[00:02:05] 3VisualEditor, VisualEditor-MediaWiki-Mobile: [Regression wmf3] Extra wide space appearing between the search box and the list of existing references inside Re-Use dialog in mobile VE - https://phabricator.wikimedia.org/T74110#808863 (10Krenair) Cannot reproduce on enwiki. Cannot access mobile VE on testwiki. [00:11:09] 3Verified-in-Phase0, VisualEditor-EditingTools, VisualEditor-MediaWiki, VisualEditor: [Regression wmf10] Cannot open any inspectors, they just flicker and get closed - https://phabricator.wikimedia.org/T76076#808884 (10Ryasmeen) Verified in en.wiki [00:11:40] 3Verified-in-Phase0, Verified-in-Phase2, VisualEditor-EditingTools, VisualEditor-MediaWiki, VisualEditor: [Regression wmf10] Cannot open any inspectors, they just flicker and get closed - https://phabricator.wikimedia.org/T76076#808887 (10Ryasmeen) [00:14:07] 3Verified-in-Phase0, VisualEditor, Verified-in-Phase2, Beta-Cluster: [Regression pre-wmf10] upload.beta.wmflabs.org is throwing 503s so all images are appearing with a broken icon inside VE - https://phabricator.wikimedia.org/T75786#808901 (10Ryasmeen) [00:15:26] 3Verified-in-Phase0, VisualEditor, Verified-in-Phase2, Beta-Cluster: [Regression pre-wmf10] upload.beta.wmflabs.org is throwing 503s so Math function parsing is completely broken inside VE - https://phabricator.wikimedia.org/T75787#808903 (10Ryasmeen) [00:16:26] 3VisualEditor, Verified-in-Phase0, Verified-in-Phase2: [Regression pre-wmf10] VisualEditor does not load on beta labs - https://phabricator.wikimedia.org/T75896#808904 (10Ryasmeen) [00:17:10] 3Verified-in-Phase0, VisualEditor, VisualEditor-EditingTools: [Regression pre-wmf10] Cannot open any Citation dialog, TypeError: this.bookletLayout.focus is not a function appears in the console - https://phabricator.wikimedia.org/T75779#808905 (10Ryasmeen) Verified in production [00:17:47] 3Verified-in-Phase0, VisualEditor, Verified-in-Phase2, VisualEditor-EditingTools: [Regression pre-wmf10] Cannot open any Citation dialog, TypeError: this.bookletLayout.focus is not a function appears in the console - https://phabricator.wikimedia.org/T75779#808906 (10Ryasmeen) [00:18:33] 3VisualEditor, VisualEditor-MediaWiki, Verified-in-Phase0: [Regression wmf10] Switch to source edit option is disabled - https://phabricator.wikimedia.org/T75965#808907 (10Ryasmeen) Verified in en.wiki [00:18:47] 3Verified-in-Phase2, VisualEditor, VisualEditor-MediaWiki, Verified-in-Phase0: [Regression wmf10] Switch to source edit option is disabled - https://phabricator.wikimedia.org/T75965#808908 (10Ryasmeen) [00:19:09] 3Verified-in-Phase0, VisualEditor, VisualEditor-MediaWiki: [Regression wmf10] Switching to Edit source mode is not keeping changes done in VE - https://phabricator.wikimedia.org/T76114#808909 (10Ryasmeen) Verified in en.wiki [00:19:20] 3Verified-in-Phase0, Verified-in-Phase2, VisualEditor, VisualEditor-MediaWiki: [Regression wmf10] Switching to Edit source mode is not keeping changes done in VE - https://phabricator.wikimedia.org/T76114#808912 (10Ryasmeen) [00:20:20] 3Verified-in-Phase0, VisualEditor-Tables, VisualEditor-EditingTools, VisualEditor, MediaWiki-extensions-UniversalLanguageSelector: IME selector shows up at the bottom-right corner of the table while selecting a cell on a table or placing a cursor into it - https://phabricator.wikimedia.org/T74824#808917 (10Ryasme... [00:20:32] 3Verified-in-Phase0, Verified-in-Phase2, VisualEditor-Tables, VisualEditor-EditingTools, VisualEditor, MediaWiki-extensions-UniversalLanguageSelector: IME selector shows up at the bottom-right corner of the table while selecting a cell on a table or placing a cursor into it - https://phabricator.wikimedia.org/T748... [00:20:55] 3VisualEditor, Verified-in-Phase0, VisualEditor-EditingTools: Add support for
s - https://phabricator.wikimedia.org/T53009#808919 (10Ryasmeen) Verified in en.wiki [00:21:07] 3Verified-in-Phase0, Verified-in-Phase2, VisualEditor-EditingTools, VisualEditor: Add support for
s - https://phabricator.wikimedia.org/T53009#808921 (10Ryasmeen) [00:21:28] 3Verified-in-Phase0, VisualEditor: [Regression pre-wmf10] Cannot close Citation dialog , error in the console:TypeError: this.outlineSelectWidget is undefined - https://phabricator.wikimedia.org/T75942#808922 (10Ryasmeen) Verified in en.wiki [00:21:40] 3Verified-in-Phase0, VisualEditor, Verified-in-Phase2: [Regression pre-wmf10] Cannot close Citation dialog , error in the console:TypeError: this.outlineSelectWidget is undefined - https://phabricator.wikimedia.org/T75942#808923 (10Ryasmeen) [00:22:02] 3VisualEditor-MediaWiki, VisualEditor: VisualEditor: Oldid in navigation links isn't updated after saving - https://phabricator.wikimedia.org/T74461#808924 (10Arlolra) [00:22:06] 3Verified-in-Phase0, VisualEditor, VisualEditor-EditingTools: [Regression pre-wmf10] Toolbar for Basic Reference and Media Settings dialog is broken - https://phabricator.wikimedia.org/T76073#808925 (10Ryasmeen) Verified in en.wiki [00:22:18] 3Verified-in-Phase0, VisualEditor, Verified-in-Phase2, VisualEditor-EditingTools: [Regression pre-wmf10] Toolbar for Basic Reference and Media Settings dialog is broken - https://phabricator.wikimedia.org/T76073#808926 (10Ryasmeen) [00:23:05] 3Verified-in-Phase0, VisualEditor, VisualEditor-MediaWiki: After updating a page
gets added after the timestamp of last edit made on that page - https://phabricator.wikimedia.org/T75571#808927 (10Ryasmeen) [00:23:16] 3Verified-in-Phase0, VisualEditor, VisualEditor-MediaWiki, Verified-in-Phase2: After updating a page
gets added after the timestamp of last edit made on that page - https://phabricator.wikimedia.org/T75571#808928 (10Ryasmeen) [00:26:01] 3VisualEditor-MediaWiki-References, Verified-in-Phase0, VisualEditor: [Regression wmf10] Basic reference that has Citation inside tries to render it as Basic Reference ({{Cite web }})and thus fails - https://phabricator.wikimedia.org/T76082#808930 (10Ryasmeen) Verified in en.wiki [00:26:22] 3Verified-in-Phase2, VisualEditor-MediaWiki-References, Verified-in-Phase0, VisualEditor: [Regression wmf10] Basic reference that has Citation inside tries to render it as Basic Reference ({{Cite web }})and thus fails - https://phabricator.wikimedia.org/T76082#808931 (10Ryasmeen) [00:29:43] (03CR) 10Esanders: [C: 032] Fix call to mw.log.warning [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/177421 (owner: 10Alex Monk) [00:29:54] 3MediaWiki-extensions-WikiEditor: Search&replace disabled in IE8, Opera - https://phabricator.wikimedia.org/T25992#808943 (10GOIII) [00:31:49] (03Merged) 10jenkins-bot: Fix call to mw.log.warning [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/177421 (owner: 10Alex Monk) [00:34:35] (03PS6) 10Esanders: [BREAKING CHANGE] Refactor for multiple surfaces [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/176952 [00:34:37] (03PS5) 10Esanders: [BREAKING CHANGE] Multiple surface support and demo [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177109 [00:34:39] (03PS5) 10Esanders: Refactor out trigger listening to allow target triggers [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/176955 [00:38:23] 3Verified-in-Phase0, OOjs-UI, Verified-in-Phase2: MessageDialog doesn't fit actions correctly - https://phabricator.wikimedia.org/T75376#808957 (10Etonkovidova) [00:49:26] 3VisualEditor-MediaWiki, VisualEditor: VisualEditor: Oldid in navigation links isn't updated after saving - https://phabricator.wikimedia.org/T74461#808967 (10Krenair) [00:57:30] (03PS2) 10Catrope: Use 'always' when removing windows [oojs/ui] - 10https://gerrit.wikimedia.org/r/177406 (owner: 10Esanders) [00:58:48] (03CR) 10Catrope: [C: 04-1] Use 'always' when removing windows (031 comment) [oojs/ui] - 10https://gerrit.wikimedia.org/r/177406 (owner: 10Esanders) [00:59:59] [15:59] Krenair RoanKattouw_away, another issue with the transclusion link existence check - mobile VE [01:00:01] Aaargh [01:00:08] Krenair: I guess we should be looking at the base URL of the document as well? [01:00:13] yeah [01:00:34] Or maybe s/as well/instead/ [01:00:41] Why do we even look at wgArticlePath [01:02:42] does parsoid consider interwiki links WikiLink? if so I think that would be affected as well [01:03:51] I think they're ExtLink insteaad [01:04:06] Let me git blame this [01:04:16] Because I wrote this and now I want to know why I did that [01:05:13] Krinkle: Srsly, could you please remove your -1 on https://gerrit.wikimedia.org/r/#/c/175180/ – it's way too late to suddenly suggest this and you're blocking development. [01:05:14] 3MediaWiki-extensions-WikiEditor: Hide signature button from WikiEditor toolbar in content namespaces - https://phabricator.wikimedia.org/T59727#808985 (10GOIII) [01:05:37] Krenair: Oh Ed wrote it; anyway yeah please feel free to use the Parsoid document base href instead of wgArticlePathh theeerre [01:05:58] I guess that would previously have been difficult because the base href wasn't the article path, it was the page URL [01:07:26] It's 1AM and I have to be up at 7. So I'll deal with this tomorrow evening [01:07:51] Yeah for sure [01:10:55] (03CR) 10Jagori79: "Hi Zeljko," (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/122400 (owner: 10Jagori79) [01:13:17] (03PS1) 10Catrope: Fix path to MobileWindowManager.js [extensions/VisualEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177453 [01:15:22] (03Abandoned) 10Jagori79: WIP [BrowserTest] Use rspec-expectations expect syntax instead of should syntax [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/146712 (https://bugzilla.wikimedia.org/66369) (owner: 10Jagori79) [01:17:10] James_F: https://gerrit.wikimedia.org/r/177453 :S [01:19:16] (03CR) 10Jforrester: [C: 031] Fix path to MobileWindowManager.js [extensions/VisualEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177453 (owner: 10Catrope) [01:31:54] James_F: There's no way that's blocking development. It's not referenced anywhere. [01:32:39] Also, if it's so close to usage, obviously there's drafts somewhere waiting for review that show how we'll use this code outside demos? [01:33:23] I've yet to see any thing that smells like production code for OOjs UI usage. I'm just suspicious about that. I'd like to see it first (the "deveopment" I'm supposedly blocking shouldn't be blocked. Let it happen. Check it out locally and develop.) [01:34:50] Krinkle: Usage is the demos, as well you know. [01:34:55] Krinkle: You're just being obtuse. [01:35:11] James_F: No I'm not. [01:35:14] James_F: I've seen the demos yet. [01:35:27] Krinkle: You have? Great! Could you please put them in gerrit? [01:35:32] (03PS1) 10Catrope: Update IE detection to include Edge [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177461 [01:35:36] yes* [01:36:02] Krinkle: 'Cos they aren't wired into MW yet because this hasn't landed, so instead they are toy ones that don't prove it works. [01:36:24] What are you referring to? [01:36:31] I assume this is meant to be used in core or some extension. [01:36:38] Which one is first waiting to adopt this? [01:36:46] Krinkle: What? [01:37:03] Krinkle: The objective is to get OOphp UI in core. [01:37:13] Krinkle: The second objective is to start to explore using it in core. [01:37:17] Not just for the sake of it, that acomplishes nothing. [01:37:21] Krinkle: And yet. [01:37:24] That can't be a goal. [01:37:29] Krinkle: And yet. [01:37:33] Where it is planned to be used? I assume there's one or many answers ready for that on your lips. [01:37:45] And in Phabricator, if you cared to look. [01:37:59] Start from https://phabricator.wikimedia.org/T74714 and have fun. [01:38:02] So explore using it in core with a commit that depends on it. Trivial. [01:38:16] I'm not saying I want to know where it'll be used. I know that. [01:38:19] I'm saying how. [01:38:25] I want to avoid the same mistakes again. [01:38:27] Fair? [01:38:39] I'm saying no-one knows, and by blocking it you're slowing that testing down. [01:38:41] One simple example of a draft in core that explores usage and shows it working. [01:38:55] If nobody knows how it'll be used yet, I"m not blocking anything. this is premature. [01:39:12] But this is the epically-confusing world of composer where everything is broken by design. [01:39:23] 3MediaWiki-extensions-WikiEditor: WikiEditor: ext.wikiEditor.toolbar.hideSig.js is loaded even when the enhanced editing toolbar is disabled - https://phabricator.wikimedia.org/T35923#809016 (10GOIII) [01:39:26] Whoever wants to use it can by using a git dependency. Tell me that's not raising the bar? [01:40:56] Does that even work? [01:41:02] Of course it does. [01:41:32] If you develop against core and you have wmf/s vendor, then it already works. If you use composer locally, then you already are in the habit of running cmposer-update after each checkout/pull/git-review-D in mediawiki-core anyway. [01:41:58] And if no-one touches this? [01:42:29] Which, as you know, is the case. [01:42:37] touches what? [01:42:47] Because Platform are too busy to help with this, except for occasional (and helpful) fly-by commits. [01:42:49] Vendor. [01:43:20] This is what Trevor wanted a Platform person to lead, but unfortunately we don't have one, so… [01:43:28] It's all just dark magic and confusion [01:43:37] If nobody is ready to start drafting patches using oojs-ui in core or an extension yet, that wouldn't change by having this commit merged. I'll merge it no questions asked as soon as there's an example usage commit in gerrit for something somewhere that works and Trevor or someone else says is the right way to use it. [01:44:07] You probably shouldn't merge things that touch vendor, surely? [01:44:34] it was already merged in mediawiki-vendor [01:44:38] Isn't it meant to go through the libraries project? I don't know. All is confused. [01:44:51] And it would have been merged in core if it weren't for your meddling. [01:44:54] I'd like to believe I can and am expected to merge changes to mediawiki-vendor and composer.json [01:45:02] OK. [01:45:09] * James_F is deeply suspicious of that area. [01:45:24] Lots of complexity and no track record of how to do things. [01:45:30] True [01:45:35] Which is why I went back and forth [01:45:35] Like with OOjs UI, where we're having to invent it on the fly. [01:45:41] a few times with bd808 to ensure I understand it [01:46:19] If you're sure. [01:46:29] and while they should approve how it works, they shouldn't make product or design decisions implicitly. Which is why bd808 waits for my -1 on that mwcore commit I think. [01:46:33] It'll be fine. [01:46:43] pretend oojs-ui is in core, ready for usage. Let those experiments begin. [01:46:51] Krinkle: Everyone waits on your -1. It's because of respect. [01:47:07] Krinkle: But in return, you should be very careful in using it. :-) [01:47:15] I understand. [01:47:23] Krinkle: Your suggestions of how to use it are welcome. [01:47:27] Maybe I should try something? [01:47:30] * James_F wories. [01:47:53] I'm not the one pushing oojs-ui php in core right now, you are. I'm saying, from mediawiki core side of things, this is ready. [01:48:02] It isn't. [01:48:24] We haven't split the styling and i18n modules out of the JS-land oojs-ui RL module. [01:48:24] But while it's possible it'll be straight forward in how we'll use it, we haven't used it yet. So I'd like to see how that goes before we add it in and start getting a mess. [01:48:42] Only the php classes are loaded by autoloader. [01:48:43] We haven't created the PHP-land not-RL-module-ish-kinda that depends on it too. [01:48:43] composer [01:48:56] Yeah, which is a problem because they also need the resource bundle. [01:49:02] No? [01:49:17] the js/css should remain in resources/ [01:49:25] And if OOphp UI is already used on the page (and so the styles are used), you don't want OOjs UI to reload the styling, do you? [01:49:36] Krinkle: What about i18n? [01:50:12] That's an excellent example of something we haven't tackled yet and why it shouldn't be merged in core as using it would instantly cause questions we haven't solved yet. Or maybe it'll work naturally and we just