ] and "merging" 1058 in non-arrivals of Messiahs into 1058. Via la difference. -- Kendrick7 15:13, 8 October 2008 (UTC)

No, but a bot which goes around removing bare year/date links will make those who want a link to consider what the guidelines are, and to work with it accordingly ;-) Ohconfucius (talk) 09:06, 9 October 2008 (UTC)

• Support due to its clarity (for editors who imitate what they see as well as for those who read the MoS) and the ease of implementing it in tandem with the process of undoing ill-conceived/depracated date links. i feel the wording of it still wants some finetuning, though. Sssoul (talk) 14:49, 8 October 2008 (UTC)

There was an e.c. Please see if it's better; I'd be pleased to hear your further suggestions. Thank you. Tony (talk) 15:03, 8 October 2008 (UTC)

i'm exhausted right now, Tony, but i'll ponder it later - thanks Sssoul (talk) 15:24, 8 October 2008 (UTC)

• Oppose both ''']''' and ''']''' being encouraged anywhere. Strongly oppose "see also" links unless in the same paragraph as the year named. If at all implemented, the link needs to be something like.

  For other notable events of 1929, see 1929.

• rather than hidden links. (Hmmm. I guess that's support with those changes and those suggested by Queerudite, but strong oppose otherwise.)
• Oh, and year links should be encouraged in the lead (birth and death years) and in the {{Birth date and age}} template. — Arthur Rubin (talk) 14:55, 8 October 2008 (UTC)

Nobody is proposing concealing links ''']'''. Quite the opposite. Have you misunderstood the proposal? Lightmouse (talk) 15:01, 8 October 2008 (UTC)

Please re-examine the tweaks to the green proposal. So you mean you don't think readers will be a lot more attracted to clicking on the explicit pipe than a plain old year link? I have to politely disagree. Tony (talk) 15:03, 8 October 2008 (UTC)

As for the paragraph-specificity, that's all good and well, but most years worth linking to would not be restricted to one paragraph. What paragraph of John F. Kennedy assassination would host the {{for|other notable events of 1963|1963}}? Background of the visit? -- Jao (talk) 15:09, 8 October 2008 (UTC)

For John F. Kennedy assassination, the {{for}} or {{seealso}} should be in the lead, or perhaps should be a specific year link inclusion, in that, for an event occuring at a specific date, the year should have an unadorned link, but only once, and only in the first sentence. — Arthur Rubin (talk) 16:26, 8 October 2008 (UTC)

Actually, I concur with the proposed text, noting reservations to the reasoning, provided that

1. The suggested location of the {{seealso}} be left unspecified, or noting that acceptable locations include the top of the article (among other {{for}} tags), the #See also section (which, by the way, requires modifying the guidelines for that section), or the section or paragraph where the year first occurs. Deprecate the "]" in favor of {{for|other notable events in 1929|1929}}, to be placed in the lead, in the "#See also" section, or in a relevant paragraph.
2. There should be occasional exceptions where a bare link is appropriate (examples being the birth year or death year of a person in an article about that person, or the year of an event in an article about the event) but almost never more than two links per article.

— Arthur Rubin (talk) 17:40, 8 October 2008 (UTC)

I believe that these demands for open-ended prominence are unreasonable. I do not recommend them. Tony (talk) 06:44, 9 October 2008 (UTC)

• Conditional support so long as this change is strictly limited to a suggestion as worded above and is not to be interpreted as a requirement. User:Masem's concern above that articles with multiple date/year links would wind up having an unwieldy "see-also" section are valid, and other conditions under which this suggestion is not favorable. I would strongly oppose any implementation of the above as a requirement or the use thereof as license to resume de-linking inline wikilinks, as any useful consensus to do either of these things requires broader attention than merely those who have an interest in watching MOS pages. Shereth 15:06, 8 October 2008 (UTC)

Shereth, adding date links to the "see also" section and simultaneously leaving in-line date links seems mighty redundant. just to make sure i've understood you properly: are you proposing that all current inline date links should be kept, including the masses of ill-conceived/now-depracated ones? (that's what your "i would strongly oppose ... resum de-linking" sounds like - but i hope i'm misinterpreting that.) either way, i agree that some wider attention/participation in this discussion would be excellent. is an RfC a good idea, or some other means of giving the proposal a wider airing? Sssoul (talk) 15:43, 8 October 2008 (UTC)

Not at all. As now, inline year-links are deprecated. The example below, although of a concealed year-in-X link, is an illustration; the principle is the same. Tony (talk) 15:56, 8 October 2008 (UTC)

(reply to Sssoul) - I'm not suggesting that we have both inline and see-also links. What I am suggesting is, as this wording is an encouragement it still leaves it up to the discretion of the editor whether or not to make the transition from inline to see-also, particularly for the cases mentioned above. To an extent calling it a suggestion is redundant - the MOS is a guideline and thus subject to interpretation/exemption. I just want to stress that it should be understood that, as a suggestion, this change remains optional and should not be enforced, especially by bot unless a broader community consensus suggests otherwise. Shereth 16:01, 8 October 2008 (UTC)

Well, no, that's not part of the deal, Shereth. Too many WPians object to inline solitary year linking. If that causes you to oppose, so be it. Tony (talk) 16:18, 8 October 2008 (UTC)

Too many Wikipedians or too many of the ones who watch this page? This whole debate was sparked by Wikipedians who have not run into this (or prior) discussions showing up to complain because Lightbot removed links on an article they maintain. File an RFC to get broader opinion on the issue of inline solitary year linking - otherwise this conversation is going to come up time and time again with editors wanting to see where the consensus for such an action was formed. Shereth 16:31, 8 October 2008 (UTC)

that objection may be just a mechanistic reaction, without due consideration of the revised MOS guideline. These days, I still come across editors who add date links to articles on my watchlist. Ohconfucius (talk) 09:12, 9 October 2008 (UTC)

Exactly -- an RfC, and probably a link in Template:Cent would seem highly appropriate to gain consensus, Tony. I've de-watchlisted this already, but if bots and script kids start mucking around again in historical articles I maintain, I'll be back. -- Kendrick7 16:49, 8 October 2008 (UTC)

I don't know how to do an RfC. Tony (talk) 16:56, 8 October 2008 (UTC)

Actually, I don't see a consensus against bare year links. (I can only see a consensus against autoformatting, which I actually don't agree with, but I can see the point.) They're usually inappropriate, per the overlinking guidelines, but not always. If this discussion were sufficiently published (RfC + {{Cent}}), and reaches consensus, then we can act on it. Not before consensus is reached. — Arthur Rubin (talk) 17:30, 8 October 2008 (UTC)

You're absolutely right on that point. There was a stable consensus to link solitary years per WP:CONTEXT only, but the deprecation of autoformatting or Lightbot's delinkings (or both) seem to have stirred up a lot of new thoughts/emotions on this, and now there are editors pulling both ends on the string, wanting anything from "link all years" to "link no years". (I'd note that the "link all years" crowd is a clear minority, at least if you discount those who would accept some non-wikilinked form of date markup. But the rest of the scale is crowded.) Personally, I mostly agree with Tony and Greg, but I'm not too blind to see that there's no consensus (yet) on the matter. And neither are they: Tony explicitly said that "This is simply a more formal and quite transparent attempt to generate the consensus that Aervanath feels hasn't yet been generated." But yes, it should be done with a larger input. And then, it will look like the birth/death date RfC section above... I'm actually having a hard time seeing how such a consensus can emerge from all this, but the MOSNUM regulars don't seem to despair just yet. -- Jao (talk) 19:38, 8 October 2008 (UTC)

(@Tony) Create a subpage under WP:MOSNUM (you could do a section here, but likely will make this talk page too long); write up the proposal you wish to add, set up a structure for "Support", "Oppose", "Neutral" and "Comments" for people to provide feedback. Then, at the top of the page, add in {{RFCstyle| section=section name !! reason=a short summary of the discussion !! time= ~~~~~ }} so that it gets listed at the RFC list. Announce that page at the various WP:VP and at WP:CENT and anywhere else you think it might help. I would also consider that this is worthy of a watchlist-notice, but I have a feeling convincing those that hate these that this should be added. --MASEM 17:35, 8 October 2008 (UTC)

• Support This is a nice, encyclopedic way of doing it. It’s a bit more of the elegant, print-way of offering up options for further reading and helps remedy the hyperlinked, blue oceans of body text that have plagued Misplaced Pages lately. Greg L (talk) 16:21, 8 October 2008 (UTC)
• Oppose; constrains editors' application of their own best judgment for no reason whatsoever. Linking inline is best for the same reason that every other relevant link in an article is made inline; Tony's logic seems to me to disagree with the general principle espoused elsewhere in the WP:MOS that links should be in the body text, rather than under See Also, where possible. Christopher Parham (talk) 02:19, 9 October 2008 (UTC)

• This is just one of a large number of examples of editors using their discretion (sic). I would say that this is probably caused by conditioning from years of linking for the purposes of Date auto-formatting, and may be difficult to overcome if in-line linking was left to individual editors without an overall policy of full deprecation with some exceptions. The article may not be from the population of articles you habitually edit, but you should be mindful of some of the problems which exist at this point in time. Ohconfucius (talk) 07:58, 9 October 2008 (UTC)

• Indeed, leaving it up to individual editors risks disorder. As Colonies Chris points out, our newer editors learn by imitating what they read, without the experience to analyse and decide whether their formatting—linking or otherwise—is appropriate. Throwing it to the wind is just what we shouldn't be doing. Parham, I forgot to type in the "Oppose" for you to make it easier. People are probably tired of the predictability in your responses to proposals, especially when your supporting reasons are, IMO, usually spurious. Tony (talk) 08:12, 9 October 2008 (UTC)

• Hmmm. He's not the only predicatable one around here. Carcharoth (talk) 04:08, 11 October 2008 (UTC)

• Oppose If a year is linked it should be to a year. If we want to link to an article about a specific aspect we should say so (see 1999 in film) - actually these links are rarely useful as the articles do not discuss context, they are simply lists of people born and dying in that year and projects started/completed or other events. More useful would be stuff like "his new film North by North West was released in 1962 (see also "Development of the thriller in the mid-twentieth century") despite the delays in editing it." Rich Farmbrough, 21:07 10 October 2008 (UTC).

