Jump to content

Wikipedia:Manual of Style/Dates and numbers/Three proposals for change to MOSNUM

From Wikipedia, the free encyclopedia
This Request for Comment is now closed. Please do not modify the archive below and instead participate in discussions at WT:MOSNUM. Thank you!
Note to readers: There are two active requests for comment. Please also see:



(moved from WT:MOSNUM)

Three issues have been causing disruption on this talk page and on MOSNUM itself, and it is high time they were dealt with more formally. Please indicate support or oppose beneath each proposal, followed by a brief comment if you wish, and your signature. Please note that any proposal to allow the linking/autoformatting of only dates of birth and death would require a change to the following requirement in MOSNUM, and is not under consideration in this RfC: "Dates in article body text should all have the same format." Tony (talk) 13:50, 23 November 2008 (UTC)[reply]

Proposal 1: A return to the linking of dates and date fragments

[edit]

That the following text in MOSNUM:

  • Linking: Dates (years, months, day and month, full dates) should not be linked, unless there is a reason to do so.

be changed to:

  • Linking: Years, months, days/months, and full dates should normally be linked.
Besides the fact that date autoformatting makes little affect on readability (the blueing of dates actually decreases the readability) and currently doesn't work for the majority of readers, these edit wars supporters of date autoformatting speak of are yet to be seen (and have little basis in any asserted facts). WP:ENGVAR has worked exceptionally well for things like color/colour, I see no reason that it shouldn't work that well for date formats. The fact that MOSNUM says that "If an article has been stable in a given style, it should not be converted without a style-independent reason" should help to prevent such wars. Dabomb87 (talk) 15:57, 30 November 2008 (UTC)[reply]
JimCubb (talk) 19:50, 2 December 2008 (UTC)[reply]

Proposal 2: A return to date autoformatting

[edit]

That the following text in MOSNUM:

Autoformatting: Dates should not be linked purely for the purpose of autoformatting (even though in the past this was considered desirable).

be changed to:

Autoformatting: Dates (containing either day, month and year, or day and month) should normally be autoformatted.

  • Oppose Autoformatting only achieves its own objective for a tiny proportion of editors and it causes lots of problems. Lightmouse (talk) 15:01, 23 November 2008 (UTC)[reply]
  • Oppose for all of the reasons that have been put forward over the past two years.Tony (talk) 15:17, 23 November 2008 (UTC)[reply]
  • Oppose Autoformatting is an attempted technical solution to a behavioural problem. We have since developed behavioural solutions, like ENGVAR, to this problem: live and let live. For its technical problems, see WP:Autoformatting. Septentrionalis PMAnderson 17:11, 23 November 2008 (UTC)[reply]
  • Oppose. Overkill solution for very small benefits. Pcap ping 18:10, 23 November 2008 (UTC)[reply]
  • Oppose per WP:CONTEXT. No reason to link dates. --John (talk) 18:17, 23 November 2008 (UTC)[reply]
  • Oppose Autoformatting benefits only A) registered editors, who have B) set their user preference. Autoformatting does not benefit regular I.P. users. Furthermore, the autoformatting masks serious shortcomings with date formats in articles; they can be inconsistent and inappropriate in the extreme but the editors who would care most (those who would set their user prefs) can’t even see that these problems exists! We editors see *pretty* proper dates but I.P. uses see “Japan attacked Pearl Harbor on 1941-12-07.” Autoformatting was a brain-damaged proposal from the beginning and should never have been ever made. For editors who might not fully understand what autoformatting has been doing to our I.P. readers, see WP:Why dates should not be linked. Greg L (talk) 19:37, 23 November 2008 (UTC)[reply]
  • Support why should we disadvantage those who choose to customise the interface by selecting a date format for their convenience. The software should be modified to handle non-logged in users and to present a unified date format for these uses. This would prevent edit wars over date formatting and all of the needless changes to date formatting as all full dates would be linked and presented to user in a consistent way no-matter which format they are in the article. Keith D (talk) 19:55, 23 November 2008 (UTC)[reply]
  • Strong support. At the moment we have a mishmash of date formats all over the Wikipedia, none of which is consistent and not always consistent within articles. If autoformatting works, then I'll support it. If not, then we should link the damn dates; I completely fail to see what the problem is with them. — OwenBlacker (Talk) 20:05, 23 November 2008 (UTC)[reply]

Proposal 3: Automated/semi-automated compliance with any particular guideline requires consensus

[edit]

That the following text be inserted into MOSNUM above the See also section, under a new title:

==Automated and semi-automated compliance==

The use of an automatic or semi-automatic process to bring article text into compliance with any particular guideline in the Manual of Style (dates and numbers) requires separate and prior consensus at [[WT:Manual of Style (dates and numbers)|the talk page]].

