[04:04:09] 6Multimedia, 10MediaWiki-File-management: SVGs without XML prolog () served with Content-Type text/html from upload.wikimedia.org - https://phabricator.wikimedia.org/T131012#2154291 (10faidon) >>! In T131012#2154137, @matmarex wrote: > @faidon, is it possible to ask Swift about the oldest SVG files... [06:34:25] 6Multimedia, 6Commons, 10MediaWiki-File-management, 10MediaWiki-Uploading, 6Zero: Prevent Wikipedia Zero exploitation of uploads to share copyrighted media - https://phabricator.wikimedia.org/T130761#2154310 (10Gunnex) > In T130761#2149875, @Bawolff wrote on Thu, Mar 24, 10:20 PM: > Could someone link to... [08:56:05] 6Multimedia, 6Commons, 10MediaWiki-File-management: Support for multilingual SVGs - https://phabricator.wikimedia.org/T18052#2154387 (10Perhelion) (As suggested above: "//it's better to discuss the implementation in one place//") The seems to be a relatively low limit on for the lang selection (the drop-down... [09:31:32] 6Multimedia, 6Commons, 10MediaWiki-File-management, 10MediaWiki-Uploading, 6Zero: Prevent Wikipedia Zero exploitation of uploads to share copyrighted media - https://phabricator.wikimedia.org/T130761#2154392 (10Vituzzu) >>! In T130761#2154310, @Gunnex wrote: > > See my comment at T129845#2153512. > > E... [14:41:13] MatmaRex: Hey, I have a pretty simple 5-digit change (starting with 1) that could use review, I've rebased it and waiting for tests now: https://gerrit.wikimedia.org/r/11063 [14:41:27] Three versions have passed since I last rebased it [14:43:16] Actually wow this must have been my first original patch to MediaWiki code, anywhere [14:43:24] Or at least one of the first ten or so [14:44:11] Unless James_F wants to take unilateral product-y action [14:44:45] Given I've tested this on three separate versions of MediaWiki over the course of its life, and it's worked every time, I feel code review is a technicality at this point [14:46:50] :-) [14:47:10] Do it. [14:47:38] I mean, I also think you should at least push the button rather than me doing it alone [14:48:22] Self-review is icky. [14:48:34] * James_F re-reviews [14:50:44] * James_F C-1s. [14:51:21] Sad days [14:51:39] Oh, the bug number changed. [14:52:17] Yeah, the default value one is the main thing. [14:52:47] Done. [14:53:38] OK, C+1'ed; leaving to MatmaRex to C+2. [14:54:02] (Do you think people will be negatively surprised and want us to set it false in WMF production?) [14:54:30] I doubt it. [14:54:48] The bug didn't seem to have any preference on that count, and you're right, intuitively this makes sense [14:54:49] Though... [14:55:14] James_F: If people have expressly unchecked watchdefault and suddenly start getting notifications for uploads... [14:55:20] I mean, they'd need to upload things, but still [14:55:37] From the bug, "unwatchlisting is easy" [14:55:42] I still agree with that [14:55:42] Shrug [14:56:59] I think we don't need special treatment for WMF config from this. [14:57:26] No, true [14:57:32] I'm just saying in general [14:59:25] * James_F agrees. [16:47:48] sure, i'll look a bit later [16:48:00] looks simple enough [16:48:51] Practically trivial. [16:49:02] Nobody has objected to it in the nearly four years it's been sitting in Gerrit :) [18:02:18] MarkTraceur: I need your advice, please [18:03:10] https://www.youtube.com/watch?v=eWsxae3a9eY has a very cool feature that you can actuly tour within the tank, i'd like to import it into commons, as it has a free license [18:03:28] but i don't think the controls will work, is there a way to make them work ? [18:24:35] matanya: Hm, I don't know what that would entail. let me try it locally and see what happens. [18:25:01] thanks MarkTraceur [18:25:05] i uploaded it to https://commons.wikimedia.org/wiki/File:Virtually_Inside_the_Tanks_-_T-55_%26_Type_59.webm [18:26:18] Yeah, it doesn't work for me on mobile [18:27:35] matanya: https://support.google.com/youtube/answer/6178631?hl=en maybe this helps? [18:27:47] I don't know if you need to re-add the metadata after you download the video... [18:28:17] Or https://support.google.com/youtube/answer/6316263 [18:55:58] 6Multimedia, 6Performance-Team: Profile single and multi-chunk upload time for end-users - https://phabricator.wikimedia.org/T130659#2142297 (10Krinkle) p:5Triage>3Normal [19:22:23] MarkTraceur: would a ticket help figuring this out ? [19:34:57] matanya: Did you check the links I sent you? [19:49:38] 6Multimedia, 10MediaWiki-Uploading, 10MediaWiki-extensions-Gadgets: HotCat in Special:Upload: double category line - https://phabricator.wikimedia.org/T131089#2155707 (10SJu) [19:53:15] MarkTraceur: got time for https://gerrit.wikimedia.org/r/#/c/277847/ today? [19:53:27] (03PS11) 10Bartosz Dziewoński: uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) [19:53:32] MatmaRex: I will look at it next! [19:55:10] (03CR) 10jenkins-bot: [V: 04-1] uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) (owner: 10Bartosz Dziewoński) [19:55:29] hmm, that's a weird -1 [19:55:30] Or not [20:00:39] oh wait, that's an embarassing mistake in the test. [20:00:39] i wonder how it ever worked. [20:02:47] 6Multimedia, 10MediaWiki-Uploading, 10MediaWiki-extensions-Gadgets: HotCat in Special:Upload: double category line - https://phabricator.wikimedia.org/T131089#2155771 (10Legoktm) 5Open>3Invalid This is a bug with the HotCat gadget, please report