[00:01:47] hmm, anybody good with git around? I have a revision which is based on another revision, and I want it to not be based on that revision [00:01:55] they touch different files [00:02:03] (03CR) 10Jeroen De Dauw: "I doubt this is coveralls specific, but rather is caused somehow by the travis build. Or is there reason to believe otherwise?" [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78155 (owner: 10Mwjames) [00:02:21] werdna: git rebase HEAD~1 [00:02:40] what's the HEAD~1 for? [00:03:16] HEAD~1 is, I assume, the commit that I don't want it to be based on [00:03:16] it's telling the rebase that you want it to substitute out that revision [00:03:20] yep yep [00:04:03] yeah see that just moved my HEAD position back one. [00:04:13] oh duh; I'm sorry; missed out an important bit [00:04:16] git rebase --onto [00:04:21] HEAD~1 [00:05:07] (03PS1) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [00:05:12] that's better. [00:05:13] thanks mw-jenkinsbot [00:05:16] also mwalker [00:05:33] np [00:05:58] (03CR) 10Parent5446: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [00:06:22] (03CR) 10Werdna: "Also, any sort of action seems to be broken now :D. I'll see what's going on there tomorrow." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 (owner: 10Werdna) [00:06:29] (03PS1) 10Mwalker: Better Method of Decoding URLs [extensions/FundraisingEmailUnsubscribe] - 10https://gerrit.wikimedia.org/r/78168 [00:06:40] (03PS2) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [00:09:46] (03CR) 10Katie Horn: [C: 032 V: 032] Better Method of Decoding URLs [extensions/FundraisingEmailUnsubscribe] - 10https://gerrit.wikimedia.org/r/78168 (owner: 10Mwalker) [00:14:11] (03CR) 10Daniel Friesen: [C: 04-1] "...I can't even begin to point out how much of this looks like a bad idea skimming over the code." [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [00:14:49] (03PS1) 10Mwalker: Updating DonationEmailUnsubscribe to Master [core] (fundraising/1.22) - 10https://gerrit.wikimedia.org/r/78169 [00:15:20] (03CR) 10Mwalker: [C: 032 V: 032] Updating DonationEmailUnsubscribe to Master [core] (fundraising/1.22) - 10https://gerrit.wikimedia.org/r/78169 (owner: 10Mwalker) [00:20:44] (03CR) 10Werdna: "That's not very productive feedback. I think it would be polite to at least explain why you feel that way." [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [00:30:22] (03CR) 10Bsitu: "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/76923 (owner: 10Matthias Mullie) [00:35:57] man. who thought that having a separate command to update submodules was a good idea [00:42:03] (03CR) 10Daniel Friesen: "$wgRequest, conditional hooks in core, and unnecessary closures make this code look like a complete mess of a hack." [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [01:06:11] (03CR) 10Aaron Schulz: "(6 comments)" [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [01:10:11] Elsie: Did you see https://gerrit.wikimedia.org/r/#/c/78047/ ? [01:10:25] Probably. [01:10:50] No, I don't think so. [01:10:59] Thoughts about it? It's based on some feedback from Lydia when I showed her the form yesterday [01:16:46] I never liked the checkbox. [01:17:29] legoktm: Killing the checkbox seems fine. [01:17:40] I guess there's somewhat of a concern that any local admin will be able to spam cross-wiki. [01:17:50] But really that's possible now, it's just more tedious. [01:18:14] OTOH, accountability may be an issue. If you get a message from a random bot, how do you know where the message came from? [01:18:28] Elsie: I think we should still require the massmessage-global right, and just remove entries when parsing the spamlist [01:18:29] Who sent it, why you're receiving it, etc. [01:18:42] I'm not sure. [01:19:17] legoktm: You can't silently remove entries. [01:19:29] You'd have to either disable the bot run or something. [01:19:33] I could do it loudly. [01:19:48] Yeah... [01:27:36] (03CR) 10Brion VIBBER: "Looks plausible (not yet tested), coupla quick notes:" [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [01:31:44] (03CR) 10MarkAHershberger: "Updating the serialized files is done in the serialized subdirectory. The make files there are obsolete. They should be updated and some" [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [01:46:31] Is there any reliable way to determine from the server side if the user viewing a page supports JavaScript? I'm working on putting together a custom EditPage, and it would be useful [01:48:22] (03CR) 10PleaseStand: "I already uploaded change I05afece3 to remove the Makefile from the serialized directory." [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [01:51:19] GorillaWarfare: the only truly reliable way is to do something from javascript that changes your server-side state [01:54:09] GorillaWarfare: Generally such a check shouldn't be necessary. [01:54:10] I think. [01:54:29] Given that we want graceful degradation if JS isn't available, it really shouldn't matter if the user supports JS. [01:54:50] You can use no-js classes if you need something client-side, of course. [01:55:06] On the tag, that is. [01:55:22] Or tag? I forget and I'm tired. [01:59:46] Yeah, I can do it by just tacking on the fancy bits with JS [01:59:51] It would just be easier to do it with PHP :P [02:01:07] (03CR) 10Aaron Schulz: [C: 032 V: 032] Use WebResponse::setcookie() [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/78147 (owner: 10Anomie) [02:01:12] (03Merged) 10jenkins-bot: Use WebResponse::setcookie() [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/78147 (owner: 10Anomie) [02:03:41] (03CR) 10Aaron Schulz: [C: 032 V: 032] Use wfResetSessionID() [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/78148 (owner: 10Anomie) [02:06:16] (03CR) 10MarkAHershberger: "Could you provide Makefile's rules for the files in the serialized directory so that we can easily update these files at release time?" [core] - 10https://gerrit.wikimedia.org/r/76882 (owner: 10PleaseStand) [02:09:57] (03CR) 10GWicke: "Do we have any stats on how common this construct is in existing content?" [core] - 10https://gerrit.wikimedia.org/r/77988 (owner: 10Cscott) [02:27:14] (03PS1) 10Aaron Schulz: Break some long lines [core] - 10https://gerrit.wikimedia.org/r/78175 [02:38:44] yuvipanda: can you double-check the LRU cache setting on the bitmap cache? if i'm reading this correctly, on nexus 1 the cache is…. 3 kilobytes. https://trello.com/c/PIkDlAKA/7-3-bug-50733-app-crashes-with-out-of-memory-error-soon-after-login [02:39:56] i'm just not sure a bitmap cache is a win anyway tho [02:44:27] (03PS1) 10Siebrand: Update formatting [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78177 [02:44:28] (03PS1) 10Siebrand: Comment out unused local variable [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78178 [02:44:29] (03PS1) 10Siebrand: Update documentation and remove use of deprecated method [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78179 [03:00:33] (03PS1) 10Siebrand: Use WikiPage::factory() instead of Article::newFromTitle() [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78181 [03:15:20] (03PS1) 10Siebrand: Add some FIXMEs based on static code analysis [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78182 [03:17:49] (03CR) 10Krinkle: "Was there no better place to put this then another global function? Meh." [core] - 10https://gerrit.wikimedia.org/r/78145 (owner: 10Anomie) [03:21:59] yuvipanda: i've added a pull req on https://trello.com/c/PIkDlAKA/7-3-bug-50733-app-crashes-with-out-of-memory-error-soon-after-login [03:22:07] please let me know if it has any negative side effects [03:22:15] i just took out the in-memory Bitmap cache [03:22:19] (03CR) 10Krinkle: [C: 04-1] "(2 