User talk:Lightmouse: Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Content deleted Content added
Chunky Rice (talk | contribs)
→‎Question: further comment
Line 177: Line 177:
::::My question is simple. Is discretion being used in these scripted edits? It's not a trick question. I don't understand what you mean by "practical means." You read the sentence (or paragraph or whatever you need for context) and form an opinion. Is the link useful? Yes or no. It's not complicated. -[[User:Chunky Rice|Chunky Rice]] ([[User talk:Chunky Rice|talk]]) 20:43, 30 September 2008 (UTC)
::::My question is simple. Is discretion being used in these scripted edits? It's not a trick question. I don't understand what you mean by "practical means." You read the sentence (or paragraph or whatever you need for context) and form an opinion. Is the link useful? Yes or no. It's not complicated. -[[User:Chunky Rice|Chunky Rice]] ([[User talk:Chunky Rice|talk]]) 20:43, 30 September 2008 (UTC)
::::Oh, and obviously, a bot is incapable of making such a determination, so I'm not sure why I'd try to suggest a way that it could. -[[User:Chunky Rice|Chunky Rice]] ([[User talk:Chunky Rice|talk]]) 20:44, 30 September 2008 (UTC)
::::Oh, and obviously, a bot is incapable of making such a determination, so I'm not sure why I'd try to suggest a way that it could. -[[User:Chunky Rice|Chunky Rice]] ([[User talk:Chunky Rice|talk]]) 20:44, 30 September 2008 (UTC)

== Birth and death years and categories ==

Hi Lightmouse. Would you have time to read the two posts I made [http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style_(dates_and_numbers)&diff=242110485&oldid=242096444 here] and [http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style_(dates_and_numbers)&diff=242116388&oldid=242110485 here]? Part of the reason I think some people (or me, at least) objected to the removal of links to birth and death years (that is year links, not the month/day links which really are trivia) was due to the potential metadata implications. Birth and death year metadata currently resides in several places on biographical articles: [[Wikipedia:Persondata]], the birth and death categories (the largest set of data), infoboxes, and links from birth and death years. The trouble is that this is not completely consistent, and if I could program a bot to do a "biographical audit" (as you've been doing a "date audit"), then one of the things I would check is whether birth and death categories existed before removing birth and death year links. One of my questions was whether your bot can tell when it is editing a biographical article and whether or not the birth and death categories are present? Would you be able to analyse the contributions log of Lightbot to see how many biographical articles it affected? Or alternatively, do you think you could do any of the other (admittedly rather large) tasks I suggested in those posts? I wouldn't normally make a personal request like this, but I've tried bot requests several times, and sometimes people have taken this on, but with mixed results. If it's not your sort of thing, don't worry - I just thought I'd ask. [[User:Carcharoth|Carcharoth]] ([[User talk:Carcharoth|talk]]) 00:09, 1 October 2008 (UTC)

Revision as of 00:09, 1 October 2008

Why the mass delinking of years?

On Onomacritus all dates were delinked. I've read the approval but it is no clearer to me what the bot is aiming to do in cases like this.Dejvid (talk) 21:10, 22 September 2008 (UTC)[reply]

I hope I'm wrong, but it seems to be unlinking all years what do not have the day and month. In the classical period we often (indeed mostly) don't know the exact date. Polybius is not a stub yet every date has been unlinked. And he's a historian. What gives?Dejvid (talk) 21:45, 22 September 2008 (UTC)[reply]

Lightmouse, the date de-linking on UK parliamentary constituencies is worse than I highlighted above. The delinking has included "created" dates (the yeat the seat was first formed). This means that users who may want to see what else happened in that year can no longer do so easily. I can understand why the UK electiom years were delinked (as I say above, really they should have been links to the UK election articles), but when the year was explicity linked to a section involving years (rather than events in that year), a mistake has been made. Could you revisit the edits made to ALL UK parliament constituency articles to ensure this edit is undone in some way? doktorb wordsdeeds 23:05, 22 September 2008 (UTC)[reply]

By the same token, years should not be delinked in articles on British peerages. Please stop your bot; this was never clearly approved, and is undesirable. Septentrionalis PMAnderson 23:13, 22 September 2008 (UTC)[reply]

