[00:00:24] Awesomeness [00:00:29] It's all working smoothly now [00:00:30] * RoanKattouw merges [00:01:38] Hmm, primaryCommandName isn't documented yet [00:01:39] I'll add that [00:02:10] Yay! [00:06:12] (03PS16) 10Catrope: Create mixin that tries to execute primary node command [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/103062 (owner: 10Robmoen) [00:14:15] (03CR) 10Catrope: [C: 032] Create mixin that tries to execute primary node command [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/103062 (owner: 10Robmoen) [00:14:44] rmoen: Woo-hoo! [00:15:12] (03Merged) 10jenkins-bot: Create mixin that tries to execute primary node command [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/103062 (owner: 10Robmoen) [00:18:05] (03PS20) 10Catrope: Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [00:18:07] (03PS1) 10Catrope: Update VE core submodule to master (9dbb70f) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 [00:19:27] (03PS21) 10Catrope: Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [00:19:29] (03PS2) 10Catrope: Update VE core submodule to master (9dbb70f) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 [00:21:14] (03PS1) 10Catrope: Actually add ClickableNode places [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119666 [00:21:49] (03CR) 10Jforrester: [C: 032] Actually add ClickableNode places [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119666 (owner: 10Catrope) [00:22:44] (03Merged) 10jenkins-bot: Actually add ClickableNode places [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119666 (owner: 10Catrope) [00:24:10] (03PS3) 10Jforrester: Update VE core submodule to master (8c7b4e2) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 (owner: 10Catrope) [00:25:28] (03PS4) 10Jforrester: Update VE core submodule to master (8c7b4e2) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 (owner: 10Catrope) [00:25:58] (03PS22) 10Jforrester: Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [00:26:22] "00:24:15 >> TypeError: 'undefined' is not an object (evaluating 'OO.ui.GroupWidget.super.prototype.setDisabled.call')" I assume this is a breaking change that we missed [00:26:53] Argh. [00:27:06] It shouldn't have been breaking. [00:27:19] But it looks like something that y'all changed? [00:27:20] RoanKattouw: ^^^^ [00:27:25] Yeah. [00:27:28] I also ran into that >_> [00:27:28] Is this because the oojs version wasn't pulled in? [00:27:38] rdwrer: What version of OOJS are you using and where does it come from? [00:27:45] It's whatever's in core [00:27:48] https://integration.wikimedia.org/ci/job/mwext-MultimediaViewer-qunit/1056/consoleFull [00:27:49] RoanKattouw: You mean OOjs UI. [00:27:50] I ran into that when I updated my core [00:27:57] James_F: No OOJS [00:28:03] I'm afraid this might be .super breaking [00:28:10] RoanKattouw: Core has the working version of OOjs. [00:28:23] Super breaking, excellent [00:28:33] Nope, super is there [00:28:37] Something wrong with GroupWidget maybe? [00:28:44] rdwrer: Are you /sure/ you're not over-riding OOjs? [00:28:52] Pretty damn sure [00:29:02] We would have seen fire 'n' flames way sooner elsewise [00:29:02] wat [00:29:02] Damn. [00:29:09] OO.ui.GroupWidget.super is OO.ui.GroupElement [00:29:18] Which doesn't have a setDisabled method [00:30:03] But presumably it was previously OO.ui.GroupElement.… and it was changed to OO.ui.GroupWidget.super… [00:30:17] My apologies [00:30:51] James_F: No it was OO.ui.Widget, that's the problem [00:30:58] GroupWidget is not an instance of Widget [00:30:58] It would be cool to have things working before the mw branch [00:31:00] Which is stupid [00:31:03] Well, then it shouldn't have been changed. [00:31:03] Yeah, wokring on it [00:31:06] But alternatively we could revert [00:32:54] (03CR) 10Jforrester: "Breakage: OO.ui.GroupWidget.super is OO.ui.GroupElement but was calling the not-its-actual-parent OO.ui.Widget.setDisabled, which got chan" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119471 (owner: 10Esanders) [00:33:40] (03PS23) 10Jforrester: Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [00:36:56] (03PS1) 10Catrope: Revert "Use super to call parent methods" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119669 [00:37:10] (03CR) 10Jforrester: [C: 032] Revert "Use super to call parent methods" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119669 (owner: 10Catrope) [00:37:31] (03Merged) 10jenkins-bot: Revert "Use super to call parent methods" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119669 (owner: 10Catrope) [00:38:44] (03PS1) 10Jforrester: Update OOjs UI to v0.1.0-pre (0f50cdc46c) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119671 [00:38:54] (03CR) 10Catrope: "Reverted in https://gerrit.wikimedia.org/r/119669 , very broken. See inline comment for details." (031 comment) [oojs/ui] - 10https://gerrit.wikimedia.org/r/119471 (owner: 10Esanders) [00:39:06] (03CR) 10Catrope: [C: 032] Update OOjs UI to v0.1.0-pre (0f50cdc46c) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119671 (owner: 10Jforrester) [00:39:34] RoanKattouw: https://gerrit.wikimedia.org/r/119672 [00:39:41] rdwrer: Sorry about this. :-( [00:39:46] 's OK [00:39:57] It really isn't. But thanks. [00:40:40] (03Merged) 10jenkins-bot: Update OOjs UI to v0.1.0-pre (0f50cdc46c) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119671 (owner: 10Jforrester) [00:42:12] (03PS1) 10Catrope: Fix messy inheritance chain in GroupWidget. [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 [00:45:43] rdwrer: The fact that your tests caught it and OOjs UI's didn't makes me sad, in particular. [00:45:57] (Oh, wait, there /are/ no actual OOjs UI tests.) [00:46:06] [01:08:39] James_F: Well there's the widgets demo, which crashed [01:08:50] The reviewer could have run the widgets demo in their browser [01:08:50] RoanKattouw: Which doesn't block merge. [01:08:54] True [01:08:57] It probably should [01:08:58] RoanKattouw: Clearly failed so to do. :-) [01:09:03] Ditto for VE-MW VPT [01:16:20] (03CR) 10Catrope: [C: 032] Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [01:17:28] (03CR) 10Catrope: [C: 032] Update VE core submodule to master (8c7b4e2) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 (owner: 10Catrope) [01:19:18] (03Merged) 10jenkins-bot: Update VE core submodule to master (8c7b4e2) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119665 (owner: 10Catrope) [01:19:29] (03Merged) 10jenkins-bot: Mixin ClickableNode and add primary command names [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/103063 (owner: 10Robmoen) [01:22:14] Krinkle: Red links commit *poke* *poke* ? [01:23:11] (03PS2) 10JGonera: [WIP] Style link inspector on mobile [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119413 [01:23:53] (03CR) 10jenkins-bot: [V: 04-1] [WIP] Style link inspector on mobile [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119413 (owner: 10JGonera) [01:33:07] (03CR) 10Physikerwelt: [C: 031] "We should make progress on that." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/105648 (owner: 10Catrope) [01:39:35] (03PS1) 10Krinkle: ToggleButtonWidget: Remove duplication in setValue [oojs/ui] - 10https://gerrit.wikimedia.org/r/119682 [01:40:03] (03PS2) 10Krinkle: ToggleButtonWidget: Remove duplication in setValue [oojs/ui] - 10https://gerrit.wikimedia.org/r/119682 [01:45:45] (03PS1) 10Krinkle: GroupWidget: Correct incorrect comment about parent method call [oojs/ui] - 10https://gerrit.wikimedia.org/r/119683 [01:46:05] Krinkle: Red links *poke* :) [01:46:20] (03CR) 10Krinkle: "What is this weird class anyway? It doesn't inherit from Widget, yet is called *Widget. Inherits from GroupElement, no futher mixins. Not " [oojs/ui] - 10https://gerrit.wikimedia.org/r/119683 (owner: 10Krinkle) [01:46:25] RoanKattouw: Yep, on it, next. [01:46:35] Krinkle: Good. :-) [01:46:35] Is there a particular schedule I should know about btw? [01:46:45] (03CR) 10Krinkle: Use super to call parent methods (031 comment) [oojs/ui] - 10https://gerrit.wikimedia.org/r/119471 (owner: 10Esanders) [01:46:47] Krinkle: We want it merged now. :-) [01:47:08] (03CR) 10Catrope: "Try rebasing onto https://gerrit.wikimedia.org/r/119675 ;)" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119683 (owner: 10Krinkle) [01:47:13] Krinkle: Specifically, we want it in master before the cut in 12 hours' time, but… [01:47:18] Krinkle: Weekly deployment on Thursdays [01:47:33] tomorrow [01:47:49] 16 hours (I think) [01:47:50] OK [01:48:11] Krinkle: The cut normally happens around 08:00 or 09:00 SF time, so definitely not 16. [01:48:31] Krinkle: But also, we need Roan to be around to respond if you've got review comments, so… :-) [01:48:42] (03CR) 10Krinkle: Fix messy inheritance chain in GroupWidget. (031 comment) [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [01:49:02] google calendar says 19:00 (my time), it's nearly 03:00 now [01:49:03] anyway [01:49:39] Krinkle: The calendar tells you when it's deployed. [01:49:47] Krinkle: The cut normally happens several hours beforehand. [01:49:48] but isn't 03:00 in real life like 19:00 in Krinkle? [01:49:51] Google knows. [01:50:07] No, I'm going to bed early today. [01:50:07] Krinkle: Sometimes Reedy cuts a corner or two, but that's not safe to rely upon. :-) [01:50:15] :) [01:50:30] Krinkle: So, just after you merge this? :-) [01:50:40] Yeah, not now. [01:50:40] Krinkle: Or do you want Trevor to merge? [01:52:28] (03CR) 10Krinkle: "Testing now.." (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [01:54:30] (03CR) 10Krinkle: Display links to nonexistent pages as red (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [01:57:00] RoanKattouw: btw, something to look into maybe (I'll check it out as well) – I noticed just now that as of today (could be Chrome update, could be us) our text nodes are being chopped up into 1-characters text nodes exploding the DOM quite a bit. [01:57:04] http://i.imgur.com/dT3qoYA.png [01:57:07] happens when cursorring around [01:57:12] each one gets chopped up [01:57:38] I cursorred from "'s" to "inf" [01:58:13] Krinkle: Chrome has always split text nodes when typing [01:58:26] We don't do that [01:58:47] In fact, you'll notice that if you press Enter or bold a word or take some other action which forces a ContentBranchNode rerender, the text nodes get merged [02:00:21] Krinkle: What does "wrap on the outside" mean? [02:00:53] iffy style, not sure where you got this one from, but we've been on "wrap iffy" instead of "wrap function" for a while now [02:01:01] include the invocation in the parenthesis [02:01:46] I think you might've started that within wikimedia conventions actually, I remember it matching crockford's anti-dogballs rule when I first joined RL mediawiki dev "all those years" ago. [02:02:30] Oh, you want me to use the illogical wrapping style [02:02:47] ( function () { ... } )() ) instead of ( function () { ... } )(); ? [02:02:58] The latter makes sense, the former pretty much doesn't [02:03:12] yep, wikimedia, mediawiki/core, ve, oojs, ve-mw, we don't wrap the function only anywhere [02:03:25] Urgh I guess you changed it to the weird one everywhere [02:03:27] So fine [02:04:22] Firstly, they're the same effectively, so consistency is the most significant factor. There's very little rationale to it, but for what its worth, I see the rationale behind it. The parenthesis aren't needed at all in the langguage in that position. They're there to indicate something special (we're returning the result of an invocation instead of the function, especially when assignging prope [02:04:22] rties or variables this is important). [02:04:30] and wrapping only the function then doesn't really make sense. [02:04:56] Just lie you wouldn't var x = (5) * 5; [02:05:02] Yeah I realize function () {} () works [02:05:14] you'd ( 5 * 5 ) (if anything) [02:05:46] It's just that it's not intuitively obvious that function () {} () would work, which is probably how ( function () {} ) () came about [02:05:56] (thing I want to execute)(); [02:06:15] Either way, I'll change it in a minute [02:06:34] (03CR) 10Catrope: Display links to nonexistent pages as red (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:06:36] RoanKattouw: I noticed one thing that could be fixed later but maybe it's easy if you know. [02:06:45] RoanKattouw: When inserting or changing a link to become a red link, it's making an api request. [02:06:58] Isn't this already known to the inspector though? The user clicks "Red link" after all. [02:07:16] "This doesn't yet integrate the LinkCache with other components like the link inspector. That should be done so we can deduplicate the existence checks." -- commit message [02:07:35] if we don't trust opensearch (which we shouldn't because there's bugs with it where even an exact match isn't showing up), we should make that request, btu make when rendering the inspector, not when saving. [02:07:51] I noticed it because there was a half second delay between clicking save and the link turning red. [02:08:06] OK [02:08:17] It kind of adds it in this commit though. [02:08:38] Krinkle: When the new search interface is actually reliably available, we