need to try. [01:50:24] And the styling needs to ship even if you're in a no-JS environment like IE6. [01:50:38] i18n would presumably be used from their own respective modules. [01:50:42] So it can't go through ResourceLoader's normal approach. [01:50:48] Assuming the php modules an have oojs-ui messages [01:50:49] can they? [01:50:49] Krinkle: So we have two copies of the i18n? [01:50:52] Yes. [01:50:53] I don't think they use it yet [01:50:57] OK [01:51:00] Nothing uses it yet. [01:51:04] Because it's not merged. :-) [01:51:11] James_F: That's not what I meant [01:51:13] James_F: We don't need two i18n copies, I've explained this before [01:51:28] James_F: I meant, oojs-ui php doesn't use its own i18n. Only dialogs use that stuff, right? [01:51:45] Krinkle: And buttons and controls and … [01:51:55] Krinkle: And Special:Version when we start listing libraries there. [01:51:58] Krinkle: And whatever. [01:52:06] James_F: They take messages, but do they have messages by default? [01:52:10] I'm pretty sure there's not *currently* anything in OOuiPHP or proposed to be ported there that uses i18n [01:52:24] But yes, this is mostly a theoretical worry of mine right now. [01:52:25] but we should probably still have a game plan for if&when that changes [01:52:29] Yeah. [01:52:31] eventually [01:53:42] James_F: As for having two copies, aside from i18n, the additional cssjs files in the vendor are just dormant. Just like any other non-php file that might exist. composer just copies down the entire repo. There' no obligation or expectation to use those files. We could even try and figure a way to exclude them from the composer install. [01:54:02] unlike npm, composer has no concept of packages. The package is the git repo. [01:54:12] so ther'es no .ignore or exclude mechanism [01:54:43] What I do expect to be a problem is the separation of the stylesheet and javascript indeed. We may have to split it up into two modules. [01:55:22] Or, as I proposed in the FSG, ensure our markup is good enough and just let the default styles of browsers and skins make something of it when in a Grade C browser [01:55:39] Krinkle: That's a laughable WONTFIX, surely? [01:55:55] Krinkle: I mean, the entire objective of this exercise is to consistently style things. [01:56:00] That's extreme, but not uncommon. The "I16 universal stylesheet" comes to mind (an initiative to use conditional comments to serve a generic stylesheet for IE6 instead of your main one) [01:56:05] Krinkle: Did someone not point that out in the meeting? [01:56:08] James_F: ... In IE6? [01:56:11] Krinkle: Yes. [01:56:16] It;ll be consisent. [01:56:23] Krinkle: Not with IE11. [01:56:26] None of them will have fancy styles. [01:56:29] in IE6 [01:56:32] Exactly. [01:56:39] So, WONTFIX. Next suggestion? ;-) [01:56:41] nobody spoke about concisetny styling between browsers. [01:56:49] Oh. They probably should have. [01:56:55] No [01:57:02] there's always degrading and fallback. This is just the last fallback. [01:57:05] It's almost the number one objective from Design, alongside "modern". [01:57:14] e.g no gradients, shadows and rounded corners in some browsers [01:57:21] I think if we go to people and say "no JS also means no styling" there'll be a revolt. [01:57:26] And we're not polyfiling those. [01:57:31] We're talking about the slab-background colours of buttons. [01:57:41] We're talking about the spacing of multiple buttons on a form. [01:57:51] We're talking about the provision of icons on buttons. [01:57:57] I believe the objective is to unify styling of the product for an individual experience. cross-browser consistency is an illusion. I think modern best practices dispute the goal of trying to match it between every device and browser. [01:58:07] yeah I get that [01:58:11] None of these are trivial nice-to-haves for old browsers. [01:58:15] They're pretty fundamental. [01:58:32] But for Grade C browsers. I think it's a realistic option to say: No mediawiki styles. Just a generic stylesheet to pretty up some semantic html. [01:58:37] If we try to merge a Special:Login where the buttons look completely different in IE6, but the same as before in IE11, it's just a regression. [01:58:58] I think that's something that would need a serious discussion and probably an RfC. [01:59:11] It's a pretty major departure from what "no JS" means this month. [01:59:14] Well, this is all RfC-grade material. [01:59:19] Yep. [01:59:40] But it comes down to a cost balance. I think we may not fully realise the cost of trying to not go that route. [01:59:49] For in many ways we've already gone down that road. [01:59:52] The RfC for front-end standardisation has already been had. [02:00:10] But it was agreed in a world where the standard didn't mean also killing styling for low-end browsers. [02:00:13] There is no such thing as an RfC for the "standardisation". There could be, but we didn't have that. [02:00:25] I don't think Trevor would be OK with re-opening that can of worms. [02:00:44] * RoanKattouw does happy dance [02:00:44] What's wrong with OOUIphp depending on oojs-ui.styling or whatever? [02:00:51] Nothing. [02:00:52] I finally got a patch working for jquery.client to recognize IE12 [02:00:54] James_F: And it will. [02:00:56] RoanKattouw: Gosh. [02:01:03] After like 4 attempts [02:01:04] Krinkle: But not for non-JS users? [02:01:07] RoanKattouw: Yay. [02:01:14] The IE12 demo is a remotely hosted app thingy so I can't point it to localhost [02:01:14] Exactly, maybe not for non-JS. [02:01:28] * James_F sighs. [02:01:33] This feels like failure. [02:01:37] The RFC was about the skin eco system to some extend, and the libraries and javascript. We dind't talk abot non-js either way. This is a new concept for MW/RL in general. Not a re-opening of anything. [02:01:45] James_F: It's reality, seriously. [02:02:02] We can do better. [02:02:03] Its' the best we can do in the long term without spending half our resources on Grade C browsers. [02:02:15] We already have the ability to ship stuff for non-JS browsers. [02:02:20] short of dropping them entirely and serving them Grade A stylesheets [02:02:31] ... in a way that bypasses caching. [02:02:42] and ends up hardcoding modules in the page. [02:05:03] I know. [02:05:05] It's for legacy modules and for basic skin styling only. Always has been. [02:05:05] * James_F sighs. [02:05:05] James_F: I'm proposing it, not because I think it's the easy way out or the better for the users. But simply as something to consider that non-engineers might deem the best use of our resources. [02:05:05] * James_F nods. [02:05:05] It's what other companies and apps have done in the past. It's realistic. [02:05:06] Its' the one-up from dropping them entirely with no css and a "please upgrade" (ala Google/YouTube) [02:05:06] True, but we've always held ourselves to much higher standards. [02:05:06] plain html with a semantic html stylesheet [02:05:16] True. [02:05:32] But I think in practise the difference would be minimal. [02:06:16] They already have no interactive dialogs (no-js). It'd just mean putting our content in a Medium.com-like layout. And form controls will be browser-native (like they have been until recently) [02:07:15] Of course the reality is, we already did this effectively. Our basic skin stylesheet is exactly that. [02:07:21] Hmm. [02:07:24] A bit more but not much. [02:08:18] The difference is us not doing something. If we continue as we are, we'll continue to load oojs-ui (And by default loading means via javascript mw.loader) and if html is output by php, that means it has no styles in no-js mode. [02:08:40] the form controls and buttons would look like they do now. [02:09:10] (excpet for the ones currently using mw-ui, those would go back to native form contorls in no-js browsers) [02:09:16] anyway, nuff said :) [02:09:21] We'll cross that bridge when you tell me to. [02:12:24] Krinkle: So, https://gerrit.wikimedia.org/r/177471 as a start. [02:12:59] Krinkle: Next thing is to use the PHP and depend on oojs-ui.styling in