Wikipedia:Bots/Requests for approval/Lightbot 3 does not give your bot permission to unlink all linked years. The approval request is misleading at best, and possibly intentionally so, if that's what your bot was designed to do. Corvus cornixtalk 23:43, 22 September 2008 (UTC)[reply]

In case you weren't aware, there is currently an ANI discussion about your bot and the delinking of years. Wikipedia:Administrators'_noticeboard/Incidents#Mass_De-linking_of_years_by_Lightbot -Chunky Rice (talk) 23:57, 22 September 2008 (UTC)[reply]

And it almost goes without saying that you should probably not restart the bot for this function until this is resolved. -Chunky Rice (talk) 00:06, 23 September 2008 (UTC)[reply]
As I understand it, if MoS guidance is wrong, the place to debate it is wt:mosnum rather than ANI. Lightmouse (talk) 10:16, 26 September 2008 (UTC)[reply]
The issuse is not with the MoS, but rather your actions. Please stop trying to deflect a discussion about your actions to the MoS. -Chunky Rice (talk) 17:02, 26 September 2008 (UTC)[reply]

Moved to: Wikipedia_talk:Manual_of_Style_(dates_and_numbers)#Wikipedia:Administrators.27_noticeboard.2FIncidents.23Lightmouse_again. Regards Lightmouse (talk) 16:21, 26 September 2008 (UTC)[reply]

Hi, I see you deleted all the date links in this article. Fair enough, given the new rules (and most of the links were junk). However, you unlinked the word "Friday" under "The start". There is a giant note embedded in the article text, and another in the talk page, explaining why this instance of Friday should be linked, so please leave it, or else explain in the talk page why you think the rationale for linking it is invalid. Cheers, — Johan the Ghost seance 23:52, 25 September 2008 (UTC)[reply]

I find it hard to believe that anyone will click on a link to a day of a week. If Wikipedia had statistics on click-through rates, it would be very useful. However, I don't mind you adding it back and I appreciate you letting me know. Thanks. Lightmouse (talk) 09:35, 26 September 2008 (UTC)[reply]
I've taken this up on Johan's talk page. Tony (talk) 16:00, 26 September 2008 (UTC)[reply]

British Rail

Please can you explain the reason for this [1]. The date was in accordance with the Manual of Style[2], so I cannot see any reason for your bot to change it. Along with others, I find your bot to be doing things that it should not, and unless you can demonstrate that your bot is operating in compliance with the MOS, then I intend to report it. Olana North (talk) 08:33, 26 September 2008 (UTC)[reply]

I have added my grave concerns to the growing list here [3]. Hopefully, commonsense will prevail and you will get a temporary block. Olana North (talk) 08:41, 26 September 2008 (UTC)[reply]

You are debating the Manual of Style and it says:

  • Dates (years, months, day and month, full dates) should not be linked, unless there is a particular reason to do so
  • The linking of dates purely for the purpose of autoformatting is now deprecated.

Discussion of the meaning of those words is best undertaken at wt:mosnum. That way, you won't just get my opinion. Lightmouse (talk) 09:33, 26 September 2008 (UTC)[reply]

No, we are debating when a bot should be permitted to enforce the Manual of Style. The obvious rule here is that bots should only act when there is clear consensus that a rule should be applied universally, without exception; that is not the case here. Septentrionalis PMAnderson 18:29, 26 September 2008 (UTC)[reply]
Anderson, please point to the consensus for applying date links and/or date autoformatting in the first place. Tony (talk) 16:25, 27 September 2008 (UTC)[reply]
I wrote an essay summarizing the discussion that lead to date autoformatting. The linking of dates for the purpose of giving people additional information about what happened in that year or that day of the year was already commonplace back then (2003). --Gerry Ashton (talk) 16:46, 27 September 2008 (UTC)[reply]
Might have been commonplace, but where was the "consensus"? Tony (talk) 02:49, 28 September 2008 (UTC)[reply]

Element order of dates

Please explain how the bot determines, when it removes linking from dates and changes the order of the day, month, and year elements, how it decides whether the finished order should be month day year or day month year. --Gerry Ashton (talk) 15:54, 26 September 2008 (UTC)[reply]

