This is a list of forums at which notification of the poll has been posted:
I think that should be adequate. Dabomb87 ( talk) 20:44, 21 March 2009 (UTC)
I see one result of the edit-war with Locke Cole has been to make this an approval poll. I regard this as most undesirable, since it will make my opinion on datelinking very difficult to express; I oppose adding any language on date-links, but I weakly oppose #3, as being as close as possible to silence. Septentrionalis PMAnderson 19:46, 26 March 2009 (UTC)
The disputed broad date links question has been removed but now added back as option 4. I would like to see the options removed as well. I think we will get contradictory data. I can see how someone could say "no we don't need special rules" and then say that "yes, we should have links for birthdates". Someone else would say "no we don't need special rules" and say "no, we should not have links for birthdates". The key dispute seems to be that we don't know what consensus is on when dates have appropriate context - the RFC should be focusing on that debate. Otherwise, if we just point people to WP:CONTEXT, we'll have the same edit wars between people who think that date links provide great context and those who don't. Karanacs ( talk) 19:58, 26 March 2009 (UTC)
The overall structure of the poll ( this version) is now, disregarding the actual text of the proposals (which I've not looked at again), very much improved in my opinion. Even though there are multiple options, it still "feels" as if I'm only being asked one question rather than four. One suggestion: I wonder if it would be worth losing the "Comments regarding the linking of month-days" and "Comments regarding the linking of years" sections. Do we really want to encourage a whole raft of undirected comments? Isn't it sufficient to just ask people if they have another, specific proposal? Matt 86.152.243.59 ( talk) 20:22, 26 March 2009 (UTC).
I ask that everyone involved here examine this version and post here how you feel about it. Note that I consolidated the year proposals and jettisoned a day-month proposal that didn’t have a snowball’s chance in hell of passing (given past RfCs). I also deleted the mini-poll within a poll (that Locke editwarred over and got himself blocked in the process). As SteveB67 stated above, this approach avoids that “retarded” structure we had before. Greg L ( talk) 20:32, 26 March 2009 (UTC)
Comment below please.
Septentrionalis PMAnderson 22:21, 26 March 2009 (UTC)
Greg, with respect to your request above, I'll try to look over the text in question. To be honest, I've been rather focused on the DA aspect of the RfC, and haven't really been following the date-linking discussions too much, so I'll have to read through the discussions first. Cheers. -- Ckatz chat spy 02:46, 27 March 2009 (UTC)
I've kept quiet regarding the wording of the proposal—until now. I'm no longer comfortable with the direction the wording of the background statement is taking. The current "Proposed options would add formatting for date ranges (e.g. "25–30 March 2009") and slashed dates ("the night of 20/21 December")" is inappropriate for a background statement. As it reads, it comes across as a done deal and/or simple, however we all know that there is debate about exactly how those additions would work (it's something that even UC_Bill didn't get right before he left the project). At best, the wording should be included as part of the "For" case, however if it is to remain, then more doubt should be cast on the ability to deliver it (in a reasonable timeframe) as part of the proposal.
In addition, the included reference to the (undocumented) new coding scheme omits the following important words: "To guarantee article consistency, this coding scheme requires all dates in an article to be coded". The current background scheme is misleading without that information.
Thanks.
HWV258 22:58, 26 March 2009 (UTC)
I notice that the wording has now been changed to "Additional proposals would seek to extend formatting to date...". Thanks for making a slight change, but the point remains that possible enhancements have no place in a background statement. I propose that the entire text "Additional proposals would seek to extend formatting to date ranges (e.g. "25–30 March 2009") and slashed dates ("the night of 20/21 December")" be deleted. In addition, I propose that the text "To guarantee article consistency, this coding scheme requires all dates in an article to be coded" be added to the Proposed replacements section of the Background statement.
Once again, thanks.
HWV258 01:41, 27 March 2009 (UTC)
I've moved this to the location you've suggested, and merged the wordings as follows:"As with the existing system, all dates in articles would need to be marked up with an autoformatting syntax."
I think this would be beneficial in that it explains that both old and new systems require markup to operate. -- Ckatz chat spy 02:21, 27 March 2009 (UTC)" As with the original system, all dates in an article would require markup in order to guarantee consistency."
I was proofreading the page (I left proposals and statements alone) and seek clarification on the following points:
I hope these will be easily addressed. I may not see the responses for 18 hours, please bear with me. Dabomb87 ( talk) 03:39, 27 March 2009 (UTC)
No problems; except that "user" is likely to encompass visitors who read the articles—they will not be able to choose a preference, for example. Can we not make it "registered users"? It's no big deal, and I'll back down if you feel strongly, but I think the distinction has to be retained throughout. Have you looked at the expanded titles for the Month-day proposals? I want to do the same thing for Years. Tony (talk) 05:29, 27 March 2009 (UTC)
The poll reads Two main date formats are used by English-speakers: March 11, 2009 (“MDY”, mainly in North America) and 11 March 2009 (“DMY”, mainly elsewhere).
Is there a place I can vote that DMY be the standard for Wikipedia (note, I am in North America)? Instead of voting for or against autoformatting, I'd rather just vote for a standard (we have one for football/soccer, so why not dates?). — Reinyday, 06:10, 27 March 2009 (UTC)
Can we go back to supporting proposals rather than just giving views? Let's be clear that this is a poll, and the header "views on proposal" will be less encouraging to users than "I support proposal x" - it's simply not as clear. Ryan PostlethwaiteSee the mess I've created or let's have banter 08:25, 27 March 2009 (UTC)
Re revert by CKatz: I believe it was 'your side' which proposed to de-emphasise any specific implementation idea, and focus on general principles. That does not, IMHO entitle you to slip in details of this revision at this stage, because it really is quite a specific proposal. It's jumping the gun back into specifics, in fact. Ohconfucius ( talk) 09:47, 27 March 2009 (UTC)
This proposal doesn't seem entirely coherent to me. The first sentence specifically says that the relevance of the two articles to each other does not matter in deciding whether to link; the second sentence says that you make the decision whether to link on a case-by-case basis. But what could you base your decision on, if not mutual relevance? This proposal should either just say 'always link the first occurrence', or be withdrawn as internally inconsistent. Colonies Chris ( talk) 10:14, 27 March 2009 (UTC)
I think I know more than most what is going on but I find the proposed RFC too complicated to understand. Please focus on trying to simplify it. The linking debate has always been troubled by differing ideas of 'relevance'. The phrase 'first occurrence' has only ever cropped up in date debates and non-date debates when people have suggested that an item satisfies some other 'relevance' criterion. Then people start talking about repeating links further down the page.
If this RFC should have a 'first occurrence' option (and I think it should not), then it seems to me to be a subset of the other proposals that purport to define 'relevance'. We should not have a proposal that allows links that do not meet a relevance criterion. Thus it would be 'Proposal 1a (conservative approach) and Proposal 1b (conversative approach with links limited to first occurrence), then the same split for Proposal 2a and Proposal 2b. Furthermore, Proposal 4 looks to me like Proposal 1. I find the current RFC text confusing in logic and in wording (e.g. similar proposals have different phrasing). Lightmouse ( talk) 18:31, 27 March 2009 (UTC)
To all: there will be intolerable editwarring if we have editors trying to neuter the other side’s autoformatting statements by hiding behind the apron strings of “ Dude, because you want it isn’t a good enough of a reason for me.” If someone has a problem with a factual inaccuracy in a Statement for/against, bring it up with Ryan. These Statements are supposed to present opposite views of the issue. I don’t expect the opposing side to like ours, just like I don’t like what you guys have in yours. Worry about your own stuff. Greg L ( talk) 16:17, 27 March 2009 (UTC)
I suggest that for the year and month-day voting, we have one vote per editor. Otherwise, we could have editors voting for three counter-proposals just to stack the deck against one they don’t like ( not that anyone would actually do such a thing). Whereas I can fully buy into the notion that one camp or another here might have a variety of views on a given subject, it doesn’t pass my ‘grin test’ here that some editor really needs to vote like “I can’t make up my mind; all three alternatives impress me soooo very very much over the option I hate.” Greg L ( talk) 20:09, 27 March 2009 (UTC)
I find this position unacceptable. Approval polls always permit multiple votes, for positions, again, like mine: my first choice is #4, but if it is impossible, #3 is much less bad than #1 or #2. Septentrionalis PMAnderson 21:04, 28 March 2009 (UTC)
If people vote against autoformatting, and in favour of linking (in whatever circumstances), then all the links disappear anyway? Should this say something like "...removed unless it has an accepted linking function, as determined by the results of the poll"?