can just use that. [02:08:38] Or is that a side effect of the link being re-created. [02:08:56] Krinkle: But it isn't, so we have to rely on OpenSearch, which sucks, yes. [02:09:23] What do you mean? OpenSearch uses whatever backend is configured, right? [02:09:48] The problem is that OpenSearching for "foo" may not return "foo" even if [[Foo]] exists [02:09:56] In fact, it may return an empty array [02:10:00] It's not a search backend afaik. And though it should be easier to fix in cirrus, I don't think cirrus is unaffected. In fact, these bugs are new in cirrus (compared to the old/default backend that is) [02:10:03] Especially when redirects are involved [02:10:12] Yes but that used to be fine iirc. [02:10:21] It started happening recently when I switched search backends. [02:10:21] Whatever the cause is, OpenSearch results are not reliable for existence checks right now [02:10:27] Yeah [02:10:30] When they are, we should integrate them with the LinkCache [02:10:48] Admittedly I didn't not do this because of the unreliability but because I was lazy [02:11:23] But you agree, that bug applies to the inspector UI more than the new class added afterwards (when the user actually is faced with "creating a red link" so we should make that request regardless of this new logic, perhaps a generator query from open search to page existence) [02:11:40] not, more, but equally so anyway. [02:11:43] But yes, you're absolutely right, when the link suggestions come back we already have existence information for certain things and that should be aggressively written into the LinkCache for later reuse [02:12:02] No, i agree open search isn't reliable right now. [02:12:05] We also already make a prop=info request for redirect status [02:12:30] Which also 1) gives us existence information and 2) gives us redirect+disambig information that we should really also be stashing in the LinkCache [02:12:45] Everything we ever find out from the API about a page should be written to the LC so we can query it later [02:13:12] Or *a* LinkCache rather, we need multiple caches for different things [02:13:19] and getting page info is page existence good (which we now do for the first time, thanks to this commit), but when we integrate this further soon, it shoudl make that request when still in the UI, not after you close it. Because now it'll create a UX regression kinda where instead of consistently being wrong about a page not existing (thx to opensearch), it'll be wrong in the inspector but be [02:13:19] right when closing it. [02:13:29] Which means I also need to split that class up into a reusable part and a per-thing part [02:13:41] (03CR) 10Krinkle: [C: 031] "Good to go (aside from the iffy)." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:13:43] Yes, that is slightly confusing, I agree [02:14:13] (03PS13) 10Catrope: Display links to nonexistent pages as red [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 [02:14:21] That fixes the iffy [02:14:32] Also, this commit has a (small) dependency [02:15:41] Hm.. Parsoid defaults to enwiki/Main_Page when providing an unknown interwiki. Disregards the title either way. [02:15:42] Interesting. [02:15:58] http://localhost:8000/_rt/localhost2/Sandbox/Worksheet -> enwiki/Main_Page [02:16:04] lolwut [02:16:18] (03CR) 10Krinkle: [C: 031] Display links to nonexistent pages as red [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:16:38] Krinkle: Yeah, it's fun, isn't it? [02:17:52] Also, it seems something changed recently. Where previously it rendered my links like [[fr:Sandbox]] as red links (which is correct and matches mediawiki because I don't have an fr interwiki on my local wiki), it now renders as hidden meta data (not in the surface) and not in the interwiki editor either. It's dissappears. [02:18:26] Probably a combination of Parsoid making bad assumptions about everyone having wmf's interwiki's and us finding out that's not true. And thus hidden on both sides [02:18:28] Known bug? [02:18:56] Krinkle: Sounds familiar. [02:19:22] (03CR) 10Krinkle: [C: 031] Add lookupTitle attribute (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:19:32] It was fine in Januari [02:19:35] January [02:20:31] Krinkle: Too much time thinking in an Scandinavian language? :-) [02:20:39] Dutch [02:20:50] That's the Dutch spelling? [02:21:00] Yep [02:21:03] Huh. [02:21:18] Is that consistent? Februari? [02:21:23] Januari Februari Maart April Mei Juni Juli Augustus September October November December [02:21:24] Yep [02:21:38] Oktober* [02:23:44] (03CR) 10Catrope: Add lookupTitle attribute (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:27:57] (03PS8) 10Catrope: Remove the math inspector code [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/105648 [02:34:49] (03PS4) 10Jforrester: Add lookupTitle attribute [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:35:56] (03PS1) 10Jforrester: Add missing ve.ce.ClickableNode.js to .docs/eg-iframe.html [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119688 [02:36:16] (03CR) 10Catrope: [C: 032] Add missing ve.ce.ClickableNode.js to .docs/eg-iframe.html [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119688 (owner: 10Jforrester) [02:37:50] (03Merged) 10jenkins-bot: Add missing ve.ce.ClickableNode.js to .docs/eg-iframe.html [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119688 (owner: 10Jforrester) [02:37:56] (03CR) 10Jforrester: Add lookupTitle attribute (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:38:53] Krinkle: ^^^ {{done}} please merge. :-) [02:40:07] (03CR) 10Catrope: Fix messy inheritance chain in GroupWidget. (031 comment) [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [02:40:52] (03PS2) 10Catrope: Fix messy inheritance chain in GroupWidget. [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 [02:42:29] (03CR) 10Krinkle: [C: 032] Add lookupTitle attribute [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:42:52] (03PS14) 10Krinkle: Display links to nonexistent pages as red [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:42:54] (03CR) 10Krinkle: [C: 032] Display links to nonexistent pages as red [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:44:01] James_F: btw, VE london before Zürich? [02:44:11] (03Merged) 10jenkins-bot: Add lookupTitle attribute [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119436 (owner: 10Catrope) [02:44:12] I need info on that weekend soon. other things start needing dates and rsvps [02:44:24] (03Merged) 10jenkins-bot: Display links to nonexistent pages as red [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118045 (owner: 10Catrope) [02:44:27] Krinkle: Option 3 seems most sane to me, but I'm not writing it, so… :-) [02:44:46] Krinkle: No; Ed doesn't leave until the weekend before the weekend before Zürich. [02:45:06] Krinkle: Given the lengthy stays in SF this quarter, I felt we didn't need it. [02:50:00] James_F: The wiki anno 2010 way for that bot would be to have a bash script running on fenari that reports every edit on every wiki to a bot-only irc channel, then another ircbot/wikibot to listen to that freenode channel and put it on a wiki, then another bot to listen to the rss feed of that wiki page and make api queries recent changes and then send e-mails to a mailing list for edits that [02:50:00] have that tag. [02:50:10] Krinkle: Indeed. :-) [02:50:15] and then an irc bot that is listening on the mail server for new emails in that inbox [02:51:35] of course having mailing threads for these might actually be useful so you can respond to them and mark them as fied. [02:51:37] fixed* [02:51:51] we should share IMAP credentials to mark this as read and archive them collectively :) [02:52:07] WONTFIX. :-P [05:28:57] (03PS1) 10Catrope: Ignore placeholder parameters in serialization [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119699 [05:31:23] (03CR) 10Jforrester: [C: 032] Ignore placeholder parameters in serialization [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119699 (owner: 10Catrope) [05:32:44] (03Merged) 10jenkins-bot: Ignore placeholder parameters in serialization [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119699 (owner: 10Catrope) [05:32:48] (03CR) 10Catrope: [C: 031] Remove the math inspector code [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/105648 (owner: 10Catrope) [05:33:08] (03CR) 10Jforrester: [C: 032] Remove the math inspector code [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/105648 (owner: 10Catrope) [05:34:32] (03Merged) 10jenkins-bot: Remove the math inspector code [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/105648 (owner: 10Catrope) [05:48:04] (03PS1) 10Catrope: Don't use search widget emptiness as a proxy for "no reusable references" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119702 [05:49:21] (03CR) 10Jforrester: [C: 032] Don't use search widget emptiness as a proxy for "no reusable references" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119702 (owner: 10Catrope) [05:50:41] (03Merged) 10jenkins-bot: Don't use search widget emptiness as a proxy for "no reusable references" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119702 (owner: 10Catrope) [06:00:09] (03PS1) 10Catrope: Followup 861d097: actually exclude reference tool in reference dialog toolbar [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119704 [06:03:24] (03CR) 10Jforrester: [C: 032] Followup 861d097: actually exclude reference tool in reference dialog toolbar [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119704 (owner: 10Catrope) [06:04:44] (03Merged) 10jenkins-bot: Followup 861d097: actually exclude reference tool in reference dialog toolbar [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119704 (owner: 10Catrope) [06:08:40] (03PS3) 10Krinkle: Fix messy inheritance chain in GroupWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [06:08:48] (03PS4) 10Krinkle: GroupWidget: Fix messy inheritance chain [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [06:11:36] (03Abandoned) 10Krinkle: GroupWidget: Correct incorrect comment about parent method call [oojs/ui] - 10https://gerrit.wikimedia.org/r/119683 (owner: 10Krinkle) [06:11:46] (03CR) 10Krinkle: [C: 032] GroupWidget: Fix messy inheritance chain [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [06:12:41] (03Merged) 10jenkins-bot: GroupWidget: Fix messy inheritance chain [oojs/ui] - 10https://gerrit.wikimedia.org/r/119675 (owner: 10Catrope) [06:14:45] (03PS1) 10Krinkle: intro/outro: Keep local reference to OO inside closure [oojs/ui] - 10https://gerrit.wikimedia.org/r/119708 [07:31:14] (03CR) 10Raimond Spekking: Disable VisualEditor on Page Translation output pages (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119427 (owner: 10Jforrester) [12:29:24] (03PS2) 10Esanders: Promote zIndex of InlineMenuWidget options list [oojs/ui] - 10https://gerrit.wikimedia.org/r/119492 [12:29:26] (03PS3) 10Esanders: Add disabled styling to InlineMenuWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/119491 [13:43:49] (03PS3) 10Esanders: Basic toolbar demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/119516 [13:43:51] (03PS3) 10Esanders: Styling for disabled toolgroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119517 [13:43:53] (03PS3) 10Esanders: Force visible label for MenuToolGroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119518 [13:43:55] (03PS1) 10Esanders: Use registry for tool group types [oojs/ui] - 10https://gerrit.wikimedia.org/r/119744 [16:21:53] (03CR) 10Jforrester: Disable VisualEditor on Page Translation output pages (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119427 (owner: 10Jforrester) [16:41:19] Hi! what is the output of window.gM = mw.jqueryMsg.getMessageFunction( {} ) ? [16:41:42] I am confused with the empty function description [16:45:07] (03PS2) 10Esanders: Use registry for tool group types [oojs/ui] - 10https://gerrit.wikimedia.org/r/119744 [16:45:09] (03PS4) 10Esanders: Basic toolbar demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/119516 [16:45:11] (03PS4) 10Esanders: Styling for disabled toolgroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119517 [16:45:13] (03PS4) 10Esanders: Force visible label for MenuToolGroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119518 [16:45:27] tonythomas: mw.jqueryMsg.getMessageFunction() will get a function suitable to be used as a replacement for mw.msg [16:45:52] tonythomas: https://doc.wikimedia.org/mediawiki-core/master/js/#!