[00:00:23] (03PS7) 10MarkTraceur: Add remote image repository information to list [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/88318 [00:03:27] (03PS3) 10Aaron Schulz: lockmanager: QuorumLockManager subclasses can get EX/SH locks at once [core] - 10https://gerrit.wikimedia.org/r/69021 [00:03:48] (03CR) 10Aaron Schulz: "Removed var_dump()" [core] - 10https://gerrit.wikimedia.org/r/69021 (owner: 10Aaron Schulz) [00:05:51] (03CR) 10MaxSem: "recheck" [core] - 10https://gerrit.wikimedia.org/r/80578 (owner: 10MaxSem) [00:06:00] (03CR) 10Tim Starling: [C: 04-1] "If HashRing::newWithoutLocation() returns false, it seems like tryJobInsertions() will give a fatal error. And the $partitionRing variable" [core] - 10https://gerrit.wikimedia.org/r/69448 (owner: 10Aaron Schulz) [00:11:11] (03PS1) 10Adamw: Update CentralNotice extension [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90459 [00:11:47] (03CR) 10Adamw: [C: 032] Update CentralNotice extension [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90459 (owner: 10Adamw) [00:11:54] (03CR) 10Adamw: [V: 032] Update CentralNotice extension [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90459 (owner: 10Adamw) [00:15:31] (03CR) 10MarkAHershberger: "When I looked, this was the primary use for an XML parser. Maybe the onl" [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [00:20:43] TimStarling: yeah it should throw some jobqueue exception in that case instead (and an "if($partitionRing)" also needed in the body of the tryInsert method) [00:32:07] (03PS1) 10Dzahn: delete Search.pm, it's not custom anymore [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90461 [00:35:43] (03PS5) 10Aaron Schulz: jobqueue: various cleanups to JobQueueFederated [core] - 10https://gerrit.wikimedia.org/r/69448 [00:36:26] (03CR) 10jenkins-bot: [V: 04-1] jobqueue: various cleanups to JobQueueFederated [core] - 10https://gerrit.wikimedia.org/r/69448 (owner: 10Aaron Schulz) [00:39:41] (03PS6) 10Aaron Schulz: jobqueue: various cleanups to JobQueueFederated [core] - 10https://gerrit.wikimedia.org/r/69448 [00:39:52] (03CR) 10Aaron Schulz: "Rebased" [core] - 10https://gerrit.wikimedia.org/r/69448 (owner: 10Aaron Schulz) [00:41:26] (03PS1) 10Aaron Schulz: Removed unused const QoS_Atomic constant (odd casing) [core] - 10https://gerrit.wikimedia.org/r/90463 [00:48:21] chrismcmahon: Don't suppose you've any experience unit testing apache config/rewrites have you? [00:49:44] Reedy: not my regular beat, but I'm pretty aware of the literature and I've never heard of anyone doing that [00:50:25] also, is there a standard for Tabs/4-spaces/other for Apache configs? [00:50:30] style guide [00:51:37] Reedy: "The maximum length of a line in the configuration file, after environment-variable substitution, joining any continued lines and removing leading and trailing white space, is 8192 characters." :) [00:51:46] bahaha [00:51:56] Lets minify our apache configs! [00:52:16] hrhr [00:52:36] but jenkins already runs apachectl configtest right [00:52:57] I think it does something like that [00:53:39] /usr/sbin/apache2 -t -d /srv/ssd/jenkins-slave/workspace/operations-apache-config-lint -C 'Include /etc/apache2/mods-enabled/*.load' -C 'Include /etc/apache2/mods-enabled/*.conf' -C 'Include /etc/apache2/mods-available/expires.load' -f all.conf [00:53:43] yep, it does [00:53:51] -t is the same thing [00:54:01] https://integration.wikimedia.org/ci/view/Operations/job/operations-apache-config-lint/ [00:54:15] yea https://integration.wikimedia.org/ci/job/operations-apache-config-lint/293/console [00:58:15] http://viget.com/extend/test-drive-mod-rewrite-rules-with-testunit [00:58:24] That involves the r word though :( [01:00:34] (03PS1) 10Adamw: Revert "Update CentralNotice extension" [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90467 [01:00:46] (03CR) 10Adamw: [C: 032 V: 032] Revert "Update CentralNotice extension" [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90467 (owner: 10Adamw) [01:13:11] (03PS1) 10Adamw: update CentralNotice extension [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90471 [01:13:29] (03CR) 10Adamw: [C: 032 V: 032] update CentralNotice extension [core] (wmf/1.22wmf21) - 10https://gerrit.wikimedia.org/r/90471 (owner: 10Adamw) [01:15:10] (03CR) 10MZMcBride: "I've reöpened bug 26792 as I can't find a use-case that justifies adding yet another global configuration variable." [core] - 10https://gerrit.wikimedia.org/r/89758 (owner: 10Reedy) [01:32:24] (03CR) 10Tim Starling: [C: 032] jobqueue: various cleanups to JobQueueFederated [core] - 10https://gerrit.wikimedia.org/r/69448 (owner: 10Aaron Schulz) [01:35:40] (03Merged) 10jenkins-bot: jobqueue: various cleanups to JobQueueFederated [core] - 10https://gerrit.wikimedia.org/r/69448 (owner: 10Aaron Schulz) [02:05:55] (03PS1) 10Legoktm: Create a new Flow topic if the page is occupied [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/90474 [02:20:46] (03CR) 10PleaseStand: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/76714 (owner: 10Ebrahim) [03:44:21] (03CR) 10Mattflaschen: "See also https://gerrit.wikimedia.org/r/#/c/79555/" [extensions/GuidedTour] - 10https://gerrit.wikimedia.org/r/82208 (owner: 10AndyRussG) [03:58:58] (03CR) 10Mattflaschen: "We are considering animations; see bug 55771. I haven't tested this particular patch. It sounds a bit different from that proposed in th" [extensions/GuidedTour] - 10https://gerrit.wikimedia.org/r/82209 (owner: 10AndyRussG) [03:59:28] (03CR) 10Mattflaschen: "Sorry, I meant describe it at the bug, not here." [extensions/GuidedTour] - 10https://gerrit.wikimedia.org/r/82209 (owner: 10AndyRussG) [04:07:27] Hi superm401, I see you're digging into some submissions I made to GuidedTour. Just thought I'd reach out and let you know I'm around in case you feel like discussing anything [04:09:14] Hey AndyRussG, wanted to make a couple notes [04:09:42] I have a few things on my plate, so it may be a little while before I can delve in and start testing and reviewing. [04:09:57] Also, I'll only be working a part day tomorrow since I'm taking some vacation time. [04:10:55] That's totally cool, I don't think it's urgent... I heard you guys were very busy [04:11:15] I'm more than happy to do the formatting cleanup [04:12:28] Re: the animation, I didn't feel I got it quite as nice as it could be, though I think the general approach (CSS3 transitions) was right [04:17:06] I don't know if you heard, I'm now working on the Education Program extension, part-time... [04:17:35] I wish you the best of luck with that :P [04:17:41] Are you converting it to ContentHandler? [04:18:11] superm401: In fact, Sage mentioned the idea of a future rewrite or replacement for the extension using GT [04:20:26] legoktm, thanks! There are a lot of bugs and feature requests that depend on converting it to ContentHandler, but I just started and am working on urgent bugs that aren't related to CH. [04:21:48] I think the next task after that will be to make a medium-term roadmap [04:22:55] Including, I expec, a decision on when and how to decide where to go with this codebase [04:23:08] Yeah, that sounds like a good idea [04:25:55] AndyRussG, no, I didn't know you were working on the Education extension. That's interesting. [04:27:58] Yeah...! That's what I said when they offered me the role [04:30:15] Since I'd applied for a job with the Growth team, I guess I assumed you guys had heard [04:33:05] Nope, we don't work on the Education extension, and didn't hear it through the grapevine earlier. [04:33:13] At least I didn't. :) [04:33:34] It's a part-time contract, now on my second week... Terry said I could still be "in the mix" for a full-time position, which I'm still quite interested in [04:35:47] Hoping to spend some hours volunteering, too, in addition to the hours in the contract... So if there are things you guys think I might be able to help out with, please don't hesitate to reach out [04:37:19] That is--if you feel it's appropriate. Once the job interview process got going, I kinda shied away from the IRC channel, since I didn't want to get in the way of any internal processes/discussions you might have had about hiring [04:38:00] We generally don't discuss individual candidates in the channel. [04:38:11] So feel free to come by if you want to talk about stuff you might want to work on. [04:39:15] OK...! That sounds fantastic [04:43:54] I have a bunch of ideas for stuff that I think might be useful to Education and maybe Growth and Features... I'll definitely stop by then [05:09:08] Elsie/MZMcBride, besides the actual issue here, your personal attacks at https://bugzilla.wikimedia.org/show_bug.cgi?id=40040 were definitely over the line. [05:09:22] superm401: Report me to my supervisor. [05:14:42] superm401: Unfortunately I can't change the reporter field on that bug. [05:14:50] Otherwise I would just ignore you and Steven completely. [05:15:16] The form is bad, but I don't need the headache of arguing about where to put wiki pages. [05:15:23] You're wrong, but I'm okay with that. [05:28:11] (03CR) 10Zfilipin: [C: 04-1] "If we decide to restore this commit, I would prefer to restore it in ULS repo (with optionally deleting existing tests there)." [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90380 (owner: 10Cmcmahon) [05:28:17] superm401: I'm off, have a great vacation if I don't see you before you go [05:33:52] AndyRussG, and you have a good weekend. [05:58:49] (03CR) 