Addition shortly after I first posted this: If there's no intention, whatever the outcome, of disabling the current autoformatting system, then I think for clarity we should say so, and say that it may have less and less effect as links deemed superfluous are removed. —Preceding unsigned comment added by 86.150.102.46 ( talk) 20:53, 27 March 2009 (UTC)
The connection with efficiency of database dumps is not at all clear.
I don't understand what this means.
The "for" statement talks about autoformatting as if it's a fully functioning feature. Is this "original" autoformatting feature something that predated the current one, or is the question of whether autoformatting succeeded or failed one of the points in dispute? A couple of words here to clarify this point would be helpful I think.
Don't understand what this is saying. At present, "the community" apparently has autoformatting. How is that "taking a conservative attitude"? Does it mean the community would be taking a conservative attitude if people vote against?
The fact that any specific date format you choose as your preference is unambiguous and non-confusing does not seem to be a demerit. At worst it seems neutral.
I don't really understand this. I wondered if it meant "Offers editors a way of directly linking to relevant content, rather than forcing readers to resort to the less precise "search" function". Even so, I don't get it, because if you "search" for "March 20", say, then you're taken straight to the article. I don't see how this is "less precise" than a link.
These sections are presented as "editorial" rather than opinion (as was the case with autoformatting), so it reads a little oddly that they directly contradict each other; for example, links to birthdates and deathdates being both "useful" and of "little or no relevance".
This does not read like an actual disadvantage of date linking.
This seems so similar to the month-day linking background statement that I wonder about the point of having two separate sections.
Matt 86.150.102.46 ( talk) 20:38, 27 March 2009 (UTC)
In what context is option 1 "conservative"? I'm not sure I follow that. I suppose it minimizes the number of links, but the old status quo is that everything should be linked, so it maximizes the changes required, making it less conservative in that sense. — Arthur Rubin (talk) 21:07, 27 March 2009 (UTC)
Sorry, 'bout this folks... I think I should shut up fairly soon. But there's one more thing that occurred to me.
Is it the case that a policy of blanket date-linking (indiscriminate linking of every single date) is not an option on the table?
One of the reasons I ask is that the purpose of the "Month-day linking -- Advantages of month-day markup" section seem to be to alert people to the idea that the markup that date linking entails might have other uses, so perhaps if they don't feel strongly one way or the other about linking per se then they might want to consider these extra benefits as a factor in their vote. However, for the first benefit ("Clearly indicates which strings refer to dates directly") to be realised through date linking, it seems to me that pretty much every "direct" date would have to be linked. If that is definitely not going to happen then this advantage will never materialise. Matt 21:45, 27 March 2009 (UTC). —Preceding unsigned comment added by 86.150.102.46 ( talk)
Regarding my recent burst of copy-edits: I started and just couldn't stop. I really tried hard to improve the readability of each section, and (apart from my final edit) had no intention of changing the meaning of any point. If I have, please revert as necessary. Cheers. HWV258 22:08, 27 March 2009 (UTC)
This was initially brought up a few days ago, but we've yet to properly address the use of Brion Vibber's quote in the "against" section of the autoformatting poll. His original comment as posted at Bugzilla in late 2007 was:
There is already consensus that his second sentence is best left out because of its sarcastic nature. However, the use of only part of his first sentence appears to take his statement out of context, a problem that is further compunded by the recent rewrite (which moves it up in the text). The way it is currently used is as follows:
"My personal recommendation would be to remove all date autoformatting."
Vibber posted his original statement over fifteen months ago, but the quote as presented suggests that he has personally endorsed the current position described in the "against" post, which calls for the removal of DA and support for the Manual of Style's multiple date format standard. This is compounded by the addition of a period after "autoformatting", which did not appear in the original statement. Vibber, however, qualified his position by in fact calling for a rewrite of the MoS to recommend one date format standard (DMY):
"My personal recommendation would be to remove all date autoformatting and let a sane manual of style recommend the fairly standard international English form, eg '4 December 2008'" (bold text added here to highlight)
There is no indication that he endorses removing autoformatting if the current mixed-format system remains. Look, I certainly don't object to the use of the statement; Vibber's opinions are as valid as any other Wikipedia contributor. I do, however, take issue with the selective use of his quote to suggest he has personally endorsed the "against" position. At the very least, someone should ask him if the wording and context accurately reflects his current point of view, and (more importantly) if he even wants to be seen as taking sides in this RfC. -- Ckatz chat spy 22:17, 27 March 2009 (UTC)
P.S. (on second thought, looking at the quote): it seems clear that Brion’s statement isn’t in the least being taken out of context. His sarcastic comment at the end (Of course that's too simple and obvious for Wikipedia) doesn’t signal to me that he was joking; only that he recognizes the same reality that I do: the absurd amount of effort to have sanity rule on Wikipedia—like the my three-month-long battle to get us back to using kilobyte (KB) instead of kibibyte (KiB). In the case of quoting Brion’s statement, this is much to do about nothing. Let’s conjecture what would happen if we quoted all of the germane portion and omitted the sarcastic part; you know what would be added(?): …and let a sane manual of style recommend the fairly standard international English form, eg '4 December 2008'. I can’t see how the full quote advances anyone’s cause here. This is what linking is for. Greg L ( talk) 01:03, 28 March 2009 (UTC)
The first half of Brion's statement sure helps that cause, but wait: at the same time, he also said that we should change the Manual of Style, do away with the regional variations, and adopt one format (DMY) across the entire encyclopedia. Seems at odds with what you've suggested he supports. Beyond that, I was of the understanding that, when quoting someone, it is customary to avoid changing the quote, even to the point of not adding extra punctuation. Brion said "Do x and then do y." Here, it has been rewritten so as to suggest that he only said "Do x." It is very easy to say that there is a link for transparency, but everyone knows full well that most readers won't bother to click through to check the link, they'll just take it at face value and assume (from the context) that Vibber is supporting your specific proposal. Nothing here indicates that he was actually speaking as part of an entirely different discussion well over a year ago. Look, if you're so confident that your use of the statement is correct, why not do as I suggested and actually ask Brion? Given the fractious nature of this whole RfC, and the fact that he is one of Wikipedia's senior officials, it would seem the courteous thing to do, rather than just presuming his support. -- Ckatz chat spy 03:34, 28 March 2009 (UTC)"Editors need to retain local control over their date formatting as they already do so under the simple, well-accepted rules at MOSNUM."
The statement is positioned and presented in a way that makes it appear as if Vibber was commenting on the recent mass delinking efforts, which is absolutely not true. -- Ckatz chat spy 07:41, 28 March 2009 (UTC)...one user has unlinked and corrected dates in more than 7,000 articles, yet has received only a handful of objections. Notably, WikiMedia's Chief Technical Officer, Brion Vibber, has stated: "My personal recommendation would be to remove all date autoformatting…".
Many thanks everyone for all your efforts with this - it's truly appreciated. I've been discussing when this should be finished by with Tony and it would be good if things could be finished by 0:00 (UTC) on Sunday morning. It's important the poll is stable for the start. I'll protect the page for 24 hours until the start time at this point. After this, the only edits to be made will be by using the edit protected template (but they should be very minor at that point in time). I've got work tomorrow, but I'll catch up with things tomorrow night. We're currently at a stage where the poll is pretty stable already and we only want minor changes from this point in time please. If everybody could leave all the proposals as they are (obviously, minor clarifications can be made to them, but please don't start removing any of the proposals at this stage) then that would be good. Many thanks, and I hope this is okay with everyone. Ryan PostlethwaiteSee the mess I've created or let's have banter 23:22, 27 March 2009 (UTC)
Ryan: You may not be aware that information about the poll appears at the very opening and in the lead to "Poll"? If retained in two separate places, they need to be consistent. Also, I feel that we should not be locking the community into yet another poll unless it's deemed necessary. I've had ago at changing inevitability into an "if necessary". Is this OK? I certainly don't want users to raise their eyes and get irritated ... "oh no, not another one after this too? I'm outa here". Some users might well do that. Tony (talk) 04:34, 28 March 2009 (UTC)
Sssoul: I really think the brief titles within parentheses will help users. It is a big ask of them to navigate through, and these titles will end up being a long way down from each other. How could one remember? I believe the brief explanatory titles need to be added to the response headers to match ... "I support Option #3 (link on first appearance)", not just the vague "I support Option #3". The poor readers will be scrolling a long way down to each.