The bot doesn't. Lightmouse (talk) 16:02, 26 September 2008 (UTC)[reply]

Sorry, Gerry, Lightmouse (the person) is very precise in his parsing of questions, so let me try to ask for the information you are seeking. How do you, Lightmouse, the person—whether through the account User:Lightmouse or the account User:Lightbot, determine when you remove linking from dates, and change the order of the day, month, and year elements, and how you decide whether the finished order should be month day year or day month year? — Bellhalla (talk) 21:09, 26 September 2008 (UTC)[reply]

Unlinking "AD" dates

The script seems to unlink "BC" dates (like 10 BC) correctly; could the same be done for 10 AD dates and such? Thanks! Gary King (talk) 18:40, 26 September 2008 (UTC)[reply]

I don't see what's wrong with linking of such dates for historical reasons. There seems little point in having articles for these years if they won't be easily found by the typical amateur historian/reader. But perhaps I am alone in such an opinion. --Candlewicke (Talk) 15:00, 29 September 2008 (UTC)[reply]

Wikipedia:Administrators'_noticeboard/Incidents#Lightmouse_again

Please refrain from continuing to unlink dates with AWB (or other scripts/automated software) prior to engaging the community in discussion regarding this matter. A large part of the reason why the bot was stopped was because the community has found this type of unilateral unlinking to be controversial. Using AWB as a substitute for your bot to perform these kinds of edits is not appropriate, and should be ceased until you are willing to discuss this matter, either here on your talk page or at Wikipedia:Administrators'_noticeboard/Incidents#Lightmouse_again. Thanks, Shereth 22:31, 25 September 2008 (UTC)[reply]

Whatever you decide to do, regarding your bot and delinking dates and whatever, it would be most helpful and appreciated if you would communicate with the other editors who have expressed concern about this issue. Right now, I think that your lack of communication is causing this to be a bigger deal than it probably should be. -Chunky Rice (talk) 23:21, 25 September 2008 (UTC)[reply]

You are debating the Manual of Style and it says:

  • Dates (years, months, day and month, full dates) should not be linked, unless there is a particular reason to do so
  • The linking of dates purely for the purpose of autoformatting is now deprecated.

Discussion of the meaning of those words is best undertaken at wt:mosnum. That way, you won't just get my opinion. Lightmouse (talk) 09:36, 26 September 2008 (UTC)[reply]

The concern is your method of enforcing the MOS, and not the MOS itself; therefore the ANI thread is valid. Again, I will ask you to refrain from continuing to engage in this method of enforcing the MOS until the concerns of multiple editors have been addressed. Continuing to do so prior to reaching a conclusion regarding this matter is disruptive and you must stop for the time being. I am not passing a judgment on your use of bots/scripts to perform date unlinking but I am stating that it is highly inappropriate to continue to do so prior to resolving the concerns over how you are going about it. It is disruptive, and if you continue to do so you run the risk of having your account temporarily blocked to prevent further disruptive editing while the discussion is ongoing. Thank you, Shereth 15:32, 26 September 2008 (UTC)[reply]

I do stop editing when some issues are raised and I did stop for a while in this case. However, it looks like the ANI discussion is being used for MoS talk. Wikipedia would grind to a halt if any editor can stop another editor on the basis that discussions are continuing. So I have decided to continue doing the same as other editors in acting in accordance with the MoS. I can't see a current question from you at the MoS. What would happen if we had this discussion at the MoS instead of on my talk page? Lightmouse (talk) 15:43, 26 September 2008 (UTC)[reply]

Sounds entirely reasonable to me. Tony (talk) 15:57, 26 September 2008 (UTC)[reply]

I don't mind if you want to start up a thread at MoS regarding the issue, but ultimately the method of enforcement is not within the remit of MoS (or its talk page) to determine. Some editors are discussing the possibility of an RFC to get a broad spectrum of community input on the matter, which, in my opinion, is the superior method. Shereth 16:07, 26 September 2008 (UTC)[reply]

