Jump to content

Wikipedia talk:Administrators' noticeboard/CXT/Pages to review/Clobbers

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Battle of the lists

[edit]

I created User:Cryptic/query/19060 before seeing this, with the specific intent to find articles that had been overwritten and not reverted. (I'll probably finish sorting through it in a day or two.) Obviously, some of the unreverted ones were fine from the start, some have been cleaned up, and some will need to be reverted. More interesting are the ones that overwrote redirects; overwrote nearly-empty pages created by the same user; or had the previous article history imported from their source wiki. Those should be treated the same as any other raw, possibly machine-, translation. —Cryptic 07:02, 1 June 2017 (UTC)[reply]

@Cryptic and Tazerdadog: Yes agreed, Cryptic; I've noticed those too. Plus, it occurred to me after going through about a quarter of the list, that it may not be necessary to go any further—depending on how the executing admin (probably Tazerdadog) implements the nuke: it could be that simply having this list is enough, and anything on it could programmatically avoid nuking regardless of CXT/PTR status.
What say you, @Tazerdadog: is the fact that an article appears on this list sufficient for you not to nuke it, even if it appears on WP:CXT/PTR, unstruck? Or, would you prefer that we strike them on CXT/PTR? Updating CXT/PTR to reflect this list is actually pretty fast, now that we have the list; so if it's easier for you to ignore this list, and just deal with presence/absence of strikeout type at CXT/PTR alone, then I'll finish the operation; it's not that hard.
Btw, there are a couple of discussions on Talk pages at a higher level than this about this exact topic; if it makes more sense to talk there, by all means pick it up at a different location, and leave breadcrumbs. Mathglot (talk) 02:57, 2 June 2017 (UTC)[reply]
Tazerdadog can't execute the nuke, since he's not an administrator.
Doing the deletes mechanically never struck me as a good idea; I'd assumed an admin would look at each unstruck article and make a decision to speedy delete or not. There's not so many (currently ~2100) that that would be unfeasible, and I'd actually planned on taking on part of that task (though I have less free time than I'd expected I would back at the start of this final stage). —Cryptic 03:09, 2 June 2017 (UTC)[reply]
Well I have a bit of a problem with any admin "making a decision" to speedy delete or not after all this effort on our part. If that was going to be the plan from the beginning, then they should have jolly well done that themselves back then, and saved me, S Marshall, Elinruby, DGG, Yngvadottir, and many others the effort of spending tens or hundreds of hours making those assessments one by one, backed with knowledge of various languages and our experience on the site. Okay, I'll be bold, here: I'd have more than "a bit of a problem" I'd be quite annoyed if an admin "makes a decision" about any article that I have looked at, after all this good-faith effort to painfully gain consensus and then to examine articles one by one, and manifest our assessment using the agreed-upon "strikeout" token (and the occasional, optional "kill" token). As far as I am concerned, the admin merely pulls the switch to implement the consensus achieved on the articles here by many editors who have diligently worked on them. And that consensus is simple: If it is enclosed in <s>-tags, save it, and if it is not, nuke it. (With already agreed-upon exceptions such as other namespaces, and maybe a couple of things I've forgotten, but basically, that's it.) Having an admin make any decision that goes against that principle had better not be what is about to happen, after a couple of months and many hours (on my part) and almost a year's worth of effort (on others' part) to deal with this issue. I don't recommend going that route without an Rfc with 3-admin closure to support it. One (monolingual?) admin deciding, after all that effort? Humbug. Find me and admin that is level-3 or higher on 25 languages, and maybe I'll vote to support you. Otherwise, you'll have to persuade me. @Cryptic and Tazerdadog: Mathglot (talk) 03:47, 2 June 2017 (UTC)[reply]
My intent with that wording was the admin would be acting as a final safety valve on the process: that the speedy deletion template had not been accidentally applied to a clobbered article, or to the target of a redirect (after reversion of a bad translation back to the redirect), etc. Certainly not something akin to someone who doesn't speak the language vetoing just because "the English looks ok now", or "the subject is notable so we're keeping this, even though I don't know how much of the article is machine-translated lies". —Cryptic 03:55, 2 June 2017 (UTC)[reply]
I could go along with "admin as final safety-valve" but in that case, I would ask that the admin concerned make up a short-list of undecidable items, and let the team go at it, and make the decision. it should be our call, not an admin's. Admin's are not more knowledgeable about content, they are more privileged about operations. There's no reason in the world for an admin to override any decision made by editors here, although I have no objection to an admin saying to the group, "Are you sure?" That would be only prudent, and welcome. Mathglot (talk) 04:00, 2 June 2017 (UTC)[reply]
Post-script to address your point about X2 "not [being] accidentally applied to a clobbered article": that's exactly what this Project page is about, and I will personally guarantee that that does not happen, by striking at CXT/PTR every article in that category. Therefore ,the implementing admin need not be concerned about that issue at all. It will be moot, in a day or two. Mathglot (talk) 04:06, 2 June 2017 (UTC)[reply]
Absolutely no intent (at least on my part) to be making content decisions. Admins substituting their (lack of) expertise for the translators' is precisely what was infuriating S Marshall at Wikipedia talk:Criteria for speedy deletion/Archive 61#X2 revision, and that's what drew my attention here in the first place. —Cryptic 04:08, 2 June 2017 (UTC)[reply]
Thank you for that. And I might add, it's what also frustrated multi-talented editor User:Yngvadottir no end, to the point where she could no longer continue in this effort, which is a great loss for all of us here. Sounds like we[a] are on the same page. :-) Mathglot (talk) 04:35, 2 June 2017 (UTC)[reply]
Sorry for alarming you. —Cryptic 04:41, 2 June 2017 (UTC)[reply]