Arthur, why not use the correct preposition, then ("On", not "In")? It would not affect your case, would it? And why not say what you mean, WRT automated recognition? The current text sends readers down a rabbit-hole, because they—like I, until a minute ago—thought you were referring to readers, not scripts/bots. Now it becomes a little clearer why "dates appearing in quotations" are relevant, but I think this is an esoteric point that doesn't help your case. Better to explain clearly to start with.
You need to insert something like "Clearly marks out date strings for automatic recognition." Is that the core of it? At the moment, we have:
"Clearly indicates which strings refer to dates directly (as opposed to, for example, dates appearing in quotations, which would not be marked up, or coincidences ("In June 19 planes departed").
Tony (talk) 07:55, 28 March 2009 (UTC)
(outdent) yes, but other people might have the same subjective opinion about other proposals; besides which, i never vote conservative and don't want to start now. (that's levity - but the word is loaded with associations that you don't mean - so why use it.) "Option 1 (link only to relevant date articles)" or "Option 1 (link dates rarely)" would parallel the other "titles". what's wrong with either of those? Sssoul ( talk) 12:31, 28 March 2009 (UTC)
Comment FYI, I might not have been clear enough (sorry) - the change to "departed" was only to avoid the "shot down" idea... it was always intended as a word choice change (i.e. original issues untouched) to avoid the idea of planes being shot down. (Might sound silly, but I thought it would come across better.) Do with it as you see fit... -- Ckatz chat spy 20:41, 28 March 2009 (UTC)
i propose conflating the two points about bots/scripts as follows: "Clearly marks out date strings for recognition by bots/scripts, which simplifies the automated processing of article text and the gathering of metadata." and then the "rejoinder" about markup not being needed to prevent ambiguity can be removed. Sssoul ( talk) 18:59, 28 March 2009 (UTC)
When this poll (finally) begins on 30 March, what will be the scope of this talk page? Will it be limited to general discussions of the poll's structure and purpose, or will the usual rants and such be permitted? I think we should set a firm limit, as this page tends to grow exponentially—I just archived half this talk page, and it is still quite long. Dabomb87 ( talk) 03:53, 28 March 2009 (UTC)
I agree with you and made a similar point above. The proposals mix three issues:
That is how it seems to me. I find it complex and confusing with at least eight permutations. Lightmouse ( talk) 10:55, 28 March 2009 (UTC)
Does 'case by case' in proposal 3 mean 'apply relevance rules in proposal 1 or proposal 2'? Lightmouse ( talk) 10:07, 28 March 2009 (UTC)
What we're desperate for is some (very?) short essays explaining the following points;
Could they please be created at User:Ryan Postlethwaite/Date strings and User:Ryan Postlethwaite/Processing article text ASAP? I'm not fussed who does it, but they need explaining. Ryan PostlethwaiteSee the mess I've created or let's have banter 19:30, 28 March 2009 (UTC)
(outdent): once again: conflating the two points makes them comprehensible, sensible-sounding, etc - can we please have: "Clearly marks out date strings for recognition by bots/scripts, which simplifies the automated processing of article text and the gathering of metadata." and then the "rejoinder" about markup not being needed to prevent ambiguity can be removed. Sssoul ( talk) 09:08, 29 March 2009 (UTC)
I've gone ahead and protected the page until the poll start time. There's 24 hours left, and this will mean the poll is stable for it starting. I'll be back from work at around 18:00 (UTC), so if there's any edits required that can wait till then I'll happily make them. If any edits are required urgently, then you can use {{ edit protected}} (but make sure you have consensus!). All the best everyone, Ryan PostlethwaiteSee the mess I've created or let's have banter 23:53, 28 March 2009 (UTC)
Is there any reason why the "month-day" question has sections for "[v]iews on Option #n", whereas the "year-linking" question has sections for "I support Option #n" with the requirement to only post to one of them? Wouldn't using the same format for both make sense? (I like more the former, but de gustibus ...) -- A. di M. ( talk) 02:07, 29 March 2009 (UTC)
I object to the claim of "relevance" for the year linking articles; option 1 is something like "link relevant year articles", while option 2 would be "link relevant years". Since we're almost certainly not going reach agreement before the deadline, I think the subtitles need to go. — Arthur Rubin (talk) 02:16, 29 March 2009 (UTC)
(outdent) here's one more try - i hope you can agree that they're quite clear, legible, sensible, etc:
please bear in mind that these subtitles don't alter the substance of the options - they're just supposed to serve as "reminders" of which option is which, to make life easier for readers/voters. Sssoul ( talk) 15:37, 29 March 2009 (UTC)
I feel dumb: I can't for the life of me see the difference in these titles, which is bad, since they need to distinguish the options for voters. What is wrong with the current two titles? Tony (talk) 16:47, 29 March 2009 (UTC)
Ryan,
The transcluded voting page for Wikipedia:Date formatting and linking poll/Month-day responses should have the same “one vote” statement, shown here…
…as does Wikipedia:Date formatting and linking poll/Year-linking responses, should it not? If so, will you please add this this code:
:''Please indicate your support vote under '''ONE''' option, accompanied by a concise explanation for your choice. Your explanation is important in determining the community consensus.''
…immediately under this code:
===Month-day responses===
… here (unless, of course, this was intentional). In advance, thanks. Greg L ( talk) 02:22, 29 March 2009 (UTC)
P.S. Oh yes, and please also change the four occurrence of Views on
with I support
in order to harmonize everything.
Greg L (
talk) 02:28, 29 March 2009 (UTC)
Before I rush into anything here, I want some more opinions - please everybody comment on what you think's best. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:17, 29 March 2009 (UTC) The more I think about this, the more I believe "One vote per section" is a bad idea. It's highly likely that some of the community won't have a preference between a number of the proposals - They may for instance broadly agree with 1 and 4 so wish to support those and I can't see a good reason to exclude them. I'm not seeing a good reason at present for one vote per section. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:24, 29 March 2009 (UTC)
The first past the post voting system (i.e. one vote per question) is simple to explain, understand, and analyse. The outcome is the most popular candidate. That doesn't always apply in preferential voting. If the outcome is not going to be the most popular candidate, then all stakeholders must accept that before the RFC starts. Lightmouse ( talk) 12:29, 29 March 2009 (UTC)
One vote only is the only sensible thing. If not, the poll is simply badly constructed. A poll should be designed to make choices mutually exclusive. Questionnaires may allow for multiple "votes" but those are not made to pick a "winner". And I guess this is what we try to attain here. So, allowing multiple votes is a no-no (unless they are required to be associated by weights summing to one, in which case nobody would bother to vote at all).-- HJensen, talk 16:14, 29 March 2009 (UTC)
You guys keep promising that “cool‑beans technology is just around the corner that will make god-damned gold dust pour out of every user’s USB port” and the community still rejects what you guys are selling. The “other” side here clearly knows what will likely happen if there is one vote per issue per editor: participants in the poll will vote for the best ones (our stuff), and you’ll get your asses kicked. You fear another repeat of “making it easy to get a clear gauge of the community consensus.” So you will do anything and everything to game the system, including try to structure the voting so A) you guys could vote for all three options in opposition to the one you fear, (and do so with absurd “4”-level votes to stack vote counts if you could), and B) make a huge cluster-pooch of voting results so it takes a team of MENSA members to divine what the community consensus is.