top? [02:13:16] James_F: (messages in js) :) [02:13:31] James_F: yeah, write some code that uses the php classes and adds the styling module to the page with addModules() [02:13:33] Krinkle: Messages are assets. :-) [02:13:43] (and position top in the styling module definition) [02:13:46] James_F: messages don't run in css [02:14:12] specifcally javascript messages dont run in css [02:14:16] Krinkle: They might be needed in PHP, as said. In future. [02:14:25] Krinkle: So we should split messages? Eww. OK. [02:14:28] No [02:14:34] James_F: Resources modules are client-side. [02:14:41] this module will be broken [02:15:03] Meh, it happens to work, but that's because .styling isn't used [02:15:05] Oh, wait, I can use the RL module definition to put things into top for server-side use? [02:15:06] it'd be loaded twice if it works [02:15:23] Or not? [02:15:25] James_F: the messages you have in Resources.php for oojs-ui belong to oojs0ui not oojs-ui stlyling, that's all [02:15:57] Better? [02:15:59] James_F: yes, if the module has position -=> top, then doing addModules() will load it from the top. [02:17:06] James_F: messages from resource modules have no relation to server-side use. wfMessage() always has the complete localisation database. [02:17:10] Krinkle: Oh, so I have to put position:top into oojs-ui.styling and that will affect every page that loads it, or only those that load it from PHP? [02:17:24] Krinkle: See updated patchset. [02:19:01] James_F: position => top is only used for modules queued server-side. Once client-side, calling mw.loader.load always loads it "now". The position isn't relevant, can't go back in time :) [02:19:22] Krinkle: Aha. OK, no worries there then. [02:39:58] Krinkle: Do we even try to upstream jquery.client changes, or have we forked too far? [02:41:13] James_F: Actually, I don't think it was a fork? [02:41:20] Didn't Neil and Trevor author that? [02:41:25] Krinkle: Did they? [02:42:12] https://github.com/wikimedia/mediawiki/commit/3bb421904a616f51918295e18f3e3e73fda986f8 [02:42:12] There's nothing on the file. [02:42:19] James_F: Yep [02:42:19] Roan and Trevor [02:42:19] formerly mw.util.client even [02:42:22] wow, I forgot about that [02:42:30] Huh [02:42:44] Krinkle: Roan denies this. [02:43:09] https://github.com/wikimedia/mediawiki/blob/4a6a90f33db58232e075a29be5a6c003a964657e/resources/mediawiki.util/mediawiki.util.client.js [02:43:18] it's been there since the resouceloader merge [02:43:53] Hmm. [02:45:21] following pre-merge [02:45:21] https://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/resources/jquery/jquery.client.js?view=log [02:46:45] RoanKattouw: I don't think I ever saw this early code of RL [02:46:47] https://svn.wikimedia.org/viewvc/mediawiki/branches/resourceloader/phase3/resources/mw/mw.js?revision=69445&view=markup&pathrev=70169 [02:46:51] mw.loader = new mw.prototypes.ResourceLoader(); [02:46:54] O_O [02:46:57] Whoa what [02:46:59] mw.msg = new mw.prototypes.Language(); [02:47:00] Wow. [02:47:03] I don't think *I* ever saw that [02:47:20] Frightening. [02:47:21] For quite a while Trevor just worked on mw.loader off in a corner and I only reviewed the semi-finished product [02:47:24] https://svn.wikimedia.org/viewvc/mediawiki/branches/resourceloader/phase3/resources/mw/mw.js?view=log&pathrev=70169 [02:47:27] yeah [02:47:29] RoanKattouw: Yay SVN. [02:47:49] Ooh right [02:47:59] I remember now that I was already making fun of Trevor back then [02:48:02] "cononical name" [02:48:40] RoanKattouw: Clearly it should have been cone-nonical? [02:48:49] RoanKattouw: You did see some of it [02:48:52] Here's an angry message from you [02:48:58] "Again, you're renaming and modifying files in one commit. Please remember not to do this next time. Did you change anything in the code you moved?" [02:49:01] https://www.mediawiki.org/wiki/Special:Code/MediaWiki/69468 [02:49:01] :P [02:49:04] Oh yeah [02:49:18] Roan? Angry in CR? NEver! [02:49:24] But yeah, he definitely wrote that code [02:49:36] Who definitely wrote what code? [02:50:01] window.mw.util.client and window.mw.util.clientTest [02:50:14] RoanKattouw: James_F and I were wondering qhere jquery.client originated. [02:50:26] Krinkle: No, RoanKattouw and I were and I asked you. :-D [02:50:26] Oh that, yeah right OK [02:50:31] So the answer to that question is "Trevor" then [02:50:35] Yup [02:51:04] 3VisualEditor, VisualEditor-MediaWiki-Mobile: VisualEditor Mobile: [Regression wmf5] The wikitext editor does not show any wikitext for any page when switching from VE in test2 and Beta Labs - https://phabricator.wikimedia.org/T74571#809069 (10Jdforrester-WMF) [02:51:38] 3VisualEditor, VisualEditor-MediaWiki-Mobile: [Regression wmf5] The wikitext editor does not show any wikitext for any page when switching from mobile VE in test2 and Beta Labs - https://phabricator.wikimedia.org/T74571#809077 (10Jdforrester-WMF) [02:51:56] Krinkle: If you've got a spare 10 minutes, can you do https://phabricator.wikimedia.org/T76069 ? [02:52:10] Krinkle: Adding CI to the services/citoid repo. [02:52:33] Krinkle: Also https://gerrit.wikimedia.org/r/#/c/177468/ [02:53:15] James_F: there's package.json [02:53:24] James_F: no script.test though, or a grunt file. [02:53:27] No grunt needed per se [02:53:33] raw test nodejs scirpt is fine [02:53:53] though "m not sure how test_files is used [02:53:57] James_F: What should it have? [02:54:06] Just jshint for now? [02:56:15] Krinkle: That'd be a start. Maybe also jscs? [02:56:59] James_F: Sure. [02:57:10] James_F: So this is prepping the repo first. I thought you meant just Jenkins pipelines. [02:57:15] LIttle more than 10, but make it 20min [02:57:28] I've just yielded so I can do it now [02:57:39] Krinkle: Awesome, thanks. [02:57:46] Krinkle: They don't technically need to pass yet. :-) [02:58:04] James_F: Well, I'd have to explicitly make it non-voting [02:58:09] new jobs require passing by default. [02:59:40] Krinkle: True, but I'd rather Marielle owned the task of making it pass a bit. [03:00:07] OK. I'm not sure what to do in that case. [03:00:16] (I think that's a good idea) [03:00:48] Add it but make it non-voting, add a commit to make it voting but don't deploy until it'd pass? [03:00:49] Or… [03:00:51] Well. [03:00:56] It's non-train deployed right now [03:01:04] So we could just break the repo and fix it tomorrow. [03:01:11] How badly does it break? [03:01:58] James_F: What I mean is, both jshint and jscs have config files. It's totally fine to have ignore rules for a few jscs rules at first (we do that even in mw, ve and oojs-ui now). But that's based on the current code. [03:02:10] Krinkle: Oh, yeah, that could work. [03:02:22] Something Marielle would need to learn and do manually. No biggie, but then I might as well help her add those two little files in the first place. [03:02:26] Krinkle: So ignore the breaking things for now and let us fix later? [03:02:37] True. [03:02:57] I'll add a non-voting npm pileline in Jenkins that will fail on npm test not existing. [03:03:00] But doing something quick now so she can do rolling improvements later (without needing you around) would help. [03:03:05] Then the rest can be done from citoid side [03:03:15] Or make it voting? [03:03:17] (03PS1) 10Sn1per: Citoid: Implement BibTeX format [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 [03:03:30] James_F: And force everything being fixed in one commit before anything else can be merged? [03:03:52] Krinkle: Force it to be fixed-or-ignored in one commit, yes. [03:03:54] (fixed as in made passing, that includes adding ignore rules instead of changing code) [03:03:57] OK [03:04:10] And I can do the initial fix and leave it to Marielle to +2. [03:04:22] And also, please remove V+2 rights from humans. :-) [03:06:54] (03PS2) 10Sn1per: Citoid: Implement BibTeX format [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 [03:07:06] 3VisualEditor, VisualEditor-MediaWiki-Mobile: [Regression wmf16] Cannot scroll the link suggestions inside link inspector in iOS Safari - https://phabricator.wikimedia.org/T71038#809086 (10Jdforrester-WMF) >>! In T71038#808723, @Krenair wrote: > Cannot scroll as in nothing happens? Or it just disappears? I think... [03:15:55] 3Continuous-Integration, Citoid, VisualEditor: Set up CI in the mediawiki/services/citoid.git repo - https://phabricator.wikimedia.org/T76069#809103 (10Krinkle) a:3Krinkle [03:16:36] James_F: It's hard to believe how hard it is to clone a repo from gerrit without making urls manually. [03:16:43] 3VisualEditor: [regression wmf11] Chrome only: Help - Keyboard shortcuts - 'Other' section is not displayed - https://phabricator.wikimedia.org/T76664#809105 (10Jdforrester-WMF) I get the same behaviour in Firefox and Safari; it's a crappy design, though. [03:17:01] Krinkle: You can clone without writing out the URLs manually?! How? [03:17:17] James_F: Find the summary page on git.wikimedia.org (I use url trickery for that, specifally using the unconventional ?r= syntax to evade url encode of slashes) and then copy the hidden gerrit-url from ssh git clone menu [03:17:26] https://git.wikimedia.org/summary/?r=mediawiki/services/citoid.git [03:17:35] repo url: 1,2,3 (?) ssh://gerrit.wikimedia.org:29418/mediawiki/services/citoid.git [03:17:56] Oh, interesting. [03:18:04] That doesn't work for me because I block Flash. [03:18:05] I've yet to be in a situation where I can have another human being clone a gerrit repo without assistence. [03:18:16] (select text and copy) [03:18:19] works without flash :) [03:18:27] Except it doesn't show the other options. [03:18:44] Because the menu closes when I try to make a selection. [03:18:48] Yeah [03:18:49] Quality JavaScript, that. [03:18:50] it does that :P [03:19:02] * James_F avoids some rude language. [03:19:23] James_F: it works if you contextmenu click and copy [03:19:33] * James_F cries. [03:19:41] (after reopening the meny first, that is) [03:19:43] So, no, not possible without manually creating the URL. [03:19:44] selection is still there [03:19:56] I was hoping it exists on gerrit somewhere [03:20:00] It does. [03:20:14] https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/services/citoid [03:20:18] Click "ssh". [03:20:21] Which is a button. [03:20:28] Even though it clearly isn't a button. [03:20:36] It's a weird fake table heading . [03:20:38] Mmm. k, slightly less bad [03:20:38] But it's a button. [03:20:40] Don't ask. [03:20:48] Welcome to Gerrit. :-) [03:20:50] yeah, those tabs are also on commit pages. seen those before [03:21:00] But no-one knows about admin/projects/… [03:21:05] So it might as well not exist. [03:21:49] * Krinkle tries; You get there from https://gerrit.wikimedia.org/ via.. [03:22:00] -> Projects -> List -> type part (citoid is enough, it does substring search) [03:22:04] (03PS1) 10Divec: Add default nodeName to annotations [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177482 [03:22:08] not bad [03:22:13] Cool. that's faster than what I've been doing [03:22:26] Tim once said something like "it's funny how the crawlers can find out gitblit URLs even though our devs can't" [03:22:31] *our [03:22:42] but not nearly as fast as github.com/wikimedia, autocomplete repo, and click ssh on the right hand side of the repo page [03:22:58] Sure. [03:23:00] RoanKattouw: So true, so true. [03:23:07] RoanKattouw: Yeah. [03:23:09] but that clones the wrong repo [03:23:14] hm.. or not [03:23:18] I have to do git review -s anyway [03:23:22] which will fix the remote [03:23:25] * James_F grins. [03:23:29] * James_F goes. [03:23:29] yay, I'll use github from now on for the initial clone [03:23:36] thanks James_F :) [03:24:15] (03CR) 10Catrope: [C: 04-1] "You can't modify these objects in this place. Let's talk later" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177482 (owner: 10Divec) [03:24:25] (03CR) 10jenkins-bot: [V: 04-1] Add default nodeName to annotations [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177482 (owner: 10Divec) [03:24:59] RoanKattouw: it may be that action=visualeditoredit is parsing twice [03:26:35] too late I guess [03:26:48] well, by tomorrow I will know one way or the other [03:40:34] James_F|Away: btw, I've also situated https://github.com/wikimedia/citoid (set url to mediawiki.org page, and alias from mediawiki/services/parsoid, copying what Parsoid did on github for themselves) [03:42:45] (03PS1) 10Krinkle: package.json: Point git-url to Git repository instead of Gerrit project page [services/citoid] - 10https://gerrit.wikimedia.org/r/177485 [03:42:47] (03PS1) 10Krinkle: package.json: Pin dependencies at a fixed version [services/citoid] - 10https://gerrit.wikimedia.org/r/177486 [03:43:00] (03CR) 10Krinkle: [C: 032 V: 032] package.json: Point git-url to Git repository instead of Gerrit project page [services/citoid] - 10https://gerrit.wikimedia.org/r/177485 (owner: 10Krinkle) [03:43:13] (03CR) 10Krinkle: [C: 032 V: 032] package.json: Pin dependencies at a fixed version [services/citoid] - 10https://gerrit.wikimedia.org/r/177486 (owner: 10Krinkle) [05:17:50] (03PS3) 10Sn1per: Citoid: Implement BibTeX format [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 [06:39:57] (03PS4) 10Jforrester: Citoid: Implement BibTeX format [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 (owner: 10Sn1per) [07:00:09] 3MediaWiki-extensions-Cite: Render references list in multiple columns based on the number of items - https://phabricator.wikimedia.org/T33597#809225 (10MZMcBride) [08:27:53] 3Parsoid, VisualEditor, VisualEditor-MediaWiki: Editing a specific article at pl.wp truncates category name - https://phabricator.wikimedia.org/T76353#809338 (10ssastry) a:3Arlolra [09:38:48] (03CR) 10Zfilipin: WIP: [BrowserTest] Rework long external link code (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/122400 (owner: 10Jagori79) [09:43:46] (03PS3) 10Bartosz Dziewoński: MediaWiki Theme: Reduce indentation in theme-oo-ui-checkboxInputWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/176396 (owner: 10Prtksxna) [09:44:38] (03CR) 10Bartosz Dziewoński: [C: 032] MediaWiki Theme: Reduce indentation in theme-oo-ui-checkboxInputWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/176396 (owner: 10Prtksxna) [09:50:55] (03Merged) 10jenkins-bot: MediaWiki Theme: Reduce indentation in theme-oo-ui-checkboxInputWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/176396 (owner: 10Prtksxna) [10:54:32] (03CR) 10Mvolz: [C: 04-1] "Great job so far!" [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 (owner: 10Sn1per) [11:06:11] (03PS1) 10Mvolz: Add contributors field to package.json [services/citoid] - 10https://gerrit.wikimedia.org/r/177516 [11:07:38] (03CR) 10Mvolz: "I wasn't sure how you wanted to be credited, so please fix as necessary." [services/citoid] - 10https://gerrit.wikimedia.org/r/177516 (owner: 10Mvolz) [11:51:45] (03PS2) 10Unicodesnowman: Add contributors field to package.json [services/citoid] - 10https://gerrit.wikimedia.org/r/177516 (owner: 10Mvolz) [11:54:16] (03CR) 10Unicodesnowman: "Thanks :) Learned a new use for Gerrit too." [services/citoid] - 10https://gerrit.wikimedia.org/r/177516 (owner: 10Mvolz) [12:47:12] 3VisualEditor-MediaWiki, VisualEditor: [Regression wmf11]Cannot set “Sort this page by default as” for the individual Categories. [12:49:21] 3VisualEditor-MediaWiki, VisualEditor: [Regression wmf11]Cannot set “Sort this page by default as” for the individual Categories. - https://phabricator.wikimedia.org/T76707#818671 (10Swainr) {F18448} [13:16:48] 3VisualEditor-MediaWiki, VisualEditor: [Regression wmf11]Cannot set “Sort this page by default as” for the individual Categories. - https://phabricator.wikimedia.org/T76707#818720 (10Swainr) Not always reproducible. [13:48:24] (03PS1) 10Esanders: Trim whitespace when annotating [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177533 [13:57:17] (03PS1) 10Esanders: Update to Unicode 7.0.0 [unicodejs] - 10https://gerrit.wikimedia.org/r/177539 [14:03:43] 3VisualEditor-MediaWiki, VisualEditor: Same Category get added multiple times ,for every change in “Sort this page by default