[09:52:46] bawolff: I mean, history is versioned right? We just need to link to pages at previous points in time? [09:52:54] And have a nice little template or some lua to facilitate that [09:53:18] I guess that means you can edit history though, but perhaps then forking to subpages if folks want to edit could be taken into account too? [15:29:48] are there any issues with WMCloud ? [15:32:28] Dragonfly6-7: none that we're aware of. are you seeing some issues? [15:35:34] taavi - on the weekend, I tried to run a query that I've been using for a few months, and got an error message. I was told that there was some ongoing database maintenance which should be done by monday [15:35:53] what error message? and where exactly are you seeing that? [15:36:14] Query status: failed [15:36:14] Error [15:36:14] Unknown column 'tl_title' in 'field list' [15:36:24] https://quarry.wmcloud.org/query/65628 [15:36:43] schema migrations are ongoing [15:36:46] so yes, that is expected [15:36:55] ah , okay [15:37:00] eh [15:37:17] I think that query needs to be updated following the changes to the templatelinks table [15:37:20] give me a minute [15:40:19] Dragonfly6-7: something like this should work https://quarry.wmcloud.org/query/67300 (once it finishes up executing) [15:41:48] is 65628 basically dead? [15:42:53] you can update that or fork my one for a new query [15:43:34] sorry, clarify: will 65628 work again, once the schema migration (or whatever) is complete? [15:43:38] no [15:43:56] okay, so 65628 is dead? [15:43:57] the tl;dr is that we changed how template transclusions are stored because the new way uses much less storage space, but as a result existing queries can unfortunately break [15:44:05] Okay. [15:44:08] by that definition, yes [15:44:48] and any repairs would effectively be a new query. Yes? [15:45:59] (assuming my question is phrased meaningfully) [15:46:13] 67300 is just 65628 updated to work with the new way [15:46:37] Sure, but that's not the question I asked. [15:47:04] i'm not exactly sure what you're asking [15:47:07] to rephrase, then [15:48:24] given query, X, with the set of parameters Q... Q cannot be modified. If Q is modified, then that creates a new query, Y. Is this a reasonably valid description of what happens? [15:50:21] also: given your parenthetical about "once it finishes executing", can 67300 be run at this time without problems? [15:50:53] given that it already finished executing, I’d answer that second question with “yes” [15:51:59] thank you [15:53:52] I'm assuming Q refers to the contents of the 'SQL' textbox on the query details page. In that case, no, the author of a query can edit the that, and subsequent runs of the query will use the new SQL instead of the old one [15:54:07] and yes, Lucas_WMDE is correct. when I linked that here it was still running [15:54:14] * Lucas_WMDE agrees with taavi [15:56:15] so, in principle, 65628 could have been edited back to functionality, if I had any idea how to use SQL instead of simply having forked that off a query that someone else wrote for me. [15:56:16] Correct? [15:57:55] yes, but you can also edit it *now* back into functionality by copying the SQL taavi wrote [16:00:53] (which is conveniently licensed under CC0, like all Quarry queries ^^) [16:02:15] and the advantage to my continuing to use 65628 is that it's mine, and therefore I can re-run it whenever I want, instead of having to come to you and nag you about it. Yes? [16:02:28] yes, correct [16:02:37] Okay, thank you. [16:03:15] related question, pertaining not so much to the query as to how the results are displayed [16:04:00] currently, pages can be sorted alphabetically (or reverse-alphabetically) by filename, or by template name [16:04:18] would it also be possible to sort them by quantity of hits [16:04:51] such that the templates that appear >100 times appear before the templates that appear <100 times, for instance [16:04:56] if this is not possible, oh well [16:07:20] in this context, templates = Creator: pages, right? [16:07:34] I think it would be possible but the query would be more complicated, not sure I could get it right [16:08:38] yes, in this context, 'templates' = Creator pages [16:09:47] I've been tracking my progress - whenever I run the template again, I make a note of the number of results [16:10:31] when I started earlier this year, it was at 119901; it's now at 78092 [16:11:13] dealing with the frequently-used ones might be easier if I knew which ones they were [16:11:28] but it's not essential, thank you just the same [16:13:17] would just the creator templates and their number be enough? [16:13:23] (without the original files linking to them) [16:13:37] that would be an easier query ^^ [16:13:54] hm [16:14:01] yes [16:14:50] alright, then that should be https://quarry.wmcloud.org/query/67302 [16:15:39] top entry: https://commons.wikimedia.org/wiki/Special:WhatLinksHere/Creator:Issued_by_William_S._Kimball_%26_Company [16:15:51] hah [16:15:54] yes, not surprised [16:16:08] I've been collecting information about the kimball company to make a wikidata entry [16:16:49] "Creator:Hulk" is there 61. [16:16:53] 61x, rather [16:17:08] problem: there was a whole family of Dutch artists with the surname "Hulk" [16:17:18] you can also add `HAVING COUNT(*) > 1` at the very end of the query (or a higher limit, if you want) to reduce the number of results returned [16:17:23] ah :/ [16:17:27] I can see how that might be an issue… [16:18:12] every now and then I have to create a Wikidata entry [16:18:15] (correction, the HAVING would need to go before the ORDER BY, not at the very end) [16:19:13] it's difficult to track information about Eugene Charpenay (341x), for instance, because he was influential enough that there is now a "Quai Eugene Charpenay" in Grenoble [16:19:28] and therefore most of the results on Google for that name are real estate ads [16:19:43] or maps, etc [16:20:08] ah, there we go. BNF has an entry. [16:20:18] thank you [16:20:25] np :) [17:56:47] has the way that email confirmation works on Wikipedia changed recently? like, are users unable to do a password reset with an unconfirmed email address? [18:01:20] no [18:01:32] in fact, it's the other way around [18:01:37] if you do an email reset [18:01:44] it automatically confirms your email address [18:01:52] but it has been doing it for a long time [18:02:41] they were thinking about enabling IPv6 for email [18:02:48] but I think it wasn't done yet [18:02:55] what's the actual problem, ragesoss ? [18:03:27] okay. i'm dealing with a user who can't get into their account, and they have a screenshot with an error message that says "Confirm email address / Invalid confirmation code. The code may have expired." I'm not sure whether they actually didn't add an email to their account, or if something else is going wrong. [18:04:29] is it on VRTS? [18:05:52] no [18:06:37] that error message makes me think that they did receive the email [18:06:49] but opened it many days later [18:06:56] or didn't copy the full url [18:07:32] so maybe they are currently just being rate-limited because of failed login attempts, and if they wait long enough they can do another password reset and have it work [18:11:17] I would check the date of the email they received [18:11:20] and try again [18:11:31] perhaps they copied the url instead of clicking [18:11:34] that's what i've advised. thanks! [18:11:36] but not full