There are already four options each on linking and they each mean something different. To help make an already-complex RfC as unambiguous as possible, voters can simply vote for the one they like best (*sound of audience gasp*). You know; like how people vote for politicians in an election (where there is more “overlap” than you can shake a stick at). It’s pretty unfortunate that all these Well… Duh! concepts are so hard to implement. As much as you want to jump up and down and scream about how one‑editor / one‑vote will revoke the Magna Carta, allow Tony and me to imprison innocent Wikipedians at Guantánamo without charges, and result in forced sterilizations of the poor, your arguments belie what you really want: to desperately seize on any trick in the book to game the system so the true community consensus on this issue is difficult to discern. We’d have to be insane to do as you ask, otherwise, we’d be at this forever. Greg L ( talk) 16:21, 29 March 2009 (UTC)
I see no problem with allowing multiple votes so long as editors provide a "first choice", "second choice" designation with each vote. Also, I must again strongly disagree with an approval voting system. Editors must be able to oppose options. Also, if you're going to use approval voting, you should use it for auto formatting as well... I very strongly disagree with the mischaracterizations of the prior poll results directly above by Greg and HJensen directly above. — Locke Cole • t • c 18:05, 29 March 2009 (UTC)
Strictly speaking, we want to determine consensus, not plurality. So the first-past-the-post is supposed to be inappropriate: we don't want to consider an outcome such as "45 votes for option A, 56 votes for option B, 51 votes for option C, and 48 votes for option D" to be consensus for option B. (I obtained these numbers by computer simulation of 200 people voting randomly with equal probability for each option.) On the other hand, it is quite unlikely that consensus by dictionary definition will very likely not be reached, and, if we aren't willing to implement a more complicated system such as elimination runoff or the Schulze method, first-past-the-post is a lesser evil than e.g. casting dice. After all, there is no such thing as a perfect voting system. -- A. di M. ( talk) 20:25, 29 March 2009 (UTC)
The sentence "even though it is occasionally celebrated on a different day due to Lent" should actually be "... due to the Holy Week". It is very common for 17 March to occur during Lent, but St Patrick's day isn't moved unless it occurs during the last week of Lent (the Holy Week), which is something very rare (it happened in 2008 and won't happen again until 2160, unless the rules for the date of Easter are changed meanwhile). -- A. di M. ( talk) 19:56, 29 March 2009 (UTC)
Ryan, at Month-day: Option #2, would you please replace the very last sentence in the paragraph that reads like this(?):
(even though it is occasionally celebrated on a different day due to Lent)
with this:
(even though it is occasionally celebrated on a different day due to Holy Week)
....I'm planning to archive this whole page. People have already said to me the hostility here is going to put them off participating. It's important we start off with a clean slate. Ryan PostlethwaiteSee the mess I've created or let's have banter 21:50, 29 March 2009 (UTC)
I am afraid that the format of this poll will not permit me, and those who agree with me (if any) to express their opinion. I therefore intend to dispute it, and present a FoF that it is unacceptable.
I thank Ryan for his efforts, and regret that they have been derailed by a successful effort to distort the results. Septentrionalis PMAnderson 23:08, 29 March 2009 (UTC)
:::While I haven't looked over this page's archives, I am somewhat confused. What about this poll do you feel is unacceptable? This poll is quite fair in identifying various viewpoints, I believe.
NuclearWarfare (
Talk) 00:54, 30 March 2009 (UTC)
Comment We all know that this poll is not perfect—far from it. However, it is better, more specific, and was worked on more than previous RfCs combined, and I think we need to appreciate that. Whatever the problems are, we must accept that this is the RfC to end it all. The community nor the editors who have debated and debated over date autoformatting and linking can take another one. All we can do now is wait for phase 1 to end. Dabomb87 ( talk) 03:19, 30 March 2009 (UTC)
Excuse me, guys above ... the instructions make it amply clear that comments are important in the interpretation of consensus. Everyone is invited to write comments after their choice (and to choose "Neutral" if they wish). There is even a separate comments subsection beneath each response section. There should be no doubt that the RfC provides lattitude for expressing individual feelings. At the same time, let's be practical: the community has to come to some kind of decision, and making it likely that the results will be melted treacle spattered all over the place is not practical. I think Ryan has come to a reasonable solution, and both camps had a lot of prep. time in which they were able to comment on the structure and shape. Tony (talk) 09:07, 30 March 2009 (UTC)
I recommend a quick trip to Wikipedia:NAM. Sincerely, GeorgeLouis ( talk)
Unfortunately, I started the poll an hour early - yup, BST started on Sunday and I've been caught out. I'm going to leave it open, because I've already made all the notifications and I would hope that 1 hour will make little difference. Apologies to all parties. Ryan PostlethwaiteSee the mess I've created or let's have banter 23:12, 29 March 2009 (UTC)
You may want to consider putting your "comment" box from the top of the page into an editnotice, so people are more likely to see it. Anomie ⚔ 23:17, 29 March 2009 (UTC)
There seem to be a series of choices missing, such as the one we used to have .. -- User:Docu
I think the main question that needs to be resolved is: who the target audience of autoformatting? If all remains as it is then we can foresee some very basic and obvious problems:
It is clear that we must move away from this current state of affairs, it does not benefit the majority. The two most appropriate solutions are the "magic word/set dates for all users" solution and the conflicting "remove all autoformatting" solution.
Which brings me to my main point: I don't see anyone supporting the use of autoformatting through linking.
It appears that options #1 and #4 of month/year linking are gaining consensus and there is much overlap between those two choices, in that #4 is in some ways the non-instruction creep version of #1. Yet, despite this growing consensus, wikilinking still seems to be a major aspect of people's reasoning to oppose autoformatting.
We need to disentangle these two issues and first of all have a vote on the deprecation of autoformatting through wikilinks. I think we will find consensus on that issue and it will mean people's responses to further polls on autoformatting will not be diluted with reference to "overlinking" etc. Sillyfolkboy ( talk) 18:35, 30 March 2009 (UTC)
I don't like overlinking any more than the next guy, but I really think that option 4 will lead to editors overreacting. We don't want to kill all year links, and if you have a special guideline for year links that says they can only be linked in X, Y, or Z situations, then people will think: "If year links have a special section, then that means they must be judged more strictly than any other links." I don't think that is what most people supporting the first proposal intend to support. I much prefer option one. Year links need to just quietly slide into the same mass of rules all other links abide by. If we overreact, then we will, I guarantee you, be back in a few months, after several edit wars, having another poll. Wrad ( talk) 17:11, 30 March 2009 (UTC)
First, I posted a notice of my suggestion here. What was the response by Ryan, the clerk overseeing this RfC? He responded fine by me. I was thinking of propsing something like that myself.
Then there is this exchange on Ryan’s talk page. He responded to you as follows: As it happens, Greg was right with the formatting - it's a poll and we need to be clear about that. … Sorry, I know that's not what you wanted to hear.
Then it was fully discussed here on the RfC talk page where many editors from all sides of the issue weighed in and the issue received a full and fair hearing.
Then Ryan, the clerk responsible for fairly overseeing and structuring this RfC, helped to restore the format to my suggested method [1].
Then, when you complained about the structure of the RfC here on the RfC talk page, Ryan responded to you as follows: Whilst I am disappointed that you dispute it, I thank you for your efforts with helping to create the poll. Ryan has shown spectacular patience and fairness in all of this. That little jewel of a response to your industrial‑strength whining comes about as close as any admin can get to “tough; go pound sand in your damned ear if you don’t like it.”