10PleaseStand: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/89610 (owner: 10PleaseStand) [06:00:41] (03PS2) 10PleaseStand: Installer: Run the LESS compiler for screen.less [core] - 10https://gerrit.wikimedia.org/r/89610 [06:01:03] (03CR) 10PleaseStand: [C: 04-1] "I haven't done the "future-proofing" yet." [core] - 10https://gerrit.wikimedia.org/r/89610 (owner: 10PleaseStand) [06:07:43] (03PS1) 10Pastakhov: Add CONTINUE and BREAK operators [extensions/Foxway] (develop) - 10https://gerrit.wikimedia.org/r/90492 [06:12:03] (03CR) 10Pastakhov: [C: 032] Add CONTINUE and BREAK operators [extensions/Foxway] (develop) - 10https://gerrit.wikimedia.org/r/90492 (owner: 10Pastakhov) [06:12:10] (03CR) 10jenkins-bot: [V: 04-1] Add CONTINUE and BREAK operators [extensions/Foxway] (develop) - 10https://gerrit.wikimedia.org/r/90492 (owner: 10Pastakhov) [06:43:35] (03CR) 10Alex Monk: [C: 04-1] "I already suggested something like this in Ib40ea868" [extensions/AbuseFilter] - 10https://gerrit.wikimedia.org/r/90457 (owner: 10Werdna) [06:54:30] (03PS1) 10Legoktm: Log rev_id to EventLogging [extensions/Echo] - 10https://gerrit.wikimedia.org/r/90497 [06:56:40] (03CR) 10Legoktm: "Schema change is https://meta.wikimedia.org/w/index.php?title=Schema%3AEcho&diff=6081131&oldid=5423520" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/90497 (owner: 10Legoktm) [07:09:05] (03PS1) 10Aaron Schulz: Small JobQueueFederated performance tweaks [core] - 10https://gerrit.wikimedia.org/r/90499 [07:10:15] (03PS3) 10Ebrahim: Tweak bidi on Special:Contributions/newbies [core] - 10https://gerrit.wikimedia.org/r/76714 [07:12:40] (03CR) 10Ebrahim: "This is just replicated what is available on https://github.com/wikimedia/mediawiki-core/blob/ebb45376df65392b78c23be813f1e34aa4eeadd4/inc" [core] - 10https://gerrit.wikimedia.org/r/76714 (owner: 10Ebrahim) [07:22:19] (03PS1) 10Legoktm: Remove "Talk:" prefix from "You have new messages" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/90500 [07:46:38] (03CR) 10Netbrain: [C: 032 V: 032] Tagged v0.1, did various cleanup and started on v 0.2 [extensions/SideBarMenu] - 10https://gerrit.wikimedia.org/r/90174 (owner: 10Netbrain) [08:20:37] (03PS1) 10Hashar: doc.wm.o: migrate index page a standalone default.html [integration/docroot] - 10https://gerrit.wikimedia.org/r/90503 [08:20:47] (03CR) 10Hashar: [C: 032] doc.wm.o: migrate index page a standalone default.html [integration/docroot] - 10https://gerrit.wikimedia.org/r/90503 (owner: 10Hashar) [08:20:49] (03Merged) 10jenkins-bot: doc.wm.o: migrate index page a standalone default.html [integration/docroot] - 10https://gerrit.wikimedia.org/r/90503 (owner: 10Hashar) [08:23:42] !deployment [08:23:43] There is no such key, you probably want to try: !deploymentqueue, !deployments, !hetdeploy, !queue, !wmf-deployment, !roadmap, [08:23:47] !deployments [08:23:47] http://wikitech.wikimedia.org/view/Software_deployments [08:23:53] !deploymentqueue [08:23:54] http://www.mediawiki.org/wiki/Deployment_queue [08:24:07] !deploymentqueue del [08:24:08] Successfully removed deploymentqueue [08:24:12] !alias deploymentqueue deployment [08:24:18] !alias deployments deployment [08:24:27] !deploymentqueue alias deployments [08:24:28] Created new alias for this key [08:24:35] !deployment alias deployments [08:24:35] Created new alias for this key [08:24:39] !deploymen [08:24:39] There is no such key, you probably want to try: !deployments, !hetdeploy, !queue, !wmf-deployment, !roadmap, [08:24:47] !wmf-deployment [08:24:47] https://gerrit.wikimedia.org/r/#/admin/groups/21,members [08:25:06] !deployers alias wmf-deployment [08:25:06] Created new alias for this key [08:25:11] !alias deployer alias deployers [08:25:16] !queue [08:25:16] Ambiguous! Try !reviewqueue, !jobqueue, or !deploymentqueue. [08:25:20] ... [08:25:27] !deploymentqueue [08:25:27] http://wikitech.wikimedia.org/view/Software_deployments [08:52:44] hashar, can you help me with a jenkins weirdness? I have a unit test that works everywhere but jenkins, as if one of files wasn't updated [08:55:24] MaxSem: in a few minutes. Can you point me the Gerrit change / job issue? [08:55:30] some workspaces are not properly cleaned out :( [08:56:12] hashar, thanks: https://gerrit.wikimedia.org/r/#/c/80578/ [08:58:16] (03PS1) 10Hashar: 'Cover' section listing code coverage reports [integration/docroot] - 10https://gerrit.wikimedia.org/r/90509 [08:58:31] (03CR) 10Hashar: [C: 032] 'Cover' section listing code coverage reports [integration/docroot] - 10https://gerrit.wikimedia.org/r/90509 (owner: 10Hashar) [08:58:33] (03Merged) 10jenkins-bot: 'Cover' section listing code coverage reports [integration/docroot] - 10https://gerrit.wikimedia.org/r/90509 (owner: 10Hashar) [09:02:40] MaxSem: looking [09:03:25] ahh [09:03:25] https://integration.wikimedia.org/ci/job/mediawiki-core-phpunit-parser/13119/console : FAILURE in 1m 57s [09:03:26] bah [09:05:02] (03PS2) 10Dan-nl: second source field mapping dissapears [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90041 [09:05:58] (03CR) 10Dan-nl: "rebased against master and updated the commit message" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90041 (owner: 10Dan-nl) [09:06:02] (03CR) 10Hoo man: [C: 04-2] "We already decided that we don't want this. If you really need this, please tell us why and point us to the bug." [extensions/AbuseFilter] - 10https://gerrit.wikimedia.org/r/90457 (owner: 10Werdna) [09:06:05] MaxSem: yeah that test pass on my local install as well [09:08:26] (03PS7) 10Hashar: Make TOC hideable [core] - 10https://gerrit.wikimedia.org/r/80578 (owner: 10MaxSem) [09:15:11] (03CR) 10jenkins-bot: [V: 04-1] Make TOC hideable [core] - 10https://gerrit.wikimedia.org/r/80578 (owner: 10MaxSem) [09:15:21] hashar, fail:P [09:15:28] Time: 1.73 minutes [09:15:29] huhu [09:19:55] MaxSem: I snapshotted the workspace and trying to reproduce locally [09:22:01] bah [09:22:02] they pass [09:22:04] wtf [09:23:56] MaxSem: so they pass with tests/parsertests.php [09:25:51] MaxSem: is "notoc" a new option in parser tests ? [09:25:59] yup [09:26:08] might need to register it in phpunit as well [09:26:17] I added a math option recentyly [09:26:21] let me look up [09:27:42] hmm [09:27:57] tests/phpunit/includes/parser/NewParserTest.php set $wgMaxTocLevel [09:28:02] 'wgMaxTocLevel' => $maxtoclevel, [09:28:04] so [09:28:13] when using the tests/parserTest.php the test is passing [09:28:21] aarghhhhhhhhhhhhhhhhhhhhhhhhhhhhhh [09:28:23] but php phpunit.php --group Parser [09:28:23] does fail on my local install [09:28:26] thanks hashar! [09:28:51] dammit, parser tests are a mess [09:28:55] you can have a look at how the math option got setup https://gerrit.wikimedia.org/r/69872 [09:29:04] the Math extension used it with https://gerrit.wikimedia.org/r/#/c/69873/ [09:29:09] core commit is 2fe1b773d3da27e379485e049c239d9ac6cd6365 [09:29:11] and yea [09:29:14] parser tests are crazy [09:30:05] MaxSem: don't we already have a magic keyword to disable toc? __NOTOC__ ? [09:30:43] hashar, this is different - the trick is to make parser cache not vary on TOC [09:31:05] for extensions interested in this stuff like CirrusSearch and MobileFrontend [09:31:08] do document that in the commit message so :-] [09:31:19] I thought it mentions that [09:31:35] ah yeah [09:32:02] I have no clue how the parser cache fragmentation works :/ [09:33:16] MaxSem: also would be nice to avoid the extra newline introduce in the HTML output [09:34:00] it doesn't change anything in page rendering [09:34:25] https://gerrit.wikimedia.org/r/#/c/80578/7/tests/parser/parserTests.txt,unified has a bunch of newlines being added [09:35:23] yup [09:35:45] but the result you see in browser is the same, just one pecularity replaced with another one [09:47:46] (03PS8) 10MaxSem: Make TOC hideable [core] - 10https://gerrit.wikimedia.org/r/80578 [10:36:54] (03PS1) 10Addshore: Add @covers for main Database test types [core] - 10https://gerrit.wikimedia.org/r/90518 [10:49:15] (03CR) 10Aklapper: [C: 031] "After the usual "bzr up" on bugzilla 4.2 upstream checkout, "git pull --rebase" on master of our modifications git repository, and diff'in" [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90461 (owner: 10Dzahn) [10:54:06] (03PS1) 10Addshore: Add @covers tags for /content files [core] - 10https://gerrit.wikimedia.org/r/90521 [11:06:56] (03PS1) 10Addshore: Add @covers for GenderCache + MessageCache tests [core] - 10https://gerrit.wikimedia.org/r/90524 [11:35:40] what needs to happen to mirror a gerrit repo at github? [11:39:40] (03PS1) 10Zfilipin: Deleted unused files [selenium] - 10https://gerrit.wikimedia.org/r/90525 [11:41:50] (03PS1) 10Dan-nl: item specific categories throws an exception [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90527 [12:00:50] (03PS1) 10Zfilipin: Upated todo with usage instructions [selenium] - 10https://gerrit.wikimedia.org/r/90531 [12:04:53] (03PS1) 10Zfilipin: Cleanup of gemspec file [selenium] - 10https://gerrit.wikimedia.org/r/90532 [12:24:28] (03CR) 10Helder.wiki: "Shoudn't also new mw.Api().