comments)" [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78181 (owner: 10Siebrand) [03:22:27] downside: pulling from volley's cache will mean re-decoding the file [03:22:39] upside: saves a few megs of memory on low-ram devices, and doesn't crash ;) [03:27:39] (03CR) 10Legoktm: [C: 032] Update formatting [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78177 (owner: 10Siebrand) [03:27:46] (03Merged) 10jenkins-bot: Update formatting [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78177 (owner: 10Siebrand) [03:29:35] (03CR) 10Legoktm: [C: 032] Comment out unused local variable [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78178 (owner: 10Siebrand) [03:29:40] (03Merged) 10jenkins-bot: Comment out unused local variable [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78178 (owner: 10Siebrand) [03:36:26] (03CR) 10Legoktm: [C: 032] "(2 comments)" [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78181 (owner: 10Siebrand) [03:37:20] (03CR) 10Legoktm: [C: 032] Update documentation and remove use of deprecated method [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78179 (owner: 10Siebrand) [03:37:25] (03Merged) 10jenkins-bot: Use WikiPage::factory() instead of Article::newFromTitle() [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78181 (owner: 10Siebrand) [03:37:26] (03Merged) 10jenkins-bot: Update documentation and remove use of deprecated method [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78179 (owner: 10Siebrand) [03:45:58] (03PS1) 10EBernhardson (WMF): various updates [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78186 [03:47:25] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] various updates [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78186 (owner: 10EBernhardson (WMF)) [03:55:17] RoanKattouw: https://gerrit.wikimedia.org/r/#/c/74570/ [04:03:58] (03PS1) 10EBernhardson (WMF): More work finishing up UUID handling [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78188 [04:16:06] (03CR) 10Parent5446: "Unfortunately not. All the session logic is in the global scope. But I'm working on it. ;)" [core] - 10https://gerrit.wikimedia.org/r/78145 (owner: 10Anomie) [04:30:41] (03PS2) 10EBernhardson (WMF): More work finishing up UUID handling [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78188 [04:31:09] (03PS1) 10Legoktm: Fix undefined $dbname variable [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78192 [04:31:22] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] More work finishing up UUID handling [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78188 (owner: 10EBernhardson (WMF)) [04:37:12] (03CR) 10Legoktm: "Basically the whole getDBname function is messed up (per our discussion we should use CentralAuth), so I'm going to re-write that in anoth" [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78182 (owner: 10Siebrand) [04:47:57] (03PS2) 10EBernhardson (WMF): Decompress revision text [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78081 (owner: 10Matthias Mullie) [04:52:24] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] "Just needed a quick rebase against todays development" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78081 (owner: 10Matthias Mullie) [04:52:38] (03CR) 10Pypt: "I've found out that the only part of MediaWiki that depends on PHP's DOM (except for a couple of maintenance scripts) is includes/parser/P" [core] - 10https://gerrit.wikimedia.org/r/76663 (owner: 10MarkAHershberger) [04:54:06] (03CR) 10Tim Starling: [C: 04-2] "I think User::loadFromSession() should not even exist (bug 41201), so if the reason for clearing $_COOKIE is to allow User::loadFromSessio" [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [05:03:17] (03CR) 10Parent5446: "I completely agree with the approach Tim mentioned. It is exactly what I envisioned us moving to when I made the ClientSession section of " [core] - 10https://gerrit.wikimedia.org/r/78146 (owner: 10Anomie) [05:29:56] (03PS1) 10PleaseStand: Remove remaining calls to MWInit methods [core] - 10https://gerrit.wikimedia.org/r/78202 [05:44:32] yuvipanda: uga? [05:58:02] thehelpfulone or any WM14 wiki bureaucrats around? [05:59:57] (03PS1) 10EBernhardson (WMF): Separate row compact/expand into class [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78203 [06:03:21] https://epl.wikimedia.org/p/ArchitectureHongKong [06:03:27] varnent: Can I help? [06:04:08] (03PS2) 10EBernhardson (WMF): Separate row compact/expand into class [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78203 [06:04:20] James_F: looking to get sysop again this year and start setting up some of the Hackathon and templates while we're here and talking/thinking about it [06:04:24] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] Separate row compact/expand into class [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78203 (owner: 10EBernhardson (WMF)) [06:04:34] if you don't wanna - that's cool - you know I'm not easily offended :) [06:11:31] varnent: For WM2014wiki? You should talk to RockDrum. [06:11:37] varnent: He's here: [06:11:42] James_F: cool beans - ty [06:11:47] RockDrum: pinger [06:12:11] varnent: He's not in this IRC channel. [06:12:16] oh - ha - like physically here [06:12:18] varnent: I meant here as in on my desk. :-) [06:12:23] a crazy concept.... [06:12:30] Yeah. [06:13:40] kart_: ping? [06:15:44] (03PS3) 10EBernhardson (WMF): Revision cleanup [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78057 (owner: 10Matthias Mullie) [06:16:08] (03CR) 10Wizardist: [C: 032 V: 032] Broken tabs at target NS if key == NS_name [extensions/NamespaceRelations] - 10https://gerrit.wikimedia.org/r/66333 (owner: 10Wizardist) [06:16:39] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] "Needed a rebase, but otherwise excellent" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78057 (owner: 10Matthias Mullie) [06:17:32] that was almost more in-person and non-IRC interaction than my brain can handle :) [07:04:24] (03CR) 10Tpt: [C: 04-1] "Some comments on patch set 26 haven't an answer + 2 remaining indentation problems (use of spaces instead of tabs)" [extensions/ProofreadPage] (pagePagesRefactoring) - 10https://gerrit.wikimedia.org/r/74850 (owner: 10Rtdwivedi) [07:12:40] (03CR) 10Tpt: [C: 04-1] "(5 comments)" [extensions/ProofreadPage] (pagePagesRefactoring) - 10https://gerrit.wikimedia.org/r/76891 (owner: 10Rtdwivedi) [07:13:12] (03CR) 10Siebrand: [C: 032] "Sorry for taking this long. Going over your open patch sets now to see what needs love and what I can deal with immediately." [core] - 10https://gerrit.wikimedia.org/r/55454 (owner: 10Liangent) [07:16:07] (03Merged) 10jenkins-bot: (bug 46463) Mark edit preview as in page view language [core] - 10https://gerrit.wikimedia.org/r/55454 (owner: 10Liangent) [07:16:36] (03CR) 10Siebrand: "This one's been inactive for a long time. Maybe work it out now that you're both in Hong Kong?" [core] - 10https://gerrit.wikimedia.org/r/22699 (owner: 10Liangent) [07:18:54] (03CR) 10Siebrand: "(1 comment)" [extensions/SpamBlacklist] - 10https://gerrit.wikimedia.org/r/43496 (owner: 10Liangent) [07:26:55] any tool labs folks here? hello? [07:29:28] You should really use #wikimedia-labs [07:39:30] (03CR) 10Siebrand: "Niklas, Tim, can you please revisit this patch set?" [core] - 10https://gerrit.wikimedia.org/r/10243 (owner: 10Liangent) [07:49:24] (03CR) 10Nikerabbit: [C: 04-1] "(2 comments)" [extensions/cldr] - 10https://gerrit.wikimedia.org/r/45499 (owner: 10Kaldari) [07:52:09] (03PS6) 10Siebrand: Change tag improvements. [core] - 10https://gerrit.wikimedia.org/r/45506 (owner: 10Liangent) [07:52:38] (03CR) 10Siebrand: "PS6 addresses comments by Niklas" [core] - 10https://gerrit.wikimedia.org/r/45506 (owner: 10Liangent) [07:57:30] (03PS14) 10Santhosh: Adding hook handler for