In agreement with User:Shereth, I have brought this discussion here because the section and the related ANI thread have a lot of MoS related issues. Any thoughts? Lightmouse (talk) 16:20, 26 September 2008 (UTC)[reply]

My objections (noted at the ANI discussion) have nothing to do with the MOS issues noted above, but rather the seemingly arbitrary, inflexible approach that Lightmouse (here used to mean the person controlling User:Lightmouse and User:Lightbot) has used in enforcing his/her interpretations of those guidelines. — Bellhalla (talk) 16:31, 26 September 2008 (UTC)[reply]
"Arbitrary"—can you explain why the use of a human-supervised automation to spare editors the manual labour of complying with MoS is this?
"Inflexible"—Lightmouse has shown admirable flexibility in updating the bot/script in response to critical feedback. He is polite and sensitive in his operation of the bot. What is inflexible about this? I'm surprised you're not thanking him for his careful work in improving WP.
"Enforcing" sounds like a spin-word for "implementing", for that is, in effect, what Lightmouse has been doing. It's hard to see a sizeable lattitude for "interpretations of those guidelines": they're quite clear, and LM is doing us a favour. Tony (talk) 17:36, 26 September 2008 (UTC)[reply]
That's the heart of the problem, Tony. You see no latitude for interpretation of a guideline that in no way says all dates must be unlinked. Should many dates be unlinked? Probably. Can a bot or super incredibly fast (as Gerry Ashton noted at the ANI discussion, Lightmouse made 8 edits during the minute 22:25, 25 September 2008, for example) "human-supervised automation" make the determination of whether there's a "particular reason" for the date? No, that's a task left to editors who work on individual articles.
Nor does the guideline in no way say that all instances of "14:35 p.m." must be corrected, but I haven't heard you objecting to editors who do make those corrections. Tony (talk) 10:09, 27 September 2008 (UTC)[reply]
Further, as I read the MOS, what I see says that auto-formatted dates are "deprecated", which, to my knowledge, does not mean banned or prohibited. If you genuinely believe that there is consensus for banning all date links, be bold and rewrite MOSNUM so that it says that. And should that occur, you still need to remember that the MOS is still only a guideline not an iron-clad proscription. — Bellhalla (talk) 18:13, 26 September 2008 (UTC)[reply]
Please point me towards any consensus that all dates should be unlinked. As far as I can tell there is not one. Right now, the current consensus is that dates should not be linked unless there is a good reason to do so. This is the kind of distinction that can only be made by a human being, not a bot. -Chunky Rice (talk) 17:38, 26 September 2008 (UTC)[reply]
"Human-supervised automation"? The number of edits would seem to preclude individual examination of each article. So the supervision seems to consist of compiling a list of articles with certain keywords in the title and turning AWB loose on them, with Lightmouse deciding what date format is appropriate for every article with the keywords present in the title. Also, a question about whether the bot can prevent the alteration of dates within quotations has gone unanswered, which I take to mean, "no, it can't". Many people don't put much value on removing date links, so if we discount that benefit, what is the ratio of beneficial to erroneous edits? --Gerry Ashton (talk) 17:46, 26 September 2008 (UTC)[reply]
  • What Lightmouse is doing is correct, appropriate, and desperately needed. The MOSNUM policy stating that dates should not be linked was thoroughly debated for a long time and is the result of a properly-arrived-at consensus. Links should be strictly limited to topics that are topical and germane to the article. Links to rambling lists of mindless trivia are virtually never topical and germane and just clutter up articles with excess blue that anesthetizes the mind. There are simply far too many of these links in far too many articles for any human to possibly hunt them all down and correct them; automated tools are the only way.