... i'm puzzled what you're opposed to, in that case - the proposal is to make truly useful links to years explicit in the "see also" section, where they'd be listed as ], ] or ] (if anyone really considers those useful), and to eliminate the ones that are not useful or exist only for now-deprecated autoformatting purposes. Sssoul (talk) 21:17, 10 October 2008 (UTC)

Yeah, I'm puzzled too. Rich, shouldn't that be a "Support"? Perhaps you missed something vital in the proposal. Tony (talk) 04:08, 11 October 2008 (UTC)

• Support : Excellent proposal. Headbomb {ταλκ – WP Physics: PotW} 03:04, 11 October 2008 (UTC)

I want to make the point here that the linking of solitary years has been on a significant declining pathway for several years, with reactions to blue years among WPs and readers ranging from quizzical to derisive. The style guides recommend against the practice. I put this proposal as a way to ensure that focused, explicit gateways into year article can be inserted into WP's article without upsetting the applecart. The "See also" section is the ideal location, and I don't understand those here who believe that readers will be more likely to interrupt their reading suddently to go to 1999 than to take up the offer through a prominent, explicit invitation at the bottom (Other major world events in November 1999). I do believe that for those who invest their time and energy in year articles to oppose this is akin to shooting themselves in the foot. Tony (talk) 04:17, 11 October 2008 (UTC)

It might also be worth pointing out that such linking of year-specific topics and years in "See also" is something that people can do anyway. If it catches on (and you know what they say about good practice - if it's truly good, it spreads by imitation), and more people use this style (I have used it in the past and will do so in the future), then this can be recorded in the MoS. I personally do lots of linking that people might consider irrelevant in the main flow of an article, but I put them in the footnotes to the article, as an aside to the main text, so while an article might look overlinked, or irrelevantly linked, once you read the context and the footnote, it usually becomes clear. Carcharoth (talk) 04:52, 11 October 2008 (UTC)

• Question - Would all seasons mentioned in a sports article have to be placed in a See also section, or just the first relevent one? If the former, I don't think this is an improvement. Having 10 season links cluttering a See also section is not somthing I'd want in an article. If only the first relevant one would be placed there, the idea makes much more sense to me. An even better idea: give a See also link to a list of seasons, when avaliable. That way, an interested reader would have access to all of them at once, and it would be easier than locating another season in a template. Giants2008 (17-14) 01:49, 12 October 2008 (UTC)

Giants, this is an excellent suggestion, and shows how we can be a little more creative in doing this. But let's remember that all "year-in-X" articles contain lists at the top—conveniently setting out for the reader all of the available sibling articles, for each year, linked. That is why linking just one, or perhaps two or three, of the most important year-in-Xs for the article at hand rather than 10 or, as in many list articles with tables, 50, has two advantages: (1) you can make a statement to the reader about which year-in-X is best to start with as the gateway, and (2) you don't flood them with choice in the main text or tables, which every marketer knows is a bad thing, whether on the internet or the supermarket shelf. People get bamboozled once choice is more than a certain amount. Tony (talk) 02:26, 12 October 2008 (UTC)

• Oppose. If the proposal were to make Sssoul's solution one way to deal with dates and years which may interest the reader, I would support. But to make it the only way is misguided: one size does not fit all. Septentrionalis PMAnderson 22:03, 12 October 2008 (UTC)

Anderson, at the moment, the linking of solitary year links is strongly discouraged (see MOSLINK/CONTEXT), and the use of "concealed" year-in-X links discouraged except in certain circumstances, such as tables. This proposal does not change those guidelines, but encourages what many people see here as a valuable option. It's not the only option either. Why do you oppose? Tony (talk) 07:18, 13 October 2008 (UTC)

• I would support Where there is good reason to link to a year-article, editors can insert a piped link into the "See also" section (other notable events in 1929) rather than linking the item in the main text (1929); this is a question of convenience. A "concealed" year-in-X link (1998) should be avoided in the main text; it should be clear to the reader where links lead. This recommendation does not apply to articles on years, other chronological items such as decades and centuries, and year-in-X articles.
• I hope the amendments will explain themselves. Septentrionalis PMAnderson 19:39, 13 October 2008 (UTC)

• If my two cents are of interest, I'd support PMAnderson's wording just as I'd support the original proposal, and pretty much couldn't care less which one is chosen. -- Jao (talk) 19:55, 13 October 2008 (UTC)

• On a related topic, I oppose a ban on day links, at least one to be imposed by bot; there are occasions where they justify themselves. The most obvious is the link to 23 April in Saint George: one sample of a wide class. Septentrionalis PMAnderson 19:39, 13 October 2008 (UTC)

I struggled to see the difference in Anderson's version. Ah, "this is a matter of convenience" (I think that begs questions and is an unnecessary second-guessing of what an editor's motive could be), the removal of the noiwikis (why? It just makes it harder to understand), and "strong reason" --> "good reason" (well, hardly much different, but I can cope with that). I still wonder why you opposed, and why what appear to be trivial differences would make you support. I can't agree with your reasoning for linking 23 April in that or any other article. Sure, in one year of the 13th century, it provides a St-George-related factoid of about eight words, which are surely in the article itself (if not, why not?). Why bother the reader with the hundreds of other absolutely irrelevant events on 23 April in other years? Tony (talk) 08:16, 15 October 2008 (UTC)

Indeed. A prominent link to St George's Day would be immensely more helpful than any link to 23 April. The lead now says "His memorial is celebrated on 23 April", and I think that should be "His memorial, St George's Day, is celebrated on 23 April" but not "His memorial is celebrated on 23 April" or "His memorial, St George's Day, is celebrated on 23 April". Of course the date is very relevant to the topic (much more so than your average birth/death date), but the date article contains almost no information relevant to the topic. -- Jao (talk) 14:30, 15 October 2008 (UTC)

• Support I see it as very positive that editors be obliged to think about what important links are and be forced to prioritise them rather than linking in a moronic and mechanistic fashion. The sea of blue years in the body of the text would thus be banished for good. Ohconfucius (talk) 07:04, 13 October 2008 (UTC)

Examples

First example: This is of a concealed year-in-X link (]) at the opening of Jaws: The Revenge. In fact, it broke the guideline at MOSLINK against the linking of adjacent items, and occurred in an already densely linked sentence:

Jaws: The Revenge (a.k.a Jaws 4) is a 1987 horror–thriller film directed by Joseph Sargent. It is the third and final sequel to Steven Spielberg's 1975 Oscar winning classic Jaws.

I've made the year plain black here, which I believe loses no clicks, reduces the detraction from the other links a little, and improves the look:

Jaws: The Revenge (a.k.a Jaws 4) is a 1987 horror–thriller film directed by Joseph Sargent. It is the third and final sequel to Steven Spielberg's 1975 Oscar winning classic Jaws.

Please now inspect the "See also" section that I've added after the main text, in place of what was a dubious inline link. I do believe, on balance, it's far more likely to attract the interest of the reader. Tony (talk) 15:56, 8 October 2008 (UTC)

Second example: Replacing what might otherwise have been a solitary year link with a "See also" explicit link to a major year in the life of Queen_Elisabeth_II.

Let's pretend that solitary year links were not deprecated, and that the year of her coronation (1953) was linked at the start of the second para here. Now it's not linked, and instead an explicit reference (]) has been added to the "See also" section. Nice, huh?

The two advantages are (1) selective focusing of the reader on a single (or even two or three) major years in her life, from which they can further explore sibling articles, (2) explicitness, and (3) prominent location, even though underneath the main text. As well, it neatly sidesteps all of the complaints of WPians who want a more selective approach to linking in the main text. Tony (talk) 16:13, 8 October 2008 (UTC)

• I think this is an outstanding idea. I hope others here click on the "See also" section link to the Jaws: The Revenge article. You don’t clutter up the main body text with more blue and the link isn’t an Easter egg and it’s a perfect way to let readers know these 1987 in film-type articles exist. Readers already tune out these 2000  links (you’re going to have to click the link or click the 2000 link to find out if my link takes you to 2000 (number) or to 2000 in film or 2000-the year.) That parenthetical clause illustrates why readers often don’t click these link: they violate WP: Principle of least astonishment and waste time. What an excellent way to let readers know that there is a 1974 in architecture article (the year of Frank Gehry’ first major design). Tony’s suggestion solves everything and improves learning and discovery. I think this method should be quickly adopted. Greg L (talk) 16:45, 11 October 2008 (UTC)

• Well, on most systems you'd get warning ahead, by means of a tooltip or status bar or something similar. But still, we can't expect readers to have those tools (considering WP:ACCESS) and know that they have them, and we certainly can't expect readers to want to check them each time they see a year link. I agree that this is a great solution. -- Jao (talk) 17:00, 11 October 2008 (UTC)

• Good point. I keep forgetting that trick: dwell on the link for about 1.3 seconds (with Mac OS X anyway) and it shows where you will go. I usually go too fast for that to happen and forget about it. And, yes, the availability of that feature doesn’t detract whatsoever from the virtues of Tony’s proposal. We’d break Misplaced Pages beyond all comprehension if its interface were predicated upon readers dwelling over and interrogating every link to find out what it would take them to in an effort to avoid suprises. Support. Greg L (talk) 17:06, 11 October 2008 (UTC)

I support the Jaws film example. The "other notable events in the coronation year" is naff. Who cares what other notable events took place that year? Anyway, by the time the readers get to the end of that article, they will have forgotten the bit they read about the coronation. I would prefer such links to be from the sections of the article, but that would quickly get silly. For biographies, I still maintain that the only years worth linking are birth and death years, to show the "state of the world" at the time of birth and death. For Queen Elizabeth II, you have: Overview of world events in the year Elizabeth II was born. For, say Queen Victoria, you have: Overview of world events in the year Queen Victoria was born and Overview of world events in the year Queen Victoria died. Ideally, a well-written biography would make the contrast already, but if not, have a look and see the changes. Carcharoth (talk) 19:48, 11 October 2008 (UTC)

Oppose When reading about the film Jaws, I don't want to just see what other films were made in 1975, I want to go to a full review of 1975 to understand the cultural setting it in which it was made and in which audiences flocked to see it. So keep links to year pages . Only link key years, not every mention of a year in an article. By all means put a link to 1975 in film in somewhere as well. Misplaced Pages is a tool to massively cross link cultural events otherwise seen in different boxes. The proposal puts everything back in the "Film " box. Lumos3 (talk) 10:37, 16 October 2008 (UTC)

Temperature units

There has been discussion over on Talk:Mercury (planet) about including Fahrenheit temperatures (along with Kelvin and Celsius) so that the encyclopedia is more readily understandable to readers from the U.S.A. One editor suggested that the appropriate place for this discussion was over here. The opinions presented so far can be viewed on that talk page under the subheading "Editing?" Tuna Night (talk) 17:09, 10 October 2008 (UTC)

• I think a Fahrenheit conversion would be reasonable in Mercury (planet), especially in the lead section. It's the kind of article that school kids read, not just planetary scientists – probably why it gets so much vandalism! In WP:CHEM, we often provide Fahrenheit conversions in articles about common or household chemicals, see acetic acid (featured article) for an example. Physchim62 (talk) 15:15, 10 October 2008 (UTC)
• I don't see any harm in having the conversions available. As Physchim says, the article isn't just read by scientists. --Tango (talk) 16:25, 10 October 2008 (UTC)
• If it's a general article like Mercury, then by all means it should have Fahrenheit in it. —MJCdetroit 16:56, 10 October 2008 (UTC)
• There was a discussion some time back (either earlier this year, or late last year) amongst Astronomy editors which resulted in a decision to to use only the metric values. I'll try to track it down. --Ckatz_spy 19:02, 10 October 2008 (UTC)
  • Really? And this from a discipline which has at least four widely-used non-SI units of distance! ;) Physchim62 (talk) 19:29, 10 October 2008 (UTC)
• I agree with the first three editors here. Provide the conversion. Though the Planet Mercury article is clearly scientific in nature, all the planets are of interest to US students still in primary school. A conversion—particularly in the lead section—will help make it more accessible for a general U.S. readership with minimal impact on the quality of the article for everyone else. Greg L (talk) 04:39, 11 October 2008 (UTC)
• Support; opposing a conversion that communicates and has a trivial cost in complication would be making the encyclopedia worse. What have we said that would discourage such a conversion? And can we get rid of it? Septentrionalis PMAnderson 21:59, 12 October 2008 (UTC)
• Oppose. (1) US grade-school students, nowadays, are just the ones to know Centigrade, because they are taught it systematically. When they get to high school science, they become yet more familiar with the scale. WP would do them a favor by reinforcing what the schools teach them. (2) Planetary surfaces are rarely at room temperature (perhaps Mars and Mercury once in a blue moon); their temperature ranges are typically at extremes where there's either little significant difference between the two scales (which converge at −40) or the reader is likely to think simply "ah, very hot" or "ah, very cold", without a conversion. (3) We already allow editors of science-based articles to choose, if they wish, not to convert units; won't it look odd to except just one? Why, the school-child may ask. Tony (talk) 22:17, 12 October 2008 (UTC)