/api/mw.jqueryMsg-method-getMessageFunction [16:50:26] rdwrer: ok. I wanted the window.gM be replaced with mw.message(' ').parse(); [16:50:45] so something like [16:51:31] mw.message( mw.jqueryMsg.getMessageFunction( {} )).parse(); ? [16:51:46] ...no? [16:52:07] tonythomas: getMessageFunction doesn't return a message key, it returns a function for *handling* message keys [16:52:28] Oh, wait. [16:52:41] tonythomas: If you're getting rid of window.gM, just delete the line that sets it maybe [16:52:58] rdwrer: actually, I want to get rid of window.gM [16:53:03] its deprecated right ? [16:53:10] Yeah, definitely [16:53:12] as per https://gerrit.wikimedia.org/r/#/c/113159/2/resources/mediawiki/mediawiki.jqueryMsg.js,unified [16:54:48] rdwrer: and https://www.mediawiki.org/w/index.php?title=Manual:Messages_API#Using_messages_in_JavaScript says something like "There are some more or less used aliases like gM (legacy shortcut for mw.jqueryMsg.getMessageFunction()) but its use is deprecated in favour of mw.message( ... ).parse()" [16:55:04] Right [16:56:06] rdwrer: so, it needs to be replaced by mw.message(something).parse() rightt ? [16:56:25] but, since the code shows only {} passed as argument, I am confuesd [16:56:25] Where can i find visualeditor documentation, structure for a quick task? [16:56:29] tonythomas: The assignment itself? No. [16:56:57] tonythomas: You can delete that line, I think, if it's in an extension. If it's the one in jquerymsg itself, leave it alone. [16:58:11] rdwrer: Oh. I forgot to give you the bug link ( https://bugzilla.wikimedia.org/show_bug.cgi?id=62825) [16:58:22] its in the Timemediahandler extension [16:59:41] rdwrer: in that case, what really happen when only {} is passed as function args ? [17:00:53] rmoen: https://gerrit.wikimedia.org/r/#/q/status:open+project:mediawiki/extensions/UploadWizard+topic:bug/47161,n,z [17:01:42] rdwrer: womp womp [17:01:52] And how. [17:02:08] tonythomas: So this is a TMH bug [17:02:12] Why are we talking in -visualeditor [17:03:43] rdwrer: oh! Looks like I'm in the wrong place, and since it was JS. can we switch to #wikimedia-dev ? I have already posted the question there some minutes ago [17:12:45] Hi, I'd like to participate in GSoC 2014. And this project need some changes to the parser and Parsoid. I wonder how can I get to know the parser. https://www.mediawiki.org/wiki/GSOC_2014#Visual_translation:_Integration_of_page_translation_with_VisualEditor [17:22:23] fantasticfears, translation can be done on the HTML; I don't think much parser work is needed [17:23:19] there are some gadgets that already do that using google translate, and https://www.mediawiki.org/wiki/Content_translation is aiming to generalize this [17:23:30] parsoid is dealing with parsing and serializing [17:24:29] yep, I know. I listen the VisualEditor team talk in wikimania [17:24:46] i am going to check the link for more info [17:25:11] RoanKattouw, are you back at your keyboard yet? I've got a redlink bug for you... [17:32:37] RoanKattouw, I filed https://bugzilla.wikimedia.org/show_bug.cgi?id=62877 for you. NB that it's related to 59196, which is open but can't be reproduced any longer. It's now only affecting redlinks. [17:34:55] gwicke: that's pretty cool and a more concise workflow for translation work. it seems we can avoid doing work for change Extension:Translate for it [17:41:48] fantasticfears, the existing gadget is linked from https://www.mediawiki.org/wiki/Parsoid/Users#Gadgets [17:54:38] gwicke: how this extension take care of multi-language page, like on meta which use [17:58:29] gwicke: what Niklas want to do in this project is actually to change the way we tackle the page translation, not related to Content translation which is used for other language wiki. [18:02:20] whatami: Back now [18:03:24] fantasticfears, so far I managed to stay away from the stuff [18:03:29] I need to buy you and ryasmeen a round of drinks or whatever some time. You're both disturbingly good at doing weird things to software that I'd never imagined could be done, then finding that they break [18:03:40] I don't know what the longer-term plans are there [18:04:04] just got the impression that it's pretty messy from an architectural POV [18:04:12] whatami: Hmm, so is that bug specific to red links? From the description I suspect it might also apply to blue links [18:04:42] Blue links were working okay for me… but that problem was a bit intermittent. [18:05:03] You probably need to be on a subpage [18:05:30] Of course, and then it seems to me that there was another detail that mattered. [18:05:32] What I wonder is, if you are on [[User:Example/Subpage]] and create a link to [[Nonexistent page]] versus [[Existing page]], do they behave differently [18:05:42] But today I was only able to get redlinks to fail. [18:05:49] (Links on subpages are known to have and have had all sorts of issues) [18:11:49] RoanKattouw, I think it's probably a duplicate. A bluelink to [[Time]] at beta has the same problem. [18:11:57] Aha [18:12:03] I don't know that we have a bug for this specifically [18:12:09] The problem at Beta is finding something that will actually make a blue link. [18:12:14] haha [18:12:21] bug 59196. [18:12:43] Which also contains exact instructions, from Codename Lisa, about how to dupe it. [18:12:45] whatami: One more thing: can you check if there is a behavior difference between a link already present in the wikitext vs a link inserted you just inserted using VE? [18:12:53] There is. [18:13:03] For blue links, at least. [18:13:06] New ones are broken but pre-existing ones work fine, I take it? [18:13:13] For blue links, at least. [18:13:19] Let me check on a red link. [18:15:00] Right. Pre-existing redlinks work like pre-existing blue links. Red or blue, just-created links take you to a subpage rather than the correct/linked page. [18:16:12] (03CR) 10Krinkle: Add new 'vehidebetadialog' GET parameter to not show the beta dialog (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 (owner: 10Jforrester) [18:19:06] Hm.. ve integration is consistently broken on laster master for me [18:19:07] Krinkle: Well, the browser tests were definitely having multiple beta dialogs showing… [18:19:17] Krinkle: oh? [18:19:23] TypeError: Cannot read property 'tabMessages' of null TypeError {stack: (...), message: "Cannot read property 'tabMessages' of null"} [18:19:31] Exception thrown by ext.visualEditor.viewPageTarget.init [18:19:42] WE used to sometimes get that in the unit tests, never on a view page before though [18:19:44] Krinkle: Is wgVisualEditorConfig not set or something? [18:21:14] whatami: So this is the same bug that we've had with subpages for a while, it's been vaguely on my radar [18:21:40] I haven't had an epiphany for a perfectly clean solution yet [18:22:22] I'll put "one epiphany for Roan" on my list of things to request if I find a genie in a bottle today. [18:22:44] :-) [18:23:38] This, and the search-engine-caused fact that it doesn't know that some pages really do exist, are the only redlink-related problems I've seen. [18:23:57] That's good :) [18:24:01] I'm overall pretty happy with this feature. Thanks. [18:24:08] Especially since I can blame both of them on other things ;) [18:24:17] (Although the