creating human-friendly relative timestamp [extensions/cldr] - 10https://gerrit.wikimedia.org/r/45499 (owner: 10Kaldari) [07:58:29] (03CR) 10Siebrand: "Can this be a starting point? http://www.mediawiki.org/wiki/Manual:PHP_unit_testing/Writing_unit_tests_for_extensions" [extensions/ExpandTemplates] - 10https://gerrit.wikimedia.org/r/47333 (owner: 10Liangent) [07:59:27] (03PS4) 10Umherirrender: Add more parser tests for magic words. [core] - 10https://gerrit.wikimedia.org/r/76493 [07:59:59] (03PS15) 10Nikerabbit: Adding hook handler for creating human-friendly relative timestamp [extensions/cldr] - 10https://gerrit.wikimedia.org/r/45499 (owner: 10Kaldari) [08:00:11] (03CR) 10Nikerabbit: [C: 032] Adding hook handler for creating human-friendly relative timestamp [extensions/cldr] - 10https://gerrit.wikimedia.org/r/45499 (owner: 10Kaldari) [08:00:20] (03Merged) 10jenkins-bot: Adding hook handler for creating human-friendly relative timestamp [extensions/cldr] - 10https://gerrit.wikimedia.org/r/45499 (owner: 10Kaldari) [08:01:41] (03CR) 10Siebrand: [C: 032] New variable: new_pst which provides a PSTed version of new_wikitext [extensions/AbuseFilter] - 10https://gerrit.wikimedia.org/r/47582 (owner: 10Liangent) [08:01:49] (03Merged) 10jenkins-bot: New variable: new_pst which provides a PSTed version of new_wikitext [extensions/AbuseFilter] - 10https://gerrit.wikimedia.org/r/47582 (owner: 10Liangent) [08:05:45] (03CR) 10Siebrand: "Can http://www.mediawiki.org/wiki/Manual:PHP_unit_testing/Writing_unit_tests be a start? I'd advise you to use the current hackathon as an" [core] - 10https://gerrit.wikimedia.org/r/53740 (owner: 10Liangent) [08:10:03] (03PS12) 10Matthias Mullie: Send Echo notifications for AFTv5 [extensions/ArticleFeedbackv5] - 10https://gerrit.wikimedia.org/r/74166 [08:10:34] (03PS5) 10Umherirrender: Add more parser tests for magic words. [core] - 10https://gerrit.wikimedia.org/r/76493 [08:11:29] (03Abandoned) 10Matthias Mullie: If there's a fragment, it's probably important enough to include it [extensions/Echo] - 10https://gerrit.wikimedia.org/r/76923 (owner: 10Matthias Mullie) [08:11:48] (03CR) 10Umherirrender: "Patch Set 5: Added a TALKSPACE and SUBJECTSPACE test for main namespace" [core] - 10https://gerrit.wikimedia.org/r/76493 (owner: 10Umherirrender) [08:12:10] (03CR) 10Liangent: "Siebrand: Can you make it a separate patchset to do rebase next time? This makes diff between patches easier to read." [core] - 10https://gerrit.wikimedia.org/r/45506 (owner: 10Liangent) [08:14:07] (03CR) 10Liangent: "(1 comment)" [extensions/SpamBlacklist] - 10https://gerrit.wikimedia.org/r/43496 (owner: 10Liangent) [08:18:11] (03PS2) 10Siebrand: Apply conversion on namespace names on Special:Search [core] - 10https://gerrit.wikimedia.org/r/53756 (owner: 10Liangent) [08:18:51] (03CR) 10Siebrand: [C: 032] "PS2 resolves merge conflict and adds a few newlines." [core] - 10https://gerrit.wikimedia.org/r/53756 (owner: 10Liangent) [08:18:54] (03PS1) 10Ryan Lane: htmlform muliselect chosen: convert each list [core] - 10https://gerrit.wikimedia.org/r/78213 [08:19:11] Krinkle|detached, RoanKattouw_away: ^^ [08:21:36] (03CR) 10Siebrand: "Daniel, Niklas, any idea who can/should review this?" [core] - 10https://gerrit.wikimedia.org/r/59793 (owner: 10Liangent) [08:23:22] (03Merged) 10jenkins-bot: Apply conversion on namespace names on Special:Search [core] - 10https://gerrit.wikimedia.org/r/53756 (owner: 10Liangent) [08:24:51] (03PS1) 10Ryan Lane: htmlform multiselect chosen: Replace lists [core] - 10https://gerrit.wikimedia.org/r/78214 [08:31:54] (03CR) 10Siebrand: [V: 04-1] "Tyler, aside from the merge problem that is fixable, can you please assess if this can be merged in the current state?" [core] - 10https://gerrit.wikimedia.org/r/65477 (owner: 10Liangent) [08:32:49] (03PS2) 10Ryan Lane: Clean up service group interface [extensions/OpenStackManager] - 10https://gerrit.wikimedia.org/r/77854 [08:33:36] (03CR) 10Siebrand: [C: 031] "Added Aaron as reviewer. Hope he feels comfortable to merge." [core] - 10https://gerrit.wikimedia.org/r/65489 (owner: 10Liangent) [08:36:16] (03CR) 10Siebrand: [C: 031] "Added Aaron and Tim. Hope one of them can sign off on the performance impact." [extensions/ParserFunctions] - 10https://gerrit.wikimedia.org/r/65828 (owner: 10Liangent) [08:38:35] (03CR) 10Siebrand: "Brad, can you please revisit this patch set? It's been quiet for 2 months..." [extensions/Scribunto] - 10https://gerrit.wikimedia.org/r/65998 (owner: 10Liangent) [08:38:57] (03CR) 10Nikerabbit: [C: 04-1] "(5 comments)" [core] - 10https://gerrit.wikimedia.org/r/62600 (owner: 10Liangent) [08:39:46] (03PS1) 10Matthias Mullie: Fix schema [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78219 [08:41:49] (03CR) 10Siebrand: "I added a few platform engineers as reviewers. Hope they can provide sane comments." [core] - 10https://gerrit.wikimedia.org/r/66338 (owner: 10Liangent) [09:10:29] Nikerabbit: are we closer to having proper date formats in MoodBar etc. and to gettin rid of MediaWiki:Ago ? https://gerrit.wikimedia.org/r/#/c/45499/ [09:28:10] (03PS1) 10Legoktm: Fix getDBName to read from $wgConf and cache if possible [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78222 [09:28:54] (03CR) 10Legoktm: "Everything should be fixed in https://gerrit.wikimedia.org/r/#/c/78222/" [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78182 (owner: 10Siebrand) [09:43:54] (03PS1) 10Mwjames: (Bug 52642) Add query duration as meta data [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78223 [10:08:10] (03CR) 10Nemo bis: "Thanks Siebrand for the review. I've fixed the offending piece of the docs: https://www.mediawiki.org/w/index.php?title=Manual%3AMessages_" [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/77898 (owner: 10Nemo bis) [11:55:00] !log deployment-prep all search indecies have finished building except simplewiki [11:55:00] There is no such key, you probably want to try: !anonnotice, !api, !chansearch, !cms, !debug, !errors, !gitaccount, !logo, !logs, !revdelete, !sal, !sharedlogins, !sqllog, !sul, !tautology, [11:56:12] thanks wm-bot. I've logged correctly at #-labs. Where the bot is nicer. [11:56:12] Hey manybubbles, you are welcome! [12:22:39] (03CR) 10Parent5446: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/65477 (owner: 10Liangent) [13:18:39] (03CR) 10Jeroen De Dauw: "(5 comments)" [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78223 (owner: 10Mwjames) [13:20:31] (03PS1) 10Zfilipin: Updated Ruby gems [qa/browsertests] - 10https://gerrit.wikimedia.org/r/78233 [13:30:43] (03PS1) 10Zfilipin: WIP cache test [qa/browsertests] - 10https://gerrit.wikimedia.org/r/78234 [13:33:13] (03CR) 10Ori.livneh: [C: 032] htmlform multiselect chosen: Replace lists [core] - 10https://gerrit.wikimedia.org/r/78214 (owner: 10Ryan Lane) [13:34:55] (03Abandoned) 10Manybubbles: Add tests for CirrusSearch. [qa/browsertests] - 10https://gerrit.wikimedia.org/r/75793 (owner: 10Manybubbles) [13:43:53] (03CR) 10Zfilipin: "Could you please provide link to the new location, or add me as a reviewer there?" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/75793 (owner: 10Manybubbles) [13:44:31] (03CR) 10Manybubbles: "https://gerrit.wikimedia.org/r/#/c/78118/" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/75793 (owner: 10Manybubbles) [13:45:05] (03PS2) 10Zfilipin: WIP Feature: Cache [qa/browsertests] - 10https://gerrit.wikimedia.org/r/78234 [13:54:01] (03CR) 10Zfilipin: "Not related to this commit, but you should not commit from master branch, you should always create a topic branch and commit there. Please" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [13:57:04] (03PS1) 10J: Parse Ogg Opus files and enable playback if supported [extensions/TimedMediaHandler] - 10https://gerrit.wikimedia.org/r/78235 [13:57:14] (03PS1) 10Matthias Mullie: Various small changes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78236 [13:58:09] (03CR) 10Manybubbles: "In this case I didn't do this because I only use this checkout (on my laptop) for the browser tests and I won't be working on any more bro" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [14:02:51] (03CR) 10Reedy: "What difference does it make?" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [14:07:54] (03CR) 10Manybubbles: "He is right that committing on master can be trouble because you can end up with those long unrelated commit chains so you shouldn't do it" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [14:17:38] (03PS2) 10Matthias Mullie: Various small changes [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78236 [14:46:41] (03CR) 10Zfilipin: "(10 comments)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [14:49:17] (03PS1) 10Manybubbles: Use the archive table to find deletes. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78237 [14:50:32] (03CR) 10Zfilipin: "Reedy, Manybubbles: when I see somebody committing to master I assume that they are not familiar with Git. Committing to master (and pushi" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [14:52:06] (03CR) 10Mwjames: "(5 comments)" [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78223 (owner: 10Mwjames) [14:54:07] (03CR) 10Zfilipin: [C: 031] "Running tests on en.wikipedia.beta.wmflabs.org." [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [15:05:15] (03PS2) 10Manybubbles: Add browser tests. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 [15:12:56] (03CR) 10Manybubbles: "(11 comments)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [15:16:54] (03CR) 10Demon: [C: 032] Fix broken analyzer customization. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78151 (owner: 10Manybubbles) [15:17:01] (03Merged) 10jenkins-bot: Fix broken analyzer customization. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78151 (owner: 10Manybubbles) [15:17:23] Thank you, ^d. [15:17:31] <^d> You're welcome. [15:17:39] <^d> I see what you were *trying* to do. [15:17:46] <^d> Array references don't quite work like that in PHP :) [15:17:57] <^d> New version is a bit more clear and will obviously work :) [15:18:49] ^d: yeah - that was unuxpected. [15:20:58] (03CR) 10Cscott: [C: 031] "Works for me." [core] - 10https://gerrit.wikimedia.org/r/76493 (owner: 10Umherirrender) [15:27:00] <^d> manybubbles: Ouch, that query on archive filesorts :\ [15:27:31] <^d> http://p.defau.lt/?oQpRQ4oK1EFqsnQdZL1GKA [15:27:35] ^d: all the ones based on timestamp do that. that is why we use the id based one when we can. [15:28:12] ^d: better than using temporary, but yeah, it isn't good. [15:28:32] * apergos peeks in [15:29:29] fyi googlebot is blocked temporarily cause even with reduction in requests it was still killing us... [15:29:35] <^d> :( [15:29:38] (03PS3) 10Cscott: New parser test: check localization of magic words. [core] - 10https://gerrit.wikimedia.org/r/57538 [15:30:00] I passed it on to them, but when we're good to go I should let em know (I think people are not on line yet fwiw) [15:30:01] <^d> Let's fix robots.txt [15:30:06] yep [15:30:33] so you rewrite everything as 'have gitblit serve this' eh? [15:31:51] <^d> RewriteRule ^/(.*) http://127.0.0.1:8080/$1 [NE,P] [15:31:52] <^d> Yep [15:32:16] <^d> Maybe should just add a RewriteCond above it to skip robots.txt [15:32:34] sure sounds like the simple way [15:32:42] I mean how many of these are we gonna have [15:33:07] <^d> Good question :\ [15:33:19] uh oh :-D [15:33:24] <^d> I'm thinking of moving svn r/o to antimony as well. [15:33:34] <^d> Which means we'd need /svnroot/ and /viewvc/ available. [15:33:34] that's logical [15:33:50] ok so there's three. [15:33:51] <^d> Actually, different vhost. [15:33:52] <^d> Nvm. [15:33:56] even better [15:34:11] <^d> So just robots.txt I suppose. [15:34:15] yup [15:40:13] <^d> Hmm, dunno why bot didn't post to ops, but we're talking here anyway. [15:40:26] <^d> https://gerrit.wikimedia.org/r/#/c/78239/ [15:40:55] (03PS6) 10Frédéric Wang: Math: Enable MathJax's Safe mode [extensions/Math] - 10https://gerrit.wikimedia.org/r/64523 (owner: 10TheDJ) [15:40:56] (03CR) 10jenkins-bot: [V: 04-1] Math: Enable MathJax's Safe mode [extensions/Math] - 10https://gerrit.wikimedia.org/r/64523 (owner: 10TheDJ) [15:41:38] I think it's without the = [15:42:34] ^d: [15:46:06] <^d> That's what google told me to copy+paste ;-) [15:46:13] :-D [15:46:24] didn't google get us into this mess? :-P [15:46:56] <^d> silly googlebot. [15:47:15] trix are for kids... [15:47:47] (03CR) 10Physikerwelt: "The "build failure" is:" [extensions/Math] (LaTeXML) - 10https://gerrit.wikimedia.org/r/74524 (owner: 10Physikerwelt) [15:48:11] <^d> Anyway, easy enough to test :) [15:48:11] (03PS1) 10Manybubbles: Imporve reindex logging. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78241 [15:49:31] (I looked at a bunch of other examples in the puppet manifests) [15:51:35] <^d> Mmm, bagels. Now I can think. [15:51:38] (03CR) 10Frédéric Wang: "MathJax.Hub.config contains the current config so you can use that to preserve the extensions array. I've tried a minimal change to give a" [extensions/Math] - 10https://gerrit.wikimedia.org/r/64523 (owner: 10TheDJ) [15:53:00] <^d> apergos: Amended, you're right. [15:53:04] cool [15:55:38] do you need me to run puppet over there? [15:56:05] oh well I'm there now and you aren't so [15:59:03] (03PS3) 10Physikerwelt: allow for pages with a combination of mathml and tex elements [extensions/Math] (LaTeXML) - 10https://gerrit.wikimedia.org/r/74524 [16:00:29] (03CR) 10Physikerwelt: [C: 032] "However, thanks for fixing the change." [extensions/Math] (LaTeXML) - 10https://gerrit.wikimedia.org/r/74524 (owner: 10Physikerwelt) [16:02:35] <^d> apergos: I dunno if puppet will restart apache, may have to give it a whack. [16:03:10] oh it didn't, I had to do so but robots.txt has content now [16:03:29] we need to unblock google next [16:03:36] or they will never see the robots.txt file :-D [16:05:11] (03CR) 10Physikerwelt: "See" [extensions/Math] (LaTeXML) - 10https://gerrit.wikimedia.org/r/74524 (owner: 10Physikerwelt) [16:06:59] you are faster than me [16:07:03] it took me longer to find the commit [16:08:28] <^d> It was on my dashboard under recently closed ;-) [16:08:56] bah needs rebase :-P [16:09:16] see I remembered mutante merging it but I thought he had written it too so [16:09:48] <^d> Rebased. [16:09:52] <^d> Damn I love that button [16:09:55] yep [16:10:03] best thing since sliced bread [16:10:57] ff hung [16:10:58] sec [16:12:25] and rerunning puppet [16:13:15] so if we still have issues after a little bit, (once I know they've gotten the new robots.txt file), we might consider adding blobdiff or whatever it is to that list [16:43:28] (03CR) 10Frédéric Wang: "OK, so I guess the skipTags should still have "annotation" in the list." [extensions/Math] (LaTeXML) - 10https://gerrit.wikimedia.org/r/74524 (owner: 10Physikerwelt) [16:45:35] (03PS3) 10Manybubbles: Add browser tests. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 [16:46:50] (03CR) 10Manybubbles: [C: 032] "I'm seeing two failure against beta, both of which look to be related to the super long cache on prefix suggestions. I have started a con" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [16:46:52] (03Merged) 10jenkins-bot: Add browser tests. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78118 (owner: 10Manybubbles) [16:51:55] (03CR) 10Cscott: "Good question, I'm grepping enwiki now." [core] - 10https://gerrit.wikimedia.org/r/77988 (owner: 10Cscott) [16:54:09] (03CR) 10Parent5446: [C: 032] Add more parser tests for magic words. [core] - 10https://gerrit.wikimedia.org/r/76493 (owner: 10Umherirrender) [16:56:33] (03Merged) 10jenkins-bot: Add more parser tests for magic words. [core] - 10https://gerrit.wikimedia.org/r/76493 (owner: 10Umherirrender) [16:59:36] (03PS1) 10EBernhardson (WMF): Remove unnecessary round trips to memcache [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78249 [17:02:02] (03CR) 10Aaron Schulz: [C: 032] Have separate timeouts for upload by url ($wgCopyUpload[Async]Timeout) [core] - 10https://gerrit.wikimedia.org/r/67229 (owner: 10Brian Wolff) [17:05:49] (03Merged) 10jenkins-bot: Have separate timeouts for upload by url ($wgCopyUpload[Async]Timeout) [core] - 10https://gerrit.wikimedia.org/r/67229 (owner: 10Brian Wolff) [17:14:16] (03PS1) 10Rjain: Added support for creating annotations by logged out users [extensions/Annotator] - 10https://gerrit.wikimedia.org/r/78251 [17:24:36] (03PS3) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [17:28:18] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] Fix schema [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78219 (owner: 10Matthias Mullie) [17:34:36] (03PS3) 10Nemo bis: Add GENDER support to pagetriage-byline and friend [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/77898 [17:35:12] (03CR) 10Aaron Schulz: [C: 031] (bug 35607) Fix OXML type detection [core] - 10https://gerrit.wikimedia.org/r/70830 (owner: 10VitaliyFilippov) [17:35:57] (03PS4) 10Nemo bis: Add GENDER support to pagetriage-byline and friend [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/77898 [17:36:42] (03PS1) 10Manybubbles: Add three more small test cases. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78253 [17:36:48] (03PS5) 10Nemo bis: Add GENDER support to pagetriage-byline and friend [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/77898 [17:37:39] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78236 (owner: 10Matthias Mullie) [17:39:18] (03CR) 10Se4598: "back to Patch Set 2 then?" [extensions/AbuseFilter] - 10https://gerrit.wikimedia.org/r/71291 (owner: 10Se4598) [17:46:45] (03PS2) 10Demon: Imporve reindex logging. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78241 (owner: 10Manybubbles) [17:46:51] (03CR) 10Demon: [C: 032] Improve reindex logging. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78241 (owner: 10Manybubbles) [17:46:58] (03Merged) 10jenkins-bot: Improve reindex logging. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78241 (owner: 10Manybubbles) [17:48:40] (03CR) 10Demon: [C: 032] "This query is nassttyyy :\" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78237 (owner: 10Manybubbles) [17:48:42] (03Merged) 10jenkins-bot: Use the archive table to find deletes. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78237 (owner: 10Manybubbles) [17:57:23] (03PS1) 10Mwjames: Resolve some formatting issues in tests [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78254 [17:59:27] (03CR) 10Mwjames: [C: 032] Resolve some formatting issues in tests [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78254 (owner: 10Mwjames) [18:00:47] (03Merged) 10jenkins-bot: Resolve some formatting issues in tests [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78254 (owner: 10Mwjames) [18:16:02] ^d: maybe blobdiffs? [18:16:35] <^d> Maybe. mark is setting up some varnish thing we're gonna stuff it behind. [18:16:56] what a great idea, a lot of that stuff can be cached forever [18:17:15] and frankly the whole thing can fit i na varnish cache with lots o room to spare so why not [18:17:27] let google reget as much stuff as it likes [18:17:34] <^d> He'd been talking about a "misc varnish cluster" for random misc services like this to sit behind. [18:17:46] sounds reasonable [18:18:33] I haven't heard back from my google contact yet [18:18:55] did you verify that they got a copy of the robots file since we unblocked them? [18:22:19] <^d> apergos: I thinkkk so. [18:22:23] nm I did: about 15 minutes after, it looks like [18:22:42] so yeah try blobdiff and whats the other stuff he suggested [18:22:43] um [18:23:43] /patch, /blame,and /history, depending on how useful you think it is to have these indexed or not [18:24:04] first time I typed that I didn't leave a leading space and my irc client threw it away as an unknown command >_M [18:24:07] >_< [18:24:09] <^d> Let's do blobdiff. [18:24:23] sunds great [18:24:55] it gets a fair number of em, 20k in today's logs [18:25:23] I was gonna say we'll have a much clearer picture tomorrow but [18:25:41] tomorrow I will have an Internet-free picture (going on a mini vacation, back late monday) [18:27:04] <^d> apergos: https://gerrit.wikimedia.org/r/78256 [18:30:37] just waiting on puppet [18:30:40] * werdna blobs and diffs [18:31:18] scary [18:31:30] so sloooow [18:33:58] ok that's done [18:34:33] and it serves it. I left a message for my google contact [18:34:51] if I hear anything I'll pass it on [18:35:13] (03PS7) 10TheDJ: Math: Enable MathJax's Safe mode [extensions/Math] - 10https://gerrit.wikimedia.org/r/64523 [18:35:43] <^d> I e-mailed upstream to get him to add more caching :) [18:36:09] hey ^d [18:36:18] <^d> ohai! [18:37:34] ^d: is there any way to give out rights to create repos to people based on the prefix of the repo? [18:37:44] (03PS2) 10EBernhardson (WMF): Remove unnecessary round trips to memcache [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78249 [18:37:46] like, 'admins on toollabs can create repos under labs/tools/*' [18:37:47] ? [18:37:49] <^d> No, it's a long-standing request. [18:38:25] ^d: https://gerrit.wikimedia.org/r/#/c/74678/ has been waiting for two weeks, does your previous -2 review still apply? [18:38:25] does upstream care? [18:38:54] <^d> They'd probably merge it if someone wrote good code for it. [18:39:09] <^d> Krenair: I haven't had the time to look at it I'm afraid. [18:39:10] (03PS4) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [18:39:32] * yuvipanda sniggers about GWT and 'good code' [18:39:46] <^d> Oh I see what you did there! [18:40:03] :P [18:40:33] (03PS1) 10Cscott: DO NOT PUSH. This is a test [vagrant] - 10https://gerrit.wikimedia.org/r/78258 [18:41:18] <^d> yuvipanda: So basically, I had this dream of writing an extension for wikitech/labsconsole where you could request a repo, someone comes along and presses "Boom!" and you'd get a repo setup with all the basic stuff like .gitreview, .gitignore in place for you. [18:41:44] <^d> Proper groups & acls. [18:41:50] ^d: yeah [18:41:50] * ^d continues to dream [18:41:55] not to mention a LICENSE [18:41:56] :P [18:42:55] <^d> Or a "I already have code at this other git repo, import it for me" [18:43:00] <^d> Oh man, I was dreaming. [18:43:35] (03CR) 10Yuvipanda: "Code-Review: +2" [vagrant] - 10https://gerrit.wikimedia.org/r/78258 (owner: 10Cscott) [18:43:38] (03Abandoned) 10Cscott: DO NOT PUSH. This is a test [vagrant] - 10https://gerrit.wikimedia.org/r/78258 (owner: 10Cscott) [18:43:44] (03CR) 10Frédéric Wang: "BTW, the Safe options and default values:" [extensions/Math] - 10https://gerrit.wikimedia.org/r/64523 (owner: 10TheDJ) [18:44:26] ^d: hehe, dream on :P [18:44:56] <^d> Oh. Ooh oh oh. [18:45:00] <^d> That import thing. [18:45:06] <^d> Might totally be doable in core gerrit. [18:45:14] write GWT! [18:45:14] :P [18:45:27] (03PS1) 10EBernhardson (WMF): LocalBufferedCache make consistent with BagOStuff multiGet [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78259 [18:45:28] ^d: but I think I finally have a proper gerrit workflow that doesn't drive me mad :) [18:45:38] involves rebase -i to edit older commits [18:45:39] yuvipanda: lies! [18:45:55] but means I dont' have to squash them all the time! [18:45:58] or do -d [18:46:06] <^d> yuvipanda: ohwaitiforgot! [18:46:20] <^d> buck