[00:00:08] (03PS4) 10Dr0ptp4kt: WIP: Allow Skins to define other HTML tag attributes [core] - 10https://gerrit.wikimedia.org/r/100920 (owner: 10Jdlrobson) [00:00:30] (03CR) 10Cmcmahon: [C: 032] "run tests on test2wiki" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100932 (owner: 10Cmcmahon) [00:01:27] (03CR) 10Parent5446: [C: 04-2] "http://www.gossamer-threads.com/lists/wiki/wikitech/404912" [core] - 10https://gerrit.wikimedia.org/r/100924 (owner: 10Dr0ptp4kt) [00:03:30] (03CR) 10Chad: "Has hard-coded WMF stuff. Inappropriate for core as-is." [core] - 10https://gerrit.wikimedia.org/r/100924 (owner: 10Dr0ptp4kt) [00:05:05] (03Abandoned) 10Dr0ptp4kt: Dynamically generate HTML5 webapp manifests and cache manifests. [core] - 10https://gerrit.wikimedia.org/r/100924 (owner: 10Dr0ptp4kt) [00:05:20] (03PS1) 10Mattflaschen: Whitelist 'useskin' and 'uselang' when choosing buttons to show [extensions/GettingStarted] - 10https://gerrit.wikimedia.org/r/100934 [00:06:12] ^ spagewmf, have a moment for a very short GettingStarted review (https://gerrit.wikimedia.org/r/100934)? [00:08:40] (03CR) 10Catrope: [C: 032 V: 032] CSSMin: Fix remapOne() for URLs that are proto-relative or have query part [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100918 (owner: 10Jforrester) [00:10:34] (03PS1) 10Catrope: Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100936 [00:10:44] (03CR) 10Catrope: [C: 032 V: 032] Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100936 (owner: 10Catrope) [00:12:27] bd808, not sure yet how to best address the jobs run against a glam server … from what i understand thus far based on what you said … running n jobs and then waiting t minutes before running another n jobs is not a good idea? [00:13:36] Waiting a bit is good for our side (keeps you from monopolizing the queue runners) but it won't really do much to help the GLAM side. [00:13:38] it would be better to run n jobs continuously making sure only p parallel jobs were running at the same time? [00:14:28] If you could find a clean way to limit the number of active jobs per domain that would be good, yes. [00:15:39] (03CR) 10Aaron Schulz: [C: 032] Always give sane output for Title::getUserCaseDBkey() [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [00:15:46] (03CR) 10Parent5446: "Wouldn't a better solution be to just make sure that mUserDBKey is actually set all the time? In reality all properties of an object *shou" [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [00:16:03] k, there's no obvious way to do that atm? [00:16:26] Aaron|home: Is there a job runner setting to throttle the maximum number of jobs of a given type that are run in parallel? [00:16:27] (03CR) 10Aaron Schulz: Always give sane output for Title::getUserCaseDBkey() [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [00:16:36] * bd808 asks the authority [00:16:53] (03CR) 10Aaron Schulz: "Removing CR due to comment race." [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [00:17:43] Aaron|home: dan-nl and I are wondering about DOS controls from our end when pulling thousands of images off a GLAM server [00:21:36] (03PS2) 10Krinkle: No 'Show extended details' for file metadata without extra details [core] - 10https://gerrit.wikimedia.org/r/99744 (owner: 10Umherirrender) [00:22:40] (03CR) 10Brian Wolff: "Personally I think it makes more sense (or at least equal sense) to check in the getter. Lots of properties of titles aren't set during c" [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [00:23:58] (03PS2) 10Krinkle: Remove -o and -moz vendor prefixes from transition LESS mixin [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [00:24:16] (03PS3) 10Krinkle: Remove -o and -moz vendor prefixes from transition LESS mixin [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [00:24:44] (03CR) 10Krinkle: [C: 031] mediawiki.mixins: Remove -o and -moz vendor prefixes for .transition [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [00:28:17] bd808: fyi 10169 records from rijksmuseum from 2013-12-11 20:18:16 to 2013-12-12 00:20:43 … looks like it's only missing 4 http://commons.wikimedia.beta.wmflabs.org/w/index.php?title=Category:Rijksmuseum [00:30:14] dan-nl: Awesome. There are some orphans in /tmp but not too many. I'll add them to the bug so we don't lose track [00:30:28] k [00:31:09] gwicke: http://parsoid.wmflabs.org/_html/ seems broken [00:36:38] (03PS3) 10Krinkle: Only show 'Show extended details' for file metadata having extra rows [core] - 10https://gerrit.wikimedia.org/r/99744 (owner: 10Umherirrender) [00:36:45] (03CR) 10Krinkle: [C: 031] Only show 'Show extended details' for file metadata having extra rows [core] - 10https://gerrit.wikimedia.org/r/99744 (owner: 10Umherirrender) [00:37:03] (03PS4) 10Krinkle: mediawiki.mixins: Remove -o and -moz vendor prefixes for .transition [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [00:37:08] (03CR) 10Krinkle: [C: 031] mediawiki.mixins: Remove -o and -moz vendor prefixes for transition [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [00:46:02] mutante: broken in which way? [00:47:17] (03PS1) 10MaxSem: Set default for nickname preference [core] - 10https://gerrit.wikimedia.org/r/100938 [00:47:25] (03PS7) 10Werdna: Add new jquery.scroll.js plugin [core] - 10https://gerrit.wikimedia.org/r/96949 [00:49:48] (03CR) 10Krinkle: "I noticed one small bug. When hovering the SVG watched star (filled blue) and the gold border appears, the star shrinks down by 1 or 2 pix" [core] - 10https://gerrit.wikimedia.org/r/99258 (owner: 10M4tx) [00:51:26] (03PS1) 10Bsitu: Add Flow to extension array in default.conf [tools/release] - 10https://gerrit.wikimedia.org/r/100940 [00:51:35] gwicke: nevermind, it's just the mime type change [00:56:32] (03PS1) 10Dan-nl: consistent-job-delay [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100942 [00:57:50] (03PS1) 10Catrope: Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100944 [00:58:05] (03CR) 10Catrope: [C: 032 V: 032] Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100944 (owner: 10Catrope) [00:59:06] bd808, just committed https://gerrit.wikimedia.org/r/100942. in order to further limit the number of jobs it puts into the job queue at a given time, [01:00:14] bd808, do you want someone else to check the code before a merge? [01:00:45] dan-nl: Looking now [01:06:20] (03CR) 10Mattflaschen: ""it is likely that transitions are buggy and not very smooth in older browsers anyway."" [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [01:07:10] (03CR) 10BryanDavis: [C: 032] consistent-job-delay [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100942 (owner: 10Dan-nl) [01:07:15] (03Merged) 10jenkins-bot: consistent-job-delay [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100942 (owner: 10Dan-nl) [01:07:38] bd808 thanks! [01:09:59] (03CR) 10JGonera: "I don't have much to support that. I know that transitions are buggy in Android 2.x browser, but on the other hand what isn't buggy there." [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [01:19:34] dan-nl: I think I found the source of your leak [01:19:46] cool [01:19:47] (03PS2) 10Leaves in Motion: Use hooks instead of $wgOut when $wgCategoryTreeForceHeaders is set [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/99649 (owner: 10Gerrit Patch Uploader) [01:20:08] what do you think it is? [01:20:55] https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FGWToolset/7b2cfa5edfd2cee682112e4efe661fbec14ee635/includes%2FHandlers%2FUploadHandler.php#L569 [01:21:33] When you don't like the result of verify upload you don't call $Upload->cleanupTempFile() [01:21:55] Normally $Upload->performUpload() cleans up for you [01:22:13] So the leaked files are ones that have verify problems [01:23:18] k [01:24:02] i'll add that $Upload->cleanupTempFile() to the if statement and look closer at that code as well [01:24:26] I'd say that just to be safe you should call $Upload->cleanupTempFile() at every exit point in uploadMediaFileViaUploadFromUrl() [01:24:39] still, some of those orphans, i believe, were not from gwtoolset … some were large tiff files [01:24:41] * bd808 wishes that php had a finally clause [01:24:54] :) [01:25:00] Some of the vary large tiff files were actually from you [01:25:10] That's why verify failed [01:25:23] the url says jpg but the file is actually tiff [01:25:24] (03CR) 10Werdna: "I've addressed your concern. Could we push this through, please?" [core] - 10https://gerrit.wikimedia.org/r/96949 (owner: 10Werdna) [01:25:51] ah, strange … but not surprised ... [01:25:56] "UploadBase::verifyExtension: mime type image/tiff mismatches file extension jpeg, rejecting file" [01:26:38] right … not surprised that their server may do that … was probably the maps from the lv server [01:26:52] btw, the new job queue delay works well now [01:26:59] excellent [01:27:03] just tried it out with a small batch [01:27:41] (03PS5) 10Werdna: Consolidate and clean up "scroll to this element" code. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 [01:27:56] hopefully that will also help with dos issues [01:28:08] So here's what I'm thinking: patch for the tmp leak and run another few batches. Optimally with known "bad" files in there (the ones that were rejected today?) [01:28:10] going to add that $Upload->cleanupTempFile() to the code now [01:28:31] that lv map server seems to be a good candidate for that [01:28:43] i'll try that one ... [01:28:53] Then if everything looks good we can probably deploy on Tuesday. I'm voting that tomorrow is off the table at this point [01:29:02] +1's bryan [01:29:08] :) [01:29:14] * greg-g is on another call.... [01:29:34] greg-g: Don't you want to be a manager when you grow up :) [01:30:42] (03PS2) 10Krinkle: mediawiki.js: Minor clean up of comments and code [core] - 10https://gerrit.wikimedia.org/r/100518 [01:31:17] Does someone want to merge https://gerrit.wikimedia.org/r/100801 please? Otherwise I'm going to have to port it out of the branch for wmf7... [01:32:38] (03PS6) 10Werdna: Consolidate and clean up "scroll to this element" code. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 [01:32:54] Reedy: I'll look at it [01:33:24] It's a revert of a recent commit that breaks stuff [01:33:28] (03CR) 10Mattflaschen: "Can you remove it from the commit message, then?" [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [01:33:57] Need to do something about https://gerrit.wikimedia.org/r/#/c/100619/ too, but I'm still not sure about it. So will likely just wmf7 it pending a better fix.. [01:34:23] (03CR) 10Werdna: [C: 032] Revert "(bug 54951) Add Malayalam aliases for 828 and 829" [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100801 (owner: 10Reedy) [01:34:49] Thanks [01:35:06] (03CR) 10Leaves in Motion: [C: 031] "I tested this new changeset on a local MediaWiki installation and it seems to work well (the bug is resolved and it does not seem to have " [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/99649 (owner: 10Gerrit Patch Uploader) [01:35:12] (03PS5) 10JGonera: mediawiki.mixins: Remove -o and -moz vendor prefixes for transition [core] - 10https://gerrit.wikimedia.org/r/99779 [01:39:50] (03Merged) 10jenkins-bot: Revert "(bug 54951) Add Malayalam aliases for 828 and 829" [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100801 (owner: 10Reedy) [01:41:20] (03CR) 10Bartosz Dziewoński: "I really wanted this merged by now so that I could start using it :(" [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [01:41:40] anyone wanna review? ^ "SpecialWantedcategories: Show more current information when in cached mode" [01:44:15] sure [01:44:38] (03CR) 10Krinkle: [C: 031] mediawiki.mixins: Remove -o and -moz vendor prefixes for transition [core] - 10https://gerrit.wikimedia.org/r/99779 (owner: 10JGonera) [01:45:33] (03PS1) 10Dan-nl: bug-58299 [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 [01:47:00] (03CR) 10Ori.livneh: "ping" [core] - 10https://gerrit.wikimedia.org/r/99547 (owner: 10Ori.livneh) [01:47:23] bd808: https://gerrit.wikimedia.org/r/#/c/100946/ … i didn't add the cleanup on success … only the non-successful exit points ... [01:54:26] (03PS7) 10Werdna: Consolidate and clean up "scroll to this element" code. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 [01:55:01] (03CR) 10BryanDavis: [C: 04-1] "Tiny tweak needed for the commit message." (031 comment) [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 (owner: 10Dan-nl) [01:56:07] (03PS2) 10Dan-nl: bug-58299 [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 [01:57:22] (03PS8) 10Werdna: Consolidate and clean up "scroll to this element" code. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 [01:58:31] (03PS1) 10Yurik: Better warning for Message object [core] - 10https://gerrit.wikimedia.org/r/100948 [01:58:58] (03CR) 10Werdna: "PS5: Rebase" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 (owner: 10Werdna) [01:59:57] (03PS3) 10Dan-nl: bug-58299 [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 [02:00:29] (03PS9) 10Werdna: Consolidate and clean up "scroll to this element" code [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 [02:00:44] (03CR) 10Werdna: "PS9: Commit message." