[00:00:04] (03PS1) 10Krinkle: [WIP] Wrap up legacy javascript stuff [core] - 10https://gerrit.wikimedia.org/r/92457 [00:07:01] (03PS1) 10Werdna: Fix paging bugs caused by 6f99d757 and another rev I cannot find. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92458 [00:08:21] (03PS2) 10Werdna: Fix paging bug 56075 caused by 6f99d757 and another rev I cannot find. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92458 [00:09:41] (03CR) 10Jforrester: [C: 032] "Per discussion with Roan, and Timo's +1." [core] - 10https://gerrit.wikimedia.org/r/86867 (owner: 10Ori.livneh) [00:10:01] OMG [00:10:33] "I felt a great disturbance in the Force, as if millions of voices suddenly cried out in terror, and were suddenly silenced. I fear something terrible has happened." [00:11:19] Silence in the library [00:11:39] Krinkle: Well, you weren't +2'ing. :-P [00:12:14] bd808: it might be nice to schedule some sort of code walkthrough for GWToolset [00:12:49] which incidentally is also the name of a Doctor Who episode. [00:12:53] it's a bit terse and Chris and I saw some strange bits looking at it already [00:12:54] AaronSchulz: Yup. I've been poking dan-nl to get some docs together for that [00:13:02] "Vashta Nerada" like how tennant pronounces that [00:13:11] (03Merged) 10jenkins-bot: Cache ResourceLoader modules in localStorage [core] - 10https://gerrit.wikimedia.org/r/86867 (owner: 10Ori.livneh) [00:14:13] (03CR) 10Brian Wolff: "Hmm, I wonder if instead of just documenting this, if the function should be instead changed to append a "c" to the end of the string in t" [core] - 10https://gerrit.wikimedia.org/r/92368 (owner: 10BryanDavis) [00:14:18] ori-l: So, the commit of doom I referred to earlier: https://gerrit.wikimedia.org/r/#/c/92457/ pushed it up to get a glimpse [00:14:50] hope to land it in 1.22.0 still, but we'll see. [00:15:12] bd808: I mean maybe a method-by-method walkthrough of parts of the code and time for some usage pattern questions [00:15:18] (03PS4) 10CSteipp: Password Expiration (WIP) [core] - 10https://gerrit.wikimedia.org/r/92037 [00:15:24] Krinkle: yay! I'll review tonight (PDT, so 5-6 hours from now) [00:15:31] cool [00:15:35] (03PS1) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92459 [00:16:00] ori-l: There's a couple of judgement calls in there that I'm open for to discuss. e.g. sometimes dummy, sometimes noop, sometimes forward to new, sometimes kept. [00:16:04] (03CR) 10jenkins-bot: [V: 04-1] Password Expiration (WIP) [core] - 10https://gerrit.wikimedia.org/r/92037 (owner: 10CSteipp) [00:16:10] AaronSchulz: That sounds reasonable. I want him to get it documented on wiki somewhere for posterity. [00:16:38] (03CR) 10jenkins-bot: [V: 04-1] installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92459 (owner: 10MarkAHershberger) [00:17:00] bd808: nah, global warming will take care of posterity, no need to worry [00:17:03] AaronSchulz: I have to admit to not having looked at the whole thing with any rigor yet. Mostly I've looked at patches to see if they were totally insane or not. [00:17:24] (03CR) 10Krinkle: "Duplicate of Iab78a60209." [core] - 10https://gerrit.wikimedia.org/r/92459 (owner: 10MarkAHershberger) [00:17:27] (03CR) 10Adamw: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/86350 (owner: 10Yuvipanda) [00:17:53] (03PS20) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [00:17:54] AaronSchulz: Well maybe more to tgr doesn't get stuck maintaining something incomprehensible than future generations [00:17:54] (03PS12) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [00:18:03] s/to/so/ [00:18:17] (03Abandoned) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92459 (owner: 10MarkAHershberger) [00:18:36] (03CR) 10Krinkle: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92037 (owner: 10CSteipp) [00:19:26] hexmode: Use the "Cherry-Pick To" button (if it applies cleanly) and it'll automatically preserve the Change-Id and create a change for a different REL branch from master. [00:19:54] (03PS5) 10CSteipp: Password Expiration (WIP) [core] - 10https://gerrit.wikimedia.org/r/92037 [00:19:58] saved me a fair amount of gerrit/git command-line wrestling (and consequently mistakes from it) [00:20:16] Krinkle: thanks for the reminder [00:20:23] * hexmode headdesks [00:20:44] (03CR) 10TTO: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92457 (owner: 10Krinkle) [00:21:00] hexmode: if it doesn't apply cleanly, well, then you'll need to checkout origin/REL1_22 manually locally and git-cherry-pick hash-of-commit-in-master [00:21:09] AaronSchulz: when would you and Chris have time for a meeting like you propose? This week, next week? Dan is in Amsterdam I think so some advanced notice needs to be given to get a meeting on SF time. [00:22:39] (03PS1) 10Reedy: Update CentralAuth recent change notification code [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92461 [00:24:16] (03PS1) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92462 [00:24:30] bd808: this week, next whatever works [00:24:36] not sure about chris though [00:25:07] (03CR) 10MarkAHershberger: [C: 032] installer: Run the LESS compiler for *.less files [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92462 (owner: 10MarkAHershberger) [00:25:41] AaronSchulz: I'll send Dan an email and see when we can get times to sync up. Thanks for bringing it up. [00:26:17] (03CR) 10jenkins-bot: [V: 04-1] Password Expiration (WIP) [core] - 10https://gerrit.wikimedia.org/r/92037 (owner: 10CSteipp) [00:26:18] (03PS2) 10Krinkle: [WIP] Wrap up legacy javascript stuff [core] - 10https://gerrit.wikimedia.org/r/92457 [00:26:45] csteipp: Any preference of this week or next for a conf call with Dan about the GLAM app? [00:27:35] bd808: Not really. My calendar is pretty up to date [00:28:02] csteipp: Okey doke. I'll start talk to Dan and see what we can work out. [00:28:12] Cool [00:30:10] (03PS1) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92464 [00:30:38] (03PS1) 10Krinkle: mediawiki.js: Fix unformatted code and unclosed HTML tag [core] - 10https://gerrit.wikimedia.org/r/92465 [00:30:56] (03PS6) 10CSteipp: Password Expiration (WIP) [core] - 10https://gerrit.wikimedia.org/r/92037 [00:31:24] (03CR) 10jenkins-bot: [V: 04-1] installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92464 (owner: 10MarkAHershberger) [00:33:59] (03PS1) 10saper: Add uoi_user_registration for PostgreSQL [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 [00:34:02] ori-l: just a fyi, I don't have a phone number of any sort [00:34:02] (03PS1) 10Krinkle: Revert "mediawiki.log: Do not dynamically redefine mw.log.deprecate" [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92467 [00:34:22] (03CR) 10Krinkle: [C: 032] "Not sure why this was backported." [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92467 (owner: 10Krinkle) [00:36:15] YuviPanda: irc! [00:36:22] ori-l: yeah, so far :D [00:36:25] i should get a sim tonight [00:36:26] hopefully [00:36:32] (03PS1) 10Legoktm: [WIP] Allow for global renaming of users [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 [00:37:58] (03CR) 10saper: [C: 031] "Now I can connect to my own OpenID provider as well to Google, register, login as well as use MediaWiki as OpenID provider to my Gerrit i" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [00:40:40] (03PS2) 10Krinkle: mediawiki.js: Fix docucumentation breakage [core] - 10https://gerrit.wikimedia.org/r/92465 [00:41:17] (03Merged) 10jenkins-bot: Revert "mediawiki.log: Do not dynamically redefine mw.log.deprecate" [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92467 (owner: 10Krinkle) [00:41:48] (03CR) 10Krinkle: "See:" [core] - 10https://gerrit.wikimedia.org/r/92465 (owner: 10Krinkle) [00:42:13] hexmode: deja vu? [00:42:45] (03CR) 10MZMcBride: "I agree with the approach being taken here: I think a simpler user interface is better for now and I'm glad we're retaining the ability to" [extensions/OAuth] - 10https://gerrit.wikimedia.org/r/91488 (owner: 10Aaron Schulz) [00:43:06] (03CR) 10Krinkle: "Duplicate in master of Iab78a6020 which is already merged." [core] - 10https://gerrit.wikimedia.org/r/92464 (owner: 10MarkAHershberger) [00:43:14] enough PostgreSQL love for one evening [00:43:14] Krinkle: ugh, yeah, maybe [00:43:45] it was pushed as a change to master, which conflicts since that change already exists in master [00:44:03] also, it has a different change-id, which makes it hard to find the original. [00:44:57] (03CR) 10MZMcBride: "Why not allow self-renames?" [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [00:44:58] need to update .gitreview? [00:46:02] Hey hexmode. [00:46:12] Elsie: ? [00:46:20] https://en.wiktionary.org/wiki/hey#English [00:46:47] I was going for sense four, I suppose. [00:46:59] legoktm: I apparently don't subscribe to the theory that Gerrit comments be used strictly for code review. ;-) [00:48:09] Elsie: of many nicks [00:49:07] <^d> Elsie: #6 would've been better. [00:49:20] Heh. [00:49:34] hexmode: Just trying to hide from you. :-) [00:49:47] hexmode: I had a blocker to 1.22 that I'm not sure is going to get fixed. :-( [00:50:08] Elsie: Bug#? [00:50:10] Probably not much to be done, but I'm still a bit disappointed. [00:50:23] hexmode: https://bugzilla.wikimedia.org/show_bug.cgi?id=424#c37 [00:50:27] <^d> Oh yeah, I need to write up that patch for rel1_22. [00:50:37] even if it doesn't make RC0, it is still possible for 1.22 [00:51:12] Yeah, I can't seem to find anyone (myself included) willing to code it. [00:51:15] Old bug is old. [00:59:27] (03CR) 10MZMcBride: "This seems like a small bug that could probably make it into 1.22." [core] - 10https://gerrit.wikimedia.org/r/39845 (owner: 10Legoktm) [01:00:59] (03PS1) 10Reedy: Update CodeReview RC2UDP code to match core changes [extensions/CodeReview] - 10https://gerrit.wikimedia.org/r/92471 [01:08:12] (03Abandoned) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] - 10https://gerrit.wikimedia.org/r/92464 (owner: 10MarkAHershberger) [01:09:57] (03PS1) 10Chad: Revert "Add a way to redact certain function parameters from exception stack traces" [core] - 10https://gerrit.wikimedia.org/r/92472 [01:13:28] (03CR) 10MZMcBride: "Proposed for reversion in ." [core] - 10https://gerrit.wikimedia.org/r/64450 (owner: 10Alex Monk) [01:14:21] (03CR) 10Jforrester: "Load." [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [01:14:35] <^d> Unfortunately I couldn't use the "revert" button. [01:14:41] <^d> It wasn't clean and had to do it by hand :\ [01:14:52] <^d> Otherwise it would've noted automatically. [01:14:54] <^d> Thanks Elsie [01:15:07] No problem. [01:15:12] I thought it might note it post-merge. [01:15:19] But while I had everything open, I figured I'd just comment. [01:16:56] AWB is listed to have over twice as much code as MediaWiki [01:17:01] (03CR) 10Alex Monk: "Why was this not just a new patch set in I6f8584f2?" [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [01:17:02] But I suspect designer files are polluting htat [01:18:17] Designer files? [01:18:49] The files that make up the forms [01:18:57] It's all code but mostly automatically created/modified [01:19:12] *make up the forms appearence [01:19:31] ohloh is stupid [01:19:33] "Wildcard characters (*) and regular expressions are not supported." [01:19:38] Hrm, I've reached over 10,000 unread e-mails in my Wikipedia folder. [01:19:49] ohLOLh [01:19:52] <^d> Reedy: ohloh's metrics have always been kinda bogus. [01:20:06] You'd hope they could at least count lines of code... [01:20:13] <^d> Counting is hard, yo. [01:21:17] <^d> I heard healthcare.gov has 500 MILLION lines of code ;-) [01:21:30] <^d> Maybe they had Ohloh count too? [01:24:42] (03PS1) 10Reedy: Fix double space in trigger_error() call [core] - 10https://gerrit.wikimedia.org/r/92473 [01:24:56] * Reedy ticks [01:28:12] (03CR) 10Chad: [C: 032] "Remove ALL the spaces" [core] - 10https://gerrit.wikimedia.org/r/92473 (owner: 10Reedy) [01:31:17] (03Merged) 10jenkins-bot: Fix double space in trigger_error() call [core] - 10https://gerrit.wikimedia.org/r/92473 (owner: 10Reedy) [01:40:26] (03PS2) 10Tim Starling: Remove poorly-phrased doc comment and the FIXME which went with it [core] - 10https://gerrit.wikimedia.org/r/91346 [01:40:31] (03CR) 10Tim Starling: [C: 032] Remove poorly-phrased doc comment and the FIXME which went with it [core] - 10https://gerrit.wikimedia.org/r/91346 (owner: 10Tim Starling) [01:42:09] <^d> Reedy: Easy change is easy: https://gerrit.wikimedia.org/r/#/c/92447/ :) [01:43:10] (03CR) 10PleaseStand: [C: 04-1] "(1 comment)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/92447 (owner: 10Chad) [01:44:47] (03PS2) 10Chad: Don't mention Solr since we don't use it [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/92447 [01:45:00] <^d> PleaseStand: Thanks :) [01:45:07] (03CR) 10Tim Starling: [C: 04-1] "What about $clearState = false? Should it be respected, or deprecated? It is meant to allow recursive calls to Parser::parse()." [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [01:45:13] (03Merged) 10jenkins-bot: Remove poorly-phrased doc comment and the FIXME which went with it [core] - 10https://gerrit.wikimedia.org/r/91346 (owner: 10Tim Starling) [01:47:47] (03PS1) 10MarkAHershberger: installer: Run the LESS compiler for *.less files [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92475 [01:49:30] (03CR) 10MarkAHershberger: [C: 032] installer: Run the LESS compiler for *.less files [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92475 (owner: 10MarkAHershberger) [01:52:55] (03Merged) 10jenkins-bot: installer: Run the LESS compiler for *.less files [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92475 (owner: 10MarkAHershberger) [01:55:42] (03CR) 10Tim Starling: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92253 (owner: 10Brian Wolff) [01:55:45] It's scary how many sites still don't seem to have SSL [01:55:58] (03CR) 10Reedy: [C: 032] Don't mention Solr since we don't use it [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/92447 (owner: 10Chad) [01:56:04] (03Merged) 10jenkins-bot: Don't mention Solr since we don't use it [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/92447 (owner: 10Chad) [01:56:17] Reedy: Like reedyboy.net! [01:56:35] True [01:56:42] But I've nothing on it in either [02:08:44] (03CR) 10Parent5446: [C: 04-1] "(2 comments)" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [02:15:15] (03CR) 10Parent5446: "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/92037 (owner: 10CSteipp) [02:21:32] Elsie: I'm not sure. When I wrote that part it was because the user needed to already be logged out at that point, except they couldn't be if they were themselves and blegh it was confusing. [02:23:54] (03CR) 10Legoktm: "The reason I initially didn't allow self-renames was because the user will be forced logged out, which could cause issues if that was the " [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [02:25:47] (03CR) 10Reedy: "Certainly a limit would have to be enforced if we allowed self renames. Both from concurrent renames, and number of renames in X period..." [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [02:26:07] You can't have concurrent renames [02:26:21] Good [02:26:22] Well, not of the same user [02:26:28] User1 -> User2 [02:26:32] User1-> User3 [02:26:38] User1 -> User4 [02:26:39] yeah, that can't happen [02:26:42] Thread safety for CentralAuth :P [02:26:44] GO AWAY STOP RENAMING YOURSELF [02:26:57] legoktm: Can't? Or shouldn't? ;) [02:27:02] Can't. [02:27:09] There's a call to CentralAuthUser::renameInProgress [02:27:40] legoktm: Thanks for taking that one other... you should have joined Chris and me when we did pair code review in HK [02:27:45] * one over [02:27:49] I didn't look at the code [02:28:12] Neither did I, I only know the old one [02:28:23] Architectured summit => CA sucks [02:29:20] But compared to EditPage it's still charming [02:29:47] I'm still in therapy from my last time touching EditPage [02:29:53] Heh. [02:30:17] hoo: Well I talked with Chris about it today and that's kinda a rough skeleton of what we came up with [02:30:33] I still strongly think we should fix the DB schema first. [02:30:36] LocalRenameUserJob isn't in autoloader [02:30:46] From the summary it looks like what we came up with in HK... if it's not, I'm probably going to -1 you at some point [02:30:49] Reedy: [WIP]! [02:30:53] The whole idea of potentially updating millions of rows for a user rename is fucking insane. [02:30:58] heh [02:31:09] if id > 0 name = '' [02:31:17] just use the text field to keep track of anon edits [02:31:22] and add loads of user joins [02:31:22] Exactly. [02:31:24] EVERYWHERE [02:31:26] Not loads. [02:31:41] But those joins are cheap in any case. [02:31:51] (03PS1) 10MarkTraceur: Add "use this file" link and dialog [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 [02:31:57] if id > 0 name is NULL [02:32:01] '' is silly. [02:32:18] We should have a user row for everyone who has log entry/ edit/ whatever [02:32:44] Maybe. [02:32:49] That's another approach to take. [02:32:53] Though I don't like it as much. [02:33:35] I like it the most as you can take an approach very similar to the current one (only on more INNER JOIN) and you keep user stuff central [02:34:38] ori-l: around? [02:35:56] (03CR) 10Legoktm: "@Alex: I uploaded a different patchset because this uses a different approach. Also because gerrit doesn't let you change the owner of a p" [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/92468 (owner: 10Legoktm) [02:42:01] (03Abandoned) 10Alex Monk: Global user renaming [extensions/CentralAuth] - 10https://gerrit.wikimedia.org/r/39171 (owner: 10Alex Monk) [02:54:51] (03PS2) 10Reedy: Update CodeReview RC2UDP code to match core changes [extensions/CodeReview] - 10https://gerrit.wikimedia.org/r/92471 [02:55:46] (03PS2) 10MarkTraceur: Add "use this file" link and dialog [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 [03:45:57] (03PS1) 10Werdna: Fixes for recursive functions. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92480 [04:27:37] (03PS3) 10Jforrester: New icons [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/91040 (owner: 10Esanders) [04:47:01] (03PS1) 10Legoktm: Remove call to deprecated Job::safeBatchInsert [extensions/Renameuser] - 10https://gerrit.wikimedia.org/r/92481 [05:03:32] (03CR) 10Legoktm: [C: 04-1] "https://gerrit.wikimedia.org/r/#/c/92268/ is probably a better idea, so we don't run into this issue in the future." [extensions/Echo] - 10https://gerrit.wikimedia.org/r/76648 (owner: 10EBernhardson) [05:06:44] (03CR) 10Legoktm: [C: 04-1] "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/81157 (owner: 10Kaldari) [05:09:58] (03CR) 10Legoktm: [C: 032] Remove class constant variable reference from Echo.php [extensions/Echo] - 10https://gerrit.wikimedia.org/r/91485 (owner: 10Bsitu) [05:10:07] (03Merged) 10jenkins-bot: Remove class constant variable reference from Echo.php [extensions/Echo] - 10https://gerrit.wikimedia.org/r/91485 (owner: 10Bsitu) [06:07:21] (03PS1) 10Hiong3-eng5: API add_annotations:added test mode;ow_downloads:DfM [extensions/WikiLexicalData] - 10https://gerrit.wikimedia.org/r/92485 [06:26:45] (03PS1) 10PleaseStand: FormatJson: Remove whitespace from empty arrays and objects [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92489 [06:35:24] (03PS1) 10Werdna: Improve JS initialisation function handling. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92490 [06:46:19] (03CR) 10PleaseStand: "Reason for backporting:" [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92489 (owner: 10PleaseStand) [06:51:20] (03PS1) 10Werdna: (bug 56132) Hide hover-over for posts that are being edited [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92492 [07:03:40] (03PS4) 10Legoktm: [WIP] Add Redis backend [extensions/Echo] - 10https://gerrit.wikimedia.org/r/92248 [07:14:37] (03PS1) 10Werdna: Add a Cancel button to topic-level reply [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92496 [07:17:07] (03CR) 10Matthias Mullie: [C: 032] Revert .htaccess permissions back to 0644 [core] - 10https://gerrit.wikimedia.org/r/92429 (owner: 10EBernhardson) [07:21:16] (03Merged) 10jenkins-bot: Revert .htaccess permissions back to 0644 [core] - 10https://gerrit.wikimedia.org/r/92429 (owner: 10EBernhardson) [07:38:32] (03CR) 10PleaseStand: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/76882 (owner: 10PleaseStand) [08:15:20] (03PS5) 10Marco: Adding description how to use Special:Redirect [core] - 10https://gerrit.wikimedia.org/r/86350 (owner: 10Yuvipanda) [08:20:46] looking for a front-end dev with +2 privileges who can review, https://gerrit.wikimedia.org/r/#/c/90041/. marktraceur's initial comments are in patch set 2. tia! [08:43:57] (03CR) 10Matthias Mullie: [C: 032] Fix paging bug 56075 caused by 6f99d757 and another rev I cannot find. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92458 (owner: 10Werdna) [08:44:07] (03Merged) 10jenkins-bot: Fix paging bug 56075 caused by 6f99d757 and another rev I cannot find. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92458 (owner: 10Werdna) [08:51:06] (03PS1) 10Ori.livneh: localStorage: one eval per batch, rather than per module [core] - 10https://gerrit.wikimedia.org/r/92499 [09:12:33] (03PS1) 10Matthias Mullie: Fix issue in recursive functions [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92500 [09:16:10] (03CR) 10Matthias Mullie: [C: 04-1] "(2 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92480 (owner: 10Werdna) [09:22:57] (03CR) 10Mattflaschen: [C: 032] "I'm glad to see some more styles being rolled out to the whole site." [core] - 10https://gerrit.wikimedia.org/r/82828 (owner: 10Bartosz Dziewoński) [09:26:03] (03Merged) 10jenkins-bot: Backport information boxes' styles from vforms to shared CSS [core] - 10https://gerrit.wikimedia.org/r/82828 (owner: 10Bartosz Dziewoński) [09:26:53] (03CR) 10Matthias Mullie: [C: 031] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92437 (owner: 10Werdna) [09:29:38] (03CR) 10Matthias Mullie: [C: 032] Spacing between consecutive collapsed topic cards [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92442 (owner: 10EBernhardson) [09:37:27] (03PS2) 10Hashar: add Tpt, trusted long-time contributor to Wikibase, ProofreadPage, etc. [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92332 (owner: 10Aude) [09:37:33] (03CR) 10Hashar: [C: 032] whitelist Thomas PT (Tpt) [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92332 (owner: 10Aude) [09:37:48] (03Merged) 10jenkins-bot: whitelist Thomas PT (Tpt) [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92332 (owner: 10Aude) [09:38:30] (03CR) 10Hashar: "deployed" [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92332 (owner: 10Aude) [09:39:06] (03CR) 10Ori.livneh: [C: 032] mediawiki.js: Fix docucumentation breakage [core] - 10https://gerrit.wikimedia.org/r/92465 (owner: 10Krinkle) [09:42:07] (03Merged) 10jenkins-bot: mediawiki.js: Fix docucumentation breakage [core] - 10https://gerrit.wikimedia.org/r/92465 (owner: 10Krinkle) [09:43:06] thanks hashar :) [09:57:43] (03CR) 10Hashar: "> I'd rather see a "getRedactedTrace" as a replacement for getTrace, instead of having every caller have to call getTrace and then iterate" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [10:00:17] (03CR) 10Hashar: [C: 04-1] "missing test autoloader entry I guess" [core] - 10https://gerrit.wikimedia.org/r/92328 (owner: 10Hashar) [10:10:46] (03CR) 10MarkAHershberger: "I'd be happy to +2 this if we can get the makefile." [core] - 10https://gerrit.wikimedia.org/r/76882 (owner: 10PleaseStand) [10:11:21] hashar: do you mind if I update your patch? [10:11:29] ori-l: which one ? [10:11:36] https://gerrit.wikimedia.org/r/92334 [10:11:38] the new static in the exception? I am working on it [10:12:00] want me to send it ? [10:12:04] oh, I have something too [10:12:05] I am writing test for it [10:12:07] oh [10:12:16] i'll pastebin the diff [10:12:21] see what you think [10:12:23] I though you were so angry/unhappy that you happily delegated to us :-] [10:12:40] i was pretty unhappy [10:12:48] * hashar sends a draft [10:12:57] (03PS2) 10Hashar: MWExceptionHandler::expandArgumentsInCall [core] - 10https://gerrit.wikimedia.org/r/92334 [10:13:13] ori-l: https://gerrit.wikimedia.org/r/#/c/92334/2/includes/Exception.php [10:13:30] merely took Anomie suggestion on PS8 of your json serialize patch [10:13:50] I think we agreed to get rid of redaction, though. Chad has a patch (https://gerrit.wikimedia.org/r/#/c/92472/) [10:13:54] aka moved most of the redaction logic to a new static MWExceptionHandler::expandArgumentsInCall [10:14:07] ohhh [10:14:18] but actually I think we can reuse quite a lot of Krenair's code [10:14:21] so check this out [10:14:25] I haven't followed that discussion [10:14:41] I guess it happened during the weekly meeting which I left earlier than expected [10:15:00] https://dpaste.de/GAZ3 [10:15:21] basically, with Exception objects you have Exception::getTrace and Exception::getTraceAsString [10:15:53] this patch adds MWExceptionHandler::getRedactedTrace and MWExceptionHandler::getRedactedTraceAsString [10:16:12] which behave the same way as the PHP builtins, except that they replace argument values with argument class/type name [10:16:38] looking for a front-end dev with +2 privileges who can review, https://gerrit.wikimedia.org/r/#/c/90041/. marktraceur's initial comments are in patch set 2. tia! [10:20:55] dan-nl: can't review that sorry :/ [10:21:22] ori-l: that is essentially what I am doing :D [10:21:33] ori-l: probably want to revert the redacted system first [10:22:06] no, because we end up retaining a big part of that patch [10:22:16] np hashar thanks for answering [10:22:40] The output format of Exception::getTraceAsString() isn't configurable [10:22:50] so to implement argument redaction, Krenair had to reproduce the code that generates it in PHP [10:23:19] so that still gets utilized here, except not to do the whole 'REDACT' thing, but instead to transform *all* arguments into their class/type name [10:23:56] my draft at https://gerrit.wikimedia.org/r/#/c/92334/2/includes/Exception.php merely takes krenair redacted stuff to make it output an array [10:24:05] so you can later format it as a string or json [10:24:29] but if we get rid of the redaction system, I guess we can go with your diff [10:24:45] yes, but the array format is your own invention [10:24:51] based on the string format code [10:25:14] whereas in the diff it's the same as you'd get from Exception::getTrace() [10:25:21] ahhh [10:25:24] except the arguments array in each stack level is modified [10:25:55] and in PHP's code, getTraceAsString calls getTrace behind the scenes [10:25:59] so the symmetry is preserved here [10:26:14] i.e., getRedactedTraceAsString calls getRedactedTrace [10:27:02] anyways, i was going to ask for your permission to submit it as an update to your patchset rather than a new patch [10:27:19] yeah go ahead although: you should take care of arrays in args [10:27:36] and I would avoid the formatRedactedTrace -> getRedactedTraceAsString renaming [10:27:48] once you push it, I guess we can abandon Chad revert commit [10:27:59] since your patchset does the same / enhance [10:28:09] what about arrays? [10:28:25] https://dpaste.de/GAZ3#L198 [10:28:31] $arg = is_object( $arg ) ? 'Object(' . get_class( $arg ) . ')' : gettype( $arg ); [10:28:45] in getRedactedTrace() [10:28:52] yes [10:28:56] if $arg = array(); [10:28:56] don't you need to get rid of the array as well ? [10:29:00] the result is 'array' [10:29:26] because is_object( array() ) -> false, gettype( array() ) -> 'array' [10:29:36] oh you gettype [10:29:41] sorry [10:29:56] i can't concentrate, had an horrible evening yesterday [10:30:18] sorry to hear that, everything ok? [10:30:27] ori-l: so yeah go ahead, will amend with some test cases [10:34:44] ori-l: so basically push your patch :D [10:34:54] I will abandon Chad reversion in favor of your [10:35:02] or you can do it [10:35:12] ok, as an update to your patch, yes? [10:35:48] this way i can say that it's your fault if it causes crashes in production [10:37:06] ori-l: You reported problems with mathoid. How did you test that? [10:37:17] on vagrant [10:38:32] with that method https://gerrit.wikimedia.org/r/#/c/90702/ [10:39:54] ori-l: yeah do it [10:40:07] ori-l: it is fine, I can always amend the Author: field before +2 ing :]]] [10:44:13] (03PS3) 10Ori.livneh: MWExceptionHandler::formatArgumentsInCall [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [10:47:25] (03CR) 10Ori.livneh: "I think we should go with I3d570a638 instead. We actually want to preserve the code that generates the exception string in PHP; it ends up" [core] - 10https://gerrit.wikimedia.org/r/92472 (owner: 10Chad) [10:49:21] (03CR) 10Ori.livneh: "Well, generating the string in PHP code ends up being highly useful anyway, so only a partial revert." [core] - 10https://gerrit.wikimedia.org/r/64450 (owner: 10Alex Monk) [10:49:45] (03CR) 10Hashar: [C: 04-1] "We chatted about this earlier with Ori. His point, and I agree, is to keep some part of Alex Monk patch while still reverting the manually" [core] - 10https://gerrit.wikimedia.org/r/92472 (owner: 10Chad) [10:50:33] (03CR) 10Hashar: "Ori follow up is in https://gerrit.wikimedia.org/r/#/c/92334/" [core] - 10https://gerrit.wikimedia.org/r/64450 (owner: 10Alex Monk) [10:56:14] (03CR) 10Werdna: [C: 032] "Better implementation of change 92480." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92500 (owner: 10Matthias Mullie) [10:56:23] (03Merged) 10jenkins-bot: Fix issue in recursive functions [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92500 (owner: 10Matthias Mullie) [10:56:44] (03Abandoned) 10Werdna: Fixes for recursive functions. [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92480 (owner: 10Werdna) [10:59:50] (03CR) 10Addshore: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [11:05:56] (03CR) 10Addshore: [C: 04-1] "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/92168 (owner: 10Legoktm) [11:07:38] (03PS1) 10PleaseStand: Fix problems when upgrading from MediaWiki 1.1 schema [core] - 10https://gerrit.wikimedia.org/r/92502 [11:09:51] (03CR) 10jenkins-bot: [V: 04-1] Fix problems when upgrading from MediaWiki 1.1 schema [core] - 10https://gerrit.wikimedia.org/r/92502 (owner: 10PleaseStand) [11:12:35] (03CR) 10Addshore: [C: 04-1] "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/22167 (owner: 10Parent5446) [11:15:02] (03CR) 10Addshore: [C: 04-1] "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/34677 (owner: 10Parent5446) [11:15:30] (03PS1) 10Matthias Mullie: Give extensions a chance to hide their (unrelated) log entries [core] - 10https://gerrit.wikimedia.org/r/92504 [11:17:46] (03CR) 10Addshore: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:19:00] (03PS2) 10Matthias Mullie: Flow moderation actions in logs [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92346 [11:19:40] (03PS2) 10PleaseStand: Fix problems when upgrading from MediaWiki 1.1 schema [core] - 10https://gerrit.wikimedia.org/r/92502 [11:21:57] (03CR) 10Brian Wolff: "Is there docs somewhere that explains what should and should not have @covers ?" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:26:38] (03CR) 10Addshore: "Not on the mediawiki side :/" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:29:51] (03PS1) 10Matthias Mullie: Make recursive identifiers less opaque [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92505 [11:34:50] (03Merged) 10jenkins-bot: Make recursive identifiers less opaque [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92505 (owner: 10Matthias Mullie) [11:37:20] (03PS3) 10Brian Wolff: Work around weird issue with Parser::mStripState. [core] - 10https://gerrit.wikimedia.org/r/92253 [11:39:15] (03CR) 10Brian Wolff: "Ok, moving to wip. Sorry about jumping the gun on this. I got a little over excited when I managed to get something that made my test case" [core] - 10https://gerrit.wikimedia.org/r/92253 (owner: 10Brian Wolff) [11:41:44] (03PS1) 10Hashar: .gitreview file for integration/phpunit [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92507 [11:41:46] (03PS1) 10Hashar: README and composer.json [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92508 [11:41:48] (03PS1) 10Hashar: phpunit 3.7.28 [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92509 [11:41:55] (03CR) 10Hashar: [C: 032] .gitreview file for integration/phpunit [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92507 (owner: 10Hashar) [11:41:57] (03CR) 10Hashar: [C: 032] phpunit 3.7.28 [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92509 (owner: 10Hashar) [11:41:59] (03CR) 10Hashar: [C: 032] README and composer.json [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92508 (owner: 10Hashar) [11:42:27] (03CR) 10Brian Wolff: "Yeah, there's no specific method doing this. - but this def does not cover parse in its entirety." [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:42:39] (03CR) 10Hashar: [V: 032] README and composer.json [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92508 (owner: 10Hashar) [11:42:41] (03CR) 10Hashar: [V: 032] .gitreview file for integration/phpunit [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92507 (owner: 10Hashar) [11:42:43] (03CR) 10Hashar: [V: 032] phpunit 3.7.28 [integration/phpunit] - 10https://gerrit.wikimedia.org/r/92509 (owner: 10Hashar) [11:43:09] (03CR) 10Gergő Tisza: "File metadata property names are now sanitized to make sure they make valid XML tag names. Property values are not sanitized, but they are" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [11:44:41] (03CR) 10Addshore: "My imediate thought then is that the parse method is doing too much :)" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:57:45] (03CR) 10Addshore: "Infact taking a second look at the change to teh parse method adding the covers tag to parse should be fine as the coverage will only go u" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [11:59:00] (03PS21) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [11:59:01] (03PS14) 10Gergő Tisza: Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [11:59:02] (03PS13) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [12:00:08] (03CR) 10Gergő Tisza: "Rebased." [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [12:00:21] (03CR) 10Gergő Tisza: "Rebased." [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [12:03:22] (03CR) 10Gergő Tisza: "Added cover annotations." [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [12:07:30] (03PS1) 10Hashar: recurse git submodules [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/92511 [12:08:21] (03CR) 10Addshore: [C: 031] recurse git submodules [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/92511 (owner: 10Hashar) [12:09:39] (03CR) 10Hashar: [C: 032] recurse git submodules [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/92511 (owner: 10Hashar) [12:09:58] (03Merged) 10jenkins-bot: recurse git submodules [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/92511 (owner: 10Hashar) [12:10:30] (03CR) 10Brian Wolff: "That's not the case - the test has to call the parser first in order for the exception line to trigger second time around." [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [12:12:38] 757 unreviewed commits, we're going up again [12:14:34] (03CR) 10Addshore: "Yep, thats the best option :)" [core] - 10https://gerrit.wikimedia.org/r/92247 (owner: 10Brian Wolff) [12:19:46] (03PS1) 10FreedomFighterSparrow: Update to work with MW1.21.x & 1.22alpha [extensions/AutoCreateCategoryPages] - 10https://gerrit.wikimedia.org/r/92512 [12:39:16] (03CR) 10Esanders: "(2 comments)" [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 (owner: 10Esanders) [12:39:38] (03PS8) 10Esanders: Compatibility filters for BetaFeatures [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 [12:39:39] (03CR) 10jenkins-bot: [V: 04-1] Compatibility filters for BetaFeatures [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 (owner: 10Esanders) [13:15:07] (03CR) 