[00:01:50] 3Multimedia, Mobile-Apps, MediaWiki-extensions-MultimediaViewer: Add a data- attribute with the file's title to [[File: elements - https://phabricator.wikimedia.org/T88825#1021987 (10Tgr) To reiterate the points from the discussion of the linked patch, the right format of a relative link to the file page can var... [00:06:43] bd808: thanks, changed [00:06:51] cool [00:07:44] Sentry is collecting errors from beta now [00:07:52] JS errors [00:34:22] sweet [00:37:50] tgr: are those 2 "foo is not defined" errors distinct or is sentry not quite rolling things up? [00:39:47] oh. it's trends and events [00:39:55] one event shown in 2 columns [00:40:21] yeah, confused me as well at first [00:42:53] click on an error, then on the stream tab, and you get a cleaner view [00:45:25] bd808, marxarelli: is there a way to import wikitext into vagrant without dump files? and if not, should be? [00:45:34] they kind of suck for code review [00:46:00] I only know of dumps [00:46:44] Yurik had a patch to make a new way to import at one point. we told him to make it a core feature and I think it got -1'd there and he abandoned it [00:48:19] Ah, not abandoned jsut stalled -- https://gerrit.wikimedia.org/r/#/c/154704/ [00:48:22] I was thinking of using maintenance/edit.php [00:48:57] Tim apparently didn't love the idea [00:49:32] yeah if you can come up with a way to do it with existing maintenance scripts that would be cool [00:54:47] bd808: what's the best way to handle updates to such a page? I imagine checking for differences via the API on every provision is not a great idea [00:55:12] just stick the textfiles in some directory somewhere and check against those? [00:55:21] the current dump loader just checks for page existance [00:55:44] yes, but that's a different use case [00:55:44] yeah doing a diff like that would work. [00:56:37] so you'd have a file resource in puppet to copy the file somewhere and then have it trigger running the import script if the file was updated [00:56:51] seems like a reasonable workaround [00:57:10] Put the files in /srv/ [00:58:00] what's the convention for multiwiki stuff? default wiki in top dir, others in ./wiki/ ? [01:04:34] the stuff in the top dir is "global" and then there is a wiki/ for each wiki in the farm [01:04:50] Ori hates that I made it that complicated [01:08:34] 3Multimedia, MediaWiki-extensions-Sentry: Catch uncaught exceptions with Sentry - https://phabricator.wikimedia.org/T88874#1022186 (10Tgr) 3NEW [01:10:50] 3Multimedia, MediaWiki-extensions-Sentry: Catch uncaught exceptions with Sentry - https://phabricator.wikimedia.org/T88874#1022197 (10Tgr) [01:10:51] 3Multimedia, MediaWiki-extensions-Sentry: Improve Sentry coverage - https://phabricator.wikimedia.org/T88399#1022196 (10Tgr) [01:17:39] 3Multimedia: Set up and test Sentry on Labs for JS error logging - https://phabricator.wikimedia.org/T1345#1022215 (10Tgr) [01:17:42] 3Multimedia, Beta-Cluster, MediaWiki-extensions-Sentry: Deploy Sentry extension on beta cluster - https://phabricator.wikimedia.org/T78807#1022212 (10Tgr) 5Open>3Resolved Done (thanks @MarkTraceur!) [01:22:09] 3Multimedia, WMF-Legal: Set up and test Sentry on Labs for JS error logging - https://phabricator.wikimedia.org/T1345#1022228 (10Tgr) Sentry is now live on the beta cluster and collecting JS errors (well, some of them - it will be more useful once T78809 is merged). Pinging #wmf-legal so they are aware as this i... [01:26:42] 3Multimedia, MediaWiki-Vagrant: Include documentation/test/showcase wiki pages in Vagrant roles - https://phabricator.wikimedia.org/T88356#1022233 (10Tgr) [01:26:48] 3Multimedia, MediaWiki-Vagrant: Import string into Vagrant - https://phabricator.wikimedia.org/T88872#1022234 (10Tgr) [01:50:34] 3Multimedia: Make a real, bonafide Flickr UI class - https://phabricator.wikimedia.org/T88880#1022289 (10MarkTraceur) 3NEW a:3MarkTraceur [01:50:52] 3Multimedia, MediaWiki-extensions-UploadWizard: Make a real, bonafide Flickr UI class - https://phabricator.wikimedia.org/T88880#1022298 (10MarkTraceur) [02:00:35] so marktraceur when is this rolling out for everyone? O_O [02:03:01] comets: Well...I dunno [02:03:06] How badly does Commons want it? [02:03:18] Because I can stop doing random refactors and fix Flickr [02:03:26] Well, pseudorandom [02:03:46] i don't want it to be released fully until all bugs are smashed.. [02:03:50] But I need to discuss it with the team, and if we decide to wait until next cycle it'll be a month or two [02:04:09] comets: Mostly I'm concerned about the brokenness of the system...there really needs to be a Flickr upload API module [02:04:34] Right now all the license checking happens on the frontend [02:04:36] Which is DUMB. [02:05:15] well they got rid of that i think.. [02:05:31] Did they? [02:06:17] lol i actually used the up/f for that ..i could upload licenses image and cro and do the next one, nowi have to wait an HOUR for a reviewbot to pass the image before i can "touch" it :/ [02:06:37] comets: I think we reverted that change [02:06:39] Try again? [02:06:51] reviewbots cannot process 'cropped' images :/ [02:07:09] it hasn;'t been reverted, i last used upwiz/f 2 days ago.. [02:07:55] Oh. Huh. [02:08:07] It should have been reverted...maybe I didn't backport it to wmf14 [02:08:09] Try now? [02:08:22] I think Commons is at wmf15 again (though actually let me check) [02:08:36] Yeah, try again [02:09:32] lemme find a nice image to upload and get back to you :D [02:10:07] i no longer use that on my firefox because it "forcedhttps" on all wikis and u know i only use http :) [02:16:57] Oh, news to me [02:17:13] comets: Actually Flickr made a change to use SSL only on their API [02:17:25] We straight up crashed the Flickr import for like two weeks before we figured that out [02:18:03] https on flickr is fine..i think i rememebr that day lol :P [02:36:52] 3Multimedia, MediaWiki-Vagrant, MediaWiki-extensions-Sentry: Clear SENTRY_DISALLOWED_IPS in the Sentry vagrant role - https://phabricator.wikimedia.org/T88068#1022368 (10Tgr) 5Open>3Resolved [17:38:36] 3Multimedia, MediaWiki-extensions-UploadWizard: Special:UploadWizard gives undescriptive error message upon too long filename - https://phabricator.wikimedia.org/T87314#1022831 (10Sn1per) So maybe the logic would be: Use newFromTitle() to check the title, if newFromTitle() returns false, use a validator function...