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/95975 (owner: 10Werdna) [02:01:08] bd808, i think that took care of it. gerrit notification bot didn't like Bug: 58299 in the message twice [02:02:43] (03PS4) 10BryanDavis: Call UploadBase::cleanupTempFile before exiting [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 (owner: 10Dan-nl) [02:03:32] (03CR) 10BryanDavis: [C: 032] Call UploadBase::cleanupTempFile before exiting [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 (owner: 10Dan-nl) [02:03:38] (03Merged) 10jenkins-bot: Call UploadBase::cleanupTempFile before exiting [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 (owner: 10Dan-nl) [02:03:41] thanks :) [02:04:40] deployment-jobrunner08:/tmp is empty now, so hopefully it will stay that way :) [02:06:00] prepping the lv map upload [02:06:50] I'm going to have to head out to dinner pretty soon, but I'll check in on things in the morning. [02:08:37] sounds good [02:08:43] thanks for all your help [02:08:54] will head to bed once the batch job gets started [02:10:06] (03PS1) 10Ori.livneh: Add 'mediaWikiLoadComplete' measurement [extensions/NavigationTiming] - 10https://gerrit.wikimedia.org/r/100951 [02:10:50] (03CR) 10Ori.livneh: "@Krinkle: I933a1e3a2 introduces usage." [core] - 10https://gerrit.wikimedia.org/r/95070 (owner: 10Ori.livneh) [02:12:18] great, got a 503 ... [02:15:59] (03PS1) 10Leaves in Motion: Use Title instead of Request for the canonical url [core] - 10https://gerrit.wikimedia.org/r/100952 [02:19:53] ori-l, do you have a sec to +2 https://gerrit.wikimedia.org/r/#/c/100948/ [02:20:21] any brave soul? Its a oneliner :) [02:21:09] I'm reviewing it [02:22:25] yurik: can you add literal quotes around the message key? [02:22:46] that is, change ' . $this->getKey() . ' to "' . $this->getKey() . '" [02:22:51] ori-l, you mean "msg" ? [02:22:52] sec [02:23:58] (03PS2) 10Yurik: Better warning for Message object [core] - 10https://gerrit.wikimedia.org/r/100948 [02:23:59] ori-l, ^ [02:24:26] funny how i was debating if i should do it... [02:24:32] should always go with the gut feeling )) [02:25:42] (03CR) 10Ori.livneh: [C: 032] Better warning for Message object [core] - 10https://gerrit.wikimedia.org/r/100948 (owner: 10Yurik) [02:25:56] thx [02:29:10] (03Merged) 10jenkins-bot: Better warning for Message object [core] - 10https://gerrit.wikimedia.org/r/100948 (owner: 10Yurik) [02:41:54] bd808|BUFFER: got the batch job started … results will show on this page http://commons.wikimedia.beta.wmflabs.org/w/index.php?title=Category:National_Library_of_Latvia [02:46:26] (03CR) 10Krinkle: [C: 04-1] "The update to the page load queue will take ~ 30 days to propagate. However the deletion of the module will be instantaneous." [extensions/WikimediaShopLink] - 10https://gerrit.wikimedia.org/r/98885 (owner: 10Ori.livneh) [02:48:11] hey greg-g, things seem to be progressing well [02:48:39] (03CR) 10Krinkle: "(the mw.config globals and js will likely have to be modified though, can be simplified to create the link unconditionally)." [extensions/WikimediaShopLink] - 10https://gerrit.wikimedia.org/r/98885 (owner: 10Ori.livneh) [02:49:01] greg-g any concerns on your side that i should look at later today? [02:50:15] (03CR) 10Daniel Friesen: [C: 04-1] "The title's local url is not always the correct canonical url." [core] - 10https://gerrit.wikimedia.org/r/100952 (owner: 10Leaves in Motion) [03:18:10] (03CR) 10Aaron Schulz: [C: 032] Fix QueryPage transaction plan [core] - 10https://gerrit.wikimedia.org/r/100728 (owner: 10Tim Starling) [03:18:21] ori-l, so i ran into an unexpected issue - could be workedaround but dirty: apparently Content object has no notion of the title [03:18:36] content does not correlate in any case with the title [03:18:56] (03CR) 10Leaves in Motion: [C: 04-1] "Oops! I hadn't checked that." [core] - 10https://gerrit.wikimedia.org/r/100952 (owner: 10Leaves in Motion) [03:19:17] which could have been fine, except that if I want to be able to have sub-namespaces, i somehow have to pass the current title around (or use wgTitle... not my fav) [03:21:14] (03Merged) 10jenkins-bot: Fix QueryPage transaction plan [core] - 10https://gerrit.wikimedia.org/r/100728 (owner: 10Tim Starling) [03:23:04] (03PS1) 10Werdna: Convert moderation success dialogs to inline boxes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100953 [03:27:02] (03CR) 10Daniel Friesen: "Old revisions have different content than the current rev so they shouldn't share the canonical URL." [core] - 10https://gerrit.wikimedia.org/r/100952 (owner: 10Leaves in Motion) [03:29:09] (03Abandoned) 10Werdna: Make Flow use core version of scrollIntoView. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/96950 (owner: 10Werdna) [04:09:11] (03PS1) 10Shirayuki: i18n: localize with "gwtoolset-step-1-heading" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100955 [04:14:26] (03Abandoned) 10Legoktm: Implement global user CSS/JS module [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/7274 (owner: 10Szymon Świerkosz) [04:32:19] (03CR) 10Legoktm: "Minor thing, looks good otherwise." (031 comment) [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [04:32:34] (03CR) 10Tim Starling: [C: 031] "Better than it was, so +1 from me. But please consider Brad's suggestion." [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100709 (owner: 10Krinkle) [04:35:48] (03CR) 10Tim Starling: [C: 032] Be slightly more user friendly in generate.php. [extensions/TrustedXFF] - 10https://gerrit.wikimedia.org/r/99709 (owner: 10Dr0ptp4kt) [04:35:51] (03Merged) 10jenkins-bot: Be slightly more user friendly in generate.php. [extensions/TrustedXFF] - 10https://gerrit.wikimedia.org/r/99709 (owner: 10Dr0ptp4kt) [04:37:22] (03PS6) 10Werdna: [WIP] Add Flow events to Special:Contributions. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 [04:38:15] (03CR) 10Werdna: "PS4 and PS5 were building on PS2, rather than PS3. So PS3's changes were clobbered." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 (owner: 10Werdna) [04:39:41] (03PS6) 10Tim Starling: IcuCollation: Sort digits under localised digits' headings [core] - 10https://gerrit.wikimedia.org/r/89488 (owner: 10Bartosz Dziewoński) [04:39:47] (03CR) 10Tim Starling: [C: 032] IcuCollation: Sort digits under localised digits' headings [core] - 10https://gerrit.wikimedia.org/r/89488 (owner: 10Bartosz Dziewoński) [04:43:42] hello yurik [04:43:53] hi diwanship [04:44:34] (03PS4) 10Tim Starling: IcuCollation: Add CollationCkb subclass for Sorani Kurdish [core] - 10https://gerrit.wikimedia.org/r/95867 (owner: 10Bartosz Dziewoński) [04:45:00] (03CR) 10Tim Starling: [C: 032] IcuCollation: Add CollationCkb subclass for Sorani Kurdish [core] - 10https://gerrit.wikimedia.org/r/95867 (owner: 10Bartosz Dziewoński) [04:45:02] i have some queries about the project can i ask you now if you hav time now? [04:45:09] sure thing [04:45:14] (03Merged) 10jenkins-bot: IcuCollation: Sort digits under localised digits' headings [core] - 10https://gerrit.wikimedia.org/r/89488 (owner: 10Bartosz Dziewoński) [04:46:34] (03PS3) 10Ori.livneh: mediawiki.js: Minor clean up of comments and code [core] - 10https://gerrit.wikimedia.org/r/100518 (owner: 10Krinkle) [04:47:29] about the action=parse module what exactly i should implement like different examples related to action=parse? [04:48:12] well, lets see what action=parse has [04:49:48] (03Merged) 10jenkins-bot: IcuCollation: Add CollationCkb subclass for Sorani Kurdish [core] - 10https://gerrit.wikimedia.org/r/95867 (owner: 10Bartosz Dziewoński) [04:49:52] i am using sandbox for that [04:51:18] (03PS2) 10Shirayuki: i18n tweaks [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 [04:51:41] diwanship, sure, but there are two more sources: the api raw help and and doc @ mediawiki [04:51:53] diwanship, https://www.mediawiki.org/wiki/API:Parsing_wikitext [04:51:59] (03Abandoned) 10Shirayuki: i18n: localize with "gwtoolset-step-1-heading" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100955 (owner: 10Shirayuki) [04:52:03] yes yes i am refering them too [04:53:06] diwanship, but of course the http://en.wikipedia.org/w/api.php?action=help&modules=parse is much more consise [04:54:27] what parts would you like to look at in parse? [04:55:01] yurik there are 3 ways for parse module [04:55:05] i mean i could always suggest things, but i might miss a good way to use a feature due to the blinds i have on :) [04:56:44] yurik so i was thinking that i'l implement all with one or two examples of each.. how does that sound? [04:57:04] which 3 ways are you thinking about ? [04:57:57] yurik 1)Specify a page or revision, using page, pageid, or oldid. 2) Specify content explicitly, using text, title, and contentmodel. 3)Specify only a summary to parse. prop should be given an empty value. [04:58:13] (03CR) 10Ori.livneh: [C: 032] "I dislike this quiet way of phasing out in-line comments. They have their place. Forbidding them adds another item to an already long list" [core] - 10https://gerrit.wikimedia.org/r/100518 (owner: 10Krinkle) [05:00:27] yurik also i have made 2 more tutorials to show you and confirm that i am making them right please have a look and tell me 1) http://www.codecademy.com/courses/web-intermediate-en-BLea4/0/7 2) http://www.codecademy.com/courses/web-intermediate-en-BLea4/0/8 [05:00:50] diwanship, sounds very good - except that people might not know how to get pageid - so might not make sense to demo ID-related ways of querying (sure can mention it) [05:01:06] so i think its better to use it by title [05:01:16] (03Merged) 10jenkins-bot: mediawiki.js: Minor clean up of comments and code [core] - 10https://gerrit.wikimedia.org/r/100518 (owner: 10Krinkle) [05:01:23] (03CR) 10Ori.livneh: "Neither the commit message nor the file header explain what it is exactly that this plugin does, so it's difficult to assess its generalit" [core] - 10https://gerrit.wikimedia.org/r/96949 (owner: 10Werdna) [05:02:50] Ok i'l go with other two ways then [05:03:17] ori-l: Just to double check, which inline comments were you referring to? Afaik all comments were preserved. [05:03:34] also yurik one example of each is ok or shall i implement more examples [05:04:09] Krinkle: I mean comments that follow code on the same line [05:04:52] I moved them up to the line over (per code conventions I did not introduce) [05:05:12] inline comments go over the code they are about [05:05:15] I didn't remove them. [05:05:27] Passive-aggressive +2ing, nice. [05:05:34] you mean trailing line comments, right? [05:05:48] yes [05:05:55] One instance: [05:05:56] mw.loader.store.enabled = false; // localStorage failed; disable store. [05:05:58] where is the convention documented? [05:06:09] Needs more "oy gevalt!" [05:06:41] there's actually an example of it in https://www.mediawiki.org/wiki/Manual:Coding_conventions/JavaScript#Creating_elements [05:06:59] diwanship, looks good to me (title should be already normalized - we should explain other things like normalization, etc in another part of the tutorial) [05:07:54] Hm.. apparently it is no longer on the CC/JS page, I think it used to be there and on the CC/PHP one. [05:07:58] ori-l: I've updated the commit message + file header, but couldn't you tell by reading the code? [05:08:01] I mean, you have to read the code anyway. [05:08:13] I do take the point that file headers are important, though. [05:08:31] It's here though (also not written by me) http://contribute.jquery.org/style-guide/js/#comments [05:08:33] (03PS8) 10Werdna: Add new jquery.scroll.js plugin [core] - 10https://gerrit.wikimedia.org/r/96949 [05:09:01] ori-l: okay, but would you oppose introducing such a convention? [05:11:31] well, I still don't know what the rationale is [05:13:05] I'm mostly just confused by the way you described that 1 line change from the larger commit. I'm still not sure whether I understand what you meant by 'quiet way of phasing out in-line comments'. [05:13:27] I am definitely on the side of ori-l with respect to coding conventions. [05:13:35] I'm fine with reasonable levels of coding conventions [05:13:44] werdna: I could tell, but in putting this in core rather than the Flow extension, you're making the case that the plugin is sufficiently useful and generic that it should be made part of MediaWiki's JavaScript API, and the argument for that needs to be made explicitly, IMO [05:13:53] both for the purposes of code review, and for aiding discoverability [05:14:50] fair. At the risk of nit-picking the nit-pick, I would appreciate that concern made more explicit rather than "I can't tell what this is for, because there is no header and I cannot read code". [05:14:50] I could make up some rationales: Consistency (why some online and some not?), less likely to grow too long. Often misused (after the code it executes, e.g. in your specific case I'd argue it should've been over the code regardless because it reports what before these 2 lines (localstorage failed), and announces what to do next (disable store) which the line below would do. [05:15:11] Krinkle: can we just let people comment code in the way that makes sense to them and others in the situation? [05:15:20] if something's unclear, we can pick them up on that [05:15:21] werdna: fair! :) [05:15:39] these two conversations are quite related :p [05:15:50] and, yeah, basically what werdna said [05:16:02] rather than "You must use this in all situations" [05:16:28] I've found the experience of developing core MediaWiki JS endlessly frustrating because of the multitude of rules I've never heard of doing any other MW development. [05:16:41] in fairness, plenty of them are reasonable. [05:16:47] and I'm happy to fix reasonable issues. [05:16:58] Krinkle: pace Gloria, I +2'd it because I thought it was a nice improvement overall [05:17:04] I guess that depends on your observation of what makes sense. I wouldn't put this in the category of conventions that require extra patterns, it removes a pattern and with that more ways to do the same thing that have no clear benefit. I'd say it improves code review and readability if there are less variations on the same thing, especially for something simple as a comment (e.g. we always use // for i [05:17:04] nline comments and /* for blocks, not "whenever" the author feels like it). [05:17:22] yurik ok i'l start working for that and also shall i divide the whole tutorial in 3 sections like for parsing,querying ,modifying respectively [05:17:27] Anyway, I won't make commits that just change this. [05:17:31] but, for example, I spent a nontrivial amount of time wondering if my formatting for the file header addition in that last revision was up to whitespace standards. [05:17:40] Krinkle: you can make commits that just change commenting style. That doesn't matter to me. [05:17:47] Got better things to do. I just did it while reading over the code as I thought it was in our conventions. I'll drop it for now. [05:17:53] It's when you start proposing that we -1 changes because they have the wrong commenting style that my nose gets out of joint. [05:18:06] but the problem is that the variations exist whether you like them or not, since people have differing styles, and bringing them in line means pricking them repeatedly with "you're doing x wrong" [05:18:21] http://inkdroid.org/creepy-polaroid/ [05:18:40] what the [05:18:49] diwanship, sounds good. Start on the parse, and we will see how that's progressing and what would be easier for the users to understand and what might be more problematic [05:19:06] Gloria: wow! [05:19:09] ori-l: Not sure it works for you, but it was very creepy for me! [05:19:20] It knew my exact address and provided an image. [05:19:49] yes, ditto. I thought I had the location API disabled. Blah! [05:20:08] Heh. I had to temporarily disable that. [05:20:13] I totally agree there should not be too many conventions, and overal I think our conventions are very minimal. I've drastically reduced the complexity of the CC page at the very least (there's at least 3 sections I'd like to remove or move to a "Learn javascript" page instead, such as the iteration conventions, that's a case where I agree having no convention is liberating and productive) [05:20:32] yurik ok thanks for the information [05:20:46] Krinkle: Cool, so we're on the same page. I'm sorry if my comment was melodramatic. [05:21:56] one thing I like about ori-l: he owns up to it when he's been even slightly out of line. [05:21:57] Props for that. [05:22:45] i get a lot of practice :P [05:23:35] ori-l: werdna: one more thing, I'm getting very close to having a working configuration of node-jscs and esformatter. node-jscs is like phpcs (coding style, not code quality like jshint, jshint has deprecated its sporadic style-related options already), and esformatter allows one (you'll love this werdna) to automatically format code to a certain style guide. [05:23:46] ooh I do like that. [05:24:05] means less being dinged, more my editor telling me what things I'm not allowed to do. [05:25:16] we'd run node-jscs on jenkins to assert coding style (right now jshint does warn for some things like var and single quotes, but very minimal, it doesn't enforce much actual 'style' like whitespace for certain operators and other things I tend to point out). And esformatter would allow you to locally just reformat it and be done with it (ideally). [05:25:33] Krinkle: like gofmt [05:26:46] * Krinkle googled it [05:26:51] It would seem so, yes. [05:27:58] > Gofmt formats Go programs. [05:30:07] (03PS7) 10Werdna: [WIP] Add Flow events to Special:Contributions. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 [05:31:16] Gloria: http://snapjudgment.org/evil-d [05:32:15] What is this? [05:42:34] oh Gloria is MZMcBride [05:43:43] * Krinkle drops a quarter, and expects werdna to understand the reference. [05:43:57] :p [05:44:25] Such a silly expression, "Het kwartje valt". Gotta love Dutch expressions. [05:45:46] werdna: Can you fix the red link here? https://www.mediawiki.org/wiki/Manual:Coding_conventions/JavaScript#Closure [05:46:12] also, your conclusion is wrong. [05:46:13] Krinkle: we have "the penny drops" in English [05:46:19] I didn't actually know the Dutch expression [05:48:07] Krinkle: what conclusion? [05:48:28] werdna: about which one the mediawiki convention prefers and is used in the example above that paragraph [05:48:40] we wrap the entire expression, including invocation [05:48:47] well, santhosh had a go at me for wrapping the entire expression [05:49:00] so I just documented that. [05:50:11] until very recently jshint even throw an error if one used the other format because there were some concerns about older browsers where, if combined with other tricks and concatenated files, wrapping only the function can break a file. [05:50:47] so I'm pretty sure all mediawiki code uses the wrapping format if they were using jshint. Last year it become an option, and it has been removed entirely since. [05:50:59] (from jshint) [05:51:15] (03PS1) 10AndyRussG: Update EducationPage for change in IContextSource [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 [05:53:07] * werdna throws arms up in dispare [05:53:09] dispair* [05:53:12] despair* [05:59:38] $ ack '\}\s*\([^)]+\)\s*\)' --js resources/mediawiki* resources/oojs/ resources/jquery [05:59:38] $ ack '\}\s*\)\s*\(' --js resources/mediawiki* resources/oojs/ resources/jquery [05:59:50] the former interestingly returns all our libs [05:59:52] the latter all upstream libs [05:59:53] :P [06:00:02] (given 1 or 2 exceptions) [06:06:27] (03CR) 10Werdna: "PS7 is a reapplication of my changes in PS4, plus others, namely:" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 (owner: 10Werdna) [06:10:05] Wow, a commit with a fair number of digits in front of the hash: https://github.com/bennoleslie/jsjpegmeta/commit/19393281446f276d8eeb1ca7e0a4b621c7c74b85 [06:10:07] rare :) [06:10:26] 12 until the first non-09 index [06:34:58] (03PS3) 10Shirayuki: i18n tweaks [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 [06:41:46] (03PS8) 10Werdna: [WIP] Add Flow events to Special:Contributions. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 [06:45:23] (03CR) 10Werdna: "Some bugs ironed, some bugs not." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99038 (owner: 10Werdna) [06:47:16] (03CR) 10Santhosh: [C: 04-1] "Thanks. I noticed only one problem. CLDR 24 explicitly defines the 'other' plural form in the plurals.xml, like" [core] - 10https://gerrit.wikimedia.org/r/100931 (owner: 10Tim Starling) [07:01:37] how do I add a field to an HTMLForm object that already exists? [07:01:52] I don't see an easy way to do that [07:02:11] well, I don't see a way at all [07:46:39] (03PS1) 10Nemo bis: Make preference "Email me when a page or file on my watchlist is changed" true by default [core] - 10https://gerrit.wikimedia.org/r/100959 [09:14:40] (03CR) 10Parent5446: [C: 032] "Fair enough. I'll mirror Aaron's +2 then." [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [09:17:45] (03Merged) 10jenkins-bot: Always give sane output for Title::getUserCaseDBkey() [core] - 10https://gerrit.wikimedia.org/r/100917 (owner: 10Brian Wolff) [09:21:04] (03PS1) 10Nikerabbit: Make (Title|User)Array implements Countable [core] - 10https://gerrit.wikimedia.org/r/100963 [09:22:03] (03CR) 10Parent5446: [C: 04-1] SpecialWantedcategories: Show more current information when in cached mode (032 comments) [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [09:35:13] (03CR) 10Siebrand: [C: 032] "Thanks. This makes the tests pass again." [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 (owner: 10AndyRussG) [09:35:18] (03Merged) 10jenkins-bot: Update EducationPage for change in IContextSource [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 (owner: 10AndyRussG) [09:40:31] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [09:46:36] (03CR) 10Siebrand: [C: 04-1] "i18n/L10n reviewed." (035 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100953 (owner: 10Werdna) [09:46:38] (03CR) 10Parent5446: [C: 032] Make (Title|User)Array implements Countable [core] - 10https://gerrit.wikimedia.org/r/100963 (owner: 10Nikerabbit) [09:49:40] (03Merged) 10jenkins-bot: Make (Title|User)Array implements Countable [core] - 10https://gerrit.wikimedia.org/r/100963 (owner: 10Nikerabbit) [09:50:27] thanks for the merge! [10:14:28] (03CR) 10Siebrand: "Two questions:" [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100517 (owner: 10BryanDavis) [10:20:48] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 (owner: 10Matthias Mullie) [10:29:30] (03CR) 10Siebrand: "Is this work in progress or can it be merged?" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100800 (owner: 10Hashar) [10:34:27] (03CR) 10Siebrand: [C: 032] i18n tweaks [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 (owner: 10Shirayuki) [10:34:32] (03Merged) 10jenkins-bot: i18n tweaks [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 (owner: 10Shirayuki) [10:54:56] (03CR) 10Helder.wiki: mediawiki.js: Minor clean up of comments and code (031 comment) [core] - 10https://gerrit.wikimedia.org/r/100518 (owner: 10Krinkle) [11:00:18] (03CR) 10Raimond Spekking: Add message for 1.22 now that it is in Special:ExtensionDistributor (031 comment) [extensions/WikimediaMessages] - 10https://gerrit.wikimedia.org/r/100803 (owner: 10Raimond Spekking) [11:03:12] (03CR) 10Matthias Mullie: "In this patch, I'm changing one more to
  • . All 4 ocurrences of looping all error messages are now
  • ." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100387 (owner: 10Matthias Mullie) [11:08:21] (03CR) 10Siebrand: Add message for 1.22 now that it is in Special:ExtensionDistributor (031 comment) [extensions/WikimediaMessages] - 10https://gerrit.wikimedia.org/r/100803 (owner: 10Raimond Spekking) [11:15:39] (03CR) 10Nikerabbit: Update EducationPage for change in IContextSource (031 comment) [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 (owner: 10AndyRussG) [11:16:19] (03CR) 10Nikerabbit: i18n tweaks (031 comment) [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 (owner: 10Shirayuki) [11:26:32] (03PS3) 10Matthias Mullie: Refactor header JavaScript [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100788 [11:26:34] (03PS7) 10Matthias Mullie: [WIP] Edit conflicts [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 [11:26:36] (03PS5) 10Matthias Mullie: Use addError() to add errors, instead of adding to array ourselves [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100387 [11:26:38] (03PS2) 10Matthias Mullie: Add callback when form & editor have completed loading [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100830 [11:27:44] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 (owner: 10Matthias Mullie) [11:30:35] (03CR) 10Shirayuki: i18n tweaks (031 comment) [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100783 (owner: 10Shirayuki) [11:38:38] (03PS1) 10Shirayuki: add explicit "escaped()" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100972 [11:54:16] (03Abandoned) 10Theopolisme: Move cursor to end of textarea [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98245 (owner: 10Theopolisme) [12:23:53] Namaste! [12:28:17] hi Niharika [12:30:05] (03CR) 10Reedy: [C: 032] Add Flow to extension array in default.conf [tools/release] - 10https://gerrit.wikimedia.org/r/100940 (owner: 10Bsitu) [12:30:08] (03Merged) 10jenkins-bot: Add Flow to extension array in default.conf [tools/release] - 10https://gerrit.wikimedia.org/r/100940 (owner: 10Bsitu) [12:34:41] (03PS7) 10Siebrand: Replace and by and in i18n [core] - 10https://gerrit.wikimedia.org/r/99195 [12:38:15] (03PS1) 10Siebrand: Update comment for wfMsgExt() [core] - 10https://gerrit.wikimedia.org/r/100981 [12:40:40] (03PS1) 10Reedy: Use mw1.23-wmf6 branches for Wikidata [tools/release] - 10https://gerrit.wikimedia.org/r/100982 [12:42:01] (03CR) 10Reedy: [C: 032] Use mw1.23-wmf6 branches for Wikidata [tools/release] - 10https://gerrit.wikimedia.org/r/100982 (owner: 10Reedy) [12:42:02] (03Merged) 10jenkins-bot: Use mw1.23-wmf6 branches for