subpage thing is also my fault so that doesn't really help me) [18:24:30] RoanKattouw: Using OpenSearch is also your fault. :-) [18:24:42] No, no, it's my husband's fault. Everything is his fault. [18:24:46] RoanKattouw: lol [18:24:47] $wgLegacyJavaScriptGlobals = 4; [18:24:54] mw.config.values === 4 [18:25:03] 1) not interperted as boolean [18:25:10] 2) mw.Map doesn't ensure values is an object [18:25:19] Local config snafu [18:25:30] from when I was testing https://gerrit.wikimedia.org/r/#/c/119706/ [18:25:32] nvm :) [18:26:20] Timo's rule for the day: Don't use === when == will do. ;-) [18:26:21] Krinkle: It cracks me up that you actually had a boolean config variable set to 4 in your config :) [18:26:34] [18:26:42] It was an easy way to bump a value that was part of the hash without it changing behaviour [18:26:50] since tit'd be truethy [18:26:52] Or so you thought [18:26:56] eah [18:26:57] yeah [18:26:57] (re without changing behavior) [18:31:46] hm [18:31:53] I just ran into some issues with parsoid [18:34:02] Pulled the latest version, got issues about missing simplediff. Installed it and everything seems to be fine until I save an edit [18:34:23] Krenair: Ah, yeah. [18:34:28] hi [18:35:10] Bold, category and section markers going missing, link targets getting replaced with null [18:35:11] Krenair: You need to do npm install after each pull. [18:35:32] Krenair: 'Cos their local modules might have changed. [18:35:34] Hey Elitre. [18:35:35] and also restart Parsoid itself [18:35:38] That too. :-) [18:35:41] do you know when the next quarterly review for VE and Parsoid is going to happen? [18:35:57] Also, if npm install misbehaves or weird things happen, you could try rm -rf node_modules ; npm install to scorch the earth and start over [18:36:00] Elitre: Wednesday (6 days' time) for VE, Friday (a week tomorrow) for Parsoid. [18:36:31] Hm. Seems to be fixed now [18:36:53] Thanks James_F [18:37:05] (03CR) 10Krinkle: [C: 04-1] "Where previously this cookie was only set if the wiki has showBetaWelcome enabled in LocalSettings, it is not set always." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 (owner: 10Jforrester) [18:37:32] (03CR) 10Krinkle: "it is non / it is now set *" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 (owner: 10Jforrester) [18:38:08] thanks, PM! is that streamed or office-houred (new word!) ? [18:38:30] Elitre: No, but the minutes are published after the fact. [18:39:25] ok. tnx! [18:40:47] RoanKattouw: https://bugzilla.wikimedia.org/buglist.cgi?columnlist=bug_severity%2Cpriority%2Ccomponent%2Cassigned_to%2Ctarget_milestone%2Cbug_status%2Cresolution%2Cshort_desc&list_id=288922&order=priority%2Cbug_status%20DESC%2Cbug_severity&product=VisualEditor&query_format=advanced&target_milestone=VE-deploy-2014-03-27 [18:41:51] (03CR) 10Krinkle: Follow redirects in the category widgets (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 (owner: 10Alex Monk) [18:43:13] Krinkle, just prepend 'Category:' and it will be localised for you? [18:43:48] Yep [18:43:56] localised or normalised, both. [18:44:09] (e.g. Image -> File, File -> Bestand (nlwiki)) [18:46:28] (03PS2) 10Jforrester: Add new 'vehidebetadialog' GET parameter to not show the beta dialog [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 [18:46:35] (03CR) 10Jforrester: "PS2 is a rebase." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 (owner: 10Jforrester) [18:49:12] (03PS3) 10Jforrester: Add new 'vehidebetadialog' GET parameter to not show the beta dialog [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 [18:49:37] (03CR) 10Jforrester: "PS3 addresses the not-in-line code review. :-)" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/119217 (owner: 10Jforrester) [18:58:42] Krinkle: mmm "There was an error sending mail from 'bugzilla-daemon@wikimedia.org' to 'krinklemail@gmail.com': error when closing pipe to /usr/lib/sendmail: Unknown Error (1)" [18:58:48] error on bugzilla i just got [18:58:57] Yeah BZ is just broken [18:59:03] James got the same error with Moriel's address [18:59:09] booo [18:59:09] why is it using sendmail [18:59:13] Krinkle: broke bugzilla [18:59:26] it's supposed to be on smtp last I checked [19:07:32] (03PS1) 10Catrope: Shut up warnings about calling cloneNode() with no parameters [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119799 [19:11:49] RoanKattouw: Tell TrevorP|Away "Thanks, Obama" [19:13:05] rdwrer: I don't follow [19:14:45] (03PS3) 10Krinkle: Use registry for tool group types [oojs/ui] - 10https://gerrit.wikimedia.org/r/119744 (owner: 10Esanders) [19:17:53] (03CR) 10Krinkle: "Breaking Change: Incompatible signature for constructor of public class OO.ui.Toolbar." [oojs/ui] - 10https://gerrit.wikimedia.org/r/119744 (owner: 10Esanders) [19:21:26] RoanKattouw: Re bash update [19:21:41] Aaah :D [19:23:44] Krinkle: +1 to adding [BREAKING CHANGE] to OOjs UI changes that are breaking. [19:23:48] Krinkle: We did it before. [19:23:56] k [19:24:07] Krinkle: If I add it, happy to +2? [19:24:27] Or should we let Ed get back tomorrow and give input? [19:25:34] James_F: haven't reviewed the rest yet [19:25:55] Krinkle: I meant whether you felt that you + me + previous use was enough for me to change it now? [19:26:03] James_F: I noticed it seems to introduce a new pattern to registering in factories. Might want a second opinion for arch from Roan. [19:26:08] Sure [19:26:14] Kk, will change. [19:26:33] (03PS4) 10Jforrester: [BREAKING CHANGE] Use registry for tool group types [oojs/ui] - 10https://gerrit.wikimedia.org/r/119744 (owner: 10Esanders) [19:27:49] We usually register from the class file, which is ugly, but conventional. Aside from the downside (loading class means it self-registers in one of the factories, which might be the wrong one). The other thing is that that pattern requires there be a global and public singleton instance of the factory, whereas this commit makes the factory a parameter (I like). [19:28:30] But it means that the registering has to happen in the constructor of this class, we'd have to repeat that when we use it elsewhere. And it has to do each of them conditionally because the factory already exists and we don't want to override a custom one. [19:28:40] Krinkle: Don't tell me, tell edsanders|away. :-) [19:28:49] So nothing bad, just a couple new patterns we might want to discuss a bit more before settling :) [19:28:56] RoanKattouw: ^ [19:28:57] * James_F nods. [19:29:10] Krinkle: He's fixing prod after the wmf19 melt-down, bit busy. [19:30:52] Krinkle: We should discuss that later [19:31:10] Krinkle: Also, did