[00:00:35] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf3] Changing a citation in the citation tool inserts a new citation, doesn't replace the old one - 10https://bugzilla.wikimedia.org/64717 (10James Forrester) 5PAT>3RES/FIX [00:06:58] James_F|Away: btw, in case you didn't know, you an set sublime as text editor for commit messages where the git command line pops up an editor. [00:07:43] Sublime has a cli bin that supports 'blocking' and 'new window' mode which opens a new window (not a tab in one of your existing sublime windows) and holds back cli process until you close that new window. kind of like vim/nano [00:09:53] https://github.com/Krinkle/dotfiles/commit/5fa0216620b4d0a4f0683901000b4530c7ecdc8f#diff-4 gitconfig [00:10:08] supported in ST2 as well btw [00:10:41] (requires you have Sublime Text.app/Contents/SharedSupport/bin in your PATH https://github.com/Krinkle/dotfiles/blob/master/hosts/KrinkleMac/modules/setup.sh#L16 [00:13:18] Krinkle: What do we need to do to use GitInfo for VE's version checking? [00:13:39] Nothing [00:14:08] What do you mean exactly? [00:14:23] We already use GitInfo in VEDataModule [00:14:47] Hmm. So it will at some point Just Work™ in Beta Labs? [00:14:55] broken in production because GitInfo is broken in production because of scap and submodules, people are fixing that with the json cache you might have seen [00:15:17] 'Cos the change to use JSON cache was merged and yet it's not running yet, or VE isn't getting it, or something. [00:15:34] Hm.. [00:15:48] Or maybe it will work if I just wait another 10 minutes? :-) [00:15:55] * James_F is impatient. :-) [00:16:53] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression pre-wmf4] There is an extra space appearing in media search result dialog in betalabs - 10https://bugzilla.wikimedia.org/64911 (10ryasmeen) a:5Juliusz Gonera>3Trevor Parscal [00:16:56] If it works on http://en.wikipedia.beta.wmflabs.org/wiki/Special:Version it should work for us [00:17:02] we're not doing anything special [00:17:14] GitINfo works for mw itself because that's a plain git directory [00:17:19] * James_F nods. [00:17:50] Well, http://en.wikipedia.beta.wmflabs.org/wiki/Special:Version doesn't have git hashes on it any more. [00:17:55] s/any more/still/ [00:18:13] its broken for extensions because those 1) those are submodules, and in case of submodules .git is a txt file that serves as a poor mans symlink, and 2) git puts an absolute path in that txt file, which doesnt' exist on apaches, (/usr/local/apache vs. /a/common) [00:18:51] Oh, reading 130498 again it doesn't actually have the scap changes to start populating it yet. [00:18:55] * James_F is a fool. [00:18:59] Never mind. :-) [00:19:15] so at that point there is no way for it to find out how and what. even if we'd shell out instead of reverse-engineering git via the file system git command give an error when done on an apache in an extension direcotry [00:19:16] That's 130560. [00:20:41] zh wikipedia main page rules [00:20:44] https://zh.wikipedia.org/wiki/ [00:20:46] They're doing things right [00:20:57] Indeed. Yay for Liangent. [00:27:10] Krinkle: How do I get the CLI for Sublime/ [00:27:33] Krinkle: (requires you have Sublime Text.app/Contents/SharedSupport/bin in your PATH https://github.com/Krinkle/dotfiles/blob/master/hosts/KrinkleMac/modules/setup.sh#L16 [00:27:55] s/2/3/ if needed, both support it [00:27:57] Aha. [00:27:59] Thanks. [00:28:29] It also takes a directory (e.g. I do "subl ." a lot when working on extensions I don't want to permanently add to my workspace) [00:28:42] Nice! [00:28:58] That's easily one of the things I type all day long when checking out random patches from gerrit [00:29:23] like "open ." but for Sublime instead of Finder [00:29:34] * James_F nods. [00:31:50] 3VisualEditor / 3Editing Tools: VisualEditor: Citation tool adds stray line break at the end of the template - 10https://bugzilla.wikimedia.org/63128#c3 (10James Forrester) 5ASS>3RES/FIX This may possibly have been fixed? Seems to now work correctly (on one line) for me, see e.g. http://en.wikipedia.bet... [00:33:35] 3VisualEditor / 3Editing Tools: VisualEditor: Citation tool opens if you click the icon, but if you double-click the ref or press the return key, the Reference dialog opens instead - 10https://bugzilla.wikimedia.org/64712#c2 (10James Forrester) In talking with Trevor this is going to take quite some time to... [00:45:26] James_F, do you have Firefox installed and a minute to check something for me? [00:46:18] I can't get [[Candy]] to open in VisualEditor in Firefox. It pops right open in Safari, but not Firefox. So I thought it might be some odd template or something. [00:46:51] I copied it to my sandbox and started stripping stuff out. It's down to three sentences and one section heading… and it still won't open. [00:47:00] So could you click https://en.wikipedia.org/wiki/User:WhatamIdoing/Sandbox?veaction=edit and tell me if it works for you? [00:47:17] (I can open [[Special:Random]] in Firefox, not just this page.) [00:47:28] whatami: Hey. [00:48:10] whatami: Both https://en.wikipedia.org/wiki/Candy?veaction=edit and https://en.wikipedia.org/wiki/User:WhatamIdoing/Sandbox?veaction=edit work for me in Firefox. [00:48:26] Thanks. It must be time to restart Firefox or something. [00:48:38] I've spent ten minutes trying to get it to open on either page with no success. [00:48:48] Hmm. Very odd. :-( [00:49:36] Maybe it'll work after a restart. [00:49:44] Thanks for letting me know that it's just me. [00:49:47] TTYL [00:51:52] 3VisualEditor / 3ContentEditable: VisualEditor: Support magic pasting of HTML that contains a reference to a wiki image - 10https://bugzilla.wikimedia.org/64935 (10James Forrester) 3ASS p:3Unprio s:3enhanc a:3Ed Sanders Right now if you have some rich content with an image and paste it into VisualEdi... [00:52:05] 3VisualEditor / 3ContentEditable: VisualEditor: Support magic pasting of HTML that contains a reference to a wiki image - 