Replied to below. I should add here that I would not make an exception, but permit editors one more choice: include both as a conversion. Septentrionalis PMAnderson 03:32, 13 October 2008 (UTC)

• Oppose – I'm with Tony. Everyone who has taken a science class in the last few decades in the US has been taught Celsius and Kelvin temperatures. And there's no useful intuition for Fahrenheit numbers of the sort in the Mercury article. Conversion might make sense in the context of weather and climate of places of earth, but nowhere else. There's no reason to dumb down the[REDACTED] to match the average IQ of the red states. Dicklyon (talk) 00:44, 13 October 2008 (UTC)
  • Anyone who has taken a science class since the adoption of the names has been taught about Celsius and Kelvin; it has been part of general education since the Sputnik panic - heck, it's part of the daily weather report. That's half a century now, and most living Americans have been affected by it.
  • This half-century effort has notably failed: Fahrenheit remains American idiom, and the form most readily "understanded of the people". That is therefore why we should be willing to include it where editors think it may help; I am not supporting a mandatory conversion clause here either. Septentrionalis PMAnderson 03:32, 13 October 2008 (UTC)

"Where editors think it will help" is OK. But if editors disagree, that's not to say that those who think it will help automatically win. Dicklyon (talk) 04:57, 13 October 2008 (UTC)

But the post that was brought here is a claim that those that think it will help automatically lose. That is not supported by the text of MOSNUM, and is an ENGVAR violation. Septentrionalis PMAnderson 05:54, 13 October 2008 (UTC)

• Oppose per Dicklyon. On a case-by-case basis it can maybe be argued that providing Fahrenheit temperatures will be useful to some readers. A look at temperature reveals however that "he entire scientific world (the U.S. included) measures temperature using the Celsius scale and thermodynamic temperature using the kelvin scale...". Later we see "... the United States is the last major country in which the degree Fahrenheit temperature scale is used by most lay people, industry, popular meteorology, and government. Other engineering fields in the U.S. also rely upon the Rankine scale (a shifted Fahrenheit scale) when working in thermodynamic-related disciplines such as combustion." If we allow Fahrenheit in the MoS, we are opening the door for Fahrenheit conversions in every single article which mentions temperatures, and where next? Rankine units? Réaumur scale? Rømer scale even? I'm for sticking with Celsius, or kelvin where appropriate. --John (talk) 05:14, 13 October 2008 (UTC)
  • the United States is the last major country in which the degree Fahrenheit temperature scale is used by most lay people, industry, popular meteorology, and government. The key point here is that it is used by most lay people, industry, popular meteorology, and government in a major English-speaking country. In short, this change is required by WP:ENGVAR, which is long-established and widespread consensus. Septentrionalis PMAnderson 05:51, 13 October 2008 (UTC)
    • That's a nice link, with ENGVAR, which I almost but don't quite buy. The difference is that on spelling conventions, even educated people use the specialized spelling which the US has adopted (as have I as I live in the US), but as the first quote makes clear, in the US, educated scientific use favors the SI units. As an encyclopedia we aim to inform, but we also aim to model educated usage. On this issue, as a science educator in the US, I strongly oppose this dumbing down for our supposed American audience which can't understand Celsius. I think we need to give Americans more credit for intelligence and resourcefulness than this proposal does, and I don't want to see Fahrenheit conversions going into every single article, which is what would happen if this proposal became part of the guideline. --John (talk) 06:09, 13 October 2008 (UTC)
      • The link to ENGVAR is, now that I think of it, actually to the little-quoted statement of principle: We are not here to declare one variety of English more correct than another; in this case, the use of Celsius for all purposes characteristic of some, but not all, Commonwealth countries.
      • I do not believe that conveying information to one's students is dumbing them down; I trust our science educator does not actually teach as he recommends. It is the mark of a good teacher, one secure about his knowledge of the subject, that he avoids uplifting his students with his command of technicalities. For more, see Richard Feynman, passim. Septentrionalis PMAnderson 18:14, 13 October 2008 (UTC)
        • I'm sorry, but in spite of reading this comment several times and being familiar with Feynman's work, I have no idea what you are trying to say here. --John (talk) 19:24, 14 October 2008 (UTC)
• Support. Even if the US is just one country, it happens to contain maybe one third of the world's English-speaking population (depending on how you count), and probably an even higher fraction of the Misplaced Pages readership. Making articles less accessible to so many people due to SI self-righteousness helps no one. I don't buy the slippery slope arguments; this only occurs in "popular science" articles and the Reamur example is a caricature. --Itub (talk) 07:40, 13 October 2008 (UTC)

This is all a storm in a teacup. Editors currently decide themselves (by consensus on a science-based article talk page, if it's under contention), whether to convert to imperial units. Why are we talking about this? Tony (talk) 09:13, 13 October 2008 (UTC)

I agree. I think there is a tendency to overlegislate in this page when local consensus works perfectly fine. --Itub (talk) 09:21, 13 October 2008 (UTC)

Disagree. I think that a general rule would benefit everyone by avoiding consensus discussions on each article. Looks like so much a waste of time to me. Let's discuss this once for all. --Cyclopia (talk) 10:09, 13 October 2008 (UTC)

I'd agree with Tony and Itub, except that the editors on the article talk page seemed to think that the simply because it was a "scientific" article meant that there should be no conversion. That's not the rule that we use on WP:CHEM, and I don't think it was what was intended when the guideline was written. There's really no need to have Fahrenheit conversion on Lanthanum(III) oxide, but it helps accessibility on acetic acid, IMHO. Physchim62 (talk) 11:56, 13 October 2008 (UTC)

I see no sign that that argument was the intent of the rule, and it's not what it says. Physchim62's examples are both to the point, and we could do worse than to add them; although since the only temperature mentioned in acetic acid is its flash point, the melting point of gallium may be a better example in that direction. Septentrionalis PMAnderson 18:14, 13 October 2008 (UTC)

We probably could do worse than to add them, indeed. However, as the arguments against adding them outweigh those for, I do not think we will. --John (talk) 19:24, 14 October 2008 (UTC)

Erm, hold on, am I reading you both correctly there? The only temperature quoted in acetic acid is its flash point? The argument against adding outweigh those for? Physchim62 (talk) 19:55, 14 October 2008 (UTC)

The first statement is clearly untrue. The second I believe to be true; the disruption caused by attempting to add Fahrenheit conversions to thousands of articles will outweigh any benefit gained. Instead I think we should seek local consensus to add the conversions on a case-by-case basis where circumstances demand it, rather than adding to MoS. --John (talk) 20:50, 14 October 2008 (UTC)

My error. That's what I thought the search results found; I'm not sure how I got that impression. Do we want to add conversions for, for example, the normalizations to 25 °C? to one of them? Septentrionalis PMAnderson 17:29, 15 October 2008 (UTC)

• No one is suggesting that thousands of articles be altered; merely that any local group of editors be free to add Fahrenheit when they judge it likely to be helpful. Fortunately, the existing guidance does permit this, although clarification would prevent our having this discussion every time some well-meaning editor decides that only Celsius and Kelvin are pure enough to include in scientific articles. Septentrionalis PMAnderson 17:29, 15 October 2008 (UTC)

• If that were merely the case there would be no need to have a discussion here about it. I recognize the right for local consensuses to decide minor things on a local basis; this MoS never was a binding, legal document. I would definitely be against amending the MoS towards any sort of recommendation to use Fahrenheit temperatures more widely than they already are in science articles though. The initial request wasn't very clear I suppose; I assumed from the fact they posted at a MoS talk page they wanted to amend the MoS. --John (talk) 05:03, 16 October 2008 (UTC)

←

• Too many spurious arguments are being employed here to argue against a Fahrenheit conversion for the planet Mercury—as if the opponents fear the general discipline of science will be set on a track back to the age of human sacrifices and blood letting if we darken the doorstep of an article by using an occasional non-SI unit of measure in (*gasp*) a science article. The Fahrenheit conversion doesn’t “dumb down” the article and it doesn’t detract from the article to any appreciable degree.

  I live in the U.S. and am mature enough to know full well that pretty much everyone in America has an oven in their kitchen that is calibrated in Fahrenheit; this is how Americans intuit the relative magnitude of “really hot.” Showing a Fahrenheit conversion of 801 °F is extraordinarily helpful to intuitively communicate to an American reader just how hot Mercury is—much hotter than the highest temperature on their oven. Most American’s can’t relate 427 °C to any familiar experience. It doesn’t matter that Celsius is taught in American schools because the scale isn’t used in daily life.

  We’ve got to stop thinking that we somehow serve a meritorious and greater good by using Misplaced Pages as a vehicle to promote the adoption and understanding of modern systems of measurement via the omission of familiar conversions. If an article is on a scientific subject that is of broad interest to a generally non-scientific audience, there is no compelling reason to not provide a conversion to help American readers to understand the magnitude of an important measure. We aren’t here to help serve up autumn offerings to the SI gods; we’re here to communicate and educate via interesting text that invites exploration and learning—our American readership included. Greg L (talk) 01:39, 18 October 2008 (UTC)

  P.S. I didn’t precede my comment with a “support” because I think it is wholly inappropriate to even pretend that this common sense, fundamental point of doing proper technical writing for a world-wide readership can be subject to a referendum vote of a small handful of editors who frequent Talk:MOSNUM. And “pardon me all over the place” for honestly expressing my observation that I detect a bit of an arrogant attitude, from some editors here, that if America is too retarded to adopt the SI into their daily life, then they loose the right to intuitively and easily learn. That’s sort of an attitude of “we’re all too stuck up here to provide a simple temperature conversion for Americans.” And lest any proponent of the SI here think I’m soft on the adoption of the SI, I’m not. I imported a Celsius thermostat from Canada for my house when I was in my 30s. I’ve simply wised up enough since then to realize that just because I like the SI system, doesn’t mean we have to be an ass about it. Greg L (talk) 01:54, 18 October 2008 (UTC)

  • Without wanting to be an ass about it, this is not Talk:Mercury (planet) but Misplaced Pages talk:Manual of Style (dates and numbers). I don't have a problem with adding a conversion to the Mercury (planet) page, but I do have a problem with altering the MoS guidance to open the way to adding conversions to an obsolete unit to potentially thousands of articles. I do hope you will be able to grasp the difference. --John (talk) 17:31, 19 October 2008 (UTC)

• John: It doesn’t matter if Fahrenheit is an “obsolete” temperature scale. It only matters that it is the scale that 300 million Americans are familiar with. Proper, modern science is now standardized on the SI and that is as it should be. A properly written MOSNUM guideline shouldn’t “open the door” to conversions like milliliters to fluid ounces in an article on titrating. I would propose wording along the lines of:

If an article is on a scientific subject that is of broad interest to a non-technical audience (e.g. the surface temperature of the planet Mercury), editors should provide conversions to non-SI units where doing so will help American readers to understand the magnitude of important measures.