Finally, In my ANI against you for slapping {disputed tags} all over an ongoing RfC that had been thoroughly and fairly developed and properly supervised, you were advised there in the ANI as follows: I strongly urge all contributors who have been involved in this dispute to not make any further modification to the RFC page. None! You have had your chance to have a say in how the RFC is conducted. If you have concerns, please voice them on the talk page and let someone who is uninvolved make any changes deemed necessary. If you, PMAnderson, haven’t yet learned how to heed advise in the face of not always getting what you want, I suggest you start with that tidbit. Greg L ( talk) 20:27, 30 March 2009 (UTC)
“ | The more I think about this, the more I believe "One vote per section" is a bad idea. It's highly likely that some of the community won't have a preference between a number of the proposals - They may for instance broadly agree with 1 and 4 so wish to support those and I can't see a good reason to exclude them. I'm not seeing a good reason at present for one vote per section. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:24, 29 March 2009 (UTC) | ” |
And cease with this “Greg is a witch who caused our crops to fail and our midwives to weep”-horsecrap. The villagers just might look at each other with that wide-eyed look of epiphany, shrug their shoulders, and decide to burn your hut down so they don’t have to listen to your “waaa-waaa.” Greg L ( talk) 02:22, 31 March 2009 (UTC)
This is a list of forums at which notification of the poll has been posted:
I think that should be adequate. Dabomb87 ( talk) 20:44, 21 March 2009 (UTC)
I see one result of the edit-war with Locke Cole has been to make this an approval poll. I regard this as most undesirable, since it will make my opinion on datelinking very difficult to express; I oppose adding any language on date-links, but I weakly oppose #3, as being as close as possible to silence. Septentrionalis PMAnderson 19:46, 26 March 2009 (UTC)
The disputed broad date links question has been removed but now added back as option 4. I would like to see the options removed as well. I think we will get contradictory data. I can see how someone could say "no we don't need special rules" and then say that "yes, we should have links for birthdates". Someone else would say "no we don't need special rules" and say "no, we should not have links for birthdates". The key dispute seems to be that we don't know what consensus is on when dates have appropriate context - the RFC should be focusing on that debate. Otherwise, if we just point people to WP:CONTEXT, we'll have the same edit wars between people who think that date links provide great context and those who don't. Karanacs ( talk) 19:58, 26 March 2009 (UTC)
The overall structure of the poll ( this version) is now, disregarding the actual text of the proposals (which I've not looked at again), very much improved in my opinion. Even though there are multiple options, it still "feels" as if I'm only being asked one question rather than four. One suggestion: I wonder if it would be worth losing the "Comments regarding the linking of month-days" and "Comments regarding the linking of years" sections. Do we really want to encourage a whole raft of undirected comments? Isn't it sufficient to just ask people if they have another, specific proposal? Matt 86.152.243.59 ( talk) 20:22, 26 March 2009 (UTC).
I ask that everyone involved here examine this version and post here how you feel about it. Note that I consolidated the year proposals and jettisoned a day-month proposal that didn’t have a snowball’s chance in hell of passing (given past RfCs). I also deleted the mini-poll within a poll (that Locke editwarred over and got himself blocked in the process). As SteveB67 stated above, this approach avoids that “retarded” structure we had before. Greg L ( talk) 20:32, 26 March 2009 (UTC)
Comment below please.
Septentrionalis PMAnderson 22:21, 26 March 2009 (UTC)
Greg, with respect to your request above, I'll try to look over the text in question. To be honest, I've been rather focused on the DA aspect of the RfC, and haven't really been following the date-linking discussions too much, so I'll have to read through the discussions first. Cheers. -- Ckatz chat spy 02:46, 27 March 2009 (UTC)
I've kept quiet regarding the wording of the proposal—until now. I'm no longer comfortable with the direction the wording of the background statement is taking. The current "Proposed options would add formatting for date ranges (e.g. "25–30 March 2009") and slashed dates ("the night of 20/21 December")" is inappropriate for a background statement. As it reads, it comes across as a done deal and/or simple, however we all know that there is debate about exactly how those additions would work (it's something that even UC_Bill didn't get right before he left the project). At best, the wording should be included as part of the "For" case, however if it is to remain, then more doubt should be cast on the ability to deliver it (in a reasonable timeframe) as part of the proposal.
In addition, the included reference to the (undocumented) new coding scheme omits the following important words: "To guarantee article consistency, this coding scheme requires all dates in an article to be coded". The current background scheme is misleading without that information.
Thanks.
HWV258 22:58, 26 March 2009 (UTC)
I notice that the wording has now been changed to "Additional proposals would seek to extend formatting to date...". Thanks for making a slight change, but the point remains that possible enhancements have no place in a background statement. I propose that the entire text "Additional proposals would seek to extend formatting to date ranges (e.g. "25–30 March 2009") and slashed dates ("the night of 20/21 December")" be deleted. In addition, I propose that the text "To guarantee article consistency, this coding scheme requires all dates in an article to be coded" be added to the Proposed replacements section of the Background statement.
Once again, thanks.
HWV258 01:41, 27 March 2009 (UTC)
I've moved this to the location you've suggested, and merged the wordings as follows:"As with the existing system, all dates in articles would need to be marked up with an autoformatting syntax."
I think this would be beneficial in that it explains that both old and new systems require markup to operate. -- Ckatz chat spy 02:21, 27 March 2009 (UTC)" As with the original system, all dates in an article would require markup in order to guarantee consistency."
I was proofreading the page (I left proposals and statements alone) and seek clarification on the following points:
I hope these will be easily addressed. I may not see the responses for 18 hours, please bear with me. Dabomb87 ( talk) 03:39, 27 March 2009 (UTC)
No problems; except that "user" is likely to encompass visitors who read the articles—they will not be able to choose a preference, for example. Can we not make it "registered users"? It's no big deal, and I'll back down if you feel strongly, but I think the distinction has to be retained throughout. Have you looked at the expanded titles for the Month-day proposals? I want to do the same thing for Years. Tony (talk) 05:29, 27 March 2009 (UTC)
The poll reads Two main date formats are used by English-speakers: March 11, 2009 (“MDY”, mainly in North America) and 11 March 2009 (“DMY”, mainly elsewhere).
Is there a place I can vote that DMY be the standard for Wikipedia (note, I am in North America)? Instead of voting for or against autoformatting, I'd rather just vote for a standard (we have one for football/soccer, so why not dates?). — Reinyday, 06:10, 27 March 2009 (UTC)
Can we go back to supporting proposals rather than just giving views? Let's be clear that this is a poll, and the header "views on proposal" will be less encouraging to users than "I support proposal x" - it's simply not as clear. Ryan PostlethwaiteSee the mess I've created or let's have banter 08:25, 27 March 2009 (UTC)
Re revert by CKatz: I believe it was 'your side' which proposed to de-emphasise any specific implementation idea, and focus on general principles. That does not, IMHO entitle you to slip in details of this revision at this stage, because it really is quite a specific proposal. It's jumping the gun back into specifics, in fact. Ohconfucius ( talk) 09:47, 27 March 2009 (UTC)
This proposal doesn't seem entirely coherent to me. The first sentence specifically says that the relevance of the two articles to each other does not matter in deciding whether to link; the second sentence says that you make the decision whether to link on a case-by-case basis. But what could you base your decision on, if not mutual relevance? This proposal should either just say 'always link the first occurrence', or be withdrawn as internally inconsistent. Colonies Chris ( talk) 10:14, 27 March 2009 (UTC)
I think I know more than most what is going on but I find the proposed RFC too complicated to understand. Please focus on trying to simplify it. The linking debate has always been troubled by differing ideas of 'relevance'. The phrase 'first occurrence' has only ever cropped up in date debates and non-date debates when people have suggested that an item satisfies some other 'relevance' criterion. Then people start talking about repeating links further down the page.
If this RFC should have a 'first occurrence' option (and I think it should not), then it seems to me to be a subset of the other proposals that purport to define 'relevance'. We should not have a proposal that allows links that do not meet a relevance criterion. Thus it would be 'Proposal 1a (conservative approach) and Proposal 1b (conversative approach with links limited to first occurrence), then the same split for Proposal 2a and Proposal 2b. Furthermore, Proposal 4 looks to me like Proposal 1. I find the current RFC text confusing in logic and in wording (e.g. similar proposals have different phrasing). Lightmouse ( talk) 18:31, 27 March 2009 (UTC)
To all: there will be intolerable editwarring if we have editors trying to neuter the other side’s autoformatting statements by hiding behind the apron strings of “ Dude, because you want it isn’t a good enough of a reason for me.” If someone has a problem with a factual inaccuracy in a Statement for/against, bring it up with Ryan. These Statements are supposed to present opposite views of the issue. I don’t expect the opposing side to like ours, just like I don’t like what you guys have in yours. Worry about your own stuff. Greg L ( talk) 16:17, 27 March 2009 (UTC)
I suggest that for the year and month-day voting, we have one vote per editor. Otherwise, we could have editors voting for three counter-proposals just to stack the deck against one they don’t like ( not that anyone would actually do such a thing). Whereas I can fully buy into the notion that one camp or another here might have a variety of views on a given subject, it doesn’t pass my ‘grin test’ here that some editor really needs to vote like “I can’t make up my mind; all three alternatives impress me soooo very very much over the option I hate.” Greg L ( talk) 20:09, 27 March 2009 (UTC)
I find this position unacceptable. Approval polls always permit multiple votes, for positions, again, like mine: my first choice is #4, but if it is impossible, #3 is much less bad than #1 or #2. Septentrionalis PMAnderson 21:04, 28 March 2009 (UTC)
If people vote against autoformatting, and in favour of linking (in whatever circumstances), then all the links disappear anyway? Should this say something like "...removed unless it has an accepted linking function, as determined by the results of the poll"?