you see I reverted the commit that edsanders|away wrote and you merged that completely broke oojs-ui? [19:31:33] It turns out that using classname.super.prototype.methodname.call doesn't work for mixin methods ;) [19:31:56] Relatedly, I think we should have Jenkins/Grunt jobs that run the OOJS-UI demo in phantomjs and assert that there are no JS errors [19:32:33] RoanKattouw: Yep, saw that yesterday [19:32:38] OK [19:32:46] Just making sure you're aware and understood what went wrong [19:32:56] Relatedly relatedly, we shout have a jenkins job that takes a screenshot and diffs it with previous one to highlight UI changes to widgets. [19:32:56] RoanKattouw: I blame the comments [19:33:13] Yeah, "parent method", I feel ya [19:33:25] RoanKattouw:https://test2.wikipedia.org/wiki/March19th [19:33:44] RoanKattouw: Did you see I squashed all those? We no longer call mixin methods anywhere anymore. [19:33:57] Krinkle: Wat? [19:33:58] I was going to fix the comments, but ended up improving the code instead [19:34:37] Link? [19:34:52] Also, I realized that my changes to GroupWidget (which you merged) are completely wrong and I need to revert them [19:34:57] GroupWidget is a mixin meant to be used for Widgets [19:35:09] So it hijacking a Widget method makes sense , and it extending another mixin also makes sense [19:35:13] It just needs to be documented better [19:35:16] It's an abstract class [19:35:22] It did say that [19:35:33] But that can mean it is either extended or mixed in [19:36:07] RoanKattouw: I must\ve dreamed that i did that [19:36:08] (03PS1) 10Catrope: Revert "GroupWidget: Fix messy inheritance chain" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119805 [19:36:24] I was going to remove OO.ui.GroupWidget.prototype.setDisabled, and let the mixinClass stay as-is [19:36:35] and instead do the little thing via an event listener [19:37:07] wait, that was a different class [19:37:09] ryasmeen: I just successfully saved an edit to that page. What do you have to do to break it? [19:37:11] Krinkle: Dream-code can't get merged. :-) [19:37:53] RoanKattouw: Yeah that's what I was telling you , it worked for me later :) [19:39:08] (03CR) 10Krinkle: "How do I reproduce that warning? Opening op veaction=edit on latest master in Firefox, and fiddling around with inspectors and dialogs doe" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119799 (owner: 10Catrope) [19:39:55] (03PS1) 10Catrope: Add documentation explaining what's going on in GroupWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/119806 [19:41:00] (03CR) 10Catrope: "It may be version-dependent. Moriel reproduced it, I don't know how. You have to trigger a call to ve.createDocumentFromHtml which means y" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119799 (owner: 10Catrope) [19:45:18] (03CR) 10Krinkle: [C: 032] Revert "GroupWidget: Fix messy inheritance chain" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119805 (owner: 10Catrope) [19:45:40] (03CR) 10Krinkle: [C: 031] Add documentation explaining what's going on in GroupWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/119806 (owner: 10Catrope) [19:45:49] (03Merged) 10jenkins-bot: Revert "GroupWidget: Fix messy inheritance chain" [oojs/ui] - 10https://gerrit.wikimedia.org/r/119805 (owner: 10Catrope) [19:47:09] James_F: https://gerrit.wikimedia.org/r/119807 [19:47:56] rdwrer: Do you remember our chat about wgHiddenPrefs and Beta Features? I can't remember whether we worked out if it worked… [19:48:27] Heh, not sure [19:48:33] Don't remember [19:55:34] rdwrer: Oh well. [20:04:24] (03PS6) 10Alex Monk: Follow redirects in the category widgets [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 [20:05:27] RoanKattouw, Krinkle: ^ [20:05:40] Yay [20:06:01] Oh, wait, one more issue [20:06:12] (03CR) 10jenkins-bot: [V: 04-1] Follow redirects in the category widgets [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 (owner: 10Alex Monk) [20:06:15] When you type in the exact name of a redirect you should have an exact match. ugh this is fiddly [20:07:01] ... ugh, and that [20:09:05] (03CR) 10Krinkle: Follow redirects in the category widgets (032 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 (owner: 10Alex Monk) [20:23:47] (03PS7) 10Alex Monk: Follow redirects in the category widgets [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 [20:24:43] RoanKattouw_away, okay, I think it works basically bug-free now. [20:25:18] James_F: thanks for the red links fix [20:25:26] can it be tested on mediwiki.org? [20:25:53] matanya: Should be live there now. [20:26:06] matanya: Also, I didn't write the code, so… :-) [20:26:16] * matanya thanks RoanKattouw_away  [20:26:21] :-) [20:26:32] (03CR) 10Krinkle: Follow redirects in the category widgets (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/118218 (owner: 10Alex Monk) [20:26:52] Krenair: in is not indexOf or inArray, it's hasKey [20:27:02] Krinkle, what the hell was I thinking when I wrote that [20:27:13] :) [20:27:25] Krinkle, I was probably thinking of python or something. wtf. [20:29:20] James_F: works great! even in RTL :) [20:31:37] matanya: But of course. mooeypoo keeps us honest about RTL. :-) [20:43:16] Krinkle, actually I'm wondering what the proper way to do that check is [20:44:17] Krenair: ve.indexOf ? [20:44:26] If its an array that is [20:44:31] It's an array of objects [20:45:49] Krenair: Each object is one from/to pair? [20:46:09] Yeah [20:46:21] Krenair: I'd merge them, like you had. [20:46:32] Why did you move to this? [20:46:34] >>> var a = [{'a': 'b'}]; a.indexOf( {'a': 'b'} ) [20:46:34] Krenair: (number) -1 [20:47:00] reference [20:47:05] yeah [20:47:35] Krinkle, merge them so I had a single object of from -> to? [20:48:42] { a: b, c: d, a: d } [20:48:43] yeah [20:48:49] redirects[ from ] = to [20:50:01] Krenair: Array.indexOf and lastIndexOf is ES5 I'm afraid. Be sure to add them to the featur etest, or use jQuery methods instead. [20:50:10] ve.indexOf instead of array.indexOf (for now) [20:50:22] lastindexOf we dont have an alternative for, so if you want to adopt that, be sure to add it to the feature test [20:50:26] it's fine, we require es5 [20:50:47] Krinkle, this code was already using lastIndexOf [20:50:49] just making sure we keep track of which features we use so that we don't break when we polyfil one day [20:50:56] Bad code, good point. [20:52:39] Krenair: String or Array [20:52:51] string indexof is js 1.0, but array indexof and lastindexof are new in es5 [20:53:03] string lastindexof is es3 [20:53:40] I don't think it was using it for arrays beofre, right? [20:54:23] Krinkle, no, it wasn't... hm [21:00:02] RoanKattouw_away: lolwut [21:00:03] Firefox [21:00:05] 21:59:18.871 ve.createDocumentFromHtml('