I don’t see this as opening the door to anything but understanding. Greg L (talk) 02:00, 20 October 2008 (UTC)

I respect your intentions of course, but I still don't agree. I guess we have different views on the exact location of the fine line between aiding understanding and a patronizing dumbing down of our project. --John (talk) 02:10, 20 October 2008 (UTC)

• That’s OK. I knew going into this that it would be an uphill battle trying to encroach barbarian units into science. America still hasn’t adopted the SI into daily life. (Sigh.) Oh well, at least I’ve had grand leadership over the last eight years by one of the great thinkers of our time—George Bush—as a consolation prize. Take that England and your “Winston Churchill.” (*whimper*) Greg L (talk) 02:50, 20 October 2008 (UTC)

• You're a good sport and a good guy and I just know your intentions are good on this, I just fear it would destabilize things more than it would add light. If individual articles (or indeed projects) want to convert I wouldn't quarrel with that, so long as they don't go too far. As for Bush, I voluntarily moved to the US while he was leader, so I must like him, right? Hey, he taught the world how to pronounce noo-ku-lar correctly, what's not to like? Churchill and Maggie Thatcher was the original quote I think, and it's an excellent allusion to make in the context of this discussion. Churchill is actually a hero of mine, although I am Scottish and therefore supposed to hate English people; but then his mum was American. Complicated world, isn't it? :) --John (talk) 04:50, 20 October 2008 (UTC)

• There's already provisions under this version of the MoS to have non-SI units in articles. See the sections on familiarity and the section on conversions. If there's consensus on Talk:Mercury (planet), then that's all there needs to be. And if there isn't (for lack of discussion), then WP:BOLD. Headbomb {ταλκ – WP Physics: PotW} 13:29, 20 October 2008 (UTC)

'Under construction' templates

People put 'under construction' on a page to say that they are actively working on articles. But then the templates seem to hang around on the article for hours, days, and weeks. It is sometimes difficult to get them removed because people say 'I haven't finished yet, I just have to get round to it'. Sometimes people even revert removal of the template.

A reasonable proposal (I think) is to merge the templates into one. The one template would have a fixed expiry time visible in read mode. If it expires, the editor would be able to refresh it but would have to take positive action to refresh it. The current fire-and-forget design would be gone. The expiry time would be consistent with real tappity-tap editing where fingers are pounding the keyboard e.g. 15 minutes to an hour.

For more details, debate, and voting, see: Misplaced Pages:Templates for deletion/Log/2008 October 13. Regards Lightmouse (talk) 16:41, 13 October 2008 (UTC)

• Sell-by date: Another thing we need to do is clarify—if it hasn’t already been—in the guidelines that when {disputed} tags and similar tags are added to articles, they should be removed if the issue hasn’t been actively worked after a reasonable period of time. I’ve seen stale, old {disputed} tags on articles and when I looked at the talk page, the issue had been discussed for a few days six months prior and hadn’t been worked since.

  Perhaps these tags could expire and vaporize on their own after a week. Whoever placed the tag in the first place would have to be active enough on the debate to notice and put it back in. Maybe a bot can troll the tag’s “what links here” and delete week-old ones. This would obviously get rid of tags on abandoned issues. And in some cases where there are vitriolic disputes, such a bot can help it to die a natural death while giving parties a way to save face. Greg L (talk) 17:50, 13 October 2008 (UTC)

Quite. Bots are indeed involved in purging stale 'under-construction' tags but without 'sell-by' date (nice metaphor), the bot programmer has to invent their own generous period of grace or risk of getting into a dispute with the person that wants the tag there. If you run through 1000 articles removing tags like that, you are bound to come into dispute with somebody. Believe me, I know. I am not sure if the tags can disappear without bot involvement but some bots run on the server and can look like there is no bot. The technical means to remove tags are easy. All we need to do is focus on how long one editor can reasonably freeze out others and can we all accept the end of unlimited sell-by periods. As you suggest, if we are successful with 'active editing' tags, we can extend it to other types of tags. Your comments on Misplaced Pages:Templates for deletion/Log/2008 October 13 will be welcome. Lightmouse (talk) 18:06, 13 October 2008 (UTC)

• {{in-use}} is a freeze-out. {{Underconstruction}} is not; indeed, it invites other editors to pitch in. It's a warning that there may be major changes. We should probably encourage posting plans of action on the talk page. Septentrionalis PMAnderson 18:17, 13 October 2008 (UTC)

Sounds good. Could the sell-by date be set by the editor when they add the template? It could be subst'ed , for example {{inuse|2 hours}} would solve the problem of allowing the editor an arbitrary fixed period. It would give greater freedom and would force the editor to estimate his editing time. Would a bot/troll then be able to calculate the time and remove it? –MDCollins (talk) 22:49, 13 October 2008 (UTC)

Amd writing the template so that there was a default time when none was specified (3 hours, say, for {{in-use}}) would encourage editors to set them. Septentrionalis PMAnderson 23:05, 13 October 2008 (UTC)

One minor point - I've occasionally prodded or speedied an article, only to get a complaint from the editor shortly afterwards (by which I mean less than 24 hours) that either they're a) still working on it, or b) asking how they can improve the article to pass WP standards. I've genrally given a few hints and/or pointed them in the direction of WP policies and guidelines, and I've also occasionally marked the article as 'under-construction' so they get a breathing space to read the guidelines and improve the article. Of course, you could say that the article should be substantially complete before it's committed to WP, but from the number of articles I've seen that were built up from repeated edits from the same editor over a short period of time, I'd say this method of editing was common. Anyway, I've generally assumed that the 'under construction' tag would give something like 4 - 5 days 'grace period' - which I would have though seems reasonable - especially for those who've never contributed before and may not know exactly what's expected. In brief, I'm all in favour of some 'automatic expiry time' on these tags - but I'd hope that any default time on the 'under construction' tag would be in terms of days, rather than hours.

Re monitoring of the tags by whoever placed them - this would be (yet another) admin overhead if it had to be done manually, but wouldn't an 'automatic removal of a tag because it had expired' show up on the watchlist (if that page was being watched) ? CultureDrone (talk) 07:24, 14 October 2008 (UTC)

Therein lies the difference between underconstruction and inuse - the latter implies that somebody is currently very active and wishes to prevent edit conflicts, the former implies that anyone can edit, and is primarily a warning against deletion. Each template can then have a different default expiry period.–MDCollins (talk) 00:13, 15 October 2008 (UTC)

• I agree. All tags should have an automatic expiration date. The time period can vary, but there is no point requiring Wikipedians to go to an article’s talk page and hunt down what some {dispute} tag was about in order to discover the issue had been worked for only three or four days and then had been abandoned for six months. There are simply too many of these stale tags junking up Misplaced Pages and it is too tedious of a task for humans to delete them all. That’s why we all often rubber-neck as we “drive by” these tags rather than do anything about them. If some editor wants to place {disputed} tags on an article and enjoy the satisfaction that it might still be there a year later, I think that’s probably too bad; our {disputed} tags should be reserved only for active disputes. If a bot deletes a tag and there is a live human who still cares, he or she can put it back in; it’s just that simple. Further, having an automatic expiration can serve as a pressure relief valve for many disputes. Greg L (talk) 22:42, 16 October 2008 (UTC)

• I've always wondered why there's a bot going around dating certain tags, but failed to see why this needed to be done. Perhaps we need to define specifically which tags date and can be automatically removed, so that the bot can operate optimally. Those other tags which do not age need not to be dated. Ohconfucius (talk) 03:13, 17 October 2008 (UTC)

• • I disagree with All tags should have an automatic expiration date. That is an extraordinarily strong position, and I continue to dispute it, as above. But if you really want such guidance, this is not the place to achieve it; I would suggest WP:VPP. Septentrionalis PMAnderson 02:22, 17 October 2008 (UTC)
  • For example, {{cn}} tags should not expire. I came across a claim that women voted in Liege, in what is now Belgium, in 1792. It is unsourced, and it seems unlikely - but not impossible. The tag should stay until a source is found, or an editor has sufficient confidence in its falsity to remove the claim; having the tag expire is worse than either. Septentrionalis PMAnderson 02:31, 17 October 2008 (UTC)

• No, those {{cn}} tags should not be automatically retired. They validly signal that someone challenges the assertion and seeks proof. What that needs is for some editor to come along and source it, or delete the entire unsourced phrase or section say after a month has elapsed since tagging. I don't see another way. Ohconfucius (talk) 03:03, 17 October 2008 (UTC)

• Agreed; my all-encompassing “all tags should automatically expire” is exceedingly broad. I can however, think of at least two tags that should have “sell-by” expiration dates: {disputed} tags and {under construction} tags. I’m sure there are more. If organic neurons haven’t cleaned up many of these languishing tags after a certain period of time, it’s time to give silicon transistors a crack at the task. Greg L (talk)
  • While I hate to continue to disagree with someone being so agreeable, I would again say it depends. Some {{disputed}} tags are effectively {{cn}}: "say what? Show me." A bot-assisted effort to settle these would be a good thing, and would remove many that are redundant; but some require an intelligence to dispose of. On {{under construction}} I agree, although a sell-by date from the last major edit might be preferable. Septentrionalis PMAnderson 04:35, 17 October 2008 (UTC)
• Yes, {{cn}} should expire: if a statement stays in an article for a year or so without anyone being able to find a decent reference for it, it should be deleted. (It can always be readded if a proper source is found.) -- Army1987 (t — c) 12:56, 18 October 2008 (UTC)

• I think we are talking about 'automatic' (ie bot-assisted) obsolescence. Such obsolence is by definition a manual process as no bot will ever be capable of cleanly removing such unsourced material or assertions. If the tags are automatically retired in such cases, no-one will know that there is/was a contention and that the request for citations has not been fulfilled. Perhaps what is needed is alert pages which age {{cn}} tags (such as Category:citation-needed tags aged more than 3 months, ], ], so that editors may tackle this work systematically if so inclined to do. Ohconfucius (talk) 13:26, 18 October 2008 (UTC)

• I agree with Army1987, for the same reason. A six-month old {{cn}} tag is like that 1991 calendar in an old farmer’s barn: everyone who goes in there notices the worthless thing, but it has almost become part of the barn and no one bothers to take the time to pull it down. Not all tags should expire at the same age; a {pardon our dust ‘cause we’re under construction} tag should expire faster than many other tags. But no tag should stay until the heat death of the Universe. If a {cn} tag still hasn’t been addressed after nine months, it’s still likely going to be the same old wart on the article after a full year.

  And it’s not like an automatic expiration douches a great deal of intellectual effort here (“My manuscript!”). It’s awfully easy to ralph these {cn} onto a page IMO. I’ve seen them attached at the end of values like, electron mass. I thought to myself “oh for God’s sake,” went to the NIST, confirmed the value on Misplaced Pages was correct, and then added a reference that wasn’t too much longer than the {cn} tag it replaced. And it took about two minutes to do it. I suspect there are editors who do nothing more than peruse Misplaced Pages with a high brow and pouted lip and do nothing more than “that ought to be cited… that one too… prove it… that one too… that one too.”

  It doesn’t matter what technology (bot?) we employ to get these to automatically vaporize, all we need to do is decide on an appropriate “sell-by” date for each tag.

  I’m very pleased that we have some proactive editors here that have a bot time-stamping current tags. This is the necessary bit of infrastructure going into the decisions regarding tag life. Greg L (talk) 16:51, 18 October 2008 (UTC)

Fractions

I thought there used to be a statement to not use unicode fractions for the same reasoning against the uses of unicode superscripts. --—— Gadget850 (Ed)  - 10:38, 15 October 2008 (UTC)

Maybe, but it must have been deleted quite a long time ago. It was me who added the present section on Fractions, at a time when the subject was not addressed at all. Originally I mentioned Unicode fractions as an alternative, but people didn't like that, so I removed the reference to them. If you want to (re?)add explicit guidance against using Unicode fractions, I'm not aware of anyone who would object.--Kotniski (talk) 11:35, 15 October 2008 (UTC)

• Before about six months ago, Misplaced Pages’s page rendering engine added lots of extra leading (line spacing) whenever someone used superscripted or subscripted characters. This made paragraphs really ugly; what with the leading jumping around all over the place. Further, this greatly expanded leading made it difficult to see where one paragraph ended and a new one began if there were successive lines in a paragraph containing super/subscripted characters. The Unicode fractions (½, ¼) were a quick work-around to this problem.