(un)watch use the new getToken method?" [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [12:27:50] (03PS1) 10Zfilipin: Moving documenation to mediawiki/selenium repository [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90534 [12:29:23] (03PS1) 10Zfilipin: Moved documentation from qa/browsertests repository [selenium] - 10https://gerrit.wikimedia.org/r/90535 [12:30:35] (03PS2) 10Zfilipin: Moving documenation to mediawiki/selenium repository [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90534 [12:42:29] (03PS1) 10Netbrain: Added #subpage functionality Given the subpages A/B/C [extensions/SideBarMenu] - 10https://gerrit.wikimedia.org/r/90536 [12:50:11] (03PS2) 10Manybubbles: Optionally enable sloppy prefix matching [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/90150 [13:21:59] (03CR) 10Krinkle: [C: 031] mediawiki/conf.d/_join.php minor fix + comment update [integration/jenkins] - 10https://gerrit.wikimedia.org/r/90114 (owner: 10Hashar) [13:36:23] (03PS1) 10Krinkle: resourceloader: Add definition hashing to improve cache invalidation [core] - 10https://gerrit.wikimedia.org/r/90541 [13:38:06] (03PS2) 10Krinkle: resourceloader: Add definition hashing to improve cache invalidation [core] - 10https://gerrit.wikimedia.org/r/90541 [13:39:33] (03CR) 10MarkAHershberger: "Waldir, do you have plans to finish this?" [core] - 10https://gerrit.wikimedia.org/r/65613 (owner: 10Waldir) [13:42:17] (03PS3) 10Krinkle: resourceloader: Add definition hashing to improve cache invalidation [core] - 10https://gerrit.wikimedia.org/r/90541 [13:42:31] (03PS1) 10Anomie: Vary on X-Forwarded-Proto for changing-protocol redirects [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/90543 [13:44:13] (03CR) 10Bartosz Dziewoński: [C: 031] "Looks good to me, but I'll defer to parser experts with merging (added some people as reviewers)." [core] - 10https://gerrit.wikimedia.org/r/80578 (owner: 10MaxSem) [13:46:32] (03CR) 10Krinkle: "Added test to assert 'class' is being included. Especially important for subclasses that change logic but didn't need to override the getD" [core] - 10https://gerrit.wikimedia.org/r/90541 (owner: 10Krinkle) [13:48:06] (03PS9) 10MaxSem: Make TOC hideable [core] - 10https://gerrit.wikimedia.org/r/80578 [13:53:48] (03CR) 10Krinkle: [C: 04-1] "Not all modules need it, but we should at least add it to file module and wiki module." [core] - 10https://gerrit.wikimedia.org/r/90541 (owner: 10Krinkle) [13:59:21] hashar: yo [14:00:18] (03CR) 10Bartosz Dziewoński: "Shouldn't includes/RecentChange.php be moved to that directory as well? RCCacheEntry which you moved inherits from it." [core] - 10https://gerrit.wikimedia.org/r/90299 (owner: 10Aude) [14:02:18] (03PS1) 10Aklapper: Work around access restriction for image attachments [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 [14:03:23] (03CR) 10Mark Bergsma: [C: 031] Vary on X-Forwarded-Proto for changing-protocol redirects [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/90543 (owner: 10Anomie) [14:04:14] (03PS1) 10Cmcmahon: only run on test2 until Bug 54886 resolved [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90547 [14:04:43] (03CR) 10Aklapper: [C: 031] "Tested on boogs.wmflabs.org, see http://boogs.wmflabs.org/show_bug.cgi?id=13 - would prefer an okay from csteipp before getting this in." [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [14:05:29] (03CR) 10Cmcmahon: "this should turn the beta labs builds green" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90547 (owner: 10Cmcmahon) [14:05:45] !g Ie94dba1047c9842ce1a4524e9e494b543409efa3 [14:05:46] https://gerrit.wikimedia.org/r/#q,Ie94dba1047c9842ce1a4524e9e494b543409efa3,n,z [14:05:53] AzaToth: hey :) [14:06:05] AzaToth: just noticed your ping, just when I wanted to start tweaking the debian jobs [14:06:10] only got an hour or so though [14:06:11] (03PS2) 10Aklapper: Work around access restriction for image attachments [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 [14:07:04] (03CR) 10Aklapper: [C: 031] "Garr. Fix indentation. Apart from that, same as comment for patchset 1." [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [14:07:52] AzaToth: I got the hack in a test job https://integration.wikimedia.org/ci/job/test-hashar-copyartifact-per-parameter/ [14:08:33] AzaToth: it copies artifact from another job, get the last successful build which has the same ZUUL_REF (which is the change) [14:09:35] k [14:10:25] updating the job [14:10:39] but ZUUL_REF is refs/tags/1.21.2 [14:10:57] I thought you where going to use the commit (sha1) directly [14:11:06] ahh true [14:11:34] can't really guarantee that no one will remake a tag [14:11:48] parameters examples: [14:11:51] https://integration.wikimedia.org/ci/job/operations-debs-pybal-source/33/parameters/? [14:11:54] https://integration.wikimedia.org/ci/job/operations-debs-pybal-binaries/14/parameters/? [14:12:11] oh [14:12:26] so [14:12:50] Zuul listens for events in Gerrit. When a patchset is send, it pick it and merge it on the latest version of the branc [14:12:51] h [14:13:00] that might craft a merge commit [14:13:11] Zuul forge a reference (prefixed with a Z) [14:13:19] and send that to Jenkins + the commit sha1 [14:13:53] https://integration.wikimedia.org/ci/job/test-hashar-copyartifact-per-parameter/4/parameters/? [14:13:57] that value is guaranted to be the same for all jobs triggered [14:13:58] had a strange ZUUL_REF [14:14:02] since it is crafted before triggering anything [14:14:11] AzaToth: yeah that one is merely a test [14:14:18] I assume all ZUUL_REF will be refs/zuul/master/Zef353427df814c339445729178a55d0c [14:14:22] I reused builds from https://integration.wikimedia.org/ci/job/mediawiki-core-release/ [14:14:36] the build 20 has ZUUL_REF=refs/tags/1.20.7 [14:14:43] hmm [14:14:59] but the test-hashar-copyartifact-per-parameter fetched the artifacts from build #19 of mediawiki-core-release [14:15:11] because that is the last successful build to match ZUUL_REF=refs/tags/1.20.7 [14:15:18] but why is ZUUL_REF a tag there? [14:15:40] the mediawiki-core-release job react on ref-updated events, which would be a tag :-] [14:15:44] will ZUUL_REF be a tag if there is a non-merge and tagged? [14:15:52] the purpose of that job is to build the tar ball whenever a new tag is made on mediawiki/core [14:16:16] so ZUUL_REF can't be guaranteed to be unique? [14:16:27] in this case no [14:16:38] but insideZuul that is different events so it is not going to be an issue [14:16:55] I mean, if someone tag a new version in some operations/debs/ repository, Zuul is not configured to do anything yet [14:16:56] k [14:17:24] and I don't really care about producing debian packages on tags [14:17:29] nor on merge [14:17:40] because ops will rebuild the package from scratch anyway [14:17:44] BUT [14:17:56] true [14:17:56] it is definitely interesting to build package on patch submission so people can easily try them out [14:18:09] so will probably add the jobs to trigger when a change is merged as well [14:18:24] sorry that might be a bit complicated :( [14:18:30] no problem [14:19:17] so it should work now? [14:20:23] hashar: haven't seen any updated at https://gerrit.wikimedia.org/r/#/c/71245/ so I don't know [14:23:15] hashar: aint jenkins 1.509 a bit old? [14:23:40] I am writing the change [14:23:47] it's like from april [14:23:52] 1.509.x are the Long Term Support version [14:24:06] I will get the latest one deployed on monday [14:24:41] (03PS13) 10Hashar: debian-glue macros and jobs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/71245 (owner: 10AzaToth) [14:24:45] rebase [14:24:52] (03CR) 10Hashar: "rebased" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/71245 (owner: 10AzaToth) [14:26:00] hashar: ah [14:26:14] (though lts is 1.509.4 and you are using 1.509.3 :-P [14:28:10] - [14:28:11] - [14:28:12] - true [14:28:13] + ZUUL_REF=$ZUUL_REF [14:28:14] + [14:28:15] + false [14:28:24] (03PS14) 10Hashar: debian-glue macros and jobs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/71245 (owner: 10AzaToth) [14:28:42] hashar: was the macro.yaml meant to be added? [14:28:47] updating jobs [14:29:03] ?? [14:29:22] patchset 14 does not have that [14:29:35] diff https://gerrit.wikimedia.org/r/#/c/71245/13..14/operations-debs.yaml,unified [14:29:37] ah, it's the rebase [14:29:58] ah yeah PS13 is a rebase [14:30:03] that is why I usually rebase first [14:30:04] then commit [14:30:16] then send the commit [14:30:22] lets spam [14:31:11] which-build: last-successful? [14:32:21] shouldn't it be "which-build: build-param"? [14:32:54] looking at https://integration.wikimedia.org/ci/computer/integration-debian-builder/ [14:33:52] I send 3 patchsets [14:33:56] https://integration.wikimedia.org/ci/job/operations-debs-pybal-source/36/ is the source of the third [14:33:59] patchset [14:34:03] (03PS1) 10Anomie: Use restriction-level-* messages in Linker::formatTemplates [core] - 10https://gerrit.wikimedia.org/r/90551 [14:34:19] when the binaries run of the first change is finished, the binary run of the second change should fetch the source of the second change [14:35:21] hmm [14:35:35] it did run puiparts before binaries was done [14:36:32] how many "binaries" are you running? [14:37:28] bahh [14:37:31] the jobs did not update [14:37:39] there should be 3 [14:37:41] one for each change [14:39:04] ah, but there are three jobs [14:40:48] 14:40:25 Copied 2 artifacts from "ha:AAAAoR+LCAAAAAAAAABb85aBtbiIQSajNKU4P08vOT+vOD8nVc+jsiC1KCczL9svvyTVzHb1RttJBUeZGJg8GdhyUvPSSzJ8GJhLi3JKGIR8shLLEvVzEvPS9YNLijLz0q0rihik0IxzhtAgwxgggJGJgaGiAMiQKWFQ0s/KT9LPB9qYWJIJVKmbkppUrFtQmZSYo1ucX1qUnKoPALhNQ4enAAAAoperations-debs-pybal-source" build number [14:40:49] ha:AAAApB+LCAAAAAAAAABb85aBtbiIQSajNKU4P08vOT+vOD8nVc+jsiC1KCczL9svvyTVzHb1RttJBUeZGJg8GdhyUvPSSzJ8GJhLi3JKGIR8shLLEvVzEvPS9YNLijLz0q0rihik0IxzhtAgwxgggJGJgaGiAMhgKmFQ1c/KT9LPB9qYWJIJVKmbkppUrFtQmZSYo1ucX1qUnKpvbKYPAIeQ3eaqAAAA36 [14:40:50] yeahhh [14:40:52] thank