Addition shortly after I first posted this: If there's no intention, whatever the outcome, of disabling the current autoformatting system, then I think for clarity we should say so, and say that it may have less and less effect as links deemed superfluous are removed. —Preceding unsigned comment added by 86.150.102.46 ( talk) 20:53, 27 March 2009 (UTC)
The connection with efficiency of database dumps is not at all clear.
I don't understand what this means.
The "for" statement talks about autoformatting as if it's a fully functioning feature. Is this "original" autoformatting feature something that predated the current one, or is the question of whether autoformatting succeeded or failed one of the points in dispute? A couple of words here to clarify this point would be helpful I think.
Don't understand what this is saying. At present, "the community" apparently has autoformatting. How is that "taking a conservative attitude"? Does it mean the community would be taking a conservative attitude if people vote against?
The fact that any specific date format you choose as your preference is unambiguous and non-confusing does not seem to be a demerit. At worst it seems neutral.
I don't really understand this. I wondered if it meant "Offers editors a way of directly linking to relevant content, rather than forcing readers to resort to the less precise "search" function". Even so, I don't get it, because if you "search" for "March 20", say, then you're taken straight to the article. I don't see how this is "less precise" than a link.
These sections are presented as "editorial" rather than opinion (as was the case with autoformatting), so it reads a little oddly that they directly contradict each other; for example, links to birthdates and deathdates being both "useful" and of "little or no relevance".
This does not read like an actual disadvantage of date linking.
This seems so similar to the month-day linking background statement that I wonder about the point of having two separate sections.
Matt 86.150.102.46 ( talk) 20:38, 27 March 2009 (UTC)
In what context is option 1 "conservative"? I'm not sure I follow that. I suppose it minimizes the number of links, but the old status quo is that everything should be linked, so it maximizes the changes required, making it less conservative in that sense. — Arthur Rubin (talk) 21:07, 27 March 2009 (UTC)
Sorry, 'bout this folks... I think I should shut up fairly soon. But there's one more thing that occurred to me.
Is it the case that a policy of blanket date-linking (indiscriminate linking of every single date) is not an option on the table?
One of the reasons I ask is that the purpose of the "Month-day linking -- Advantages of month-day markup" section seem to be to alert people to the idea that the markup that date linking entails might have other uses, so perhaps if they don't feel strongly one way or the other about linking per se then they might want to consider these extra benefits as a factor in their vote. However, for the first benefit ("Clearly indicates which strings refer to dates directly") to be realised through date linking, it seems to me that pretty much every "direct" date would have to be linked. If that is definitely not going to happen then this advantage will never materialise. Matt 21:45, 27 March 2009 (UTC). —Preceding unsigned comment added by 86.150.102.46 ( talk)
Regarding my recent burst of copy-edits: I started and just couldn't stop. I really tried hard to improve the readability of each section, and (apart from my final edit) had no intention of changing the meaning of any point. If I have, please revert as necessary. Cheers. HWV258 22:08, 27 March 2009 (UTC)
This was initially brought up a few days ago, but we've yet to properly address the use of Brion Vibber's quote in the "against" section of the autoformatting poll. His original comment as posted at Bugzilla in late 2007 was:
There is already consensus that his second sentence is best left out because of its sarcastic nature. However, the use of only part of his first sentence appears to take his statement out of context, a problem that is further compunded by the recent rewrite (which moves it up in the text). The way it is currently used is as follows:
"My personal recommendation would be to remove all date autoformatting."
Vibber posted his original statement over fifteen months ago, but the quote as presented suggests that he has personally endorsed the current position described in the "against" post, which calls for the removal of DA and support for the Manual of Style's multiple date format standard. This is compounded by the addition of a period after "autoformatting", which did not appear in the original statement. Vibber, however, qualified his position by in fact calling for a rewrite of the MoS to recommend one date format standard (DMY):
"My personal recommendation would be to remove all date autoformatting and let a sane manual of style recommend the fairly standard international English form, eg '4 December 2008'" (bold text added here to highlight)
There is no indication that he endorses removing autoformatting if the current mixed-format system remains. Look, I certainly don't object to the use of the statement; Vibber's opinions are as valid as any other Wikipedia contributor. I do, however, take issue with the selective use of his quote to suggest he has personally endorsed the "against" position. At the very least, someone should ask him if the wording and context accurately reflects his current point of view, and (more importantly) if he even wants to be seen as taking sides in this RfC. -- Ckatz chat spy 22:17, 27 March 2009 (UTC)
P.S. (on second thought, looking at the quote): it seems clear that Brion’s statement isn’t in the least being taken out of context. His sarcastic comment at the end (Of course that's too simple and obvious for Wikipedia) doesn’t signal to me that he was joking; only that he recognizes the same reality that I do: the absurd amount of effort to have sanity rule on Wikipedia—like the my three-month-long battle to get us back to using kilobyte (KB) instead of kibibyte (KiB). In the case of quoting Brion’s statement, this is much to do about nothing. Let’s conjecture what would happen if we quoted all of the germane portion and omitted the sarcastic part; you know what would be added(?): …and let a sane manual of style recommend the fairly standard international English form, eg '4 December 2008'. I can’t see how the full quote advances anyone’s cause here. This is what linking is for. Greg L ( talk) 01:03, 28 March 2009 (UTC)
The first half of Brion's statement sure helps that cause, but wait: at the same time, he also said that we should change the Manual of Style, do away with the regional variations, and adopt one format (DMY) across the entire encyclopedia. Seems at odds with what you've suggested he supports. Beyond that, I was of the understanding that, when quoting someone, it is customary to avoid changing the quote, even to the point of not adding extra punctuation. Brion said "Do x and then do y." Here, it has been rewritten so as to suggest that he only said "Do x." It is very easy to say that there is a link for transparency, but everyone knows full well that most readers won't bother to click through to check the link, they'll just take it at face value and assume (from the context) that Vibber is supporting your specific proposal. Nothing here indicates that he was actually speaking as part of an entirely different discussion well over a year ago. Look, if you're so confident that your use of the statement is correct, why not do as I suggested and actually ask Brion? Given the fractious nature of this whole RfC, and the fact that he is one of Wikipedia's senior officials, it would seem the courteous thing to do, rather than just presuming his support. -- Ckatz chat spy 03:34, 28 March 2009 (UTC)"Editors need to retain local control over their date formatting as they already do so under the simple, well-accepted rules at MOSNUM."
The statement is positioned and presented in a way that makes it appear as if Vibber was commenting on the recent mass delinking efforts, which is absolutely not true. -- Ckatz chat spy 07:41, 28 March 2009 (UTC)...one user has unlinked and corrected dates in more than 7,000 articles, yet has received only a handful of objections. Notably, WikiMedia's Chief Technical Officer, Brion Vibber, has stated: "My personal recommendation would be to remove all date autoformatting…".
Many thanks everyone for all your efforts with this - it's truly appreciated. I've been discussing when this should be finished by with Tony and it would be good if things could be finished by 0:00 (UTC) on Sunday morning. It's important the poll is stable for the start. I'll protect the page for 24 hours until the start time at this point. After this, the only edits to be made will be by using the edit protected template (but they should be very minor at that point in time). I've got work tomorrow, but I'll catch up with things tomorrow night. We're currently at a stage where the poll is pretty stable already and we only want minor changes from this point in time please. If everybody could leave all the proposals as they are (obviously, minor clarifications can be made to them, but please don't start removing any of the proposals at this stage) then that would be good. Many thanks, and I hope this is okay with everyone. Ryan PostlethwaiteSee the mess I've created or let's have banter 23:22, 27 March 2009 (UTC)
Ryan: You may not be aware that information about the poll appears at the very opening and in the lead to "Poll"? If retained in two separate places, they need to be consistent. Also, I feel that we should not be locking the community into yet another poll unless it's deemed necessary. I've had ago at changing inevitability into an "if necessary". Is this OK? I certainly don't want users to raise their eyes and get irritated ... "oh no, not another one after this too? I'm outa here". Some users might well do that. Tony (talk) 04:34, 28 March 2009 (UTC)
Sssoul: I really think the brief titles within parentheses will help users. It is a big ask of them to navigate through, and these titles will end up being a long way down from each other. How could one remember? I believe the brief explanatory titles need to be added to the response headers to match ... "I support Option #3 (link on first appearance)", not just the vague "I support Option #3". The poor readers will be scrolling a long way down to each.
