[01:13:41] Hey gang, I'm not terribly familiar with the phab/release system, where does this one lie at the moment? https://phabricator.wikimedia.org/T180678 [01:13:54] It's making the new source editor unusable for me :( [01:15:37] drewmutt: we were hoping that it would be fixed when we fixed the subtask. if it's still broken for you, please comment so on the task :( [01:16:02] (i don't think we've re-tested it yet) [01:17:25] Okie, thanks for the quick response! Fwiw, it seems it's related to bolding pushing the line to where there's some disagreement between if a word should remain on that line or drop. [08:48:31] Hi, if somebody present here is responsible for outreachdashboard, it is down (502 Bad Gateway). Does somebody know why? Thanks in advance [09:53:55] Urbanecm: #wikimedia-cloud is the channel your looking for [11:20:57] https://en.wikisource.org/wiki/Author:Carrie_Jacobs-Bond#Sheet_music [11:21:06] Where's the whtiespace coming from? [15:30:48] Technical Advice IRC meeting starting in 30 minutes in channel #wikimedia-tech, hosts: Amir1 and Lucas_WMDE – all questions welcome, more infos: https://www.mediawiki.org/wiki/Technical_Advice_IRC_Meeting [15:52:56] Technical Advice IRC meeting starting in 10 minutes in channel #wikimedia-tech, hosts: @amir1 & @Lucas_WMDE - all questions welcome, more infos: https://www.mediawiki.org/wiki/Technical_Advice_IRC_Meeting [15:53:00] :) [15:55:45] can I talk about user-rights in the tech meeting? [15:55:54] specifically making new ones [15:56:29] Dysklyver: what wiki? [15:56:40] Or on your own [15:56:54] En wikipedia [15:57:04] I have a bunch on my own [15:58:16] I want to know why adding a new user-right is billed as "hard" and "taking lots of dev time", when you can add one to a standard MW build with half a dozen lines of code... [16:00:54] \o [16:00:56] alright, I think we can start with the Technical Advice IRC Meeting! Welcome everyone :) [16:01:00] Hello! [16:01:04] hi [16:01:14] o/ [16:01:35] I’ll just drop a link to the board for dev questions that we have since a few weeks ago: https://discourse-mediawiki.wmflabs.org/ [16:02:50] I think we also have already one question by ? [16:03:00] also, reminder that the deadline for applying for scholarships for the Hackathon 2018 is January 30th: https://www.mediawiki.org/wiki/Wikimedia_Hackathon_2018 [16:03:02] So for the benefit of the records, I will repeat my question in the meeting, one sec [16:03:24] how difficult would it be to add a new useright to wikipedia? [16:04:07] Dysklyver: In technical terms, it's very easy, you need to make a patch against operations/mediawiki-config [16:04:23] the useright would allow editors to move a draft to article space, which would otherwise only be allowed for admins [16:04:34] https://www.mediawiki.org/wiki/Manual:User_rights [16:04:36] would that make it harder? [16:05:35] (Also bear in mind that I have tested this on a standard MW, this is just to see how it could be added to wikipedia) [16:05:37] Dysklyver: anyone can move Rights from draft to article space on en-wiki [16:05:42] It can become complex because you are moving between two different namespaces and not inside a namespace [16:05:44] I mean move pages [16:06:05] Adotchar, yes this would change that [16:06:21] I would guess the hard part would be the discussions around it… [16:06:24] Dysklyver: If you already done it, it's very easy [16:06:54] Lucas_WMDE: Yeah. I’d oppose it on the basis everyone would make articles straight to main space and skip the AfC submission process. [16:07:00] Dysklyver: https://github.com/wikimedia/operations-mediawiki-config/blob/master/wmf-config/InitialiseSettings.php [16:07:25] This is the place you need to make a patch [16:07:37] ok [16:07:53] but overall, I agree, this needs community consensus before moving to technical side [16:07:54] (warning, that’s a huge file :) ) [16:08:02] so from a technical viewpoint, this is very easy to implement? [16:08:25] “a web page is slowing down your browser…” just fyi :) [16:09:06] Dysklyver: yes [16:09:17] ok thanks [16:10:14] is there any news about the new search index for wikidata? any idea when it is going to be rolled out? [16:10:43] (not