Wikipedia:Bots/Requests for approval/AnomieBOT 61
- The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was Approved.
Time filed: 21:01, Tuesday March 13, 2012 (UTC)
Automatic, Supervised, or Manual: Automatic
Programming language(s): Perl
Source code available: User:AnomieBOT/source/tasks/ShowByDateSubster.pm
Function overview: Replace or remove instances of templates that "expire", as appropriate.
Links to relevant discussions (where appropriate): Wikipedia:Bot requests/Archive 46#Removal of redundant date templates, Template:Show by date/doc
Edit period(s): More or less daily
Estimated number of pages affected: Depends how many use {{show by date}}. Current backlog is 31 pages.
Exclusion compliant (Y/N): Yes
Already has a bot flag (Y/N): Yes
Function details: Certain templates, such as {{show by date}} or {{citation needed by}}, "expire" after a period of time and should be replaced or removed. The bot will go through transclusions of these templates, calculate the expiration date, and if the date is past replace or remove the invocation of the template as appropriate.
Current target templates are:
- {{show by date}} – Replaced with the value of parameter 5, after the date defined by parameters 1-3.
- {{citation needed by}} – Removed, 14 days after the date defined by parameters 1-3.
Other such templates may be added as needed.
Discussion
editWhile I have no problem getting rid of old "show by date" via bot, I'm not sure removing material via bot is all that good an idea. Seems that it could very-easily be abused/gamed or something. At the very least, any bot removing content should not do it while using its bot flag (so the edit shows on watchlists), and the edit should not be marked as minor, and the edit summary should explicitly ask editors to review the edit.
I'd like some other BAG input on this before going forward with a trial. Headbomb {talk / contribs / physics / books} 00:35, 14 March 2012 (UTC)[reply]
- With the caveats of watchlist and editsummary notice, I would be fine approving this. Seems like a fairly tiny usage. MBisanz talk 02:46, 14 March 2012 (UTC)[reply]
- So the concern is someone adds {{citation needed by}} with a past date just to get the bot to do the actual removal of the content? Seems a bit paranoid to me, but there you go. Anomie⚔ 03:03, 14 March 2012 (UTC)[reply]
- My concern was more removing the content entirely rather than letting it just be hidden by the template, but reviewing the documentation I see many warnings that this may happen so I'm really not at all concerned now as long as bot=0 and minor=0; I wouldn't even flag it for review. This would of course change if it was abused in the future but really, there are much better ways to abuse the wiki. — madman 03:21, 14 March 2012 (UTC)[reply]
- So the concern is someone adds {{citation needed by}} with a past date just to get the bot to do the actual removal of the content? Seems a bit paranoid to me, but there you go. Anomie⚔ 03:03, 14 March 2012 (UTC)[reply]
- I'd agree on not editing with the bot or minor flags when removing {{citation needed by}}, or perhaps leaving the text for which citation was required as a comment, but otherwise this task is just doing what MediaWiki does by itself anyway. Reviewing the source code, I highly approve of checking the last revision ID of the templates, which is an excellent example of defensive programming; I'm not sure about a hard-coded list of pages to skip, though. Are these templates commonly used inside the Template namespace? If not, I'd just exclude the Template namespace. Finally, I love having a "whine" function and wish to build one into my own framework now. — madman 03:14, 14 March 2012 (UTC)[reply]
- (edit conflict) I see Headbomb's concern has been addressed satisfactorily; Approved for trial (30 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. — madman 03:14, 14 March 2012 (UTC)[reply]
- Trial complete. [1]. Everything appears to have worked as intended. Anomie⚔ 16:16, 14 March 2012 (UTC)[reply]
- (edit conflict) I see Headbomb's concern has been addressed satisfactorily; Approved for trial (30 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. — madman 03:14, 14 March 2012 (UTC)[reply]
I don't see any effort to detect if a citation is provided (still tagged, in error). Is it worthwhile checking, or is that a warren we ought not be going down? Josh Parris 13:27, 14 March 2012 (UTC)[reply]
- Mmm, that's a fair point; it can't hurt to check for <ref /> tags in parameter four. — madman 13:41, 14 March 2012 (UTC)[reply]
- Is the concern here something like this situation? Ever since the new parser was introduced years ago, a reference in a template parameter that doesn't wind up being output to the page is never processed, so removing it entirely doesn't change anything. Wikipedia:Bots/Requests for approval/AnomieBOT 3 is for fixing this issue.
- Or is the concern that someone would have used {{citation needed by}}, and then someone else added a
<ref>
but didn't remove the {{citation needed by}}? Which would have resulted in something along the lines of "text.[1][citation needed]" in the article? I'm not sure what exactly the bot can do about that, and note that in the current configuration the text will already have been hidden for 2 weeks. Anomie⚔ 16:16, 14 March 2012 (UTC)[reply]- The latter hypothetical was the one I was considering; if an article included
{{citation needed by|2012|6|5|This sentence is not referenced!}}
, later an editor added a reference:{{citation needed by|2012|6|5|This sentence is not referenced! [1]}}
, then the template expires and two weeks later the bot removes it. Yes, the problem would have existed for two weeks as the reference was hidden, but the page might not be watched closely and the problem is recoverable; in that situation, I should think the bot could just check for "<ref" within parameter four and skip or flag the page. It's an edge case, to be sure, and perhaps a paranoid one, but it can't hurt to cover it. — madman 18:27, 14 March 2012 (UTC)[reply]- The rabbit warren I was referring to is: inline citations are just the tip of the iceberg, and the one best known by most editors. But Wikipedia explicitly doesn't dictate a citation style, so there's any number of them, some of them tough to automagically detect. Josh Parris 22:28, 14 March 2012 (UTC)[reply]
- Then perhaps it's not worth checking at all. If it were to be checked at all, I'd say just check for "<ref", as I believe that's what most bots do, AWB does, etc. when checking to see whether an article is referenced or not. But this may be an edge case that's going to be rare enough there's no need to code for it; human editors should be able to spot it. — madman 23:39, 14 March 2012 (UTC)[reply]
- The rabbit warren I was referring to is: inline citations are just the tip of the iceberg, and the one best known by most editors. But Wikipedia explicitly doesn't dictate a citation style, so there's any number of them, some of them tough to automagically detect. Josh Parris 22:28, 14 March 2012 (UTC)[reply]
- The latter hypothetical was the one I was considering; if an article included
Failing dissent, Approved. Josh Parris 13:15, 21 March 2012 (UTC)[reply]
- The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.
- ^ foo