  • Oppose Bots and semi-automated processes are very useful to save effort, particularly for MOS janitorial edits. Lightmouse (talk) 15:01, 23 November 2008 (UTC)[reply]
  • Oppose because the use of bots/scripts to spare editors at large the labour of manual clean-ups to comply with the style guides is well-established and of significant benefit to the project. WP requires a massive amount of ongoing house-cleaning that is performed superbly well by bots/scripts, despite the small number of false positives. Requiring special consensus here to cover activity WRT specific guidelines would be impossibly cumbersome and bureaucratic, and unnecessary given that there's already a process for bots. Tony (talk) 15:15, 23 November 2008 (UTC)[reply]
  • Strongly support for automatic tools; support for semi-automatic tools. Editing text without understanding it, as bots do, is a recipe for doing harm to Wikipedia; our current disruptive bot has changed book titles; changed dates in quotations, and User talk:Lightmouse#Error in prose introduced by Lightbot. A human editor could have avoided all of these with the slightest degree of attention; a bot cannot do so unless it can recognize all the ways we indicate titles and quotations. Septentrionalis PMAnderson 17:11, 23 November 2008 (UTC)[reply]
  • Oppose. Why would this be special? There is a process in place for approving bots already. Pcap ping 18:10, 23 November 2008 (UTC)[reply]
  • Oppose Per Pohta ce-am pohtit just above. --John (talk) 18:17, 23 November 2008 (UTC)[reply]
  • Strongly Oppose What the hell? This isn’t the Soviet Union, where comedians had to literally run their material by the Thought Police before doing a standup comedy routine. Editors who run bots (in order to be prolific and productive) only need to ensure that their bot activity is compliant with MOSNUM guidelines. They absolutely do not need to get permission from other editors before they can be permitted to modify their bots to do this or that. Such an absurd policy would do nothing more than create yet another step of wikilawering to drag things out until editors just want to puke. If we can obtain a true and proper consensus on a guideline, then that’s the end of the story. Greg L (talk) 19:48, 23 November 2008 (UTC)[reply]
  • Oppose as per Greg L — OwenBlacker (Talk) 20:05, 23 November 2008 (UTC)[reply]
  • Oppose per Tony and Greg L. DOUBLEBLUE (talk) 21:29, 23 November 2008 (UTC)[reply]
  • Support, despite opposing the two above propositions. Bots should not be used to force through anything controversial, even when I agree with the actual change. --Apoc2400 (talk) 23:33, 23 November 2008 (UTC)[reply]
  • Oppose for semi-auto bots/scripts - semi-auto bots bring articles into line with all other aspects of the MOS, so why not this one. Fully-automatic bots might not be able to tell varieties of WP:ENGVAR, and so should be used which much caution. We can't get consensus here, so requesting consensus to semi-automatically correct 3 dates in an article is a waste of everybody's time, making it quicker to do them by hand. Let the scripts be bold, they can always be reverted if they make a genuine mistake.—MDCollins 23:43, 23 November 2008 (UTC)[reply]
  • Support - the unleashing of a bot to unlink dates before consensus had been reached at the previous RFC was unacceptable (it also had a misleading description: audits do not change, they count). dramatic (talk) 01:07, 24 November 2008 (UTC)[reply]
You mean that you think that it is fine to use a bot to jump the gun and run a bot to implement one possible outcome of a debate that hasn't been decidded yet? I think not! To all the others defending bots - this proposal doesn't say that they shouldn't be used for MOS tasks. Simply that there shpould be some clear finality on a policy before the bots should be unleashed. dramatic (talk) 08:03, 24 November 2008 (UTC)[reply]
The bots already have enough controls, and they were only editing articles to comply with MOS. Dabomb87 (talk) 17:16, 24 November 2008 (UTC)[reply]
No, the last time this was debated (September?) a bot was already stipping dates while the debate was still quite young and far from consensus (that's how I found the debate, following a link in a bot edit summary). dramatic (talk) 23:51, 3 December 2008 (UTC)[reply]
LeadSongDog: it refers to the use of bots. I hope that enables you to declare now. Tony (talk) 14:41, 24 November 2008 (UTC)[reply]
Oppose provided that this interpretation is captured clearly. The fuss wasn't about using unauthorized bots, but about using them to implement a questionable element of a guideline, in that it was lacking the clear concensus that this RFC's proposals seek to establish.LeadSongDog (talk) 16:52, 24 November 2008 (UTC)[reply]
  • Oppose and question purpose of this proposal. I don't get it — what is the supposed benefit of this? —Josiah Rowe (talkcontribs) 04:55, 24 November 2008 (UTC)[reply]
  • Oppose: Bots are already subject to stringent requirements of the WP:BAG, so further consensus from here is just bureaucracy. Delinking is drudge work, and semi-automated processes bring greater consistency save considerable time and effort. Ohconfucius (talk) 05:04, 24 November 2008 (UTC)[reply]
  • Oppose per above. G.A.Stalk 06:26, 24 November 2008 (UTC)[reply]
  • Oppose for reasons already given. Sssoul (talk) 06:42, 24 November 2008 (UTC)[reply]
  • Oppose There is no reason to prohibit scripts and bots from doing what would be acceptable if done with a manual edit. That is the whole purpose of having these tools.--User:2008Olympianchitchatseemywork 07:17, 24 November 2008 (UTC)[reply]
  • Srong Support, as there was no consensus for the "guideline" wording, there should be no consensus for bots. Furthermore, there is no way for a bot to determine whether a change should be made, as even Tony's false consensus says dates should "usually" not be linked. The determination of unusual conditions cannot be made by a bot. — Arthur Rubin (talk) 09:54, 24 November 2008 (UTC)[reply]
