[00:00:09] http://paste.tstarling.com/p/QoIpea.html [00:00:33] 10758 orphaned revisions on dewiki, looks like more than the rest put together [00:00:49] Krinkle: I think it should be a static method on MWExceptionHandler; MWExceptionHandler::printError is a close analog in terms of purpose [00:01:39] this is a list of revisions which have a rev_page which the updates table indicates was deleted [00:01:52] ori-l: yeah, MWExceptionHandler looks right [00:01:55] (03PS3) 10EBernhardson (WMF): Rename summaries to headers [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 (owner: 10Werdna) [00:01:55] thx [00:01:58] undeletion uses a new page_id so this is always an error [00:03:00] ori-l: Looks like getLogMessage is used in a few places outside RL, but those all call debuglog('exception') [00:03:12] so we can probably move that without a trace? [00:03:20] * Krinkle checks extension usage [00:03:27] and when it was introduced [00:03:51] <^d> I wonder if we could use pdo_mysql. Then we'd work on both. [00:04:17] 2006 by TimStarling, okay so we can't remove that method. I'll deprecate it instead and forward to our new static one [00:04:35] https://github.com/wikimedia/mediawiki-core/commit/bcb9f9e1c0dda36bfe2096e92de0ad77e21b38f8 old times :) [00:06:16] <^d> Meh, two bugs requesting mysqli support and hhvm said it's not a priority :\ [00:08:05] (03CR) 10Tim Starling: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [00:09:54] number of affected pages by year of deletion: http://paste.tstarling.com/p/LOLBWn.html [00:10:08] (03PS1) 10Robmoen: Update VisualEditor to wmf19 branch for cherry-picks [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88675 [00:10:17] [old_text] => DB://cluster20/0 [00:10:26] (03CR) 10Chad: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [00:10:29] yeah, that's broken [00:10:33] what revision is that for? [00:10:40] ahh, rev 16265152 on plwiki has bad text [00:10:45] I remember a bug that caused that [00:10:53] but getText() returns "" instead of false due to some ContentHandler logic somewhere [00:11:03] I was wondering why it wasn't false yet the sha1 was broken [00:11:44] TimStarling: the revisions use restored has non-broken text though right? [00:12:10] the one page I have investigated so far had non-broken text [00:12:21] we now have this: [00:12:23] $dbw->insert( $this->getTable( $dbw ), [00:12:23] array( 'blob_id' => $id, 'blob_text' => $data ), [00:12:23] __METHOD__ ); [00:12:23] $id = $dbw->insertId(); [00:12:23] if ( !$id ) { [00:12:25] throw new MWException( __METHOD__ . ': no insert ID' ); [00:12:27] } [00:12:37] that exception was added to fix the DB://cluster20/0 problem [00:14:48] anyway, bug 53687 is probably a race condition, the edit in the same second as the deletion seems very suspicious [00:15:31] but it seems like all the relevant places are loading with FOR UPDATE [00:17:34] hmm... [00:17:53] (03CR) 10Krinkle: [C: 031] "LGTM." [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88675 (owner: 10Robmoen) [00:20:23] not seeing anything obvious [00:21:29] both Revision::loadText() and ExternalStoreDB::fetchBlob() have a fallback to master if the slave doesn't have the row [00:21:46] which should work as long as the transaction isolation level is appropriate, and the insert is committed [00:22:09] (03PS1) 10JGonera: Don't cache LESS files with errors when debug=1 [core] - 10https://gerrit.wikimedia.org/r/88677 [00:24:21] tx_isolation is REPEATABLE-READ on both the ES masters, and all the ES slaves too [00:25:37] (03CR) 10GWicke: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/85926 (owner: 10MarkTraceur) [00:27:18] heh, I was about to think of something but say my comment in loadLastEdit()...can't be that [00:27:27] (03CR) 10Krinkle: "See also I6510b43f02cf82 which for me solved the main problem with debug mode being hard to catch less errors." [core] - 10https://gerrit.wikimedia.org/r/88677 (owner: 10JGonera) [00:29:16] if ExternalStoreDB::fetchBlob() failed then there would be an entry in external.log [00:30:07] (03CR) 10GWicke: [C: 04-1] "Adding minus one as the implementation here is lacking. If the resource attribute is used proper URLs should be returned." [core] - 10https://gerrit.wikimedia.org/r/85926 (owner: 10MarkTraceur) [00:30:30] gwicke: FWIW we were talking about it in here and we may add a data-mw-title attribute instead/first [00:31:19] (03PS1) 10GWicke: Fix SVG glyphs and make the service more robust [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 [00:32:08] which there isn't [00:32:09] (03CR) 10Bsitu: [C: 032] Use standard humanized timestamps for moderation [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88637 (owner: 10EBernhardson (WMF)) [00:32:13] (03Merged) 10jenkins-bot: Use standard humanized timestamps for moderation [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88637 (owner: 10EBernhardson (WMF)) [00:33:10] TimStarling: btw, it would be nice if ES worked in autocommit mode [00:34:04] for optimal performance? [00:34:33] performance aside, it would avoid any issue of being stuck with an old ES trx snapshot and newer DB://cluster urls (especially when FOR UPDATE is used) [00:34:57] this would only matter if a previous blob was fetched from an ES master of course [00:35:25] since you said there was no external.log entry that can't be the problem in this case though [00:36:30] 'externalTemplateOverrides' => array( [00:36:30] 'flags' => 0, // No transactions [00:36:30] ), [00:36:40] someone must have thought of this already [00:38:03] (03CR) 10Werdna: [C: 032] Generate recentchanges rows from Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/81017 (owner: 10EBernhardson (WMF)) [00:38:08] (03Merged) 10jenkins-bot: Generate recentchanges rows from Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/81017 (owner: 10EBernhardson (WMF)) [00:38:19] (03PS1) 10Krinkle: Update VisualEditor submodule to latest wmf/1.22wmf20 head [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88679 [00:38:51] (03CR) 10Krinkle: [C: 032] Update VisualEditor to wmf19 branch for cherry-picks [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88675 (owner: 10Robmoen) [00:38:57] (03CR) 10Krinkle: [C: 032] Update VisualEditor submodule to latest wmf/1.22wmf20 head [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88679 (owner: 10Krinkle) [00:39:32] TimStarling: heh, so the commit() call in the ES store function is just for sanity [00:39:51] I guess [00:40:30] if ( $id == 0 ) { [00:40:30] $this->loadPageData( 'forupdate' ); [00:40:37] is $id always zero? [00:41:28] well, it's a bug in any case [00:41:37] it shouldn't fail just because someone passed in the page_id [00:43:16] (03Merged) 10jenkins-bot: Update VisualEditor to wmf19 branch for cherry-picks [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88675 (owner: 10Robmoen) [00:44:51] (03Merged) 10jenkins-bot: Update VisualEditor submodule to latest wmf/1.22wmf20 head [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88679 (owner: 10Krinkle) [00:44:55] the comment on loadLastEdit() deals with fetching the revision row, though not the text table row [00:45:02] in theory that could be a problem too [00:45:17] (03CR) 10JGonera: "I think that your patchset is a good change, but I'm not sure if it's solving my particular problem." [core] - 10https://gerrit.wikimedia.org/r/88677 (owner: 10JGonera) [00:45:37] if the revision row is correct then it should get the text row from the master if necessary [00:45:38] TimStarling: $id seems to be 0 as used [00:45:46] TimStarling: but it would need FOR UPDATE [00:45:58] I grepped the extensions directory, they all use 0 except for Wikidata [00:46:15] I don't have time to work out whether Wikidata uses 0 [00:46:28] passing in $id definitely is more race prone probably [00:47:03] anyway, the whole loadLastEdit() comment is about the fact that using the master is not enough [00:47:38] odd that the problems seemed to go away after I changed the logic though [00:47:39] true [00:48:40] so the idea is that the snapshot on the master is opened prior to the edit [00:48:53] then the edit is committed [00:48:57] <^d> php.net has two MYSQLI_* constants that don't seem to exist in php 5.3-5.5 or master. Other than the single page on php.net, I can find no evidence they actually do exist. [00:49:10] <^d> And of course that page doesn't indicate if they were removed. [00:49:13] <^d> *sigh* [00:49:14] then the revision row is fetched from the DB with FOR UPDATE which gets it from after the snapshot [00:49:24] then the text row is fetched without FOR UPDATE which fails [00:51:13] if Revision::loadText() fails, we should log that [00:51:54] ori-l: Perfect is the enemy of the done. [00:52:11] ori-l: If there are better ways to measure and address technical debt, I have no doubt you'll implement them. :-) [00:53:51] yeah, I was going to say so but I got distracted by work. I think your list will be useful even if it remains necessarily incomplete. Rock on, etc. [00:54:21] yeah, that should really be logged [00:55:51] ori-l: Congrats on the promotion. :-) [00:59:33] thanks [01:03:19] TimStarling: maybe READ_LOCKING for revision should grab and prefill mTextRow [01:03:49] yes, that would do it [01:34:00] (03PS3) 10Werdna: Change rev_comment to rev_change_type and adjust references to "comments" to instead refer to "change types". [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88398 [01:34:26] (03CR) 10Werdna: "Rebased" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88398 (owner: 10Werdna) [01:38:20] (03CR) 10Werdna: [C: 031] "Looks fine to me, but I don't know shit about LESS." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88663 (owner: 10EBernhardson (WMF)) [01:45:03] (03CR) 10Krinkle: "OK, well, if that's the reason, that seems like something that shouldn't be solved in this method but in general." [core] - 10https://gerrit.wikimedia.org/r/88677 (owner: 10JGonera) [01:47:31] (03CR) 10Krinkle: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88663 (owner: 10EBernhardson (WMF)) [02:11:26] (03CR) 10Krinkle: "I've stared at this regex for a bit and trying to figure out what is keeping it from touching string values that look like json." [core] - 10https://gerrit.wikimedia.org/r/88651 (owner: 10PleaseStand) [02:27:22] (03PS4) 10Werdna: Rename summaries to headers [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 [02:27:30] (03CR) 10jenkins-bot: [V: 04-1] Rename summaries to headers [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 (owner: 10Werdna) [02:29:41] (03CR) 10Parent5446: [C: 031] FormatJson: minor cleanup [core] - 10https://gerrit.wikimedia.org/r/88650 (owner: 10PleaseStand) [02:29:57] (03PS5) 10Werdna: Rename summaries to headers [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 [02:47:47] (03PS5) 10thibaultmarin: (extension/PGFTikZ) PGFTikZ extension: initial import [extensions/PGFTikZ] - 10https://gerrit.wikimedia.org/r/38487 [02:49:43] (03CR) 10thibaultmarin: "(1 comment)" [extensions/PGFTikZ] - 10https://gerrit.wikimedia.org/r/38487 (owner: 10thibaultmarin) [03:01:04] (03PS1) 10Legoktm: Kill a global [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88680 [03:20:31] (03PS6) 10thibaultmarin: (extension/TimelineTable) Refactoring and cleanup [extensions/TimelineTable] - 10https://gerrit.wikimedia.org/r/38485 [03:41:43] (03PS4) 10Werdna: Add AJAX header modification [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 [03:41:51] (03CR) 10jenkins-bot: [V: 04-1] Add AJAX header modification [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [03:42:47] (03PS5) 10Werdna: Add AJAX header modification [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 [03:43:49] (03CR) 10Werdna: "PS5: Rebase and a bit more futzing around in JS" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [04:04:27] (03CR) 10Daniel Friesen: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/85926 (owner: 10MarkTraceur) [04:09:00] (03PS8) 10MarkAHershberger: Install without XML support [core] - 10https://gerrit.wikimedia.org/r/76663 [04:15:33] (03PS9) 10MarkAHershberger: Install without XML support [core] - 10https://gerrit.wikimedia.org/r/76663 [04:35:57] (03CR) 10Tim Starling: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [04:59:36] (03PS1) 10Legoktm: Support global group change notifications [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 [05:00:02] (03CR) 10Legoktm: [C: 04-1] "Still needs i18n messages." [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 (owner: 10Legoktm) [05:00:32] (03PS1) 10Werdna: Fix some VisualEditor annoyances: * Made the editor restore focus to whatever had focus, instead of allowing VE to steal it. * Hid the textarea that is being converted into a VE instance, replaced with a spinner so people do not start typing in the wron [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88684 [05:00:47] (03PS2) 10Werdna: Fix some VisualEditor annoyances: [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88684 [05:07:05] (03CR) 10Nikerabbit: "So is it a parser bug or not?" [core] - 10https://gerrit.wikimedia.org/r/88266 (owner: 10Krinkle) [05:08:53] (03CR) 10Nikerabbit: [C: 032] Native digits on 'onGetHumanTimestamp' [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88136 (owner: 10Ebrahim) [05:09:00] (03Merged) 10jenkins-bot: Native digits on 'onGetHumanTimestamp' [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88136 (owner: 10Ebrahim) [05:40:46] (03CR) 10GWicke: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/85926 (owner: 10MarkTraceur) [05:46:28] (03PS2) 10Legoktm: Support global group change notifications [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 [05:53:13] (03PS2) 10Werdna: Hide icons unless the post/topic/summary is being hovered over. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88663 (owner: 10EBernhardson (WMF)) [05:53:44] (03CR) 10Werdna: [C: 032] "Learned shit about LESS, updated commit message." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88663 (owner: 10EBernhardson (WMF)) [05:53:50] (03Merged) 10jenkins-bot: Hide icons unless the post/topic/summary is being hovered over. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88663 (owner: 10EBernhardson (WMF)) [05:57:50] (03PS1) 10Aaron Schulz: Fix revision table cleanup on delete [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88686 [05:58:17] (03CR) 10Aaron Schulz: [C: 032] Fix revision table cleanup on delete [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88686 (owner: 10Aaron Schulz) [06:02:06] (03PS1) 10Werdna: Add mw-ui-input class to VisualEditor surface. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88687 [06:02:33] (03Merged) 10jenkins-bot: Fix revision table cleanup on delete [core] (wmf/1.22wmf19) - 10https://gerrit.wikimedia.org/r/88686 (owner: 10Aaron Schulz) [06:03:57] (03PS1) 10Reviewer-bot: Test commit, version 23 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88688 [06:09:45] Oh great, if password-hashing ever gets rebased it needs to be updated to handle wgRedactedFunctionArguments [06:10:55] Dantman: I was bitching about wgRedactedFunctionArguments IRL, heh [06:11:21] I was thinking it could maybe be autogenerated from a custom phpdoc annotation of sorts [06:11:31] * Aaron|home should find time to mess with stuff like this one day [06:13:18] Heh, find time... [06:29:09] (03PS3) 10Werdna: Fix some VisualEditor annoyances: [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88684 [06:36:58] (03PS4) 10Nemo bis: No error anymore when uploading files with same base name [core] - 10https://gerrit.wikimedia.org/r/87020 (owner: 10Marco) [06:41:50] (03CR) 10Daniel Friesen: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/85926 (owner: 10MarkTraceur) [06:44:34] (03CR) 10Rfaulk: [C: 032 V: 032] add - test for test_dulwich_tag. [sartoris] - 10https://gerrit.wikimedia.org/r/88690 (owner: 10Rfaulk) [07:07:25] (03PS6) 10SuchetaG: mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [07:08:23] (03CR) 10jenkins-bot: [V: 04-1] mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [07:58:04] rebooting Jenkins server now, that will take 1+ hour. [08:31:52] Is it bad form to class UserRegistry extends RegistryPattern {} in order to have a contextual task name for the class? [08:36:49] Amgine: we do that for jobs iirc [08:37:00] Amgine: we have a base job class and each job extends it [08:37:19] lass DoubleRedirectJob extends Job {} [08:37:39] and since profiling calls use __METHOD__ that let us have different profiling per jobs [08:46:03] Cool, thanks Hashar [08:48:04] Amgine: Similar is done for exceptions too... [09:00:25] good point. [09:03:05] (03PS2) 10Reedy: Add Option to Enable Keyword Links [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/82854 [09:03:21] (03CR) 10Reedy: [C: 032] Add Option to Enable Keyword Links [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/82854 (owner: 10Reedy) [09:05:56] (03PS3) 10Reedy: Add UserGetReservedNames hook subscriber [extensions/ArticleFeedbackv5] - 10https://gerrit.wikimedia.org/r/84476 [09:07:24] (03CR) 10Physikerwelt: [C: 04-1] "for me it does not work" [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 (owner: 10GWicke) [09:08:12] (03PS4) 10Reedy: Add UserGetReservedNames hook subscriber [extensions/ArticleFeedbackv5] - 10https://gerrit.wikimedia.org/r/84476 [09:16:16] (03PS1) 10Rfaulk: mod - 1. remove obsolete tests. \n2. Add some setup to existing tests. [sartoris] - 10https://gerrit.wikimedia.org/r/88696 [09:16:44] (03CR) 10Rfaulk: [C: 032 V: 032] mod - 1. remove obsolete tests. \n2. Add some setup to existing tests. [sartoris] - 10https://gerrit.wikimedia.org/r/88696 (owner: 10Rfaulk) [09:21:44] (03CR) 10Reedy: [V: 032] Add Option to Enable Keyword Links [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/82854 (owner: 10Reedy) [09:24:07] (03CR) 10Reedy: "I wonder if we should add a minor/dash or whatever" [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/86395 (owner: 10Ori.livneh) [09:32:07] (03CR) 10Addshore: [C: 031] Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [09:35:21] (03PS2) 10Reedy: Update formatting [core] - 10https://gerrit.wikimedia.org/r/88465 (owner: 10Siebrand) [09:35:26] (03CR) 10Reedy: [C: 032] Update formatting [core] - 10https://gerrit.wikimedia.org/r/88465 (owner: 10Siebrand) [09:39:28] (03CR) 10Ori.livneh: "Sure. Maybe '~wmf-1' or something." [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/86395 (owner: 10Ori.livneh) [09:41:05] andre__: thanks for the extension rename [09:42:27] (03Merged) 10jenkins-bot: Update formatting [core] - 10https://gerrit.wikimedia.org/r/88465 (owner: 10Siebrand) [09:45:52] (03CR) 10Mattflaschen: [C: 04-1] "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/65867 (owner: 10Parent5446) [09:48:55] (03PS2) 10Ori.livneh: Haskell syntax: omit single quotes from QUOTEMARKS [extensions/SyntaxHighlight_GeSHi] - 10https://gerrit.wikimedia.org/r/86395 [09:49:47] ori-l: sure. And congrats to be the one looking at bugs with the "performance" keyword now. :P [09:51:55] heh, thanks [09:57:36] +1 [09:57:44] * Nemo_bis runs to add the keyword to dozens bugs [09:59:55] $wgMiserMode && die() [09:59:56] DONE [10:00:08] ^^ [10:00:23] ; [10:00:36] (03PS1) 10Hashar: zuul: metrics for test/gate pipelines [integration/docroot] - 10https://gerrit.wikimedia.org/r/88700 [10:00:38] redirecting pk.wikimedia.org to a page on meta, it's a user group on the way of becoming a chapter. it will be somehow temporary but noone knows how long until they are full chapter. do i use 301 or 302 ?:) [10:00:45] (03CR) 10Hashar: [C: 032] zuul: metrics for test/gate pipelines [integration/docroot] - 10https://gerrit.wikimedia.org/r/88700 (owner: 10Hashar) [10:00:47] (03Merged) 10jenkins-bot: zuul: metrics for test/gate pipelines [integration/docroot] - 10https://gerrit.wikimedia.org/r/88700 (owner: 10Hashar) [10:02:16] mutante: 307 [10:03:07] " [10:03:08] Apache HTTP client doesn't respect HTTP status code 307 [10:03:17] eh, random result :p [10:04:47] 308 - (approved as experimental) .. reads [10:16:16] (03PS1) 10Siebrand: Run PHP CS strict on 4 extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88703 [10:19:48] Nikerabbit: font URLs reference wmf branch [10:20:03] e.g. https://bits.wikimedia.org/static-1.22wmf20/extensions/UniversalLanguageSelector/data/fontrepo/fonts/amiri/amiri-regular.woff?version=1.0.2&20120101 [10:21:10] ori-l: do you have better way? [10:21:20] Reedy, hadn't you sent him to bed? [10:21:37] i protested [10:21:43] are fonts allowed on commons? [10:22:10] I believe they aren't [10:22:16] DENIED [10:23:09] ori-l: not sure whether commons wants to serve fonts for all mediawiki users or whether all mediawiki users want to rely on commons for serving fonts [10:24:01] neither seems problematic to me, tbh [10:24:18] it's not like people don't hotlink commons all over the place [10:24:44] the only precedent is Roan temporarily whitelisting the format and uploading a font locally on some wiki which was loading it from an external website against privacy policy [10:25:10] the format is not allowed, so of course there isn't any policy on the topic [10:25:14] could upload it on mediawiki.org [10:25:22] s/on/to [10:25:24] sounds like a nightmare [10:25:34] who's going to maintain that stuff? [10:26:39] Nemo_bis: That precedent isn't on commons either, that's local uploads [10:27:05] I think I might have done that on wikis that disallowed local uploads, too [10:28:00] ori-l: so what's the issue you want to tackle? changing urls? we don't necessarily need to use commons to solve that [10:28:56] I figured that would be better than just removing the font and wait for the community to put the privacy policy-violating font back. Better to keep their font working so they potentially don't even notice / find out [10:29:40] Nikerabbit: the fonts don't change that much, right? stable url + far future expires would be nice [10:31:26] (03PS1) 10Siebrand: Add PHP CodeSniffer jobs for 4 extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 [10:31:40] it's 591k on https://commons.wikimedia.org/wiki/Main_Page , though that's probably the langbar, right? [10:32:06] (03CR) 10Matthias Mullie: [C: 032] Add UserGetReservedNames hook subscriber [extensions/ArticleFeedbackv5] - 10https://gerrit.wikimedia.org/r/84476 (owner: 10Reedy) [10:32:10] (03Merged) 10jenkins-bot: Add UserGetReservedNames hook subscriber [extensions/ArticleFeedbackv5] - 10https://gerrit.wikimedia.org/r/84476 (owner: 10Reedy) [10:32:41] (03PS2) 10Siebrand: Run PHP CS strict on 4 extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88703 [10:32:47] ori-l: https://github.com/wikimedia/mediawiki-extensions-UniversalLanguageSelector/commits/master/data/fontrepo no they don't change often [10:33:09] and they are versioned separately already [10:34:05] maybe we should have a https://bits.wikimedia.org/static-current symlink pointing at the latest branch? (maybe this exists already) [10:34:42] (03PS3) 10Siebrand: Run PHP CS strict on 4 extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88703 [10:34:52] RoanKattouw: sure, it was an emergency :) [10:35:03] (does anybody sleep here??) [10:35:20] Nemo_bis: ori-l is up in the middle of the night but I'm in Europe this week [10:35:20] Sleep is overrated [10:35:32] ah good :) [10:36:05] (and I only woke up like an hour ago) [10:38:06] (03CR) 10Hashar: "There is already a project named mwext-UniversalLanguageSelector :" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 (owner: 10Siebrand) [10:38:11] ori-l: we are currently using wgExtensionAssetsPath... I do not think there is any clean way to make exceptions [10:38:18] (03PS2) 10Siebrand: Add PHP CodeSniffer jobs for 4 extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 [10:38:54] there's a https://bits.wikimedia.org/static-master/extensions/UniversalLanguageSelector/data/fontrepo/fonts/amiri/amiri-regular.woff?version=1.0.2&20120101 [10:39:09] but the perms are wrong [10:39:41] plus it's pointing at /usr/local/apache/common-local/php-master/resources, which doesn't exist [10:39:51] but other than that..! :D [10:40:14] (03CR) 10Hashar: [C: 032] Add PHP CodeSniffer jobs for 4 extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 (owner: 10Siebrand) [10:40:32] (03Merged) 10jenkins-bot: Add PHP CodeSniffer jobs for 4 extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 (owner: 10Siebrand) [10:40:57] (03CR) 10Hashar: "Created jobs:" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88704 (owner: 10Siebrand) [10:41:10] Nikerabbit: just have $wgULSFontAssetsPath = $wgExtensionAssetsPath; in UniversalLanguageSelector.php and override it for WMF prod in CommonSettings.php [10:41:32] (03PS1) 10Siebrand: Update grouping and sorting [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88707 [10:42:11] ori-l: and deliver that to JS [10:42:27] ori-l: but yeah I can make that code if you make stable paths available [10:44:20] $wgHooks[ 'ResourceLoaderGetConfigVars' ] = function ( &$vars ) { global $wgULSAssetsPath; $vars['wgULSAssetsPath'] = $wgULSAssetsPath; } [10:45:09] Reedy: do you know how to do that? presumably your make-wmf-branch scripts would need to be updated somewhere so that they update the symlink [10:46:05] (03CR) 10Hashar: [C: 032] Run PHP CS strict on 4 extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88703 (owner: 10Siebrand) [10:46:13] (03Merged) 10jenkins-bot: Run PHP CS strict on 4 extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88703 (owner: 10Siebrand) [10:52:32] hashar: is the 'static-master' used by labs? [10:52:36] *symlink [10:53:25] ori-l: I guess so [10:53:31] probably for bits.beta.wmflabs.org [10:54:10] ah, yeah [10:54:22] https://gerrit.wikimedia.org/r/#/c/28337/ [10:54:28] and yet another revert later on https://gerrit.wikimedia.org/r/#/c/42526/ [10:54:35] sam deleted them a couple time iirc [10:55:03] haha [10:55:57] (03CR) 10Hashar: [C: 032] "No jobs changed/deleted. Excellent work :-]" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88707 (owner: 10Siebrand) [10:56:15] (03Merged) 10jenkins-bot: Update grouping and sorting [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88707 (owner: 10Siebrand) [10:56:59] well, i guess static-master isn't the same as static-current [10:57:39] Reedy: would you be opposed to a static-current that points at $latestDeploymentBranch? [10:58:11] Brion's topic on wiki hosting and reading all these GAE/AWS docs on things like scaling, etc... Is giving me ideas and reviving my interest in writing a powerful self-serve MediaWiki hosting provider. [10:59:00] DanielFriesen: are you on the wikifarms mailing list? [10:59:21] because I don't quite get what they're doing, but maybe that's what orain is doing and you can contribute to their code [10:59:47] I think so. [11:00:05] But it's mostly empty. [11:03:59] Nemo_bis: Orain looks like yet another wiki farm. The self-serve I'm talking about is extremely different