you jenkins [14:40:58] that is not helpful :D [14:41:16] hehe [14:41:29] (03CR) 10Brion VIBBER: [C: 031] "Provisional +1 from me certainly. :)" [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [14:41:52] I notice the piuparts never ran [14:42:36] (03CR) 10Zfilipin: [C: 032] only run on test2 until Bug 54886 resolved [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90547 (owner: 10Cmcmahon) [14:42:50] 14:37:48 *** SCRIPTSDIR variable is NOT set - using default [/etc/piuparts/scripts/] for piuparts run *** [14:42:50] 14:37:49 Guessed: ubuntu [14:42:50] 14:37:49 0m0.0s ERROR: Scripts directory is not a directory: /etc/piuparts/scripts/ [14:43:23] hashar: I assume piuparts aint installed on the computer [14:43:40] i.e. apt-get install piuparts [14:45:03] argh [14:45:28] so another issue [14:45:42] the binaries job keep all the stuff in its workspace [14:45:45] though I assume you would want to disable piuparts for now, as it can take 20min [14:45:49] and operations-debs-pybal-binaries attempts to find the latest version [14:45:53] which is not what we want :-] [14:46:07] offcoursew [14:46:20] which-build: last-successful [14:46:28] na that is different [14:46:34] that is the artifact jenkins should fetch [14:46:43] but the workspace is never cleared [14:46:48] ah [14:46:54] so we might end up with a more recent package made by a previous build [14:48:02] wrappers: [14:48:02] - workspace-cleanup [14:48:49] "Requires the Jenkins `Workspace Cleanup Plugin" [14:49:10] "The post-build workspace-cleanup is available as a publisher" [14:50:01] moaaaaa [14:50:05] I HATE JENKINS [14:50:17] https://github.com/openstack-infra/jenkins-job-builder/blob/master/jenkins_jobs/modules/publishers.py#L1777 [14:50:22] there is an issue in the job config [14:50:30] where? [14:50:48] no clue [14:50:58] what is the issue? [14:51:01] but in the config history I can see it is deleting the modification :( [14:51:16] (03CR) 10Bartosz Dziewoński: "Are the 'template-(semi)protected' and 'restriction-level-*' messages equivalent in all translations? If not, I'd avoid constructing the t" [core] - 10https://gerrit.wikimedia.org/r/90551 (owner: 10Anomie) [14:51:27] linky? [14:55:22] https://integration.wikimedia.org/ci/job/operations-debs-pybal-binaries/jobConfigHistory/? [14:55:25] not sure you can access it [14:56:12] AzaToth: http://imgur.com/EOytA5M [14:58:02] The page you are trying to view cannot be shown because it uses an invalid or unsupported form of compression. [14:58:04] :-P [14:58:10] uhuh [14:59:41] hashar: well, here is the generated binaries xml: http://paste.debian.net/59218/ [15:00:20] yeah [15:00:24] need to find a way to validate it [15:00:45] blockBuildWhenDownstreamBuilding false? [15:03:28] (03PS1) 10Anomie: Add $wgSemiprotectedRestrictionLevels [core] - 10https://gerrit.wikimedia.org/r/90552 [15:03:35] hmm [15:03:49] I have no clue which configuration it is falling back to [15:05:52] it's job: block-(downstream|upstram): true|false [15:06:48] (03CR) 10Anomie: "It's the difference between "(protected)" and "(fully protected)" and "(semi-protected)" and "(semi protected)" for English. Not identical" [core] - 10https://gerrit.wikimedia.org/r/90551 (owner: 10Anomie) [15:08:29] AzaToth: config history for the operations-debs-pybla-binaries http://paste.debian.net/59219/ :D [15:14:42] the only difference I can spot is the build timeout plugin [15:15:00] err [15:15:03] only suspicious [15:15:16] hmmzors [15:15:19] (03CR) 10CSteipp: "This make the caching headers more accurate, but I think we actually can't cache this redirect at all since we also essentially vary on th" [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/90543 (owner: 10Anomie) [15:15:26] and the block downstream/upstream [15:15:29] it is by default [15:15:47] I mean, by default JJB set them to false [15:15:57] what does a diff -u give? [15:16:20] http://paste.debian.net/59221/ [15:16:51] ... [15:16:52] ffs [15:17:04] when applying manually the ZUUL_REF=$ZUUL_REF filter [15:17:05] that works [15:17:35] and of course jenkins kindly reorder the XML elements [15:18:02] I can try deleting them and rebuilding them [15:18:13] jenkins should be prevented to modify the config by itself [15:19:51] but it seems it want to revert the change [15:20:35] yup [15:20:37] something invalid [15:20:42] and no log / error / anything [15:23:21] AzaToth: I am deleting the jobs [15:23:25] k [15:27:50] recreating them [15:28:00] then I will trigger a run [15:28:03] (03CR) 10BryanDavis: [C: 031] "Seems reasonable. It's not used in core for sure. Could there be extensions that still use the old const? And if so does this change need " [core] - 10https://gerrit.wikimedia.org/r/90463 (owner: 10Aaron Schulz) [15:28:03] and rush out to home :/ [15:29:32] it did it again :-((((((((( [15:29:33] no clue [15:30:03] (03CR) 10Anomie: "Good point about the IP, although you did the same thing in Gerrit change 86697 (see the change to includes/specials/SpecialUserlogin.php)" [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/90543 (owner: 10Anomie) [15:30:53] AzaToth: jenkins does strip the ZUUL_REF=$ZUUL_REF parameter filters :( [15:31:42] lol [15:32:02] well, I'm off now (mom just arrived so...) [15:33:39] maybe the copy artifact does not support it :( [15:33:49] would look at it later tonight or over the weekend [15:33:51] sorry :( [15:38:17] (03PS1) 10Diederik: Remove

wrapping around label that shows in tooltip [extensions/SecurePoll] - 10https://gerrit.wikimedia.org/r/90555 [15:39:33] (03PS1) 10Cicalese: Added version warning. [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/90556 [15:39:57] off [15:39:59] good weekend [15:40:11] (03CR) 10Cicalese: [C: 032 V: 032] Added version warning. [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/90556 (owner: 10Cicalese) [15:46:58] andre__: ping? [15:47:16] hexmode, pong [15:48:48] andre__: Today is T-6weeks for releasing 1.22. Can you create 1.22 versions for MW and extensions in bugzilla? cf https://www.mediawiki.org/wiki/Project:Release_management/Release_timeline [15:49:01] (I can if you want me to) [15:49:18] R-6weeks [15:49:31] hexmode: 1.22.0 for MW? Shouldn't that be something like 1.22rcX or so? [15:50:35] andre__: Whatever you prefer. "1.22rcX" (no # for X since there will be a few) is ok with me [15:51:10] hexmode, I'd start with that. Slightly afraid that people might be confused otherwise (with 1.22-git). [15:51:19] hexmode, could you file a ticket in Bugzilla against Bugzilla for it? :) [15:51:35] andre__: Also, if you could update the release timeline wiki doc for the version numbers [15:51:38] will do [15:51:47] (03PS1) 10Zfilipin: WIP Moved config files to mediawiki/selenium repository [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90559 [15:51:48] ah, sure [15:52:51] (03PS1) 10Zfilipin: WIP Moved confing files from qa/browsertests repository [selenium] - 10https://gerrit.wikimedia.org/r/90560 [15:53:44] (03PS2) 10Zfilipin: WIP Moved config files to mediawiki/selenium repository [qa/browsertests] - 10https://gerrit.wikimedia.org/r/90559 [15:54:09] (03PS2) 10Zfilipin: WIP Moved confing files from qa/browsertests repository [selenium] - 10https://gerrit.wikimedia.org/r/90560 [15:56:31] andre__: bug created :) https://bugzilla.wikimedia.org/show_bug.cgi?id=55879 [15:57:18] (03PS4) 10Krinkle: resourceloader: Add definition hashing to improve cache invalidation [core] - 10https://gerrit.wikimedia.org/r/90541 [15:57:33] hexmode, thanks! [16:04:05] (03PS1) 10Zfilipin: Relevant links [selenium] - 10https://gerrit.wikimedia.org/r/90562 [16:04:57] (03CR) 10CSteipp: [C: 04-1] "I'm not entirely familiar with bugzilla, but I don't think this is safe for IE users." [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [16:09:54] (03PS1) 10Bartosz Dziewoński: Make #contentSub slightly darker [core] - 10https://gerrit.wikimedia.org/r/90563 [16:15:19] Class let out. [16:16:32] (03CR) 10Waldir: "Mark, yes. I haven0t been able to work much in MW development lately, but in this case I just need some help with the documentation commen" [core] - 10https://gerrit.wikimedia.org/r/65613 (owner: 10Waldir) [16:23:51] (03CR) 10Chad: "How about compared to json, like I suggested above?" [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [16:27:23] (03CR) 10DarTar: [C: 031 V: 031] "Note to self: a schema change will break all echo dashboards that currently pull data from scid:5423520, for example:" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/90497 (owner: 10Legoktm) [16:30:15] (03CR) 10MarkAHershberger: "I was sure I had edited that comment and said something slightly different." [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [16:35:35] (03CR) 10Kipcool: [C: 032 V: 032] added check mode and corrections API add_syntrans [extensions/WikiLexicalData] - 10https://gerrit.wikimedia.org/r/90080 (owner: 10Hiong3-eng5) [16:37:56] (03CR) 10Chad: [C: 032] Optionally enable sloppy prefix matching [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/90150 (owner: 10Manybubbles) [16:38:00] (03Merged) 10jenkins-bot: Optionally enable sloppy prefix matching [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/90150 (owner: 10Manybubbles) [16:38:48] (03PS1) 10Aklapper: Add support for Trello URLs to Bugzilla's 'See Also' field. [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90565 [17:04:24] (03PS1) 10Bartosz Dziewoński: Rename mw.util.wikiGetlink to getUrl [core] - 10https://gerrit.wikimedia.org/r/90569 [17:08:41] (03PS2) 10Bartosz