  • Oppose: there should be and are processes for approving bots and guidelines about the use of scripts and so on. While there should be such regulation, there is absolutely no reason to put it on the MOSNUM page, and anyway no reason why consensus needs to be gained separately for putting into practice a rule that has already gained consensus once. This would lead to disruptive churning of the sort that we have suffered almost to the limits of tolerance over the date linking thing.--Kotniski (talk) 11:13, 24 November 2008 (UTC)[reply]
  • Oppose. I think it is not helpful to single out bots and tools and ban them. However I think all such bots and semi-automated scripts should be thoroughly tested, and then only run under the tight supervision. Ruslik (talk) 13:47, 24 November 2008 (UTC)[reply]
  • Oppose. Karanacs (talk) 14:52, 24 November 2008 (UTC)[reply]
  • Oppose. If the community deems that articles are better without these links, then they should be removed with all available means. If someone misuses an automated tool, take it up with that user. -- Jao (talk) 15:10, 24 November 2008 (UTC)[reply]
  • Oppose. To the extent that this is about Lightbot, the bot employed by Lightmouse to do useful work, I am very impressed with how responsive Lightmouse is to complaints. More generally, as others have pointed out above, the use of bots to do the drudge work is a boon and not a problem.--Goodmorningworld (talk) 15:30, 24 November 2008 (UTC)[reply]
  • Oppose for semi-automatic edits. Neutral for fully automatic bots. I'm skeptical of the ability of fully automatic bots to be properly programmed, except in very narrow situations. --Gerry Ashton (talk) 16:43, 24 November 2008 (UTC)[reply]
  • If consensus for removing date links is achieved above, I'm happy for the bot approval group to be responsible for the testing and approval of relevant automatic tools. Agree that the higher the level of human intervention on these edits, the higher the accuracy of the edits; accuracy is to be demanded over volume. Knepflerle (talk) 17:05, 24 November 2008 (UTC)[reply]
    • That is to say, I oppose the creation of a separate approval process and instead support close cooperation - timely and clear notification of any MOS-relevant discussions at the bot-approval group to be made here, and notification to the bot-approval group of any relevant discussions on MOS automation that occur here. Knepflerle (talk) 17:10, 24 November 2008 (UTC)[reply]
  • Oppose. There's nothing wrong with the existing bot approval process. HiDrNick! 17:23, 24 November 2008 (UTC)[reply]
  • Support Some of the bots are very annoying.--Toddy1 (talk) 18:24, 24 November 2008 (UTC)[reply]
  • Oppose. Pointless wikiwonkery. --Malleus Fatuorum 20:11, 24 November 2008 (UTC)[reply]
  • Oppose, per Tony1, Greg L, et al. It Is Me Here t / c 20:40, 24 November 2008 (UTC)[reply]
  • Oppose, there's no reason for this to even be raised on style pages. However, it should be expected that in the bot approval process or after approval that bot operators and reviewers are aware of what may be disputed sections of MOS guidelines and thus not initiate or halt bot operations until such disputes are over. This might suggest that talk pages of MOS should have a permalist of bots that act upon it and what they are doing such that if a section is disputed, the bot operator can be notified. (this is not a style guide issue, more a general bot operation one) --MASEM
  • Oppose And what is wrong with the current process. If you feel that a bot's rights should be revoked, take it up with them. - NuclearWarfare contact meMy work 23:50, 24 November 2008 (UTC)[reply]
  • support. At present a bot's Talk page is is the property of the bot's operator, who is therefore free to remove complaints. Some form of public accountability is needed. In addition any professional software operation agrees requirements, tests that the requirments are met and only then goes into production. --Philcha (talk) 23:58, 24 November 2008 (UTC)[reply]
    • If you have a problem with a bot operator it can be taken to BAG and the BON, there is a form of public accountability through the approvals group and noticeboard. At BRFA vots are tested on requirements, code is generally checked over by other programmers, and then when everything is deemed okay processes can begin. §hep¡Talk to me! 01:00, 29 November 2008 (UTC)[reply]
  • Oppose: That sort of bureaucracy nearly cost us the war! Ryan4314 (talk) 00:56, 25 November 2008 (UTC)[reply]
  • Oppose: Unworkable. There is already an enormous amount of business to transact at these MOS talkpages. The process of deliberation is already seriously compromised. Don't add to the chaos.–¡ɐɔıʇǝoNoetica!T01:43, 25 November 2008 (UTC)[reply]
  • oppose: such links adds nothing but clutter to articles and serves no useful user purposee. Any such links that currently exist should be removed by any and all means, manually or in bulk, the faster the better. There should be more bots like Lightbot. There should be more coders, like Lightmouse, who does outstanding and commendable work, under very trying circumstances. Hmains (talk) 04:11, 25 November 2008 (UTC)[reply]
  • Oppose, we have guidelines to centralise such discussions, any such requirement would be unworkable. Tim Vickers (talk) 04:39, 25 November 2008 (UTC)[reply]
  • Oppose Requests for Bot tasks should have the final say on this issue. Not a "MOS talk" issue. Sillyfolkboy (talk) 17:41, 25 November 2008 (UTC)[reply]
  • Oppose - I see no reason not to use properly approved processes, and no reason they need to be approved twice. Ealdgyth - Talk 22:15, 25 November 2008 (UTC)[reply]
  • Oppose: No reason to make an exception. There's already the BRFA process, which is plenty enough.Headbomb {ταλκκοντριβςWP Physics} 01:57, 26 November 2008 (UTC)[reply]