10https://bugzilla.wikimedia.org/64935 (10James Forrester) p:5Unprio>3High [01:08:05] 3VisualEditor / 3ContentEditable: VisualEditor: Support magic pasting of HTML that contains a reference to a wiki image - 10https://bugzilla.wikimedia.org/64935#c1 (10Tisza Gergő) Granted, MediaViewer's HTML format is total crap, but there isn't a clean, semantic way to mark up image captions that works in H... [01:18:50] 3VisualEditor / 3ContentEditable: VisualEditor: Support magic pasting of HTML that contains a reference to a wiki image - 10https://bugzilla.wikimedia.org/64935#c2 (10James Forrester) (In reply to Tisza Gergő from comment #1) > At any rate, whether it is in a figure tag figure or something else, users > woul... [01:49:05] RoanKattouw: https://gerrit.wikimedia.org/r/#/c/131243/ and https://gerrit.wikimedia.org/r/#/c/131071/ [02:42:46] (03PS4) 10Catrope: ve.init.mw.ViewPageTarget: Actually animate the toolbar during setup [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131071 (https://bugzilla.wikimedia.org/64751) (owner: 10Krinkle) [02:43:01] (03CR) 10Catrope: [C: 032] ve.init.mw.ViewPageTarget: Actually animate the toolbar during setup [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131071 (https://bugzilla.wikimedia.org/64751) (owner: 10Krinkle) [02:43:39] (03PS5) 10Catrope: ve.init.mw.ViewPageTarget: Actually animate the toolbar during setup [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131071 (https://bugzilla.wikimedia.org/64751) (owner: 10Krinkle) [02:43:45] (03CR) 10Catrope: [C: 032] ve.init.mw.ViewPageTarget: Actually animate the toolbar during setup [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131071 (https://bugzilla.wikimedia.org/64751) (owner: 10Krinkle) [02:45:30] (03Merged) 10jenkins-bot: ve.init.mw.ViewPageTarget: Actually animate the toolbar during setup [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131071 (https://bugzilla.wikimedia.org/64751) (owner: 10Krinkle) [02:46:05] 3VisualEditor / 3MediaWiki integration: VisualEditor: Transition between read mode and edit mode should have minimal visual shifts - 10https://bugzilla.wikimedia.org/64751#c5 (10James Forrester) 5PAT>3RES/FIX Enough now that I'm going to call this done. [02:47:13] (03CR) 10Catrope: [C: 04-1] "Needs deduplication between MobileContext and MobileSurface CSS, otherwise good to go." (031 comment) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131243 (owner: 10JGonera) [03:52:01] (03CR) 10Catrope: [C: 031] "LGTM, not merging because it's marked as WIP and the schema definition stuff presumably hasn't been done yet." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130543 (owner: 10GWicke) [09:44:42] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130328 (https://bugzilla.wikimedia.org/61073) (owner: 10Alex Monk) [10:10:48] (03CR) 10Ori.livneh: "I wouldn't use EventLogging here: the only thing you're doing is logging a timing measurement with a key from PHP, and MediaWiki's own fac" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130543 (owner: 10GWicke) [10:29:37] (03CR) 10GWicke: "Can you sketch tho pros / cons of EventLogging vs. ProfilerSimpleUDP? Is it just that one is built in while the other relies on the extens" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130543 (owner: 10GWicke) [11:07:21] 3VisualEditor / 3Editing Tools: VisualEditor: ve.ui.MWMediaSearchWidget fails on private wikis due to forced use of JSON-P (logged-out API) - 10https://bugzilla.wikimedia.org/64822#c2 (10Barry Coughlan) The crossDomain logic will switch between the 'script' and 'xhr' transports, but both requests will still... [11:50:16] James_F|Away: Krinkle|detached: what happened to zh wikipedia main page? [11:50:19] (03PS1) 10Esanders: DebugBar tidy up [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131683 [11:50:52] (03CR) 10jenkins-bot: [V: 04-1] DebugBar tidy up [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131683 (owner: 10Esanders) [12:02:37] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130394 (https://bugzilla.wikimedia.org/50955) (owner: 10Alex Monk) [12:16:56] looks like latest visual editor does not work with mw 1.22.6 anymore [12:17:00] "PHP Fatal error: Call to a member function getPageLanguage() on a non-object in includes/OutputPage.php" [12:17:04] is that... ok ? :) [12:25:09] (03PS1) 10Esanders: Cache references in handleDelete [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131687 [12:25:11] (03PS1) 10Esanders: Special case for deleting the entire document [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131688 (https://bugzilla.wikimedia.org/64917) [13:53:05] (03CR) 10Siebrand: "i18n/L10n reviewed." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/123093 (https://bugzilla.wikimedia.org/51260) (owner: 10Alex Monk) [14:56:47] liangent: I think Krinkle|detached was talking about the layout with
s not s. :-) [14:56:49] Richlv: Hmm. [14:57:04] Richlv: VE master needs MW 1.23, yes – but it should also fail gracefully when you try to start it, rather than have random fatal errors. :-( [15:28:22] James_F, it did work a couple of weeks ago, though [15:29:03] James_F, in any case, should i report that fatal error ? [15:30:19] Richlv: Yeah, https://gerrit.wikimedia.org/r/#/c/129567/ bumped wfUseMW to 1.23. [15:30:24] Richlv: If you could that'd be great, thanks [15:36:28] James_F, dang, can't help with further testing then much (regarding 1.23 req) - will report the fatal error, though [15:36:42] Richlv: :-( [15:37:01] Richlv: There was a breaking change in MW 1.23 that we needed to support, sadly. [15:40:43] "The 1.23.0 stable release is expected to come out somewhere in 2014." - heh, not too specific :) [15:41:48] ah, schedule says "2014-05-29" [15:41:58] so i might get a working ve in one month [15:42:02] thanks :) [15:48:28] (03PS1) 10Esanders: Unset fixed height from slideDown in SA demo [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131711 [15:48:53] Richlv: There's a "0" release candidate out already; not sure when