[06:21:58] Good morning Prague [06:23:01] good morning [07:30:54] Dobré ráno - Help desk is open 🎉 [08:24:35] Is it possible to get a Wikidata diff as a json? https://www.wikidata.org/w/index.php?title=Q1980992&curid=1909765&diff=944125992&oldid=932465242 [08:34:18] What's the best way to try to get someone to review something in gerrit? [08:35:04] tobias47n9e: not as far as I know [08:40:01] Complain a lot :P [08:40:53] tobias47n9e: https://phabricator.wikimedia.org/T218779 [08:40:56] :/ [08:44:18] Hey zhuyifei1999_, could you send me the link to the milestone about deleting tools please? [08:45:56] waldir: https://phabricator.wikimedia.org/project/view/3975/ [08:46:19] Cool, thanks! [08:54:52] zhuyifei1999_: am I supposed to remove myself as a maintainer of the tool once I add "Owner of abandoned tools" as maintainer? [09:05:30] Phabricator intro talk starting in Education Center, second floor [09:07:20] hey @eprodromou are you still 'having problems with my Wikidata SPARQL query'? [09:08:00] Can anyone tell me about MCR? [09:10:41] Isarra: Daniel Kinzler might be a good person to start with...but others can also talk about MCR, just depends on the questions you have :) [09:11:50] zeljkof: oh oh. Forgot about our 11 am thingy on https://phabricator.wikimedia.org/T212558 [09:11:52] still around ? [09:12:26] My questions are largely 'why it be like that'. [09:12:34] So he probably would be the one. [09:12:49] Based on the rfc. [09:14:39] So yeah. Thanks. [09:14:45] Isarra: I haven't seen Daniel yet this morning, but will try and track him down :) [09:19:11] zeljkof: on the bright side, I cannot find you :D [09:53:11] Intro to Cloud Services starting in about 10 minutes in the Education Center, second floor [09:55:49] Krinkle: longma and I are having some issues with building images with docker-pkg on macOS, would you have some time to help us or could you point us to someone who can? [09:56:10] I'm getting "OSError: Could not find a suitable TLS CA certificate bundle, invalid path: /etc/ssl/certs/ca-certificates.crt", and I had some workaround for this in the past but have no idea what I did [09:56:20] Let's say "Ask me anything about Cloud Services" too! Come talk to me about Toolforge, Wiki Replicas, Cloud VPS, etc! [09:56:22] or maybe James_F knows? [10:03:58] c++ problem arrived at help desk. anyone a master of c++? [10:05:41] bawolff, Ostrzyciel: can I just assign figuring out how to handle the threads as a block (in terms of content model, slots, tables, whatever) to you guys? >.> [10:05:57] tobias47n9e: I'm not at the hackathon, but I might be able to help [10:06:04] tobias47n9e: depends a bit on what the question is :P [10:07:23] valhallasw`cloud: Thanks for the response. Someone is already helping. 🏅 [10:07:29] (y) [10:07:33] bawolff, Ostrzyciel: With the single requirement that it make sense? [10:07:35] I mean maybe...im not sure if ill actually be working on that after this hackathon [10:08:01] Honestly if we can work out the ridiculous storage crap, I can probably take over most of how we interact with it later? >.> [10:09:56] suuure [10:10:08] we just need a whiteboard or something probably [10:10:18] and a lot of coffee/beer [10:13:17] i think i get the general idea now [10:14:22] Yay! Because I don't. Backend is scary. [10:14:53] DBs are easy [10:15:18] What about abstraction? You know folks are doing away with page and stuff, or something? [10:38:47] \o/ Can I get some help on deleting one of these OAuth consumers on https://meta.wikimedia.org/wiki/Special:OAuthConsumerRegistration/list [10:39:18] tonythomas: which one? [10:40:03] tgr_: both of them there :) [10:41:13] can you link to them? [10:41:47] 79566565b55f9a12e4586ed784b68359 and cab61c5eb153595cbde11cf9550ec17c I guess? [10:42:39] tonythomas: I can reject them, is that what you mean by deleting? [10:46:47] tgr_: exactly. thank you [10:47:17] is there an efficient way to get all translations of a single interface message? [10:47:54] tonythomas: does it matter which one? they are not quite identical [10:48:24] I know. cab61c5eb153595cbde11cf9550ec17c and 79566565b55f9a12e4586ed784b68359 were unplanned improvements. [10:48:43] in fact either of them had the right configuration I was looking for :-( [10:48:51] would be creating one straightaway once this is clean. [11:02:01] Krinkle / James_F more questions about profilers, if you have a few minutes to discuss with me and longma [11:37:27] tgr_: status of both still remains 'proposed' btw. [12:03:22] tonythomas: sorry, got distracted. Done. [12:03:28] tack. [12:04:47] INaturalist and Wikidata session on 4th floor [12:17:38] help again: Is it possible for me to use a database for my app on tools server ? Couldnt find any straightforward documentation [12:17:57] tonythomas: yes [12:18:05] chicocvenancio: the best. thanks! [12:18:08] there is a mysql isntance provided [12:18:29] where are you? I'm in the chill out zone by the stairs [12:18:41] coming over there in a minute then. [13:05:14] Psssst! https://t.me/whatshackening for some low-volume updates [13:05:27] Also, if there's anything you want me to share on here, let me know! [13:20:25] Reedy: https://gerrit.wikimedia.org/r/c/mediawiki/core/+/491455 [13:32:08] Guys, we need help. Naming our content models. [13:32:50] THICC apparently has two things: sections (like on a talkpage), and... like, the entire talkpage. With all the sections, [13:32:51] . [13:33:05] But what do we call them? I can't think of anything sufficiently tasteless. [13:33:40] Isarra: Flow uses "conversation" and "board", iirc. [13:33:49] Bear in mind the... entire talkpage equivalent, here, is not intended to actualle be the entire page, just a slot on the page. So it's literally just a bunch of sections. [13:33:57] Flow is the entire page. This is just... uuuuh part of it? [13:34:08] Isarra: by the way, this seems to confirm that each conversation should be on a separate page: https://www.mediawiki.org/wiki/Talk_pages_consultation_2019/Phase_1_report#Experienced_contributors [13:35:03] Well, that's good, since that seems to be the only way to do it from a technical standpoint regardless. >.