sure this is the right meeting to ask that actually) [16:11:51] pintoch: this is the right place, at the moment, the elastic backend for search has been rolled out in Wikidata (it was rolled out in the Wikidata's birthday) [16:12:05] but there is an ongoing process to improve the index [16:12:39] I don't know the details of that but SMalyshev is the person is working on it [16:14:56] okay! I was referring to http://wikidata-wdsearch.wmflabs.org/wiki/Special:Search [16:15:20] it looks like this experimental one is still much better than the production one [16:15:38] I’m not sure if this wiki is still in use [16:15:50] Special:Search on Wikidata seems to be using Cirrus at least [16:16:58] It has been since the begging [16:17:14] *beginning [16:25:54] oh I have a question for Lucas_WMDE actually. Are distinct value constraints meant to be enforced in references? [16:26:19] pintoch: not sure about “meant to”… but we do check them there, yes [16:26:23] see for instance the references of https://www.wikidata.org/wiki/Q815352#P355 which are flagged [16:26:45] I have been using ids in references like this and thought it was fine [16:27:07] hm, I’m not sure how to interpret that reference tbh [16:27:33] the idea is just to say "that relationship is stated in this GRID record" [16:27:35] shouldn’t that be “stated in: GRID” or something like that? [16:27:55] the issue with that is that it is less specific, and harder to verify [16:27:59] but anyways, that’s not my place to judge :) [16:28:11] if the distinct values constraint should not be checked on references, then you can specify that [16:28:17] with the brand new “constraint scope” parameter [16:28:26] amazing \o/ [16:28:28] you’d say that this constraint should only be checked on the main value (and perhaps on qualifiers) [16:28:31] and then it would be ignored on references [16:28:39] I just need to check if support for that is already deployed or not… [16:28:54] should be deployed, yes [16:28:58] cool, let me try [16:31:28] > constraint scope is not a valid qualifier for property constraint [16:31:31] oops :D [16:31:35] I just noticed that [16:32:16] haha :) [16:32:21] fixed :) [16:33:04] so, unless a constraint scope is defined, the distinct values constraint is applied everywhere [16:33:10] yes [16:33:27] but it only looks inside the same scope – having the same value in a qualifier and a reference is not a violation [16:33:29] but so, if multiple references use an ID like this one in various references, this will not be flagged [16:34:05] I mean if the multiple references are on the same item, sorry [16:34:13] I’m not sure, actually, let me see [16:34:41] okay, yes, other uses on the same entity are not a violation [16:34:50] example: https://www.wikidata.org/wiki/Q42297317 [16:36:16] okay - anyway now that we have the constraint scope parameter, I just need to start a discussion on Project chat about this sort of references [16:36:27] sounds good :) [16:36:32] I believe all ids should have "constraint scope: main snak" [16:36:39] I’m not sure if this behavior for the distinct values constraint is the best, either [16:36:48] perhaps it should be changed [16:37:17] we have some other constraint types which are by default checked only on the main statement [16:37:30] and where you have to opt-in to have them checked elsewhere [16:37:35] “type”, for example [16:37:44] perhaps that would make sense for “distinct values” as well… [16:39:47] Hey everyone! [16:40:06] Hey :) [16:40:22] Hi! [16:42:11] I got this patch I'm working on but stuck somewhere [16:42:25] https://gerrit.wikimedia.org/r/#/c/405281/ [16:42:43] I'm a little confused on how to get the new names added in the .mailmap file [16:43:13] The issue I'm facing is, when I look at .mailmap, I see canonical names mapped to different email instances [16:43:27] I've not seen different carnonical names mapped to same email address [16:43:50] Can someone clarify me please? [16:44:05] so you have some commits with canonical email addresses but names that need to be mapped? [16:44:38] in that case AFAIK 'canonical name ' in the .mailmap should be enough [16:45:35] Lucas_WMDE: Yeah, I'm having something like this [16:45:56] Pppery [16:46:09] pppery [16:46:16] Is that correct to be in the .mailmap file? [16:46:26] well, one of those should be in the mailmap file [16:46:27] Those are 2 different names that map to the same email [16:46:30] I don’t know which one :) [16:46:41] Yeah, that's where my confusion is [16:46:43] but then one of the names would be mapped to the other one [16:46:52] Which one exactly will be in .mailmap [16:47:34] Lucas_WMDE: Looking at that patch, @Umh left some comments [16:47:44] I didn't really understand the comments TBH [16:49:36] if I understand it correctly, Umherirrender is saying that these entries from the git log should not be added to the CREDITS file [16:49:47] instead, the .mailmap file should be updated so that the names don’t appear in the git log anymore [16:49:55] (instead, the fixed form of the name is used) [16:50:27] and the only thing you need to do is to edit .mailmap and add the added name [16:50:43] I’m not sure if you’re familiar with the .mailmap format – documentation is at https://git-scm.com/docs/git-shortlog#_mapping_authors [16:50:56] specifically, the “John Developer” part of the first example is the same situation as here [16:50:59] so for example for Pppery add it with capital P [16:51:21] actually, if “pppery” with lowercase p already exists in the CREDITS file, I guess it would be better to keep that form [16:51:26] so that the CREDITS entry doesn’t change [16:51:53] yes [16:52:22] unless pppery prefers the uppercase form, I guess :) [16:53:04] Lucas_WMDE: That makes sense, very clear now! :) [16:53:09] Amir1: Thank you too :) [16:53:24] alright, great :) [16:53:26] I'll just do a quick PS and see if this solves the issue (as you explained) [16:53:41] Lucas_WMDE: Will you be willing to idle while I do it and check back here? :) [16:53:47] sure :) [16:58:26] Lucas_WMDE: Ahhh, forgot to ask something. What about the comment Not same as next line? [16:58:40] Normally .mailmap is used when there are duplicate users right? [16:58:45] yes [16:58:49] But in this case, users are different [16:58:55] So I didn't also get that comment [16:59:06] I think this just means that the line looks similar to the next one, but is not the same, so it shouldn’t use mailmap [16:59:10] d3r1ck_: it was just to assure others we don't need to check if they are duplicate [16:59:11] just as a note to the next reviewer [16:59:20] you don't need to do anything [16:59:29] Perfect! [16:59:36] Amir1, Lucas_WMDE, thanks :) [16:59:47] Let me not waste so much time and submit a quick patch [17:00:00] Sorry to be late for the meeting. Facing some issues with ZNC [17:01:51] Patch set is up! [17:02:01] https://gerrit.wikimedia.org/r/#/c/405281/4 [17:02:12] Just to make sure I'm doing the correct thing :) [17:03:07] Lucas_WMDE, Amir1, ^ [17:03:24] Checking [17:03:29] ooh, I see some familiar names in that list! :D [17:04:32] Is it possible for the updateCredits.php script to do this automatically? Or does it already use the .mailmap? I see it seems this data in .mailmap needs to be added manually? [17:04:37] Lucas_WMDE: Okay! :D [17:05:06] d3r1ck_: it already uses the mailmap… the remaining part, keeping the mailmap up to date, needs to be done manually [17:05:30] Alright! Understood [17:12:22] Amir1: At least Jenkins is happy :) [17:12:47] that's good :P [17:15:56] :) [17:16:16] Lucas_WMDE, Amir1, thanks so much for the clarification. I'm grateful :) [17:16:32] you are welcome [17:16:41] you’re welcome! [17:16:42] feel free to reach out to us at any time [17:17:15] Okay Amir1. I love this IRC meetings, they are really a catalyst in clarifying problems :) [17:17:19] and I just noticed that we’re already 15 minutes past the end of the meeting :D [17:17:23] *these [17:17:26] any final questions? :) [17:17:35] Lucas_WMDE: Yes, you are correct. I don't have any questions for now :) [17:18:52] okay, then we’re done! thank you everyone :) [22:09:39] not going to create a task on phab yet, but new Firefox 58 sometimes won't finish loading (a page is shown correctly but it still says loading), it happens just randomly sometimes. I noticed this already last year when I used Firefox 58 as beta. Firefox 57 worked fine.