Arthur, why not use the correct preposition, then ("On", not "In")? It would not affect your case, would it? And why not say what you mean, WRT automated recognition? The current text sends readers down a rabbit-hole, because they—like I, until a minute ago—thought you were referring to readers, not scripts/bots. Now it becomes a little clearer why "dates appearing in quotations" are relevant, but I think this is an esoteric point that doesn't help your case. Better to explain clearly to start with.
You need to insert something like "Clearly marks out date strings for automatic recognition." Is that the core of it? At the moment, we have:
"Clearly indicates which strings refer to dates directly (as opposed to, for example, dates appearing in quotations, which would not be marked up, or coincidences ("In June 19 planes departed").
Tony (talk) 07:55, 28 March 2009 (UTC)
(outdent) yes, but other people might have the same subjective opinion about other proposals; besides which, i never vote conservative and don't want to start now. (that's levity - but the word is loaded with associations that you don't mean - so why use it.) "Option 1 (link only to relevant date articles)" or "Option 1 (link dates rarely)" would parallel the other "titles". what's wrong with either of those? Sssoul ( talk) 12:31, 28 March 2009 (UTC)
Comment FYI, I might not have been clear enough (sorry) - the change to "departed" was only to avoid the "shot down" idea... it was always intended as a word choice change (i.e. original issues untouched) to avoid the idea of planes being shot down. (Might sound silly, but I thought it would come across better.) Do with it as you see fit... -- Ckatz chat spy 20:41, 28 March 2009 (UTC)
i propose conflating the two points about bots/scripts as follows: "Clearly marks out date strings for recognition by bots/scripts, which simplifies the automated processing of article text and the gathering of metadata." and then the "rejoinder" about markup not being needed to prevent ambiguity can be removed. Sssoul ( talk) 18:59, 28 March 2009 (UTC)
When this poll (finally) begins on 30 March, what will be the scope of this talk page? Will it be limited to general discussions of the poll's structure and purpose, or will the usual rants and such be permitted? I think we should set a firm limit, as this page tends to grow exponentially—I just archived half this talk page, and it is still quite long. Dabomb87 ( talk) 03:53, 28 March 2009 (UTC)
I agree with you and made a similar point above. The proposals mix three issues:
That is how it seems to me. I find it complex and confusing with at least eight permutations. Lightmouse ( talk) 10:55, 28 March 2009 (UTC)
Does 'case by case' in proposal 3 mean 'apply relevance rules in proposal 1 or proposal 2'? Lightmouse ( talk) 10:07, 28 March 2009 (UTC)
What we're desperate for is some (very?) short essays explaining the following points;
Could they please be created at User:Ryan Postlethwaite/Date strings and User:Ryan Postlethwaite/Processing article text ASAP? I'm not fussed who does it, but they need explaining. Ryan PostlethwaiteSee the mess I've created or let's have banter 19:30, 28 March 2009 (UTC)
(outdent): once again: conflating the two points makes them comprehensible, sensible-sounding, etc - can we please have: "Clearly marks out date strings for recognition by bots/scripts, which simplifies the automated processing of article text and the gathering of metadata." and then the "rejoinder" about markup not being needed to prevent ambiguity can be removed. Sssoul ( talk) 09:08, 29 March 2009 (UTC)
I've gone ahead and protected the page until the poll start time. There's 24 hours left, and this will mean the poll is stable for it starting. I'll be back from work at around 18:00 (UTC), so if there's any edits required that can wait till then I'll happily make them. If any edits are required urgently, then you can use {{ edit protected}} (but make sure you have consensus!). All the best everyone, Ryan PostlethwaiteSee the mess I've created or let's have banter 23:53, 28 March 2009 (UTC)
Is there any reason why the "month-day" question has sections for "[v]iews on Option #n", whereas the "year-linking" question has sections for "I support Option #n" with the requirement to only post to one of them? Wouldn't using the same format for both make sense? (I like more the former, but de gustibus ...) -- A. di M. ( talk) 02:07, 29 March 2009 (UTC)
I object to the claim of "relevance" for the year linking articles; option 1 is something like "link relevant year articles", while option 2 would be "link relevant years". Since we're almost certainly not going reach agreement before the deadline, I think the subtitles need to go. — Arthur Rubin (talk) 02:16, 29 March 2009 (UTC)
(outdent) here's one more try - i hope you can agree that they're quite clear, legible, sensible, etc:
please bear in mind that these subtitles don't alter the substance of the options - they're just supposed to serve as "reminders" of which option is which, to make life easier for readers/voters. Sssoul ( talk) 15:37, 29 March 2009 (UTC)
I feel dumb: I can't for the life of me see the difference in these titles, which is bad, since they need to distinguish the options for voters. What is wrong with the current two titles? Tony (talk) 16:47, 29 March 2009 (UTC)
Ryan,
The transcluded voting page for Wikipedia:Date formatting and linking poll/Month-day responses should have the same “one vote” statement, shown here…
…as does Wikipedia:Date formatting and linking poll/Year-linking responses, should it not? If so, will you please add this this code:
:''Please indicate your support vote under '''ONE''' option, accompanied by a concise explanation for your choice. Your explanation is important in determining the community consensus.''
…immediately under this code:
===Month-day responses===
… here (unless, of course, this was intentional). In advance, thanks. Greg L ( talk) 02:22, 29 March 2009 (UTC)
P.S. Oh yes, and please also change the four occurrence of Views on
with I support
in order to harmonize everything.
Greg L (
talk) 02:28, 29 March 2009 (UTC)
Before I rush into anything here, I want some more opinions - please everybody comment on what you think's best. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:17, 29 March 2009 (UTC) The more I think about this, the more I believe "One vote per section" is a bad idea. It's highly likely that some of the community won't have a preference between a number of the proposals - They may for instance broadly agree with 1 and 4 so wish to support those and I can't see a good reason to exclude them. I'm not seeing a good reason at present for one vote per section. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:24, 29 March 2009 (UTC)
The first past the post voting system (i.e. one vote per question) is simple to explain, understand, and analyse. The outcome is the most popular candidate. That doesn't always apply in preferential voting. If the outcome is not going to be the most popular candidate, then all stakeholders must accept that before the RFC starts. Lightmouse ( talk) 12:29, 29 March 2009 (UTC)
One vote only is the only sensible thing. If not, the poll is simply badly constructed. A poll should be designed to make choices mutually exclusive. Questionnaires may allow for multiple "votes" but those are not made to pick a "winner". And I guess this is what we try to attain here. So, allowing multiple votes is a no-no (unless they are required to be associated by weights summing to one, in which case nobody would bother to vote at all).-- HJensen, talk 16:14, 29 March 2009 (UTC)
You guys keep promising that “cool‑beans technology is just around the corner that will make god-damned gold dust pour out of every user’s USB port” and the community still rejects what you guys are selling. The “other” side here clearly knows what will likely happen if there is one vote per issue per editor: participants in the poll will vote for the best ones (our stuff), and you’ll get your asses kicked. You fear another repeat of “making it easy to get a clear gauge of the community consensus.” So you will do anything and everything to game the system, including try to structure the voting so A) you guys could vote for all three options in opposition to the one you fear, (and do so with absurd “4”-level votes to stack vote counts if you could), and B) make a huge cluster-pooch of voting results so it takes a team of MENSA members to divine what the community consensus is.