    Jimbo himself posted the most important rule of all on Wikipedia: Wikipedia:Ignore all rules: “If a rule prevents you from improving or maintaining Wikipedia, ignore it.” Well, in this case, Lightmouse has followed all the rules in his effort to improve Wikipedia. He just shouldn’t have to put up with any more flack from people who flat disagree with MOSNUM, just love their links to trivia, and want to drag this out with even more debate; such views have been discredited and Wikipedia is now well on the road to improvement. Greg L (talk) 18:01, 26 September 2008 (UTC)[reply]

    • I don't disagree with MOSNUM. At least I don't think I do. I agree that most date links are not useful. I do not agree that date links are never useful or allowed. And as far as I can tell MOSNUM doesn't say that either. Do you think it says that? -Chunky Rice (talk) 18:06, 26 September 2008 (UTC)[reply]
    • You state that links should be strictly limited to topics that are topical and germane to the article. Does Lightbot (or any bot), or Lightmouse (or any user) making semi-automated edits at the rate of 4-5 per minute, have the ability to determine whether or not the links being removed are topical and germane to the article? I think not. Ultimately that is the problem; it is not that people are disagreeing with MOSNUM but that they are disagreeing with this method of enforcement thereof. Shereth 18:16, 26 September 2008 (UTC)[reply]
    • The suggestion that each and every date link should be individually scrutinised before deciding whether to unlink it would make sense if each and every date had been individually scrutinised to decide whether to link it in the first place. But of course they weren't - they were done unthinkingly en masse because that was what the MoS recommended at the time. It's entirely reasonable for Lightmouse and other editors - including myself - to unlink dates routinely. In those very few cases where there's an identifiable benefit to linking a date, it's a simple matter to re-link it. Date linking is deprecated (though permitted) by the Manual of Style. To complain that someone is changing articles to fit the MoS is simply refusing to accept that there has been a change of policy, and this persecution of LM over it is shameful. Colonies Chris (talk) 18:27, 26 September 2008 (UTC)[reply]
        • Using your own reasoning, Colonies Chris, why should we allow a bot continue to change articles that, by your own admission, are using a permitted style under the MOS? — Bellhalla (talk) 19:32, 26 September 2008 (UTC)[reply]
          • Because date linking is deprecated. That means there's a presumption against it. Anyone who thinks a date should be linked needs to make a case for linking it. The default position is 'unlinked'. Colonies Chris (talk) 21:21, 26 September 2008 (UTC)[reply]
            • But you yourself admitted that linked dates are permissible under the current MOS guidelines. I'm not meaning to be contentious, but I don't understand the inconsistency in your reasoning. — Bellhalla (talk) 21:44, 26 September 2008 (UTC)[reply]
              • There is no inconsistency. Linking dates is permissible, but it's deprecated. In almost all cases it's a bad idea, I'm not saying that no-one should be allowed to link dates, I'm saying that linking a date is only rarely a good thing to do. The onus is one those who wish to link a particular date to put forward a case for departing from the general MoS guideline, not on unlinkers to justify bringing articles in line with the MoS. Colonies Chris (talk) 00:43, 27 September 2008 (UTC)[reply]
                • Just so you know, deprecatedbanned. If the outright banning of all linked dates was the intention, then MOSNUM needs to be updated to be explicitly reflect that intent. Until then, as a somewhat ambiguously worded guideline, it is open to different interpretations. If, as you state, "Linking dates is permissible", then the onus should be on the editor going against the de facto consensus in each article. — Bellhalla (talk) 05:02, 27 September 2008 (UTC)[reply]
                  • The key point here is the mistaken idea that there is a a 'consensus' to link dates. It never worked that way. Dates were linked en masse because the MoS said so, not because the editors of each article talked together and made a careful reasoned decision on whether or not to link. As far as I'm concerned, if on any given article there is a consensus among the editors to put back linking (it only takes a single mouse click to revert) then I'm not going to object. But the point is that linking was never something that editors decided to use. Most editors don't care either way - I've unlinked hundreds of articles and in all but a handful of cases there has been no reaction whatever, pro or anti.Colonies Chris (talk) 11:00, 27 September 2008 (UTC)[reply]