from a wiki farm. [11:06:11] volunteer reviews welcome https://gerrit.wikimedia.org/r/#/q/owner:dzahn+status:open,n,z [11:07:29] Speaking of "yet another wiki farm" ;) I like the name. If I ever setup a host that offered managed and self-serve MW hosting and I decided to create an extra method of wiki hosting (same style as the other wiki farms) I should call it YAWF.net (Yet Another Wiki Farm) [11:20:36] (03PS1) 10Siebrand: Update formatting [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88712 [11:21:53] (03CR) 10jenkins-bot: [V: 04-1] Update formatting [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88712 (owner: 10Siebrand) [11:32:41] DanielFriesen: well, maybe they'd like to adopt your concept if you explained it to them :) or maybe they have some code you can work on, dunno... just saying [11:33:56] No no, I really mean that what I'm talking about is quite different. [11:34:48] I had this idea a long time ago. It's closer to a real hosting provider than it is to a farm. [11:35:52] (03PS1) 10Hashar: TranslationNotifications check-only -> check [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88716 [11:35:57] Databases, some server resources, load balancing, reverse proxies, etc... are naturally shared. [11:36:14] DanielFriesen: fantastico installs on drugs? [11:36:29] (03CR) 10Hashar: [C: 032] TranslationNotifications check-only -> check [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88716 (owner: 10Hashar) [11:36:34] (03Merged) 10jenkins-bot: TranslationNotifications check-only -> check [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88716 (owner: 10Hashar) [11:37:53] Well you're a little closer... [11:39:15] There's a central dashboard/console to control things from. Naturally add new hosted wikis. [11:40:42] A huge amount of MW config is controllable through that console for the wiki. I had actually implemented a pretty nice screen for the license choosing. [11:41:06] (03PS4) 10Jeroen De Dauw: Change course status "Passed" to "Ended" to avoid ambiguity about "passing" a course. [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/78392 (owner: 10Ragesoss) [11:41:17] But the method of integration with MW is completely different. [11:41:19] (03CR) 10Jeroen De Dauw: [C: 032] "CR by Siebrand" [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/78392 (owner: 10Ragesoss) [11:41:25] (03Merged) 10jenkins-bot: Change course status "Passed" to "Ended" to avoid ambiguity about "passing" a course. [extensions/EducationProgram] - 10https://gerrit.wikimedia.org/r/78392 (owner: 10Ragesoss) [11:41:53] On a farm you have one MW source, you configure that, and every site uses that. [11:42:35] And wiki creation, configuration, etc... is implemented as an extension to the running MW site. [11:44:26] The self-serve hosting is different. Control is more akin to a virtual machine. MediaWiki is a mere guest. It runs completely vanilla being controlled from the outside by the self-serve system. [11:47:54] You create a wiki with a form; An initial database is created, initialized, and associated with the wiki, a bucket of config is associated with it, as well as a reference to a version of MediaWiki that determines the source files to use. [11:51:39] (03PS1) 10Siebrand: Remove stray newline [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88717 [11:53:07] The servers serving the wikis look up the wiki they're going to serve and get a blob of user config, an identifier of a MW version, and a reference to a database. It then serves out the wiki using the source files for the specified version of MW and gives it config based on the user config and what database is associated with the wiki. [12:01:45] The console gives control over many other things; Extensions (which have been reviewed and added to the system along with some metadata on how to handle the extension) can be configured withi a few button presses and can be configured, disabled, uninstalled, etc...; Various tasks can be ordered to run on the wiki many making mantenance scripts available; The wiki/database can be marked as read-only. And the database can be snapshotted, old [12:02:55] * And the database can be snapshotted, old snapshots can be restored from the console. With a few commands a wiki can switch between running off a new DB to an archive of the wiki. [12:04:59] (03PS1) 10Siebrand: Cleanup jobs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88718 [12:05:42] (03PS2) 10Siebrand: Cleanup jobs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88718 [12:06:10] (03CR) 10Zfilipin: "(1 comment)" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:06:55] (03CR) 10Bartosz Dziewoński: "Why is this in core if it requires an extension to work?" [core] - 10https://gerrit.wikimedia.org/r/79948 (owner: 10Jdlrobson) [12:09:11] (03PS4) 10Zfilipin: Adding new CREDITS and LICENSE files, and license headers to all files. [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:09:37] (03PS1) 10Raimond Spekking: Consistency tweaks in preparation for adding extension to translatewiki.net [extensions/VectorBeta] - 10https://gerrit.wikimedia.org/r/88719 [12:10:01] (03CR) 10Zfilipin: [C: 031] "I have implemented the suggested changes. Luis, could you please take a look ? Feel free to merge it into master if it looks good to you." [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:10:13] (03PS5) 10Zfilipin: Adding new CREDITS and LICENSE files, and license headers to all files. [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:14:40] (03CR) 10Bartosz Dziewoński: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88650 (owner: 10PleaseStand) [12:14:44] (03PS6) 10Zfilipin: Adding new CREDITS and LICENSE files, and license headers to all files. [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:15:12] (03CR) 10Zfilipin: [C: 031] "Added headers to new files." [qa/browsertests] - 10https://gerrit.wikimedia.org/r/86355 (owner: 10Yuvipanda) [12:15:41] "Upgrading a wiki" basically: Places the wiki into read-only mode. Creates a snapshot of the wiki + a live running copy of the database. Sets the live wiki to run in read-only mode off the copy/snapshot. On the non-live database/version of the wiki config it is set to use the files for a new version of MediaWiki. The database upgrade tasks are then run on that database. After this the switch is flipped and suddenly the new version of the wi [12:16:08] * After this the switch is flipped and suddenly the new version of the wiki is being served. After confirming the wiki is running OK read-only is dropped. And the old database goes offline. [12:30:53] (03PS3) 10Siebrand: Cleanup jobs for LangEng extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88718 [12:31:00] (03PS1) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:31:08] (03CR) 10jenkins-bot: [V: 04-1] Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 (owner: 10Siebrand) [12:34:05] (03CR) 10Siebrand: "recheck" [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 (owner: 10Siebrand) [12:36:29] commuting back home for conf call with hexmode [12:37:53] (03PS2) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:38:00] (03CR) 10jenkins-bot: [V: 04-1] Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 (owner: 10Siebrand) [12:42:40] (03PS3) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:42:48] (03CR) 10jenkins-bot: [V: 04-1] Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 (owner: 10Siebrand) [12:43:56] (03PS4) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:46:03] (03PS5) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:46:09] (03CR) 10jenkins-bot: [V: 04-1] Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 (owner: 10Siebrand) [12:46:43] (03PS6) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:47:28] (03PS7) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:48:21] (03PS1) 10Matthias Mullie: Various small design fixes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88725 [12:48:27] (03PS8) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:50:02] (03PS9) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:52:30] (03PS10) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [12:53:12] (03CR) 10Werdna: [C: 031] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88725 (owner: 10Matthias Mullie) [12:54:25] (03PS11) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [13:00:19] (03CR) 10Matthias Mullie: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88725 (owner: 10Matthias Mullie) [13:04:07] hexmode: sorry i am late [13:08:46] (03PS12) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [13:09:27] (03PS13) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [13:18:45] (03PS4) 10Siebrand: Cleanup jobs for LangEng extensions [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88718 [13:25:12] (03PS14) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [13:25:19] (03PS15) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [13:42:55] (03PS1) 10Manybubbles: Allow reindex count to differ from old count [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/88729 [13:45:17] (03CR) 10Hashar: "that is a rough experiment to play with tags. Got to enhance this job to use make-release repository. Merges pending there." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/86160 (owner: 10Hashar) [13:51:07] (03CR) 10Siebrand: "This isn't going to pass PHP strict as long as bug 55508 isn't resolved. The Jenkins job will be non-voting once Change-Id: I4c2c6c5ae5567" [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88712 (owner: 10Siebrand) [13:54:16] siebrand: moaar spam in my inbox :D [13:57:35] (03PS1) 10Cmcmahon: id value for Search button changed [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88736 [13:58:03] (03CR) 10Hashar: [C: 032] Remove stray newline [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88717 (owner: 10Siebrand) [13:58:21] (03Merged) 10jenkins-bot: Remove stray newline [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88717 (owner: 10Siebrand) [13:58:26] hashar: Yeah, sorry for that many tries on the patch set. There were 3 syntax issues in it... [13:58:35] siebrand: it is ok :-) [13:58:41] siebrand: so phpcs-strict-HEAD everywhere ? [13:58:42] hashar: I don't do that much yaml. [13:58:51] hashar: That's what I've decided at least. [13:58:55] (03CR) 10Zfilipin: [C: 032] id value for Search button changed [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88736 (owner: 10Cmcmahon) [13:58:58] siebrand: I thought we could use phpcs-strict on gate-and-submit to make sure phpcs is working on everything [13:59:10] siebrand: what ever you want, you are the beta tester / primary "customer" [13:59:11] hehe [13:59:20] hashar: it would be cool to at least expose the issues if they are there. [13:59:54] I added HEAD jobs to make the test faster [14:00:04] for gate-and-submit, I guess we can wait a bit longer [14:01:49] (03CR) 10Matthias Mullie: [C: 04-1] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88398 (owner: 10Werdna) [14:01:52] (03PS7) 10SuchetaG: mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [14:02:58] (03CR) 10jenkins-bot: [V: 04-1] mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [14:04:18] hasharCall: Should we allow $options as global in CodeSiffer? It's used in command line scripts. [14:04:42] legacy ones [14:05:15] hasharCall: I came across it in https://gerrit.wikimedia.org/r/#/c/88616/3/scripts/cli.inc [14:05:20] (03PS1) 10Cmcmahon: two preferences removed, collapsible and experiments [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88738 [14:05:23] * siebrand nods at Nikerabbit. [14:05:32] Nikerabbit: Is there another way to not use that global? [14:06:58] Currentl we're passing these mediawiki specific globals: messageMemc, parserMemc, IP. [14:08:06] siebrand: maintenance class can access options without the global [14:08:10] (03CR) 10Zfilipin: [C: 032] two preferences removed, collapsible and experiments [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88738 (owner: 10Cmcmahon) [14:08:18] some cli scripts haven't been converted to that in Translate yet [14:09:08] Nikerabbit: STDOUT() I pointed to, appears to be in global scope. [14:09:26] (03PS8) 10SuchetaG: mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [14:10:25] (03CR) 10jenkins-bot: [V: 04-1] mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [14:10:27] siebrand: we could get rid of it with some refactoring effort [14:10:39] Nikerabbit: k. I'll report a bug. [14:12:51] (03PS1) 10Cmcmahon: this fixes the Math test and is needed for further Flow updates [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88741 [14:13:17] Nikerabbit: Done> https://bugzilla.wikimedia.org/show_bug.cgi?id=55517 [14:14:59] I creeped some scope in, if you don't mind [14:15:34] Nikerabbit: Not at this point in time :) [14:15:45] (03PS9) 10SuchetaG: mediawiki.api: add postWithToken and getToken [core] - 10https://gerrit.wikimedia.org/r/64582 (owner: 10Nikerabbit) [14:16:18] Nikerabbit: Translate has cli.php because not all scripts use Maintenance? [14:16:34] (03CR) 10Zfilipin: [C: 032] this fixes the Math test and is needed for further Flow updates [qa/browsertests] - 10https://gerrit.wikimedia.org/r/88741 (owner: 10Cmcmahon) [14:18:17] qchris: how do I leave a comment at https://www.mediawiki.org/wiki/Git/New_repositories/Requests ? [14:20:53] (03PS1) 10Siebrand: Specify encoding of source files as utf-8 [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 [14:22:08] (03PS2) 