rc1 is going to come out. [17:31:59] (03PS1) 10Cmcmahon: [browser test] Refactor markup test to use API [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131738 [17:33:49] (03PS2) 10Cmcmahon: [browser test] Refactor markup test to use API [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131738 [17:58:18] Krinkle: Fun tea party? :-) [18:21:04] James_F: liangent: I was referring to zh wiki main page not being ugly. [18:21:45] James_F: liangent: Proper colours, not a circus of borders and columns, mostly responsive/fluid, calm, simple. And the header integrates nicely. [18:21:52] It doesnt look like an article. [18:22:39] Indeed. [18:24:35] https://nl.wikipedia.org/wiki/Main_Page [18:24:40] Also not an article [18:24:46] but even worse than enwiki if you ask me [18:27:27] (03PS1) 10Esanders: Dialog demos [oojs/ui] - 10https://gerrit.wikimedia.org/r/131752 [18:29:09] edsanders: You seen https://gerrit.wikimedia.org/r/#/c/109006/ already? [18:30:11] That's getting old though. We should merge 'em. [18:30:23] so, the tricky little thing done in LinkInspector which skips over AnnotationInspector's setup and teardown in certain cases (when editing link nodes) has gotta change because I'm implementing setup/teardown processes and this skipping over parent method thing make a mess of it [18:30:23] Or abandon it. [18:30:31] :( [18:30:34] WIPs for three months… [18:30:47] TrevorParscal: Fun. [18:30:56] i know, right?! [18:31:14] quite, mine actually works so.... [18:31:22] I probably want too much out of the dialog demo. My specific use case / intent with it is to see what the different layouts look like, not really the dialogs themselves. [18:31:34] That should be a separate demo (layouts.html0 [18:31:47] this is just to make sure they open and their options work [18:31:59] I'll take my bits and put them into this one assuming this one is minimal and actually working (unlike mine which was still a WIP) [18:32:09] (cf recent breaking of footless dialogs) [18:32:09] Easier to iterate on [18:32:15] Yeah [18:32:32] edsanders: Any thoughts about writing a browser test for the OOjs UI demos? [18:32:35] (03PS13) 10Krinkle: [WIP] Add demo for dialogs and layouts [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 [18:32:52] edsanders: Something that automatically confirms that, indeed, text appears and buttons work? [18:33:04] (03CR) 10jenkins-bot: [V: 04-1] [WIP] Add demo for dialogs and layouts [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 (owner: 10Krinkle) [18:33:41] Krinkle: Ha, new JSCS errors since your last. :-) [18:34:14] (03PS14) 10Krinkle: demo: Make demo console more generic [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 [18:34:47] (03PS1) 10Esanders: Minor refactor of widgets demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 [18:34:49] (03CR) 10Krinkle: "Took the 'dialogs' part of it out. Keeping only the improvements to the demo framework in general. This will allow Ifffd31225dace65c to ma" [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 (owner: 10Krinkle) [18:34:54] edsanders: ^ [18:35:47] Hm.. [18:36:07] Where did the consoles go? [18:36:08] (03CR) 10Esanders: [C: 032] demo: Make demo console more generic [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 (owner: 10Krinkle) [18:36:18] Thx [18:36:28] Strange though, that line got commented out in widgets.js [18:36:40] (03Merged) 10jenkins-bot: demo: Make demo console more generic [oojs/ui] - 10https://gerrit.wikimedia.org/r/109006 (owner: 10Krinkle) [18:36:51] Krinkle: regarding being responsive/fluid, I guess it's possible to make it turn to be one-column and hide some parts in the header if the screen is too narrow, instead of specifying a min-width, as what it does now [18:37:10] liangent: I'm not complaining, I'm saying it's great. [18:37:26] Sure, it could be even more fluid, but the point is, that it's nice. [18:40:36] Krinkle: and I'm talking about improvements :) [18:41:02] * Krinkle is fixing widgets demo to use consoles again [18:41:08] Why was that just commented out? [18:41:20] Sounds like Trevor :P [18:41:24] Krinkle: You broke it IIRC. [18:41:37] Krinkle: And you said you didn't have time to fix it, so it was commented out by ?Roan. [18:42:12] https://github.com/wikimedia/oojs-ui/commit/424b40373e4731585cd99adccfd765c869256755 [18:42:30] No, it was changed from separate layouts to be one large fieldlayout. And it was not migrated along with the rest [18:42:34] Why would I have to fix it? It wasn't broken. [18:42:48] If you change one part, stands to reason to fix the other part as well. We migrated the labels as well. [18:43:06] * James_F shrugs. [18:49:21] (03PS1) 10Krinkle: demo/widgets: Bring dev console back [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 [18:49:46] (03CR) 10jenkins-bot: [V: 04-1] demo/widgets: Bring dev console back [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 (owner: 10Krinkle) [18:52:14] (03CR) 10Krinkle: "Screenshot: http://imgur.com/3qiZEEp" [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 (owner: 10Krinkle) [18:53:23] Krinkle: Neat. [18:53:59] (03PS2) 10Krinkle: demo/widgets: Bring dev console back [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 [19:08:29] ryasmeen: https://gerrit.wikimedia.org/r/#/projects/mediawiki/extensions/VisualEditor,dashboards/default [19:08:47] 3VisualEditor: can't use save form checkboxes in opera - 10https://bugzilla.wikimedia.org/63615#c2 (10richlv) sorry, with ve currently not working with 1.22, i'll have to wait for the release of 1.23 to test this again [19:10:03] Richlv: https://www.mediawiki.org/wiki/VisualEditor?veaction=edit works as a test-bed if nothing else. [19:11:33] James_F, ah, good point, let me try that [19:12:15] James_F, https://commons.wikimedia.org/wiki/File:VisualEditor-logo-pacifico.svg [19:16:49] (03PS2) 10Esanders: Dialog demos [oojs/ui] - 10https://gerrit.wikimedia.org/r/131752 [19:16:51] (03PS2) 10Esanders: Minor refactor of widgets demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 [19:16:56] edsanders: Nice. [19:17:04] 3VisualEditor / 3Editing Tools: VisualEditor: A straight line appears above the image while making it (Frameless/Basic-Left/Right-border) - 10https://bugzilla.wikimedia.org/64966 (10ryasmeen) 3NEW p:3Unprio s:3normal a:3None Created attachment 15306 --> https://bugzilla.wikimedia.org/attachment.cg... [19:18:45] (03PS3) 10Krinkle: demo/widgets: Bring dev console back [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 [19:18:47] 3VisualEditor: can't use save form checkboxes in opera - 10https://bugzilla.wikimedia.org/63615#c3 (10richlv) for the record, this still does not work for me at https://www.mediawiki.org/w/index.php?title=VisualEditor&veaction=edit - but i'll wait for 1.23 to be out before complaining more about it :) [19:19:24] (03CR) 10Jforrester: [C: 032] "Yay." [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 (owner: 10Krinkle) [19:19:39] Krinkle: Nice to have it back; worked great in my testing. [19:19:53] (03Merged) 10jenkins-bot: demo/widgets: Bring dev console back [oojs/ui] - 10https://gerrit.wikimedia.org/r/131757 (owner: 10Krinkle) [19:20:32] edsanders: I'll rebase 131754 onto it. Was about to merge yours when JAmes merged mine. [19:22:41] Krinkle: Sorry for merging your code and saying thank you. In future I won't. :-P [19:22:57] It's fine, either way it needed a rebase. [19:23:02] * James_F nods. [19:23:03] Only like 1 line [19:23:35] (03PS3) 10Krinkle: Minor refactor of widgets demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 (owner: 10Esanders) [19:24:08] (03CR) 10Krinkle: [C: 031] "Rebased to resolve conflict with I405f76456e13ae4, and removed dependency on dialogs demo (doesn't seem to be dependent)." [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 (owner: 10Esanders) [19:24:49] (03CR) 10Krinkle: [C: 032] Minor refactor of widgets demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 (owner: 10Esanders) [19:25:04] Krinkle: Apparently http://i.imgur.com/G0Zp0cA.jpg is an important way to say this nowadays. [19:25:21] hehe [19:25:48] Do you have enough pixels per pixel there? [19:25:53] darn, that thing is blown [19:26:12] what's that capture from? I don't recognise it.. [19:26:16] 3VisualEditor / 3Language: VisualEditor: Ctrl-Arrow moves the caret too far in RTL wikis - 10https://bugzilla.wikimedia.org/64872#c2 (10Ed Sanders) This looks like just a problem with Hebrew. The new Unicode spec introduced a "HebrewLetter" group which we didn't add to the WordBreak checker. [19:26:21] (03Merged) 10jenkins-bot: Minor refactor of widgets demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/131754 (owner: 10Esanders) [19:27:24] The hangover? the office? [19:27:37] office it is [19:27:43] I don't recall exactly. [19:27:54] thx image search, https://www.youtube.com/watch?v=fiXze0535As [19:28:10] bang, and we've removed another 100 lines from mediawiki.js this week [19:28:36] Krinkle: When are you going to fix the regression re. collapsibleNav, BTW? [19:28:56] I'm not. [19:29:01] You aren't? [19:29:06] Depending which one you mean. [19:29:16] 'Cos it was merged on the basis that it would be done in CSS, right? [19:29:35] I did one major fix related to it already (bringing back the visual changes it did that weren't related to collapsible functinality) [19:29:36] (03PS1) 10Esanders: Update moveBreakOffset to include 'HebrewLetter' category [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131767 (https://bugzilla.wikimedia.org/64872) [19:30:14] OK. [19:30:19] But what about the changes that are? [19:30:31] making it collapsible nav is up to someone else to implement. This is removed under the flag of, very badly performant code that is unmaintained and not responding to pokes from performance engineering (aka ori) to address it, get nuked. [19:30:54] (03PS2) 10Jforrester: unicodejs: Update moveBreakOffset to include 'HebrewLetter' category [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131767 (https://bugzilla.wikimedia.org/64872) (owner: 10Esanders) [19:31:11] Very boldly put, If it's important, someone ought to be assigned to implement it (whether or not by re-using the old code from git history) [19:31:16] Krinkle: So major UX regressions are just a casualty of war? [19:31:26] It doesn't scale to have ori or me fix everthing that doesn't perform. [19:31:30] No, there are obvious exceptions. [19:31:38] Except when you don't care about them enough? [19:31:47] But this seemed worth the trade off, considering what it was doing on a lot of clients. It was causing serious delays and reflows. [19:31:50] Design are quite upset about this. [19:31:54] Basically, yes. [19:32:14] And there doesn't appear to be any work by you to actually ask for Product or Design input on whether this was an OK feature to just kill. [19:32:33] for the record though, I observed this one from the side line. [19:32:54] I agree with the principle, But I neither drafted, reviewed or deployed it. [19:33:44] You plural. [19:34:16] 3VisualEditor / 3Editing Tools: can't use save form checkboxes in opera - 10https://bugzilla.wikimedia.org/63615#c4 (10James Forrester) 5RES/WOR>3REO Not good. [19:35:02] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression] Can't use save form checkboxes in Opera - 10https://bugzilla.wikimedia.org/63615 (10James Forrester) 5REO>3ASS p:5Unprio>3High s:5normal>3major [19:36:06] I agree with the principle that stuff that is below what we'd approve for new code needs to get fixed in due time. And, if it isn't an isolated feature that can't hurt stuff, it may have to get killed if it is causing more UX trouble in the overal experience than the feature is adding. I think in this case, the benefit of having a sidebar less cluttered was smaller than the reflow, transisions [19:36:06] , and delay it was causing in the first palce. [19:36:32] for example, many abandoned components exist in core, but those won't get killed I thinik because they only hurt themselves. [19:36:42] Krinkle: So, given this, you