10Manybubbles: [C: 032] "(1 comment)" [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/91119 (owner: 10Chad) [13:15:13] (03Merged) 10jenkins-bot: General code cleanup [extensions/CirrusSearch] - 10https://gerrit.wikimedia.org/r/91119 (owner: 10Chad) [13:17:48] (03CR) 10Anomie: "Code seems fine, haven't fully tested." [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [13:19:53] (03CR) 10Anomie: ">> I'd rather see a "getRedactedTrace" as a replacement for getTrace, instead of having every caller have to call getTrace and then iterat" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [13:36:39] (03Abandoned) 10Hashar: Revert "Add a way to redact certain function parameters from exception stack traces" [core] - 10https://gerrit.wikimedia.org/r/92472 (owner: 10Chad) [13:37:34] manybubbles: looks like Search suggestions are gone from beta labs again, but they're working in mw.o [13:37:50] chrismcmahon: I'll have a look [13:44:05] (03PS2) 10Manybubbles: Fix ElasticaConnection to not care what wiki we're on [extensions/Elastica] - 10https://gerrit.wikimedia.org/r/91137 (owner: 10Chad) [13:44:11] (03CR) 10Manybubbles: [C: 032 V: 032] Fix ElasticaConnection to not care what wiki we're on [extensions/Elastica] - 10https://gerrit.wikimedia.org/r/91137 (owner: 10Chad) [13:45:17] anomie: around ? [13:45:25] chrismcmahon: merging ^^^^ should do it. Beta should see the update in a few minutes. [13:45:25] hashar: yes [13:45:35] what do you think of https://gerrit.wikimedia.org/r/#/c/92334/ ? [13:45:51] I will amend it to remove some obsolete release note, but I think it is fine to go in [13:46:17] awesome thanks manybubbles. (I <3 my little Search test, I should probably update myself on the current status of your search tests) [13:46:38] chrismcmahon: they are a bit sprawling at this point [13:46:50] hashar: Besides what I said in the code review comments? I wish that wikitech-l thread had more discussion, especially since it seems that it might be under the misapprehension that the idea was to be able to enable $wgShowExceptionDetails. [13:46:56] (03PS4) 10Hashar: redact exception traces and abstract getTrace [core] - 10https://gerrit.wikimedia.org/r/92334 [13:47:20] anomie: and if we want to enable $wgShowExceptionDetails, we want to remove arguments :/ [13:47:22] (03PS1) 10Esanders: Standardise and minimise licence headers [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92525 [13:47:37] (03CR) 10jenkins-bot: [V: 04-1] Standardise and minimise licence headers [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92525 (owner: 10Esanders) [13:47:39] anomie: I can live with it, not sure how helpfull it is to have arguments in the exception log [13:47:43] (03CR) 10PleaseStand: [C: 04-1] "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [13:48:06] (03CR) 10Hashar: "PS4:" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [13:48:27] anomie: another possible problem is that the original exception trace is modified by reference [13:48:43] (03CR) 10PleaseStand: [C: 04-1] "Per comments on PS3." [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [13:49:42] PleaseStand: ++$i; ... [13:50:04] I am not sure how it matter :-] [13:50:19] in C any compiler would optimize that for us already [13:50:45] though for php .. that is unsure :-] Maybe with Zend Cache that is landing with PHP5.5 [13:50:56] hashar: Hmm. Yeah, the mod-by-reference might be an issue. But that should be easy to fix. [13:52:24] Also, "WTF? You can have a class named 'boolean'?" [13:53:43] ori-l: are you around? got a question about a tricky bug that might be something of interest to you. [13:53:48] class boolean { } [13:53:50] $bool = new boolean(); [13:53:51] var_dump(get_class($bool)); [13:53:52] var_dump(gettype(true)); [13:54:00] both would yield: string(7) "boolean" [13:54:00] :-D [13:56:55] (03CR) 10MarkTraceur: [C: 04-1] "What Jenkins said." [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92525 (owner: 10Esanders) [13:59:19] (03CR) 10Hashar: "(4 comments)" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:03:19] (03PS5) 10Hashar: redact exception traces and abstract getTrace [core] - 10https://gerrit.wikimedia.org/r/92334 [14:05:02] (03CR) 10Hashar: [C: 031] "* got rid of the referencing in getRedactedTrace() foreach loops to avoid modifying the original exception when one would expect a get() t" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:05:14] anomie: I am not sure why ori used references to the exception frames :/ [14:05:50] (03PS9) 10Esanders: Compatibility filters for BetaFeatures [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 [14:05:51] (03PS2) 10Esanders: Standardise and minimise licence headers [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92525 [14:06:28] hashar: wouldn't that only change the copy of the array that existed in getRedactedTrace()? [14:06:34] anomie: he must have though of something to redact the exception inplace and then call another method to format it [14:07:02] oh a brion! happy coffee :) [14:08:53] (03CR) 10Brion VIBBER: "Arrays are copied by value in PHP, not by reference; the reference usage in the foreach would change only the copy of the arrays in the ar" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:08:59] brion: arch now I am not sure anymore :/ [14:09:22] got to test it out again [14:09:33] :) [14:09:39] my job is to make you doubt your reality [14:09:41] I should have copy/pasted [14:09:46] THAT's what we call code review \o/ [14:09:58] I doubt there is a reality even if it is often mentionned [14:10:08] it's turtles all the way down [14:12:59] ARGHHH [14:13:05] I need a new career path really [14:13:13] computer engineering is too complicated [14:17:21] brion: so if I understand correctly referencing in foreach is to prevent PHP from doing a copy [14:17:43] I wish it was smart enough to always use a reference and dereference it for us when it is altered [14:18:01] Unless, of course, PHP returns arrays containing array-references from $exception->getTrace(). [14:20:59] no clue how to check that [14:23:18] anomie: https://gist.github.com/hashar/7215603 :D [14:23:40] though that is an anonymous array [14:24:26] hashar: Yeah, right there it looks like it's not returning references. Unless PHP is doing something weird behind the scenes. [14:24:53] * anomie would prefer to just not use references in that function so we don't have to worry about it [14:25:46] is there any chance to make an object behave as if it's a string? [14:25:56] (03CR) 10Hashar: "So I did my tests in the scope of MWExceptionHandler::getRedactedTrace(), double checked and the exception trace remains unchanged." [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:26:00] anomie: I guess so [14:26:17] anomie: part of the reason is optimizing I guess [14:26:33] if so, we can make a class Password or class SensitiveString, and special case for it in backtrace output [14:26:36] liangent: a class can get a __tostring() magic call [14:26:38] hashar: Is that function a critical path though? [14:26:45] liangent: so you could print SensitiveString(); [14:27:22] anomie: I am not sure, but you potentially have huge arguments passed to methods. [14:27:33] (03PS22) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [14:27:34] (03PS14) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [14:27:35] (03PS1) 10Gergő Tisza: Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 [14:27:39] and require any component that fetches a password from *external* environment (eg, webquest) wraps that into such an object immediately [14:27:50] *webrequest [14:31:14] liangent: there is a PHP extension for that , can't remember the name though [14:31:32] liangent: basically a variable coming from $_GET would be tinted [14:31:53] if it ends up being used in a database call, an exception is raised asking for it to be escaped [14:33:26] (03CR) 10Gergő Tisza: "Small language handling bugfix." [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [14:34:00] anomie: so if we don't care of perf we can use the copy as the latest patchset of https://gerrit.wikimedia.org/r/92334 [14:34:20] anomie: or wait for ori to explain is rationale :-] I don't mind using reference, might use a comment though [14:34:35] (03CR) 10Gergő Tisza: "Rebased." [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [14:34:38] hashar: I think we should care about perf there if it proves to be a problem [14:35:58] (03CR) 10PleaseStand: [C: 04-1] "MWExceptionHandler::getRedactedTrace will not work as written in this patch set." [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:36:09] hashar: This is the reason I said you cannot modify the arguments in place like in PS3: https://gist.github.com/plstand/7215695 [14:37:42] ah nice [14:37:59] hah [14:38:09] references within references man [14:38:18] that's some effed up stuff [14:38:32] but I don't think the argument would be changed during the trace formatting so it is probably not an issue [14:38:44] yeah i think "avoid references for sanity" is good enough reason [14:39:55] (03CR) 10Brion VIBBER: "So apparently the argument arrays can contain references if they arguments were passed by reference. Awesome!" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [14:40:21] damn Safari 6.1 on OS X 1.7.x is fully bugged [14:41:09] upgrade to 10.9! it *might* not break [14:41:26] my laptop is 2 years old or so, I am too afraid that will slow it down [14:41:47] i had good experience on my spare laptop (mid-2010 15" mbp) [14:42:09] but [14:42:17] if you have home-brew and such installed you'll have to reinstall everything [14:42:24] cause they changed the standard c++ library and crap [14:43:08] PleaseStand: yeah I am not careful [14:43:30] I have been mostly doing python for the last 3 Months or so :/ [14:51:55] ok writing tests now [14:58:09] (03PS1) 10Gergő Tisza: Make sure CMD doesn't overwrite itself [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/92537 [14:58:57] (03CR) 10MarkTraceur: [C: 04-1] "Still need the full license headers in each file." [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92525 (owner: 10Esanders) [15:01:44] (03CR) 10MarkTraceur: [C: 04-1] "These still look significantly darker than the mockup..." [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/91040 (owner: 10Esanders) [15:07:34] I HATE YOU PHP [15:07:57] hashar: Right back at you [15:08:03] php > $a = true; $a ||= false; [15:08:03] Parse error: parse error in php shell code on line 1 [15:08:21] : looking for a front-end dev with +2 privileges who can review, https://gerrit.wikimedia.org/r/#/c/90041/. marktraceur's initial comments are in patch set 2. tia! [15:08:42] dan-nl: No luck yet? :( [15:08:56] not yet … [15:09:33] dan-nl: Ask MatmaRex or Krenair to hop on it - maybe one of them has +2, but if not, either one of their +1s would very seriously lower my barrier to merging [15:09:44] (03CR) 10Anomie: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [15:13:01] (03PS6) 10Hashar: redact exception traces and abstract getTrace [core] - 10https://gerrit.wikimedia.org/r/92334 [15:14:36] (03CR) 10Hashar: [V: 031] "Fixed the demo code to make it work (aka use foreach( $f as $k => $v );" [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [15:14:45] I am done with that patch for today [15:16:33] thanks marktraceur. Krenair or MatmaRex let me know if you have any questions [15:17:16] (03CR) 10Krinkle: "Oops, sorry about the docucu spelling" [core] - 10https://gerrit.wikimedia.org/r/92465 (owner: 10Krinkle) [15:17:21] dan-nl: I don't think MatmaRex is onlyne :) [15:17:26] online* wow [15:20:13] (03CR) 10Anomie: "(6 comments)" [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [15:21:57] (03CR) 10MarkTraceur: "(4 comments)" [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 (owner: 10Esanders) [15:23:33] (03CR) 10MarkTraceur: [C: 032] "All right!" [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 (owner: 10Esanders) [15:23:43] (03Merged) 10jenkins-bot: Compatibility filters for BetaFeatures [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/88987 (owner: 10Esanders) [15:25:31] (03CR) 10MarkTraceur: "Cleverrrrr" [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/92537 (owner: 10Gergő Tisza) [15:25:51] (03CR) 10MarkTraceur: [C: 032] "Sorry, had a bit of a mouse spasm there. This is what I meant to do." [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/92537 (owner: 10Gergő Tisza) [15:25:54] (03Merged) 10jenkins-bot: Make sure CMD doesn't overwrite itself [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/92537 (owner: 10Gergő Tisza) [15:28:04] (03PS1) 10Sarniakjr: chris_file: substantial edit [test/mediawiki/extensions/examples] - 10https://gerrit.wikimedia.org/r/92540 [15:29:35] (03CR) 10Matthias Mullie: [C: 032] Add a Cancel button to topic-level reply [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92496 (owner: 10Werdna) [15:29:50] (03Merged) 10jenkins-bot: Add a Cancel button to topic-level reply [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92496 (owner: 10Werdna) [15:29:58] (03CR) 10MZMcBride: "(1 comment)" [extensions/TimedMediaHandler] - 10https://gerrit.wikimedia.org/r/88944 (owner: 10Mattflaschen) [15:32:22] (03CR) 10Erjiang: [C: 04-1] "(1 comment)" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [15:32:29] (03CR) 10Matthias Mullie: [C: 04-1] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92492 (owner: 10Werdna) [15:36:00] (03CR) 10Matthias Mullie: [C: 031] "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92490 (owner: 10Werdna) [15:36:03] (03CR) 10Chad: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92004 (owner: 10Chad) [15:36:53] (03CR) 10MarkTraceur: [C: 04-1] "(3 comments)" [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/90919 (owner: 10Gergő Tisza) [15:37:13] (03CR) 10Anomie: [C: 04-1] "(5 comments)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [15:37:18] Turns out I don't actually have much of a threshold for -1ing patches. [15:37:53] (03CR) 10Matthias Mullie: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92454 (owner: 10EBernhardson) [15:38:00] (03CR) 10Anomie: [C: 031] "Code here seems sane. Haven't tested." [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [15:38:13] \o/ [15:38:20] I love it when I'm able to fake sanity [15:39:21] <^d> marktraceur: #everyday [15:39:36] ^d: Wow, I didn't think I was that convincing [15:39:43] I need to wear clown shoes to work more often [15:39:47] i.e. ever [15:41:12] reedy@ubuntu64-web-esxi:~/git/mediawiki/core$ git branch -D wmf/1.22wmf11 [15:41:12] Deleted branch wmf/1.22wmf11 (was d6ed1a2). [15:41:15] nice hash [15:41:30] (03CR) 10Anomie: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [15:49:06] <^d> andre__: I've got an idea for a bug day :) [15:49:13] (03CR) 10Anomie: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [15:49:15] RESOLVED WONTFIX bug 1 [15:49:25] ^d: Share it! [15:49:47] <^d> I've been trying to get through the backlog of old MWSearch/lucene-search-2 bugs. Some are worth turning into enhancement requests for Cirrus. [15:50:05] (03CR) 10Anomie: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [15:50:27] <^d> Others we need to make sure we're taking care of in Cirrus. [15:50:33] <^d> The remainder can all be WONTFIX'd. [15:50:44] <^d> I've gone through about ~20 so far, but there's 66 to go. [15:52:15] wow, ^d I did that once before CirrusSearch was even a glimmer in its pappy's eye, that's probably worth doing again now, +1 [15:52:41] (03CR) 10Krinkle: [C: 031] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh) [15:53:15] ^d, yeah, I saw your cleanup, thanks for that :) [15:53:37] ^d: If we can define a "how-to" on a wikipage how anybody can help and retest those tickets, that would be great [15:54:17] (03CR) 10Krinkle: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh) [15:55:45] (03CR) 10MarkTraceur: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [15:56:08] (03CR) 10MarkTraceur: [C: 04-1] "For easy-state-checking" [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [15:56:24] (03CR) 10Krinkle: [C: 031] FormatJson: Remove whitespace from empty arrays and objects [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92489 (owner: 10PleaseStand) [15:56:35] Krinkle: if you are ready, we can get our checkin now :] [15:56:40] Krinkle: I finished my call with rob [15:56:56] I'll be there in ~ 5min [15:57:21] ^d: Probably that even I myself often have is already understand which package/layer provides what in our Search setup. I know there are wikipages describing it, but still :-/ [15:57:26] * greg-g waits for the ECT meeting to start [15:57:38] greg-g: http://lists.wikimedia.org/pipermail/qa/2013-October/000611.html :-D [15:57:47] greg-g: deployment system is super easy to use :] [15:57:55] ^d: So if we can say "Test it on test2.wikimedia.org and if it does not happen anymore add a comment to the report" or so, that would be cool [15:57:57] hasharCall: :) yeah! I read! [15:58:09] yay ryan-deploy! [15:58:20] huhu [15:58:26] ori-l is going to kick me for saying that [15:58:32] andre__, guillom: ECT meeting starts in a few minutes? [15:58:40] greg-g: not seeing any hangout info for ECT thingie [15:58:55] chrismcmahon: yeah, not yet, waiting for quim to show up [15:59:05] there he is [15:59:28] (03PS1) 10Anomie: UploadStash::removeFileNoAuth shouldn't need auth [core] - 10https://gerrit.wikimedia.org/r/92542 [15:59:51] GET HIIIIMMMM [16:00:09] zeljkof: I think so, I don't have a link yet though [16:00:15] qgil, I need a hangout URL :) [16:00:24] guillom: thanks [16:00:35] (if I got it right and if we're streaming. If this is IRC only, also fine) [16:00:39] (03CR) 10Reedy: [C: 032] UploadStash::removeFileNoAuth shouldn't need auth [core] - 10https://gerrit.wikimedia.org/r/92542 (owner: 10Anomie) [16:00:43] zeljkof: also, this week, it's not a regular meeting, it's a "showcase", meaning it's recorded and available publicly as streaming [16:00:51] (03PS1) 10Reedy: UploadStash::removeFileNoAuth shouldn't need auth [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92543 [16:01:05] andre__: I think you're right. Or I'm wrong too :) [16:01:09] guillom: yes, I understood that, all I need is youtube link :) [16:01:12] greg-g: :D [16:01:15] (03PS1) 10Reedy: UploadStash::removeFileNoAuth shouldn't need auth [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/92544 [16:01:30] this thing of naming projects after the person who wrote it is awkward and funny [16:01:35] (03PS1) 10Reedy: UploadStash::removeFileNoAuth shouldn't need auth [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92545 [16:01:40] now we have yuvi-proxy and ryan-deploy? [16:01:42] All of the cherry picks [16:01:48] and who was this Mr. Cirrus? [16:01:58] Why so...cirrus [16:02:17] Ryan_Lane: We even have Wiki Mark-up [16:02:18] When 1 becomes 4! [16:02:26] It's pandemic [16:02:28] Ryan_Lane: reminds me of my buddy who names his software projects things like "Steve" and "Frank" :) [16:02:36] <^d> Huh? Cirrus? [16:02:49] <^d> Needless pings are needless :p [16:03:04] ^d ^d ^d [16:03:05] Ryan_Lane: https://github.com/willkg "richard" "dennis" "steve" [16:03:11] Hah, willkg. [16:03:21] :) [16:03:24] * ^d stabs marktraceur [16:03:31] * ^d stabs marktraceur a few more times [16:03:52] ^d: You're only eligible for license to stab me if you come into the office [16:04:08] zing [16:04:41] ^d: See, greg-g has an excuse - there's a body of water between him and the office. You have but to march a few blocks. [16:04:48] eek, they're in R36, that's not a good room for this (which might be why it's taking a while for the hangout url) [16:04:49] *maybe* bike. [16:04:56] (03Merged) 10jenkins-bot: UploadStash::removeFileNoAuth shouldn't need auth [core] - 10https://gerrit.wikimedia.org/r/92542 (owner: 10Anomie) [16:04:58] greg-g: No, they changed rooms [16:05:00] 34 I think [16:05:01] oh good [16:05:06] <^d> marktraceur: But...tuesdays... [16:05:13] qgil: around? [16:05:39] (03CR) 10Reedy: [C: 032 V: 032] UploadStash::removeFileNoAuth shouldn't need auth [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/92544 (owner: 10Reedy) [16:05:47] (03CR) 10Reedy: [C: 032 V: 032] UploadStash::removeFileNoAuth shouldn't need auth [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92543 (owner: 10Reedy) [16:05:57] ^d: Not gonna lie, that may be the worst possible reason for WFHing [16:06:08] zeljkof: hangout is being set up [16:06:10] Right, so if people could stop doing stuff in gerrit for this meeting, that'd be great. [16:06:17] (03CR) 10Gergő Tisza: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [16:06:22] hasharCall: calling [16:06:31] guillom: great, thanks [16:06:50] Krinkle: joining [16:07:08] poor hangout [16:07:08] andre__: are you seeing the invite? [16:07:26] <^d> marktraceur: Or the best reason ever ;-) [16:07:35] hasharCall: there's no calendar event, calling your account directly [16:07:37] (03CR) 10Gergő Tisza: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [16:07:59] Hi, the URL of the ECT showcase is http://www.youtube.com/watch?v=Y3Gl-oR2ucU - more info at https://www.mediawiki.org/wiki/Engineering_Community_Team/Meetings#2013-10-29 [16:12:24] watching the youtube stream [16:15:16] who's on the far left of the youtube stream? [16:15:40] no mouse hover over :/ [16:16:47] (03PS10) 10Physikerwelt: Mark Math-specific functions in core as deprecated [core] - 10https://gerrit.wikimedia.org/r/87479 [16:17:15] qgil: Click your screen to force it as primary. [16:17:46] greg-g, it's Alvaro working on Community metrics [16:17:57] andre__: cool, thanks! [16:18:39] protip: what bd808 said. :) [16:19:24] (03PS1) 10Aude: Update Wikibase [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92546 [16:19:38] robla: seeing you and Quim on screen rather than slides [16:20:03] focus follows voice by default [16:20:30] (03PS1) 10Mwalker: Removing Adyen from random selection [core] (fundraising/1.22) - 10https://gerrit.wikimedia.org/r/92547 [16:20:36] bd808: better? [16:20:41] (03CR) 10Mwalker: [C: 032 V: 032] Removing Adyen from random selection [core] (fundraising/1.22) - 10https://gerrit.wikimedia.org/r/92547 (owner: 10Mwalker) [16:20:58] ahhh, there's a big delay, maybe [16:21:14] infinite zoom [16:21:17] better now [16:21:21] thanks robla [16:21:52] there is a big delay...probably 15-20 sec [16:22:14] robla: Yup. That's typical for the youtube recast [16:22:46] robla: it is probably a minute or two [16:23:31] (03PS2) 10Reedy: Update Wikibase [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92546 (owner: 10Aude) [16:23:39] (03CR) 10Reedy: [C: 032 V: 032] Update Wikibase [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92546 (owner: 10Aude) [16:25:50] robla: mute your mics, there's some echo... and now he's gone [16:25:55] live failures are live [16:26:42] muted [16:26:57] thanks, andre__ sounds good [16:27:45] yay inline history :) [16:27:53] guillom: yay! [16:28:02] (not there yet on the livecast ;) ) [16:28:08] heh [16:28:27] there now [16:28:35] greg-g: you're a time traveler! [16:28:44] * greg-g puts on a fedora [16:28:57] (03PS2) 10EBernhardson: Post topic without comment [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92454 [16:29:09] (03CR) 10EBernhardson: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92454 (owner: 10EBernhardson) [16:30:20] oh, nice Fx theme, guillom [16:32:06] Will TechNews use the new MassMessage extension? :) [16:34:03] Great! legoktm would be happy to hear that. [16:34:25] ugh: ^^ [16:34:38] greg-g: He's apparently disgusted with us right now. :P [16:35:06] greg-g: I love Solarized :) I'm rebooting my personal website and building the theme on Solarized colors. [16:35:24] guillom: I was referring to the WP logo :) [16:35:28] marktraceur: well, that Wikitech-l thread probaly didn't help. [16:35:57] greg-g: oh! Ok :) I didn't make the Fx = Firefox connection. [16:36:11] guillom: oh yeah, I'm old school like that. [16:36:11] greg-g: if uses the old logo though. [16:36:58] it* [16:37:20] It's the Debil! [16:37:33] marktraceur: 'ugh' is me hating webchat >.