Dziewoński: Rename mw.util.wikiGetlink to getUrl [core] - 10https://gerrit.wikimedia.org/r/90569 [17:08:49] not bad, hhvm 2.2.0 released yesterday. no time to fully test yet, but Main_Page loads (in half the time), even loads up Flow reasonably well(but needs some sort of configuration for assets and pretty-urls and usch) [17:11:28] ebernhardson: nice1 [17:11:29] NICE! [17:11:31] gah [17:11:32] nice! [17:11:46] :) [17:13:10] anomie: FYI, tgr and I have been working on the core patch for extended metadata and we'd like you to take another look at it very soon (he's in the middle of uploading a new patchset) [17:13:35] marktraceur: ok [17:13:46] anomie: It's sort of eating into the launch time for the MultimediaViewer extension, and we're hoping to get that on mediawiki.org on Halloween or so [17:14:24] Could use help on CommonsMetadata too, but one step at a time :) [17:14:29] ori-l: You too. :) [17:14:46] (03PS1) 10Reedy: When using importImages.php, if the summary parameter is not set, then use the page text of the given page, NOT that of the first image uploaded. [core] - 10https://gerrit.wikimedia.org/r/90571 [17:17:19] (03PS2) 10Bartosz Dziewoński: When using importImages.php, if the summary parameter is not set, then use the page text of the given page, NOT that of the first image uploaded. [core] - 10https://gerrit.wikimedia.org/r/90571 (owner: 10Reedy) [17:17:30] (03CR) 10Bartosz Dziewoński: [C: 031] importImages.php: Correctly generate automatic summary [core] - 10https://gerrit.wikimedia.org/r/90571 (owner: 10Reedy) [17:22:13] * Reedy waits for jenkins [17:24:25] (03CR) 10BryanDavis: "(2 comments)" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90149 (owner: 10Dan-nl) [17:29:54] ^d: What needs to be done to get tgr +2? He's Gergő Tisza, the newest member of the Multimedia team. [17:30:42] <^d> He's FT, right? [17:31:08] ^d: Yes. RobLa announced him in the mailing list this week [17:31:16] <^d> Done [17:32:06] tgr: No more hiding behind the "I don't have +2" shield :) [17:32:40] yay [17:32:51] thx [17:33:07] (03CR) 10Aaron Schulz: "I don't see any" [core] - 10https://gerrit.wikimedia.org/r/90463 (owner: 10Aaron Schulz) [17:33:57] (03PS1) 10Reedy: Don't blindly overwrite $options in importImages.php [core] - 10https://gerrit.wikimedia.org/r/90573 [17:34:38] MatmaRex: The bugfix on 90571 won't actually work without what I just changed in 90573 to fix another bug [17:34:39] :D [17:34:54] Except I fail at spelling [17:34:57] Reedy: rebase all the things [17:35:26] just merge them all [17:35:41] * Reedy waits a second [17:35:41] also, lol php scoping and variable shadowing [17:35:42] (03PS2) 10Reedy: Don't blindly overwrite $options in importImages.php [core] - 10https://gerrit.wikimedia.org/r/90573 [17:35:49] Exactly [17:36:00] Reedy: make it a class, will you? [17:36:11] these .inc files really get on my nerves [17:36:14] lol [17:36:40] We can kill the .inc by just moving its functions into importImages.php [17:36:46] But yeah, really should be refactored [17:37:26] (03CR) 10Jdlrobson: [C: 031] "Looks good and works fine but I'm not seeing the deprecation notice for some reason. I'm not familiar with how deprecation works so I'm no" [core] - 10https://gerrit.wikimedia.org/r/90569 (owner: 10Bartosz Dziewoński) [17:38:45] (03CR) 10Bartosz Dziewoński: [C: 032] Don't blindly overwrite $options in importImages.php [core] - 10https://gerrit.wikimedia.org/r/90573 (owner: 10Reedy) [17:38:53] Reedy: i trust you that works ^ [17:39:42] So do I [17:43:47] (03CR) 10BryanDavis: [C: 032] Removed unused const QoS_Atomic constant (odd casing) [core] - 10https://gerrit.wikimedia.org/r/90463 (owner: 10Aaron Schulz) [17:43:58] (03Merged) 10jenkins-bot: Don't blindly overwrite $options in importImages.php [core] - 10https://gerrit.wikimedia.org/r/90573 (owner: 10Reedy) [17:44:30] (03CR) 10Bartosz Dziewoński: "It's only shown in debug mode ?debug=true (mediawiki.log.js is registered amond the 'debugScripts' in Resources.php), when you call mw.uti" [core] - 10https://gerrit.wikimedia.org/r/90569 (owner: 10Bartosz Dziewoński) [17:44:49] (03PS3) 10Bartosz Dziewoński: importImages.php: Correctly generate automatic summary [core] - 10https://gerrit.wikimedia.org/r/90571 (owner: 10Reedy) [17:45:35] (03CR) 10Bartosz Dziewoński: [C: 032] importImages.php: Correctly generate automatic summary [core] - 10https://gerrit.wikimedia.org/r/90571 (owner: 10Reedy) [17:47:55] (03PS1) 10Chad: Remove duplicates from search engine results in tests [core] - 10https://gerrit.wikimedia.org/r/90575 [17:48:11] (03CR) 10Umherirrender: "Nikerabbit, something unclear?" [core] - 10https://gerrit.wikimedia.org/r/84138 (owner: 10Umherirrender) [17:48:15] <^d> AaronSchulz: ^ [17:48:27] (03Merged) 10jenkins-bot: Removed unused const QoS_Atomic constant (odd casing) [core] - 10https://gerrit.wikimedia.org/r/90463 (owner: 10Aaron Schulz) [17:50:16] (03CR) 10Nikerabbit: "Nope. Just didn't see your reply." [core] - 10https://gerrit.wikimedia.org/r/84138 (owner: 10Umherirrender) [17:51:04] ^d: ok the tests pass for me without that [17:51:13] (03CR) 10AndyRussG: "I can definitely clean up the formatting and describe this in bug 55771." [extensions/GuidedTour] - 10https://gerrit.wikimedia.org/r/82209 (owner: 10AndyRussG) [17:51:23] <^d> AaronSchulz: Indeed, and they've been passing for me on jenkins. [17:51:33] <^d> And...in some previous life, they did. [17:51:45] <^d> But these have been pretty consistently breaking for me recently. [17:51:58] on hiphop? ;) [17:52:43] Reedy: merge something for me now? https://gerrit.wikimedia.org/r/#/c/84315/ :> [17:53:21] (03PS5) 10Reedy: Correctly update wl_notificationtimestamp when viewing old revisions [core] - 10https://gerrit.wikimedia.org/r/84315 (owner: 10Bartosz Dziewoński) [17:53:27] omfg it rebased [17:53:34] (03Merged) 10jenkins-bot: importImages.php: Correctly generate automatic summary [core] - 10https://gerrit.wikimedia.org/r/90571 (owner: 10Reedy) [17:54:41] (03CR) 10Jeroen De Dauw: "Well this is certainly an original approach!" [extensions/SideBarMenu] - 10https://gerrit.wikimedia.org/r/90536 (owner: 10Netbrain) [17:54:50] (03PS6) 10Bartosz Dziewoński: Correctly update wl_notificationtimestamp when viewing old revisions [core] - 10https://gerrit.wikimedia.org/r/84315 [17:54:53] (03CR) 10Umherirrender: "Due to wikidata some other api modules where added, which does edits. So it should be possible to add assertion also to other modules. One" [core] - 10https://gerrit.wikimedia.org/r/90263 (owner: 10Legoktm) [17:54:53] you did not rebase it hard enough! [17:56:43] <^d> AaronSchulz: hhvm and zend. [17:57:45] (03CR) 10Bartosz Dziewoński: "(Shameless plug: you can avoid release notes rebase hell with https://github.com/MatmaRex/mediawikireleasenotes-driver (you still need to " [core] - 10https://gerrit.wikimedia.org/r/90263 (owner: 10Legoktm) [17:59:29] Reedy: could you look over https://gerrit.wikimedia.org/r/#/c/90490/ ? [18:02:03] ^d: someone needs to clean up the DatabaseBase constructor :) [18:02:05] btw [18:02:19] <^d> Not it! [18:02:21] <^d> :) [18:02:52] https://integration.wikimedia.org/cover/mediawiki-core/master/php/index.dashboard.html "Top Project Risks: Parser (840552)" [18:02:55] well don't you say. [18:06:06] (03CR) 10Jeroen De Dauw: "(1 comment)" [extensions/SideBarMenu] - 10https://gerrit.wikimedia.org/r/90536 (owner: 10Netbrain) [18:28:17] (03CR) 10Jeroen De Dauw: "(1 comment)" [extensions/SideBarMenu] - 10https://gerrit.wikimedia.org/r/90536 (owner: 10Netbrain) [18:31:56] (03CR) 10MarkTraceur: [C: 04-1] "(22 comments)" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90041 (owner: 10Dan-nl) [18:34:41] DarTar: ping [18:34:50] hey [18:35:21] DarTar: so why could changing the schema for echo break the current dashboards? i'm not very familiar with how eventlogging works [18:36:00] because they currently pull data from a single SQL table, defined by {schema, rev_id} [18:36:13] oh :/ [18:36:31] maintaining dashboards across rebids would require unions across multiple tables [18:36:47] keeping track of what revid's are relevant [18:36:54] s/rebids/revids [18:37:22] also, is the data collected by eventlogging all public? because thanks rev_ids are not supposed to be public, only the person recieving should know what edit it is for [18:37:31] https://bugzilla.wikimedia.org/show_bug.cgi?id=49087 [18:37:34] (03CR) 10Dzahn: [C: 032] delete Search.pm, it's not custom anymore [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90461 (owner: 10Dzahn) [18:37:36] no, it's private data [18:37:42] (03CR) 10Dzahn: [V: 032] delete Search.pm, it's not custom anymore [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90461 (owner: 10Dzahn) [18:37:50] ok [18:37:51] I'm in a meeting at the moment but can chat later [18:37:56] sure [18:42:01] (03CR) 10Mwalker: [C: 032] Add @covers for main Database test types [core] - 10https://gerrit.wikimedia.org/r/90518 (owner: 10Addshore) [18:47:51] (03CR) 10Bsitu: [C: 04-1] Only allow Thanks for public revisions [extensions/Thanks] - 10https://gerrit.wikimedia.org/r/88223 (owner: 10Legoktm) [18:48:01] (03Abandoned) 10Chad: Remove duplicates from search engine results in tests [core] - 10https://gerrit.wikimedia.org/r/90575 (owner: 10Chad) [18:48:10] ok who has the wgEnableZeroRatedMobileAccessTesting changes unmerged on tin? [18:48:17] (03CR) 10Mwalker: [C: 032] Add @covers for GenderCache + MessageCache tests [core] - 10https://gerrit.wikimedia.org/r/90524 (owner: 10Addshore) [18:48:22] I'm