> [13:35:14] Without it getting really messy, anyway. [14:46:39] tonythomas: did it work? [15:10:29] chicocvenancio: well, we got 3.4 up and running. However, my app has some requirements conflict on that version. [15:10:37] Probably downgrading some dependencies would work. [15:10:40] will do it soon. [15:10:52] thanks for the help anyway. [15:11:07] and we need Python > 3.4 on toolforgee pleeeeeseee [15:19:10] tonythomas: https://gerrit.wikimedia.org/r/#/c/operations/software/tools-webservice/+/496265/ [15:19:13] It's coming... [15:19:27] I think bd808 didn't get through all the deploy process [15:19:35] wohoooo [15:19:46] because it's a lot of stuff rebuilding containers and stuff [15:20:23] tonythomas: Oh, apparently, it should be all out now... [15:20:46] happy I postponed downgrading to 3.4 for post-beers. [15:26:03] hey, does anyone know of any extensions that have namespace aliases? [15:27:44] JADE seemingly [15:28:12] it's now Jade :P (i think). [15:35:17] huh [15:35:23] nvm, we figured that out [15:36:02] but another thing is, me named the namespace THICC: and it shows up like that in URL and everywhere except page title [15:36:06] not sure why [15:36:18] may be hardcoded into core or something... [15:36:33] (in page title it's Thicc:) [15:39:10] Reedy: I have this vague understanding (based on a report Ladsgroup wrote for me) that there is this schema drift between what's in the Jade repository and what's in Beta Cluster. What... would cause that? [15:39:28] Bad schema updater code in Jade? [15:40:31] What's missing? [15:40:39] Possibly. Anyways, we are changing the namespace name for a third time, and it sounds like things are all messy anyway, and Beta Cluster is Beta Cluster, so I am wondering if it would be worthwhile to undeploy Jade and re-deploy it at such a time [15:41:04] How would that necessarily help? [15:42:49] In that I don't really know what is causing the problem. If it's not that complicated an issue I guess it could be fixed with a patch. (Please, please do not make this your project.) [15:43:07] I'm doing whatever I feel like [15:43:10] Is there a task on it? [15:46:18] ...it would help if I looked at our actively curated team backlog and not our not-curated Jade backlog :P [15:46:49] Holy cow, the backlog exploded overnight. Hackathons! [15:50:10] it doesn't look like there is one. i will make quickly [15:50:22] I imagine it's probably easy enough to fix [17:13:07] yurb: you are not around at the moment? [18:00:07] To anyone I didnt say goodbye: bye [18:00:37] *everyone [18:01:09] ? [18:01:35] * Dvorapa|hack & his English struggle [18:03:40] I think either works :) [18:04:02] 😁 thx [18:09:23] Reedy: sorry for the delay, https://phabricator.wikimedia.org/T223747 [18:39:02] hare: https://gerrit.wikimedia.org/r/#/c/mediawiki/extensions/Jade/+/511080/ [18:46:05] So... I have the Wikipedia iOS app on my iPad... and I go to the... whatever the screen left of the home screen is called... and I have those widgets that give me the Wikipedia Top Read articles, as furnished by the app. But I click on one of the links and it opens Wikipedia in Safari. I wonder what the deal is. [19:23:50] I will let you know when I see tonythomas and I will deliver that message to them [19:23:50] @notify tonythomas did you get the python3.5 version issue sorted out? [20:01:00] zhuyifei1999_, https://github.com/wikimedia/revscoring/pull/438 [20:01:02] <3 [21:26:02] tramm: I'm back [22:48:53] yurb: just looked at eventmetrics which seems to use some direct wikimedia database connection and wikigrades doesn't use any of the existing apis either, why so? [22:49:57] tramm: wikigrade uses the standard mediawiki api - /w/api.php [22:50:26] * tramm meant api wrappers for python [22:52:06] tramm: I see. Well, for my task (running simple queries against mediawiki api) pywikibot was too heavy, so I wrote my own little wrapper: https://gitlab.com/yurb/wikigrade/blob/master/evaluate/littlemw.py [22:53:23] As for why I didn't use mwapi, this pull request should explain it;) https://github.com/mediawiki-utilities/python-mwapi/pull/40 [22:54:03] I think I will replace it with wmapi eventually, as they ended up veeery similar to use [22:55:22] (it'd be greate if wmapi had unit tests though...) [22:55:33] s/greate/great/ [22:55:56] i wonder if i should do the same, just that in the end i need quite many different queries about number of authors, discussion page lengths etc, so i feel for clarity it would be good to use the most standard lightweight wrapper [22:57:02] tramm: I think the idea behind mwapi is to allow any kind of query that the api provides [22:57:55] but there are some 4-5 different wrappers available [22:58:05] yeah [22:59:29] but since the mediawiki-utilities project (one of which is mwapi) is used by people at WMF, I think it should be the safest bet (I may be wrong though) [23:19:19] ok, thanks, yurb, will see about that [23:19:36] np. what is your project by the way, tramm?