There are already four options each on linking and they each mean something different. To help make an already-complex RfC as unambiguous as possible, voters can simply vote for the one they like best (*sound of audience gasp*). You know; like how people vote for politicians in an election (where there is more “overlap” than you can shake a stick at). It’s pretty unfortunate that all these Well… Duh! concepts are so hard to implement. As much as you want to jump up and down and scream about how one‑editor / one‑vote will revoke the Magna Carta, allow Tony and me to imprison innocent Wikipedians at Guantánamo without charges, and result in forced sterilizations of the poor, your arguments belie what you really want: to desperately seize on any trick in the book to game the system so the true community consensus on this issue is difficult to discern. We’d have to be insane to do as you ask, otherwise, we’d be at this forever. Greg L ( talk) 16:21, 29 March 2009 (UTC)
I see no problem with allowing multiple votes so long as editors provide a "first choice", "second choice" designation with each vote. Also, I must again strongly disagree with an approval voting system. Editors must be able to oppose options. Also, if you're going to use approval voting, you should use it for auto formatting as well... I very strongly disagree with the mischaracterizations of the prior poll results directly above by Greg and HJensen directly above. — Locke Cole • t • c 18:05, 29 March 2009 (UTC)
Strictly speaking, we want to determine consensus, not plurality. So the first-past-the-post is supposed to be inappropriate: we don't want to consider an outcome such as "45 votes for option A, 56 votes for option B, 51 votes for option C, and 48 votes for option D" to be consensus for option B. (I obtained these numbers by computer simulation of 200 people voting randomly with equal probability for each option.) On the other hand, it is quite unlikely that consensus by dictionary definition will very likely not be reached, and, if we aren't willing to implement a more complicated system such as elimination runoff or the Schulze method, first-past-the-post is a lesser evil than e.g. casting dice. After all, there is no such thing as a perfect voting system. -- A. di M. ( talk) 20:25, 29 March 2009 (UTC)
The sentence "even though it is occasionally celebrated on a different day due to Lent" should actually be "... due to the Holy Week". It is very common for 17 March to occur during Lent, but St Patrick's day isn't moved unless it occurs during the last week of Lent (the Holy Week), which is something very rare (it happened in 2008 and won't happen again until 2160, unless the rules for the date of Easter are changed meanwhile). -- A. di M. ( talk) 19:56, 29 March 2009 (UTC)
Ryan, at Month-day: Option #2, would you please replace the very last sentence in the paragraph that reads like this(?):
(even though it is occasionally celebrated on a different day due to Lent)
with this:
(even though it is occasionally celebrated on a different day due to Holy Week)
....I'm planning to archive this whole page. People have already said to me the hostility here is going to put them off participating. It's important we start off with a clean slate. Ryan PostlethwaiteSee the mess I've created or let's have banter 21:50, 29 March 2009 (UTC)
I am afraid that the format of this poll will not permit me, and those who agree with me (if any) to express their opinion. I therefore intend to dispute it, and present a FoF that it is unacceptable.
I thank Ryan for his efforts, and regret that they have been derailed by a successful effort to distort the results. Septentrionalis PMAnderson 23:08, 29 March 2009 (UTC)
:::While I haven't looked over this page's archives, I am somewhat confused. What about this poll do you feel is unacceptable? This poll is quite fair in identifying various viewpoints, I believe.
NuclearWarfare (
Talk) 00:54, 30 March 2009 (UTC)
Comment We all know that this poll is not perfect—far from it. However, it is better, more specific, and was worked on more than previous RfCs combined, and I think we need to appreciate that. Whatever the problems are, we must accept that this is the RfC to end it all. The community nor the editors who have debated and debated over date autoformatting and linking can take another one. All we can do now is wait for phase 1 to end. Dabomb87 ( talk) 03:19, 30 March 2009 (UTC)
Excuse me, guys above ... the instructions make it amply clear that comments are important in the interpretation of consensus. Everyone is invited to write comments after their choice (and to choose "Neutral" if they wish). There is even a separate comments subsection beneath each response section. There should be no doubt that the RfC provides lattitude for expressing individual feelings. At the same time, let's be practical: the community has to come to some kind of decision, and making it likely that the results will be melted treacle spattered all over the place is not practical. I think Ryan has come to a reasonable solution, and both camps had a lot of prep. time in which they were able to comment on the structure and shape. Tony (talk) 09:07, 30 March 2009 (UTC)
I recommend a quick trip to Wikipedia:NAM. Sincerely, GeorgeLouis ( talk)
Unfortunately, I started the poll an hour early - yup, BST started on Sunday and I've been caught out. I'm going to leave it open, because I've already made all the notifications and I would hope that 1 hour will make little difference. Apologies to all parties. Ryan PostlethwaiteSee the mess I've created or let's have banter 23:12, 29 March 2009 (UTC)
You may want to consider putting your "comment" box from the top of the page into an editnotice, so people are more likely to see it. Anomie ⚔ 23:17, 29 March 2009 (UTC)
There seem to be a series of choices missing, such as the one we used to have .. -- User:Docu
I think the main question that needs to be resolved is: who the target audience of autoformatting? If all remains as it is then we can foresee some very basic and obvious problems:
It is clear that we must move away from this current state of affairs, it does not benefit the majority. The two most appropriate solutions are the "magic word/set dates for all users" solution and the conflicting "remove all autoformatting" solution.
Which brings me to my main point: I don't see anyone supporting the use of autoformatting through linking.
It appears that options #1 and #4 of month/year linking are gaining consensus and there is much overlap between those two choices, in that #4 is in some ways the non-instruction creep version of #1. Yet, despite this growing consensus, wikilinking still seems to be a major aspect of people's reasoning to oppose autoformatting.
We need to disentangle these two issues and first of all have a vote on the deprecation of autoformatting through wikilinks. I think we will find consensus on that issue and it will mean people's responses to further polls on autoformatting will not be diluted with reference to "overlinking" etc. Sillyfolkboy ( talk) 18:35, 30 March 2009 (UTC)
I don't like overlinking any more than the next guy, but I really think that option 4 will lead to editors overreacting. We don't want to kill all year links, and if you have a special guideline for year links that says they can only be linked in X, Y, or Z situations, then people will think: "If year links have a special section, then that means they must be judged more strictly than any other links." I don't think that is what most people supporting the first proposal intend to support. I much prefer option one. Year links need to just quietly slide into the same mass of rules all other links abide by. If we overreact, then we will, I guarantee you, be back in a few months, after several edit wars, having another poll. Wrad ( talk) 17:11, 30 March 2009 (UTC)
First, I posted a notice of my suggestion here. What was the response by Ryan, the clerk overseeing this RfC? He responded fine by me. I was thinking of propsing something like that myself.
Then there is this exchange on Ryan’s talk page. He responded to you as follows: As it happens, Greg was right with the formatting - it's a poll and we need to be clear about that. … Sorry, I know that's not what you wanted to hear.
Then it was fully discussed here on the RfC talk page where many editors from all sides of the issue weighed in and the issue received a full and fair hearing.
Then Ryan, the clerk responsible for fairly overseeing and structuring this RfC, helped to restore the format to my suggested method [1].
Then, when you complained about the structure of the RfC here on the RfC talk page, Ryan responded to you as follows: Whilst I am disappointed that you dispute it, I thank you for your efforts with helping to create the poll. Ryan has shown spectacular patience and fairness in all of this. That little jewel of a response to your industrial‑strength whining comes about as close as any admin can get to “tough; go pound sand in your damned ear if you don’t like it.”
Finally, In my ANI against you for slapping {disputed tags} all over an ongoing RfC that had been thoroughly and fairly developed and properly supervised, you were advised there in the ANI as follows: I strongly urge all contributors who have been involved in this dispute to not make any further modification to the RFC page. None! You have had your chance to have a say in how the RFC is conducted. If you have concerns, please voice them on the talk page and let someone who is uninvolved make any changes deemed necessary. If you, PMAnderson, haven’t yet learned how to heed advise in the face of not always getting what you want, I suggest you start with that tidbit. Greg L ( talk) 20:27, 30 March 2009 (UTC)
“ | The more I think about this, the more I believe "One vote per section" is a bad idea. It's highly likely that some of the community won't have a preference between a number of the proposals - They may for instance broadly agree with 1 and 4 so wish to support those and I can't see a good reason to exclude them. I'm not seeing a good reason at present for one vote per section. Ryan PostlethwaiteSee the mess I've created or let's have banter 03:24, 29 March 2009 (UTC) | ” |
And cease with this “Greg is a witch who caused our crops to fail and our midwives to weep”-horsecrap. The villagers just might look at each other with that wide-eyed look of epiphany, shrug their shoulders, and decide to burn your hut down so they don’t have to listen to your “waaa-waaa.” Greg L ( talk) 02:22, 31 March 2009 (UTC)