  Since then, the developers have fixed that problem and only one extra point of leading is added for super/subscripting. Since the need for this workaround no longer exists, I believe the Unicode fractions should no longer be used because their very small characters are hard to read on some OS/browser/hardware combinations; particularly if they appear in the small text-style used in many of our References & Notes sections (⅜, ⅝). Our {{frac}} template, e.g. {{frac|3|8}} → 3⁄8 and {{frac|4}} → 1⁄4, is a fine tool. Greg L (talk) 17:07, 15 October 2008 (UTC)

  • But, as often, it's a trade-off. Some editors will dislike the large amounts of white space introduced by {{frac}}, which this very section displays. Present the cases on both sides, and let editors decide; we cannot envisage all their circumstances. (White space is particularly bad on tables, and so forth...) Septentrionalis PMAnderson 17:34, 15 October 2008 (UTC)
    • Any accessibility implications? Which will screen readers do better with?LeadSongDog (talk) 17:41, 15 October 2008 (UTC)

• That’s an interesting question. With Mac OS X (10.5.5), the built-in speech services pronounces 3⁄8 as two digits: “three eight”. It is wholly silent with the Unicode ⅜. It is also silent for ½ and ⅓. I have no idea how other, fine operating systems perform. You really have to use a lot of interpretation with OS X’s built-in reader; standard techno lingo is handled rather well but it starts getting stilted with advanced technical matters. For instance, it doesn’t know that “9.81 N” should be pronounced “nine point eight one newtons.” Note still, that the Mac is better than other free alternatives in its ability to parse English meaning—it’s almost clairvoyant. For instance, it correctly pronounces “I project that you will be done with that project on Feb. 13th” in that it properly pronounces the two different meanings of “project” (v. tr. and n.), and it says “February”, not “Feb”. And on routine technical matters, OS X is pretty good; for instance, “1000 kg” is pronounced “one thousand kilograms.” So I rather expected it would pronounce the Unicode fractions even better: as a true fraction (“three eighths”).

  I would agree with PMAnderson above, that there are places (tabular charts for instance) where the Unicode fractions might be better suited, and the best guideline is to provide usage guidance for editors and let them make the choice. Greg L (talk) 22:30, 15 October 2008 (UTC)

• The {{frac}} template looks horrible to me (the denominator is on a completely different line than the rest of the fraction). What's wrong with using "1/2" or "5/8" like we have for the past 141 years? If you really need something fancy, you can always resort to math mark-up. Personally I think using {{frac}} and unicode fractions are clumsy solutions to a non-existent problem. Kaldari (talk) 23:05, 15 October 2008 (UTC)

I don't understand, the denominator looks fine to me:

• One half, 1/2, 1⁄2, ½
• One quarter, 1/4, 1⁄4, ¼

--—— Gadget850 (Ed)  - 23:20, 15 October 2008 (UTC)

• Yes, the denominator produced by {frac} looks fine with me too. These sort of issues can often be traced to a combination of OS/browser/hardware setup. I’m on Safari, which—like the latest version of Firefox—anti-aliases its fonts. So I personally don’t have any difficulty reading even the micro-size Unicode fractions.

  But I do appreciate fine typography and among the “easy” ways to produce nice-looking fractions, the results of {frac} come closest to what you’d get in professionally typeset materials. Ease-of-use and attractive output probably underlies why {frac} has proved so popular. Even an IBM Selectric typewriter from the mid-70s could produce fractions that looked midway between what {frac} produces and the Unicode characters. While a fraction like 3/4 could have indeed been produced on a 1940s Underwood typewriter in 7th grade (or “141 years ago” as Kaldari wrote), the typography world has fortunately moved on.

  Were it I who had designed the {frac} template, I think I would have tried one more notch smaller of text, such as ⁄₃, v.s. 2⁄3 with {frac} and ⅔ via Unicode. I don’t know what it looks like when rendered on other people’s computers, but my hand-tuned version looks just fine on Safari on a Mac. But, it’s a pain to code:
  <font size="-1"><sup>2</sup>⁄<sub>3</sub></font size>
  Greg L (talk) 02:55, 16 October 2008 (UTC)

• I also like the {{fraction}} template. If the unicode fractions are deprecated, they should also be removed from the symbols editing palette too. Their existence there encourages their use. Ohconfucius (talk) 04:19, 16 October 2008 (UTC)
• For me (I'm on WinXP) it looks fine in Firefox, a little less fine in IE7 and pretty bad in Opera 9.10 (haven't tried later versions). I have no idea why, or whether we should consider that Opera's problem. There are no legibility issues though, which is the most important thing. -- Jao (talk) 12:11, 16 October 2008 (UTC)

As for me, I like 1/4 more than 1⁄4 for fractions in running text, but maybe that's just me.-- Army1987 (t — c) 13:28, 17 October 2008 (UTC)

1⁄4 is obnoxious in some contexts. For example, consider the expression √1⁄4. Students write this in a way that makes it unclear whether the radical covers the whole fraction or only the numerator, and then you need to tell them it's illegible on that account and deduct points for it and they act surprised that anyone cares about it, so you have to be really emphatic in forbidding such things. So we shouldn't encourage it. Michael Hardy (talk) 17:41, 17 October 2008 (UTC)

The fraction template looks horrible to me. I'm on a Mac running Firefox 3 (it looks fine in Safari though). Instead of using a template that looks great for 90% of users and terrible for 10% of users, why don't we stick with "2/3" which looks great for 100% of users and causes no problems? Kaldari (talk) 18:51, 17 October 2008 (UTC)

• I’m sure there is a wide variety of opinion on this subject. I think all that is being established here is that there is no fits-all solution for fractions. Tabular tables, for instance, might benefit from using the compact Unicode fractions like ¾. As for radics in math, do you think √9/17 is in any fashion clearer than √9⁄17? And isn’t all that ambiguity solved by being unambiguous, e.g. √(9⁄17) or √9⁄17 ? Further, once one is getting into math and radics, shouldn’t one be using Misplaced Pages’s math markup or TeX anyway? The expression ${\displaystyle {\sqrt {9 \over 17}}}$ is absolutely unambiguous. Regular text is unsuitable, really, for math.

  I think we’re getting off onto a tangent. This discussion started with whether the Unicode symbols, like ⅛, should be encouraged by MOSNUM for use in regular in-line text when {{frac}} (for producing fractions like 1⁄8 ) is available. Then the discussion expanded to “typewriter” fractions like 1/2. Whereas typewriter fractions are an alternative, well done typography doesn’t do it that way… and hasn’t since the 1880s, when, even lead Linotype had special characters for fractions. Now that we have computerized, proportionally spaced typefaces and browses that anti-alias fonts, I think we can can afford to do better than recreate what a simple typewriter could do back when Abraham Lincoln wasn’t yet old enough to drink. Greg L (talk) 19:38, 17 October 2008 (UTC)

• Well if it's a choice solely between {{frac}} and unicode fractions, I support unicode fractions. Since opinion seems to be divided, however, why don't we list both as options in the MoS? Kaldari (talk) 19:45, 17 October 2008 (UTC)
• I have created the template {{radic}} for stuff like √9/17, though I do prefer TeX for anything much more complicated than that. And I've seen fractions written as 9/17 in many places, most of them much more recent than the 1880s, including for example the IUPAP Red Book; I really can see nothing wrong with them, especially when used in-line within a sentence. -- Army1987 (t — c) 20:16, 17 October 2008 (UTC)
• Btw, what about 9 ⁄ 17 (9 ⁄ 17)? -- Army1987 (t — c) 20:18, 17 October 2008 (UTC) (Never mind, very ugly. -- Army1987 (t — c) 20:23, 17 October 2008 (UTC))

• That is what PMAnderson suggested in his 17:34, 15 October 2008 post, above: just present the pros and cons to editors and let them do as they see fit. I can certainly see that {frac}-created fractions look stupid on Firefox 3.0.1. I simply use Safari on a Mac and pretty much everything looks good on it (on a modern LCD monitor); I can clearly read the Unicode fractions. The trouble is, some OS/browser/Preferences settings/hardware combos make the Unicode fractions truly unreadable because they are too small. This is a more serious problem than merely having ungainly leading or inconsistent vertical offsets. It’s terribly hard for editors to really believe that utter junk they might be looking at actually and truly looks fine for many other users. In sorting through the best compromise setting on {{val}} for delimiting numbers (such as 6.022365875×10), we had to exchange screen shots between various editors so we can see what they were seeing and understand why they wanted something tightened up or expanded. We even had an editor looking at the results on an iPhone. There is a new benchmark for browsers to follow and to get a perfect score, they have to *exactly* match—pixel for pixel—a certain standard. As I recall, the newest version of Safari is compliant. I think we can assume that technology will catch up with our aspirations. In the mean time, we can muddle through with alternative tools and differing opinions as to what is best. All we can do on MOSNUM is provide information. Greg L (talk) 20:28, 17 October 2008 (UTC)

• I'm thinking that {{frac}} may be best for general text and unicode may be best for tight areas such as tables. --—— Gadget850 (Ed)  - 20:55, 17 October 2008 (UTC)

• Seems that way to me too. Greg L (talk) 00:53, 18 October 2008 (UTC)

• I'm thinking that "horizontal" fractions (1/4) is best for general text and tight areas such as tables, and "vertical" fractions ${\displaystyle {\frac {1}{4}}}$ is (usually) best for formulas on their own line. I can't see any place in which "diagonal" fractions (1⁄4 or ¼) look best.{{frac}} looks best for cases such as "Aged 13+3⁄4, and Unicode fractions in places where {{frac}} can't be used, such as the title The Secret Diary of Adrian Mole, Aged 13¾. -- Army1987 (t — c) 10:25, 18 October 2008 (UTC)

• These differences of opinion are sometimes purely a matter of personal preference. Often though, there are surprisingly great differences in visual appearance due to an interaction of OS, browser, preferences settings, and hardware. Some browsers, notably Firefox, are currently mucking up the placement of subscripts. We know what OS/browser Jao is using. So too for me and Kaldari (who, like me, finds that diagonals look fine on a Mac using Safari but not Firefox). What are you running? Greg L (talk) 16:21, 18 October 2008 (UTC)

← I didn't say that they aren't displayed correctly, only that I don't like them. I use Epiphany on Ubuntu, and both the numerator and the denominator are legible in 1⁄4 and ¼. (But I used Internet Explorer on Windows XP recently, and it crops the bottom part of denominators of 1⁄3 in some places (e.g. tables), and the denominator of ⅓ almost looks like a 2 at some font sizes. But if M$ stuff is broken, that's their fault, not ours, right?) OTOH, 1/3 looks just fine to me, and it's impossible for even the most brain-dead browser to get it wrong, so why not use it? -- Army1987 (t — c) 19:09, 18 October 2008 (UTC)

• I don’t know what would be the wisest thing to do here. If Microshaft is producing software that cuts off the bottom of some fractions, like 1⁄3, then they actually are not being displayed correctly. Unfortunately, many readers use Internet Explorer. It would be most unfortunate if we were to optimize Misplaced Pages for lowest-common-denominator technology. I remember there was an editor who was shrinking pictures really small because he had a 640 × 480 monitor. The default standard for Web pages is to now optimize for 1024 pixels across (below that, scrolling is required). And it’s been that way for a long time. The same issue applies to page sizes (not the size the edit preview says, but the actual HTML download, including pictures): many of our articles require impractically long waits on dial-up service.