wanting to push through a db-eqiad change, see... [18:49:29] that's usually yurik & dr0ptp4kt(sp?) [18:49:43] i don't see either of them [18:50:16] (03CR) 10Mwalker: [C: 032] Add @covers tags for /content files [core] - 10https://gerrit.wikimedia.org/r/90521 (owner: 10Addshore) [18:50:39] coren +2'ed it, I pinged them both [18:57:38] (03CR) 10Mwalker: [C: 032] "The SQL files are direct copies; and I tested that this didn't break the updater. Did not attempt to install MW 1.14 to see if the upgrade" [core] - 10https://gerrit.wikimedia.org/r/80367 (owner: 10Addshore) [19:00:05] (03PS4) 10MarkTraceur: Add source and authour to the title div [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89128 [19:01:48] (03PS2) 10MarkTraceur: Add creation time to list [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90040 [19:03:30] Reedy: There: https://www.mediawiki.org/wiki/Architecture_guidelines/Meetings/Architecture_Summit_2014#Requests - now all structured in a table. Happy? [19:05:05] (03PS3) 10Gergő Tisza: Return categories and license [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/90213 (owner: 10MarkTraceur) [19:05:06] (03PS6) 10Gergő Tisza: Extension to screen scrape metadata out of commons [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/80403 (owner: 10Brian Wolff) [19:05:14] (03PS4) 10MarkTraceur: Add description to the DOM [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89242 [19:06:31] (03CR) 10Gergő Tisza: [C: 031] "Split up into scary HTML scraper part (seems to work, will be replaced with Wikidata soon-ish) and rest." [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/80403 (owner: 10Brian Wolff) [19:07:15] (03CR) 10Gergő Tisza: [C: 031] "Rebased." [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/90213 (owner: 10MarkTraceur) [19:11:09] (03PS16) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [19:11:10] (03PS6) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [19:12:33] Nice James_F [19:12:35] (03PS1) 10Cicalese: Don't overwrite title when not using display title (to avoid conflict with SemanticTitle). [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/90583 [19:12:54] Reedy: Sadly you can't add rows with VisualEditor yet, though. :-( [19:12:54] (03CR) 10Gergő Tisza: "Some refactoring, use similar array format as metadata prop, removed dead code, re-added caching" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [19:13:13] (03CR) 10Cicalese: [C: 032 V: 032] Don't overwrite title when not using display title (to avoid conflict with SemanticTitle). [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/90583 (owner: 10Cicalese) [19:13:40] (03CR) 10Gergő Tisza: [C: 031] "Rebased." [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [19:13:43] (03PS1) 10JanZerebecki: Use PHPUnit filter/whitelist to select the files that need testing. [core] - 10https://gerrit.wikimedia.org/r/90584 [19:16:44] DarTar: Did an RT ticket ever get filed for me and stat1? [19:16:51] Not that I'm impatient, just curious [19:19:20] mark I think I did, let me check [19:19:58] arg no, I posted requests for someone else and for some reason I thought you were in that list :( [19:20:04] sorry man [19:20:09] I'll do it now [19:35:14] marktraceur: preferred shell user name? [19:36:27] (03CR) 10Jeroen De Dauw: [C: 031] Use PHPUnit filter/whitelist to select the files that need testing. [core] - 10https://gerrit.wikimedia.org/r/90584 (owner: 10JanZerebecki) [19:38:32] DarTar: marktraceur please :) [19:38:59] (03CR) 10Aklapper: [C: 04-1] "csteipp: Thanks!! Confirming the problem after testing with IE6 in emulator and a fake PNG. Going to test more..." [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [19:39:08] k [19:40:39] manybubbles & ^d, I think I've made the revision of our mutual interest, https://gerrit.wikimedia.org/r/80578 work right, can you guys review it? [19:42:46] (03CR) 10Jdlrobson: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [19:46:34] how can i make an extension add to the header that EditPage::showHeader() writes? [19:47:11] do i have to make a new hook? [19:48:02] Dunno [19:48:06] Is there a hook in that function? [19:50:08] i don't see one. i do see TitleGetEditNotices in Title::getEditNotices(), which that function calls. would it be "clean" to use it? [19:57:44] (03CR) 10Brion VIBBER: "We can simply disable IE's type sniffing: http://msdn.microsoft.com/en-us/library/ie/gg622941%28v=vs.85%29.aspx" [wikimedia/bugzilla/modifications] - 10https://gerrit.wikimedia.org/r/90546 (owner: 10Aklapper) [20:16:24] (03PS2) 10BryanDavis: Reduce excess lock contention and chance for failure on rename [core] - 10https://gerrit.wikimedia.org/r/90385 (owner: 10Aaron Schulz) [20:20:19] (03CR) 10BryanDavis: [C: 032] "Not only is this better for performance, I think the code reads better too." [core] - 10https://gerrit.wikimedia.org/r/90385 (owner: 10Aaron Schulz) [20:23:23] (03Merged) 10jenkins-bot: Reduce excess lock contention and chance for failure on rename [core] - 10https://gerrit.wikimedia.org/r/90385 (owner: 10Aaron Schulz) [20:26:12] mwalker: so yeah Zuul broken somehow :( [20:26:24] I am missing the HEAD remote referene :( [20:27:07] that's interesitng though that git can't find it [20:28:27] (03PS1) 10Bartosz Dziewoński: Rename Toolbox to Tools [core] - 10https://gerrit.wikimedia.org/r/90645 [20:28:40] so; I just did a test clone of wmf://wikimedia/fundraising/crm [20:28:41] $ git remote set-head origin -a [20:28:42] origin/HEAD set to master [20:28:43] \O/ [20:28:48] something wrong happened eventually [20:30:00] miauhahaha [20:30:02] whooo [20:30:02] mwalker: https://integration.wikimedia.org/ci/job/wikimedia-fundraising-crm-phplint/1/console [20:30:05] mwalker: Does wmf:// work for you? [20:30:12] Because that's kinda awesome [20:30:14] it does; but I added it by hand [20:30:14] so that is really crazy [20:30:23] mwalker: How? [20:30:33] in ~/.gitconfig add [20:30:34] [url "ssh://mwalker@gerrit.wikimedia.org:29418/"] [20:30:34] insteadOf = "wmf:" [20:30:49] My hero [20:30:54] hashar: so basically the git repo was just borked? [20:31:02] well; the clone was borked somehow? [20:31:06] reading https://bugs.launchpad.net/zuul/+bug/1235005 [20:31:11] marktraceur did everything basically [20:31:22] I mean, he completed the investigation [20:31:32] damn that is old [20:31:32] Yarp [20:31:46] how the hell did I miss such an issue? Have you guys filled something in bugzilla already ? [20:32:23] dont think so -- and we commit pretty rarely to the crm repo so it was annoying; but not enough to spur immediate aggghhhhh [20:34:45] ahh [20:35:01] well you could test the module as well` [20:35:02] but IIRC, Zuul does not update submodules [20:35:04] so you should be fine [20:35:17] operations/puppet/* could have a similar 'issue' [20:35:26] since we now have operations/puppet/modules/* :-] [20:35:43] I'm suprised it doesn't update submodules [20:35:50] marktraceur: the upstream bug, I might eventually fix it one day [20:35:57] marktraceur: I got to upgrade our Zuul install first [20:36:06] mwalker: I don't think it needs them [20:36:09] not sure [20:36:26] I am leaving https://bugzilla.wikimedia.org/show_bug.cgi?id=55896 open [20:36:30] well; if you were testing a core change against all the submodules; it would [20:36:32] feel free to close if that wokrks for you [20:37:02] off If I google for zuul submodules [20:37:15] I mostly have bug report I have filled to myself in our bugzilla installation [20:38:12] hashar: Sounds good [20:38:19] well; it's no longer throwing any errors and we don't have a cross module test suite -- so do you want me to close it? [20:39:16] I think you can close it indeed [20:39:27] note that it is only used internally by Zuul to craft a commit [20:39:41] the jenkins jobs then clone that commit in a workspace and might do the submodule update if asked to [20:39:51] for linting job, the submodules are not processed [20:43:44] (03CR) 10Bartosz Dziewoński: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [20:46:01] (03PS4) 10BryanDavis: lockmanager: QuorumLockManager subclasses can get EX/SH locks at once [core] - 10https://gerrit.wikimedia.org/r/69021 (owner: 10Aaron Schulz) [20:46:26] (03PS1) 10Bartosz Dziewoński: BaseTemplate: Exclude two more attributes in makeLink() [core] - 10https://gerrit.wikimedia.org/r/90652 [20:47:20] marktraceur: kudos on finding the root cause for the crm repo :-] [20:47:46] hashar: Thanks, I try. :) [20:49:59] (03PS1) 10Aaron Schulz: Cleaned up DatabaseBase constructor to use an array [core] - 10https://gerrit.wikimedia.org/r/90653 [20:51:28] (03CR) 10jenkins-bot: [V: 04-1] Cleaned up DatabaseBase constructor to use an array [core] - 10https://gerrit.wikimedia.org/r/90653 (owner: 10Aaron Schulz) [20:53:34] hehe [20:55:56] (03CR) 10Swalling: [C: 031] "Yay! Simplification for the win." [core] - 10https://gerrit.wikimedia.org/r/90645 (owner: 10Bartosz Dziewoński) [20:58:13] (03PS1) 10Aaron Schulz: Fixed a bunch of annoying deprecation notices [core] - 10https://gerrit.wikimedia.org/r/90655 [20:59:36] ^d: https://gerrit.wikimedia.org/r/#/c/90655/ [20:59:47] hashar: tests don't like me using wfWarn() for deprecation notices :) [20:59:53] * AaronSchulz will have to find something else [21:00:03] wfDeprecated() ? [21:00:34] it's not really the method but a certain calling signature [21:01:00] (03PS1) 10MaxSem: Moar fixage [extensions/Minifier] - 10https://gerrit.wikimedia.org/r/90656 [21:01:16] (03CR) 10MaxSem: [C: 032 V: 032] Moar fixage [extensions/Minifier] - 10https://gerrit.wikimedia.org/r/90656 (owner: 10MaxSem) [21:02:17] AaronSchulz: I wish wfDeprecated() would accept an additional parameter with a longer description :( [21:02:28] maybe you can abuse $function and pass a longer text ? [21:04:00] (03CR) 10Chad: [C: 032] Fixed a bunch of annoying deprecation notices [core] - 10https://gerrit.wikimedia.org/r/90655 (owner: 10Aaron Schulz) [21:05:07] (03CR) 10Hashar: [C: 032] Fixed a bunch of annoying deprecation notices [core] - 10https://gerrit.wikimedia.org/r/90655 (owner: 10Aaron Schulz) [21:07:00] (03Merged) 10jenkins-bot: Fixed a bunch of annoying deprecation notices [core] - 10https://gerrit.wikimedia.org/r/90655 (owner: 10Aaron Schulz) [21:25:55] (03PS3) 10Dan-nl: corrected issue where batch upload breaks off early [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90149 [21:39:28] (03CR) 10Dan-nl: "(2 comments)" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90149 (owner: 10Dan-nl) [21:49:00] (03CR) 10BryanDavis: "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/69021 (owner: 10Aaron Schulz) [21:49:51] (03CR) 10Brion VIBBER: [C: 04-1] "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90073 (owner: 10MarkTraceur) [21:50:29] <^d> Every time I see GWToolset I read it as "GWT oolset" rather than "GW Toolset" [21:52:06] (03PS1) 10MarkTraceur: Fix a few fullscreen bugs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90660 [21:53:57] brion: DarTar wrote that schema [21:54:08] I have to figure they've been over that question before [21:54:14] commuting on that, brine's concerns are legitimate [21:54:23] s/commuting/commenting [21:54:55] (03CR) 10Brion VIBBER: "(2 comments)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 (owner: 10MarkTraceur) [21:55:00] Actually less so [21:55:09] (03CR) 10MarkTraceur: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90073 (owner: 10MarkTraceur) [21:55:10] spiff [21:55:24] * marktraceur <3 brion [21:55:27] (03CR) 10DarTar: [C: 04-1] "Instrumentation involving clicks on UI elements has been used pretty extensively in Product, but I agree that this is getting dangerously " [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90073 (owner: 10MarkTraceur) [21:55:52] brion: I have NO IDEA what's going on in that SVG. Blame violetto or Jared I think. [21:56:02] They used Illustrator to generate it [21:56:12] dunno what my autocorrect is doing today [21:56:19] commuting, brine [21:56:33] ok if it's just measuring number of clicks that's probably not so bad [21:56:36] i just hate recording user ids :P [21:56:45] No no no [21:56:49] it's not just counting [21:56:53] hmmm wait [21:57:00] DarTar: We don't store data about which article or picture is being viewed [21:57:01] it associates user metadata with each event [21:57:02] lemme double-check that schema [21:57:16] Only user ID and edit count and action [21:57:23] marktraceur: isn't there a rev_id ? [21:57:30] Er [21:57:46] https://meta.wikimedia.org/wiki/Schema:MediaViewer ok i don't see a rev id... [21:57:47] sorry, I must be confusing too many schemas I looked at today [21:57:52] unless there's something else in there [21:57:52] DarTar: https://meta.wikimedia.org/wiki/Schema:MediaViewer no [21:58:16] k let me see it again [21:58:54] (03CR) 10Brion VIBBER: "Ok double-checking https://meta.wikimedia.org/wiki/Schema:MediaViewer there is indeed no recording of which file is being loaded, so it's " [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90073 (owner: 10MarkTraceur) [21:59:10] brion: action is actually used as the rev_id, except we're only tracking stats for 6 articles. :P [21:59:14] 'user_id' just gave me a heart attack [21:59:15] heh [21:59:36] that's interesting, my understanding from Fabrice's request is that he wished to understand how clicks would perform on an article basis [21:59:43] Was it? [21:59:47] hmm [21:59:50] which would have been creepy, privacy wise [21:59:52] Maybe we should revisit that then [21:59:53] heh [22:00:02] but there's no trace of page_id or rev_id in the schema [22:00:27] so let's double check with him on the requirements before we move on with this patch [22:00:46] marktraceur: on the SVG, I think the line can just be removed, it was probably a temporary item that got traced :) [22:00:46] dropping him a line now [22:00:51] Can I just say that, having come from a place where my job was to figure out how track everything everyone did everywhere they might eventually use a credit card, the respect for privacy here is inspiring. [22:01:12] brion: Yeah, probably [22:01:21] brion: I'm just gonna compress a bunch of the whitespace too [22:01:24] heh [22:01:29] whitespace is for HUMANS [22:01:30] Fucking illustrator and its fucking bullshit svg export [22:01:43] * marktraceur has a few opinions on this [22:01:57] ELEMENTS FOR EVERYTHING [22:02:05] Style them? WHY WOULD WE!? [22:02:14] if i had time i'd retool inkscape to look like illustrator and try feeding it to the designers instead :) [22:02:31] * brion replaces their coffee with folger's crystals [22:02:32] NO DUH [22:02:36] lol [22:04:02] marktraceur: You wait until you try creating an element in inline HTML that covers the entire page, then measuring it in JS [22:04:34] ( https://github.com/trevorparscal/stalk/commit/80a5a93ca938ba1871fdfdb1e1f5cf3f2a21ee65 ) [22:05:02] RoanKattouw: Yeah, I've been there (orgcharts) [22:05:53] <^d> My solution is to stay as far from the browser as I can. [22:06:02] <^d> backend4ever [22:09:23] (03PS1) 10Aaron Schulz: Try to make sure jobs will not get pushed twice on failures [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90663 [22:10:25] (03CR) 10Aaron Schulz: [C: 032] Try to make sure jobs will not get pushed twice on failures [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90663 (owner: 10Aaron Schulz) [22:12:35] (03PS9) 10Bartosz Dziewoński: Rewrite rendering of Vector simple search [core] - 10https://gerrit.wikimedia.org/r/82100 [22:12:51] (03PS10) 10Bartosz Dziewoński: Rewrite rendering of Vector simple search [core] - 10https://gerrit.wikimedia.org/r/82100 [22:13:29] (03CR) 10Bartosz Dziewoński: [C: 04-1] "Mentioned two more bugs this fixes, rebased, TODO the JS thing." [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [22:14:31] (03CR) 10Brion VIBBER: "(2 comments)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90267 (owner: 10MarkTraceur) [22:14:56] (03Merged) 10jenkins-bot: Try to make sure jobs will not get pushed twice on failures [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90663 (owner: 10Aaron Schulz) [22:14:57] (03PS2) 10Aaron Schulz: Cleaned up DatabaseBase constructor to use an array [core] - 10https://gerrit.wikimedia.org/r/90653 [22:15:40] (03PS1) 10Aaron Schulz: jobqueue: various cleanups to JobQueueFederated [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90664 [22:15:52] (03CR) 10Aaron Schulz: [C: 032] jobqueue: various cleanups to JobQueueFederated [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90664 (owner: 10Aaron Schulz) [22:16:26] (03PS4) 10MarkTraceur: Add logo to feature preference, fix copy [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 [22:16:29] brion: Got you a nice compressed SVG now [22:16:35] Still a bunch of nasty paths, but better anyway [22:16:42] whee :D [22:17:02] now can you put some RDFa metadata into that SVG? let's make it semantic! [22:17:05] re: cc.svg I blame greg-g [22:17:13] * brion wonders how far he can troll marktraceur ;) [22:17:20] brion: Thiiiiiiis far [22:17:22] hehe [22:17:54] brion: I dream of going back to rewrite viewer.svg so it's polygons and not paths. It may be a while. :) [22:18:06] nom nom nom [22:18:33] that's what we need -- an in-browser tool for normalizing SVGs to make them human-readable source [22:18:38] csteipp: so, is https://gerrit.wikimedia.org/r/#/c/89107/ ok? [22:19:05] (03CR) 10jenkins-bot: [V: 04-1] Cleaned up DatabaseBase constructor to use an array [core] - 10https://gerrit.wikimedia.org/r/90653 (owner: 10Aaron Schulz) [22:19:18] AaronSchulz: You had an undefined variable when I tried to test it [22:19:25] See comment [22:19:56] huh, all I saw was "Also, we'll figure out a better name in a followup patch, so that part is ok." [22:19:58] * AaronSchulz looks [22:20:58] brion: Any XML viewer should do [22:21:22] marktraceur: that SVG logo doesn't actually get shown yet does it? i see the updated copy on Special:Version, but no logo [22:21:38] aaand netbeans code intelligence is broken [22:21:47] why oh why did I have to upgrade? [22:21:53] * AaronSchulz watches the cpu waste time [22:22:00] *that* is what i want on my tombstone [22:22:05] "why oh why did I have to upgrade?" [22:22:26] brion 2.0: Now with more dead [22:22:41] bd808: "lower resting energy usage" [22:22:46] <^d> I've been waiting on brion 2.0 for ages. [22:22:50] "higher entropy"? [22:22:51] <^d> brion 2.0 implements __clone() [22:22:53] :D [22:23:21] (03Merged) 10jenkins-bot: jobqueue: various cleanups to JobQueueFederated [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/90664 (owner: 10Aaron Schulz) [22:23:28] (03CR) 10Brion VIBBER: "Copy looks good, but the logo doesn't actually show anywhere that I can see. Is it just stashed for later or is it meant to show?" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 (owner: 10MarkTraceur) [22:23:38] 7.4 has all the old 7.1 bugs again it feels like [22:23:52] freezing on prefs too [22:24:50] brion: It should show up on the betafeatures preference page [22:24:51] nothing kill -9 can't solve [22:24:57] ah that's where it is [22:25:00] lemme pull that up [22:25:30] yeah confirmed it shows \o/ [22:25:46] nice and sharp on retina display too :D [22:25:48] (03CR) 10MarkTraceur: "It's on the BetaFeatures page in preferences." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 (owner: 10MarkTraceur) [22:25:55] brion: SVG huzzah. :) [22:26:16] (03CR) 10Brion VIBBER: [C: 032 V: 032] "Confirmed image displays and looks beautiful on high-res screen thanks to SVG! \o/" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 (owner: 10MarkTraceur) [22:26:27] i forget if i'm supposed to push 'submit' on this repo or not [22:26:35] ori-l: is there a way for an extension to bundle a schema? I'm thinking I could do pageedit request in the update hook; but that seems awfully hacky [22:26:38] (03PS2) 10Aaron Schulz: Avoid duplicate key error on /authorize page [extensions/OAuth] - 10https://gerrit.wikimedia.org/r/89107 [22:26:43] (03CR) 10MarkTraceur: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90267 (owner: 10MarkTraceur) [22:27:06] brion: I think I enabled gate-and-submit on it [22:27:15] yeah i see that running [22:27:23] ok i'll avoid pushing 'submit' on the next one so it doesn't step on jenkins' toes [22:27:27] brion: Anyway, sorry about not explaining how to test, I have a basic information template that I'm using locally that I'll give you [22:27:34] ah awesome [22:27:54] brion: http://paste.marktraceur.info/378 four pages including the file description [22:28:08] brrrrrion. [22:28:35] * MatmaRex needs someone to review https://gerrit.wikimedia.org/r/#/c/84315/ at last [22:28:42] And presumably you have CommonsMetadata installed, because we're assuming that [22:28:53] * AaronSchulz loves it when car horns are stuck on [22:31:15] lemme make sure that's on :D [22:31:19] ahh, deprecation notices make tests fail too [22:32:37] * AaronSchulz kicks testAddIdentifierQuotes [22:32:42] (03CR) 10BryanDavis: [C: 032] corrected issue where batch upload breaks off early [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90149 (owner: 10Dan-nl) [22:32:48] (03Merged) 10jenkins-bot: corrected issue where batch upload breaks off early [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/90149 (owner: 10Dan-nl) [22:32:54] marktraceur: CommonsMetadata in the extensions metarepository isn't checking out a submodule for me :( [22:33:54] hmm [22:34:13] marktraceur: nope it's checking out, it's just empty [22:35:38] Right, you need the first commit...maybe the second [22:35:52] brion: https://gerrit.wikimedia.org/r/#/q/status:open+project:mediawiki/extensions/CommonsMetadata,n,z [22:36:05] (03CR) 10Brion VIBBER: "Can you list the external changesets this depends on? Can't test without something from CommonsMetadata extension apparently..." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90267 (owner: 10MarkTraceur) [22:36:14] * marktraceur is on it [22:36:17] tx [22:37:40] (03PS3) 10Aaron Schulz: Cleaned up DatabaseBase constructor to use an array [core] - 10https://gerrit.wikimedia.org/r/90653 [22:37:56] (03PS5) 10MarkTraceur: Add license to the header [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90267 [22:38:10] (03PS2) 10BryanDavis: Small JobQueueFederated performance tweaks [core] - 10https://gerrit.wikimedia.org/r/90499 (owner: 10Aaron Schulz) [22:38:36] (03CR) 10BryanDavis: [C: 032] Small JobQueueFederated performance tweaks [core] - 10https://gerrit.wikimedia.org/r/90499 (owner: 10Aaron Schulz) [22:38:42] "Unrecognized value for parameter 'iiprop': extmetadata" [22:38:43] hrm [22:39:04] (03PS3) 10MarkTraceur: Fetch extended metadata from the API [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89103 [22:39:21] (03PS4) 10MarkTraceur: Fetch extended metadata from the API [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89103 [22:39:36] ah a core change too [22:39:44] ? [22:39:53] (03PS5) 10MarkTraceur: Add source and authour to the title div [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89128 [22:40:13] (03PS5) 10MarkTraceur: Add description to the DOM [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89242 [22:40:33] (03PS3) 10MarkTraceur: Add creation time to list [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90040 [22:40:53] brion: Yeah, it adds a hook for CommonsMetadata [22:40:56] ok [22:41:02] brion: Sorry it's so complicated :( [22:41:14] tgr and I are trying to get it sorted, but it's taking a while [22:43:14] so... now i've got extmetadata entries, but no License in there [22:44:00] (03Merged) 10jenkins-bot: Small JobQueueFederated performance tweaks [core] - 10https://gerrit.wikimedia.org/r/90499 (owner: 10Aaron Schulz) [22:45:54] brion: i think the pastebin code does not contain a valid license [22:46:17] (03CR) 10Brion VIBBER: [C: 04-1] "With the dependent core changes and unmerged changesets to CommonsMetadata pulled down and enabled, I get 'extmetadata' in the imageinfo b" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/89103 (owner: 10MarkTraceur) [22:46:21] some of the subtemplates needed are missing [22:47:23] you can just add a category with a license name like Category:CC-BY-SA-3.0 as a crappy temporary way of testing it [22:47:48] i'll review it again when there's enough information in the commit summary to reproduce the functionality :) [22:47:48] Urgh [22:48:03] I think that you've actually found me out [22:48:09] I haven't updated for tgr's latest work :) [22:48:27] hehe [22:48:46] brion: I might just have tgr start reviewing them, I know he's got the setup [22:49:05] that way lies "oops it worked when we tested it" when you set it up somewhere else... [22:49:08] And a lot of this is meant to be pretty commons-specific, though I think we're hiding it fairly well [22:49:09] (03CR) 10CSteipp: [C: 032] "Working in test" [extensions/OAuth] - 10https://gerrit.wikimedia.org/r/89107 (owner: 10Aaron Schulz) [22:49:14] (03Merged) 10jenkins-bot: Avoid duplicate key error on /authorize page [extensions/OAuth] - 10https://gerrit.wikimedia.org/r/89107 (owner: 10Aaron Schulz) [22:49:26] Well, OK [22:56:51] (03PS2) 10BryanDavis: filebackend: try to combine SH and EX lock acquisition [core] - 10https://gerrit.wikimedia.org/r/68220 (owner: 10Aaron Schulz) [22:57:24] (03CR) 10BryanDavis: "Manually rebased. Only conflict was in phpdoc comments." [core] - 10https://gerrit.wikimedia.org/r/68220 (owner: 10Aaron Schulz) [22:57:40] (03PS3) 10BryanDavis: filebackend: try to combine SH and EX lock acquisition [core] - 10https://gerrit.wikimedia.org/r/68220 (owner: 10Aaron Schulz) [22:59:32] (03CR) 10Brion VIBBER: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90455 (owner: 10MarkTraceur) [22:59:39] (03CR) 10Jdlrobson: [C: 04-1] "The two button approach should have further discussion especially considering Jared's comments. We can't assume they are a power user. For" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [23:00:37] brion: https://www.mediawiki.org/wiki/Extension:MultimediaViewer#Setting_up_context_for_metadata helpful? [23:00:49] * brion reads [23:01:25] (03CR) 10Bartosz Dziewoński: "Other skins will continue to use "Go" as default, so we'd have to add a special case for Vector, which will suck… eh." [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [23:02:16] almost :D [23:02:18] lemme poke one more ting [23:02:39] (03CR) 10MarkTraceur: "Well, or video formats, or audio formats, for which the interface is nowhere near ready. There are a bunch of possible extensions that won" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90455 (owner: 10MarkTraceur) [23:03:10] marktraceur: maybe you could dump your test wikis into xml and attach to the extension page? that way anyone can set it up quickly via special:import [23:03:27] Eughhhh [23:03:48] I'll do something like that, sure [23:04:09] that might be the way to go :) [23:04:10] or i can dump mine, it only has a single image though [23:04:12] greg-g: That awesome link I mentioned thanks to ori-l? http://ganglia.wikimedia.org/latest/?r=day&cs=&ce=&tab=v&vn=VisualEditor ;-) [23:04:17] (03CR) 10Bartosz Dziewoński: "And then we need to switch the tooltip and the (hidden) text (for accessibility) on the button if JS is enabled, and we can't do that sane" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [23:04:44] weird, i added [[Category:CC-BY-SA-3.0]] specifically and it's still not picking it up [23:04:54] our skinning everything sucks so fucking much. [23:05:17] brion: it is cached probably [23:05:25] James_F: where's the load time? [23:05:29] * brion stabs $wgCacheEpoch :D [23:06:06] (03CR) 10Jdlrobson: "I'm still confused by "Go" - this is when you go directly to page?" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [23:06:43] greg-g: Right-hand graph; blue is cached, orange is uncached. [23:07:25] James_F: ah, but no "time to load", right? [23:07:58] (03CR) 10Bartosz Dziewoński: "> If I click the search icon or press enter without a search result selected i do a full text search" [core] - 10https://gerrit.wikimedia.org/r/82100 (owner: 10Bartosz Dziewoński) [23:08:09] ok it's time to debug CommonsMetadata and find out what's going on inside [23:08:24] greg-g: It's time-to-load metrics for inside VE. [23:08:30] greg-g: But we could do the same for other loads. [23:09:15] oh, maybe I'm confused, one sec [23:09:22] * AaronSchulz rotates brion's cache epoch [23:09:53] is "Avg 9xx" saying average 900ms and "max 1.2k" max 1.2 second? [23:10:21] (03CR) 10Bartosz Dziewoński: "OK, actually I can just use