sucks and I haven't been able to build gerrit master for over a week :p [18:46:34] you forgot buck sucks? [18:46:35] pfft [18:46:36] :P [18:46:38] (03PS2) 10EBernhardson (WMF): Reusable external store insertion with fallback [core] - 10https://gerrit.wikimedia.org/r/77979 [18:47:07] (03PS1) 10BryanDavis: Add Puppet role for thumb.php 404 handler. [vagrant] - 10https://gerrit.wikimedia.org/r/78260 [18:47:10] <^d> yuvipanda: No I knew that. I forgot to mention the bit about not being able to use master :p [18:47:23] ^d: :P [18:48:01] ^d: but you aren't really working much on gerrit now, are you? [18:48:05] you're the search dude now [18:48:06] <^d> No. [18:48:11] <^d> But I like to see when they break things. [18:48:44] heh [18:49:48] ^d: also - we have a new debian/ubuntu packager on staff, but he's in language engineering so most people don't know. KartikMistry is the nginx/other-stuff package maintainer for deb/ub, and I guess that's one more person to poke for deb stuff [18:49:52] that's kart_ [18:50:03] not sure if that's useful, but I guess I've been telling everyone since I found out :P [18:50:11] lol [18:50:14] man, people just keep appearing [18:50:18] did you ask his permission :P [18:50:20] I don't know anyone anymore [18:50:30] werdna: are you in HK? [18:50:32] yeah, a werdna (re)appeared recently too, I heard [18:50:34] nay [18:50:41] aw [18:50:41] (03PS3) 10EBernhardson (WMF): Remove unnecessary round trips to memcache [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78249 [18:50:45] I'm in San Francisco of all places [18:50:55] ah, cold-place [18:51:23] it is pretty cold [18:51:24] * yuvipanda makes werdna deploy LQT2 [18:51:30] werdna: have you moved there? [18:51:47] nah [18:51:49] just visiting [18:52:16] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] Remove unnecessary round trips to memcache [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78249 (owner: 10EBernhardson (WMF)) [18:53:05] (03CR) 10EBernhardson (WMF): [C: 032 V: 032] LocalBufferedCache make consistent with BagOStuff multiGet [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78259 (owner: 10EBernhardson (WMF)) [18:56:00] (03CR) 10Umherirrender: "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/57538 (owner: 10Cscott) [18:56:28] hmm, does anybody here love the API? [18:56:41] don't we all? [18:58:10] well, anybody want to talk about it? I want to accept key/value parameters in a parameter and need somebody to convince me either that it's not so bad or that there is a better way [19:01:17] * werdna waves K4-713  [19:01:28] oh hey, werdna. [19:01:42] how's hong kong? are you hong konging? [19:01:49] honging kong? is that a thing? [19:01:51] K4-713 is in Hong Kong? [19:01:56] ha, no. Sadly. [19:02:01] ah me neither [19:02:06] however I am in san francisco [19:02:12] Actually, I'm okay not being there. Lots already happened this year. [19:02:35] nod, I feel bad for missing out but actually I'm happy to be in one place for three weeks. even if it isn't quite home. [19:02:39] werdna: orly. Is the office, like, empty? [19:02:45] completely. [19:02:50] from tech it's me, S, and Erik (B) [19:02:52] that's it. [19:03:20] I'm totally wfh until they have a desk for me on 3. [19:04:13] there are plenty of desks, they just don't belong to anybody in particular [19:04:24] (03PS4) 10Cscott: New parser test: check localization of magic words. [core] - 10https://gerrit.wikimedia.org/r/57538 [19:04:27] (03CR) 10Mattflaschen: "(2 comments)" [extensions/Annotator] - 10https://gerrit.wikimedia.org/r/75645 (owner: 10Rjain) [19:07:54] (03CR) 10Mattflaschen: [C: 032] "At least anyone looking at the log will be able to see what didn't work." [core] - 10https://gerrit.wikimedia.org/r/77490 (owner: 10Ori.livneh) [19:19:20] (03PS2) 10Mwjames: (Bug 52642) Add query duration as meta data [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78223 [19:21:49] (03Merged) 10jenkins-bot: Revert "Pedantic tweak to README" [core] - 10https://gerrit.wikimedia.org/r/77490 (owner: 10Ori.livneh) [19:27:40] mwalker: Just added you as a reviewer for a pretty huge commit (https://gerrit.wikimedia.org/r/#/c/78264/) [19:27:44] (03CR) 10Nemo bis: "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/78020 (owner: 10Bsitu) [19:27:53] It's not done, but I figured I should push it to at least get some feedback on it [19:28:02] kk; I'll take a peak a bit later [19:28:16] Yeah, no rush at all [19:30:13] mwalker: Also, any idea what this error is? tools.wmflabs.org/bookmanagerv2/wiki/Book:The_Interpretation_of_Dreams It happens even with the BookManagerv2 extension disabled, but it's been like this for a few days [19:33:22] (03PS3) 10Mwjames: \SMW\StoreUpdater (compartmentalize ParserData) [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/78154 [19:36:53] GorillaWarfare: looks like a memcache error [19:37:08] hurm [19:37:13] Hmm... No idea how I'd fix that [19:37:25] restart memcache typically :p [19:37:29] but you're on a shared server [19:37:41] since it's part of translate you might try updating that [19:37:49] in case it's a bug that just got introduced [19:39:48] (03PS8) 10Mwjames: \SMW\SimpleDependencyBuilder + \SMW\DependencyInjector [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/76956 [19:41:52] GorillaWarfare: what version of core are you running? [19:42:01] (03PS2) 10Mwjames: Remove addPropertyValue logic from ParserData (further compartmentalize of ParserData) [extensions/SemanticMediaWiki] - 10https://gerrit.wikimedia.org/r/75295 [19:42:02] (03PS14) 10Mwalker: Update Special:Version licensing for extensions [core] - 10https://gerrit.wikimedia.org/r/65299 [19:42:30] (03CR) 10Mwalker: [C: 04-2] "Mid patch; don't review -- had to commit so I didn't lose my place" [core] - 10https://gerrit.wikimedia.org/r/65299 (owner: 10Mwalker) [19:43:22] (03CR) 10Yuvipanda: "Change Submitted" [core] - 10https://gerrit.wikimedia.org/r/65299 (owner: 10Mwalker) [19:44:09] yuvipanda: eeeevil! [19:44:18] mwalker: did it trick you enough? :P [19:44:22] yes [19:44:26] yw [19:44:35] :D [19:44:40] you get a cookie of ++ dark powers [19:44:46] :D :D :D :D [19:45:21] mwalker: now I hope everyone else on that reviewers doesn't kill me [19:45:42] hah [19:46:47] GorillaWarfare: actually; try running maintenance/rebuildLocalisationCache.php [19:47:15] werdna: btw there is still this open for you https://translatewiki.net/wiki/Thread:Support/About_MediaWiki:Action-block [19:54:15] (03CR) 10Mattflaschen: [C: 04-1] "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/74662 (owner: 10MarkTraceur) [20:02:31] mwalker|away: It's 1.21. I tried the rebuild; no go [20:03:41] (03CR) 10Mollywhite: [C: 04-2] "This shouldn't be merged yet." [extensions/BookManagerv2] - 10https://gerrit.wikimedia.org/r/78264 (owner: 10Mollywhite) [20:04:38] (03CR) 10Parent5446: [C: 031] Reusable external store insertion with fallback [core] - 10https://gerrit.wikimedia.org/r/77979 (owner: 10EBernhardson (WMF)) [20:08:55] (03CR) 10Parent5446: [C: 04-1] "(3 comments)" [extensions/Annotator] - 10https://gerrit.wikimedia.org/r/78251 (owner: 10Rjain) [20:11:26] (03CR) 10Umherirrender: [C: 031] New parser test: check localization of magic words. [core] - 10https://gerrit.wikimedia.org/r/57538 (owner: 10Cscott) [20:13:25] (03PS3) 10Umherirrender: Do not parse param of newsectionsummary on new section preview [core] - 10https://gerrit.wikimedia.org/r/73519 [20:32:30] yuvipanda: i've created a monster [20:41:01] (03PS1) 10Manybubbles: Improve suggestion configuration. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78311 [20:42:47] (03PS1) 10Nemo bis: Make PLURAL support on foxway-php-fatal-error-max-execution-time explicit [extensions/Foxway] - 10https://gerrit.wikimedia.org/r/78312 [20:52:46] (03CR) 10Ori.livneh: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/76304 (owner: 10Ori.livneh) [20:56:37] (03PS2) 10Manybubbles: Improve suggestion configuration. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78311 [20:59:41] @seen rillke [20:59:41] MatmaRex: Last time I saw rillke they were leaving the channel #wikimedia-commons at 8/5/2013 8:03:19 PM (3d56m22s ago) [20:59:49] hmph. [21:00:01] any commons admins? hotcat needs fixin'. https://commons.wikimedia.org/wiki/MediaWiki_talk:Gadget-HotCat.js#Autosaving_problems [21:00:08] and i ain't touching that code with a stick [21:01:55] GorillaWarfare: I would send an email to the labs list then; it's going to involve admin intervention and I don't know how that gets coordinated on toollabs [21:02:31] (03PS1) 10EBernhardson (WMF): Remove flow_text table [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78313 [21:02:48] mwalker: Sounds good [21:03:32] might cc Niklas as well in case he has any clue of what to do beyond restarting the cache [21:06:52] actually; thinking a bit more about it -- I'm pretty sure I was reading somewhere that toollabs was going to retire memcache in favour of redis; you might check your local settings to see if you can get it to use redis -- but I'm not sure of the settings you need [21:07:00] so it'll probably still require an email :p [21:07:40] Yaeh, I was just reading that.. [21:08:23] It said memcached went away today, though, and this error has been showing for a few days [21:09:09] (03PS4) 10Ori.livneh: Optionally encode exception log entries as JSON documents [core] - 10https://gerrit.wikimedia.org/r/76304 [21:09:32] (03CR) 10jenkins-bot: [V: 04-1] Optionally encode exception log entries as JSON documents [core] - 10https://gerrit.wikimedia.org/r/76304 (owner: 10Ori.livneh) [21:13:20] (03PS5) 10Ori.livneh: Optionally encode exception log entries as JSON documents [core] - 10https://gerrit.wikimedia.org/r/76304 [21:13:39] speaking of redis, is there a fully configured client which shards across however many servers are running redis? I saw the RedisBagOStuff, but i am looking for a full client that can use the various redis data structures through [21:14:25] a look through mediawiki-config seems to imply any redis usage needs to instantiate their own client and collect their own serverconfig information ? [21:14:47] (03PS1) 10Manybubbles: Fix the suggest test. [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/78316 [21:18:03] ebernhardson: that's probably true; but if you get an answer otherwise I'd be interested [21:18:08] ebernhardson: $pool = RedisConnectionPool::singleton(); $pool->getConnection(); [21:18:22] cool [21:18:36] ebernhardson: they're replicated, though, not sharded [21:18:46] i don't think we do any sharding at the moment [21:18:57] ori-l: awsome [21:20:52] ebernhardson: I don't there ever will be though [21:21:05] maybe one supporting a subset of commands (single key commands) [21:21:23] even redis cluster will avoid those [21:21:49] redis cluster is probably >1yr out stil though [21:21:52] (multi-key commands like rpoplpush) [21:36:35] (03PS1) 10Jqnatividad: Expanded connect string comments; fully implemented Aggregation Framework [extensions/ExternalData] - 10https://gerrit.wikimedia.org/r/78319 [21:38:33] (03Abandoned) 10Jqnatividad: Expanded MongoDB connect string comments; fully implemented MongoDB aggregation framework support [extensions/ExternalData] - 10https://gerrit.wikimedia.org/r/77825 (owner: 10Jqnatividad) [21:39:49] (03CR) 10Yaron Koren: [C: 032 V: 031] Expanded connect string comments; fully implemented Aggregation Framework [extensions/ExternalData] - 10https://gerrit.wikimedia.org/r/78319 (owner: 10Jqnatividad) [21:41:00] (03CR) 10Yaron Koren: [V: 032] Expanded connect string comments; fully implemented Aggregation Framework [extensions/ExternalData] - 10https://gerrit.wikimedia.org/r/78319 (owner: 10Jqnatividad) [21:48:00] (03CR) 10Ori.livneh: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/78213 (owner: 10Ryan Lane) [21:49:01] liangent: what does -{R|...}- do? [21:49:34] liangent: that's the markup which is screwing up links in galleries, but i'm trying to figure out why the bug doesn't affect [[File:...|link=]] as well. [21:52:36] (03PS5) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [22:03:18] (03PS6) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [22:04:25] cscott: looks like variant related stuff [22:05:10] (03CR) 10Ori.livneh: htmlform multiselect chosen: Replace lists [core] - 10https://gerrit.wikimedia.org/r/78214 (owner: 10Ryan Lane) [22:09:37] werdna: yes, the question is: what does that variant syntax actually do, and why is it being generated? [22:10:19] I believe it's there to override automatic language variant conversion [22:10:23] but I may be making shit up. [22:11:13] https://www.mediawiki.org/wiki/LanguageConverter#LanguageConverter [22:14:23] (03CR) 10Matthias Mullie: "(3 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78313 (owner: 10EBernhardson (WMF)) [22:19:22] cscott: TimStarling wrote the language converter, he might have more insight [22:19:58] I didn't start it, although I have done a lot of maintenance on it [22:20:33] I certainly didn't invent that single-letter syntax [22:20:48] 'R' => 'R', // raw content [22:21:22] case 'R': [22:21:22] // if we don't do content convert, still strip the -{}- tags [22:21:22] $this->mRuleDisplay = $rules; [22:24:29] TimStarling: so what triggers the addition of -{R|...}- around external links? It doesn't seem to happen for [[File:...|link=http://...]], only for external links within [22:24:44] (03CR) 10Ori.livneh: [C: 04-1] "(6 comments)" [vagrant] - 10https://gerrit.wikimedia.org/r/78143 (owner: 10BryanDavis) [22:25:19] you're saying VE/parsoid is adding them? [22:30:23] is there a bug number? [22:33:42] (03CR) 10Cscott: "Details in bug 41756 comment 6. Short story: only 419 articles affected by this change." [core] - 10https://gerrit.wikimedia.org/r/77988 (owner: 10Cscott) [22:40:37] (03PS7) 10Werdna: Add API modules to Flow. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78167 [22:41:04] \o/ JS wrappers [22:49:03] TimStarling: bug 52190. it's the php parser, actually. [22:52:01] (03CR) 10BryanDavis: "(6 comments)" [vagrant] - 10https://gerrit.wikimedia.org/r/78143 (owner: 10BryanDavis) [22:54:59] cscott: the -{R}- is not really relevant to that bug [22:55:43] the problem is that DML is running on the gallery source when it shouldn't be [22:56:09] TimStarling: i'd like to understand why it doesn't show up for [[File:...|link=...]]. That is, what is the proper ordering of the DML passes? [22:56:13] actually maybe the link= parameter is just ignored and passed through to DML [22:57:36] if i understood what triggers the -{R}- i thought i could try to construct a [[File:...