guys asked on wikitech-l for someone to take care of it, or it would be removed in a month? Right? [19:36:50] I don't know [19:36:54] (No.) [19:36:56] I would say that's how it shoudl go [19:37:00] There was a bug [19:37:02] It would have been fine. [19:37:11] There are thousands of bugs. [19:37:13] Almost all of which are unwatched. [19:37:19] Then I saw a week later it was removed, and I was upset, because it screwed up the design. So I restored the design. [19:37:35] Not every bit of the code has someone like me. [19:38:08] Are you saying I should tell ori to do it differently next time? I'm agreeing with you. [19:38:27] Yes, but don't just tell, do it differently. [19:38:42] Pick five things you hate right now, send a quick e-mail to wikitech-l about them, and see if you get any bites. [19:39:09] Worst case scenario, you waste 30 minutes writing an e-mail, Ori doesn't care to follow your lead, and they get removed anyway. [19:39:27] Best case, you get interested developers, Ori agrees it's a good way to go forward, etc. etc. [19:43:22] Thoughts? [19:50:33] James_F: wanna know my thoughts? MW core doesn't have enough people working on it to respond to all concerns. if the design people are "upset", they should probably get someone to fix it for them. [19:50:47] 3VisualEditor / 3Editing Tools: VisualEditor: 0,0 initial selection in Firefox when opening a new or pre-existing references means inserting template causes fatal JS error - 10https://bugzilla.wikimedia.org/64434#c6 (10ryasmeen) Verified the fix on test2 [19:51:13] (on a side note, WMF's design dep't is continuously upset or upsetting someone else. that might be a problem on their side.) [19:51:31] MatmaRex: Failure to communicate upsets everyone unnecessarily. [19:51:40] you tell me [19:52:23] * MatmaRex conjectures that the design dep't is itself full of stories of communication failures [19:59:29] (03CR) 10Jforrester: [C: 032] Dialog demos [oojs/ui] - 10https://gerrit.wikimedia.org/r/131752 (owner: 10Esanders) [19:59:57] (03Merged) 10jenkins-bot: Dialog demos [oojs/ui] - 10https://gerrit.wikimedia.org/r/131752 (owner: 10Esanders) [20:05:32] 3VisualEditor / 3ContentEditable: VisualEditor: Links should follow the local CSS rules as if they were rendered in the view mode (e.g. external links shown with the external icon) - 10https://bugzilla.wikimedia.org/38726#c4 (10ryasmeen) Verified the fix in Betalabs and test2 [20:11:25] (03PS2) 10JGonera: Fix Dialog support on mobile [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131243 [20:11:53] (03CR) 10Jforrester: [C: 032] "Per Roan." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131243 (owner: 10JGonera) [20:11:56] (03CR) 10JGonera: Fix Dialog support on mobile (031 comment) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131243 (owner: 10JGonera) [20:13:44] (03Merged) 10jenkins-bot: Fix Dialog support on mobile [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131243 (owner: 10JGonera) [20:15:22] (03PS2) 10JGonera: Update VE core [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131245 [20:19:32] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression] Can't use save form checkboxes in Opera - 10https://bugzilla.wikimedia.org/63615#c5 (10Bartosz Dziewoński) Can not reproduce using Opera 12.17 on Windows 7, on the page from comment 3. Clicking on both the checkboxes and their labels works as expecte... [20:21:03] (03PS3) 10Jforrester: Update VE core submodule to master (f855dbb) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131245 (owner: 10JGonera) [20:21:25] (03CR) 10Jforrester: [C: 032] Update VE core submodule to master (f855dbb) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131245 (owner: 10JGonera) [20:23:34] (03Merged) 10jenkins-bot: Update VE core submodule to master (f855dbb) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131245 (owner: 10JGonera) [20:33:02] 3VisualEditor / 3Language: VisualEditor: Ctrl+cursor (Opt+cursor on Mac) moves the caret too far in Hebrew - 10https://bugzilla.wikimedia.org/64872 (10James Forrester) a:3Ed Sanders [21:02:17] 3VisualEditor / 3MediaWiki integration: VisualEditor: [Regression wmf22] "Edit source" tab for wikitext no longer selected on action=edit - 10https://bugzilla.wikimedia.org/64646#c8 (10ryasmeen) Verified the fix on Production, Betalabs and Test2 [21:20:43] (03CR) 10Esanders: [C: 04-1] "We should put the confirm dialog in ve core." (034 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130394 (https://bugzilla.wikimedia.org/50955) (owner: 10Alex Monk) [21:25:16] (03CR) 10Esanders: Swap from using native confirm() to using an OOUI dialog when cancelling (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/130394 (https://bugzilla.wikimedia.org/50955) (owner: 10Alex Monk) [21:25:42] thanks edsanders [21:25:55] welcome [21:26:17] I just wrote a very similar class in another project I'm using OOUI in [21:26:33] was initially wondering where to put it, eventually decided to upload it as part of mediawiki/extensions/VisualEditor [21:27:54] Krenair, https://gist.github.com/edg2s/66ae98e1e57343312c0d [21:28:14] a confirm dialog seems like a fairly useful re-usable component to me [21:28:36] could even argue to push it one further up to OOUI [21:28:42] yeah [21:28:56] I was thinking about OOUI as another option [21:29:40] also TrevorP|Away is working on stuff do with dialogs return promises [21:30:58] edsanders: Let's put it in OOUI then. [21:31:12] what does TrevorParscal think? [21:31:15] edsanders: If it's useful to a non-VE project JFDI. [21:31:17] yes [21:31:54] then it shall be [21:31:55] make it so [21:31:59] ok [21:32:00] edsanders: it's nice to see that James hasn't lost his nuanced approach to collective decision making [21:32:20] I wonder where I left my ooui repo... [21:32:52] El Segundo [21:33:26] I'm sort of considering this idea where open() returns a promise (instead of just being chainable) which is resolved/rejected when the dialog is closed (reject for cancel, resolve for