  Now that there is a new Web browser gold standard that developers are trying to certify their products as being compliant with (Safari is one of them as I recall), it should be soon enough that browser technology can keep up with the modest typography demands that some of our Misplaced Pages templates throw at them. Rather than put an explicit statement in MOSNUM that would place a premium on using barbarian-style typewriter-style fractions like 1/2 in order to make things look better on sub-standard browsers, I think the better solution at the moment would be to remain silent until the technology catches up; I don’t think the wait will be too long. Greg L (talk) 22:00, 18 October 2008 (UTC)

  Which pixel-by-pixel standards are you talking about? Is it the Acid tests? Those tests may check for pixel-perfect positioning of page elements such as boxes, but they say nothing about the exact rendering of fonts. I think it would be unrealistic to ask for a standard that required everyone in the world to use the exact same font and have it rendered the same pixel by pixel. --Itub (talk) 07:14, 19 October 2008 (UTC)

  I see "barbarian-style typewriter-style fractions" practically anywhere, even in professionally typeset books. Just to pick an example, on The Feynman Lectures on Physics you see ${\displaystyle \displaystyle m={\frac {m_{0}}{\sqrt {1-v^{2}/c^{2}}}}}$, with both v and c on the same baseline as the 1. (Very slightly below, but that's a common thing to do with glyphs with curved bottoms, so the symbols are at the same relative heights as if 1, v, and c appeared in any one line of a sentence.) It appears to be much more common than diagonal-style fractons. If TeX allows you to do ${\displaystyle \displaystyle n/m}$ and ${\displaystyle \displaystyle {\frac {n}{m}}}$ directly but requires a kludge to do ${\displaystyle ^{n}/_{m}}$, do you think there's a reason for it? -- Army1987 (t — c) 08:48, 19 October 2008 (UTC)

Can I just throw into the discussion the question of mixed numbers? While I might be able to accept 3/4 for 3⁄4, I would find it harder to accept 1 3/4 or 1+3/4 for 1+3⁄4. Can those who don't like {frac} say what they would do with these?--Kotniski (talk) 09:19, 19 October 2008 (UTC)

7/4. -- Army1987 (t — c) 09:43, 19 October 2008 (UTC)

Oh all right, 13+3⁄4 then (as in The Secret Diary of Adrian Mole, Aged 13¾ - though I don't think we have any alternative to Unicode in article titles).--Kotniski (talk) 09:54, 19 October 2008 (UTC)

Ok, ok, so there is a case in which {{frac}} is useful. But extending this to cases like "Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. The speed of light is 299,792,458 m⁄s. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat." (I have seen something similar to that – including the link – in a real article) is not a very good idea. -- Army1987 (t — c) 10:13, 19 October 2008 (UTC)

I haven't read all of this yet, but I want to make three quick points. (1) There are extended ASCII characters for ½, ¼, and ¾. (2) On the edit screen, where you have "insert". "Wiki markup", and "symbols", there are symbols for several common fractions that is pretty convenient to use. Right now they insert Unicode. Those need to still work to give what ever is decided. (3) I strongly dislike "2/3". Bubba73 (talk), 14:36, 19 October 2008 (UTC)

• Army1987: I believe the discussion is whether MOSNUM should remain silent on the use of in-line body text fractions like ¾, ⁄₄, 3⁄4, or 3/4. A mathematical expression like ${\displaystyle \displaystyle m={\frac {m_{0}}{\sqrt {1-v^{2}/c^{2}}}}}$, with both v and c on the same baseline

  …has nothing to do with it. For over a hundred years, fine typography—even in newspapers at the turn of the century—have had single-symbol, over/under fractions. I see no compelling reason to deprecate their use because you think “3/4” is beautiful. Clearly, other editors prefer over/unders. Why are we even debating this? Greg L (talk) 15:29, 19 October 2008 (UTC)

  I agree that the MOS should stay silent – istruction creep is evil. But I don't understand why that expression "has nothing to do with it". You say that fractions on one line haven't been used "since the 1880s", but that book was published in the 1960s; and that was only the first example I thought of; there are many others, even more recent than that. But I'm just not bored enough to go around making a list of such examples. I think that most books I've seen recently use such fractions (except books which use few or no fractions at all, such as some novels.) -- Army1987 (t — c) 15:47, 19 October 2008 (UTC)

  But one of these examples springs to my mind immediately. The SI brochure says: (1.1 Quantities and units) " particle may be expressed in the form v = 25 m/s = 90 km/h, where metre per second "; the m, the s, the km, and the h are on the same line as all other characters; a margin note on 5.1 Unit symbols says "m/s or ${\displaystyle {\frac {m}{s}}}$ or m s, for metre per second"; the form m⁄s is not even mentioned. 5.3.6 Multiplying or dividing quantity symbols, the values of quantities, or numbers says: "When multiplying or dividing quantity symbols any of the following methods may be used: ab, a b, a⋅ b, a × b, a/b, ${\displaystyle {\frac {a}{b}}}$, a b." And even there, the a and the b in a/b are on the same line as any other character (except the −1 exponent and the ${\displaystyle {\frac {a}{b}}}$), and a⁄b isn't mentioned. The edition I am quoting was published in 2006, a hell of a long time later than Abraham Lincoln's legal drinking age. So, honestly, I can't understand where you get the impression that n/m fractions are so old-fashioned. -- Army1987 (t — c) 16:03, 19 October 2008 (UTC)

• You raise a number of good points and it got me thinking about how I manage these issues. Indeed, I always use expressions like 5.2 m/s and not 5.2 m⁄s. I do the following for stand-alone fractions, compound fractions, and complex fractions:

• ...there are 1000⁄123.562134 (≈8.093 moles)

• The test voltage must be no less than 1 7⁄8 the operating...

Greg L (talk) 18:51, 19 October 2008 (UTC)

Yes, maybe we disagree on fewer things than we thought, because 1000/123.562134 would look not very clear to me. (As for the voltage, I'd probably say "no less than 187.5%", but that's another matter.) Furthermore, we both agree that there's no point in adding an item on the MoS about this, so... -- Army1987 (t — c) 12:04, 21 October 2008 (UTC)

Historical years (again)

What's the deal people? I thought we'd been thru this two weeks ago -- bots and script kiddies (yes, you know who you are) have no cause to be auto-delinking historical years from articles. As noted above, 1066, a year largely considered a turning point in English civilisation had less that a few hundred "what links here" results. I understand the issue with recent years per WP:RECENT, but as a general rule, articles relating to olden days should be allowed to have years link so as to provide an easy historical context for the casual reader. Again, what's the problem? Where's the RfC myself and others demanded be called before implementing this? -- Kendrick7 05:45, 21 October 2008 (UTC)

Please would you kindly explain the distinction between a 'year' and a 'historic year'?? I could be wrong in interpreting it as a year not in living memory when something important happened. Er, It strikes me that the above concept may be extremely subjective. You mentioned 1066. Presumably you would also include 1776, 1789, 1780, 1937, 1914, 1918, 1939, 1945, 1949. How would you define 'important', and where would you advise the community to draw the line, if only to avoid the plague of blue numbered years from 0 to year infinity? Ohconfucius (talk) 06:13, 21 October 2008 (UTC)

Kendrick7, could you let us know what specific edits/pages you're talking about, so the rest of us can follow the discussion? thanks. Sssoul (talk) 07:42, 21 October 2008 (UTC)

I'm understanding "historic" to mean something like "memorable" in the Sellar and Yeatman sense. I don't think it's a tenable concept in a worldwide project like ours. If you think it's significant that event X happened in the year of event Y, then say so explicitly and link to Y ("She was born in 1066, the year of the Norman Conquest.") If you just link to the year, readers a) likely won't follow the link; and b) even if they do, likely won't know which of the hodge-podge of facts they find there is supposed to be relevant.--Kotniski (talk) 10:25, 21 October 2008 (UTC)

Lovely, thanks for that link. Excellent book, by all accounts! Ohconfucius (talk) 10:31, 21 October 2008 (UTC)

Date links per wp:mosnum?

Moved from Lightmouse talk page: begin
Why are you still delinking historical years? You know darned well there was no consensus at WP:MOSNUM supporting this. You edit comments are disingenuous. -- Kendrick7 19:24, 20 October 2008 (UTC)

Please don't be so aggressive. If you would like to debate wp:mosnum issues, I suggest the talk page of wp:mosnum is the best place. Lightmouse (talk) 20:52, 20 October 2008 (UTC)

Please reread the comment. Also consider: by making the edits you are being aggressive. I have looked hard at several of your year delinks, and agreed with them, but you seem to be getting a little overboard, or perhaps are not checking your robot closely enough. Perhaps I am reading Kendrick7's comment incorrectly, but I think the point is that your remarks on your edits need be be "more"... at least in some cases. I would agree with that. But perhaps I am incorrect. sinneed (talk) 02:18, 21 October 2008 (UTC)

If I wanted to be aggressive, I would have called for your AWB (sp?) scripting privileges to have been revoked. I am mad, but not yet completely pissed, as I'm not sure of your degree of caution with your ongoing efforts here. I had thought we had a sort of understanding that you were to gather wider consensus before auto-delinking years from articles, and I am, as such, suspicious. WP:CONSENSUS does not form simply because people on one side tire of re-arguing their point. -- Kendrick7 06:10, 21 October 2008 (UTC)

Which "historical year" is at issue, Kendrick? I'm interested, because I don't think I've ever seen one that provides useful information that will increase the reader's understanding of the topic at hand, as required by the MoS. Tony (talk) 03:51, 21 October 2008 (UTC)

Er, and what pray, is the distinction between a 'year' and a 'historic year', and how is anyone likely to tell the difference here on WP?? Ohconfucius (talk) 06:08, 21 October 2008 (UTC)

(sorry for stomping your sig) I had drawn the line at years within living memory. The point is, everyone understands culture in, say, the 1960s was different, generally, from culture in the 1980s, or even the 1930s, and there's a context which we all understand to events within each milieu, even if we've only seen late night ads for Time-Life CD collections. Anyone who thinks there's not just as much of a difference between the 430s, 460s, and 480s is ignorant. I understand that too many articles link to recent years, but that's no reason to delink the historical context from years whose links measure in the dozens and not the hundred thousands. -- Kendrick7 06:21, 21 October 2008 (UTC)

Are you kidding me, Tony? Maybe you are such an avid student of human history that you have at your immediate disposal the context for any given year, but I assure you for our readers that is not the case. What is your hang up about not providing temporal context? The are four dimensions, get used to it. I don't really feel like re-arguing this just for arguments sake, but OK, we can redux this on Talk:MOSNUM. -- Kendrick7 06:10, 21 October 2008 (UTC)

One more vote to stop this work till the consensus. I'm looking on the page, where the birthdate was delinked. My personal beleif, is that the page did lose some important information.--Alogrin (talk) 07:20, 21 October 2008 (UTC)

There's no vote here; only discussion. Kendrick, I'm still eagerly waiting examples of how a "historical" year page can "increase the reader's understanding" of the topic at hand. Alogrin, are you referring to the removal of date autoformatting, per MOSNUM's deprecation of it? DA was never intended to provide links to dm/md pages or to year pages. But just supposing it was, I'd like examples of how August 8 or 1977 provides focused and relevant information warranting a link from a particular article. This is an old argument that has been resolved yet rehashed many times; I'm still willing to go through it again. Tony (talk) 10:35, 21 October 2008 (UTC)

Moved from Lightmouse talk page: end

If your point is that most year articles are listcruft, the solution is to fix them, as was done to 1345, not to orphan them. (If someone believes that the current listcruft is of any use, they can simply copy it to List of 1913 events, for example.) -- Army1987 (t — c) 12:45, 21 October 2008 (UTC)