Wikidata [tools/release] - 10https://gerrit.wikimedia.org/r/100982 (owner: 10Reedy) [12:49:19] (03PS2) 10Reedy: Update Scribunto to fix ml namespace aliases [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100806 [12:49:26] (03CR) 10Reedy: [C: 032 V: 032] Update Scribunto to fix ml namespace aliases [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100806 (owner: 10Reedy) [12:49:33] (03PS2) 10Reedy: Update Scribunto to fix ml namespace aliases [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/100807 [12:49:42] (03CR) 10Reedy: [C: 032 V: 032] Update Scribunto to fix ml namespace aliases [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/100807 (owner: 10Reedy) [12:51:20] (03PS1) 10Reedy: Pass $title parameter to getCredits() call [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/100983 [12:51:35] (03CR) 10Reedy: [C: 032] Pass $title parameter to getCredits() call [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/100983 (owner: 10Reedy) [12:55:49] (03CR) 10Reedy: [V: 032] Pass $title parameter to getCredits() call [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/100983 (owner: 10Reedy) [13:05:41] reedy@ubuntu64-web-esxi:~/git/mediawiki/core$ git submodule update --init --recursive extensions/ [13:05:41] fatal: reference is not a tree: e36213362ef459f92e8b09167e537e9312f55ae3 [13:05:41] Unable to checkout 'e36213362ef459f92e8b09167e537e9312f55ae3' in submodule path 'extensions/Elastica/Elastica' [13:05:41] Failed to recurse into submodule path 'extensions/Elastica' [13:07:07] Local issues it seems... [13:08:33] hi.... [13:59:57] Iḿ trying to figure out which JS file in the ULS is responsible for the positioning of the ULS window. Any ideas? [14:08:44] Niharika1: hey! I think santhosh_ wrote most of it, so he might know [14:09:02] Niharika1: if he's not able to respond, do send an email to the mediawiki-i18n mailing list [14:09:09] (https://lists.wikimedia.org/mailman/listinfo/mediawiki-i18n) [14:09:41] YuviPanda: Okay. I have a silly-ish question. What does i18n mean? :P [14:10:09] Niharika1: Internationalization. It's the process of writing software so that it can be used with multiple languages without issue [14:10:19] Niharika1: it's called 'i18n' because there are 18 chars between i and n :D [14:10:53] YuviPanda: Cool. :D [14:10:58] :) [14:22:51] l10n [14:22:55] t12n :D [14:23:04] h4r [14:23:36] Nice naming convention. :D [14:26:22] https://en.wiktionary.org/wiki/i18n [14:26:45] https://en.wiktionary.org/wiki/Category:English_numerical_contractions [14:27:54] Nemo_bis: awesome [14:28:11] Nemo_bis: how do you manage to find all that knowledge !? [14:28:27] hashar: shhhh it's a firefox extension [14:28:57] Firefox > extensions > AutomaticallyLookSmartOnIrc [14:29:21] I'm sure there's an irssi plugin for that too [14:38:13] hashar: t12n? [14:38:41] might have been t10n [14:38:46] anyway something for: translation [14:38:48] ah [14:38:49] right [14:38:54] agilization [14:38:56] yeah confusing definitely [14:38:58] :D [14:39:05] scrummization [14:39:10] alienization [14:39:11] I don't even know if they are different or same [14:39:16] hashar: good fit :P [14:39:20] mingleization [14:39:34] nodejsization [14:39:57] dephpization [14:40:00] na that one has no luck to win against pythonization [14:40:05] unmediwikification [14:40:47] keep it up and I'll go all rhcp on you guys [14:40:51] wikitextification [14:41:11] apergos: red hat's fork of cp? :) [14:41:18] red hot chili peppers [14:41:21] ah [14:41:25] californication :-P [14:41:32] * YuviPanda googles [14:41:33] YuviPanda: naah wikitext is demode [14:43:43] unfortunately Ward Cunningham has taken the name "wiki" in the npm module namespace for his own invention :) [15:04:27] (03PS1) 10Anomie: Add ApiBeforeMain hook [core] - 10https://gerrit.wikimedia.org/r/100996 [15:22:02] (03PS1) 10Manybubbles: More BetaFeatures check to hooks called later [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 [15:27:42] (03PS1) 10Matthias Mullie: Fetch latest title content from API instead of relying on (old) html [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101001 [15:28:33] (03PS2) 10Manybubbles: Move BetaFeatures check to hooks called later [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 [15:30:24] (03CR) 10Anomie: [C: 031] "Looks sane. Haven't tested." [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [15:31:25] (03CR) 10Manybubbles: [C: 031] "Tested and it works for me but I'm gun shy on core." [core] - 10https://gerrit.wikimedia.org/r/100996 (owner: 10Anomie) [15:33:17] (03PS1) 10Hashar: integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 [15:33:23] (03CR) 10jenkins-bot: [V: 04-1] integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 (owner: 10Hashar) [15:33:50] (03PS2) 10Hashar: integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 [15:33:55] (03CR) 10jenkins-bot: [V: 04-1] integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 (owner: 10Hashar) [15:37:21] (03PS3) 10Hashar: integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 [15:38:22] (03CR) 10Hashar: [C: 032] "deployed" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 (owner: 10Hashar) [15:38:38] (03Merged) 10jenkins-bot: integration-jjb-config-test now points to git:// Zuul URL [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101003 (owner: 10Hashar) [15:38:48] (03PS2) 10Hashar: Enable voting phpcs-strict-HEAD for LocalisationUpdate [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100785 (owner: 10Siebrand) [15:39:31] (03PS3) 10Hashar: Add Python 3.3 syntax checker [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/98275 (owner: 10Merlijn van Deen) [15:39:41] (03CR) 10jenkins-bot: [V: 04-1] Add Python 3.3 syntax checker [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/98275 (owner: 10Merlijn van Deen) [15:40:11] (03PS1) 10Cmcmahon: focus on the Login button before clicking it [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101004 [15:41:01] (03PS8) 10Matthias Mullie: [WIP] Edit conflicts [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 [15:41:03] (03PS2) 10Matthias Mullie: Fetch latest title content from API instead of relying on (old) html [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101001 [15:41:14] (03CR) 10jenkins-bot: [V: 04-1] [WIP] Edit conflicts [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 (owner: 10Matthias Mullie) [15:42:13] (03PS9) 10Matthias Mullie: [WIP] Edit conflicts [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 [15:42:15] (03CR) 10Cmcmahon: [C: 032] focus on the Login button before clicking it [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101004 (owner: 10Cmcmahon) [15:43:09] (03CR) 10Bartosz Dziewoński: "Thanks for review." (033 comments) [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [15:44:55] (03PS2) 10Bartosz Dziewoński: SpecialWantedcategories: Show more current information when in cached mode [core] - 10https://gerrit.wikimedia.org/r/100210 [15:46:07] (03CR) 10Brian Wolff: [C: 04-1] "There's also a:" [core] - 10https://gerrit.wikimedia.org/r/100825 (owner: 10Aaron Schulz) [15:46:20] (03CR) 10Brian Wolff: "s/unrelated/related" [core] - 10https://gerrit.wikimedia.org/r/100825 (owner: 10Aaron Schulz) [15:57:15] (03CR) 10AndyRussG: Update EducationPage for change in IContextSource (031 comment) [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 (owner: 10AndyRussG) [15:57:52] (03CR) 10Hashar: "> Is this work in progress or can it be merged?" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100800 (owner: 10Hashar) [16:01:26] (03PS1) 10Cicalese: Fixed URL [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/101006 [16:02:30] (03CR) 10BryanDavis: "Sebrand: Information on the project is at https://www.mediawiki.org/wiki/Wikimania_Scholarships_app." [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100517 (owner: 10BryanDavis) [16:03:02] (03CR) 10Cicalese: [C: 032 V: 032] Fixed URL [extensions/TitleIcon] - 10https://gerrit.wikimedia.org/r/101006 (owner: 10Cicalese) [16:03:58] (03PS1) 10Njw: Release v1.7 [skins/erudite] - 10https://gerrit.wikimedia.org/r/101007 [16:05:03] (03CR) 10Njw: [C: 032 V: 032] Release v1.7 [skins/erudite] - 10https://gerrit.wikimedia.org/r/101007 (owner: 10Njw) [16:13:11] (03CR) 10Hashar: "Thank you! Been looking at it yesterday :-D" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/100946 (owner: 10Dan-nl) [16:14:48] (03PS1) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [16:18:53] (03CR) 10Parent5446: [C: 031] "Haven't tested." [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [16:19:19] (03PS2) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [16:19:45] (03PS1) 10Cmcmahon: get the latest mediawiki/selenium gem [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101009 [16:20:28] (03CR) 10Cmcmahon: [C: 032] "maintenance" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101009 (owner: 10Cmcmahon) [16:24:19] (03CR) 10Anomie: "The commit message here is wrong: you didn't make UserArray and TitleArray implement Countable, you made UserArrayFromResult and TitleArra" [core] - 10https://gerrit.wikimedia.org/r/100963 (owner: 10Nikerabbit) [16:26:44] (03CR) 10Hashar: "So I got hit by that issue (log dir not being writable by apache). I thought about creating a simple shell script that would create both l" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/95309 (owner: 10Krinkle) [16:28:25] hey bd808 i discussed with david the thoughts we had on the fact that commons may create a dos for a glam server. after discussing it we decided to allow the user to set the media file throttle, thus this change set https://gerrit.wikimedia.org/r/#/c/101008/. hopefully we've covered enough in that for you and tgr to be happy with it [16:29:24] bd808, also when get a chance, i'd like to know what concerns you have about not releasing gwtoolset to production today. just need a list of things to make sure i can tick them off so that we can deploy it this month [16:29:27] dan-nl: I'll take a look in a bit. Did you put a maximum in there so that a GLAM can't set the limit too high? [16:29:35] yes 1-60 [16:29:49] default of 10 [16:30:24] dan-nl: My concern was just stability. I'm not a fun of making changes right up until the deadline. [16:30:32] :) [16:31:06] The lack of /tmp leaks last night was a good sign that things are smoothing out [16:31:22] definitely that was a good catch on your part! [16:31:55] I'm going to ask around and see if anyone agrees with me about that being a flaw in the core classes [16:31:58] i have some more thoughts on job throttling but i'd like to discuss them with you after you've had a chance to review this latest change-set and … hopefully merge :) [16:32:45] i think you're definitely on target with a destructor … that's a perfect place for it [16:34:12] (03PS1) 10Yaron Koren: Changed SQL select call to work with SQLite [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101013 [16:34:24] (03CR) 10Nikerabbit: "Is there practical different on doing it on those classes instead?" [core] - 10https://gerrit.wikimedia.org/r/100963 (owner: 10Nikerabbit) [16:38:08] (03PS1) 10Hashar: register some more extensions [extensions] - 10https://gerrit.wikimedia.org/r/101014 [16:38:21] (03CR) 10Hashar: [C: 032 V: 032] register some more extensions [extensions] - 10https://gerrit.wikimedia.org/r/101014 (owner: 10Hashar) [16:40:57] (03CR) 10Anomie: "Doing it on UserArray and TitleArray means that someday if someone decides to implement a new subclass, then it too will work correctly wi" [core] - 10https://gerrit.wikimedia.org/r/100963 (owner: 10Nikerabbit) [16:57:35] (03CR) 10Hashar: [C: 032] "Job mwext-LocalisationUpdate-phpcs-strict-HEAD created in Jenkins. Will need Zuul triggers now :-]" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100785 (owner: 10Siebrand) [16:57:51] (03Merged) 10jenkins-bot: Enable voting phpcs-strict-HEAD for LocalisationUpdate [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100785 (owner: 10Siebrand) [16:58:14] (03PS5) 10Hashar: (WIP) (WIP) visualeditor-doitall (WIP) (WIP) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100800 [16:58:16] (03PS2) 10Hashar: make browsertest runner a macro [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100798 [16:58:41] (03CR) 10Yaron Koren: [C: 032 V: 032] Changed SQL select call to work with SQLite [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101013 (owner: 10Yaron Koren) [16:59:49] (03PS6) 10Nemo bis: Add direct links to translation interface on Special:AllMessages [core] - 10https://gerrit.wikimedia.org/r/94064 [17:01:13] (03PS1) 10Yaron Koren: Fix for display of spaces in history page [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101016 [17:02:13] (03CR) 10Yaron Koren: [C: 032 V: 032] Fix for display of spaces in history page [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101016 (owner: 10Yaron Koren) [17:02:44] (03PS1) 10AndyRussG: Fix EducationPage::getConfig() [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/101018 [17:05:49] (03CR) 10Siebrand: "@hashar: I think you have been trolled :D." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/100800 (owner: 10Hashar) [17:06:20] (03CR) 10CSteipp: [C: 031] "I think a hook at this point is useful. Manybubbles, can you confirm that Cirrus can use this hook (I think that's Brad's intent)?" [core] - 10https://gerrit.wikimedia.org/r/100996 (owner: 10Anomie) [17:06:57] (03CR) 10Manybubbles: "Confirmed: https://gerrit.wikimedia.org/r/#/c/100999/" [core] - 10https://gerrit.wikimedia.org/r/100996 (owner: 10Anomie) [17:08:16] siebrand: thanks :-D (WIP) [17:08:33] made me laugh reading about trolling notifications :D [17:11:08] (03CR) 10BryanDavis: [C: 031] user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [17:11:25] (03PS4) 10Aude: Update Wikibase and DataValues [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100701 [17:11:28] hello yurik r u der? [17:12:11] * siebrand grins at hashar. That commit summary was asking for it :) [17:12:21] hashar: points for originality! [17:12:52] [WIP] Invent louder WIP notice [17:14:16] (03CR) 10Siebrand: [C: 04-1] "6 keys were added, and qqq for only 4 of them. Also left inline comment about trailing whitespace in L10n." (031 comment) [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [17:14:52] (03CR) 10Chad: [C: 031] "lgtm, will merge after core change lands." [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [17:14:59] best way to prevent a patch to be merged is probably to write a failing test in it :-D [17:15:11] :D [17:15:13] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [17:15:34] one day we will probably prevent humans from voting Verified [17:15:51] (03CR) 10Chad: [C: 032] Add ApiBeforeMain hook [core] - 10https://gerrit.wikimedia.org/r/100996 (owner: 10Anomie) [17:16:11] (03CR) 10AndyRussG: "Despite the mistake noted by Nikerabbit (thanks, Nikerabbit!!), this patch does fix things following a change in core (the addition of a n" [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/100956 (owner: 10AndyRussG) [17:17:07] Hah, I was typing a commit message to +2 that, but chad beat me to it [17:18:13] (03CR) 10Siebrand: "That's clarification makes my jaw drop, Bryan. I'm very disappointed that a Wikimedia Foundation product manager thinks localisation and i" [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100517 (owner: 10BryanDavis) [17:18:23] <^d> csteipp: You can't out-gerrit the gerrit man! [17:18:51] (03Merged) 10jenkins-bot: Add ApiBeforeMain hook [core] - 10https://gerrit.wikimedia.org/r/100996 (owner: 10Anomie) [17:20:00] (03CR) 10Ragesoss: [C: 032] Fix EducationPage::getConfig() [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/101018 (owner: 10AndyRussG) [17:20:06] (03Merged) 10jenkins-bot: Fix EducationPage::getConfig() [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/101018 (owner: 10AndyRussG) [17:21:24] i wonder, do requests to /w/index.php and /w/api.php get routed to a different set of servers? [17:22:01] (03PS3) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [17:22:36] ebernhardson: i think so. I know we have an 'API cluster' [17:22:36] siebrand: Sorry about the i18n issues in Scholarships. It's just a raw matter of person hours available for getting the app ready to run. From what I've been able to understand historically the community developed app sometimes had translations and sometimes didn't. [17:22:48] (03CR) 10Dan-nl: "- addressed siebrand’s inline comment" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [17:23:20] (03PS1) 10Aude: Update Wikibase, improve JS performance [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101020 [17:23:31] In all the work I did to bring the code up to the level where we could run it on the prod cluster I tried to keep i18n support a possibility, but when we got down to the last sprint it was cut based on time available before the launch deadline. [17:23:35] ebernhardson: why? [17:23:47] YuviPanda: hmm, ok. The reason i ask is that requesting the same information in flow takes 1s via /w/index.php and ~5s via /w/api.php [17:23:51] YuviPanda: but its the exact same code path [17:23:58] (03CR) 10Chad: [C: 032] Move BetaFeatures check to hooks called later [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [17:24:06] (03Merged) 10jenkins-bot: Move BetaFeatures check to hooks called later [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [17:24:07] YuviPanda: only difference is how it bootstraps [17:24:21] bd808: siebrand would be good to address for next year but not needed this time [17:24:38] ebernhardson: hmm, perhaps performance issues in the bootstrapping code? [17:24:38] what to do when people submit applications in languages the scholarship committee doesn't understand? [17:24:52] ebernhardson: if you are in the office one of the opsen might have a clearer idea about the API cluster [17:24:53] YuviPanda: nope, the bootstraping doesn't touch network or anything like that [17:24:53] and I am off to home *wave* [17:24:54] (03CR) 10Legoktm: Move BetaFeatures check to hooks called later (032 comments) [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [17:25:06] aude: Agreed. It may actually be a good GSOC or similar project [17:25:11] but in the future the app might be used by conferences with multiple languages [17:25:15] sure :) [17:25:26] YuviPanda: hmm, who would i talk to? i never know who in ops does what(and i think there are only 2 or 3 in the office these dayS?) [17:25:47] ebernhardson: well, it is a well known fact that ori-l knows everything :D [17:25:57] especially, UK being the location this year.... [17:26:16] * aude doesn't expect bilingual tracks [17:26:32] "Nobody expects the Spanish bilingual tracks!" [17:26:46] :) [17:26:47] * bd808 slaps YuviPanda with a fish [17:27:07] * YuviPanda orders some spam for bd808 [17:27:30] * bd808 wants an albatross [17:27:43] ebernhardson: TimStarling would know too, perhaps. A cursory search on wikitech gave me nothing, though. [17:28:29] YuviPanda: ok, i'll ping them once they start the day [17:28:31] YuviPanda: hanks [17:28:31] ebernhardson: usually I just randomly find the first non-frontend person I can see and ask them and go from there. [17:28:35] heh [17:28:36] ebernhardson: yw :) [17:29:11] * YuviPanda sends bd808 an unladen swallow [17:30:36] (03PS3) 10BryanDavis: Use select to chose wikis [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100730 [17:34:21] (03PS4) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [17:35:10] (03CR) 10Dan-nl: "removed an extra ‘the’ in one of the qqq descriptions." [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [17:35:25] hey siebrand hopefully this covers it [17:36:03] (03PS1) 10Chad: Add ApiBeforeMain hook [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/101022 [17:36:07] (03PS1) 10Chad: Updating Cirrus to master [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/101023 [17:36:27] (03PS10) 10Matthias Mullie: [WIP] Edit conflicts [extensions/Flow] - 10https://gerrit.wikimedia.org/r/98836 [17:37:36] (03PS1) 10UltrasonicNXT: Fix messages not scrolling sometimes [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101024 [17:38:12] (03CR) 10UltrasonicNXT: [C: 032 V: 032] Fix messages not scrolling sometimes [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101024 (owner: 10UltrasonicNXT) [17:40:47] ebernhardson: api.php reqs are uncached, IIRC [17:41:45] (03CR) 10Manybubbles: Move BetaFeatures check to hooks called later (032 comments) [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [17:41:53] ori-l: Some should be.. [17:42:38] you can curl -v and compare the headers [17:44:18] ori-l: Numerous have anon-public-user-private or similar [17:44:33] search is all public [17:44:59] (03PS1) 10Chad: Add ApiBeforeMain hook [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101025 [17:45:02] ebernhardson: yeah, I forgot about that too. Unless explicitly asked, they are uncached [17:45:04] (03PS1) 10Chad: Updating Cirrus to master [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101026 [17:47:50] (03PS2) 10Chad: Add ApiBeforeMain hook [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/101022 [17:48:16] (03Abandoned) 10Inchikutty: Replacing 'show on map' button with a link having same functionality [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/96973 (owner: 10Inchikutty) [17:49:35] (03PS1) 10Inchikutty: Replacing 'show on map' button with a link [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/101029 [17:50:26] (03CR) 10Ori.livneh: "Have you verified that the module continues to load correctly on desktop?" [extensions/WikimediaEvents] - 10https://gerrit.wikimedia.org/r/100835 (owner: 10Robmoen) [17:51:38] ori-l: hmm, for logged in users i was under the impression everything is uncached? [17:52:15] stuff like parser cache is still cached [17:52:18] * aude thinks [17:52:27] (03PS1) 10Chad: Add ApiBeforeMain hook [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101031 [17:52:31] (03PS1) 10Chad: Updating Cirrus to master [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101032 [17:52:38] aude: yup, but we dont use that one :) just trying to figure out why the same code path takes ~3-4x longer on api.php [17:52:52] * aude nods [17:57:53] (03CR) 10Robmoen: "Yes indeed." [extensions/WikimediaEvents] - 10https://gerrit.wikimedia.org/r/100835 (owner: 10Robmoen) [18:03:14] (03PS1) 10UltrasonicNXT: Fix JS errors [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101033 [18:03:38] (03CR) 10UltrasonicNXT: [C: 032 V: 032] Fix JS errors [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101033 (owner: 10UltrasonicNXT) [18:04:35] (03CR) 10Anomie: [C: 032] Add ApiBeforeMain hook [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101031 (owner: 10Chad) [18:05:09] (03CR) 10Anomie: [C: 032] Add ApiBeforeMain hook [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101025 (owner: 10Chad) [18:06:14] (03CR) 10Legoktm: Move BetaFeatures check to hooks called later (031 comment) [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/100999 (owner: 10Manybubbles) [18:07:43] (03Merged) 10jenkins-bot: Add ApiBeforeMain hook [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101031 (owner: 10Chad) [18:08:02] (03PS3) 10Ori.livneh: Configure ext.wikimediaEvents.ve to be compatible with mobile [extensions/WikimediaEvents] - 10https://gerrit.wikimedia.org/r/100835 (owner: 10Robmoen) [18:08:09] (03CR) 10Ori.livneh: [C: 032] Configure ext.wikimediaEvents.ve to be compatible with mobile [extensions/WikimediaEvents] - 10https://gerrit.wikimedia.org/r/100835 (owner: 10Robmoen) [18:08:52] (03Merged) 10jenkins-bot: Configure ext.wikimediaEvents.ve to be compatible with mobile [extensions/WikimediaEvents] - 10https://gerrit.wikimedia.org/r/100835 (owner: 10Robmoen) [18:09:21] (03Merged) 10jenkins-bot: Add ApiBeforeMain hook [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101025 (owner: 10Chad) [18:25:33] (03PS2) 10Aaron Schulz: Avoid calling protected File field [core] - 10https://gerrit.wikimedia.org/r/100825 [18:27:13] hey siebrand is https://gerrit.wikimedia.org/r/#/c/101008/ okay now? [18:27:23] (03CR) 10Anomie: [C: 032] Updating Cirrus to master [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101032 (owner: 10Chad) [18:27:33] (03CR) 10Anomie: [C: 032] Updating Cirrus to master [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101026 (owner: 10Chad) [18:30:50] (03Merged) 10jenkins-bot: Updating Cirrus to master [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101032 (owner: 10Chad) [18:32:28] (03Merged) 10jenkins-bot: Updating Cirrus to master [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101026 (owner: 10Chad) [18:38:09] (03PS3) 10Cmcmahon: add test for Topic actions to previous Post actions [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100818 [18:38:32] (03CR) 10Cmcmahon: "merging after permissions fixed for Selenium_user on test2wiki" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100818 (owner: 10Cmcmahon) [18:41:47] (03PS3) 10Brian Wolff: SpecialWantedcategories: Show more current information when in cached mode [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [18:41:57] (03CR) 10Brian Wolff: [C: 032] "I'm not entirely convinced that an arrow is the best way to show number changed (Especially given there's a chance the old value is more a" [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [18:43:24] * MatmaRex wants ^ in wmf7 :( [18:43:28] (03CR) 10Jhall: [C: 031] add test for Topic actions to previous Post actions [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100818 (owner: 10Cmcmahon) [18:43:34] will there be deployments during christmas? probably not? [18:44:00] !roadmap [18:44:01] For long-term engineering goals roadmap, see . For short-term WMF deployments roadmap, see . [18:44:02] MatmaRex: probably not [18:44:18] https://wikitech.wikimedia.org/wiki/Deployments#Week_of_December_23 "NO DEPLOYS - Holiday vacations" :( [18:44:28] Reedy: did the enwiki deployments go out today? [18:47:39] (03Merged) 10jenkins-bot: SpecialWantedcategories: Show more current information when in cached mode [core] - 10https://gerrit.wikimedia.org/r/100210 (owner: 10Bartosz Dziewoński) [18:51:57] (03CR) 10Brian