it at (03PS12) 10Bartosz Dziewoński: uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) [20:16:57] MarkTraceur: i'm not entirely sure if the uw.ConcurrentQueue tests are stable – if UW tests start randomly failing a lot after we merge this, check if it's these tests first :D [20:17:04] (it passed this time though) [20:21:00] 6Multimedia, 10MediaWiki-Uploading, 10MediaWiki-extensions-Gadgets: HotCat in Special:Upload: double category line - https://phabricator.wikimedia.org/T131089#2155707 (10matmarex) Previously reported: https://commons.wikimedia.org/wiki/Commons:Village_pump#Bug_on_Commons:Upload [20:32:23] 6Multimedia: Missing pronunciation file (Nl-gegourmet.ogg) - https://phabricator.wikimedia.org/T131096#2155883 (10StephenLester) [20:41:48] MatmaRex: What might cause them to be unstable? [20:42:25] (hm, randomness in tests makes me think we aren't testing everything) [20:44:19] MarkTraceur: i'm not sure if the order of the setTimeout() calls from ConcurrentQueue and setTimeout() calls from the tests is deterministic, and some tests depend on it [20:44:45] the random delay time shouldn't matter, but i guess i could just put some small number in there [20:45:35] I mean, why use a random time? Shouldn't it be fine to use a specific timeout so the tests are the same each time? [20:45:59] it should [20:46:04] ok, let's do that [20:47:26] At least that way it will fail every time if something is wrong [20:48:04] If different timing causes different results, we should be testing different timing, not depending on randomness to pick up on each possible combination [20:50:02] hmm, actually. it might not be fine :P i've written some of the tests so that they assume that the actions don't all finish at the same time. bahhh [20:50:04] * MatmaRex grumbles [20:50:22] i really don't wanna rewrite them all. [20:50:23] hmm [20:50:24] Heh [20:50:40] The "deleting items while the queue is running" one looks pretty suspect [20:52:35] how so? [20:52:50] (03PS13) 10Bartosz Dziewoński: uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) [20:52:58] Hmm maybe not [20:52:59] ^ ok, the delay is deterministic now [20:53:14] MatmaRex: I'm still tying strings together in my head over here [20:56:09] (03PS14) 10Bartosz Dziewoński: uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) [20:56:09] I need two coat racks, three different colours of post-its, and about 200 metres of string [20:56:09] Then maybe these tests will be easy to understand [20:56:09] (03PS15) 10Bartosz Dziewoński: uw.controller.Step: Refactor simultaneous transitions [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/277847 (https://phabricator.wikimedia.org/T92809) [20:56:09] (one last comment tweak) [20:56:14] MarkTraceur: yeah. they're horrible. [20:56:26] They're not bad, just complex [20:57:26] MatmaRex: OK, so we assume in these tests that the tasks added first will finish first, which is not necessarily the case in MTV's The Real World. [20:57:53] MatmaRex: And if they don't finish in order, the tests would break (bad) but the real code would not (so it's OK we don't test it)? [20:57:54] MarkTraceur: no, we don't. we only assume that tasks added first will start first. [20:58:13] they can finish in any order and the tests should pass fine [20:58:35] So the comment at lines 23-24 is just for show? [20:59:08] i could change incrementallyDelayedPromise() to decrementallyDelayedPromise() or something. [20:59:21] that could actually be a better test, heh [20:59:23] The only thing that matters is they don't all finish at once? [21:00:46] yes. some of these tests want to do stuff between two subsequent 'progress' events [21:00:48] e.g. "Deleting items while queue running" [21:13:03] Oh, and with a tie, it might turn out that the process you want to delete is the one that wins the tie [21:13:30] Or...something [21:14:23] * MarkTraceur gets caffeine [21:23:13] OK this is making way more sense to me now [21:23:17] I'm not sure what I missed before. [21:25:02] OK, the tests seem fine to me [21:25:06] I doubt they will be unstable [22:49:55] MatmaRex: You'll have to forgive 2012 MarkTraceur for not understanding MediaWiki's upload system [22:50:03] ;) [22:52:01] nobody understands MediaWiki's upload system [22:52:10] because once you understand it [22:52:11] you go mad [22:52:35] it's kind of like embracing cthulhu [22:53:23] That explains why I haven't seen many FB posts from AaronSchulz lately [22:53:40] :) [23:00:56] we stole your room, james is coming [23:02:21] Gosh quiddity [23:04:45] I'm running a little behind for standup [23:04:58] but it occurs to me that perhaps we should shift the timing this week [23:05:11] unless MatmaRex likes it this late? :) [23:05:16] MatmaRex: Yunocome? [23:05:55] i'm not here [23:06:01] * MatmaRex makes the not-the-droids motion [23:12:53] MarkTraceur: https://www.youtube.com/watch?v=NJ6kJ7GWtv0 [23:13:15] Oh my. [23:14:02] Indeed. [23:14:14] The bridge is fantastic. [23:20:31] ok, looks like flights should work out for the iiif thing \o/ [23:53:19] 6Multimedia, 6Commons, 6Design-Research-Backlog, 10UploadWizard, 7Voice & Tone: Review UploadWizard UX for "fatal user errors" - https://phabricator.wikimedia.org/T77820#830918 (10Jdforrester-WMF) >>! In T77820#2151464, @ggellerman wrote: > Is there a timeframe in which you would want Design Research to... [23:53:36] brion: (Yay.)