  • Oppose this specific vetting option. I see little value in forcing every case of compliance cleanup to be run through the editors here. However, in the case of large-scale, wiki-wide changes (such as automated or semi-automated mass bot delinkings), the potential for large-scale disruption calls for some general community approval (perhaps at WP:VPP) would seem advisable and desirable – especially given recent events here. Askari Mark (Talk) 03:16, 26 November 2008 (UTC)[reply]
  • Oppose there is a whole lot of work to be done cleaning up this date-linking mess. It would be ideal if it were done in a timely manner. If editors disagree on what a bot has done, then let it be discussed on talk. There's nothing special about this clean up job. JIMp talk·cont 10:19, 26 November 2008 (UTC)[reply]
  • Oppose; we have limited resources here, and the benefits of using automated tools (time-saving, consistency) far outweigh the very occasional mistakes that are made (and which are quickly brought to light and fixed manually). We don't need yet another approvals process either. Steve TC 10:52, 26 November 2008 (UTC)[reply]
  • Oppose Bots and automated edits do huge good all over the encyclopedia. That they occasionally screw-up is a truism and unfortunate, but hugely outweighed by the good. The key is to feed back to the processes, to get scripts and usage tightened up, and where that's not possible to stop that individual script or user. This proposal is nasal trimming. --Dweller (talk) 12:06, 26 November 2008 (UTC)[reply]
  • Oppose. We do want to comply with our own MOS, don't we?  Sandstein  14:36, 26 November 2008 (UTC)[reply]
  • Oppose I see it as an unnecessary bureaucratic hoop; if a bot goes off and makes changes that are against consenus, then a bot can fix 'em. That this is beyond the skillset of most editors isn't important; we have far and away enough bot maintainers to do this. Bots shouldn't be an exception to normal processes in this particular regard. Studerby (talk) 19:47, 26 November 2008 (UTC)[reply]
  • Oppose. Per Tony, Greg L et al. Ben MacDui 20:01, 26 November 2008 (UTC)[reply]
  • Oppose. Nobody should make mass edits based on MOS rules that are constantly changing, but that's common sense, and the proposed wording is way too strong. --Hans Adler (talk) 23:12, 26 November 2008 (UTC)[reply]
  • Oppose. For both automatic and semi-automatic, per the reasoning used in MDCollins' last sentence. --RexxS (talk) 05:06, 28 November 2008 (UTC)[reply]
  • Oppose Someone needs to do the work, and it's a bit too much to expect editors to do it manually. If a bot screws up (i.e. contains bugs), it can be cleaned up after, but if it's doing legitimate cleanup work and making Wikipedia more useful, I see no harm at all. Orderinchaos 15:46, 28 November 2008 (UTC)[reply]
  • Oppose Sounds like a hurdle designed with the single purpose of delaying and hindering implementation of MOS. Either editors do it or bots do it. The rules apply equally to both, and bots do dull jobs a lot faster. --Escape Orbit (Talk) 16:08, 28 November 2008 (UTC)[reply]
  • Oppose: the same process should appliy to bots and users; a separate bot consensus destroys the bots' utility. Coemgenus 17:11, 28 November 2008 (UTC)[reply]
  • Oppose per Tony, Greg L etc. Pfainuk talk 18:20, 28 November 2008 (UTC)[reply]
  • Oppose: There's already a mechanism in place to rein in bots that aren't doing the right thing. Binksternet (talk) 18:21, 28 November 2008 (UTC)[reply]
  • Strong Oppose 21stCenturyGreenstuff (talk) 21:05, 28 November 2008 (UTC)[reply]
  • Oppose Per Greg L above Joe Nutter 21:22, 28 November 2008 (UTC)[reply]
  • Oppose If a bot is properly constructed and tested according to the rules for bots, there is no need for further consensus on whether it is desirable that articles comply with the MOS. Poltair (talk) 22:47, 28 November 2008 (UTC)[reply]
  • Support Bots are totally indiscriminate. G-Man ? 23:14, 28 November 2008 (UTC)[reply]
  • Oppose. --Srleffler (talk) 23:22, 28 November 2008 (UTC)[reply]
  • Oppose We already have BRFA, there is no need to make botops jump through so many hoops. Approved means approved, there should be only one venue for bots to be approved and we have that at BRFA. §hep¡Talk to me! 00:57, 29 November 2008 (UTC)[reply]
  • Support for automatic tools, neutral for semi-automatic tools (that require human intervention). Bwrs (talk) 02:36, 29 November 2008 (UTC)[reply]
  • Strong oppose - I don't have to get approval to make these changes manually, so why should anyone have to get approval to do it with AWB or a bot? لennavecia 02:48, 29 November 2008 (UTC)[reply]
  • Support There is a three-revert rule. This is in place to prevent excess load being placed on the servers. Having bots fighting over style will put worse load on the servers. I think letting bots do this would be a big mistake for performance reasons. -BarkerJr (talk) 04:43, 29 November 2008 (UTC)[reply]
    • Um... 3RR is to prevent disruption not for WP:PERFORMANCE reasons, which we are explicitly told not to worry about. There are numerous bots that put more load on the servers than a bot altering date formatting. Assuming this RFC defines what the format should be then the bots won't be "fighting" as any bot that does the opposite will be blocked for breaking a style-guideline that was formed with an overwhelming consensus. ~ User:Ameliorate! (with the !) (talk) 05:41, 29 November 2008 (UTC)[reply]
  • Oppose per Phil Bridger's statement "We should be using the most efficient methods possible to update Wikipedia. If a task can be done by a bot then this frees up human editors' time for other tasks." RainbowOfLight Talk 05:46, 29 November 2008 (UTC)[reply]
  • Oppose - use of an automatic or semi-automatic process ... requires separate and prior consensus at [MOSNUM/talk] is unnecessary bureaucracy. The bot approval process is sufficient. -- Fullstop (talk) 09:21, 29 November 2008 (UTC)[reply]
  • Oppose per earlier comments. Manxruler (talk) 09:27, 29 November 2008 (UTC)[reply]