Wolff: [C: 032] "ArchiveFile::getUser is broken, but that's unrelated to this commit. I'll approve this and submit another commit to fix that." [core] - 10https://gerrit.wikimedia.org/r/100825 (owner: 10Aaron Schulz) [18:58:52] (03PS1) 10UltrasonicNXT: Add internalationisation for chat-yesterday [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101037 [18:59:10] (03CR) 10UltrasonicNXT: [C: 032 V: 032] Add internalationisation for chat-yesterday [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101037 (owner: 10UltrasonicNXT) [19:01:40] (03CR) 10Bsitu: "Looks good to me. A couple of minor issues." (033 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/99789 (owner: 10EBernhardson) [19:02:19] (03PS1) 10Brian Wolff: Fix inconsistency in ArchivedFile::getUser [core] - 10https://gerrit.wikimedia.org/r/101038 [19:03:35] qgil: is there hangout now? [19:03:54] (03PS2) 10Brian Wolff: Fix inconsistency in ArchivedFile::getUser [core] - 10https://gerrit.wikimedia.org/r/101038 [19:04:03] ah no, it's tomorrow, good. [19:04:14] saper, ufff :) [19:04:42] it's now 12DEC here and tomorrow we have 13DEC, Friday in most parts of the world [19:04:42] (03CR) 10Brian Wolff: [C: 04-1] "I should always test" [core] - 10https://gerrit.wikimedia.org/r/101038 (owner: 10Brian Wolff) [19:05:07] saper, yes, this is a true statement. :) [19:05:12] thanks. [19:05:28] good to have read-eval-print loop on IRC :) [19:05:34] (03CR) 10Chad: "Well, everybody should :)" [core] - 10https://gerrit.wikimedia.org/r/101038 (owner: 10Brian Wolff) [19:05:35] :D [19:05:46] >> var i = 1; i++; [19:05:46] marktraceur: (number) 1 [19:05:54] It really is [19:06:14] (03CR) 10Chad: [C: 032] Use select to chose wikis [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100730 (owner: 10BryanDavis) [19:06:15] >> date(); [19:06:15] saper: ReferenceError: date is not defined [19:06:23] >> Date.now(); [19:06:23] marktraceur: (number) 1386875183265 [19:06:30] >> go_to_wolfram_alpha() [19:06:31] saper: ReferenceError: go_to_wolfram_alpha is not defined [19:06:32] * bawolff wonders at the File class hierarchy. ArchivedFile is not an instance of File! [19:06:41] saper: It's ECMAscript [19:07:06] bawolff: Of course not. They're TOTALLY DIFFERENT. [19:07:08] bawolff: and ForeignDBRepo is a LocalRepo [19:07:14] nothing confusing there [19:07:16] (03Merged) 10jenkins-bot: Use select to chose wikis [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/100730 (owner: 10BryanDavis) [19:07:18] nothing at all [19:07:20] Snrk. [19:08:04] >> var page=require('webpage'); [19:08:04] saper: ReferenceError: require is not defined [19:08:09] Aaron|home: And half of ForeignAPIFile is just restoring behaviour from the File baseclass that got overridden by LocalFile... [19:08:58] we should use hooks, not subclasses :) [19:11:10] (03PS3) 10Brian Wolff: Fix inconsistency in ArchivedFile::getUser [core] - 10https://gerrit.wikimedia.org/r/101038 [19:11:39] qgil: will you be sending invitations to the hangout tomorrow? I'm loonytune997@gmail.com (this is a throwaway account meant not to infect myself with Google+ or Hangouts) [19:12:21] So in 6 years, nobody noticed that we returned the wrong username on deleted files... [19:12:54] saper, be here 10-15 minutes before the tech talk, please [19:13:02] (or 5-10) :) [19:13:15] [18:44:25] Reedy: did the enwiki deployments go out today? [19:13:27] It wasn't even the window time yet [19:13:35] what's the window time, Reedy? [19:13:43] Started 13 minutes ago [19:13:52] ah [19:13:56] noon PST [19:13:57] err [19:14:00] 11AM PST [19:14:01] got it [19:18:55] (03PS5) 10Reedy: Update Wikibase and DataValues [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100701 (owner: 10Aude) [19:19:01] (03CR) 10Reedy: [C: 032 V: 032] Update Wikibase and DataValues [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/100701 (owner: 10Aude) [19:19:13] (03PS2) 10Reedy: Update Wikibase, improve JS performance [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101020 (owner: 10Aude) [19:19:22] (03CR) 10Reedy: [C: 032 V: 032] Update Wikibase, improve JS performance [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101020 (owner: 10Aude) [19:20:07] (03CR) 10Ragesoss: [C: 032] Disable institution-page cache when logged in [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/92335 (owner: 10AndyRussG) [19:21:42] gwicke: any time to look at https://gerrit.wikimedia.org/r/#/c/99033/ ? [19:22:51] siebrand: do you have time to re-check https://gerrit.wikimedia.org/r/#/c/101008/? [19:23:12] bd808: are you only waiting on siebrand's review before merging? [19:23:46] dan-nl: I'd like Chris to take a look too. I poke him about it earlier but haven't heard back [19:24:06] ah [19:24:33] bd808: do you have time to chat about the job queue throttle a bit? [19:25:12] !seen andre__ [19:25:12] Did you mean @seen andre__? [19:25:13] dan-nl: Sure. I was about to get lunch but that can wait a bit [19:25:18] (03Merged) 10jenkins-bot: Disable institution-page cache when logged in [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/92335 (owner: 10AndyRussG) [19:25:19] @seen andre__ [19:25:19] mutante: Last time I saw andre__ they were quitting the network with reason: Quit: Cheers & thanks for the fish. N/A at 12/12/2013 4:48:09 PM (2h37m9s ago) [19:25:50] bd808: just let me know when you want to go … no worries ... [19:26:11] i was thinking about how to control the server load … if J job runners are running [19:26:45] would it make sense to create a php array and store it in the filebackend and update it [19:27:05] the array would contain the nr of jobs placed in the queue for a specific domain [19:27:31] maybe add a time stamp to it for when to release another set of jobs or a count [19:28:23] dan-nl: I don't think you can assume shared file system access for the job runners [19:28:30] that way if 16 job runners are running they would first check that file with the php array and look up the domain of the job they plan to run … if it's count or timestamp says don't run yet they would delay the job [19:28:42] And tracking something like that in a file smells bad to me [19:28:48] ah, that's one thing i was concerned about … k, what about a db? [19:29:24] maybe. That seems like the sort of thing that memcached and redis are for more than mysql [19:29:39] ephemeral data about a particular job [19:29:42] (03PS1) 10Leaves in Motion: Use ResourceLoader to load CSS [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101044 [19:29:56] well, db or file … i'd rather collect the nr of jobs for a domain … but not sure where i can store that … was thinking of adding it to the job name but i don't think that would fit into the current workflow of the job queue [19:29:57] Warning: DOMDocument::loadHTML() [domdocument.loadhtml]: Empty string supplied as input in /usr/local/apache/common-local/php-1.23wmf6/extensions/Flow/includes/ParsoidUtils.php on line 186 [19:30:02] A few of those occuring [19:30:18] ebernhardson: ^^ [19:30:57] all GWT jobs should be in $wgJobTypesExcludedFromDefaultQueue anyway, the maximum number of runners for those could easily be controlled via puppet [19:32:32] (03CR) 10Yaron Koren: [C: 032 V: 031] Use ResourceLoader to load CSS [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101044 (owner: 10Leaves in Motion) [19:32:33] Aaron|home: is there a way to find out how many are in the queue for a specific domain? [19:33:04] e.g., 60 jobs for rijksmuseum.nl [19:33:05] That would seem like a reasonable way to manage it for now. If GWToolset was limited to say 4 runners I would hope that would be ok for any GLAM to deal with [19:33:20] (03Merged) 10jenkins-bot: Use ResourceLoader to load CSS [extensions/ApprovedRevs] - 10https://gerrit.wikimedia.org/r/101044 (owner: 10Leaves in Motion) [19:33:33] k, how do we do that? [19:33:56] * dan-nl looks for $wgJobTypesExcludedFromDefaultQueue in config [19:35:25] $wgJobTypesExcludedFromDefaultQueue[] = 'gwtoolsetUploadMetadataJob'; [19:35:25] $wgJobTypeConf['gwtoolsetUploadMetadataJob'] = array( 'checkDelay' => true ) + $wgJobTypeConf['default']; [19:35:43] in CommonSettings.php ? [19:35:57] then how do we control the number of runners? [19:36:10] the first line should really be in the extension [19:37:26] k … [19:37:52] how do we control the number of runners via puppet? [19:37:59] controlling of runners is in modules/mediawiki/templates/jobrunner/jobs-loop.sh.erb [19:38:30] near the very bottom [19:38:47] * dan-nl looks for the file ... [19:39:00] dan-nl: That would be in the operations/puppet git repo [19:39:09] thanks :) [19:39:43] (03PS1) 10BryanDavis: Remove closed, private and fishbowl wikis [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 [19:41:24] Argh, stop spamming. [19:41:49] Aaron|home: so adding those jobs to wgJobTypesExcludedFromDefaultQueue requires that they be added to a job runner directly? [19:42:18] for us, it means there has to a be a line in that puppet file, yes [19:44:10] dan-nl: I'm going to grab some lunch. [19:44:27] bon apetito! [19:46:58] (03CR) 10MZMcBride: Remove closed, private and fishbowl wikis (031 comment) [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 (owner: 10BryanDavis) [19:47:11] (03CR) 10Aaron Schulz: [C: 032] Improve redis connection error log [core] - 10https://gerrit.wikimedia.org/r/95579 (owner: 10Tim Starling) [19:47:42] (03CR) 10jenkins-bot: [V: 04-1] Improve redis connection error log [core] - 10https://gerrit.wikimedia.org/r/95579 (owner: 10Tim Starling) [19:47:42] Aaron|home, let me try to squeeze it in today [19:48:26] (03CR) 10Aaron Schulz: "What places?" [core] - 10https://gerrit.wikimedia.org/r/101038 (owner: 10Brian Wolff) [19:48:51] Aaron|home: so i need to add an entry under # Start a loop for default high and low priority jobs? [19:48:52] hpriotypes="$hpriotypes gwtoolsetUploadMetadataJob gwtoolsetUploadMediafileJob" # GWToolset [19:49:27] no, I'd just add one like the "ParsoidCacheUpdateJobOnEdit"entry [19:49:39] k and * 4 for 4 runners? [19:49:55] lets start with 1 per server [19:51:06] so ... [19:51:07] (runJobsLoopService "gwtoolsetUploadMetadataJob" "n" <%= procs_per_iobound_type %> ) & [19:51:07] (runJobsLoopService "gwtoolsetUploadMediafileJob" "n" <%= procs_per_iobound_type %> ) & [19:51:49] maybe do (runJobsLoopService "gwtoolsetUploadMetadataJob gwtoolsetUploadMediafileJob" "n" <%= procs_per_iobound_type %> ) & [19:52:02] cool, if i add that change will it affect beta and production? [19:52:31] not sure about beta, maybe hashar would know [19:54:19] Aaron|home: do we need to wait for the extension to be on production before adding this puppet config change? [20:00:41] (03CR) 10Nikerabbit: "Dunno which warnings this is about as it is not mentioned, but I would recommend throwing exception in numParams instead as passing null t" [core] - 10https://gerrit.wikimedia.org/r/100948 (owner: 10Yurik) [20:00:57] (03PS1) 10Leaves in Motion: Use LEFT JOIN for SQLite compatibility [extensions/LiquidThreads] - 10https://gerrit.wikimedia.org/r/101055 [20:03:18] Nikerabbit, there are some messages bubbling up in the fatalmonitor, no idea which message is causing it :) [20:03:20] Aaron|home: how do i control the number of runners that run the GWT jobs? [20:03:43] <%= procs_per_iobound_type %> * N where N is the number of runners? [20:04:17] i guess that would be $((<%= procs_per_iobound_type %> * N)) [20:08:19] yurik: what I know, the most likely cause is null being passed to xxxParams method, and you really want to know the caller to fix those [20:08:37] greg-g: may I have a spot in the LD today to pin collection to a working version? [20:08:55] Nikerabbit, i would rather not crash production by introducing an exception into the message object :) [20:11:34] (03CR) 10Leaves in Motion: "I am new to SQL explicit joins and MediaWiki in general, so I'm not sure it's a good fix. But it appears to work. At least it doesn't show" [extensions/LiquidThreads] - 10https://gerrit.wikimedia.org/r/101055 (owner: 10Leaves in Motion) [20:14:18] yurik: well, just log the backtrace or something [20:14:22] (03PS1) 10Bsitu: History in '4 hours ago' should not be included in 'Today' [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101059 [20:15:15] Nikerabbit, i will if message key doesn't give enough info [20:17:10] (03CR) 10TMg: [C: 04-1] "Sorry for being a little nitpicking. :-) I would like to suggest two improvements. Overall the solution looks very good to me. Thanks." (032 comments) [core] - 10https://gerrit.wikimedia.org/r/100930 (owner: 10Bartosz Dziewoński) [20:17:37] Reedy: for justice: https://gerrit.wikimedia.org/r/#/c/98074/ [20:18:17] plop [20:18:47] (03CR) 10Anomie: [WIP] Make API follow file redirects like normal redirects (033 comments) [core] - 10https://gerrit.wikimedia.org/r/100921 (owner: 10Brian Wolff) [20:18:57] mwalker: yeah, plz add yourself to calendar, thanks! [20:21:11] dan-nl: procs_per_iobound_type is 1 [20:21:16] there are 16 runners [20:21:48] how do i limit the number of runners that will pickup the get jobs? [20:23:56] it's trickier if you want less than 16 for some reason (something