as” value. - https://phabricator.wikimedia.org/T76715 (10Swainr) 3NEW p:3Normal [14:04:16] James_F|Away, https://doc.wikimedia.org/UnicodeJS/master/ :( [14:08:43] 3VisualEditor-MediaWiki, VisualEditor: Same Category get added multiple times ,for every change in “Sort this page by default as” value. - https://phabricator.wikimedia.org/T76715#818800 (10Swainr) Screencast at http://youtu.be/EbQoK4-3saI Screenshot {F18460} [14:33:06] 3VisualEditor-MediaWiki, VisualEditor: [Regression pre-wmf12]- Uncaught Error: “Offset could not be translated to a DOM element and offset: 2” when selecting a heading format with language inspector open. - https://phabricator.wikimedia.org/T76717#818828 (10Swainr) [14:34:00] 3VisualEditor-MediaWiki, VisualEditor: [Regression pre-wmf12]- Uncaught Error: “Offset could not be translated to a DOM element and offset: 2” when selecting a heading format with language inspector open. - https://phabricator.wikimedia.org/T76717#818828 (10Swainr) {F18467} [14:35:13] 3VisualEditor-MediaWiki, VisualEditor: [Regression pre-wmf12]- Uncaught Error: “Offset could not be translated to a DOM element and offset: 2” when selecting a heading format with language inspector open. - https://phabricator.wikimedia.org/T76717#818838 (10Swainr) Browsers-Chrome 39,FF 33.1.1 [15:37:27] 3VisualEditor-MediaWiki, VisualEditor: VisualEditor: VE does not check for global blocks or title blacklist on page edit/creation - https://phabricator.wikimedia.org/T40268#818940 (10MarcoAurelio) [15:50:04] 3VisualEditor: Newly made links are red until page is saved - https://phabricator.wikimedia.org/T75741#818981 (10Elitre) Updating [[ https://en.wikipedia.org/wiki/Wikipedia:VisualEditor/Feedback/Archive_2014_4#Newly_made_links_are_red_until_page_is_saved | feedback page ]] link. This bug looks like a regression... [16:00:15] (03PS2) 10Esanders: Update to Unicode 7.0.0 [unicodejs] - 10https://gerrit.wikimedia.org/r/177539 [16:04:14] (03PS3) 10Esanders: Remove window even if closing promise rejects [oojs/ui] - 10https://gerrit.wikimedia.org/r/177406 [17:21:12] (03PS1) 10Paladox: Remove $ from wgExtensionCredits [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/177566 [17:21:26] (03PS2) 10Paladox: Remove $ from wgExtensionCredits [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/177566 [17:22:20] edsanders: Will get that published. [17:22:22] edsanders: Oh, hmm, there's already a task for it? [17:22:26] edsanders: Aha. It's case sensitive. https://doc.wikimedia.org/unicodejs/master/ [17:24:29] (03PS2) 10Paladox: Convert entry point to use explicit globals [extensions/WikiEditor] (REL1_24) - 10https://gerrit.wikimedia.org/r/175296 [17:25:39] (03PS3) 10Paladox: Convert entry point to use explicit globals [extensions/WikiEditor] (REL1_24) - 10https://gerrit.wikimedia.org/r/175296 [17:29:41] edsanders: https://gerrit.wikimedia.org/r/#/c/177570/ [17:30:36] Corruption alert: visualeditor-needcheck on svwiki: https://sv.wikipedia.org/?diff=29043494 [17:30:36] Corruption alert: visualeditor-needcheck on svwiki: https://sv.wikipedia.org/?diff=29043519 [17:30:36] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=109653524 [17:30:36] Corruption alert: visualeditor-needcheck on ukwiki: https://uk.wikipedia.org/?diff=15231197 [17:30:36] Corruption alert: visualeditor-needcheck on ruwiki: https://ru.wikipedia.org/?diff=67149778 [17:30:36] Corruption alert: visualeditor-needcheck on ruwiki: https://ru.wikipedia.org/?diff=67149883 [17:30:36] Corruption alert: visualeditor-needcheck on ruwiki: https://ru.wikipedia.org/?diff=67154716 [17:30:37] Corruption alert: visualeditor-needcheck on ptwiki: https://pt.wikipedia.org/?diff=40769130 [17:30:37] Corruption alert: visualeditor-needcheck on enwiki: https://en.wikipedia.org/?diff=636592704 [17:30:43] James_F, link was wrong mw.org will fix - also not listed on https://doc.wikimedia.org/ [17:31:11] edsanders: My gerrit patch adds it to the docroot repo. [17:31:17] edsanders: Thanks for fixing on MW.org. [17:32:47] (03CR) 10Jforrester: [C: 032] Update to Unicode 7.0.0 [unicodejs] - 10https://gerrit.wikimedia.org/r/177539 (owner: 10Esanders) [17:32:58] I don't have +2 on docroot [17:33:07] edsanders: Of course not. It's a production repo. [17:33:15] edsanders: Timo can push it when he's back. [17:33:27] (03Merged) 10jenkins-bot: Update to Unicode 7.0.0 [unicodejs] - 10https://gerrit.wikimedia.org/r/177539 (owner: 10Esanders) [17:33:35] Yes - how could I possibly not know that [17:33:56] All of integration/* is off-limits to anyone except hashar, Krinkle and cscott. [17:38:53] hi. [17:42:13] (03CR) 10Bartosz Dziewoński: "It's some Composery junk." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/175214 (owner: 10Paladox) [17:42:49] (03CR) 10Bartosz Dziewoński: Convert entry point to use explicit globals (031 comment) [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/175214 (owner: 10Paladox) [17:42:55] (03CR) 10Bartosz Dziewoński: [C: 032] "Whoops." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/177566 (owner: 10Paladox) [17:43:00] (03Merged) 10jenkins-bot: Remove $ from wgExtensionCredits [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/177566 (owner: 10Paladox) [17:43:48] James_F: might want to SWAT that ^, not sure if anyone cares. WikiEditor is currently not visible on Special:Version on the cluster. [17:44:06] MatmaRex: Ouch. When did it break? [17:44:11] two versions ago [17:44:15] wmf versions* [17:44:24] Did anyone other than developers notice? [17:44:40] paladox broke it and it seems he noticed himself, and just submitted that patch [17:44:44] i don't think anyone else did [17:44:49] so, probably no one cares [17:48:38] (03CR) 10Paladox: "Ok thanks." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/175214 (owner: 10Paladox) [17:48:52] (03CR) 10Paladox: "Thanks." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/177566 (owner: 10Paladox) [17:53:18] (03PS1) 10Paladox: Remove $ from wgExtensionCredits [extensions/WikiEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177577 [17:53:57] (03CR) 10Paladox: "When I try to look through https://www.mediawiki.org/w/api.php?action=query&meta=siteinfo&siprop=extensions for wikieditor I carnt find it" [extensions/WikiEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177577 (owner: 10Paladox) [17:54:49] * cscott is an integration/* stud [17:55:00] (03CR) 10Reedy: [C: 031] Remove $ from wgExtensionCredits [extensions/WikiEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177577 (owner: 10Paladox) [17:55:03] (03PS1) 10Jforrester: Tag v0.1.2 [unicodejs] - 10https://gerrit.wikimedia.org/r/177578 [17:55:42] (03PS1) 10Paladox: Remove $ from wgExtensionCredits [extensions/WikiEditor] (wmf/1.25wmf10) - 10https://gerrit.wikimedia.org/r/177579 [17:56:01] (03CR) 10Paladox: "When I try to look through https://www.mediawiki.org/w/api.php?action=query&meta=siteinfo&siprop=extensions for wikieditor I carnt find it" [extensions/WikiEditor] (wmf/1.25wmf10) - 10https://gerrit.wikimedia.org/r/177579 (owner: 10Paladox) [18:00:48] edsanders: https://gerrit.wikimedia.org/r/177578 +2 please. [18:14:34] (03CR) 10Esanders: [C: 032] Tag v0.1.2 [unicodejs] - 10https://gerrit.wikimedia.org/r/177578 (owner: 10Jforrester) [18:15:16] (03Merged) 10jenkins-bot: Tag v0.1.2 [unicodejs] - 10https://gerrit.wikimedia.org/r/177578 (owner: 10Jforrester) [18:23:50] edsanders: Thanks! [18:33:08] (03PS1) 10Jforrester: Update UnicodeJS to v0.1.2 [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177582 [18:36:53] (03PS1) 10Jforrester: bin: Update update-unicodejs.sh to parity with update-oojs.sh [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177583 [18:39:09] edsanders: BTW https://doc.wikimedia.org/ now has UnicodeJS listed, thanks to Timo. [18:40:12] (03CR) 10Jforrester: "Worth it? The train will fix this in a couple of weeks' time…" [extensions/WikiEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177577 (owner: 10Paladox) [18:47:08] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819244 (10Catrope) [18:51:22] RoanKattouw: review of