10Siebrand: Specify encoding of source files as utf-8 [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 [14:22:41] siebrand: cli used to be some collection of useful stuff, but it's role needs to change now [14:25:35] and I have finished conf call [14:26:03] qchris: I could not find a way how to edit https://www.mediawiki.org/wiki/Git/New_repositories/Requests so I have replied at https://www.mediawiki.org/wiki/Talk:Git/New_repositories/Requests#mediawiki-selenium [14:30:16] zeljkof: Sorry. I've been in a meeting. Let me look ... [14:30:56] zeljkof: You should be able to edit https://www.mediawiki.org/wiki/Git/New_repositories/Requests/Entries [14:31:18] qchris: but I am not [14:31:35] zeljkof: Note the trailing "/Entries" [14:31:51] Anyways ... I'll make it mediawiki/selenium then. [14:31:56] I get only this text when I open the page: "Click here to visit the main Git repositories request page. Your edit has been added there. [14:31:56] This page is meant only to feed entries into the primary new Git repositories request page. The wikitable syntax is intentionally left "open" to allow for new entries submitted via the Request button. Closing the table here will break it on other wikipages." [14:32:15] qchris: thanks [14:32:40] zeljkof: Yes, the page shows that text ... But if you click the Edit tab it takes you to the Wikitext: [14:32:51] https://www.mediawiki.org/w/index.php?title=Git/New_repositories/Requests/Entries&action=edit [14:33:48] qchris: I see… :) sorry, did not know that would work [14:33:52] (03PS1) 10Hashar: run-phpcs-mw: Specify encoding of source files as utf-8 [integration/jenkins] - 10https://gerrit.wikimedia.org/r/88745 [14:34:05] (03CR) 10Hashar: [C: 04-1] "That only updated mediawiki-core-phpcs-HEAD" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 (owner: 10Siebrand) [14:34:09] zeljkof: Neither do I know how it works ... I came across it by pure chance :-) [14:34:48] (03PS3) 10Hashar: Specify encoding of source files as utf-8 [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 (owner: 10Siebrand) [14:34:59] (03CR) 10Hashar: [C: 032] run-phpcs-mw: Specify encoding of source files as utf-8 [integration/jenkins] - 10https://gerrit.wikimedia.org/r/88745 (owner: 10Hashar) [14:35:26] (03CR) 10Hashar: [V: 032] run-phpcs-mw: Specify encoding of source files as utf-8 [integration/jenkins] - 10https://gerrit.wikimedia.org/r/88745 (owner: 10Hashar) [14:37:48] gods [14:37:55] the preferences handling is so messed up [14:38:13] i was like, who wrote all that damn code [14:38:21] and then i was like, i wrote some of that damn code myself [14:39:40] (03CR) 10Hashar: [C: 032] "being deployed" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 (owner: 10Siebrand) [14:39:58] (03CR) 10Matthias Mullie: [C: 04-1] "(5 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 (owner: 10Werdna) [14:40:01] (03Merged) 10jenkins-bot: Specify encoding of source files as utf-8 [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/88743 (owner: 10Siebrand) [14:40:42] MatmaRex: Many people wrote some code in there... [14:40:57] MatmaRex: And IIRC it was the first use of the HTMLForm class. [14:41:11] ahhh HTMLForm [14:41:18] I wish we could split it in some tiny files [14:41:21] and get more doc / tutorial [14:41:28] and make it independent from the rest of mw :-D [14:41:29] siebrand: it still is the *ultimate* use [14:41:35] oh, yes. [14:41:43] siebrand: you know how the options api validates the data passed to it? [14:41:51] (you probably know, you touched that stuff too) [14:41:58] with a htmlform, naturally! [14:41:59] (03CR) 10Anomie: [C: 04-1] "Restoring -1 after rebases, see PS1" [core] - 10https://gerrit.wikimedia.org/r/86393 (owner: 10MarkTraceur) [14:42:08] MatmaRex: IIRC it stuffs it into some methods of SpecialPreferences.php (or Preferences.php) [14:42:33] $field = HTMLForm::loadInputFromParameters( $key, $prefs[$key] ); [14:42:33] $validation = $field->validate( $value, $user->getOptions() ); [14:43:11] i am responsible of the idiotic separate handling of multiselect preferences when validating and doing stuff to them [14:43:13] ah, yes I've seen that before. Interesting and intuitive, indeed. [14:43:30] and partially for the handling of checkmatrix preferences, since it was copied from mine and search&replaced, it seems [14:43:46] MatmaRex: start writing unit tests, and then untangle? [14:44:13] siebrand: i'm currently trying to make it all only use htmlform for validation in one place in Preferences themselves [14:44:26] would be rather easy [14:44:39] except that you might have unvalidatable fields saved in the database, it seems [14:44:48] (this case is handled now) [14:44:53] hasharCall: Is that UFT-8 command deployed (wanna test it...) [14:44:54] ? [14:44:57] so i'm reorganizing some stuff [14:44:58] bleh [14:45:03] (03Abandoned) 10MaxSem: Prohibit blocked users from purging [core] - 10https://gerrit.wikimedia.org/r/87268 (owner: 10MaxSem) [14:45:12] I3140f15586a67e5938e1c6bb658d9025b705e87f is the wip in gerrit [14:46:46] (03CR) 10Siebrand: "recheck" [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88712 (owner: 10Siebrand) [14:49:00] siebrand: I think it is already deployed for -strict checks [14:49:08] siebrand: they are using a shell script [14:49:17] siebrand: instead of an inline shell command in side the job [14:49:26] siebrand: so whenever I update the shell script, all jobs are magically migrated [14:49:49] siebrand: if you look at https://integration.wikimedia.org/ci/job/mwext-cldr-phpcs-strict/2/console [14:49:58] siebrand: it has a line saying: /var/lib/jenkins/bin/run-phpcs-mw.sh [14:50:09] that is the wrapper, it is in integration/jenkins.git under /bin/ [14:50:13] I know, that is messy :-((( [14:50:19] but I am a messy person! [14:50:32] ahhh cldr pass phpcs [14:50:34] awesome [14:50:42] * hasharCall does the CI dance [14:51:01] (03PS2) 10Hashar: Update formatting [extensions/cldr] - 10https://gerrit.wikimedia.org/r/88712 (owner: 10Siebrand) [14:51:50] hasharCall: Cool! Let me update another patch set, because PHPCS can be voting now, too :) [14:57:36] (03PS16) 10Siebrand: Update jobs for LangEng extensions [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/88722 [14:59:45] arhhrhr [14:59:46] ;-D [15:00:13] siebrand: the changes that were needed to add --encoding=utf-8 : https://gerrit.wikimedia.org/r/#/q/I44f42262d709b2b9ad876bdf9fc44fb5e3dfff09,n,z [15:00:50] siebrand: I will eventually migrate shell commands to actually shell scripts so we will just have to update a single file and have it deployed on jenkins slave [15:00:57] that would save us the "update a thousands jobs" step [15:01:01] hasharCall: Oh, right, I missed that second one. [15:01:11] that was hard to spot to be honest [15:01:19] I don't even know how I have spot it [15:01:29] something clicked in my brain when I have seen only one job got updated [15:01:38] anyway done now :-] [15:03:03] I got two pages of incoming reviews :( [15:03:11] and 3 pages of outgoing reviews [15:03:37] hasharCall: Taking a look... [15:03:59] hasharCall: I see only one page. Possibly because I have a 100 line limit. [15:04:22] https://gerrit.wikimedia.org/r/#/settings/preferences [15:04:23] I have a small screen [15:05:15] siebrand: https://gerrit.wikimedia.org/r/#/q/owner:hashar+is:open,n,z [15:05:28] hasharCall: or https://gerrit.wikimedia.org/r/#/dashboard/24 [15:05:42] yeah that one [15:05:45] feel my pain ? [15:08:08] (03PS3) 10JanZerebecki: Make items in the StringUtilsTest data provider easier to identify by adding names. [core] - 10https://gerrit.wikimedia.org/r/65300 [15:08:25] (03CR) 10Hashar: [C: 032] "that is slightly changing the $minor default in doQuickEditcontent() but that is fine given 0 was compared with 1 below that does not chan" [core] - 10https://gerrit.wikimedia.org/r/88519 (owner: 10Siebrand) [15:08:32] (03CR) 10Siebrand: [V: 04-1] "Doesn't merge cleanly. Please rebase." [core] - 10https://gerrit.wikimedia.org/r/64941 (owner: 10Hashar) [15:08:36] (03PS1) 10Reviewer-bot: Test commit #23 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88749 [15:09:33] (03PS1) 10Anomie: Handle raw when calculating $rawtoc [core] - 10https://gerrit.wikimedia.org/r/88750 [15:10:13] legoktm, did you forget to add your Echo.php changes to https://gerrit.wikimedia.org/r/#/c/88683/ ? [15:10:21] hasharCall: I observe that a good number of your patch sets have no reviewers at all. My advice would be to add at least one reviewer right after submitting. [15:11:08] (03CR) 10Merlijn van Deen: "Oh, great, now I can't abandon this change. That was not something I had thought of...." [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88688 (owner: 10Reviewer-bot) [15:11:11] siebrand: we have a bug for that [15:11:23] siebrand: https://bugzilla.wikimedia.org/show_bug.cgi?id=52329 [15:11:38] if that bug was fixed, we could go around adding reviewers where needed [15:11:42] legoktm, also that patch is for bug 51541 not 51540 [15:11:53] hasharCall: Your incoming reviews don't look that bad. About a third of them have a -1, some 15-20 are really open for you. [15:11:54] (03PS3) 10Alex Monk: Support global group change notifications [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 (owner: 10Legoktm) [15:12:01] (03PS2) 10Matthias Mullie: Various small design fixes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88725 [15:12:42] Nemo_bis: That looks like a feature request that should be reported upstream. [15:12:58] siebrand: for now I'm ensuring these patches have reviewers https://www.mediawiki.org/wiki/Gerrit/Reports/Open_changesets_by_newbie_owner [15:13:24] siebrand: the API does provide that info, it's just not super-practical to refetch it for all changesets every time [15:13:25] That's a great service, Federico. [15:14:01] (03CR) 10Hashar: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88513 (owner: 10Siebrand) [15:14:11] actually it seems they usually are reviewed rather quickly :) because they're mostly small patches [15:14:19] but the report has existed only for a few days ;) [15:14:35] siebrand: I still have to many bold entries pending actions :( [15:15:12] hashar: Bold sometimes is a false positive. It becomes bold for example if someone replies to your -1 with "Yes, I'll fix that"> [15:16:25] (03Abandoned) 10Siebrand: (bug 37713) Fix documentation for `MWHttpRequest::execute` [core] - 10https://gerrit.wikimedia.org/r/47712 (owner: 10Abhshkdz) [15:16:29] siebrand: ah that is true [15:18:02] (03Merged) 10jenkins-bot: Fix PHP CodeSniffer errors and warnings [core] - 10https://gerrit.wikimedia.org/r/88519 (owner: 10Siebrand) [15:18:03] is there a way to unbold that? [15:18:19] not that I know. [15:18:24] (I remember there are some reports upstream about bold but not exactly what) [15:18:32] Only when "you" are the last commenter. [15:18:47] I'd expect it to be "read on seen". [15:18:47] which however would bold it again for all others? :) [15:19:34] I sometimes remove myself as reviewer if a patch appears to be going nowheree. [15:19:43] No point in keeping it om my review list. [15:20:49] (03CR) 10Siebrand: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/64533 (owner: 10Balloonguy) [15:21:48] Yes, and you're also nice enough to tell them to that they're not clueless :) [15:21:59] Does this match what you see? https://code.google.com/p/gerrit/issues/detail?id=1936#c1 (comment 1) [15:22:08] (03CR) 10Siebrand: [C: 04-1] "Submitter, are you going to follow up on this? If this remains open for too long without follow-up, we will abandon it. It can always be r" [core] - 10https://gerrit.wikimedia.org/r/47373 (owner: 10Bouron) [15:23:08] (03CR) 10Siebrand: [C: 04-1] "Submitter, are you going to follow up on this? If this remains open for too long without follow-up, we will abandon it. It can always be r" [core] - 10https://gerrit.wikimedia.org/r/8784 (owner: 10Dan Collins) [15:23:42] (03CR) 10Siebrand: [C: 04-1] "Submitter, are you going to follow up on this? If this remains open for too long without follow-up, we will abandon it. It can always be r" [core] - 10https://gerrit.wikimedia.org/r/20013 (owner: 10Dash1291) [15:24:55] (03CR) 10Siebrand: "Niklas, your comments were addressed. Can you give this another look?" [core] - 10https://gerrit.wikimedia.org/r/84475 (owner: 10Ejegg) [15:27:42] Nemo_bis: Comes close I guess. [15:29:32] siebrand: should I "hijack" that bug or open a new one, to ask for the bold to be removed when the changeset is visited (and last patch has been already reviewed)? [15:29:52] Nemo_bis: I think you should report a new one. [15:29:55] (It seems a reasonable request, they already have the green checkmark for patch files.) [15:30:02] Nemo_bis: Different feature. [15:30:07] yeah, give me a moment [15:30:16] Krenair: ah thanks [15:31:15] lemme see what needed changing in Echo.php... [15:31:38] (03CR) 10Siebrand: [C: 032] Make items in the StringUtilsTest data provider easier to identify by adding names. [core] - 10https://gerrit.wikimedia.org/r/65300 (owner: 10JanZerebecki) [15:32:26] (03CR) 10Hashar: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88628 (owner: 10Dereckson) [15:33:15] (03CR) 10Siebrand: "What needs to be done to get this patch set going again?" [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [15:33:31] (03CR) 10Siebrand: [V: 04-1] "Can't be merged cleanly." [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [15:34:22] (03Merged) 10jenkins-bot: Make items in the StringUtilsTest data provider easier to identify by adding names. [core] - 10https://gerrit.wikimedia.org/r/65300 (owner: 10JanZerebecki) [15:34:40] (03CR) 10Siebrand: [C: 032] tests: remove date from @author docs statements [core] - 10https://gerrit.wikimedia.org/r/87323 (owner: 10Hashar) [15:36:52] siebrand: https://code.google.com/p/gerrit/issues/detail?id=2174 [15:37:14] (03PS3) 10Dereckson: Fixes stacktrace output when no file context is available. [core] - 10https://gerrit.wikimedia.org/r/88628 [15:37:41] (03CR) 10Dereckson: "PS3: fixing typo" [core] - 10https://gerrit.wikimedia.org/r/88628 (owner: 10Dereckson) [15:38:01] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [15:38:43] huh. [15:38:54] php has a array_change_key_case() function, but not array_map_keys() function [15:38:54] (03Merged) 10jenkins-bot: tests: remove date from @author docs statements [core] - 10https://gerrit.wikimedia.org/r/87323 (owner: 10Hashar) [15:40:15] also, the "User Contributed Notes" on http://pl1.php.net/manual/en/function.array-change-key-case.php are hilarious. [15:50:46] (03PS2) 10Hashar: Address PHP CodeSniffer errors and warnings [core] - 10https://gerrit.wikimedia.org/r/88466 (owner: 10Siebrand) [15:53:50] (03CR) 10Siebrand: [C: 04-1] "Submitter, are you going to follow up on this? If this remains open for too long without follow-up, we will abandon it. It can always be r" [core] - 10https://gerrit.wikimedia.org/r/63395 (owner: 10Jérémie Roquet) [15:54:20] (03CR) 10Siebrand: [C: 04-1] "Submitter, are you going to follow up on this? If this remains open for too long without follow-up, we will abandon it. It can always be r" [core] - 10https://gerrit.wikimedia.org/r/61390 (owner: 10Kai Nissen (WMDE)) [15:54:29] (03CR) 10Matthias Mullie: "(5 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [15:54:43] (03PS6) 10Matthias Mullie: Add AJAX header modification [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [15:55:07] (03CR) 10Cscott: "Test case?" [core] - 10https://gerrit.wikimedia.org/r/88750 (owner: 10Anomie) [15:55:27] if i wanted to file a wishlist bug for internationalized timestamp rendering, is cldr the proper extension? Basically if a revision was created and then hidden(user level moderation in flow) would like to say "Hidden after 7 minutes" but with cldr the two outputs are "7 minutes ago" and "In 7 minutes" [15:56:06] what's wrong with 'Hidden 7 minutes ago'? [15:56:07] ebernhardson: CLDR extension should only provide functionality that can be provided with the data from CLDR. [15:56:22] legoktm: nothing wrong with it, the problem is we cant get cldr to say that :) [15:56:35] siebrand: ahh, any better suggestions for where to file the wishlist? [15:56:46] ebernhardson: have you looked at getHumanTimestamp? [15:56:56] legoktm: yes, it outputs "7 minutes ago" and "In 7 minutes" [15:57:08] legoktm: (its overridden by cldr with hooks) [15:57:12] hm. [15:57:28] siebrand: or perhaps actually internationalizing that is a bit bigger of a problem then i'm thinking? [15:57:46] (03PS1) 10Matthias Mullie: Fix double-compression [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88756 [16:00:57] legoktm: oh, my bad i misinterpreted your response, the concern with "7 minutes ago" vs "After 7 minutes" is mostly just a ui thing that design had thought up, thinking it would be more usefull to know how long a post was visible rather than how long ago it was hidden [16:01:35] oh [16:01:41] (03PS3) 10Matthias Mullie: Various small design fixes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88725 [16:01:48] from a user point of view, thats not what i'd like other users to see [16:01:55] if i hid something for privacy reasons [16:02:05] i dont want other users to know it was visible for 2 hours [16:02:27] legoktm: in flow hiding is something anyone can do, its user-level moderation. The concern i think is trolls going arround hiding old posts [16:02:36] most moderation should occur fairly soon, within a day or two i imagine [16:02:46] (03CR) 10Matthias Mullie: "Just fixed my remarks myself" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [16:02:49] oh. so hiding != suppression? [16:03:00] who can unhide? [16:03:03] no, hiding is autoconfirmed, or if the wiki wants literaly anyone [16:03:24] anyone can unhide [16:03:54] (03CR) 10Matthias Mullie: [C: 032] Add mw-ui-input class to VisualEditor surface. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88687 (owner: 10Werdna) [16:04:03] (03Merged) 10jenkins-bot: Add mw-ui-input class to VisualEditor surface. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88687 (owner: 10Werdna) [16:04:15] ah [16:04:15] there are (currently, not final) 3 levels of moderation in flow, the different levels just signify who can see/restore the content. Hide is the base level done by any user, Delete is sysop, and Suppress is oversighters. [16:04:35] but they all do the same thing, the only difference is you have to have the right permissions to see/restore content [16:04:36] gotcha [16:04:39] makes sense [16:06:29] (03CR) 10Matthias Mullie: [C: 032] "That's one hideously large spinner :)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88684 (owner: 10Werdna) [16:06:39] (03Merged) 10jenkins-bot: Fix some VisualEditor annoyances: [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88684 (owner: 10Werdna) [16:09:46] (03CR) 10Hashar: [C: 04-1] "(8 comments)" [core] - 10https://gerrit.wikimedia.org/r/88466 (owner: 10Siebrand) [16:11:22] (03PS2) 10Siebrand: Remove unused local variable [core] - 10https://gerrit.wikimedia.org/r/88496 [16:12:24] (03CR) 10Hashar: [C: 032] "Got rid of the dependency on another change, so basically cherry-picked it on top of master." [core] - 10https://gerrit.wikimedia.org/r/88496 (owner: 10Siebrand) [16:12:39] (03CR) 10Chad: [C: 032] Remove unused local variable [core] - 10https://gerrit.wikimedia.org/r/88496 (owner: 10Siebrand) [16:14:42] (03CR) 10Hashar: [C: 04-1] "Ideally we would to use HTML templates and feed them with the content but we dont have support for that." [core] - 10https://gerrit.wikimedia.org/r/88500 (owner: 10Siebrand) [16:16:22] (03PS2) 10Siebrand: Remove incorrect span in translation [core] - 10https://gerrit.wikimedia.org/r/88506 [16:16:23] (03Merged) 10jenkins-bot: Remove unused local variable [core] - 10https://gerrit.wikimedia.org/r/88496 (owner: 10Siebrand) [16:17:38] (03CR) 10Hashar: [C: 032] Remove incorrect span in translation [core] - 10https://gerrit.wikimedia.org/r/88506 (owner: 10Siebrand) [16:21:20] (03Merged) 10jenkins-bot: Remove incorrect span in translation [core] - 10https://gerrit.wikimedia.org/r/88506 (owner: 10Siebrand) [16:22:34] (03CR) 10Nikerabbit: "+0 -0?" [core] - 10https://gerrit.wikimedia.org/r/88506 (owner: 10Siebrand) [16:23:02] (03PS2) 10Siebrand: Remove deprecated HTML tag [core] - 10https://gerrit.wikimedia.org/r/88501 [16:23:04] (03CR) 10Chad: "(1 comment)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/88729 (owner: 10Manybubbles) [16:24:33] (03CR) 10Siebrand: "I think this has been exported from twn before it was fixed here." [core] - 10https://gerrit.wikimedia.org/r/88506 (owner: 10Siebrand) [16:25:20] (03Abandoned) 10Hashar: Remove deprecated HTML tag [core] - 10https://gerrit.wikimedia.org/r/88501 (owner: 10Siebrand) [16:25:59] siebrand: reviewed some of your style fixes [16:26:53] hashar: Thanks a lot. I'm replying. [16:26:56] legoktm, so what happened to Echo.php? [16:27:03] siebrand: would follow up tomorrow [16:27:06] class happened [16:27:10] i'm fixing it up right now [16:27:24] siebrand: trivia: I used objective CAML as an example [16:28:02] (03PS3) 10Siebrand: Address PHP CodeSniffer errors and warnings [core] - 10https://gerrit.wikimedia.org/r/88466 [16:28:42] (03PS3) 10Hashar: Declare visibility on class properties [core] - 10https://gerrit.wikimedia.org/r/87510 (owner: 10Siebrand) [16:28:59] (03CR) 10Hashar: "removed link to jenkins build, copy pasted phpcs warning instead." [core] - 10https://gerrit.wikimedia.org/r/87510 (owner: 10Siebrand) [16:29:06] (03PS4) 10Hashar: Declare visibility on class properties [core] - 10https://gerrit.wikimedia.org/r/87510 (owner: 10Siebrand) [16:29:11] hashar: I don't understand what you want to turn this into: [16:29:17] return $this->db->tableExists( 'cur', __METHOD__ ) || $this->db->tableExists( 'revision', __METHOD__ ); [16:29:24] sorry [16:29:36] return $this->db->tableExists( 'cur', __METHOD__ ) [16:29:37] || $this->db->tableExists( 'revision', __METHOD__ ); [16:29:42] Ah, that. [16:29:50] this way when you read the return, you instantly see there is another condition below [16:30:07] we could even enclose it in () but that is overkill hehe [16:30:11] There's a check that will fail when that's done with &/&& at least. [16:30:22] return ( [16:30:22] So that would require a change in a CS rule. [16:30:23] condition1 [16:30:24] || condition2 [16:30:24] ); // end of return [16:30:38] aohh :( [16:30:44] so yeah we might want to adjust our CS rule [16:30:53] or fix upstream / write a new sniff in upstream [16:31:35] (03PS11) 10Chad: Add support for mysqli extension [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [16:31:39] (03CR) 10Hashar: [C: 031] Declare visibility on class properties [core] - 10https://gerrit.wikimedia.org/r/87510 (owner: 10Siebrand) [16:31:49] (03PS4) 10Legoktm: Support global group change notifications [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 [16:31:51] Krenair: ^ [16:32:39] greg-g: I added you to https://bugzilla.wikimedia.org/show_bug.cgi?id=54886 , not sure who else should look at that [16:34:42] (03CR) 10Alex Monk: "Just needs rebase and review AFAIK" [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [16:37:22] Krinkle: do you know why git review would say that https://gerrit.wikimedia.org/r/#/c/88652/ is an 'outstanding commit' when I do git review but when I try to amend that branch git tells me "(change 88652 closed)" [16:37:59] I am off [16:38:02] see you tomorrow [16:38:34] (03CR) 10Siebrand: "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/88466 (owner: 10Siebrand) [16:38:55] (03PS9) 10Alex Monk: Add auto-save feature for edit textareas, via localStorage [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [16:39:15] (03PS4) 10Siebrand: Address PHP CodeSniffer errors and warnings [core] - 10https://gerrit.wikimedia.org/r/88466 [16:40:19] (03CR) 10Siebrand: "PHPStorm warned on these two tags, because there were no opening tags. That's the primary reason for this change." [core] - 10https://gerrit.wikimedia.org/r/88500 (owner: 10Siebrand) [16:41:28] siebrand, have you tested that commit (the one you just commented on)? [16:41:57] Krenair: 88500? [16:42:01] yeah [16:42:06] Because it looks like it's only going to generate the text, not do anything with it... [16:42:40] Krenair: oops. [16:42:42] Shouldn't they also be echoed? [16:42:48] Krenair: That's useful input :) [16:42:50] yeah, exactly [16:43:08] Krinkle: flagging https://gerrit.wikimedia.org/r/#/c/88244/ & https://gerrit.wikimedia.org/r/#/c/88408/ , for when you have a chance to look [16:45:05] (03CR) 10Siebrand: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88513 (owner: 10Siebrand) [16:45:30] (03PS3) 10Siebrand: Fix PHP CodeSniffer warnings and errors [core] - 10https://gerrit.wikimedia.org/r/88513 [16:47:16] (03PS2) 10Siebrand: Add close elements through code [core] - 10https://gerrit.wikimedia.org/r/88500 [16:48:42] (03CR) 10Siebrand: "PS2 is a rebase to get rid of the dependency." [core] - 10https://gerrit.wikimedia.org/r/88500 (owner: 10Siebrand) [16:49:47] I'd like to backport a patch set to wmf20. What do I put in the "Cherry Pick to Branch" dialog in gerrit? [16:49:50] wmf/1.22wmf20 ? [16:51:11] yes [16:51:30] (03PS1) 10BryanDavis: Add script to purge changed pages from cache [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 [16:52:18] AaronSchulz: Thanks. And also ^ [16:52:27] (03PS3) 10Siebrand: Add close elements through code [core] - 10https://gerrit.wikimedia.org/r/88500 [16:52:36] (03CR) 10Jdlrobson: "Essentially it comes down to file paths and shared resources. If the resource was packaged in an extension its impossible for it to access" [core] - 10https://gerrit.wikimedia.org/r/79948 (owner: 10Jdlrobson) [16:53:12] Krenair: Fix. Thanks for spotting that. [16:53:18] (03PS37) 10Jdlrobson: Vector: New beta module with new typography styles [core] - 10https://gerrit.wikimedia.org/r/79948 [16:54:10] MaxSem: these keep coming up : http://www.mediawiki.org/w/index.php?title=Extension_talk:MobileFrontend&offset=20131009144612&lqt_mustshow=34183#Html_Formatter_Not_Found_34183 is there any way to tell users when they have installed an incompatible version? [16:54:23] (03CR) 10BryanDavis: "Backporting new maintenance script to 1.22wmf20 to make running it next week in production easier." [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [16:58:25] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/5130 (owner: 10Joancreus) [16:59:31] chrismcmahon: You probably have an outdated git clone. [16:59:54] Krinkle: thanks, got it resolved, but I'd not seen that before [17:00:00] chrismcmahon: When you do "git review", git-review compares your current head with the local copy of origin/master, and the differrence is what you're sending to gerrit [17:00:17] (03CR) 10Siebrand: [C: 031] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88412 (owner: 10Werdna) [17:01:06] doing