No one is talking of orphaning year pages, although frankly they're widely recognised as being substandard collections of ragbag facts. 1345 proves the point, although it's far from perfect and presents another challenge: it has sucked in much of the overview material that would be more appropriate to the whole decade; pages such as 1345 can't just be conjured up for all of the years in the vicinity. My suggestion at the WikiProject Years to conflate year pages into decade pages before a certain century was met with utter silence, as though not one person cared. Apart from this issue, which will be a gargantuan task to fix (there are thousands of pages, and the WikiProject is apparently moribund), there is the popular proposal here that chronological pages be linked in the "See also" section, where readers are more likely to be attracted to them. And let's not forget that they all link to each other like a huge tree, so one gateway in an article, or several gateways, will lead to all sibling pages. Linking years systemically throughout articles is not now an option, and linking just a few selected years is an ineffective way of attracting readers and will open a slippery slope to mass linkings. WP has moved on. Tony (talk) 13:00, 21 October 2008 (UTC)

"Linking just a few selected years" is the right thing to do if they are "selected" in the right way. See my 10 October comment at the end of #User:Lightbot_paused. -- Army1987 (t — c) 08:55, 22 October 2008 (UTC)

Linking years systematically through articles is still an option, and apparently a very popular one at that. Just because you (or I incidentally) don't find these links to be of great value, doesn't mean that loads of other people don't either. Deprecating date auto-formatting was a huge mistake in my opinion, the functionality was very helpful to a lot of users and the amount of edit wars it prevented must have been immense. Let's not make any more decisions that will inhibit the value of Misplaced Pages articles to users. - fchd (talk) 05:32, 22 October 2008 (UTC)

I'm sure it was nice for a lot of editors, but most readers are neither editors, nor registered users. Maybe the number of edit wars prevented was immense (since people without accounts don't seem to be interested in edit wars about date format; they'd rather edit war about how to spell metre; don't ask me why, I don't know). For whatever reason, present-day editors no longer seem interested in edit-waring over date format. --Gerry Ashton (talk) 05:41, 22 October 2008 (UTC)

Yup, concur with Gerry. It's an unexpected and very pleasing evolution that there's systemic acceptance of binary systems in both spelling and date formats. I think the last we saw of mutual irritations underlay some of the interactions in the debate on MOSNUM's rules for choosing date formats for articles that are unrelated to anglophone countries. Even that has settled. Tony (talk) 05:46, 22 October 2008 (UTC)

Bot proposal: remove autoformatted dates and leave solitary years alone

There is a bot proposal to remove autoformatted dates and leave solitary years alone. Please see Misplaced Pages:Bots/Requests for approval/Cleanbot. Regards Lightmouse (talk) 13:17, 21 October 2008 (UTC)

Unwikilinking full dates?

I have to say, I'm seriously thoroughly ticked off over this whole recent move to unwikilink full dates. Not so much over the fact that it was done, but moreso over the fact that I can't see that it was debated or discussed very well, or openly. Sure, I see several discussions at various pages, but I still can't see any single debate of "yes, do it" versus "no, don't". I mostly see several isolated discussions by very few editors that somehow someone seems to think represents the consensus of everyone. Seriously, what happened is not "consensus" -- it looks more to me like the cabal dictating its way on everyone.

What I really dislike about not having full dates wikilinked isn't so much the autoformatting issue, although I think we've opened up a HUGE can of worms on that one since, without autoformatting, there's inevitably going to be more edit wars between the British/European date format versus the American one. But what I really don't like about non-linked full dates is the fact that, in text, the date is not highlighted in any way, and it doesn't stand out like it did with the link there. I think the wikilinked full dates added some emphasis on the date in the article and, with particular attention to articles about historical topics, this was especially helpful. Taking these wikilinked dates away is doing a serious disservice to our readers.

I'll probably just get overruled by the cabal for this, if not outright banned for being an ass ;-), but I would strongly urge the community to reverse this decision. Dr. Cash (talk) 22:23, 21 October 2008 (UTC)

I didn't know about this change either, but I firmly agree with it. It always annoyed me to see those blue dates for very little reason that I could discern. I assume that casual visitors to WP were equally annoyed — or at least confused. I say, let it ride for a while and see if the sky falls on us. Sincerely, your friend, GeorgeLouis (talk) 22:31, 21 October 2008 (UTC)

The discussions about this issue were ongoing for YEARS. If it was decided by the Cabal, they certainly took their time. Kaldari (talk) 22:47, 21 October 2008 (UTC)

• Derek: Yes, what Kaldari said. With what must be tens of thousands, if not hundreds of thousands of Wikipedians (registered and not), there is no way that I know of to communicate to them all and keep them apprised of what is being contemplated here. For what it is worth, there are many editors who frequent this venue and many more still who occasionally come here to chip in with their views on maters. As a result, there has been—and continues to be—a wide variety of views expressed here. Note too, that this move was very thoroughly discussed (albeit in your absence). And whereas this issue certainly hadn’t been discussed out on the curb in front of your house, I can assure you that it was discussed “openly”: here on Talk:MOSNUM. I can also assure you that there have been a number of editors here acting as an ambassador of sorts who have been advocating your position.

  I believe too, that the recent move is very wise. Even though these blue links for days of the year and years are available to link to, doesn’t mean it is a good idea to actually link to them and turn yet another bit of body text to blue; far too few readers are interested in actually reading more than four or five line items in these random lists of pure trivia the first time they click on such links. From thereafter, they rarely click on them again. Links within Misplaced Pages articles should always be topical and germane. Properly chosen links anticipate what the readership of any given article would likely be interested in further reading. If you disagree on this fundamental point (whether these trivia articles are of sufficient interest to our readership to merit linking to them), I encourage you to accept the Sewer cover barnstar challenge. If you can actually earn such a barnstar, then I think you will have a stronger case for advocating the linking of these dates. Greg L (talk) 22:53, 21 October 2008 (UTC)

• As the change, if fully implemented, would effect the majority of articles on Misplaced Pages (although many would be improved), it should have been discussed on the Village Pump (policy) before implementation was done. It wasn't. It's now (in the past month or two, long after the guidelines change) being discussed in relevant WikiProjects, namely Misplaced Pages:WikiProject Years and Misplaced Pages:WikiProject Days of the year. — Arthur Rubin (talk) 01:37, 22 October 2008 (UTC)

• Some aspects of this change were discussed at Misplaced Pages:Village pump (technical)/Archive 46#Autoformatting standard violation . --Gerry Ashton (talk) 01:45, 22 October 2008 (UTC)

• True, but that's the technical aspects. The consensus was that autoformatting is deprecated, not date linking, although there is now probably a consensus that dates should almost never be linked (although that wasn't pointed to by the VP), but there's no similiar consensus for years. — Arthur Rubin (talk) 01:55, 22 October 2008 (UTC)

• The policy has been for years that full dates are linked always, only because of autoformatting. Without autoformatting, the policy on full date links is the same as the policy on other date links and linking in general, that they are linked when especially relevant to the context of the article. —Centrx→talk • 03:58, 22 October 2008 (UTC)

Derek Cashman, I've decided not to ban you "for being an ass". :-) Instead, I simply want to point out the dilution principle (one of four or five significant disadvantages that were never properly considered when we were fooled into adopting this autoformatting thing back in 2003). Every additional bright-bluing of text comes at a cost, which is that it vies for the reader's attention with links in the vicinity. That cost needs to be balanced against the benefit of the blue. There's no benefit at all to readers, since they're not logged in. Tony (talk) 02:06, 22 October 2008 (UTC)

Arthur, the disparagement of chronological links has been at CONTEXT for some time; it certainly pre-dates the current flurry of queries an the popular recommendation to highlight selected links in the "See also" section of an article. Tony (talk) 02:06, 22 October 2008 (UTC)

There's a difference between "don't link unless there's a reason" (presently here, and at WP:CONTEXT), "don't link unless there's a specific reason" (formerly here, added without consensus), and "don't link unless there's an overwhelming reason" (what you seem to be saying), and "don't link unless the article falls into a specific type in which date links are appropriate"" (what Lightbot is enforcing). — Arthur Rubin (talk) 02:14, 22 October 2008 (UTC)

• Are you asserting that "reason" by itself, unqualified, doesn't mean "useful" or "arguable" or "supportable" reason? The onus is still on an editor who wants to link a solitary year to demonstrate why it is useful in the context. This is hard to do, as suggested above when my repeated requests of certain complainants to provide examples of the solitary year-links or anniversary links at issue are met with silence. Methinks this is just a case of WP:IDONTLIKEIT, without having though through and properly weighed up the evidence. Tony (talk) 03:01, 22 October 2008 (UTC)

• As Tony suggested some number of threads above, an outstanding method (IMO) to provide historical trivia to readers is to provide a well-aliased link in the See also section. If there is an article on the War of 1812, then a link that is aliased to read “Notable events of 1812”, is a wonderful, least-intrusive way to let readers know of the availability of these lists without cluttering up the body text with more blue. Implicit here, is that these sort of aliased links in the See also section would be best limited to intrinsically historical articles. There’s no point providing a link in the See also section of Angela Lansbury so trivia on her birthdate (October 16) can be researched. If you actually read the Oct. 16  article, you can wade through *fascinating* bits like this…

Really, if we’re going to be providing links in an article on a famous female actresses like Angela Lansbury, then keep the links germane. A rather germane link would be “Robo-biscuit actresses of the 21st century.” Greg L (talk) 03:58, 22 October 2008 (UTC)

• The interesting new thing about this section is that Derek wants to have dates stand out in the articles. I haven't seen that argument before; of course, for a person who wants this, it is a compelling one. But then, someone might want to have all personal names bolded, all negative numbers in red, etc., and I simply don't think these things are appropriate for article prose. At any rate, our links should be used for linking. Using them for something else (autoformatting) was a bad idea from the start and now that we can avoid it, let's. -- Jao (talk) 05:28, 22 October 2008 (UTC)

• thanks Jao - i agree that links should be used for linking, not for "highlighting" purposes. Sssoul (talk) 06:47, 22 October 2008 (UTC)

• To Tony above; for the years of birth and death in articles about a person, we've established a supportable reason. You seem to be saying it's not a reason, rather than it's not a good reason. — Arthur Rubin (talk) 14:28, 22 October 2008 (UTC)

Arthur: I know that you're keen to see biographical articles start with a splash of bright blue, but nowhere do I see evidence of this "supportable reason" you talk of. Are you sure this is not just wishful thinking? Tony (talk) 14:51, 22 October 2008 (UTC)

• We’re covering old ground and discussing unwise new concepts, like using color for emphasis. It’s not hard (unless you add the factor of “change on Misplaced Pages”, which is rarely easy). In‑text blue links should be strictly limited to links that are particularly relevant to the article and will be of likely interest to that readership. This isn’t rocket science here. Links to random trivia are just that: links to lists of random trivia. Honestly, I wouldn’t even link dates in our Trivia article. Too few readers ever reads more than five or six entries in these articles after their first encounter with them. As for using text color for emphasis, that’s easy. Unwise, IMO, but easy. Greg L (talk) 16:19, 22 October 2008 (UTC)

The heart of the matter: no strong demonstration of consensus

First off, I completely agree about delinking (I'd rather see a scheme done in a way we can recover autoformatting w/o linking, but that's not critical). However, when I read this page, the thing that stands out is that while the effort to remove date linking due to its horrendous implementation has been a process that has taken a couple years, it has been a slow, glacial change. However, this last change - and the most drastic, going from "data linking is depreciated" to "date linking is not to be used except in limited cases", which removes the concept completely - seems to have been pushed through by a handful of editors with minimal time for discussion at a broader level. I know Tony put messages on several top Wikiprojects to explain why date linking was being removed, but this is not the same as a centralized discussion or an RFC to determine, even with the strong body of reasons to get rid of it, if date linking should be abolished. The fact that editors are coming here and not finding such a consensus even with what are in the archives, is why there is still a point of contention - it feels like the discussion was all done behind closed doors and suddenly put into place with no chance for further rebuttal.