  • Oppose per above, the bot is usually the best bet and any mistakes can be handled by humans quickly. --Banime (talk) 14:53, 29 November 2008 (UTC)[reply]
  • Neutral The bot and semi-automated scripts have caused a fair number of problems, some due to the underlying data, others due to people not paying attention and some to mistakes in the code. In this particular case, removing the linking did something specific and unfortunate, it removed metadata about the date which cant easily be replaced and is at this point, already gone for a large number of articles. dm (talk) 16:46, 29 November 2008 (UTC)[reply]
  • Oppose: the existance of a guideline surely represents consensus for its enforcement, whether automatic or manual. Of course, bots should go through the WP:BRFA to ensure they aren't going to be inadvertantly damaging. Martin (Smith609 – Talk) 21:23, 29 November 2008 (UTC)[reply]
  • Oppose That's what Wikipedia:Bot Approvals Group and Wikipedia:Bots/Requests for approval are for right? No need for redundancy. --Patar knight - chat/contributions 23:47, 29 November 2008 (UTC)[reply]
  • Weak Support - BAG is not a substitute for community consensus. BAG mainly evaluates bots for technical suitability. The last thing we need is more controversial bots running around because no one spoke up on the BRFA about potential problems. (BAG member) Mr.Z-man 00:55, 30 November 2008 (UTC)[reply]
  • Oppose per Greg L. Mentioning relevant bot approval requests on MOS talk would be polite, though. -- Avenue (talk) 04:18, 30 November 2008 (UTC)[reply]
  • Support: The use of bots in this area has lead to cold and careless edits to articles. This is being made into an argument about bots in general, when it is not. There should be a human touch to the process before allowing the bot to indiscriminately plough into an article in relation to the wording of the proposal. No one is asking for a go ahead every time a bot in employed. There must be some oversight in areas such as this where bots have done some damage. I am again disheartened by good editors basing their oppose on per and as above. Nja247 (talkcontribs) 12:21, 30 November 2008 (UTC)[reply]
  • Oppose Whenever a bot can do boring MOS-related jobs, let them do it by allo means! Then editors can focus on contents.--HJensen, talk 22:18, 30 November 2008 (UTC)[reply]
  • Support because it would address legitimate concerns people have about these bots, although I would prefer that only an announcement be made (with a link to BAG) to the MoS talk page. Wronkiew (talk) 23:17, 30 November 2008 (UTC)[reply]
  • Strongest possible oppose - Why should every single bot run that brings articles in compliance with our style guidelines require prior approval from the MOSNUM talk page? That's ridiculous. Date-linking removal in particular is a perfect task for a bot, as it automatically handles a boring job without affecting an article's content. Giants2008 (17-14) 01:51, 1 December 2008 (UTC)[reply]
  • Bad question. Not well enough defined. "What is consensus" does not have consensus (help at Wikipedia:What is consensus?), and this proposal would only create roadblocks, and is not the wiki way. --SmokeyJoe (talk) 01:55, 1 December 2008 (UTC)[reply]
  • Oppose Bots do lots of useful work and help created a harmonious look and feel to wikipedia. Gaius Cornelius (talk) 15:45, 1 December 2008 (UTC)[reply]
  • Oppose Simply not required. Hohum (talk) 16:55, 1 December 2008 (UTC)[reply]
  • Support: Automated formatting brings helpful consistency to Wikipedia, and releases editors from the need to find the relevant Wikipedia formatting policy out of the crowd. Libcub (talk) 17:02, 1 December 2008 (UTC)[reply]
  • Oppose Overly bureaucratic. Problems do occur occasionally, but they are generally resolved without too much pain. This would eliminate a lot of good work to stop a small amount of problems. --Merlinme (talk) 18:02, 1 December 2008 (UTC)[reply]
  • Support, conditional: I'd unequivocally support it if clarified as "The use of an automatic or semi-automatic process to bring article text into compliance with any particular guideline in the Manual of Style (dates and numbers) requires clear consensus at the talk page, if there is a reasonable possibility that an ambiguous interpretation is being used." (For the record, I don't think such a guideline would affect the debate at hand.) arimareiji (talk) 18:29, 1 December 2008 (UTC)[reply]
  • Strongly oppose I strongly oppose having a special rule in place for this single issue. We have perfectly good rules for bots and don't need a separate process just for this one issue. I particularly think that the idea of requiring a special process for semi-automated (that is, with a live editor doing and reviewing the work) is stupid. Editors should be allowed to edit. They should not have to get special permission from a special group of people to do so efficiently. I specifically oppose this because the people originally pushing for this were doing so largely because they were on the losing side of the consensus, and their solution seems to be "Well, you can change the guidelines to support de-linking inappropriately linked dates -- but if you actually edit articles to make them conform with the new guidelines, then we're going to pitch a fit." (See for example Wikipedia:Requests for comment/Tony1.) Looking over these pages, there's a very strong consensus in favor of only selective, contextually appropriate linking of dates. This issue has been discussed for what, two years now? I've exceeded my patience with this frivolous and vexatious use of process to delay implementation of the consensus. Can we be done with this issue now? WhatamIdoing (talk) 01:20, 2 December 2008 (UTC)[reply]
  • Support These are style guidelines. Writing style is too complex to be “generated” by computer software. Running a search-and-replace operation is not “writing”. Michael Z. 2008-12-02 01:22 z
  • Oppose - a different rule for bots compared to manual editors makes no sense. Peter Ballard (talk) 01:44, 2 December 2008 (UTC)[reply]