git remote update, git rebase origin/master should fix that in the future. Be sure to verify `git log` afterwards and confirm you're not re-sending an old version of a patch you depended on. [17:01:23] chrismcmahon: in your case you likely branched a new commit off of 'iframe' before it landed, so it was a dependency [17:01:40] (03CR) 10Kipcool: [C: 032 V: 032] found \r bug in formatting [extensions/WikiLexicalData] - 10https://gerrit.wikimedia.org/r/87999 (owner: 10Hiong3-eng5) [17:02:28] (03CR) 10Tpt: [C: 032] "Minor change that fix a fatal error in some pages (merge now in order to be deployed with 1.22wmf21)" [extensions/ProofreadPage] - 10https://gerrit.wikimedia.org/r/88217 (owner: 10Tpt) [17:02:41] (03Merged) 10jenkins-bot: (bug 55439) Fix an error when an image has an unknown media type [extensions/ProofreadPage] - 10https://gerrit.wikimedia.org/r/88217 (owner: 10Tpt) [17:05:04] (03CR) 10Siebrand: [C: 04-1] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88407 (owner: 10Werdna) [17:09:02] (03CR) 10Siebrand: [C: 04-1] "(1 comment)" [extensions/VectorBeta] - 10https://gerrit.wikimedia.org/r/88719 (owner: 10Raimond Spekking) [17:09:18] (03CR) 10Aaron Schulz: "All wikis will be on wmf22 by Thursday, October 10, 2013" [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [17:10:24] (03CR) 10Siebrand: [C: 04-1 V: 04-1] Refactor ipboptions and Xml::listDropDown() [core] - 10https://gerrit.wikimedia.org/r/22699 (owner: 10Liangent) [17:11:34] (03CR) 10Anomie: "Looking at this a year later, I wonder if its use case is covered by OAuth or OpenID." [core] - 10https://gerrit.wikimedia.org/r/20013 (owner: 10Dash1291) [17:12:53] (03CR) 10BryanDavis: "RobLa wants me to run it "next week". He thought backporting would be a good idea for a couple of reasons:" [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [17:13:10] (03PS1) 10Aaron Schulz: Fixed trx isolation bug that could cause text not to be found [core] - 10https://gerrit.wikimedia.org/r/88761 [17:13:20] (03CR) 10BryanDavis: "See also: https://www.mediawiki.org/wiki/User:BDavis_(WMF)/Notes/Finding_Files_To_Purge#Production_purge" [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [17:13:43] bd808: nice docs! [17:14:19] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed. No opinion on the change itself." [core] - 10https://gerrit.wikimedia.org/r/84194 (owner: 10TTO) [17:14:42] * bd808 blushes [17:14:54] * siebrand grins. [17:14:55] (03PS2) 10Aaron Schulz: Fixed trx isolation bug that could cause text not to be found [core] - 10https://gerrit.wikimedia.org/r/88761 [17:15:04] (03CR) 10Aaron Schulz: "Fixed bug #" [core] - 10https://gerrit.wikimedia.org/r/88761 (owner: 10Aaron Schulz) [17:15:04] ori-l: thanks. I forget things if I don't make plans. [17:16:10] bd808: how do you know? :P [17:16:22] (03CR) 10Siebrand: [C: 04-1] Add logo to feature preference, fix copy [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/87276 (owner: 10MarkTraceur) [17:19:49] (03CR) 10Siebrand: "Can we merge this again? It's really hard to see the forest from the trees." [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/82619 (owner: 10Dan-nl) [17:21:02] (03CR) 10Siebrand: [C: 031] addressing mark traceur’s comments [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/82619 (owner: 10Dan-nl) [17:21:44] (03CR) 10Siebrand: [C: 031] "L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [17:22:21] (03CR) 10Dash1291: "Sorry for not being able to follow up on this. Totally slipped out of my mind, so thanks for bringing it up." [core] - 10https://gerrit.wikimedia.org/r/20013 (owner: 10Dash1291) [17:22:33] (03PS6) 10Siebrand: (extension/PGFTikZ) PGFTikZ extension: initial import [extensions/PGFTikZ] - 10https://gerrit.wikimedia.org/r/38487 (owner: 10thibaultmarin) [17:23:56] (03CR) 10Siebrand: [C: 031] "i18n/L10n is fine." [extensions/PGFTikZ] - 10https://gerrit.wikimedia.org/r/38487 (owner: 10thibaultmarin) [17:26:57] brion: Added you to the next MultimediaViewer patch, in case you've some time today :) [17:30:14] (03CR) 10Parent5446: [C: 04-1] "(8 comments)" [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [17:31:18] whee [17:31:31] (03PS1) 10Chad: Remove wfDl() and cleanup everything it touched [core] - 10https://gerrit.wikimedia.org/r/88764 [17:31:31] (03PS1) 10Chad: Remove deprecated parameter from MimeMagic::guessMimeType() [core] - 10https://gerrit.wikimedia.org/r/88765 [17:33:23] (03PS1) 10Anomie: Fix ForeignAPIRepo::fileExistsBatch() [core] - 10https://gerrit.wikimedia.org/r/88766 [17:33:39] (03CR) 10jenkins-bot: [V: 04-1] Remove deprecated parameter from MimeMagic::guessMimeType() [core] - 10https://gerrit.wikimedia.org/r/88765 (owner: 10Chad) [17:36:09] (03CR) 10Parent5446: "@Matt - I think I might know why it's not working. Have you tried testing as an anonymous user? Because for some stupid reason the ping li" [core] - 10https://gerrit.wikimedia.org/r/65867 (owner: 10Parent5446) [17:36:20] (03PS1) 10Ebrahim: Native digits on pref [core] - 10https://gerrit.wikimedia.org/r/88767 [17:36:23] (03PS2) 10Aaron Schulz: Deal with PHP bug 63411 and tweaked sleep waits [extensions/SwiftCloudFiles] - 10https://gerrit.wikimedia.org/r/88406 [17:37:41] (03CR) 10Aaron Schulz: "Passes unit tests against ceph" [extensions/SwiftCloudFiles] - 10https://gerrit.wikimedia.org/r/88406 (owner: 10Aaron Schulz) [17:41:36] (03Abandoned) 10Chad: Remove deprecated parameter from MimeMagic::guessMimeType() [core] - 10https://gerrit.wikimedia.org/r/88765 (owner: 10Chad) [17:42:38] is there a way to make mediawiki's tests run just one test? [17:43:28] you can make it run just one file pretty easily [17:43:42] which can come in handy. I don't know about just a single test [17:44:24] I'll take just one file [17:44:53] (03CR) 10Krinkle: [C: 04-1] "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [17:45:32] php phpunit.php /path/to/file [17:46:27] ah. surprisingly easy. thanks [17:47:14] (03CR) 10CSteipp: [C: 031] "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 (owner: 10Legoktm) [17:48:20] (03CR) 10Aaron Schulz: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88764 (owner: 10Chad) [17:48:28] (03PS2) 10Anomie: Handle raw when calculating $rawtoc [core] - 10https://gerrit.wikimedia.org/r/88750 [17:49:06] (03CR) 10Anomie: "PS2: Add a test case, as requested" [core] - 10https://gerrit.wikimedia.org/r/88750 (owner: 10Anomie) [17:49:58] (03PS5) 10Legoktm: Support global group change notifications [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 [17:50:06] csteipp: thanks for the review [17:50:15] (03CR) 10Legoktm: "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/88683 (owner: 10Legoktm) [17:50:20] manybubbles: Try the --filter option [17:50:24] (03CR) 10Ebrahim: "This patch is for fixing semilocalized messages on http://imgur.com/XuXD0A3" [core] - 10https://gerrit.wikimedia.org/r/88767 (owner: 10Ebrahim) [17:50:42] (03PS2) 10GWicke: Fix SVG glyphs and make the service more robust [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 [17:50:53] (03PS4) 10Ori.livneh: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 [17:51:01] ^ Krinkle [17:51:07] (03CR) 10Chad: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88764 (owner: 10Chad) [17:51:08] legoktm: Ping re https://gerrit.wikimedia.org/r/#/c/83219/ [17:51:26] looking [17:51:48] (03PS2) 10Chad: Remove wfDl() and cleanup everything it touched [core] - 10https://gerrit.wikimedia.org/r/88764 [17:51:51] (03CR) 10Hashar: [C: 032] "\O/" [core] - 10https://gerrit.wikimedia.org/r/88513 (owner: 10Siebrand) [17:51:56] ori-l: rebase/change :) [17:52:09] Krinkle: ajhsdakjhsdhalsjdh yes sorry, I need to unlearn that habit. [17:52:25] (03CR) 10GWicke: "It seems that setTimeout was only added recently." [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 (owner: 10GWicke) [17:52:34] Krinkle: i fixed the space and switched to $.map / $.each [17:53:02] ori-l: one should remain within (function I think, and need a check for console as well (or use mw.log) [17:53:10] space* [17:53:31] but mw.log is debug=true only, so perhaps do a check for log and use it directly [17:53:36] (03CR) 10Aaron Schulz: [C: 032] Remove wfDl() and cleanup everything it touched [core] - 10https://gerrit.wikimedia.org/r/88764 (owner: 10Chad) [17:53:38] (03CR) 10Hashar: "Sorry I should have been more careful :/" [core] - 10https://gerrit.wikimedia.org/r/88506 (owner: 10Siebrand) [17:53:50] ori-l: ah, missed the check [17:54:03] window.console is undefined, which is then a local variable that's falsey [17:54:07] OK :) [17:54:51] Krinkle: i don't mind changing it if you find an alternative more idiomatic [17:55:19] ori-l: yeah, var console = window.console; seems easier to understand (passing around through a closure is a bit hard to see) [17:55:40] Krinkle: k, i'll update [17:55:46] (03CR) 10Hashar: [C: 04-1] "Are you planning to add any tests or would you be more comfortable in having someone else writing them? (just wondering)." [core] - 10https://gerrit.wikimedia.org/r/88628 (owner: 10Dereckson) [17:56:15] ori-l: lastly, the underscore, maybe use 'i' or 'key' there? I'd expect jshint to throw on that but it seems it doesn't. [17:56:25] Krinkle: even jquery does that :P [17:57:20] (03PS3) 10Legoktm: Tests for API section editing [core] - 10https://gerrit.wikimedia.org/r/83219 [17:57:28] ori-l: I don't oppose the underscore from a logical POV. I trust jQuery to know better if it was problematic for any reason at all. [17:57:28] anomie: ^ [17:57:35] Just naming convention. [17:58:18] bleh, will rebase [17:58:51] (03Merged) 10jenkins-bot: Fix PHP CodeSniffer warnings and errors [core] - 10https://gerrit.wikimedia.org/r/88513 (owner: 10Siebrand) [17:59:57] (03CR) 10jenkins-bot: [V: 04-1] Tests for API section editing [core] - 10https://gerrit.wikimedia.org/r/83219 (owner: 10Legoktm) [18:00:46] (03Merged) 10jenkins-bot: Remove wfDl() and cleanup everything it touched [core] - 10https://gerrit.wikimedia.org/r/88764 (owner: 10Chad) [18:01:18] (03CR) 10Legoktm: "Not true, https://gerrit.wikimedia.org/r/#/c/83219/ had been open for a while." [core] - 10https://gerrit.wikimedia.org/r/87612 (owner: 10Chad) [18:03:04] (03PS5) 10Ori.livneh: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 [18:03:05] (03CR) 10Anomie: "Also, I note that the test removed in ApiWatchTest had two asserts before the markIncomplete." [core] - 10https://gerrit.wikimedia.org/r/87612 (owner: 10Chad) [18:03:31] (03PS4) 10Legoktm: Tests for API section editing [core] - 10https://gerrit.wikimedia.org/r/83219 [18:03:58] Krinkle: updated [18:04:08] anomie: ok, rebased properly ^ [18:04:37] (03CR) 10Anomie: [C: 032] Tests for API section editing [core] - 10https://gerrit.wikimedia.org/r/83219 (owner: 10Legoktm) [18:05:14] :D [18:06:24] (03CR) 10GWicke: [C: 032] Deal with PHP bug 63411 and tweaked sleep waits [extensions/SwiftCloudFiles] - 10https://gerrit.wikimedia.org/r/88406 (owner: 10Aaron Schulz) [18:09:04] ori-l: I predict jshint might throw on this one though, it's an edge case. [18:09:12] I'll hang tight for a few seconds. [18:10:04] (03CR) 10Aaron Schulz: [C: 032] "Nevermind, I was mixing wmf 19 and 20" [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [18:10:32] (03CR) 10jenkins-bot: [V: 04-1] Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [18:10:57] blargh. [18:14:10] (03Merged) 10jenkins-bot: Tests for API section editing [core] - 10https://gerrit.wikimedia.org/r/83219 (owner: 10Legoktm) [18:14:17] it isn't listed as an accepted global (shouldn't be) so though after checking window.console it is safe to assume you will get at that console if there are no other variables by that name anywhere in this or a parent scope, it's a semantic difference jshint doesnt' like being implicit. [18:15:55] yeah, i'm just whining [18:15:59] (03Merged) 10jenkins-bot: Add script to purge changed pages from cache [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88757 (owner: 10BryanDavis) [18:16:04] (03PS6) 10Ori.livneh: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 [18:16:32] ori-l: I had a similar back/forth with this one https://github.com/wikimedia/mediawiki-core/blob/master/resources/mediawiki/mediawiki.js#L783-L795 [18:16:42] the console.error one [18:17:52] I think Tim or Roan originally had console.[ console.error ? 