                    • A consensus does not have to be explicitly agreed upon. If an editor makes a change to an article and other editors are tacit as to the change, that is a de facto consensus. (Look at your date unlinking activity: articles in which there was no response are an example of a de facto consensus to not link dates.) When a mass removal of all dates—without consensus, I might add—takes place (the reason this discussion was started), there is no way a bot or an editor spending less than ten seconds on an article can tell whether the dates were linked en masse, as you say, or carefully with specific reasons. That's the problem. If a person hasn't worked on an article, they just don't know. — Bellhalla (talk) 13:04, 27 September 2008 (UTC)[reply]
                • But, rather than taking the bull-in-a-china-shop-style approach favored by Lightmouse and others, why doesn't someone with a bot start a campaign to post messages (in the form of a template, perhaps?) on article talk pages. You know, something explaining that there is a new policy (not everyone is an MOS wonk) and outlining the many valid reasons for not linking dates. That way a discussion of interested editors could be held on each article's page. Then after a specified, reasonable amount of time (like, maybe two weeks?), editors can look to see if there are any objections, and if none, unlink the dates then via script or whatever. I bet you'd find many interested editors who, once educated, might even help to do the job. Two weeks, or whatever is deemed a reasonable amount of time, is not that long to wait, because, after all, there is no deadline. The whole situation would be certainly be a lot less adversarial, that's for sure. — Bellhalla (talk) 05:02, 27 September 2008 (UTC)[reply]
    • Response to Bellhalla: Firstly, your suggestion that dates should ony be unlinked after a careful scrutiny of the article (and, presumably, its talk page, and its archives) to determine whether the linking was intended. You're right that a bot can't tell that - but in almost all cases, neither can a human editor. How many articles have any discussion whatever on that subject? Almost universally, an editor such as myself (yes, I am a former date-linker) came along and linked all the dates for autoformatting, and no-one commented on it. Secondly, the suggestion that a bot should place a note on each article's talk page. Then what? After two weeks, a human editor comes back to the article, looks through the talk page to see if there are any objections, and if not, fixes the dates? There are around two million articles to be brought into line with the MoS - this is not remotely practical. And thirdly, in response to your point that "there is no deadline", I would disagree. The present situation sees all manner of time-related terms (weekdays, month names, centuries, decades, years) being routinely linked by editors who think that's the right thing to do because that's what they see elsewhere - the root cause of all this bad practice is the confused design and confusing syntax of the autoformatting mechanism. As you rightly point out, most people don't read the MoS - they edit by example. And what they see is leading them to make things worse all the time. The longer we leave it, the longer it will take. I seriously doubt whether it could be done by manual means alone - the situation is getting worse faster than any manual editing process could catch up. Colonies Chris (talk) 19:53, 28 September 2008 (UTC)[reply]
      • Sorry, but I simply cannot condone a scorched earth policy that is clearly against current consensus simply because other people might come in and clean up behind the bot. And I'm not sure what's so shameful about that. -Chunky Rice (talk) 18:38, 26 September 2008 (UTC)[reply]
      • Also, maybe I'm missing something, but as long as the bot keeps running (and people run scripts without actually paying attention to them), won't a deliberately relinked date get indiscrimiately delinked again? -Chunky Rice (talk) 18:49, 26 September 2008 (UTC)[reply]
    • Desperately needed? Tony, where's the fire? We have more blue than we should have, but very rarely are isolated years a large part of that; very few articles have a large part of their text as years, linked or unlinked. Septentrionalis PMAnderson 18:52, 26 September 2008 (UTC)[reply]
      • I think there is a middle ground between the two alternatives described by Colonies Chris. At the one extreme is "every date link should be individually scrutinised". At the other extreme is running a bot on a list of articles that contain some keywords that give a hint about the appropriate date format, such as "British" or "United Kingdom", without any individual attention to what is in the article. The middle ground is for a person to look at the article, get a sense of what the article is really about, what the appropriate date format is, whether there are any quotes, and whether there are a ton of indiscriminate date links. Once those factors are evaluated, a script can be run to do what is right for that article, or if the article does not lend itself to a semi-automated fix, it can be skipped. --Gerry Ashton (talk) 20:32, 26 September 2008 (UTC)[reply]