  • Oppose - we already have a bot approval process. And most talk pages get ignored. Questions about content can sit there for days/weeks/months unnoticed. Then if the dates get changed, the same people that have been ignoring the talk page, will say something about the dates being changed "all of a sudden". Just do it. Dismas|(talk) 08:39, 2 December 2008 (UTC)[reply]
  • Support maybe stronger and vaguer than I would have worded it and seems worded for failure. Broad consensus should be required and that seems to be missing from the existing process. -- billinghurst (talk) 09:34, 2 December 2008 (UTC)[reply]
  • Oppose: Enforcement of accepted guidelines should be encouraged by all means possible, it makes no difference whether it is done by a bot or a human. — Emil J. 10:59, 2 December 2008 (UTC)[reply]
  • Oppose as per Greg L - there is a difference between a controversial change (which is continuing doing something against the stated wishes of other editors or against standard policy) and getting consensus (which is an extra step that is requested here for any change). We should not add such a requirement. Ingolfson (talk) 11:45, 2 December 2008 (UTC)[reply]
  • Oppose - if the edits are controversial, the problem is that too few editors are seeing the MoS discussions. Adding an extra step to the discussions will not address this problem, and will inconvenience editors who wish to use bots to make changes which have consensus. Warofdreams talk 12:11, 2 December 2008 (UTC)[reply]
  • Oppose It's a waste of human energy to enforce this guideline, and the bot makes fewer mistakes anyways. Calliopejen1 (talk) 16:26, 2 December 2008 (UTC)[reply]
  • Oppose. « Hiram111ΔTalK Δ 02:25, 3 December 2008 (UTC)[reply]
  • Oppose. Without the help of bots, this problem will never go away. Vegaswikian (talk) 04:00, 3 December 2008 (UTC)[reply]
  • Oppose. Requiring consensus before a bot is allowed to do useful work is counterproductive. It is important, though, that bot errors, when they occur, must be quickly corrected. Truthanado (talk) 04:09, 3 December 2008 (UTC)[reply]
  • Support. In response to WarofDream's point, what is happening is that individuals have found that an easy way to circumvent the process of creating a consensus on a matter is to make an edit to the MOS, wait a day or two, then claim they have the support of this document & unleash a bot to make the changes. By the time anyone who disagrees with them notices what has been done, the issue has become moot & they fight an uphill battle trying to effect a comprehensive reversion of what has been done. And I don't want to need to monitor the MoS to prevent this kind of bad-faith activity: I want to contribute content following common sense, guidelines like the MLA Handbook, & only consult the MoS when I am unsure about how to handle a specific problem. -- llywrch (talk) 21:10, 3 December 2008 (UTC)[reply]
  • Oppose Ouch! Adding extra steps beyond the processes already in place seems to serve only to add more stress and angst. If clever vandals are disrupting we can sort that out. If some mad bot activity is taking place we can address that as well. -- Banjeboi 22:00, 3 December 2008 (UTC)[reply]
  • Oppose MOSNUM compliance editing is the definition of tedious and repetitive work. It is EXACTLY what bots are designed to do. That a tiny minority of editors objects to aspects of MOSNUM does not make those aspects of MOSNUM controversial, the overwhelming consensus above indicates this, and seeing as these edits are tedious, repetitive, and non-controversial, I see no reason to place extra restrictions on bot-editing beyond the standard bot approval process. --Jayron32.talk.contribs 03:41, 4 December 2008 (UTC)[reply]
  • Oppose. Isn't the purpose of using (semi)-automated processes to avoid unnecessary lapses of time? Mandating approvals will only delay the process of (semi)-automated editing. Besides, there's already a separate approval process for bots.
  • Oppose – It would generate a workload that cannot be managed and slow things down to the speed of a very small snail. I consider the checks in place sufficient, and one must consider what is practicable, too. With over two-and-a-half million articles, we do not have the luxury of spending a week and two on each small change. Waltham, The Duke of 02:14, 6 December 2008 (UTC)[reply]
  • Strong support. This prevents vandalism of removing allegedly overlinked dates, when no human mind is used and all linked dated are simply delinked. Guy Peters TalkContributionsEdit counter 22:46, 6 December 2008 (UTC)[reply]
  • Oppose as long as the existing approval process ensures that bots do not change text which deliberately departs from guidelines. Certes (talk) 00:31, 7 December 2008 (UTC)[reply]
  • Oppose as per Tony and Greg L. Snappy (talk) 03:45, 7 December 2008 (UTC)[reply]
  • Strong Support per Arthur Rubin. Conflicts over this have caused massive amounts of edit warring and largescale bot changes across many articles. Beside that, however, bots are no substitute for human judgment. GlassCobra 11:35, 7 December 2008 (UTC)[reply]
  • Strong Oppose. Bots are supposed to do mundane tasks, like bring articles into conformity with guidelines. Slowing down this process doesn't make sense, considering that bots are not running out of control. Instead, it seems like a way to circumvent consensus: lost the battle regarding the guideline? fight on regarding its implimentation! RJC TalkContribs 16:07, 7 December 2008 (UTC)[reply]
  • Oppose per all above. Consensus ≠ Unanimous agreement. See question #1 above for demonstration of non-unanimous consensus. -- Quiddity (talk) 01:32, 8 December 2008 (UTC)[reply]