Notes

  1. ^ We—that is, "you and I"—a semantic distinction which is encompassed in a single word in numerous languages, but not in English.

Anyway, back to a question you raised before we got sidetracked - I know I wouldn't delete a clobbered article listed here, since it's something I know to watch for; but it's safer to strike them, to be sure. I can take that over the striking, by the way, if all you're doing is making sure everything on this list is struck on /PTR and not doing any analysis of the content. —Cryptic 05:13, 2 June 2017 (UTC)[reply]

Yep, that's all I'm doing on this sub-task, i.e., making sure everything on the list is struck at CXT/PTR; 'A' - 'R' are done, leaving S-Z + non-Latins. I probably won't do any more of these tonight, and either do some lang-tagging at /PTR, or find some translators for WP:CXT/PTR/BL. Or go to sleep. You know, when I go to sleep, it helps to imagine little, fluffy, unreferenced, poorly translated, Wikipedia 1- sentence stubs, monotonously frolicking up to a fence, and jumping over. Only, in this version, the fence is built parallel to, and at the edge of the cliffs of Dover, and as the little fellas jump over, well, you know what happens. Do you think I oughta worry about WP:HOLIC? Sleep well... Mathglot (talk) 06:06, 2 June 2017 (UTC)[reply]
I concur with Cryptic about how to go forward. (I've noticed that I'm often on the same page as Cryptic.)—S Marshall T/C 15:09, 2 June 2017 (UTC)[reply]

Two points: Firstly, we are going to move these to draftspace before we delete them, so everyone who has these pages on a watchlist has the opportunity to provide input, right? Pinging @MusikAnimal:, because I know you have a bot that is specifically designed to mass-draftify articles. Secondly, 2100 articles * 30 seconds of manual review per article (generous, I think it will take longer) = 17.5 hours of admin time. Cryptic, are you REALLY willing to spend that kind of time on this? If the answer is no, I'd lean towards delete/draftify first, and then undelete/undraftify later if a reviewing admin thinks a mistake has been made. Tazerdadog (talk) 19:04, 2 June 2017 (UTC)[reply]