"Deprecate"—"to express disapproval of"; orig. 17th century, from the Latin "pray against (evil)". That modern dictionary definition concurs with the consensus at MOSNUM, and the results of the preceding debate. It also concurs with the generally approving attitude to the removal of date autoformatting. Arguing the toss over how fast WP should be ridded of DA is not a good use of our time. If human-supervised automaticity is used, well and good. I don't agree with Colonies Chris that DA is "permissable", by the way, unless he means "only until removed". I think his explanation that the presumption is on an editor to justify DA in an article is probably a stretch: disapproval is disapproval. Tony (talk) 10:09, 27 September 2008 (UTC)[reply]

Thanks, Tony, but I do have access to a dictionary. Unfortunately, you explanations and rationales sounds a lot like wikilegalese. If the intent of the MOS is truly to remove all linked dates, then please make a proposal to explicitly spell that out; otherwise, it is open to interpretation. — Bellhalla (talk) 20:18, 28 September 2008 (UTC)[reply]

I can only second this proposal ("Please refrain from continuing to unlink dates with AWB"). Please repair David Strauss. Guy Peters TalkContributionsEdit counter 10:21, 27 September 2008 (UTC)[reply]

Misbehaving bot

Hi, Lightbot has been messing around with dates here. --Candlewicke (Talk) 04:50, 28 September 2008 (UTC)[reply]

Good feedback. I have updated the code and I am correcting pages where it occured. Thanks. Lightmouse (talk) 07:41, 28 September 2008 (UTC)[reply]

With respect, I gave you the same "feedback" two weeks ago. How many more articles has the bot damaged since then? Sarah777 (talk) 13:40, 28 September 2008 (UTC)[reply]

The earliest comment that I can find is ten days ago. You asked for the retention of concealed links on a matter of principle. Candlewicke reports an incorrect addition of 'xx'. In response your complaint, I believe that I told you that Lightbot has been updated and will not remove concealed links - they will continue to remain invisible to the reader. I have not gone back into the archive to check what I said, please feel free to do so yourself.

However, if you or anybody else has added a concealed link to a full date, Lightbot will continue to remove them. That is a good thing. If you still believe that is wrong, perhaps we should move the discussion to wt:mosnum where you can get third party opinions rather than just mine. I hope that helps. Lightmouse (talk) 13:57, 28 September 2008 (UTC)[reply]

Helicopter

Hi, I undid your edits to Helicopter, which were just removing links to dates and years. Linking dates and years is recommended, and I couldn't see a good reason why to remove the links. I'll watch your talk page if you want to chat about this. Petemyers (talk) 16:38, 28 September 2008 (UTC)[reply]

Hi,
Where are they recommended? The guidance at wp:mosnum says:
  • Dates (years, months, day and month, full dates) should not be linked, unless there is a particular reason to do so.[
Regards Lightmouse (talk) 16:40, 28 September 2008 (UTC)[reply]

Ah, sorry, they are recommended here: http://ang.wikipedia.org/wiki/Wikipedia:Tutorial_(Wikipedia_links) I'll reinstate your edit.

I find it a funny policy personally, it makes sense to me to link all dates, but if that's the guideline that's the guideline... I'll revert your edits back. Petemyers (talk) 16:50, 28 September 2008 (UTC)[reply]

Thank you. Lightmouse (talk) 16:59, 28 September 2008 (UTC)[reply]
I've responded on Pete's talk page with links to more information about DA and its removal. Tony (talk) 09:57, 30 September 2008 (UTC)[reply]

Quick question

Is it possible to nominate categories for the bot treatment? These ones are pretty uncontroversial - I wrote nearly all the contained articles prior to the guideline changing (hence this is the only reason why they were linked to start with) and am the only editor of most of them, and all of the dates are dmy (already correctly formatted) and modern era, etc. All in, there's about 450-500 contained articles, I believe - would do them myself but my time available is woefully short.

The categories concerned are:

  1. Category:Local Government Areas of Western Australia
  2. Category:Former Local Government Areas of Victoria (Australia)
  3. Category:Former Local Government Areas of Queensland

If not, that's OK, but never hurts to ask. :) Thanks for any help you can offer. Orderinchaos 02:57, 29 September 2008 (UTC)[reply]

RfC now open on linking dates of birth and death

Further to comments above on whether dates of birth and death at the top of a bio should or should not be linked, an RfC is now open at WT:MOSNUM#RfC: Linking of dates of birth and death Jheald (talk) 11:30, 29 September 2008 (UTC)[reply]