accept, passing data into reject and resolve to provide further details) [21:33:59] then using a confirm dialog would be like dialog.open().then( /* continue */, /* nevermind */ ); [21:34:09] we will get there [21:35:12] Krenair, that means if you do it quick enough, Trevor will have to rebase on your work instead of the other way around :) [21:35:30] heh [21:36:10] I uploaded a couple of OOUI patches a couple of months ago but forgot where I put my repository :/ [21:37:32] hmph, oh well [21:38:48] 3VisualEditor: latest ve from git gets "fatal exception" with 1.22.6 - 10https://bugzilla.wikimedia.org/64971 (10richlv) 3UNC p:3Unprio s:3minor a:3None mw 1.22.6. enabling latest visual editor from git results in : "Fatal exception of type MWException" after some testing, no specific error messages c... [21:44:14] Krinkle, why do we use recess over contrib-less ? [21:52:59] edsanders: I don't know. What's so great about contrib-less? [21:53:29] I know recess has good validation, and some prettification [21:53:36] https://github.com/ngbp/ngbp/issues/215 [21:53:39] I haven't used grunt-contrib-less [21:53:39] newer version of less [21:54:36] I'd say give it a try and diff the dist/ (or I'll do at some point). [21:54:46] * Krinkle makes note [21:55:09] TrevorParscal, https://gist.github.com/Krenair/93558f8285c72fa6f975 look about right? [21:56:13] setupDeferred is odd [21:56:23] you aren't using it within setup, you are using with initialize [21:57:03] also, you don't need to ensure the dialog is initialized inside of open, it's gauranteed [21:57:11] Krinkle, I think it removes double whitespaces [21:57:14] TrevorParscal, that was probably to do with the way I misunderstood it before [21:57:16] 3VisualEditor: latest ve from git gets "fatal exception" with 1.22.6 - 10https://bugzilla.wikimedia.org/64971#c1 (10Andre Klapper) (In reply to richlv from comment #0) > enabling latest visual editor from git Which branch? [21:57:27] and doesn't do the .....-moz-transition indentation trick [21:57:29] no worries, just giving you some review :) [21:57:38] but it appears to be under more active development [21:58:00] in case we need use of more recent less features (not that I know what they are) [21:59:27] it also re-orders the rules within a selector to be consistent [22:01:23] I need to split LinkInspector into LinkAnnotationInspector and LinkNodeInspector, but I'm unhappy about it [22:01:30] I'm going to frump for a bit [22:03:03] How do I build OOUI into something I can put into MW again? [22:03:17] 3VisualEditor: latest ve from git gets "fatal exception" with 1.22.6 - 10https://bugzilla.wikimedia.org/64971#c2 (10richlv) master [22:03:23] * Krenair will rtfm [22:07:28] (03PS1) 10Esanders: Watch dot files [oojs/ui] - 10https://gerrit.wikimedia.org/r/131873 [22:08:43] (03PS1) 10Esanders: Watch dot files [oojs/core] - 10https://gerrit.wikimedia.org/r/131874 [22:09:46] TrevorParscal, how do I turn the contents of oojs/ui into a file like mediawiki/core.git's resources/lib/oojs-ui/oojs-ui.js ? [22:11:12] (03PS1) 10Esanders: Fix watch list [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131875 [22:12:31] edsanders, ^ [22:12:35] (03CR) 10Cmcmahon: [C: 032] [browser test] Refactor markup test to use API [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131738 (owner: 10Cmcmahon) [22:13:09] grunt [22:13:35] just 'grunt dist' probably should do it [22:14:05] (03PS1) 10Esanders: Fix watch list [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131876 [22:14:05] Warning: Task "dist" not found. Use --force to continue. [22:14:07] (03Merged) 10jenkins-bot: [browser test] Refactor markup test to use API [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131738 (owner: 10Cmcmahon) [22:14:26] where will that have left the output file(s) edsanders? [22:14:43] grunt build maybe [22:14:47] in dist/ [22:15:49] Krenair: I have a shell function I can share with you [22:16:12] I think I've got it now, but thanks [22:16:58] ok [22:17:00] no worries [22:20:58] TrevorParscal, I think I need to register my dialog with something... [22:21:22] as a ve-mw module that was ve.ui.windowFactory, what do I do instead in OOUI? [22:21:33] 3VisualEditor: "Middle click" / Cmd+Click on mac does not open links while in VE edit session. - 10https://bugzilla.wikimedia.org/64974 (10Jared Zimmerman (WMF)) 3NEW p:3Unprio s:3normal a:3None open in new tab used to work for links while in VE with CMD+Click on mac, this no longer works Mac 10.9.2 /... [22:24:14] Krenair, remember you need to put your built files in MW core if you're testing ve-mw [22:24:36] yes, that's what I was asking about earlier.. [22:36:13] Krenair: no, you don't need to register it [22:36:29] a confirm dialog would just be a one-off thing [22:36:52] you would just create and open it whenever [22:37:22] we only register things so we can associate tools, or nodes, or whatever, with dialogs [22:37:41] it's a way to loosely couple them, by using symbolic names we can always override the registry to link that name to a subclass [22:38:28] TrevorParscal, think I just found out why I was doing that setupDeferred stuff earlier - what's the proper way to open a dialog like this? [22:40:40] you just call .open() [22:43:09] How do I actually create it then? [22:43:21] It's apparently unknown, so... [22:43:30] new OO.ui.ConfirmDialog() [22:43:45] i mean, the whole windowset thing is not needed for simple cases [22:43:52] right. [22:43:58] that explains where I'm going wrong [22:44:03] sure [22:44:54] attach it's $element to $( 'body' ) or an overlay, whatever makes sense to you [22:45:13] edsanders: the way you did this with the language dialog, where's that at? [22:45:21] maybe Krenair could look at that [22:45:31] 3VisualEditor / 3Editing Tools: VisualEditor: Link icon in menu shown as "selected" (but only if link was created this session?) - 10https://bugzilla.wikimedia.org/64979 (10James Forrester) p:5Unprio>3High [22:45:31] it's a one-off dialog, similar to a confirm dialog [22:45:33] 3VisualEditor / 3Editing Tools: VisualEditor: Link icon in menu shown as "selected" (but only if link was created this session?) - 10https://bugzilla.wikimedia.org/64979 (10James Forrester) 3ASS p:3Unprio s:3normal a:3None Created attachment 15309 --> https://bugzilla.wikimedia.org/attachment.cgi?i... [22:45:45] TrevorParscal: Fun bug for you if you want something that's probably(?!) "easy". ^^^ [22:45:52] My dialog code relies on initialize being called TrevorParscal [22:45:59] but it's apparently not getting called [22:46:25] Krenair: I think you follow something like, create with new, attach to DOM, call initialize, then call open/close [22:46:46] LanguageInputWidge [22:46:47] t [22:46:48] you can also have most of this be managed for you by using a windowset, but for a one-off, it's not nessecary [22:46:58] So I have to call initialize as well? [22:46:59] Krenair: see LanguageInputWidget (thanks edsanders) [22:47:50] edsanders: oooh, owch that is evil [22:47:59] you are creating a dummy surface just to launch dialogs from it? [22:48:02] this needs cleaning [22:48:24] "// HACK" [22:48:41] Krenair, you don't need the dummy surface because you're not inside an inspector [22:49:38] TrevorParscal, we also need the surface to get access to the windowset [22:51:19] but why do you need a surface? [22:51:29] can't you just open a dialog that is attached to the body? [22:52:20] edsanders, this just does surface.getDialogs().getWindow( 'languageSearch' ) and runs open... [22:53:08] TrevorParscal, it doesn't need the whole surface, that just creates an overlay and windowset [22:53:14] Krenair, yup [22:54:06] that requires it to be registered with the ve.ui.windowFactory [22:54:43] I have to do it without registering with ve stuff [22:55:24] getWindow(name) just finds the constructor by name and instantiates it [22:55:28] so just instantiate it [22:56:00] ( new OO.ui.ConfirmationDialog() ).open( ... [22:56:20] yes, although we don't have to construct it every time, they're reusable [22:56:47] provided you've got nothing instance specific in initialize [22:56:57] everything should be be in setup/teardown [22:58:14] initialize does not appear to get called [22:59:01] my initialize function hasn't changed, I still set this.$container there which I need to use in open [23:01:15] edsanders, are you saying this should be in setup? [23:01:42] (03CR) 10Krinkle: [C: 032] Watch dot files [oojs/ui] - 10https://gerrit.wikimedia.org/r/131873 (owner: 10Esanders) [23:02:09] (03PS2) 10Krinkle: build: Add dot files to grunt watch [oojs/core] - 10https://gerrit.wikimedia.org/r/131874 (owner: 10Esanders) [23:02:11] (03Merged) 10jenkins-bot: Watch dot files [oojs/ui] - 10https://gerrit.wikimedia.org/r/131873 (owner: 10Esanders) [23:02:33] your initialize looks fine as it's all generic stuff [23:02:59] (03PS3) 10Krinkle: build: Add dot files to grunt watch [oojs/core] - 10https://gerrit.wikimedia.org/r/131874 (owner: 10Esanders) [23:03:35] (03CR) 10Krinkle: [C: 032] "(updated commit message to be suitable for History.md at release, + fixed missing new lines while at it)" [oojs/core] - 10https://gerrit.wikimedia.org/r/131874 (owner: 10Esanders) [23:04:00] (03Merged) 10jenkins-bot: build: Add dot files to grunt watch [oojs/core] - 10https://gerrit.wikimedia.org/r/131874 (owner: 10Esanders) [23:04:21] (03CR) 10Krinkle: [C: 032] Fix watch list [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131875 (owner: 10Esanders) [23:04:39] Krinkle: Any new features for OOjs, BTW? [23:04:39] (03CR) 10Krinkle: [C: 032] Fix watch list [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131876 (owner: 10Esanders) [23:05:00] * James_F was thinking about whether the only thing we've got ahead of ourselves is the ES3-parser compatibility. [23:05:08] nope, nothing has changed in src/ yet since the last release [23:05:32] (03Merged) 10jenkins-bot: Fix watch list [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/131875 (owner: 10Esanders) [23:05:46] initialize gets called when the frame is ready [23:05:47] I'm prioritising the rl dependency functions first. [23:05:50] see OO.ui.Frame [23:06:01] and OO.ui.Window [23:06:13] (03Merged) 10jenkins-bot: Fix watch list [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131876 (owner: 10Esanders) [23:06:48] I think I'm just going to put the diffs on gist [23:07:05] (03PS1) 10Cmcmahon: [browser test] template workflow changed somewhat [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131895 [23:09:57] TrevorParscal, https://gist.github.com/Krenair/adbe126eae978c6e5cef https://gist.github.com/Krenair/9f23ae9062804423d912 [23:12:18] Uncaught TypeError: Cannot read property 'text' of undefined [23:13:22] Krenair, OO.ui.Dialog#open only takes config [23:13:37] no need for surface fragment, or 'dir' property [23:14:14] hm [23:14:31] I forget where I found that being done [23:15:37] Krinkle: No, I didn't mean do we have them written and merged, I meant do we have any we're expecting to do. [23:16:38] James_F: Don't think so. The only thing on my mind is a class maker that will abstract a lot of this. But can't priotise that right now [23:16:44] * James_F nods. [23:16:53] That's fine, just thinking about roadmaps. :_) [23:17:40] e.g. something like var self = ve.Foo = OO.Base.extend({ constructor: .., static: .., methods: .. }); [23:17:57] ... mixins: [] .. [23:18:20] which would set up everything at once instead of calling various OO methods in various places. [23:18:38] And also should perform slightly better by not referencing everything in full from the global scope [23:18:59] (e.g. self.super instead of ve.this.that.Foo.super.prototype. [23:20:12] Indeed. [23:24:32] 3VisualEditor / 3ContentEditable: VisualEditor: Firefox allows cursor to get into illegal position around wide objects (, , etc), leading to corruption, and prevents down arrow button navigation - 10https://bugzilla.wikimedia.org/51957#c4 (10James Forrester) 5ASS>3RES/FIX a:3Ed Sand... [23:26:32] 3VisualEditor / 