> [16:37:49] guillom, greg-g: yay :D [16:38:51] legoktm_: hey :) if MassMessage is only deployed on testwiki, can it still deliver to all wikis? [16:39:09] no [16:39:11] it needs to be deployed everywhere [16:39:12] Is that a zen question? [16:39:16] ok :/ [16:39:23] And then the sun rose. [16:39:27] from [[Deployments]]: November 14: (Pending review) MassMessage rollout to all wikis (bugzilla:52723, gerrit:91344) [16:40:15] legoktm_: I guess that's an argument for me lobbying for full deployment to happen. [16:40:23] heheh :D [16:40:34] ori-l: Got a minute to discuss https://gerrit.wikimedia.org/r/#/c/90541/ ? [16:40:45] greg-g: is that info available anywhere I can look at it? [16:40:55] Amgine: which? the deploy schedule? [16:41:01] yes, please. [16:41:06] https://wikitech.wikimedia.org/wiki/Deployments [16:41:12] thanks! [16:41:31] on wikitech as it needs to be available if the site goes down (ie: wikitech.wikimedia is hosted on a third-party service, not our cluster) [16:41:53] Krinkle: hey. I just woke up, so what I say may be stupider than usual [16:41:55] but otherwise sure [16:42:53] DGarry: ohai. I have a question for you [16:42:56] ori-l: I understand what you say about the potential to oscillate. Using the hash in the key would mitigate that since the hash is bound to arrive at once (files may arrive in chunks, but the registry does not). [16:43:05] legoktm_: Sure. :) [16:43:11] ori-l: However, I don't understand what you're proposing about the file name where it comes from. How does that help? [16:43:27] legoktm_: In a meeting right now so I may get back to you afterwards. [16:44:21] Krinkle: if all VisualEditor resources were declared in VisualEditor.resources.php (or whatever), then you'd include that file in the max mtime calculation [16:44:22] I lost sound? [16:44:30] I bet they are muted [16:44:37] they were [16:44:39] :) [16:44:41] no, Quim was muted for everybody except for Rob next to him. [16:44:42] heh [16:44:44] had to reload page [16:44:56] whoops [16:44:58] this is fun :) [16:45:00] DGarry: for global renaming, should logs be on all wikis or just meta? Personally I think it just needs to be on meta, log entries on randomwikiwith2peopleonit.wikipedia.org are useless and make it hard to see important real entries, but csteipp said I should check with you [16:45:01] :) [16:45:16] Krinkle: it would mean that you couldn't declare modules from just any file, though [16:46:08] ori-l: Hm.. I'd rather not go that route unless there is no other way. The ability to infer it from data alone seems more stable, though harder to achieve. But I think that sort of investment is worth it. [16:46:10] and you'd need a solution for dynamically-defined modules I guess [16:46:18] yeah, I'm inclined to agree [16:46:59] ori-l: and unfortantely I've already seen a few instances of module definitions being changed or added from .hooks.php [16:47:12] guilty of at least 1 myself. [16:47:38] guilty of killing quilty [16:48:12] yes, it'd be easier in one way (you wouldn't have to touch ResourceLoader much), but harder in another (have to enforce a new convention across the board) [16:48:51] but if you made it opt in (you can continue registering your modules as before, but if you do it in a $wgResourceDefinitionManifest[] = 'MyExtension.resources.php'; you get better cache invalidation behavior [16:49:19] hmm, you know what else you could do [16:49:23] Roan would shoot me for proposing this [16:49:35] ori-l: So, do you think the hash based key would address this problem? If 2 apaches have different registries, each would use one of 2 possible keys (old and new), and the first one to get the new key declares the timestamp in objectcache, then when the other apaches come to the same hash later, they'll find that timestamp in cache and use it [16:49:49] make module registration a function and use debug_backtrace() to identify the caller's file :D [16:50:12] We make an effort to invalidate as little as possible [16:51:10] Krinkle: it could work, but ideally the worst case scenario of having a memcached restart is that things are slower / more overloaded [16:51:44] legoktm, you cannot hardcode metawiki into software [16:51:47] (03CR) 10saper: "(3 comments)" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [16:52:01] ori-l: Hm.. maybe use a different storage for this one? We put module deps cache in the database as well. [16:52:02] huh, I like the use of "■" in that table [16:52:04] Krenair: don't look at EventLogging.php [16:52:24] s/cannot/mustnot [16:52:26] and you shouldn't really assume the wiki farm it's running on has a central wiki to do do this stuff on [16:52:35] Though all cache can purge, even db cache. [16:52:41] so put the log on the wiki which the rename was started from [16:53:07] Krinkle: redis would be nice too, lots of synchronization primitives [16:53:18] (03PS2) 10saper: Add uoi_user_registration for PostgreSQL [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 [16:53:35] but the problem is that yeah, if you lose the hash <-> timestamp association for whatever reason [16:53:39] you can't distinguish newer from older [16:53:53] legoktm__: did you get my messages? [16:54:18] but i think you could probably design it in a manner that makes issues extremely extremely unlikely [16:54:57] (03PS23) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [16:54:58] (03PS2) 10Gergő Tisza: Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 [16:54:59] (03PS15) 10Gergő Tisza: Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [16:55:00] (03PS15) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [16:55:01] (03PS10) 10Gergő Tisza: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [16:55:31] having issues there legoktm______? [16:55:33] [16:52:41] so put the log on the wiki which the rename was started from <-- yes, that's what i meant. [16:55:36] yeah [16:55:44] You said put it on metawiki [16:55:49] >_> [16:55:53] each _ is for every time i've been disconnected before i joined the channel [16:56:05] yeah, sorry i should have been more clear [16:56:10] Krinkle: you could also do something like.... rather than replace each hash with the next, keep the last N hashes in the value [16:56:25] office wifi? [16:56:35] Reedy: i'm at school [16:57:16] Krinkle: though i guess that could screw up rollbacks if you're not careful to update the hash the mtimes [16:59:38] (03Abandoned) 10saper: Really populate ar_len this time [core] - 10https://gerrit.wikimedia.org/r/92443 (owner: 10saper) [16:59:45] (03CR) 10Gergő Tisza: "Fixed." [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [17:00:11] qgil: andre__ guillom: good job [17:00:18] thank you. was good! [17:00:27] qgil, guillom, andre__ : nice meeting. Thanks for sharing [17:00:27] (03CR) 10EBernhardson: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/91206 (owner: 10Matthias Mullie) [17:01:22] (03CR) 10Gergő Tisza: "Fixed." [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [17:02:15] (03CR) 10Ori.livneh: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh) [17:02:28] (03CR) 10MarkTraceur: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [17:02:38] (03CR) 10MarkTraceur: [C: 031] "Anomie, it's all you." [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [17:03:39] looking for a front-end dev with +1 or +2 privileges, who can review, https://gerrit.wikimedia.org/r/#/c/90041/. marktraceur's initial comments are in patch set 2. mark recommended Krinkle, Krenair, or MatmaRex if available; tia! [17:06:53] dan-nl: a start would be to add them to reviewers, please do so :) [17:07:05] Nemo_bis: thanks [17:07:41] (03CR) 10Krinkle: "Yeah, though that's only a problem when using module names in the key. If we also include the hash itself in the key, then it is race cond" [core] - 10https://gerrit.wikimedia.org/r/90541 (owner: 10Krinkle) [17:08:09] dan-nl: also, you're the only member of https://gerrit.wikimedia.org/r/#/admin/groups/334,members hence you have to find one of these folks for +2 https://gerrit.wikimedia.org/r/#/admin/groups/11,members but +1 is free for all [17:08:27] (03PS5) 10Krinkle: resourceloader: Add definition hashing to improve cache invalidation [core] - 10https://gerrit.wikimedia.org/r/90541 [17:08:43] (03PS2) 10Krinkle: resourceloader: Use state "error" instead of "missing" in case of exceptions [core] - 10https://gerrit.wikimedia.org/r/90320 [17:09:42] Nemo_bis thanks again [17:09:47] Krinkle: btw, omg does mw.loader.work need to be refactored [17:10:22] dan-nl: finally, all thisk you can find our if you click the magnifying glass from your patch, then "Access" in the top bar, and start clicking around :) [17:11:01] s/thisk/this s/find our/find out [17:15:00] Nemo_bis: i see, that's clearer now. definitely could use some more debs for this project, but atm, i am indeed the only one ... [17:15:20] (03CR) 10jenkins-bot: [V: 04-1] resourceloader: Use state "error" instead of "missing" in case of exceptions [core] - 10https://gerrit.wikimedia.org/r/90320 (owner: 10Krinkle) [17:15:31] marktraceur, tgr: Do we have a test file with multiple languages in some bit of metadata? [17:16:59] (03CR) 10Anomie: "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [17:17:31] Nemo_bis, any quick thoughts about https://www.mediawiki.org/wiki/Talk:Requests_for_comment/MediaWiki.org_Main_Page_tweaks#OPW_project:_Redesigning_mediawiki.org_homepage.3F ? [17:18:01] anomie: Not to my knowledge, but I think only extended metadata has multiple language requirements, e.g. commons file descriptions [17:18:23] guillom, your quick thoughts about https://www.mediawiki.org/wiki/Talk:Requests_for_comment/MediaWiki.org_Main_Page_tweaks#OPW_project:_Redesigning_mediawiki.org_homepage.3F are also welcome [17:19:14] marktraceur: Hmm. I suppose I can just inject some metadata with the hook. I'll just do that to test this. [17:19:29] Sounds good [17:19:37] * anomie hates merging code that hasn't had at least a basic manual test [17:19:42] Heh [17:20:14] (03CR) 10Krinkle: [C: 04-1] "(2 comments)" [core] - 10https://gerrit.wikimedia.org/r/75511 (owner: 10Helder.wiki) [17:20:19] legoktm: I'd go with log entries on meta, yes. [17:21:07] qgil: nothing from the top of my head; main page redesign is quite outside my typical wiki interests :) [17:21:11] (03CR) 10Parent5446: [C: 04-1] "(1 comment)" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [17:21:43] (03PS24) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [17:21:44] (03PS3) 10Gergő Tisza: Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 [17:21:45] (03PS16) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [17:24:41] (03CR) 10Krinkle: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/72844 (owner: 10MarkTraceur) [17:25:06] Nemo_bis, ok thanks, I just wanted to test the waters with you, since you tend to have strong opinions about website related topics. :) [17:27:37] (03CR) 10Krinkle: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88773 (owner: 10Umherirrender) [17:28:48] (03CR) 10Gergő Tisza: "Oops, just missed the nitpicks. (getLanguage() will work even if merged separately, though.) Everything is sanitized now; it will look ugl" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [17:29:04] qgil: I saw your e-mail about it. I didn't have anything to add, I agree it could need some love. I'd suggest involving someone from the design team as co-mentor, if they're willing. [17:30:34] (03CR) 10saper: "Thanks, do you mean something as ugly as instanceof or is there a nicer way?" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [17:31:10] (03CR) 10Anomie: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [17:32:48] personally, I always think main pages need content analysis more than shiny-looking stuff [17:35:55] (03CR) 10Krinkle: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/88773 (owner: 10Umherirrender) [17:36:39] (03CR) 10Krinkle: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/90569 (owner: 10Bartosz Dziewoński) [17:46:09] (03CR) 10EBernhardson: "(5 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92159 (owner: 10Bsitu) [17:51:11] (03PS25) 10Gergő Tisza: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [17:51:12] (03PS4) 10Gergő Tisza: Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 [17:51:13] (03PS17) 10Gergő Tisza: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [17:55:46] (03PS6) 10Jdlrobson: Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 [17:56:24] (03PS3) 10Krinkle: resourceloader: Use state "error" instead of "missing" in case of exceptions [core] - 10https://gerrit.wikimedia.org/r/90320 [17:57:30] (03PS7) 10Jdlrobson: Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 [17:58:02] (03CR) 10Parent5446: "I think this is how you're supposed to do it:" [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 (owner: 10saper) [17:58:03] (03CR) 10Jdlrobson: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [17:59:21] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [core] - 10https://gerrit.wikimedia.org/r/86350 (owner: 10Yuvipanda) [17:59:41] (03CR) 10Matthias Mullie: [C: 04-1] "(13 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92159 (owner: 10Bsitu) [17:59:43] (03CR) 10Jdlrobson: [C: 031] "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/92181 (owner: 10Bartosz Dziewoński) [18:06:30] (03PS1) 10MarkAHershberger: Typo [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92550 [18:07:05] (03CR) 10MarkAHershberger: [C: 032 V: 032] Typo [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92550 (owner: 10MarkAHershberger) [18:07:36] (03CR) 10Siebrand: [C: 04-1] "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:08:30] siebrand: Since it could be on the current wiki or on a remote repo (or on a wiki that uses the current site *as* a remote repo), I think on-wiki is the best term to encompass the concept [18:10:48] (03CR) 10MarkTraceur: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:12:24] (03PS3) 10saper: Add uoi_user_registration for PostgreSQL [extensions/OpenID] - 10https://gerrit.wikimedia.org/r/92466 [18:12:25] (03CR) 10Siebrand: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:12:34] (03CR) 10MarkTraceur: [C: 031] "I don't see any issues with this; didn't test however." [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [18:13:47] (03CR) 10MarkTraceur: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:14:08] (03Merged) 10jenkins-bot: Typo [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92550 (owner: 10MarkAHershberger) [18:14:44] (03CR) 10Anomie: [C: 032] Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [18:14:47] (03CR) 10Anomie: [C: 032] Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [18:14:48] (03CR) 10Anomie: [C: 032] Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [18:14:52] (03CR) 10Anomie: [C: 032] Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [18:14:59] (03CR) 10Anomie: [C: 032] Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [18:15:21] Holy crap [18:15:28] * marktraceur hi5s tgr [18:15:32] * marktraceur hugs anomie [18:15:57] marktraceur: So let's hope it really works ;) [18:16:12] uuh [18:16:15] I've seen it working very well, so I'm going to go with "yeah" :) [18:17:46] (03CR) 10Siebrand: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:21:49] (03PS3) 10MarkTraceur: Add "use this file" link and dialog [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 [18:21:57] siebrand: Thanks [18:22:49] (03CR) 10Siebrand: "I'm disengaging from this patch set. I really don't like the "another list of licenses" duplication, and the simplification of the license" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/90267 (owner: 10MarkTraceur) [18:23:12] (03Merged) 10jenkins-bot: Add an interface for getting "standard" file metadata. [core] - 10https://gerrit.wikimedia.org/r/81598 (owner: 10Brian Wolff) [18:24:36] (03CR) 10Siebrand: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/90260 (owner: 10Werdna) [18:25:32] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed. Sorry to have to keep on nagging sometimes." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:25:58] (03Merged) 10jenkins-bot: Make FormatMetadata accept RequestContext, instead of hard coding $wgLang. [core] - 10https://gerrit.wikimedia.org/r/78162 (owner: 10Brian Wolff) [18:29:33] (03Merged) 10jenkins-bot: Add "extended" file metadata to API [core] - 10https://gerrit.wikimedia.org/r/78926 (owner: 10Brian Wolff) [18:32:55] (03Merged) 10jenkins-bot: Add extmetadata to the ForeignAPIFile iiprops [core] - 10https://gerrit.wikimedia.org/r/88889 (owner: 10MarkTraceur) [18:33:33] (03Merged) 10jenkins-bot: Add language handling to imageinfo/extmetadata API [core] - 10https://gerrit.wikimedia.org/r/92532 (owner: 10Gergő Tisza) [18:38:58] (03PS17) 10Matthias Mullie: Historical view of a topic [extensions/Flow] - 10https://gerrit.wikimedia.org/r/91206 [18:39:28] (03CR) 10Matthias Mullie: "(3 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/91206 (owner: 10Matthias Mullie) [18:39:49] (03CR) 10Bartosz Dziewoński: "Thank you Matt! I'm going to backport this to 1.22 because the release notes already landed in the 1.22 file (oh well) and because come on" [core] - 10https://gerrit.wikimedia.org/r/82828 (owner: 10Bartosz Dziewoński) [18:40:04] (03PS1) 10Bartosz Dziewoński: Backport information boxes' styles from vforms to shared CSS [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92557 [18:40:44] (03CR) 10Bartosz Dziewoński: [C: 032] "This is too pretty not to backport it." [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92557 (owner: 10Bartosz Dziewoński) [18:40:55] (03CR) 10Matthias Mullie: [C: 04-1] "@todo: fix check to display history link, now that topic first post is optional" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/91206 (owner: 10Matthias Mullie) [18:42:27] (03CR) 10CSteipp: [C: 031] "Security looks fine" [extensions/CheckUser] - 10https://gerrit.wikimedia.org/r/52237 (owner: 10saper) [18:44:23] (03CR) 10MarkTraceur: "Hey, it's helpful, we appreciate it :)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92476 (owner: 10MarkTraceur) [18:44:33] (03CR) 10EBernhardson: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/90260 (owner: 10Werdna) [18:49:41] (03Merged) 10jenkins-bot: Backport information boxes' styles from vforms to shared CSS [core] (REL1_22) - 10https://gerrit.wikimedia.org/r/92557 (owner: 10Bartosz Dziewoński) [19:05:15] (03PS2) 10Jdlrobson: Rewrite vectorTabs and #p-personal with nesting [core] - 10https://gerrit.wikimedia.org/r/90171 [19:18:02] hexmode: what is the final-final, no exceptions, oh-no-you-don't deadline for 1.22 changes? did it already pass? [19:18:29] ori-l: we literally just uploaded 1.22.0rc0 to download.wm [19:18:42] i guess that answers the question :D [19:19:03] (03CR) 10Krinkle: "Now that the standards/IE6 override rules are no longer next to each other, maybe rephrase those comments to be clearer. They don't make m" [core] - 10https://gerrit.wikimedia.org/r/90171 (owner: 10Jdlrobson) [19:19:06] http://dumps.wikimedia.org/mediawiki/1.22/ few minutes ago [19:19:39] (03CR) 10Krinkle: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/90171 (owner: 10Jdlrobson) [19:29:58] spagewmf, ebernhardson: https://gerrit.wikimedia.org/r/#/c/91321/ was just merged [19:30:12] (03Abandoned) 10Bartosz Dziewoński: Update documentation for $join_conds in DatabaseBase::select() [core] - 10https://gerrit.wikimedia.org/r/92301 (owner: 10Bartosz Dziewoński) [19:40:20] (03CR) 10Bartosz Dziewoński: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/90569 (owner: 10Bartosz Dziewoński) [19:41:37] (03CR) 10Bartosz Dziewoński: "(1 comment)" [extensions/Echo] - 10https://gerrit.wikimedia.org/r/92181 (owner: 10Bartosz Dziewoński) [19:52:14] (03CR) 10Bsitu: "(5 comments)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92159 (owner: 10Bsitu) [19:56:06] (03PS1) 10MarkTraceur: Add info and discussion links for BetaFeatures [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/92568 [19:56:29] Huh. [19:56:34] That should probably have failed. [20:02:49] greg-g: thanks for stepping in the the SSL/cert thing [20:12:29] (03PS1) 10Gergő Tisza: Use standard multilang arrays [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/92572 [20:15:59] ori-l: as the Roving Performance Fixer, can I ask your opinion about what I think might be a browser-specific possible race condition? I'm not sure if it's just one extension or if it might be a globalesque thing like ResourceLoader, it's a bit hard to pin down. [20:26:35] (03PS1) 10MarkTraceur: Add info links for Media Viewer [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92573 [20:29:22] chrismcmahon: in a meeting, will ping you shortly [20:29:36] ori-l: sure, thanks [20:30:24] Whenever I see "Roving" used in title.. reminds me of Knight Rider's KARR (Knight Automated Roving Robot). [20:39:55] i still want my talking car [20:40:05] and not that bs siri stuff :P [20:41:46] https://code.google.com/p/gerrit/issues/detail?id=2219 [20:43:19] (03PS6) 10Krinkle: Exception: Clean up html document for error pages [core] - 10https://gerrit.wikimedia.org/r/84991 [20:50:46] (03PS6) 10Krinkle: Rewrite Vector's special.preferences.less with nesting [core] - 10https://gerrit.wikimedia.org/r/87171 (owner: 10Jdlrobson) [20:50:57] (03PS7) 10Krinkle: Rewrite Vector's special.preferences.less with nesting [core] - 10https://gerrit.wikimedia.org/r/87171 (owner: 10Jdlrobson) [20:51:31] (03PS4) 10Umherirrender: Add "wgRelevantUserName" from skin to mw.config [core] - 10https://gerrit.wikimedia.org/r/88773 [20:52:51] (03CR) 10Krinkle: [C: 04-1] "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [20:52:53] (03CR) 10Umherirrender: "Patch Set 4: Fixed release notes and commit message" [core] - 10https://gerrit.wikimedia.org/r/88773 (owner: 10Umherirrender) [20:53:32] jdlrobson: ping [20:53:36] https://gerrit.wikimedia.org/r/#/c/87170/ [20:53:40] "Timo any thoughts on the controversial aspect of this. Was there any reason it was done this way historically? " [20:53:47] what controversial aspect? [20:55:29] (03PS1) 10MaxSem: Update MobileFrontend to wmf22 tip [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/92577 [20:55:46] (03CR) 10MaxSem: [C: 032 V: 032] Update MobileFrontend to wmf22 tip [core] (wmf/1.22wmf22) - 10https://gerrit.wikimedia.org/r/92577 (owner: 10MaxSem) [20:59:04] (03PS1) 10MaxSem: Update MobileFrontend to wmf1 tip [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92578 [20:59:23] (03CR) 10MaxSem: [C: 032 V: 032] Update MobileFrontend to wmf1 tip [core] (wmf/1.23wmf1) - 10https://gerrit.wikimedia.org/r/92578 (owner: 10MaxSem) [21:04:34] (03PS10) 10Umherirrender: Add expensive parser functions {{REVISION*:}} [core] - 10https://gerrit.wikimedia.org/r/76534 [21:05:32] (03CR) 10Umherirrender: "Patch Set 10: Moved to 1.23" [core] - 10https://gerrit.wikimedia.org/r/76534 (owner: 10Umherirrender) [21:05:44] marktraceur: Is there a labs instance with MultimediaViewer setup? [21:05:50] There is! [21:06:01] csteipp: http://multimedia-alpha.wmflabs.org/wiki/Lightbox_demo [21:06:23] It has an outdated version of the license patch, but it's pretty-much-what-we-expect-to-deploy [21:07:31] csteipp: Dare I ask for a projection of how we're doing on the reviews? :) [21:08:15] (03PS8) 10Krinkle: Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [21:11:18] (03PS9) 10Krinkle: Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [21:11:35] (03CR) 10Krinkle: [C: 032] Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [21:17:04] (03Merged) 10jenkins-bot: Move Vector's preferences css out of screen.less [core] - 10https://gerrit.wikimedia.org/r/87170 (owner: 10Jdlrobson) [21:17:19] (03PS8) 10Krinkle: Rewrite Vector's special.preferences.less with LESS nesting [core] - 10https://gerrit.wikimedia.org/r/87171 (owner: 10Jdlrobson) [21:18:09] Krinkle: Do you think you'll have a chance to look at the MultimediaViewer JS before tomorrowish? It's not super necessary, as I know it's working at least, but it would be good to have your opinions on the structure before it comes out of beta [21:18:32] It's going out *as* beta this