Jump to content

Wikipedia:Interface administrators' noticeboard: Difference between revisions

From Wikipedia, the free encyclopedia
Content deleted Content added
Line 55: Line 55:
::{{done}} and noted at that script talk. — [[User:Xaosflux|<span style="color:#FF9933; font-weight:bold; font-family:monotype;">xaosflux</span>]] <sup>[[User talk:Xaosflux|<span style="color:#009933;">Talk</span>]]</sup> 21:32, 17 April 2020 (UTC)
::{{done}} and noted at that script talk. — [[User:Xaosflux|<span style="color:#FF9933; font-weight:bold; font-family:monotype;">xaosflux</span>]] <sup>[[User talk:Xaosflux|<span style="color:#009933;">Talk</span>]]</sup> 21:32, 17 April 2020 (UTC)
:::Awesome, that works now correctly. [[User:Jo-Jo Eumerus|Jo-Jo Eumerus]] ([[User talk:Jo-Jo Eumerus|talk]]) 09:37, 18 April 2020 (UTC)
:::Awesome, that works now correctly. [[User:Jo-Jo Eumerus|Jo-Jo Eumerus]] ([[User talk:Jo-Jo Eumerus|talk]]) 09:37, 18 April 2020 (UTC)

== A [[WP:DRN]] script has been de-gadgeted and is ready to be renamed ==

Please see the result of [[MediaWiki talk:Gadget-DRN-wizard.js#Requested move 1 April 2020]]. This move was listed at [[WP:RMTR]] to be carried out, but a regular admin can't do it. It needs an interface admin. Judging from the comments left at [[MediaWiki talk:Gadget-DRN-wizard.js]] it seems that [[User:Xaosflux]] must be aware of the de-gadgeting of this script. I am listing this here just to get it out of the RMTR queue, not because I know what it is. [[User:EdJohnston|EdJohnston]] ([[User talk:EdJohnston|talk]]) 22:48, 18 April 2020 (UTC)

Revision as of 22:48, 18 April 2020

    Welcome to the interface administrators' noticeboard

    This is the interface administrator noticeboard, for discussion of interface administrators and coordination of edits to the interface.

    Currently only interface administrators can undelete JS/CSS pages, if you have an uncontroversial undelete or deleted version retrieval request, please list it below.

    Any administrator can delete JS/CSS/JSON pages, for speedy deletions just use a CSD template on the page or its talk page.

    Individual requests for edits to interface or user JavaScript/CSS pages should continue to be made on their respective talk pages.

    1 interface-protected edit request
    v·h
    Page Tagged since Protection level Last protection log entry
    MediaWiki:Gadget-markblocked.js (request) 2024-07-24 06:42 Site JS page (log)
    Updated as needed. Last updated: 06:43, 24 July 2024 (UTC)


    Hello, since {{REVISIONID}} does not work now (see phab:T137900), it is causing issues with prefilled reason in Template:Renameuser2 and Template:Usurp2. So please create MediaWiki:Group-user.js (see m:MediaWiki:Group-user.js) so that the solution provided at phab task can be implemented. Thanks! ‐‐1997kB (talk) 05:54, 21 February 2020 (UTC)[reply]

    I get it, but it seems rather excessive to add some javascript, even if it's tiny, to every user's page load, just so global renamers have an easier time with requests. I haven't checked, are there are other templates using this trick that would further justify making this change? ~ Amory (utc) 11:00, 21 February 2020 (UTC)[reply]
    I am only aware of usage in these two templates. If you have any better idea without creating that page, we can implement that. ‐‐1997kB (talk) 11:26, 21 February 2020 (UTC)[reply]
    I was about to say the same thing as Amorymeltzer. I think the benefit here is not worth loading this for every user. Alternative ways should be considered such getting dedicated bot to add the permalink or script like the one used at WP:PERM. – Ammarpad (talk) 11:55, 21 February 2020 (UTC)[reply]
    Found out that Template:Unblock-un is also using it. ‐‐1997kB (talk) 12:33, 21 February 2020 (UTC)[reply]
    Which is fine, but it is the same thing, this was just a shortcut for a very small group of people - if the templates are broken then fix them (turn off the bad links). — xaosflux Talk 12:52, 21 February 2020 (UTC)[reply]
    Well I have updated the templates to link with just main page (not the specific revision), which seems kinda fine to me. If anybody comes up with better solution update it please. ‐‐1997kB (talk) 13:04, 21 February 2020 (UTC)[reply]
    Seems like this would be ripe for a global-renamer user script, a la the meta example. ~ Amory (utc) 13:08, 21 February 2020 (UTC)[reply]
    Meta is pushing this on "all users" as well, which is also overkill but there are much less meta users and usage. — xaosflux Talk 13:47, 21 February 2020 (UTC)[reply]
    Exactly right. ~ Amory (utc) 17:01, 21 February 2020 (UTC)[reply]