3ContentEditable: VisualEditor: When cursor ends up outside of a content branch node, move it back into one - 10https://bugzilla.wikimedia.org/56340#c2 (10James Forrester) 5ASS>3RES/FIX a:5David Chan>3Ed Sanders I believe that this was fixed by Ed's Firefox CE re-write. [23:29:07] Krenair, ve.ui.Dialog requires it, because they interact with the surface [23:29:54] we should rename those to SurfaceDialog [23:29:59] or FragmentDialog [23:30:46] 3VisualEditor / 3ContentEditable: VisualEditor: Vertical keyboard selection (shift+up/down cursor) over a node type boundary gives unexpected results in Firefox - 10https://bugzilla.wikimedia.org/56248 (10James Forrester) 5NEW>3ASS p:5Unprio>3High [23:31:17] 3VisualEditor / 3ContentEditable: VisualEditor: Pressing a modifier key (Ctrl/Alt/Shift/Cmd/etc.) when cursor is offscreen or under the toolbar causes the page to scroll up - 10https://bugzilla.wikimedia.org/53420 (10James Forrester) [23:31:17] 3VisualEditor / 3ContentEditable: VisualEditor: Vertical keyboard selection (shift+up/down cursor) over a node type boundary gives unexpected results in Firefox - 10https://bugzilla.wikimedia.org/56248 (10James Forrester) [23:32:02] James_F: Hm.. did you notice the square brackets around edit|edit source shivver on page load? [23:32:13] they don't seem to move any where, just shrink in spaceing and then back where they were [23:32:23] Krinkle: No? Is this in master/prod? [23:32:23] can you confirm? [23:32:26] master [23:32:29] but prod as well I imagine [23:32:40] What browser? [23:32:43] chrome [23:32:50] really, English is broken. shivver should be with two v's [23:32:52] Oh, the FOUC thing? [23:33:00] Is it fouc? [23:33:03] I guess [23:33:09] but it's not any different [23:33:13] it doens't go from A to B [23:33:24] it's like A -> B -> C, where C equals A [23:33:25] It starts as "[edit source | editbeta]" and then changes to [ edit source | editbeta ] [23:33:32] For me, at least. [23:33:49] Oh, right. It looks like ABA because Im refresshing [23:33:50] Hmm. OK, now it's not doing that. [23:33:51] * James_F sighs. [23:33:53] refreshing* [23:34:07] so my visual initial state is the expanded state [23:34:13] browsers are too fast these days [23:34:17] Yeah. :-) [23:34:53] Hm.. why does it fouc though? The link is already there on page load. It just looks like it adds a space [23:35:06] Seems like some css is in the wrong module [23:35:18] why isn't it styled like that from the get go? [23:35:19] Quite possibly. [23:35:36] it's not like it needs javascript to figure it we want a space there [23:36:06] Actually it does. [23:36:16] your'e right [23:36:19] "edit" is js only [23:36:20] Remember that the normal version is [edit] not [ edit ] [23:36:33] normal is [edit source] [23:36:40] and becomes [ edit | edit source ] [23:36:41] [ edit | edit source ] (or [ edit source | edit beta ]) is JS. [23:36:43] Yeah. [23:36:44] but not quite [23:36:48] both edit links are there [23:36:52] from the php response [23:36:57] but one is hidden with css [23:37:04] .client-js, client-nojs [23:37:06] Maybe we should change Vector's styling to be like VE's, or v.v.? [23:37:08] which is good [23:37:08] Yeah. [23:37:14] that means no fouc for 'edit' appearing [23:37:37] looks like the space somehow are made by other means [23:38:18] 3VisualEditor / 3ContentEditable: VisualEditor: Page down and page up not working as expected in Firefox - 10https://bugzilla.wikimedia.org/53750#c5 (10James Forrester) 5ASS>3RES/FIX a:3Ed Sanders AFAICT this is now works in Firefox since the Firefox-related CE fixes, though there are some oddities gen... [23:38:20] Huh? It's almost as if someone wanted a fouc [23:38:29] it adds class mw-editsection-expanded which makes the spaces [23:38:38] tat class is added by js on page load [23:39:07] Helpful. [23:40:03] 3VisualEditor / 3Editing Tools: VisualEditor: The suggestion dropdown box gets stuck inside the transclusion dialog box in Firefox when you click on Add template button as soon as the dropbox appears - 10https://bugzilla.wikimedia.org/64982 (10ryasmeen) 3NEW p:3Unprio s:3normal a:3None Created attach... [23:40:20] edsanders, I don't think changing that fixes the issue.... [23:40:32] 3VisualEditor / 3Editing Tools: VisualEditor: The suggestion dropdown box gets stuck inside the transclusion dialog box in Firefox when you click on Add template button as soon as the dropdown box appears - 10https://bugzilla.wikimedia.org/64982 (10ryasmeen) [23:40:48] 3VisualEditor / 3ContentEditable: VisualEditor: Moving cursor up or down before a reference behaves as page up/down in Firefox - 10https://bugzilla.wikimedia.org/50726#c3 (10James Forrester) 5ASS>3RES/FIX a:3David Chan I believe that it's now fixed, yes. The cursor getting stuck on down is odd and I'll... [23:40:57] what is the issue? [23:42:48] 3VisualEditor / 3ContentEditable: VisualEditor: Vertical cursoring can get cursor stuck on a paragraph-terminal reference in Firefox - 10https://bugzilla.wikimedia.org/64983 (10James Forrester) 3ASS p:3Unprio s:3normal a:3None From bug 50726 comment : | Steps to reproduce: | 1. Load https://en.wiki... [23:43:01] 3VisualEditor / 3ContentEditable: VisualEditor: Moving cursor up or down before a reference behaves as page up/down in Firefox - 10https://bugzilla.wikimedia.org/50726#c4 (10James Forrester) (In reply to James Forrester from comment #3) > I believe that it's now fixed, yes. The cursor getting stuck on down i... [23:43:31] 3VisualEditor / 3ContentEditable: VisualEditor: Vertical cursoring can get cursor stuck on a paragraph-terminal reference in Firefox - 10https://bugzilla.wikimedia.org/64983 (10James Forrester) p:5Unprio>3Normal [23:56:36] (03CR) 10Cmcmahon: [C: 032] [browser test] template workflow changed somewhat [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131895 (owner: 10Cmcmahon) [23:58:04] (03Merged) 10jenkins-bot: [browser test] template workflow changed somewhat [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/131895 (owner: 10Cmcmahon)