foo

') [21:00:06] 21:59:18.873 [Exception... "The operation is insecure." code: "18" nsresult: "0x80530012 (SecurityError)" location: ""] [21:02:00] From newDocument.open(); [21:05:17] setting .src = 'about:blank'; fixes it [21:05:26] Oh, right, we don't want this path for Firefox, [21:05:30] mozilla bug 867102 [21:09:47] Hm.. no, that applies to the createHTMLDocument path [21:19:28] (03PS5) 10Jforrester: Basic toolbar demo [oojs/ui] - 10https://gerrit.wikimedia.org/r/119516 (owner: 10Esanders) [21:19:32] (03PS5) 10Jforrester: Styling for disabled toolgroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119517 (owner: 10Esanders) [21:19:36] (03PS5) 10Jforrester: Force visible label for MenuToolGroups [oojs/ui] - 10https://gerrit.wikimedia.org/r/119518 (owner: 10Esanders) [21:20:04] Krinkle: Any chance you could respond to my CR on https://gerrit.wikimedia.org/r/#/c/118607/ so I can merge (or fix)? [21:20:19] Krinkle: (It's just the TemplateData licensing note change.) [21:20:31] I think we should GNU v2 "or later" [21:20:49] OK, will change. [21:21:02] The license is the same either way. [21:21:06] Hmm. Could you state this on gerrit, given it's effectively a licensing change. [21:21:25] The license text essentially contains "If referred to this file as "or later", then the following clause takes effect" [21:21:32] Yes. [21:21:36] great licensing [21:21:39] yay [21:22:09] Yeah. [21:22:30] NOthing on https://www.mediawiki.org/wiki/Special:Version has v2+ though [21:22:33] So .. [21:23:06] That's why I assumed you actually meant GPL2, not GPL2+. [21:23:24] BUt don't we generally advocate (and use) 2+? [21:23:36] CentralNotice says v2 in that new spot [21:23:41] loks like a mistake by that recent commit [21:23:45] because it is in fact under 2 and later [21:23:59] * CentralNotice extension @license GNU General Public Licence 2.0 or later [21:24:02] I'm pretty sure we are 2, rather than 2+. [21:24:10] 2+ means we'd get GPL3 forks. [21:24:24] Let's stick to v2 then [21:24:29] I mean, that's what the code had [21:24:38] might need relicense votes for 2+ [21:24:43] * James_F nods. [21:24:43] what does mediawiki core do? [21:24:45] 2+ ? [21:24:45] OK, will +2. [21:25:03] (03CR) 10Jforrester: [C: 032] license: Add COPYING file for GNU GPL v2 [extensions/TemplateData] - 10https://gerrit.wikimedia.org/r/118607 (owner: 10Krinkle) [21:25:32] (03Merged) 10jenkins-bot: license: Add COPYING file for GNU GPL v2 [extensions/TemplateData] - 10https://gerrit.wikimedia.org/r/118607 (owner: 10Krinkle) [21:32:49] Krinkle: 2+. [22:36:30] It feels like every time I solve an issue I run into 2 more. [22:40:06] 99 little bugs in the code, 99 little bugs... Take one down, patch it around, 117 little bugs in the code [22:40:12] Krenair, ^^ [22:41:18] Wow, only 19 regressions? [22:41:26] You guys are getting better. :) [22:44:20] * James_F grins. [23:31:21] (03PS3) 10Trevor Parscal: Update context on transact, in addition to select [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119227 [23:41:25] (03PS4) 10Trevor Parscal: Update context on transact, in addition to select [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119227 [23:42:14] (03PS1) 10Trevor Parscal: Add class to internal item nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119899 [23:42:21] (03PS1) 10Trevor Parscal: Propagate destroy calls down to children of branch nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119900 [23:42:27] (03PS1) 10Trevor Parscal: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 [23:44:15] (03PS5) 10Catrope: Update context on transact, in addition to select [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119227 (owner: 10Trevor Parscal) [23:44:36] (03PS1) 10Trevor Parscal: Split autoAdd into catchall and group specific settings [oojs/ui] - 10https://gerrit.wikimedia.org/r/119902 [23:45:05] (03CR) 10Catrope: [C: 032] Update context on transact, in addition to select [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119227 (owner: 10Trevor Parscal) [23:45:08] (03PS2) 10Trevor Parscal: Split autoAdd into catchall and group specific settings [oojs/ui] - 10https://gerrit.wikimedia.org/r/119902 [23:45:14] (03PS2) 10Catrope: Add class to internal item nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119899 (owner: 10Trevor Parscal) [23:45:20] (03CR) 10Catrope: [C: 032] Add class to internal item nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119899 (owner: 10Trevor Parscal) [23:45:27] (03PS2) 10Catrope: Propagate destroy calls down to children of branch nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119900 (owner: 10Trevor Parscal) [23:45:40] (03PS3) 10Catrope: Propagate destroy calls down to children of branch nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119900 (owner: 10Trevor Parscal) [23:46:08] (03Merged) 10jenkins-bot: Update context on transact, in addition to select [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119227 (owner: 10Trevor Parscal) [23:46:16] (03Merged) 10jenkins-bot: Add class to internal item nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119899 (owner: 10Trevor Parscal) [23:46:59] (03CR) 10Catrope: [C: 032] Propagate destroy calls down to children of branch nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119900 (owner: 10Trevor Parscal) [23:47:03] (03PS2) 10Catrope: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:47:12] (03PS3) 10Catrope: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:47:56] (03Merged) 10jenkins-bot: Propagate destroy calls down to children of branch nodes [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119900 (owner: 10Trevor Parscal) [23:49:19] (03CR) 10Catrope: [C: 04-1] Use transaction comparison to decide to render or not (031 comment) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:51:35] (03PS4) 10Trevor Parscal: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 [23:51:51] (03CR) 10Trevor Parscal: [C: 032] Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:51:54] (03PS5) 10Trevor Parscal: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 [23:55:08] (03PS6) 10Catrope: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:55:14] (03CR) 10Catrope: [C: 032] Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal) [23:56:08] (03Merged) 10jenkins-bot: Use transaction comparison to decide to render or not [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/119901 (owner: 10Trevor Parscal)