    WP:Requested moves#Technical requests is another application broken by the removal of support for {{REVISIONID}}. See Wikipedia talk:Requested moves#Permalinks no longer linked in move reasons. We're all here to ask for help with MediaWiki:Group-user.js because that's the only solution presented by the developers who broke our apps, in their "migration guide". I'd be happy if the solution only worked for administrators and users with page mover rights. An installable "gadget" would do the trick. – wbm1058 (talk) 16:53, 29 February 2020 (UTC)[reply]

    @Wbm1058: your explanation above suggests that this is only needed by a very small minority of editors, so adding code for every editor on every page load is certainly overkill if it could be done in MediaWiki:Group-sysop.js or MediaWiki:Group-extendedmover.js. That being said, it also seems to be only needed for a minority of those users as well - so an opt-in gadget (Page moving helper gadget or the like) sounds like a good answer. — xaosflux Talk 17:19, 29 February 2020 (UTC)[reply]
    Thanks, I didn't even know that MediaWiki:Group-sysop.js and MediaWiki:Group-extendedmover.js existed. A gadget would be ideal if we could count on drive-by users to install the gadget before using, but getting everyone to follow instructions isn't easy. If the load caused by a few extra lines of code is significantly less than the previous load caused by {{REVISIONID}} generation, then maybe it's worth installing it in the two MediaWiki script files, for user convenience and to ensure more consistent use. wbm1058 (talk) 18:58, 29 February 2020 (UTC)[reply]
    @Wbm1058: anything in the group.js gets loaded on every single page load; the prior was only when making an edit and using that template (if I'm reading this right) so it is a much bigger difference. — xaosflux Talk 19:08, 29 February 2020 (UTC)[reply]
    (edit conflict) The "previous load caused by {{REVISIONID}} generation" was, in effect, that pages that contained the magic word were several times slower to save. That seems much less of a big deal than the new way of extra JavaScript being loaded on every logged-in or autoconfirmed page view. (Just putting it in MediaWiki:Group-extendedmover.js isn't good enough, because some RM/TR requests are made my IPs, and don't have any pages in the way, meaning an autoconfirmed user could in theory do it. Another possible reason for such a request would be the title blacklist, creating a request that could be processed by a template editor who is not a page mover). * Pppery * it has begun... 19:10, 29 February 2020 (UTC)[reply]
    So, as I said at Wikipedia talk:Requested moves#Permalinks no longer linked in move reasons, the revisionID permalinks aren't placed in the edit summaries of ordinary edits, but only the edit summaries of page moves, which are clearly more resource-intensive operations, e.g. HERE. Thus, it's the page movers who leave the permalinks – to the version of the technical requests page that was live at the time the page was moved.
    Maybe we can talk the developers into re-enabling {{REVISIONID}} during page-move operations only? If this is indeed more efficient than loading JavaScript, then it strikes me as disingenuous to suggest JavaScript as the "migration" solution. wbm1058 (talk) 20:03, 29 February 2020 (UTC)[reply]
    You're misunderstanding me. re-enabling {{REVISIONID}} during page-move operations only isn't a thing because the {{REVISIONID}} magic word is evaluated when the page Wikipedia:Requested moves/Technical requests is rendered, not when the page move happens. The long parenthetical in my previous post was explaining that one doesn't need to be a page mover to execute a request at RM/TR. * Pppery * it has begun... 20:34, 29 February 2020 (UTC)[reply]
    What is the big deal from one script not working the way it used to? Can't see this preventing anyone from actually making a request, nor from anyone actually fulfilling it correct? — xaosflux Talk 22:40, 29 February 2020 (UTC)[reply]