'error' : 'log' ]( e ); [18:18:03] hah [18:18:36] console.log | console.log doesn't work because it needs the context bound [18:18:42] error || [18:22:36] (03PS3) 10GWicke: Fix SVG glyphs and make the service more robust [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 [18:23:36] (03CR) 10Aaron Schulz: [V: 032] Deal with PHP bug 63411 and tweaked sleep waits [extensions/SwiftCloudFiles] - 10https://gerrit.wikimedia.org/r/88406 (owner: 10Aaron Schulz) [18:29:00] ori-l: hm.. something seems to be off. then numbers are too low. Some are 0 even. http://cl.ly/image/3g292h2D0c1A [18:29:27] might be the rounding after / 1024 I guess [18:29:34] (03CR) 10Aaron Schulz: Added supported for retrieving file metadata/headers [core] - 10https://gerrit.wikimedia.org/r/86642 (owner: 10Aaron Schulz) [18:30:33] (03PS5) 10Aaron Schulz: Added supported for retrieving file metadata/headers [core] - 10https://gerrit.wikimedia.org/r/86642 [18:33:10] (03PS1) 10Umherirrender: Add wgRelevantUserName from skin as js variable [core] - 10https://gerrit.wikimedia.org/r/88773 [18:33:16] Krinkle: yeah [18:33:56] ori-l: Hm.. I changed it locally to add a displaySize column (and made size the raw byte size) [18:33:57] still 0 [18:34:32] ah, no, there are still some 0s left, but different ones [18:34:37] jquery, user, site,mediawiki [18:34:42] makes sense [18:36:24] ori-l: size === 0 ? '?' : size > 1024 ? Math.round( size / 1024 ) + 'KB' : size + 'B'; [18:36:46] size / displaySize. [18:36:53] I don't like the extra column though. meh. [18:37:21] yeah, and if you append the unit size it becomes a string and console.table quotes it [18:37:25] omg, chrome can't draw a simple table. [18:37:35] yes, it's a pretty crappy API, no? [18:37:45] I mean, the blue/white alternating [18:37:51] it's completely broken [18:39:09] after a row with 2 lines, the highgliting is off [18:39:09] http://cl.ly/image/1v3j1m1I3Z0h [18:39:11] xD [18:40:33] (03CR) 10MarkTraceur: "(12 comments)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [18:40:42] heh [18:49:06] (03PS1) 10Umherirrender: Set relevant User on Special:Unblock [core] - 10https://gerrit.wikimedia.org/r/88785 [18:55:33] (03PS1) 10Manybubbles: Expose changed links in LinksUpdate. [core] - 10https://gerrit.wikimedia.org/r/88789 [18:56:18] hi qgil, hexmode, who is starting the hangout? [18:59:53] (03PS4) 10Tpt: Major refactoring of the JS part of the page edition system [extensions/ProofreadPage] (pagePagesRefactoring) - 10https://gerrit.wikimedia.org/r/87527 [19:03:40] mglaser: I think qgil needs to, I can't start a public one [19:03:45] mglaser: can you? [19:04:02] hexmode: qgil will invite you [19:04:28] (03CR) 10Tpt: "PS4: Fix the issue in labs" [extensions/ProofreadPage] (pagePagesRefactoring) - 10https://gerrit.wikimedia.org/r/87527 (owner: 10Tpt) [19:05:00] mglaser: ok.... wonder where he is. [19:08:02] (03CR) 10Aaron Schulz: [C: 032] Fix ForeignAPIRepo::fileExistsBatch() [core] - 10https://gerrit.wikimedia.org/r/88766 (owner: 10Anomie) [19:08:12] (03PS1) 10Legoktm: Avoid parser cache fragmentation by showing error messages in content language [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/88791 [19:08:23] (03PS1) 10Krinkle: deleteEqualMessages: Exclude messages that are empty by default [core] - 10https://gerrit.wikimedia.org/r/88792 [19:09:05] (03PS2) 10Legoktm: Avoid parser cache fragmentation by showing error messages in content language [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/88791 [19:10:17] https://www.mediawiki.org/wiki/Meetings/2013-10-09 has just started! [19:10:43] qgil: youtube stream looks good [19:10:44] You can follow at http://www.youtube.com/watch?v=gf0_cL--nSM&feature=youtu.be [19:10:49] thanks bd808 [19:11:14] If you want to be invited to the hangout just tell me (I need your email address) [19:11:39] If you just want to ask questions here, I will pass them along [19:11:46] (03Merged) 10jenkins-bot: Fix ForeignAPIRepo::fileExistsBatch() [core] - 10https://gerrit.wikimedia.org/r/88766 (owner: 10Anomie) [19:12:07] (03CR) 10Brian Wolff: "Technically this isn't avoiding parser cache fragmentation, since it wasn't splitting the parser cache in the first place." [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/88791 (owner: 10Legoktm) [19:12:34] bawolff: err, so what's the right term for that then? [19:12:50] duh: Good question [19:13:07] Cache pollution I think, not 100% sure [19:13:23] (03PS3) 10Legoktm: Avoid parser cache fragmentation by showing error messages in content language [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/88791 [19:13:37] "Avoid cache issues by showing error messages in content language" <-- thats what i set it to [19:13:45] that sounds good [19:13:52] mglaser is talking about features described at https://www.mediawiki.org/wiki/MediaWiki_1.22 [19:14:13] in any case, I'm probably just being annoying and pedantic ;) [19:14:24] :P [19:14:41] qgil: I can't see the stream, it says an error occured [19:15:13] (03PS1) 10Reviewer-bot: Something Something UTF-8 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88793 [19:15:15] qgil: nevermind [19:15:22] bawolff, :) [19:15:23] It likes chrome better than firefox for me [19:15:52] works on my Firefox... [19:16:36] * bawolff is using firefox 3.5, which is probably the problem [19:16:51] hexmode is referring to this page: https://www.mediawiki.org/wiki/Project:Release_management/Release_timeline [19:18:04] Any questions so far? [19:19:49] qgil: youtube stream works great for me [19:19:57] (03Abandoned) 10Reviewer-bot: Test commit, version 23 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88688 (owner: 10Reviewer-bot) [19:20:00] (03Abandoned) 10Reviewer-bot: Test commit #23 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88749 (owner: 10Reviewer-bot) [19:20:04] (03Abandoned) 10Reviewer-bot: Something Something UTF-8 [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/88793 (owner: 10Reviewer-bot) [19:25:51] (03PS1) 10TK-999: Cleanup [extensions/Tabber] - 10https://gerrit.wikimedia.org/r/88798 [19:25:53] (03PS1) 10TK-999: Refactor front-end logic [extensions/Tabber] - 10https://gerrit.wikimedia.org/r/88799 [19:25:53] (03PS1) 10TK-999: Caps [extensions/Tabber] - 10https://gerrit.wikimedia.org/r/88800 [19:30:22] (03CR) 10Hashar: [C: 04-1] "writing tests for it." [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [19:30:33] Any questions about https://www.mediawiki.org/wiki/Meetings/2013-10-09 ? [19:31:03] hashar: You're writing tests for it, or you want me to write tests? [19:31:08] hexmode: Wooo! [19:31:11] bd808: I am [19:31:15] agree +a billion [19:31:15] bd808: like right now :] [19:31:18] hashar: Awesome [19:31:35] I thought about a test and then got lazy :P [19:31:38] I am just p*** off by php but writing then nonetheless :D [19:31:39] like [19:31:51] given an abstract class, how would you call a public (non abstract) method of that class ? [19:32:15] * bawolff wonders if mark just proposed making interwiki transclusions... :) [19:32:15] I think you have to make a concrete stub class [19:32:30] yeah [19:32:33] copy pasting right now [19:35:34] questions for hexmode mglaser ? [19:36:14] qgil: Are there tasks/bugs that they need help with for the automation effort? [19:36:18] (03CR) 10Physikerwelt: [C: 04-1] "now the program produces reasonable error messages and fails kindly." [extensions/Math] - 10https://gerrit.wikimedia.org/r/88678 (owner: 10GWicke) [19:36:50] bd808: probably, ask hashar what he can't get to :) [19:36:54] * bd808 is not volunteering, but ... [19:36:58] oh [19:36:59] well then [19:39:52] (03CR) 10jenkins-bot: [V: 04-1] Localisation updates from http://translatewiki.net. [extensions/Thanks] - 10https://gerrit.wikimedia.org/r/88855 (owner: 10L10n-bot) [19:41:28] (03PS2) 10Manybubbles: Allow reindex count to differ from old count [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/88729 [19:43:08] qgil, hexmode, mglaser: thanks for the overview. It's interesting stuff for a n00b like me. [19:43:39] (03CR) 10Brian Wolff: "(7 comments)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [19:44:00] bd808 thanks for joining the talk. I hope you'll be with us during the next few weeks :) [19:44:40] bd808: :) [19:44:56] (03Merged) 10jenkins-bot: Localisation updates from http://translatewiki.net. [extensions/Thanks] - 10https://gerrit.wikimedia.org/r/88855 (owner: 10L10n-bot) [19:44:56] mglaser: I'd be glad to be added to code reviews on your tests. [19:46:29] * hashar whistles out of despair [19:46:44] why haven't Magnus chosen python/postgre ? :( [19:49:03] hashar: Did you get the abstract issue sorted out? [19:49:22] You could try getMockForAbstractClass() if not [19:49:57] See http://phpunit.de/manual/3.7/en/test-doubles.html for sketchy details [19:50:07] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/ProofreadPage] (pagePagesRefactoring) - 10https://gerrit.wikimedia.org/r/87527 (owner: 10Tpt) [19:57:03] bd808: there still ? [19:57:19] hashar: pong [20:03:05] bd808: test done :-] [20:04:24] (03PS2) 10Hashar: Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:07:22] (03CR) 10jenkins-bot: [V: 04-1] Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:07:38] (03CR) 10Manybubbles: "(1 comment)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/88729 (owner: 10Manybubbles) [20:07:59] (03PS3) 10Manybubbles: Allow reindex count to differ from old count [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/88729 [20:09:22] (03CR) 10Hashar: "added basic unit test infrastructure, that might deserve a commit summary update." [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:10:35] hashar: I'll start making the test sane. Thanks for the bootstrap [20:11:15] (03CR) 10Hashar: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:11:21] bd808: I found out that people like writing tests [20:11:40] bd808: but usually dislike writing the base stuff :-] you should be fine now [20:11:47] i have added a couple comments meanwhile [20:12:10] also note that ideally the tests should be in a different change that would cover the function as it is in master [20:12:24] and your patch should pass without having to alter the tests (aka no regression) [20:14:00] hashar: So in a perfect world you would commit broken tests and then a separate patch to fix them? [20:14:24] Wouldn't you get a -1 from jenkins for the bad tests? [20:15:01] test for tests [20:15:12] there's a yo dawg here [20:16:45] (03CR) 10Ebrahim: "From http://en.wikipedia.org/wiki/Hebrew_punctuation#Period.2C_question_mark.2C_exclamation_mark.2C_comma" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/87654 (owner: 10Ebrahim) [20:20:16] (03CR) 10BryanDavis: [C: 04-1] "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:20:36] bd808: you did a refactoring of something that had no test coverage [20:20:45] bd808: the idea is that you will first cover the existing feature and writing tests [20:20:59] bd808: then do the refactoring in a follow up. If you did it well, the tests should still pass [20:21:24] bd808: so really instead of amending, I should have send another patchset and rebased your patch on top of my tests. [20:21:44] bd808: so this way you know the tests were passing and that your refactoring is matching whatever behavior we had before. [20:22:12] hashar: Ah. Sure. Except it's not really a refactor. It's a bug fix. Current behavior is wrong. [20:22:24] refactor/bug fix. Same thing :-] [20:22:35] fixing a bug has the potential to introduce future bugs! [20:23:12] Agreed. And I should have written a test in the first place. [20:24:46] Reedy: do you have some time to review a few MassMessage patches? [20:28:38] (03CR) 10Siebrand: "(2 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/81017 (owner: 10EBernhardson (WMF)) [20:29:22] bd808: we are missing a lot of basic tests and have a terrible coverage. So you have good excuses. [20:29:52] hashar: I have excuses. Their quality is debatable. :) [20:31:29] the quality of the excuses or the quality of the tests? [20:31:32] :P [20:32:00] Krinkle: does https://gerrit.wikimedia.org/r/#/c/88408/ look ok now? [20:33:25] (03CR) 10Hashar: "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [20:33:45] what I like with tests is that it is a good to delegate -2 to Jenkins-bot :-] [20:33:55] I tend to dislike voting CR-2 [21:04:39] ori-l: yeah, though it'd be nice if it somehow made clear that it is KB, not bytes (the ones being 0 looks a bit odd). Maybe you can address that in some way? [21:05:12] the ones that dont have scrtips/styles in mw.loader being 0 makes sense, but the ones that are just < 1024 [21:07:00] (03PS4) 10Plavi: Created a test for ProveIt gadget [qa/browsertests] - 10https://gerrit.wikimedia.org/r/87436 [21:09:03] Krinkle: ok, will do [21:13:37] (03PS10) 10Bartosz Dziewoński: SpecialRecentChanges: Rewrite the options form [core] - 10https://gerrit.wikimedia.org/r/64615 [21:13:59] (03CR) 10Plavi: "Patch set 4: runs quicker (1 scenario less, but the same functionality is checked), should run on test2 env if ProveIt gadget was installe" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/87436 (owner: 10Plavi) [21:14:07] (03PD1) 10Wikinaut: adding the forgotten table row to the table schema sql [extensions/AJAXPoll] - 10https://gerrit.wikimedia.org/r/88880 [21:14:08] (03CR) 10Bartosz Dziewoński: [C: 04-1] "Only fixed merge conflicts to keep this mostly up-to-date." [core] - 10https://gerrit.wikimedia.org/r/64615 (owner: 10Bartosz Dziewoński) [21:14:36] (03CR) 10Wikinaut: [C: 032 V: 032] adding the forgotten table row to the table schema sql [extensions/AJAXPoll] - 10https://gerrit.wikimedia.org/r/88880 (owner: 10Wikinaut) [21:15:00] (03PS1) 10EBernhardson (WMF): Add bezels to .mw-ui-button classes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88881 [21:27:10] csteipp: the background on this is at https://www.mediawiki.org/wiki/User:GWicke/Notes/Storage [21:27:13] (03PS7) 10Ori.livneh: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 [21:27:20] Krinkle: ^ [21:28:24] csteipp, this is the JWT spec I was looking at: http://tools.ietf.org/html/draft-ietf-oauth-json-web-token-12 [21:36:03] * marktraceur grumbles [21:36:27] I swear, if I wind up becoming the de-facto maintainer for FileRepo, I will be very annoyed but also secretly amused. [21:37:08] Not very secretly [21:39:33] you fix it you own it [21:39:57] Eloquence: See also UploadWizard [21:40:02] indeed :) [21:40:07] (03CR) 10JGonera: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88677 (owner: 10JGonera) [21:40:19] James_F: i still can't edit on mediawiki.org with ve :( [21:40:25] keep forgetting and losing