Question

When you are running the AWB script to unlink dates and years, are you reviewing each one to make sure that it is appropriate to remove the link? -Chunky Rice (talk) 20:18, 29 September 2008 (UTC)[reply]

If I may jump in, my response to Bellhalla, above, is relevant here. How would anyone determine this? Should an editor carefuly scrutinise the article (and, presumably, its history and its talk page, and its archives) to determine whether the linking was intended? How many articles have any discussion whatever on that subject? Almost universally, an editor came along and linked all the dates for autoformatting, probably in the course of making some other change, because that was what the MoS recommended, or linked bare years because they'd seen it so many times elsewhere and thought that was what WP required; and no-one commented on it. So there's no discussion to find.
The thinking behind this question appears to be that no outsider should ever change anything in an article without consulting the existing editors first. This run quite contrary to the spirit of WP:BE BOLD. If the change turns out to be unhelpful, or misunderstands something that's not immediately obvious to an outsider, there's always the revert button. And it's reasonable to presume that a change that's in accordance with the MoS will be the right thing to do in almost all cases. Colonies Chris (talk) 08:14, 30 September 2008 (UTC)[reply]
No, changes in accordance with MoS are usually enforcing a "consensus" of a handful of opinionated cranks. MoS should respond to WP rather than the other way around. Septentrionalis PMAnderson 17:07, 30 September 2008 (UTC)[reply]
I never said anything about consulting other editors. I asked what discretion is applied to each edit. It has nothing to do with the intent of the original linker. It has to do with whether the link has any value. And why won't anybody let Lightmouse actually answer a question for himself? A lot of people have asked some very reasonable questions, but instead of responding he just defers to Tony or others to respond for him.
I have no problem with running a script to remove links AS LONG AS discretion is being applied. If no discretion is being applied, then it's no different than running a bot, which would be unacceptable. All I'm asking for is a confirmation that the user spends a couple seconds looking at each change to ensure that 1) the link doesn't serve any purpose and 2) no errors are produced. I don't think that's unreasonable to expect. -Chunky Rice (talk) 17:14, 30 September 2008 (UTC)[reply]
Anderson, will you make me a small, colourful template to label myself as "an opinionated crank"? I rather like it. Tony (talk) 18:37, 30 September 2008 (UTC)[reply]
CR, since you aren't prepared to suggest any practical means for an editor, human or bot, to make such a determination, your question is both empty and tendentious and LM is quite right not to answer it. Colonies Chris (talk) 19:38, 30 September 2008 (UTC)[reply]
My question is simple. Is discretion being used in these scripted edits? It's not a trick question. I don't understand what you mean by "practical means." You read the sentence (or paragraph or whatever you need for context) and form an opinion. Is the link useful? Yes or no. It's not complicated. -Chunky Rice (talk) 20:43, 30 September 2008 (UTC)[reply]
Oh, and obviously, a bot is incapable of making such a determination, so I'm not sure why I'd try to suggest a way that it could. -Chunky Rice (talk) 20:44, 30 September 2008 (UTC)[reply]

Birth and death years and categories

Hi Lightmouse. Would you have time to read the two posts I made here and here? Part of the reason I think some people (or me, at least) objected to the removal of links to birth and death years (that is year links, not the month/day links which really are trivia) was due to the potential metadata implications. Birth and death year metadata currently resides in several places on biographical articles: Wikipedia:Persondata, the birth and death categories (the largest set of data), infoboxes, and links from birth and death years. The trouble is that this is not completely consistent, and if I could program a bot to do a "biographical audit" (as you've been doing a "date audit"), then one of the things I would check is whether birth and death categories existed before removing birth and death year links. One of my questions was whether your bot can tell when it is editing a biographical article and whether or not the birth and death categories are present? Would you be able to analyse the contributions log of Lightbot to see how many biographical articles it affected? Or alternatively, do you think you could do any of the other (admittedly rather large) tasks I suggested in those posts? I wouldn't normally make a personal request like this, but I've tried bot requests several times, and sometimes people have taken this on, but with mixed results. If it's not your sort of thing, don't worry - I just thought I'd ask. Carcharoth (talk) 00:09, 1 October 2008 (UTC)[reply]