To me, the matter can be settled if someone would create an RFC to get it out there (specifically, the matter of removing date links completely), announcing it to the Village Pumps, WP:CENT, and if we can convince those that maintain is, watchlist-details. Run it for two weeks, and then say it's complete, archiving the page. Presumably, it will have support to remove dates, at which point, those naysayers will hopefully stop complaining now that consensus has shown to be there. This is an issue that effects every user of WP (including anons, even if for the better), so passing it as a consensus with as small an input that it seemed to have just feels wrong. But there might be significant resistance to that, and if that's the case, we then need to talk about it more. Right now, it is impossible to judge that from what we have presently, and that's my primary concern. --MASEM 16:33, 22 October 2008 (UTC)

• “The fact that editors are coming here and not finding such a consensus even with what are in the archives, is why there is still a point of contention - it feels like the discussion was all done behind closed doors and suddenly put into place with no chance for further rebuttal.” That will always be a problem on Misplaced Pages. There is a wide variety of volunteer, contributing editors. Most are unregistered and have no idea of the existence of special forums like this nor any interest in participating even if they did know. Even many registered editors are similarly inclined. It’s not at all surprising that editors wouldn’t have any knowledge of this decision to deprecate the linking and autoformatting of dates until they see their favorite articles affected by a bot.

  I don’t think the solution is yet more debate; debate would go on forever if we widened this and started all over fresh. And the conclusion would be the same: autoformatting, which can only be seen by registered editors and doesn’t benefit ordinary I.P. readers, would be deprecated. Linking of dates would also be deprecated since such links are rarely topical and germane.

  The end result of all that exercise would be an education for those who come in late to the discussion. So maybe what we need, is just that: a vehicle for education. Maybe we could have an essay explaining in detail, the rationale behind the decision. The trouble with that is, there are opponents of deprecation who would contest the “facts” in such an explanatory essay and, in the end, it wouldn’t be possible to write anything that was actually explanatory; it would just be a neutered re-hash of the nature of the debate.

  We’re just living the results of what is a pure democracy where there are no elected representatives. We rely instead on de-facto, volunteer leaders like Tony and others who care enough to stay engaged long enough to be familiar with all the details and persuade others to their point of view. It’s hard work and Tony has shown remarkable patience in the face of numerous instances of editors coming here complaining about how the linking of dates is such a splendid thing and no one told them about the change. In the end, we do have a general consensus. But a general consensus is not 100% of editors in full agreement. And it never was. Misplaced Pages would grind to a halt if every single editor who had a knee-jerk reaction to the deprecation of his or her linked dates could come here and make a politically-correct-sounding argument over how he wasn’t included in the decision so the decision is null and void.

  Now that I think about it, the best solution is to have two essays. One, written perhaps by Tony, could explain the reasoning behind the decision. If someone disagrees, they can write their own essay. It’s time to move on. The amount of debate that has gone on here over the years over just dates is nearly unimaginable. Greg L (talk) 17:02, 22 October 2008 (UTC)

Maybe Signpost can run an article about it? Also, I still think that the policy itself should be clarified. I don't think the casual reader would understand that it is saying "don't blue-link dates unless you have a specific reason why you want people to click on that link." I think that many editors continue to link dates for two reasons: 1) they like the way the blue-link LOOKS - that it gives prominence to the date. 2) newbies see other articles with dates blue-linked and think that this is the normal way to wikify articles. It takes some getting used to for those of us who have created many articles and carefully linked the dates to wrap our heads around such a big policy change, so I agree that publicity would be very helpful about the change, together with a clear explanation of WP:OVERLINK and the other reasons for the change. Happy editing! -- Ssilvers (talk) 17:07, 22 October 2008 (UTC)

If anyone wants to measure community acceptance, it is easy. Just go to Featured Articles, Good Articles etc and see for yourself. Those articles are, by definition the best that Misplaced Pages can offer. You can measure the ratio (actual date links)/(potential date links). I bet actual date links represent a tiny fraction of a percent of potential date links. Lightmouse (talk) 17:18, 22 October 2008 (UTC)

That's not really a valid method of gauging acceptance since you, Lightbot, Tony, and others seem to have already gone through and delinked dates in many featured articles. — Bellhalla (talk) 17:46, 22 October 2008 (UTC)

The way to solve that would simply be to examine the verisions at the time those articles passed FARLeadSongDog (talk) 22:25, 22 October 2008 (UTC)

Edits by bots should not count when trying to determine whether there is consensus. Tennis expert (talk) 10:12, 23 October 2008 (UTC)

• Good summary of the status quo and the nature of the problem, Ssilvers. I agree. Greg L (talk) 19:10, 22 October 2008 (UTC)

Lightmouse's suggestion assumes that FA and GA are good measures of community consensus and that most FA and GA reviews have considered this matter. Neither seems plausible. Septentrionalis PMAnderson 03:16, 23 October 2008 (UTC)

Given how recurrent the complaints about date unlinking are, it is obvious that there is no widespread consensus. I suggest running a wider RFC or poll, announced via the watchlist, community portal, etc. to try to get more people than the MOS regulars to voice their opinion. --Itub (talk) 08:52, 23 October 2008 (UTC)

• I'm not all that sure it's "obvious". What you might imply (and are you?) is that the consensus has changed. It took us two years to get this far, so I think it ain't all that likely to have changed in such a short time. The discussion here seems to be centred on some aspects of linking, but I don't see a wholescale revolt against the underlying principle of deprecating links. Anyway, how many people and how many discussions will it take to constitute a valid consensus? I believe what counts is that this was essentially a style matter, and it has reached a consensus here. In addition, the WP community is in a constant state of flux. We have to nail this thing at one point in time, and it appears to have been done. The horse is well and truly dead. Ohconfucius (talk) 09:28, 23 October 2008 (UTC)

Exactly right, Septentrionalis. Tennis expert (talk) 10:12, 23 October 2008 (UTC)

Indeed Ohconfucius. The "complaints" that Itub squeals about are coming from a tiny, loud minority of WPians who've probably been around for too long and are hard-wired against formatting improvements, because "they just don't like it". I'm sorry that they missed out on the long long debate here during 2007 and 2008, widely promulgated, at least towards the latter part.

What is telling is that when asked for examples of why the linking cancer should be allowed to persist, they are silent or offer (occasionally, if you're lucky) rather lame evidence.

What is also missing from their noisy complaining is the elephant in the living room: where was the original consensus for the linking frenzy? Please locate it for us (methinks it never existed).

It would be a breath of fresh air if detractors stopped in their tracks and thought of our readers: I hear of no readers, nor of WPians at large, lining up to bag the efforts of those who are bringing about these long-overdue improvements. Tony (talk) 09:49, 23 October 2008 (UTC)

• It's a filibuster. The conservative (used advisedly, I might add) opposition has observed a chink, and now want to exhume the putrid equine remains. Ohconfucius (talk) 10:58, 23 October 2008 (UTC)

If there ever were a "linking frenzy", that frenzy alone is evidence of consensus to have the links. You seem to be stuck on your personal dislike of date linking, in contravention of WP:IDONTLIKEIT, which you so conveniently and repetitiously cite to support your position. How ironic. Tennis expert (talk) 10:12, 23 October 2008 (UTC)

• Doesn't need to be a real frenzy, AFAICT. The guideline's been in existence for several years, and I know there are editors who refused to promote articles to WP:GA or WP:FA without rigid adherence to the linking (in the same way as there are those who insist on the use of citation templates). IT may also tend to support to view that most editors are merely law-abiding citizens who have been obediently inserting links because a guideline says so, for the benefit of DA. Or perhaps someone wrote a bot to automatically turn dates blue for the sake of the red-green colour blind? Ohconfucius (talk) 10:58, 23 October 2008 (UTC)

"The tiny and loud minority": what a perfect description of the self-selecting MOS legislature! (Of which admittedly I'm part of by posting here.) I agree that the consensus in this page changed. But the consensus out there never did, as evidenced by the reaction that often occurs whenever someone uses a bot to massively unlink dates in articles. FWIW, I don't give a damn about date links and don't use them myself. But I am interested in other issues related with units and numbers and it is annoying to have this page overflowing with WP:LAME discussion about date links for so long. --Itub (talk) 12:15, 23 October 2008 (UTC)

• Misplaced Pages is not a democracy; it operates by consensus. It is not a form of government; the chief difference is that there are times when governments must act, and doing nothing is worse than either decision. In democratic governments, the majority has a right to rule (limited by other rights); here, decisions can be put off until the eve of our publication date. Therefore a filibuster is no evil; it is a sign that some temporary and local majority (all our polls are temporary and local) prefers their own way to the compromises required to get a minority to join the consensus. Septentrionalis PMAnderson 13:39, 23 October 2008 (UTC)

• I would say that although I remain deeply unhappy about the inconsistent application of international vs American date formats in that 'no man's land of the non-British-non-American article. However, most of us are prepared to live with the decision arrived at by consensus, for the sake of harmony. Note that that debate only took place over a few weeks, if not days. How can we be sure it was reached correctly, and over a correct length of debate?? Sure, it may be OK to filibuster a bit, but when the debate has continued for two years, it is already one helluva filibuster. The battle lines are entrenched, and there is not much to suggest that minds will be changed any more. Ohconfucius (talk) 14:22, 23 October 2008 (UTC)

• No, that's not fair. The debate went on until a (possibly temporary) consensus among the people monitoring this page at the time was arrived at, for about a week. Once the changes started being made according to that "consesnus", then the complaints followed, and have been following since. I think the "consensus" needs to be revisited. — Arthur Rubin (talk) 15:02, 23 October 2008 (UTC)

• This is precisely the issue. Let's all be fair here; the goings on at WT:MOSNUM are not exactly well broadcast throughout the community as a whole. To your average Misplaced Pages editor, the methods by which these guidelines were written up and decided upon are a fairly esoteric process - it is no surprise that the discontents often refer to the regulars here as "MOS-wonks" or something less flattering. Many editors aren't even aware of the existence of the Manual of Style as a whole, let alone the obscure subsection of a section dealing with autoformatted dates. Since it's "just a guideline" it isn't that important, but when attempts are made to enforce the guideline it gets people's attention. That's why RfC exists. No one would reasonably expect that an RfC would be filed for small minute matters, but when attempting to enforce something on a Misplaced Pages-wide level it requires Misplaced Pages-wide consensus. Various avenues have been suggested here (using RfC, signpost, some other form of catching the community's attention) but for some reason have yet to be undertaken. Having been made aware of this issue for some time, I don't doubt that this back-and-forth is going to continue ad infinitum until there is some coherent consensus that can be pointed out. Shereth 15:10, 23 October 2008 (UTC)

Arthur, you say in your edit summary "Quite. Agree with title"; do you agree with the opening sentence underneath that title, naturally by the same editor? "First off, I completely agree about delinking". Tony (talk) 15:12, 23 October 2008 (UTC)

Shereth, it's all very easy—specious, actually—to keep demanding more and more consensus just because you don't agree. Perhaps your strategy is to be dissatisfied until there's a complete referendum of all WPians who've ever set foot on the project; maybe even all visitors too? It's just a spin-way of trying to discredit a well-established consensus. Purusing this line, you could batter down any consensus that has ever been established on WP, and freeze policy and practice to your liking. People won't be fooled by this. Tony (talk) 15:17, 23 October 2008 (UTC)