edits [21:40:39] did i hear there was a gadget that lets me extract the wikitext? [21:40:57] jdlrobson, https://en.wikipedia.org/wiki/User:John_Vandenberg/switch_editor [21:40:59] sorta works [21:41:01] Um wat. [21:41:06] thanks Eloquence [21:41:13] I thought there wasn't a public API yet [21:41:25] Uncaught NotFoundError: An attempt was made to reference a Node in a context where it does not exist again [21:41:34] gadget is .. an overstatement :) [21:42:14] "again" meaning this is the second NotFoundError or the Node was referenced before, and it existed, but not anymore? [21:43:02] Hah, it's a VE mod [21:43:09] I hope that becomes a thing [21:44:01] (03PS3) 10BryanDavis: Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 [21:44:40] marktraceur: i raised a bug [21:44:53] i showed James_F the other day [21:45:04] visual editor just doesn't work for me on mediawiki.org [21:45:05] raise BugInMediaWikiException [21:45:33] class BugInMediaWikiException(WhatElseIsNewException): [21:45:46] that would be good to sort out before we switch wikipedia.org to wmf20 tomorrow :) [21:45:53] Hm, yeah [21:45:59] jdlrobson: What platform? [21:46:36] marktraceur: https://bugzilla.wikimedia.org/show_bug.cgi?id=55453 [21:46:38] jdlrobson: what exactly do you mean by "doesn't work"? can't use controls? doesn't save? [21:46:51] all details there chrismcmahon marktraceur ^ [21:46:55] chrismcmahon: can't save [21:46:57] but can edit [21:47:08] although without save that edit is worthless :) [21:47:17] can confirm on that page [21:47:37] mediawiki.org has a fair bit of uncommon markup, so this issue may be narrowly manifesting there. it also has some experimental VE stuff enabled. [21:48:04] jdlrobson: nice. the contentEditable field on that particular page never gets the focus I bet [21:48:38] jdlrobson: Except your OS, and you put the platform details in a comment :) [21:48:44] jdlrobson: OS X 10.what? [21:48:47] jdlrobson: thus the Save page never gets enabled [21:48:59] jdlrobson, Eloquence: Can you try clicking out of the editor (e.g. in empty space in the sidebar) so the editor loses focus, then back in again so it regains focus? [21:49:14] and then start typing and see if the save button state changes [21:49:34] If that works then it's probably that ULS bug again [21:49:51] Maybe we neglected to backport our earlier fix for it or something [21:50:02] RoanKattouw, nope [21:50:07] RoanKattouw: Chrome/Ubunto I can't put focus on CE on that page regardless of clicking in and out [21:50:13] (03CR) 10BryanDavis: "Corrected expects for Hashar's tests and added a few more." [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [21:50:25] I get an awesome spinner of doom when mousing over the save button though [21:50:27] Oh, exception in the converter? [21:50:31] Yeah then you're screwed [21:51:07] Uncaught NotFoundError: An attempt was made to reference a Node in a context where it does not exist. [21:51:35] jdlrobson: What version of OS X are you using? [21:51:55] 10.7.5 [21:52:06] Eloquence: yeh i get that too [21:52:23] I can reproduce that [21:52:32] So, ignore all the behavior while in the editor [21:52:35] It's already crapped itself [21:52:45] when reading in the document [21:52:59] RoanKattouw, I like it when you get all technical [21:53:33] csteipp: If you'd like https://gerrit.wikimedia.org/r/#/c/88533 to be in wmf22 now is you chance to act. :) [21:55:42] ve.ce.Surface.crap() [21:55:42] (03PS8) 10Krinkle: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [21:56:03] bd808: wmf21, right? [21:56:06] I probably screwed up that joke :) [21:56:08] * RoanKattouw is making sure he didn't go crazy [21:56:26] marktraceur: Something more appropriate would be this.getSurface().getSurface().crap(); [21:56:32] (yes, we have this.surface.surface, don't ask) [21:56:41] (03CR) 10Krinkle: "* A few more double spaces" [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [21:56:58] RoanKattouw: I will some other time [21:57:06] RoanKattouw: You're right, wmf21 [21:57:24] (03CR) 10Ori.livneh: [C: 031] "Works for me; thanks! +1 for Krinkle's emendations." [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [21:57:36] RoanKattouw: Two surfaces are sitting in memory and one says to the other "I'm tired of dealing with this.crap" [21:57:44] ll [21:57:46] *lol [21:58:10] (03CR) 10Krinkle: [C: 032] Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [22:01:20] (03Merged) 10jenkins-bot: Add mediawiki.inspect module [core] - 10https://gerrit.wikimedia.org/r/88408 (owner: 10Ori.livneh) [22:02:30] Eloquence, jdlrobson: Found it, it's a bug in the support [22:02:46] It turns out that domElement.appendChild( child ); doesn't work when child === undefined , who knew [22:02:51] That doesn't manifest in tags? [22:03:02] In the tag support in VE only [22:03:04] RoanKattouw to the rescue [22:03:09] Because I was able to edit [[Extension:BetaFeatures]] without issue [22:03:43] Well, it depends on whether a certain bug in Parsoid is triggered [22:03:49] The code assumes that it is [22:03:59] So if Parsoid does indeed have the bug, you're fine. If not, it crashes [22:04:06] RoanKattouw: Assuming bugs, bad plan [22:04:20] This is trivially fixable [22:05:36] RoanKattouw, cool - that's what I suspected on that page [22:05:51] (03PS5) 10MarkTraceur: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:05:55] (03CR) 10Aaron Schulz: [C: 032] Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [22:05:57] (03PS11) 10MarkTraceur: Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [22:05:58] Also, very bad DOM hygiene in that code, it inserts DOM elements that were attached somewhere else and were attached to the wrong document [22:06:03] (03PS11) 10MarkTraceur: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [22:06:11] That was magical [22:06:16] Three rebases, no problems [22:07:44] RoanKattouw: that proliferation of iframes, then? that's how I read "wrong document" [22:08:07] No, there are multiple documents involved before we even get to the iframes [22:08:15] Krinkle: thanks again for the review [22:08:16] RoanKattouw: which bug in Parsoid are you referring to? [22:08:32] The document containing the HTML received from Parsoid, the main document used for rendering, and the document containing HTML being sent back to Parsoid [22:09:01] gwicke:
...
\n It usually does that but not alwayas [22:09:26] The code assumed that it would always happen and that there would always be a span [22:09:42] we just encapsulate the extension output [22:09:53] if the extension emits trailing newlines, they are faithfully encapsulated [22:09:58] Right [22:10:01] That makes sense [22:10:18] We should probably fix the extension not to usually output a newline except when it doesn't [22:10:23] A lot of tag extensions have this problem [22:10:52] visually it does not matter much [22:11:27] not sure if this is just an oversight or if there is a reason for it around tidy interaction or who-knows-what [22:12:28] (03Merged) 10jenkins-bot: Escape backticks when quoting MySQL identifiers [core] - 10https://gerrit.wikimedia.org/r/88533 (owner: 10BryanDavis) [22:12:51] ^d: https://gerrit.wikimedia.org/r/#/c/88399/ [22:14:03] (03CR) 10Chad: [C: 032] Made root job de-duplication work without cache setup [core] - 10https://gerrit.wikimedia.org/r/88399 (owner: 10Aaron Schulz) [22:14:24] (03PS3) 10Aaron Schulz: Made JobQueueDB use getConnectionRef [core] - 10https://gerrit.wikimedia.org/r/88128 [22:14:56] ^d: https://gerrit.wikimedia.org/r/#/c/88128/ [22:15:37] (03PS1) 10MarkTraceur: Configurablize the iiprop for ForeignAPIRepos [core] - 10https://gerrit.wikimedia.org/r/88889 [22:15:42] * marktraceur is excited about that one [22:16:33] (03PS6) 10BryanDavis: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:16:54] (03CR) 10Chad: [C: 032] Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 (owner: 10Aaron Schulz) [22:17:27] NO WAIT [22:17:31] Configurable PER REQUEST [22:17:33] Of course [22:21:22] (03Merged) 10jenkins-bot: Made root job de-duplication work without cache setup [core] - 10https://gerrit.wikimedia.org/r/88399 (owner: 10Aaron Schulz) [22:23:12] (03CR) 10jenkins-bot: [V: 04-1] Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 (owner: 10Aaron Schulz) [22:27:01] (03PS7) 10BryanDavis: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:27:47] (03CR) 10BryanDavis: "(5 comments)" [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:27:51] (03PS4) 10Aaron Schulz: Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 [22:28:07] (03CR) 10Aaron Schulz: [C: 032] "Trivial conflict merge" [core] - 10https://gerrit.wikimedia.org/r/88128 (owner: 10Aaron Schulz) [22:28:19] (03CR) 10BryanDavis: [C: 031] Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:28:47] (03CR) 10jenkins-bot: [V: 04-1] Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:31:12] (03PS6) 10Aaron Schulz: Added supported for retrieving file metadata/headers [core] - 10https://gerrit.wikimedia.org/r/86642 [22:31:30] (03CR) 10jenkins-bot: [V: 04-1] Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 (owner: 10Aaron Schulz) [22:33:13] (03PS8) 10BryanDavis: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:33:43] (03PS5) 10Aaron Schulz: Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 [22:34:20] (03PS1) 10Legoktm: Move Special:Drafts specific code into SpecialDrafts.php [extensions/Drafts] - 10https://gerrit.wikimedia.org/r/88893 [22:34:21] (03PS1) 10Legoktm: Modify Drafts::display to return HTML instead of a count [extensions/Drafts] - 10https://gerrit.wikimedia.org/r/88894 [22:34:22] (03PS1) 10Legoktm: Make [[Special:Drafts/title]] show drafts for that title [extensions/Drafts] - 10https://gerrit.wikimedia.org/r/88895 [22:35:06] (03PS6) 10Aaron Schulz: Made JobQueueDB use getConnectionRef() [core] - 10https://gerrit.wikimedia.org/r/88128 [22:35:46] (03CR) 10BryanDavis: [C: 031] "This time I actually ran the unit tests locally before saying that it looks good." [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [22:36:26] (03PS12) 10BryanDavis: Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [22:36:29] d^: :) [22:37:11] (03PS2) 10MarkTraceur: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 [22:37:12] Bettar [22:37:51] Now I just need to replicate the template structure of Commons on my local remote repository and I can start working on CommonsMetadata support on my local machine [22:38:01] "just" [22:50:40] werdna: any chance you want to review some Drafts patches? :) [22:50:48] * werdna hides [22:51:25] I have some really stupid ones open like fixing comments [22:51:35] like https://gerrit.wikimedia.org/r/#/c/88245/ [22:51:48] or https://gerrit.wikimedia.org/r/#/c/88243/1 which is equally as trivial [22:52:44] (03PS1) 10Krinkle: Minor clean up of css/js [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/88897 [22:59:26] (03CR) 10Tim Starling: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/62173 (owner: 10Jakub Vrána) [22:59:53] (03CR) 10TTO: [C: 04-1] "(1 comment)" [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/88897 (owner: 10Krinkle) [23:01:18] * marktraceur swears loudly [23:01:37] Special:Export is exporting the templates on a page, but not the templates in the templat.e [23:03:46] [x] [23:04:05] I did [23:04:56] So Special:Export/File:Foobar.jpg exports Template:Information, but...OH [23:05:03] This isn't a template issue, it's a message issue [23:05:04] OK [23:05:18] And maybe a module issue [23:06:08] (03CR) 10Krinkle: "(1 comment)" [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/88897 (owner: 10Krinkle) [23:06:43] * Reedy puts the blame wheel away [23:07:39] (03CR) 10TTO: [C: 031] "(1 comment)" [extensions/CategoryTree] - 10https://gerrit.wikimedia.org/r/88897 (owner: 10Krinkle) [23:11:31] (03PS1) 10Ryan Lane: Update OATHAuth/OpenStackManager for wikitech [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88898 [23:13:45] (03PS2) 10EBernhardson: Add bezels to .mw-ui-button classes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/88881 [23:25:07] Urgh, weird, my remote API is giving me the descriptions and everything, but the local one isn't [23:25:12] I guess there's some caching somewhere [23:25:17] Which is bad because I disabled it [23:25:59] (03CR) 10Ryan Lane: [C: 032] Update OATHAuth/OpenStackManager for wikitech [core] (wmf/1.22wmf20) - 10https://gerrit.wikimedia.org/r/88898 (owner: 10Ryan Lane) [23:26:59] Re [[Extension:CommonsMetadata]] [23:33:55] (03CR) 10MarkTraceur: [C: 04-1] "Agh" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [23:34:33] * marktraceur swears loudly at bawolff who isn't here [23:35:42] (03CR) 10MarkTraceur: [C: 04-1] "Doesn't work if the host of the image is a ForeignAPIRepo" [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/80403 (owner: 10Brian Wolff) [23:36:25] (03CR) 10MarkTraceur: "To reproduce you'll need I4a1b46669658cc7b7d074f8fb85dd172e6c45633" [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/80403 (owner: 10Brian Wolff) [23:45:09] (03CR) 10Tim Starling: [C: 032] Fixed trx isolation bug that could cause text not to be found [core] - 10https://gerrit.wikimedia.org/r/88761 (owner: 10Aaron Schulz) [23:46:15] (03PS1) 10MaxSem: Don't save secondary coordinates equal to primary one [extensions/GeoData] - 10https://gerrit.wikimedia.org/r/88903 [23:48:06] (03Merged) 10jenkins-bot: Fixed trx isolation bug that could cause text not to be found [core] - 10https://gerrit.wikimedia.org/r/88761 (owner: 10Aaron Schulz)