  • Strong support for the same sort of reasons given by Septentrionalis, dramatic, apoc and Arthur Rubin above. If a consensus develops to follow some particular format, editors will incorporate that into editing paragraphs and articles which they're already editing because they know something about the subject; they can therefore make the best judgements about relevancy and recognize something like a quotation or title when they see one. I already delink almost (but not quite) all dates in the paragraphs I run across. You can't argue with a Bot, even a relatively-benign one like Lightmousebot. —— Shakescene (talk) 04:43, 8 December 2008 (UTC)[reply]
  • Now (from this and other comments), I understand the distinction and am less puzzled. Thanks.
  • Although this is hardly a decisive argument, semi-automated editors can be careless. One benign, experienced and completely trustworthy human editor using Twinkle in pursuit of a spammer didn't see that he was wiping out two months' worth of work on New York City. —— Shakescene (talk) 02:02, 12 December 2008 (UTC)[reply]
  • Strong Oppose as this measure defies the very core of individual creativity, and would in the end destroy that most basic of human rights - The right to dissent. This would set a horrible precedent in my opinion.--Yachtsman1 (talk) 20:41, 8 December 2008 (UTC)[reply]
  • Support. I would prefer no bots were used on date formatting. Whatever the style guidelines end up as, they will involve a degree of judgement in individual cases, and so is inappropriate for for automated tools or bots. An example of a peculiar semi-automated edit which broke a redirect: [1].--Rumping (talk) 16:17, 10 December 2008 (UTC)[reply]
  • Oppose—ridiculous. If someone wants to use a semi-automated process to add non-breaking spaces between numbers and units of measurement, why should he require separate approval? Asking for consensus to adhere to a guideline? Should bots require separate approval to assume good faith as well? Pagrashtak 14:47, 9 December 2008 (UTC)[reply]
  • Somewhat support — Bots can easily make mistakes if they encounter an unexpected circumstance. I think some form of approval should be required for them. For semi-automated processes, I'm not sure. It may be fine to let them run without approval. —Celtic Minstrel (talkcontribs) 17:38, 9 December 2008 (UTC)[reply]
  • Oppose for semi-automated. So far the arguments for this have centered on bots encountering unexpected circumstances. Semi-automated scripts and tools such as AWB by definition require human intervention to work an should not cause such issues. Phatom87 (talk contribs) 16:22, 10 December 2008 (UTC)[reply]
  • Strong Oppose Bots keep pages in compliance without manual intervention. There should always be a way to tell a bot to ignore a page in the bot's coding for special cases. Timmccloud (talk) 23:07, 11 December 2008 (UTC)[reply]
  • Strongly support per the reasoning of Septentrionalis, GlassCobra, Arthur Rubin, and Llywrch. Semiautomatic edits (using AWB, for example) can be just as disruptive and dangerous as bots. For example, look at the thousands of articles recently edited by Lightmouse using AWB, often at the rate of 1 article every 4 seconds. See this discussion that was closed by an administrator (Reedy) who himself uses AWB to make rapid edits, an apparent conflict of interest. See also this discussion on Reedy's talk page. Tennis expert (talk) 06:01, 12 December 2008 (UTC)[reply]
  • Support - I see bots and "script-assisted" users making many errors as well as being unbelievably irritating in their smug belief that they are somehow "contributing" to this project just because they make lots and lots of unnecessary edits. For those of us who actually create articles, it's just another example of how wikipedia is fast losing sight of its original agenda. Deb (talk) 12:52, 12 December 2008 (UTC)[reply]
  • Strong Oppose. Editors using Bots do a great job in my opinion. We are already have an approval process for bots and guidelines about doing rapid edits using bots. It is great for the consistency of presntation of the encyclopaedia that there are people who are willing to do these tasks. They ought to be able to get on with bringing articles into line with the MOS. All edits, even from Bots, can be reverted. The worse that can happen is that there is a later consensus to revert - in which case the bot edits can be reverted by another bot! Viv Hamilton (talk) 15:22, 12 December 2008 (UTC)[reply]
Given that the MOS is merely a guideline and not a policy, why is it necessary or desirable to have a bot to "bring articles into line with the MOS"? I have never understood why people (or their bots) are allowed to force an article to comply with mere MOS recommendations in opposition to the longstanding consensus of the editors of that article. Tennis expert (talk) 22:01, 12 December 2008 (UTC)[reply]
I couldn't agree with you more. There are often reasons peculiar to a particular article, subject, field or set of data that make a format that varies from the MoS more suitable; quite often it's a rough but workable compromise after an article's editors tried other approaches (usually including those recommended by the MoS).
An extreme example, in my view, is AWB's flagging of "first and only" for redundancy, when there could be several good reasons to prefer it to "only" alone. [What a bot is good for is things like cleaning up misspellings (rather than variant spellings), although even those cases need human supervision for cases such as misspellings, antique spellings, foreign words, slang or historical punctuation in an original quotation or title.] Rage against the machine. —— Shakescene (talk) 22:54, 12 December 2008 (UTC)[reply]
Proof that the MOS is merely a guideline and not mandatory comes from the criteria for an article being classified as a "good article". Such an article has to comply with only a few aspects of the MOS. If the MOS were "mandatory", the article of course would have to comply with all aspects of the MOS. Therefore, it's bankrupt for MOS enforcers to argue that the MOS is mandatory without first actually making it mandatory. Tennis expert (talk) 05:13, 14 December 2008 (UTC)[reply]