like the transcode servers) [20:24:30] (03PS1) 10Mwalker: Revert "Rewrite of interaction with renderer" [extensions/Collection] - 10https://gerrit.wikimedia.org/r/101062 [20:24:35] if 16 runners is fine, then that line works [20:25:06] k. i'll leave it as is for now and we can address issues later as necessary ... [20:25:17] (03PS1) 10Legoktm: Add hook for modifying FormSpecialPage fields [core] - 10https://gerrit.wikimedia.org/r/101063 [20:25:18] our concern is flooding a glam with too many downloads [20:25:40] we have a throttle in place which the glam can adjust 1-60 media file jobs added to the queue per minute [20:25:46] (03PS2) 10Mwalker: Revert "Rewrite of interaction with renderer" [extensions/Collection] - 10https://gerrit.wikimedia.org/r/101062 [20:26:08] but i imagine a backlog could build up and potentially all 16 runners would run against the same glam at once [20:26:18] MaxSem: do you have time to review this: https://gerrit.wikimedia.org/r/#/c/101062/ [20:26:36] was hoping there might be a way to limit that somehow .... [20:26:47] mwalker, FUBAR?:P [20:26:57] (03PS4) 10Brian Wolff: Fix inconsistency in ArchivedFile::getUser [core] - 10https://gerrit.wikimedia.org/r/101038 [20:27:03] MaxSem: I haven't yet figured out what's wrong with it [20:27:04] (03CR) 10Brian Wolff: "It occurred to me after I wrote that, that the weird behaviour only occurred after 8bd386 - but basically if you try to oversight an alrea" [core] - 10https://gerrit.wikimedia.org/r/101038 (owner: 10Brian Wolff) [20:27:10] so I'm changing us to a deploy branch [20:27:15] and that reverts it from the deploy branch [20:27:19] master still has it in it [20:27:26] (03CR) 10MaxSem: [C: 032] Revert "Rewrite of interaction with renderer" [extensions/Collection] - 10https://gerrit.wikimedia.org/r/101062 (owner: 10Mwalker) [20:27:38] sigh [20:27:51] *nods* [20:28:25] (03CR) 10MaxSem: [V: 032] Revert "Rewrite of interaction with renderer" [extensions/Collection] - 10https://gerrit.wikimedia.org/r/101062 (owner: 10Mwalker) [20:29:12] (03Abandoned) 10Chad: Add ApiBeforeMain hook [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/101022 (owner: 10Chad) [20:29:30] (03Abandoned) 10Chad: Updating Cirrus to master [core] (wmf/1.23wmf5) - 10https://gerrit.wikimedia.org/r/101023 (owner: 10Chad) [20:30:23] greg-g: Have added an LD for VE this afternoon – hope that's OK? i18n breakage in wmf7. :-( [20:31:19] (03CR) 10BryanDavis: Remove closed, private and fishbowl wikis (031 comment) [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 (owner: 10BryanDavis) [20:31:28] (03CR) 10Chad: [C: 04-1] Remove closed, private and fishbowl wikis (031 comment) [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 (owner: 10BryanDavis) [20:34:41] (03CR) 10Legoktm: Remove closed, private and fishbowl wikis (031 comment) [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 (owner: 10BryanDavis) [20:40:39] James_F: s'all right [20:40:55] greg-g: Cool. I noted that mwalker added one in the same minute as me. :-) [20:41:17] yeah, knew about his :) [20:43:19] (03CR) 10AndyRussG: "Thanks again for working on this, and really sorry for not getting to it sooner! It seems there is a small mistake (though nothing is brok" (031 comment) [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/98567 (owner: 10Siebrand) [20:43:28] https://www.mediawiki.org/wiki/Special:Code/MediaWiki/115794#c33541 [20:43:29] Spam? :/ [20:44:39] looks like it [20:45:15] :-( [20:45:55] gone [20:47:23] greg-g: Hmm; possibly could we have a 15:00-16:00 slot instead of the LD (we're going to need to scap). [20:47:45] ah, yeah. [20:47:49] James_F: wanna edit the calendar? [20:47:54] Sure. [20:47:57] wiki that is, I'll get gcal [20:48:24] (03CR) 10Legoktm: Vector: New beta module with new typography styles (031 comment) [core] - 10https://gerrit.wikimedia.org/r/79948 (owner: 10Jdlrobson) [20:49:05] greg-g: {{done}} [20:49:22] (03PS1) 10UltrasonicNXT: Add config vars $wgChatKicks & $wgChatSocialAvatars [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101067 [20:49:41] James_F: ty [20:53:00] (03CR) 10Brian Wolff: [WIP] Make API follow file redirects like normal redirects (032 comments) [core] - 10https://gerrit.wikimedia.org/r/100921 (owner: 10Brian Wolff) [21:00:11] (03PS1) 10Legoktm: Ignore testRunLegacyHooks in findHooks.php [core] - 10https://gerrit.wikimedia.org/r/101069 [21:00:12] (03PS1) 10Legoktm: Add missing directories to findHooks.php [core] - 10https://gerrit.wikimedia.org/r/101070 [21:02:36] (03PS2) 10UltrasonicNXT: Add config vars $wgChatKicks & $wgChatSocialAvatars [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101067 [21:16:56] (03PS1) 10Aude: Update Wikibase, fix lua integration [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101095 [21:23:02] (03Abandoned) 10Aude: Update Wikibase, fix lua integration [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101095 (owner: 10Aude) [21:23:49] (03PS1) 10Aude: Update Wikibase, fix lua integration [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101103 [21:24:05] (03CR) 10CSteipp: "Is there a discussion of why this is desired? If it's to reduce the concurrent jobs so we don't kill the target, that's fine. But please h" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [21:27:51] (03CR) 10Reedy: [C: 032 V: 032] Update Wikibase, fix lua integration [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101103 (owner: 10Aude) [21:28:05] (03PS1) 10Aude: Update Wikibase, fix lua integration [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101104 [21:29:11] (03CR) 10Reedy: [C: 032 V: 032] Update Wikibase, fix lua integration [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101104 (owner: 10Aude) [21:32:05] (03PS1) 10Mwalker: Change Collection to a deploy branch [tools/release] - 10https://gerrit.wikimedia.org/r/101108 [21:32:59] Reedy: if you're around, would you kindly CR ^ [21:35:01] (03CR) 10UltrasonicNXT: [C: 032 V: 032] Add config vars $wgChatKicks & $wgChatSocialAvatars [extensions/MediaWikiChat] - 10https://gerrit.wikimedia.org/r/101067 (owner: 10UltrasonicNXT) [21:38:58] (03CR) 10Raimond Spekking: "I am not sure what happens here. Commit message say "Conflicts: Collection.i18n.php"." [extensions/Collection] - 10https://gerrit.wikimedia.org/r/101062 (owner: 10Mwalker) [21:39:54] mwalker: ping [21:40:02] gwicke: pong [21:40:09] hey bd808, saw csteipp's comments on https://gerrit.wikimedia.org/r/101008 … not sure how to refer to the discussion ... [21:40:26] I'll meet with Nuria on Monday to chat about node packaging [21:40:42] and to give her an idea of how we use node so far [21:40:54] (she is considering to use it for analytics backend stuff) [21:41:05] are you interested in joining? [21:41:25] dan-nl: Probably the best thing to do is file a feature enhancement bug against GWToolset to describe the usecase and link it to the commit [21:41:35] ya; I'll come along [21:42:04] mwalker: k, once I figure out what her nick is I'll let her know [21:46:00] (03PS4) 10Cmcmahon: add test for Topic actions to previous Post actions [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100818 [21:46:12] (03CR) 10Cmcmahon: [C: 032] "needed rebase" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100818 (owner: 10Cmcmahon) [21:47:31] (03CR) 10Dan-nl: "- discussion happened in irc and with david haskiya in skype" [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [21:47:45] bd808, k added some references ... [21:48:25] will that work? [21:49:15] (03CR) 10Aaron Schulz: Fix inconsistency in ArchivedFile::getUser (031 comment) [core] - 10https://gerrit.wikimedia.org/r/101038 (owner: 10Brian Wolff) [21:49:21] dan-nl: Ask csteipp not me :) [21:50:31] csteipp: do those comments cover your concerns? are you okay with giving https://gerrit.wikimedia.org/r/101008 a +1? [21:51:37] dan-nl: not really... i'm trying to find the answer of, "why 60?" [21:51:43] (03PS2) 10BryanDavis: Remove closed, private and fishbowl wikis [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 [21:51:55] csteipp the default is 10 [21:52:12] the max is 60 … david liked the idea of a max of 60 media file items per minute [21:52:20] as a max [21:52:46] https://gerrit.wikimedia.org/r/#/c/101008/4/includes/Config.php line 75 [21:54:21] (03PS5) 10Brian Wolff: Fix inconsistency in ArchivedFile::getUser [core] - 10https://gerrit.wikimedia.org/r/101038 [21:54:34] csteipp: i believe david came up with that max when we ran a 3000+ rijksmuseum batch upload … their server averaged 60 media files/min based on his calculations [21:58:13] csteipp: if you're not okay with 60 as a max, what max would you like to see? [21:58:32] dan-nl: What concerns me would be if someone setup files that were near the max upload size (1GB on commons, iirc). At 60/min, that's 8 Gbps.... which would likely dos any other server, and be a significant chunk of our bandwidth [21:59:13] (03CR) 10BryanDavis: Remove closed, private and fishbowl wikis (032 comments) [wikimedia/wikimania-scholarships] - 10https://gerrit.wikimedia.org/r/101051 (owner: 10BryanDavis) [21:59:17] (03PS1) 10Aaron Schulz: Removed stray ! in isQueueDeprioritized() [core] - 10https://gerrit.wikimedia.org/r/101118 [22:00:16] dan-nl: Lets set it at 20, and then someone can ask ops what they think is appropriate. [22:00:22] (03PS1) 10Aaron Schulz: Removed stray ! in isQueueDeprioritized() [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101119 [22:00:32] (03CR) 10Aaron Schulz: [C: 032] Removed stray ! in isQueueDeprioritized() [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101119 (owner: 10Aaron Schulz) [22:00:52] csteipp: cool, that's all i need … i'll add that to the assembla ticket i referred to as well [22:00:56] (03PS1) 10EBernhardson: Only roundtrip through parsoid for wikitext when necessary [extensions/Flow] - 10https://gerrit.wikimedia.org/r/101120 [22:02:41] (03CR) 10BryanDavis: [C: 032] Removed stray ! in isQueueDeprioritized() [core] - 10https://gerrit.wikimedia.org/r/101118 (owner: 10Aaron Schulz) [22:04:14] (03PS1) 10Nemo bis: Remove ArticleFeedbackTool v4 [tools/release] - 10https://gerrit.wikimedia.org/r/101122 [22:04:15] dan-nl: If you can also open a bugzilla ticket, that would help. Then we can track op's response too [22:04:54] k, do you mean create a ticket that would track requests for a higher max? [22:05:52] (03PS2) 10Nemo bis: Remove ArticleFeedbackTool v4 [tools/release] - 10https://gerrit.wikimedia.org/r/101122 [22:08:37] (03Merged) 10jenkins-bot: Removed stray ! in isQueueDeprioritized() [core] - 10https://gerrit.wikimedia.org/r/101118 (owner: 10Aaron Schulz) [22:10:47] (03PS5) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [22:10:49] (03CR) 10jenkins-bot: [V: 04-1] user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [22:12:25] (03CR) 10Nemo bis: "Raymond, I was wondering if in your opinion for such highly visible messages we should make a sort of exception and carry over the current" [core] - 10https://gerrit.wikimedia.org/r/100724 (owner: 10MegaAlex) [22:18:57] (03PS6) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [22:19:56] (03CR) 10Dan-nl: "- addressed csteipp’s concerns in ps 4." [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [22:20:56] csteipp: i think i covered everything we discussed … [22:24:36] (03CR) 10Tim Starling: "Actually, that grammar says that the condition is required, and that samples are not valid unless they follow a condition. But UTS #35 has" [core] - 10https://gerrit.wikimedia.org/r/100931 (owner: 10Tim Starling) [22:31:15] (03PS1) 10Jdlrobson: Count categories in deploy script [tools/release] - 10https://gerrit.wikimedia.org/r/101130 [22:31:21] (03CR) 10jenkins-bot: [V: 04-1] Count categories in deploy script [tools/release] - 10https://gerrit.wikimedia.org/r/101130 (owner: 10Jdlrobson) [22:32:17] (03PS2) 10Jdlrobson: Count categories in deploy script [tools/release] - 10https://gerrit.wikimedia.org/r/101130 [22:50:32] YuviPanda: Oi, staff channel. [22:50:35] (03PS7) 10Dan-nl: user-job-throttle [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 [22:51:32] YuviPanda: just was going to say that a situation in -operations would have been improved with the dashboard tool you're writing ;) [22:52:12] (03CR) 10Dan-nl: "adding a bug to the commit as requested." [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [22:52:16] csteipp: hopefully i got that right … wasn't sure if that's what you meant by adding a bug in bugzilla [22:56:16] RoanKattouw: would appreciate reviews of https://gerrit.wikimedia.org/r/#/c/95070/ and https://gerrit.wikimedia.org/r/#/c/99547/ if you have the time -- the former is one line, the latter you already know from VE [23:11:39] (03PS1) 10BryanDavis: Use TempFSFile in UploadFromUrl [core] - 10https://gerrit.wikimedia.org/r/101132 [23:13:10] ori-l: greg-g was telling me yesterday that you tried compiling LocalisationCache CDB files on the apaches and found that it was too slow or disruptive, is that correct? [23:13:28] TimStarling: no, he must have had me confused with someone else; possibly aaron [23:14:07] AaronSchulz Aaron|home : ring any