|link=]] that would trigger the same bug. [22:58:10] the parsing of gallery lines is separate from the parsing of [[File:]] links [22:58:41] TimStarling: yes, i understand. i'm hacking on that bit in Parser.php right now. but i'm trying to understand the process so i can figure out a *correct* fix. [22:59:05] also -- how is -{R|foo}- different from -{foo}- [22:59:36] I don't think it is very different [22:59:41] maybe a few punctuation details [23:00:54] presumably DML is called by recursiveTagParse, the one annotated by a FIXME comment [23:01:34] yeah, the fixme says, "this is different from makeImage" ... but makeImage doesn't call a recursiveTagParse at all. [23:01:41] well, it doesn't have to [23:01:50] makeImage isn't inside an extension tag [23:02:30] I'm setting up the test case locally [23:02:34] which circles back to my question -- how is DML invoked on makeImage? and/or can I do template expansion w/o a full recursiveTagParse [23:04:25] liangent's original patch tried to extract just the caption and pass that into recursiveTagParse, but the "extract just the caption" part of that failed in the face of links and templates with embedded pipes, like: [23:04:25] [23:04:25] File:foobar.jpg|[[File:foobar.jpg|20px|desc|alt=inneralt]]|alt=galleryalt [23:04:25] File:foobar.jpg|{{Test|unamedParam|alt=param}}|alt=galleryalt [23:04:27] [23:06:40] "and/or can I do template expansion w/o a full recursiveTagParse" [23:06:45] yes, you can call replaceVariables() [23:08:11] the function name predates templates :) [23:08:37] hm. what if i replaceVariables() before extracting the caption, then recursiveTagParse once I have the caption? [23:08:56] File:foobar.jpg|[[File:foobar.jpg|20px|desc|alt=inneralt]]|alt=galleryalt is still an issue though [23:09:00] then you would be expanding templates twice [23:09:28] which is probably not the worst of all possible failure modes, but still not quite technically correct [23:10:15] maybe the answer is just to teach enough about parsing to match [[ ]] and {{ }} ? [23:10:37] yeah... [23:10:53] if you call replaceVariables() first, then {{!}} etc. will be treated as parameter separators [23:11:05] which is how it is, I guess, but maybe not intended? [23:12:33] i'll note that [[File:foobar.jpg|[[caption|bar]]]] seems not to parse correctly, so it's not as if is unique in having embedded-pipe issues [23:15:39] no, i stand corrected: [[File:Foobar.jpg|thumb|[[caption|alt=bar]]]] parses correctly. [23:16:14] yeah, I was going to say [23:16:18] I think that is pretty commonly used [23:16:38] (03PS1) 10Werdna: Flow: Strip out the Parsoid boilerplate before saving to DB. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78323 [23:17:03] (03PS1) 10Werdna: Fix up authors: Add self and correct spelling in Matthias' name. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78324 [23:17:13] TimStarling: yeah, in that case we have [[File:Foobar.jpg|thumb|]] by the time makeImage sees it. [23:17:52] which is apparently what is *trying* to do by calling recursiveTagParse. but that's not the right function to do that. [23:18:07] due to line 2018 [23:18:17] $text = $this->replaceExternalLinks( $text ); [23:18:17] $holders->merge( $this->replaceInternalLinks2( $text ) ); [23:20:34] there's a parser test which sets out some of the goals for gallery parsing [23:20:44] yes, it's failing if you set language=zh [23:21:13] (03PS1) 10Legoktm: Rename variable to avoid confusion of what class the object is [extensions/MassMessage] - 10https://gerrit.wikimedia.org/r/78325 [23:21:13] (which exposes bug 52190) [23:22:35] so... what i really want are templates expanded, the links replaced, *then* split parameters, *then* prevent templates from being expanded in the output [23:24:02] replaceInternalLinks2 does not necessarily guard its caption text [23:24:31] you saw but that's not always how it comes out, there is a memory limit which can trigger expansion to actual tags [23:24:54] so I mean it doesn't guard its link text [23:25:06] so parameter splitting doesn't necessarily work after link expansion [23:25:33] i guess that brings me back to trying to understand the -{R|...}- substitution [23:26:15] (03PS2) 10BryanDavis: Extract mediawiki::settings from mediawiki::extension. [vagrant] - 10https://gerrit.wikimedia.org/r/78143 [23:26:56] because if [[File: ...|caption with http://google.com]] can sometimes get expanded to -{R|google.com}- then we can have this same bug with ordinary [[File:]] links [23:27:36] the alone breaks it, right? so I don't know why the -{R|...}- is important [23:27:48] it's the embedded pipe that breaks it [23:28:21] but link= won't be treated as a working link parameter [23:28:25] even if you do split it correctly [23:28:27] since the parameter parsing is for both makeImage and is very simplistic and revolves around StringUtils::explode( '|', ...) [23:28:44] TimStarling: that's why i had to put the link in the caption in the File: case. [23:29:41] but i guess making this break requires $wgAllowExternalImages=true, which is not the case in most places [23:30:10] again, unless there's some other mechanism that is going to trigger the -{R|...}- expansion, and put a pipe where it's not expected [23:30:44] i have this bad habit of trying to understand a bug well enough to fix its root cause and all its as-of-yet-undetected manifestations. ;) [23:31:50] my feeling at this point is that if the external link markup used -{..}- instead of {R|...} then we could fix this at the root cause, without necessarily having to make gallery's parsing more correct. ;-p [23:31:53] I am thinking along the lines of actually parsing with the preprocessor and matching syntax that way [23:32:40] so you want to fix the bug as reported, rather than fix the FIXME comment in renderImageGallery()? [23:33:16] TimStarling: it's one option, yes. [23:33:39] if you look at the syntax from the point of view of preprocessor DOM, it is not too complicated [23:34:27] then, the pipes can only exist in text nodes as children of the root [23:34:39] aha! [[File:foobar.jpg|-{R|rawcaption}-]] breaks [23:34:45] you can expand each node in turn with RECOVER_ORIG, skipping links [23:35:02] yes, that would not work in my scheme [23:35:17] so even the parsing in makeImage is 'broken' as far as languageconverter is concerned [23:35:21] because -{...}- is not implemented as preprocessor syntax [23:35:41] -{...}- is a very inelegant add-on to the parser [23:36:11] we haven't started implementing it in parsoid yet. one of the reasons i'm tackling this particular bug is to get my hands good and dirty before we start working on that. [23:36:15] but it wouldn't be too hard to make the preprocessor annotate it, the same way it does with links [23:36:33] you know the preprocessor is responsible for expanding templates [23:36:55] but it marks up links for the sole purpose of getting correct template DOM [23:36:56] yes. parsoid hooks up templates in a different way. fwiw. [23:37:28] e.g. for parameter splitting in {{ a [[b|c]] }} [23:37:37] (03PS2) 10EBernhardson (WMF): Remove flow_text table [extensions/Flow] - 10https://gerrit.wikimedia.org/r/78313 [23:37:45] it would probably be beneficial for -{}- to be handled in the same way [23:38:11] yes. i think i'm going to add [[File:foobar.jpg|-{R|rawcaption}-]] as a parser test and open a bugzilla for that. for the future. [23:38:15] then {{ a -{b|c}- }} would work in the intuitive way [23:39:10] yes. and if we make gallery shove its text through the preprocessor in the same way, then gallery will stay bug-compatible with [[File:]] [23:39:22] yes [23:41:36] speaking of the preprocessor: https://gerrit.wikimedia.org/r/77988 [23:44:38] yeah, I guess [23:45:08] it's one of those bugs that was carefully conserved through MW 1.12 for backwards compatibility [23:46:13] hmm, actually it wasn't, it was new in 1.12 [23:46:29] it's listed on https://meta.wikimedia.org/wiki/MNPP as a regression [23:48:02] I will look at it, I guess [23:48:16] TimStarling: could I ask you to review https://gerrit.wikimedia.org/r/#/c/76304 when you get a chance? I don't think anybody else would be totally comfortable merging it. [23:49:05] (03CR) 10Tim Starling: "This was listed as a MW 1.12 regression on https://meta.wikimedia.org/wiki/MNPP" [core] - 10https://gerrit.wikimedia.org/r/77988 (owner: 10Cscott) [23:52:36] Anyone around who knows AuthPlugin? Someone wants help debugging an issue involving it in #mediawiki... [23:57:58] (03PS1) 10Cscott: Allow lines empty but for tabs and comments to be ignored. [core] - 10https://gerrit.wikimedia.org/r/78327 [23:59:00] (03CR) 10jenkins-bot: [V: 04-1] Allow lines empty but for tabs and comments to be ignored. [core] - 10https://gerrit.wikimedia.org/r/78327 (owner: 10Cscott)