    So if I understand correctly, most of the time this functionality is used ONLY to prefill the editsummary (or the edit itself) with the revisionid, for purposes of a select group making the action ? In that case, create a gadget, with the replace functionality as described, enable by default but place a relevant rights filter in the gadget definition, so that only those with that right (and thus the only ones to be able to actually perform the action) receive the gadget. formwizard gadget does this for instance. And of course try to do it with as few gadgets possible of course. You might as well hide the action with 'sysop-show' to only make it visible to sysops for instance (ps, might be that not all groups have relevant show/hide classes, it's been a while since that was checked I think). —TheDJ (talkcontribs) 15:48, 4 March 2020 (UTC)[reply]

    Btw. putting it in a relevant group-groupname.js might be 'cheaper', as that is automatically added to the the 'user' module of the relevant user, whereas a gadget, actually adds a RL module definition, which is then included on every page for every user. Of course that is only a couple of bytes on each pageview, but still. It sort of depends on the level of control needed and the size of the group you are targeting, what the right choice is. —TheDJ (talkcontribs) 15:53, 4 March 2020 (UTC)[reply]

    Question about repairing an userscript

    There is a (now archived) bug report about User:Lingzhi2/reviewsourcecheck.js breaking the edit screen on Invisible Man. Given that Lingzhi2 hasn't been active in the past few weeks (I did ping them again, though), I wonder if it would be appropriate for an IAN to go in and fix them. Jo-Jo Eumerus (talk) 11:38, 17 April 2020 (UTC)[reply]

    @Jo-Jo Eumerus: to be clear, this isn't breaking any editing in general - it is breaking something for you because you have chosen to use Lingzhi2's script - correct? You should be able to edit the article in safemode, or by not importing the faulty script. — xaosflux Talk 14:22, 17 April 2020 (UTC)[reply]
    Yes, but I don't think that that means that whatever the issue should be left unresolved. Jo-Jo Eumerus (talk) 14:24, 17 April 2020 (UTC)[reply]
    Looks like there are about 40 people using this, so it is quite low impact. @Jo-Jo Eumerus: do you have a specific fix in mind? If not, you could post at Wikipedia:User scripts/Requests to see if someone wants to write a fix. If it is something very minor, I think we would consider forcing the update on to Lingzhi2's personal page - if it is larger you could always fork the script. — xaosflux Talk 14:28, 17 April 2020 (UTC)[reply]
    A bug-fix type fix can be proposed at User talk:Lingzhi2/reviewsourcecheck.js as an edit request, which will enqueue it for review as well. — xaosflux Talk 14:29, 17 April 2020 (UTC)[reply]
    That's part of the issue - I don't really know what the correct fix is, other than TheDJ's suggestion that It seems it modifies links in a page, without considering that not all links are part of content. Jo-Jo Eumerus (talk) 15:03, 17 April 2020 (UTC)[reply]
    @Jo-Jo Eumerus: the immediate fix for you is to edit User:Jo-Jo Eumerus/common.js and stop using that broken script until it is fixed. If this was a project wide gadget, we'd probably get more attention put on it, but with it being a private script used by a very little number of editors it is very low on the response priority. — xaosflux Talk 16:16, 17 April 2020 (UTC)[reply]
    @Jo-Jo Eumerus, Xaosflux, and Lingzhi2: I've fixed the bug at User:Wugapodes/reviewsourcecheck.js with this change. Would a interface admin be willing to make this change if Lingzhi doesn't return soon? Wug·a·po·des 20:25, 17 April 2020 (UTC)[reply]
     Done and noted at that script talk. — xaosflux Talk 21:32, 17 April 2020 (UTC)[reply]
    Awesome, that works now correctly. Jo-Jo Eumerus (talk) 09:37, 18 April 2020 (UTC)[reply]

    A WP:DRN script has been de-gadgeted and is ready to be renamed

    Please see the result of MediaWiki talk:Gadget-DRN-wizard.js#Requested move 1 April 2020. This move was listed at WP:RMTR to be carried out, but a regular admin can't do it. It needs an interface admin. Judging from the comments left at MediaWiki talk:Gadget-DRN-wizard.js it seems that User:Xaosflux must be aware of the de-gadgeting of this script. I am listing this here just to get it out of the RMTR queue, not because I know what it is. EdJohnston (talk) 22:48, 18 April 2020 (UTC)[reply]