week [21:18:38] So we have time. [21:25:20] (03CR) 10Krinkle: [C: 032] Rewrite Vector's special.preferences.less with LESS nesting [core] - 10https://gerrit.wikimedia.org/r/87171 (owner: 10Jdlrobson) [21:25:50] marktraceur: Any link, change or module in particular? [21:26:43] Krinkle: Generally the way the thing is built, but the lightbox library (that I wrote and split off as its own "upstream") could probably use some review [21:26:52] https://git.wikimedia.org/tree/mediawiki%2Fextensions%2FMultimediaViewer/e09056850b519de8bc50fbb46fc042e8be2d7123/js [21:27:51] (03CR) 10Esanders: "Fixed the colour. The speech bubble is a very similar shame to comment.svg, but I think the outlined version with lines looks better along" [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/91040 (owner: 10Esanders) [21:28:05] (03PS4) 10Esanders: New icons [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/91040 [21:29:08] (03Merged) 10jenkins-bot: Rewrite Vector's special.preferences.less with LESS nesting [core] - 10https://gerrit.wikimedia.org/r/87171 (owner: 10Jdlrobson) [21:30:10] (03CR) 10Werdna: "(1 comment)" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/92492 (owner: 10Werdna) [21:32:02] (03CR) 10MarkTraceur: [C: 031] "I would lurve to merge this." [extensions/BetaFeatures] - 10https://gerrit.wikimedia.org/r/91040 (owner: 10Esanders) [21:38:35] (03CR) 10Krinkle: "(1 comment)" [core] - 10https://gerrit.wikimedia.org/r/90569 (owner: 10Bartosz Dziewoński) [21:41:29] (03PS1) 10Hashar: mediawiki-core-release in publish pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92585 [21:42:05] (03PS2) 10Hashar: mediawiki-core-release in publish pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92585 [21:42:38] (03CR) 10Hashar: [C: 032] mediawiki-core-release in publish pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92585 (owner: 10Hashar) [21:42:48] (03Merged) 10jenkins-bot: mediawiki-core-release in publish pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/92585 (owner: 10Hashar) [21:44:11] Krinkle: sorry , took a while to connect :D [21:45:39] chrismcmahon: hey, sorry [21:45:40] here now [21:47:41] Krinkle: sorry for the delay with the legacy js patch btw, a bit swamped [21:50:50] ori-l: no problem, and this is a super-low priority, but I'd like to pick your brain about something I've been seeing over time, where some chunks of javascript just never appear. I *think* I see it sometimes in UploadWizard, in Login, in loading the wikitext editor, sometimes loading VisualEditor, but it never happens in Firefox, only Chrome and IEs. the poster child is probably UploadWizard, e.g. https://wmf.ci.cloudbees.com/job/brows [21:52:37] ori-l: in that example, the "Next" button never shows up. here's one for IE10: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_10/262/ let me find one for e.g. an editor... [21:53:58] usually 'never appears' issues are the result of some uncaught exception that prevents the rest of the script from executing [21:54:07] can you see the js console for those sessions? [21:58:21] ori-l: can you follow the links there to the SauceLabs jobs? this for example does show an error https://saucelabs.com/jobs/ef647c907b444799bc585fbe274c6fc4 " [21:58:21] network log: Sending: GET http://commons.wikimedia.beta.wmflabs.org/w/index.php?action=raw&ctype=text/css&title=MediaWiki:CollapsibleTemplates.css javascript error: Uncaught SyntaxError: Unexpected token ILLEGAL" [21:58:56] ori-l: some do, some don't. maybe I should collect the errors I do have. [22:00:22] well, if a user interface is assembled in javascript, and there is a syntax error in the javascript, then the user interface won't be assembled, so i assume there is no mystery for those cases [22:00:39] but if you have one where there doesn't appear to be an uncaught exception, point it out [22:01:01] ori-l: the mystery to me is why Firefox seems immune [22:02:22] heh [22:02:24] trailing commas? [22:02:30] could be [22:02:31] (in object literals?) [22:02:36] brb [22:02:46] chrismcmahon: x = { prop: 'value', otherProp: 'otherValue', }; [22:02:53] note trailing comma after last key/value pair [22:03:12] OK for most browsers, not so IE (<= 8, I think) [22:06:57] ori-l: could be. I was thinking about the possibility of race conditions. UploadWizard is my particular concern because a) I often (but not always) see the "Next" button not load in every browser except FF, but no consistent errors behind it and b) anecdotal evidence that UW is flaky in prod, but no one can ever repro the issue. [22:07:55] ori-l: I should probably put together an airtight history of some particular error condition, I have not yet done that. [22:08:03] yep [22:12:46] ori-l: thanks for lending an ear! I'll isolate some less hand-wavy story to check, but in the meantime if you hear any rumors of browser-specific performance problems, I probably have some history to draw on. [22:15:10] TimStarling: the "$logBuffer = array_pop( $lines );" line seems odd to me...so you want $logBuffer to have the first line of the last batch of newlines that fread() gave? [22:16:38] I want it to have the last line in the batch of lines [22:16:47] because that line is not terminated with \n [22:16:47] (03CR) 10Ori.livneh: "(3 comments)" [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh) [22:16:50] so it hasn't finished yet [22:17:00] so I want to re-add it to the log buffer and wait for the \n to come [22:17:10] (03PS2) 10Ori.livneh: mw.loader.store: do one eval per batch, rather than one per module [core] - 10https://gerrit.wikimedia.org/r/92499 [22:17:27] you are thinking of array_shift [22:19:13] ah, right [22:20:18] ok, on to testing around [22:25:52] (03PS1) 10MaxSem: Refactor diffs [core] - 10https://gerrit.wikimedia.org/r/92588 [22:26:53] (03PS1) 10Chad: Rename wfIsHipHop() to wfIsHHVM() [extensions/GWToolset] - 10https://gerrit.wikimedia.org/r/92589 [22:27:03] (03PS1) 10Chad: Rename wfIsHipHop() to wfIsHHVM() [core] - 10https://gerrit.wikimedia.org/r/92590 [22:30:19] (03CR) 10Bartosz Dziewoński: [C: 031] "Hm, now that you've factored the code out, it looks really weird in the DifferenceEngine class. Maybe it should belong to ChangeTags? Or t" [core] - 10https://gerrit.wikimedia.org/r/92348 (owner: 10Aude) [22:32:44] (03CR) 10MaxSem: "Yeah, DifferenceEngine in next on my list (evil grin)." [core] - 10https://gerrit.wikimedia.org/r/92348 (owner: 10Aude) [22:34:30] (03CR) 10MaxSem: "Err, wrong change:P" [core] - 10https://gerrit.wikimedia.org/r/92348 (owner: 10Aude) [22:35:41] :D [22:36:31] (03PS2) 10Chad: Rename wfIsHipHop() to wfIsHHVM() [core] - 10https://gerrit.wikimedia.org/r/92590 [22:38:22] (03CR) 10Aaron Schulz: [C: 032] Improve logging for wfShellExec() and ignore missing cgroup [core] - 10https://gerrit.wikimedia.org/r/91115 (owner: 10Tim Starling) [22:40:15] thanks Aaron [22:42:19] (03Merged) 10jenkins-bot: Improve logging for wfShellExec() and ignore missing cgroup [core] - 10https://gerrit.wikimedia.org/r/91115 (owner: 10Tim Starling) [22:44:21] (03CR) 10Tim Starling: [C: 032] Remove linkprefix message, add $linkPrefixCharset [core] - 10https://gerrit.wikimedia.org/r/91425 (owner: 10Anomie) [22:44:45] (03CR) 10jenkins-bot: [V: 04-1] Remove linkprefix message, add $linkPrefixCharset [core] - 10https://gerrit.wikimedia.org/r/91425 (owner: 10Anomie) [23:01:54] ^d: can you WONTFIX 30714? [23:03:11] (03CR) 10Siebrand: [C: 04-1] "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92573 (owner: 10MarkTraceur) [23:07:25] (03CR) 10Anomie: "Oh, wonderful. l10n-bot reverted most of Iaa7eaa44 in I7d7dfd1a." [core] - 10https://gerrit.wikimedia.org/r/91425 (owner: 10Anomie) [23:08:02] (03PS3) 10Anomie: Remove linkprefix message, add $linkPrefixCharset [core] - 10https://gerrit.wikimedia.org/r/91425 [23:08:19] (03PS1) 10Aaron Schulz: Various grants management message cleanups [extensions/OAuth] - 10https://gerrit.wikimedia.org/r/92594 [23:16:03] (03PS7) 10Ori.livneh: redact exception traces and abstract getTrace [core] - 10https://gerrit.wikimedia.org/r/92334 (owner: 10Hashar) [23:23:10] (03PS1) 10Catrope: Fix crash in mw.page.watch.updateWatchLink() when title not set [core] - 10https://gerrit.wikimedia.org/r/92596 [23:23:11] (03PS1) 10Krinkle: (DRAFT) stash from review conversation [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92597 [23:24:37] Krinkle: https://gerrit.wikimedia.org/r/92596 [23:24:40] (03CR) 10Krinkle: "(1 comment)" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/92597 (owner: 10Krinkle) [23:25:12] Does anyone know how I can make sure that https://gerrit.wikimedia.org/r/92596 gets picked up / backported in the 1.22 release? [23:26:18] RoanKattouw: By backporting it yourself after it is merged. Possibly leave a comment asking the merger to do so. [23:26:27] and add hexmode as a reviewer [23:26:27] To what branch? [23:26:27] we don't have a good way to ensure it or remember ourselves. [23:26:32] REL1_22 [23:26:45] .. after review, of course, otheriwse the cherry-pick will go stale [23:26:47] Krinkle: Also the reviewer would be you ;) [23:26:55] no kidding :) [23:27:03] (This is actually your code according to git blame [23:27:21] (03CR) 10Catrope: "Needs backporting to 1.22" [core] - 10https://gerrit.wikimedia.org/r/92596 (owner: 10Catrope) [23:27:28] it's been moved a lot, I don't take credit for writing it. [23:27:34] I remember moving it from legacy [23:27:37] Right, yeah [23:28:30] RoanKattouw: When does the (un)watch link not have a title attribute? [23:28:47] i'm not sure the defensiveness is entirely appropiate, depending on what triggered it. [23:29:11] It somehow crashes from within VE while saving according to someone who tried VE on his localhost [23:31:43] Oh, VE is re-using that method? Hm.. interesting. [23:31:52] Yes, in ViewPageTarget [23:32:00] Lines 311 and 316 [23:33:14] but we pass the the regular links, not our own [23:33:24] we select the native skin tabs [23:33:29] those have titles. [23:35:01] I'd prefer to document the method as requiring to be #ca-(un)watch, or look-alikes that have: title attribute, contained within 'li' and use id 'ca-watch'/'ca-unwatch'. because the other two are also requirements for this function. [23:35:16] it isn't a general purpose ajax handler for an element that should (un)watch a page. [23:35:26] I wonder how those end up without a title [23:35:50] RoanKattouw: Were you able to reproduce the error in VE? [23:37:27] I wasn't personally, no [23:37:34] He says he reproduced it by just saving [23:38:24] Maybe he has some other extension or gadget or whatever that modifies the watch link. I agree it's weird [23:38:31] I can save a watched page, unwatched page and unwatched page ticking 'watch this page' and all work as expected for me. [23:38:40] Hm.. [23:39:07] OK, lets do it. Can you modify the patch to keep the value of attr( 'title' ) the first time, though? [23:39:15] What do you mean? [23:39:24] not retreive it from the DOM twice [23:39:28] Like, store it in a variable? [23:39:38] yeah [23:39:59] looks like it is already guarded against being empty, so we don't stick 'undefined' in the title later, so that's good. [23:40:33] Yes [23:40:36] (03PS2) 10Catrope: Fix crash in mw.page.watch.updateWatchLink() when title not set [core] - 10https://gerrit.wikimedia.org/r/92596 [23:40:53] Well it's .match(), right, so that might return null [23:41:24] Oh, crap, don't merge that patchset [23:41:26] Duplicate var name [23:41:54] (03PS3) 10Catrope: Fix crash in mw.page.watch.updateWatchLink() when title not set [core] - 10https://gerrit.wikimedia.org/r/92596 [23:44:20] (03PS1) 10MaxSem: DifferenceEngine minor fixes [core] - 10https://gerrit.wikimedia.org/r/92598 [23:52:35] (03CR) 10Krinkle: [C: 032] mw.loader.store: do one eval per batch, rather than one per module [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh) [23:53:31] Whoa, hey jarry1250_ [23:55:35] (03Merged) 10jenkins-bot: mw.loader.store: do one eval per batch, rather than one per module [core] - 10https://gerrit.wikimedia.org/r/92499 (owner: 10Ori.livneh)