Comments

[edit]
There is no reason to link September 11. The correct link would be September 11 attacks, which a bot should leave alone because teh date is not surrounded by brackets. Truthanado (talk) 04:12, 3 December 2008 (UTC)[reply]
  • I'm disturbed by the large number of Opposes I see which state some variant of "because bots are useful." I don't think that's the question being asked. arimareiji (talk) 18:29, 1 December 2008 (UTC)[reply]
    • Similarly, I'm disturbed by the large number of Supports I see that appear to be completely unaware of the existing bot-approval process. No bot is allowed to run on Wikipedia to do anything without explicit permission. The question here has two parts:
      1. Should (real, live, human, non-bot) editors have to get special permission from a special-interest group at WT"MOSNUM (that is, a special exception to the normal process) to use an ergonomically friendly script (such as Twinkle) to make changes that they want to make anyway
      2. Should making dates conform with MOSNUM by a (non-human) bot be considered so incredibly special that it requires not just the normal, extensive bot approval process, but also special permission from the special-interest group at WT:MOSNUM.
      Compare this "special rules" system, for example, to the way we handle vandal fighting and external link spamming: is undoing an error made by a date-link-changing bot dramatically harder than undoing an error made by ClueBot, or any of the many other bots that run everyday on Wikipedia? WhatamIdoing (talk) 07:19, 8 December 2008 (UTC)[reply]
  • Please remember that Proposal 3 also concerns semi-automatic edits also, which are supervised by humans. Dabomb87 (talk) 01:04, 2 December 2008 (UTC)[reply]
  • Comment I expect to see this revisited in a year or two when a less obtrusive way of doing these things is devised. I think the oppose comments reflect the overuse and the obtrusiveness, not the principle. DGG (talk) 18:11, 10 December 2008 (UTC)[reply]
  • Comment. The statement that the guideline "Dates in article body text should all have the same format" somehow eliminates the possibility of linking birth and death dates in biography articles is clearly incorrect. This guideline refers to date formatting, not to linking.
  • What's this deal about ignoring birth/death dates? This does not apply to consistency because it is only within the parentheses in the first sentence, not in the article body text. I don't like date links littered everywhere, but they should be linked at the very top where it's relevant. Reywas92Talk 21:31, 16 December 2008 (UTC)[reply]
  • That's an interesting new use of the word 'relevant' that I hadn't come across before. Charles Darwin, a featured article begins "Charles Robert Darwin (12 February 1809 – 19 April 1882)". I looked at 12 February. Among 48 KB of other stuff, it turns out he shares a birthday with Abraham Lincoln and Lolly Badcock - relevant?? I then looked at 1809. It seems Darwin was born the day after Robert Fulton patented the steamboat. Same nonsense at 19 April and 1882. If anybody suggested that articles should have non-date links to that sort of trivia, they'd be laughed at. If I want to see jokes I read Icanhascheezburger, not Wikipedia. --RexxS (talk) 03:29, 24 December 2008 (UTC)[reply]

Tony (talk) 13:50, 23 November 2008 (UTC)[reply]