bells? [23:14:46] I remember that Ryan was worried that it would be too slow but I don't remember if he tested it [23:15:57] TimStarling: Ryan did, I didn't do that [23:16:15] I profiled running it on terbium or such [23:16:44] greg-g was telling me something about nice levels and CPU utilisation [23:16:55] it's hard to imagine it being especially slow or disruptive [23:17:14] the methods that read the PHP files were around like 15% time or so...I assumed a lot of that was CPU and moving stuff around in memory [23:17:14] Oh, that might have been my fault [23:17:36] if there is a lot of cpu usage, then nice will slow it down a good bit presumably (when running on apaches) [23:17:46] I replaced VisualEditor.i18n.php with something that reads i18n from a bunch of JSON files [23:18:11] in wmf7 only [23:18:18] I have a deployment window to revert that change right now [23:18:21] because it's causing other issues [23:18:37] I don't think that is related [23:20:06] apparently Erik was saying that LC update is too slow and that we have to fix it [23:20:53] I think he's counting the time it takes to transmit the CDB files [23:20:56] yeah [23:21:57] (03CR) 10EBernhardson: [C: 032] "I don't think converting an array that should always have the same keys into objects that will always have the values is in the same realm" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100387 (owner: 10Matthias Mullie) [23:22:42] We should just drop support for non en [23:22:48] way faster [23:22:49] (03PS1) 10Catrope: Update VisualEditor to wmf7 branch for cherry-picks [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101136 [23:22:51] lol [23:23:08] Reedy: Except for en-rtl, of course. [23:23:09] greg-g: you saw ori-l's comment saying that he didn't work on that? [23:23:10] right, looking at the MW Core notes, it was rlane [23:23:30] (03PS1) 10Catrope: Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101137 [23:23:36] (03CR) 10Catrope: [C: 032 V: 032] Update VisualEditor to wmf7 branch for cherry-picks [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101136 (owner: 10Catrope) [23:23:37] (03Merged) 10jenkins-bot: Use addError() to add errors, instead of adding to array ourselves [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100387 (owner: 10Matthias Mullie) [23:23:42] i only take credit for other people's work once they are successful [23:23:49] (03CR) 10Catrope: [C: 032 V: 032] Update VisualEditor to wmf6 branch for cherry-picks [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101137 (owner: 10Catrope) [23:23:51] Who is invited? [23:23:51] WMF Staff San Francisco Bay based staff and their significant others/guests [23:23:54] WMF Office Interns, Work Study Students, and Volunteers and their significant others/guests [23:23:57] What kind of event should I be expecting? [23:23:59] This is a private event for the Wikimedia Foundation and their guests. As always, there will be food and drinks. [23:24:02] All beverages will be served at the bar including juices, water, alcoholic options, and soda. You must be 21 or older to consume alcohol at this event. If you are of age and choose to partake in alcoholic beverages, please make sure you and your guest drink responsibly. [23:24:07] The food menu is below. [23:24:10] If you have any food restrictions and/or allergies please let Mel know asap. [23:24:13] How should I dress? [23:24:15] oh, menu, goodie [23:24:15] This event is private so no dress code will be enforced. Feel free to take the opportunity to dress in your holiday best! [23:24:18] ah [23:24:21] sorry! [23:24:23] * greg-g hangs head [23:24:26] * Ryan working on trebuchet patches, in-progress [23:24:28] ** including the per-apache-generated i10n cache [23:24:54] TimStarling: would it be wrong to use the objectcache table to store a "link table last updated timestamp' value for de-duplicated refreshLinks jobs? [23:25:38] judging from graphite, that could at least half the amount of parsing (if not better) [23:26:18] it would take up too much space to spam redis with that [23:27:11] Aaron|home: sounds complicated [23:28:05] how is it different to the deduplication we've already got? [23:29:53] well, the job division is a mix of and older and a newer method right now. In any case, we can de-duplicate by resolving to title jobs but it spams the queue pretty hard. [23:30:05] (03CR) 10CSteipp: "A few issues I ran into trying to test. Getting close though." (035 comments) [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [23:30:29] the only reason there are 9403279 refreshLinks jobs now on enwiki is due to a silly throttling bug that was fixed [23:30:49] (03CR) 10Gergő Tisza: [C: 04-1] "This will work correctly, but it is way more complicated than it should be. The general algorithm is right, but you are clinging too much " [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/101029 (owner: 10Inchikutty) [23:30:55] heh, there are 30mil pages on the wiki, I guess at the worst case we'd use like 30Gb/54Gb per redis server if each page had one [23:32:10] I suppose we could throw more servers at it and do it that way...though that seems kind of a waste of RAM [23:32:53] The jobs already update link tables, changing one extra row won't hurt much, and would allow for lots of jobs to abort without parsing nor spamming redis [23:32:53] i dunno, 50G of ram seems fairly cheap, compared to dev time [23:33:03] 30M pages, 1000 bytes per page? [23:33:16] if we used objectcache, this would be <10 of code or so [23:33:39] http://ganglia.wikimedia.org/latest/?c=Redis%20eqiad&h=rdb1001.eqiad.wmnet& [23:34:44] many of those jobs are parsoid ones [23:35:10] IMO it makes sense to do the dedup in the job itself [23:35:36] in the parsoid cache that could be a conditional HTTP request [23:36:14] actually there are very few of those jobs atm [23:36:22] not sure why actually... [23:36:44] interesting [23:36:46] currently, jobs are stored as refreshLinks2 jobs and progressively expanded to refreshLinks, controlled by priority [23:36:56] and refreshLinks jobs are de-duplicated, correct? [23:37:00] no new refreshLinks2 jobs are made atm [23:37:10] I saw that the jobs were processed rather quickly, probably because they were mostly commons and wiktionary (?) [23:37:21] (03CR) 10EBernhardson: [C: 032] Refactor header JavaScript [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100788 (owner: 10Matthias Mullie) [23:37:24] ^demon|lunch, sorry for asking about your reasons about !Gerrit-Gadgets... before seeing that you had replied in a forked thread. All good. :) [23:37:34] (03Merged) 10jenkins-bot: Refactor header JavaScript [extensions/Flow] - 10https://gerrit.wikimedia.org/r/100788 (owner: 10Matthias Mullie) [23:37:34] we are just inserting refreshLinks jobs directly? [23:38:09] TimStarling: the old way was refreshLinks2 => refreshLinks, de-duplicating the later...but that didn't work well at all due to the isQueueDeprioritize() hack to avoid queue explosion [23:38:47] the new way is refreshLinks base jobs => (a chunk of per-title refreshLinks jobs, a recursive refreshLinks jobs for the remainder) [23:39:18] they are pushed in that order and the queue is roughly FIFO, so the remainder jobs won't expand until the 500 per-title ones are done [23:39:28] (03PS2) 10BryanDavis: Update FSRepo to avoid the deprecated FileBackend construction style [core] - 10https://gerrit.wikimedia.org/r/100695 (owner: 10Aaron Schulz) [23:39:34] (03CR) 10BryanDavis: [C: 032] Update FSRepo to avoid the deprecated FileBackend construction style [core] - 10https://gerrit.wikimedia.org/r/100695 (owner: 10Aaron Schulz) [23:39:37] * Aaron|home plans to kill isQueueDeprioritize() eventually, which was a terrible hack [23:40:09] Aaron|home, so most jobs in the queue are refreshLinks currently? [23:40:12] TimStarling: the old way didn't really de-duplicate anything due to isQueueDeprioritized() and the new way can only do so much since it doesn't expand all the jobs either [23:40:16] (03PS1) 10Catrope: Update EducationPage for change in IContextSource [extensions/EducationProgram] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101139 [23:40:24] (03CR) 10Catrope: [C: 032 V: 032] Update EducationPage for change in IContextSource [extensions/EducationProgram] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101139 (owner: 10Catrope) [23:40:27] refreshLinks: 9403279 queued; 28 claimed (0 active, 28 abandoned) [23:40:28] refreshLinks2: 41271 queued; 0 claimed (0 active, 0 abandoned) [23:40:34] gwicke: yes [23:40:39] (03PS1) 10Catrope: Fix EducationPage::getConfig() [extensions/EducationProgram] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101140 [23:40:42] is this enwiki? [23:40:44] (03CR) 10Catrope: [C: 032 V: 032] Fix EducationPage::getConfig() [extensions/EducationProgram] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101140 (owner: 10Catrope) [23:41:05] TimStarling: what I propose is to do the de-duplication using a more appropriate storage that can fit lots of values [23:41:07] gwicke: yes [23:41:20] k [23:41:29] (03PS1) 10Catrope: Update EducationProgram to wmf7 branch for cherry-picks [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101141 [23:41:42] the parsoid queues are both practically empty [23:41:51] (03PS2) 10Krinkle: Clean up existing code and pass jshint [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100709 [23:41:55] ok, well I haven't seen this change, I see that I wasn't listed as a reviewer on it [23:41:55] gwicke: yeah, which seems odd [23:43:37] push/pop stats look okayish: https://graphite.wikimedia.org/render?from=-70hours&until=now&width=1200&height=680&target=*.job-insert-ParsoidCacheUpdateJobOnDependencyChange.count&target=*.job-pop-ParsoidCacheUpdateJobOnDependencyChange.count [23:44:38] (03Merged) 10jenkins-bot: Update FSRepo to avoid the deprecated FileBackend construction style [core] - 10https://gerrit.wikimedia.org/r/100695 (owner: 10Aaron Schulz) [23:44:45] I wonder where all the refreshLinks jobs come from though, and if those are legit why there weren't a similar number of parsoid jobs [23:45:35] gwicke: when I altered isQueueDeprioritized() there was a bug making it not work [23:45:54] that was fixed and backported, but an explosion of jobs still made it through [23:46:17] so are those legit jobs that were expanded too early, or are they just bogus jobs? [23:46:23] they are legit [23:46:29] so about Id378d47df... [23:46:32] (03CR) 10CSteipp: [C: 031] "Should be ok for now. Someone should check with ops to find out what we'll set this at for WMF." [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/101008 (owner: 10Dan-nl) [23:46:33] k [23:46:58] how does this improve deduplication? just by interleaving recursive jobs with non-recursive jobs? [23:47:27] yes, were as before (with isQueueDeprioritized() as it was), only backlinks for one template were done at a time) [23:48:14] but my point is that really good de-duplication can be done in the job run() method itself (for parsoid using headers, for refreshLinks using a last-updated value for title backlinks) [23:48:36] * gwicke nods [23:48:42] keeps the job queue simple too [23:49:10] well, 1000 bytes per title sounds like a lot to store a timestamp [23:49:56] what is that based on? [23:50:30] (03PS3) 10Krinkle: Clean up existing code and pass jshint [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100709 [23:50:42] (03CR) 10Krinkle: Clean up existing code and pass jshint (032 comments) [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/100709 (owner: 10Krinkle) [23:51:03] but in general, yes, fine, let's deduplicate at the page level by storing masses of data [23:51:04] * Aaron|home was just looking at the ganglia stats plus showJobs.php counts...of course some of the RAM usage is for root job timestamps and not everything is enwiki of course [23:51:51] 1000 bytes per title is for the RefreshLinks jobs, not for timestamps [23:52:11] used_memory in redis? [23:52:28] yes, also not that some fragmentation can inflate that a bit [23:52:33] *note that [23:53:44] I'm going to look at its stats [23:54:01] Aaron|home: I'm looking at 99598 and it's not obvious to me where the Queues discard jobs with empty titles. Can you give me a hint? [23:54:16] in theory a page_links_touched column would use very little space [23:54:24] bd808: JobQueueDB and JobQueueRedis files [23:54:39] Thanks [23:55:00] little for disk at least [23:56:16] (03PS1) 10Mwalker: Switching Collection to deploy branch [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101143 [23:56:49] (03PS1) 10Mwalker: Updating Collection to deploy branch [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101145 [23:57:00] (03CR) 10Mwalker: [C: 032] Updating Collection to deploy branch [core] (wmf/1.23wmf7) - 10https://gerrit.wikimedia.org/r/101145 (owner: 10Mwalker) [23:57:06] (03CR) 10Mwalker: [C: 032] Switching Collection to deploy branch [core] (wmf/1.23wmf6) - 10https://gerrit.wikimedia.org/r/101143 (owner: 10Mwalker) [23:58:00] mem_fragmentation_ratio:1.14 [23:58:07] which is not much [23:59:04] right, which is why I was kind of ignoring that