https://gerrit.wikimedia.org/r/#/c/177501/ appreciated! [18:51:43] ori: Ooooooh [18:51:58] Krinkle|detached: I'll look at ---^^ but you should too [18:55:49] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819269 (10Milimetric) a:3Milimetric [18:57:32] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819244 (10Milimetric) I'm making this now. FYI - this will be for all edit-team related dashboarding, it's not restricted to the edit schema. [19:26:25] 3MediaWiki-extensions-WikiEditor: Search&replace disabled in IE8, Opera - https://phabricator.wikimedia.org/T25992#819305 (10Kipod) >>! In T25992#808943, @GOIII wrote: > Well that was 4 years and at least 2 versions of IE ago. > Do you think something might have changed in the interim? > We will never know unles... [19:27:03] RoanKattouw: Want me to fix 177468 ? [19:29:07] James_F: I'm on it [19:30:10] RoanKattouw: Kk. [19:36:12] 3VisualEditor: VisualEditor inserts inexpected tags in the wikicode - https://phabricator.wikimedia.org/T76749#819322 (10Automatik) [19:37:51] 3VisualEditor-ContentEditable, VisualEditor: data-ve-clipboard-key span inserted in page - https://phabricator.wikimedia.org/T76749#819333 (10Catrope) a:3Esanders [19:44:54] (03PS18) 10Mooeypoo: Adding DraggableGroupElement and DraggableElement mixins [oojs/ui] - 10https://gerrit.wikimedia.org/r/172325 [19:44:54] RoanKattouw: should ( 'textStyle/bold', { nodeName: 'b' } ) be distinct to ( 'textStyle/bold', undefined ) in an AnnotationSet? [19:46:08] (03CR) 10Paladox: "Oh ok." [extensions/WikiEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177577 (owner: 10Paladox) [19:46:08] divec: Maaaybe [19:46:16] I haven't thought through how I want to handle that yet [19:47:34] If so, we need something like a mergeAnnotations( targetAnnotations, sourceAnnotations ) function. [19:47:46] RoanKattouw: But we probably need that for other purposes in any case [19:48:51] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819344 (10Milimetric) [19:52:49] (03CR) 10Trevor Parscal: [C: 032] Adding DraggableGroupElement and DraggableElement mixins [oojs/ui] - 10https://gerrit.wikimedia.org/r/172325 (owner: 10Mooeypoo) [19:52:59] Whee. [19:53:42] RoanKattouw: all the dashboarding stuff is set up - just waiting for you to basically say "ok". The way it's set up right now, you'd see a graph of this query: https://gerrit.wikimedia.org/r/#/c/177601/2/edit/action.sql,cm [19:53:54] (once we merge everything) [19:54:11] wheeee [19:55:07] milimetric: Some lines in that query have "event action" and some have "event_action" [19:55:13] (03Merged) 10jenkins-bot: Adding DraggableGroupElement and DraggableElement mixins [oojs/ui] - 10https://gerrit.wikimedia.org/r/172325 (owner: 10Mooeypoo) [19:55:20] \o/ [19:55:33] milimetric: Never mind [19:55:38] Stupid Gerrit display thing lying to me [19:56:21] milimetric: Wow, awesome. Thank you! [19:56:22] :) i was gonna say - i screw up a lot of things ... but i actually ran the query :) [19:56:59] lol [19:57:18] milimetric: OK so I -1ed your change because unless I'm also blind in a different way, you have "Save" throughout and "save" in one place [19:57:36] milimetric: And yeah thanks for setting this all up man! [19:57:55] veaction=edit&debug=true is fun [19:58:08] Nuria linked me to the doc page so I did step 1 (request repo) and worked on other things for a while, and now you're running through all the steps for me :) [19:58:15] hexmode: Yeah, it's what, 350 requests for JS files or something? [19:58:20] It's slow even on localhost [19:58:29] RoanKattouw: was there anything else wrong though? I'll fix the spelling and merge everything else? [19:58:44] RoanKattouw: exactly [20:00:02] milimetric: Looks fine to me [20:00:30] k - also, np, easy to do while listening to metrics [20:01:08] milimetric: We will probably want dashboards for the duration components of things , and all sorts of other stuff I haven't thought of yet, but I'll let James_F and ori handle that [20:02:59] RoanKattouw: so the process, once these pipes are laid and proven non-leaky, is for you to come up with queries that you want to run, add them much like the one I added here, and submit for review. [20:03:03] so have James_F|Away and others think about what to measure, write the query, ping me :) [20:03:12] (also i can help with SQL if you need that) [20:04:55] Yeah that's what I figured [20:04:58] Thanks so much man! [20:06:14] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819375 (10Jdforrester-WMF) [20:07:22] milimetric: Brilliant. [20:20:22] 3VisualEditor: VisualEditor: Text in Japanese disappears - https://phabricator.wikimedia.org/T74619#819396 (10Yukichi99) [20:30:40] 3VisualEditor-MediaWiki, VisualEditor: [Regression wmf11]Cannot set “Sort this page by default as” for the individual Categories. - https://phabricator.wikimedia.org/T76707#819417 (10Etonkovidova) To reproduce - "hit enter to set the value" is important. 1. Categories - type/select a category - hit Enter. 2.... [20:37:45] divec: Have you seen https://phabricator.wikimedia.org/T74619 ? [20:39:34] (03CR) 10Catrope: [C: 031] "Looks good to me, but Trevor said something about how this function was intended to do something different yesterday. I think he's wrong, " [oojs/ui] - 10https://gerrit.wikimedia.org/r/177406 (owner: 10Esanders) [20:40:48] (03PS4) 10Jforrester: Remove window even if closing promise rejects [oojs/ui] - 10https://gerrit.wikimedia.org/r/177406 (owner: 10Esanders) [20:42:35] 3OOjs-UI: clearWindows() is completely broken, only removes open window(s) - https://phabricator.wikimedia.org/T76659#819436 (10Jdforrester-WMF) [20:42:44] 3OOjs-UI: clearWindows() is completely broken, only removes open window(s) - https://phabricator.wikimedia.org/T76659#819437 (10Jdforrester-WMF) a:3Esanders [20:48:47] RoanKattouw, did someone mention VE mobile being broken on test wikis in the hangout? [20:48:56] because I ran into that yesterday [20:50:06] Krenair: Yes, Elena did. [20:50:22] Krenair: It's fixed in master but Roan needs to merge the backport. [20:50:29] ah [20:50:43] and then presumably deploy it? :p [20:51:08] Krenair: I can get other people to deploy it. :-) [20:51:33] I need to line up the backport for that, I'll do that now [20:52:11] (03CR) 10Catrope: [C: 032] Fix path to MobileWindowManager.js [extensions/VisualEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177453 (owner: 10Catrope) [20:52:14] (03PS2) 10Divec: POC: normalize annotations on insertion [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177482 [20:53:30] (03CR) 10jenkins-bot: [V: 04-1] POC: normalize annotations on insertion [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177482 (owner: 10Divec) [20:54:15] (03Merged) 10jenkins-bot: Fix path to MobileWindowManager.js [extensions/VisualEditor] (wmf/1.25wmf11) - 10https://gerrit.wikimedia.org/r/177453 (owner: 10Catrope) [20:58:09] 3VisualEditor: [Regression pre-wmf12] VE is not opening in mobile - https://phabricator.wikimedia.org/T76660#819559 (10Ryasmeen) 5Resolved>3Open [20:59:15] 3VisualEditor: [Regression pre-wmf12] VE is not opening in mobile - https://phabricator.wikimedia.org/T76660#808432 (10Ryasmeen) [21:01:20] James_F, I saw your comment on https://phabricator.wikimedia.org/T71038 - I followed the instructions (though on my laptop rather than iOS Safari) and got what appears to be an OOUI dropdown box [21:02:12] Krenair: Yes, but over a fixed full-size dialog that means you can't scroll in iOS, right? [21:02:35] Krenair: The problem AIUI is the magical trickery in MobileFrontend to put a toolbar at the top makes things not scroll when they should. [21:02:47] Krenair: It's really horrible. :-( [21:03:11] 3VisualEditor, OOjs-UI: Tools in the toolbar do not have unique attributes - https://phabricator.wikimedia.org/T76143#819578 (10TrevorParscal) [21:06:18] I still don't understand your comment. The link suggestions show in an OOUI drop down menu, even if that is inside a modal [21:07:15] 3VisualEditor: [Regression wmf11] VE is not opening in mobile - https://phabricator.wikimedia.org/T76660#819587 (10Ryasmeen) [21:08:54] I can't check iOS myself, but thought I'd point out an issue where ooui menuwidget scrollbars were not working was fixed since this got created [21:09:13] Yes, and the scroll issue is with the modal moving the OOUI widget wrongly. [21:09:18] It's not about scrollbars. [21:09:34] It makes the 5th+ item not scrollable. :-( [21:13:45] (Or whatever, depending on screen size.) [21:19:43] RoanKattouw_away / James_F: http://edit-reportcard.wmflabs.org/ [21:20:36] milimetric|post: Nice. Thank you! [21:21:25] milimetric|post: Interesting how those numbers don't quite add up. :-) [21:22:11] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819633 (10Milimetric) [21:22:13] milimetric|post: 60k inits, 52k ready => 8k clicked edit but never got the editor loaded. Definitely a good pointer for us to investigate. [21:22:21] milimetric|post: Thanks. [21:23:00] James_F: yeah, a number of things could be wrong. My query looks ok to me, but you can double check that too [21:23:31] btw, I grabbed data from two revisions of the schema [21:23:32] milimetric|post: I'm more suspicious of our instrumentation or our code just being wonky, but yes. :-) [21:24:03] With and without the new saveFailure case? Cool. [21:24:05] the 8k drop is interesting for sure [21:24:42] It's totally possible that ~12% of users abort editing before they get to the editor loaded state. [21:24:48] On long pages it can take a serious while. [21:25:03] (Timing and things like that are the first things I'm going to start investigating.) [21:25:15] well - if "data leads to questions" was our goal - success! [21:25:34] Indeed. :-) [21:25:56] cool - so this will be running hourly from now on, and it re-generates all the data each time [21:26:09] Right. [21:26:18] feel free to ping me / us if anything goes wrong or if you have trouble adding graphs, etc. [21:26:19] Is that potentially a load issue? [21:26:25] not from what i could tell [21:26:27] Or should we not worry? [21:26:28] OK. [21:26:28] 3VisualEditor, OOjs-UI: [Regression 0.2.3] While switching dialog sizes, the window opens with an animation from bottom - https://phabricator.wikimedia.org/T76371#819645 (10Ryasmeen) [21:26:57] if you notice data lagging or not being refreshed, we can set "timeboxed" mode which makes it only update the data it hasn't crunched yet [21:27:07] 3VisualEditor, Analytics-Engineering: Dashboard repository for Edit schema - https://phabricator.wikimedia.org/T76744#819648 (10Jdforrester-WMF) [21:27:09] * James_F nods. [21:45:50] 3VisualEditor, VisualEditor-ContentLanguage: Cursor error typing into end of paragraph with Ibus Malayalam Swanalekha - https://phabricator.wikimedia.org/T56057#819700 (10Jdforrester-WMF) [21:45:51] 3VisualEditor, VisualEditor-ContentLanguage: VisualEditor multilingual input / i18n issues (tracking) - https://phabricator.wikimedia.org/T35077#819702 (10Jdforrester-WMF) [21:46:39] 3VisualEditor, VisualEditor-EditingTools: Implement a find/search and replace tool - https://phabricator.wikimedia.org/T50368#819713 (10Jdforrester-WMF) [21:46:48] 3VisualEditor, VisualEditor-EditingTools: Implement a find/search and replace tool - https://phabricator.wikimedia.org/T50368#819714 (10Jdforrester-WMF) [21:48:50] 3VisualEditor, OOjs-UI: [Regression 0.2.3] While switching dialog sizes, the window opens with an animation from bottom - https://phabricator.wikimedia.org/T76371#819722 (10Jdforrester-WMF) [21:48:53] 3VisualEditor, VisualEditor-MediaWiki: [Regression wmf10] Switching to Edit source mode is not keeping changes done in VE - https://phabricator.wikimedia.org/T76114#819724 (10Jdforrester-WMF) [21:48:55] 3VisualEditor-EditingTools, VisualEditor, VisualEditor-MediaWiki: [Regression wmf10] Cannot open any inspectors, they just flicker and get closed - https://phabricator.wikimedia.org/T76076#819731 (10Jdforrester-WMF) [21:48:57] 3OOjs-UI: OOjs UI: OO.ui.ButtonWidget default target should not be "_blank" - https://phabricator.wikimedia.org/T75452#819733 (10Jdforrester-WMF) [21:48:58] 3OOjs-UI: MessageDialog does not re-fit actions on dialog resize - https://phabricator.wikimedia.org/T75783#819734 (10Jdforrester-WMF) [21:49:00] 3VisualEditor-ContentEditable, VisualEditor, VisualEditor-MediaWiki-Media: [Regression wmf10] Changes in the caption surface in MediaDialog does not enable the "Apply" button - https://phabricator.wikimedia.org/T76074#819735 (10Jdforrester-WMF) [21:49:01] 3Wikimedia-Extension-setup, VisualEditor: VisualEditor has been disabled unexpectedly on fr.wiktionary - https://phabricator.wikimedia.org/T76099#819736 (10Jdforrester-WMF) [21:49:04] 3VisualEditor, VisualEditor-MediaWiki: The message visualeditor-recreate says "Press continue", but it's unclear what does "continue" refer to - https://phabricator.wikimedia.org/T75971#819737 (10Jdforrester-WMF) [21:49:05] 3OOjs-UI: Issues with dialog sizing on first open - https://phabricator.wikimedia.org/T72061#819738 (10Jdforrester-WMF) [21:54:57] 3VisualEditor-CopyPaste, VisualEditor: Clear the paste target once it's been used, lest the user be confused by too many copies in browser search - https://phabricator.wikimedia.org/T76787 (10Jdforrester-WMF) 3NEW p:3Normal [22:17:17] when i paste a three-row table in VE, I'm getting only two rows in the output -- the first and second row are being run together. known bug? [22:18:03] also, there seems to be no way to apply a class to the table as a whole, so things look very different after I hit save. [22:19:30] edsanders|away, ^ [22:20:06] RoanKattouw, how am I supposed to access parsoid's document base href from MWInternalLinkAnnotation? [22:20:37] There is a thing, let me find it [22:21:04] hello VE folks [22:21:04] cscott: Did you paste a table that has a ? When you say "output", do you mean wikitetxt? [22:21:14] Morning TimStarling [22:21:30] sorry I didn't get a patch out yesterday, but hopefully today I can make VE page saves 4x faster [22:23:10] ok maybe that is a slight exaggeration [22:23:10] TimStarling: Sounds great. :-) [22:23:14] let's say 3x [22:23:33] TimStarling: I'll settle for 2x. [22:23:51] I think 2x would be a disappointment [22:23:53] TimStarling: Thanks for finding that, I read your email last night and my mind was blown [22:23:57] TimStarling: It will also make non-autoconfirmed wikitext API saves faster, right? [22:24:05] yes [22:24:10] Good. [22:24:13] What was found? [22:24:15] We care about speed for all users, not just VE. [22:24:18] RoanKattouw: yes, i mean wikitext. i don't know exactly what i pasted, i copied it from an html page. let me check. [22:24:35] That displaytitle thing was something where I could swear the code tried to reuse the ParserOutput from the parser cache, but maybe we didn't do that at all or else I wasn't diligent enough reviewing it [22:24:41] Krenair: TimStarling found that ConfirmEdit isn't very efficient. [22:24:41] I don't think I saw anything on editing-team@ [22:24:56] Krenair: In extreme cases, saving a page via VE parses it four times [22:25:00] ouch [22:25:10] Once for good reasons, once for the displaytitle alone, and twice for CAPTCHA stuff [22:25:40] The displaytitle thing is VE-specific code that IIRC you wrote, and the CAPTCHA thing is in MW core's edit API module [22:25:51] I thought about the displaytitle thing but assumed it would come from the cache [22:26:00] Yeah I remember thinking the same thing [22:26:04] I probably should have checked that. [22:26:06] assumed [22:26:15] I don't remember it very well but I do remember spending some minimal amount of time convincing myself that this would hit pcache [22:26:19] I know, I assumed something about the parser.. [22:26:29] So I feel like most likely the code at least looked to me like it was trying to use the cache [22:26:49] Good intentions, etc [22:28:23] Does legal have its own IRC channel? [22:28:35] T13|mobile: Doubt it. Why? [22:29:20] Because I was told I needed to get in touch with legal to get a phab request filled. [22:29:36] TimStarling: In other news, my new shiny instrumentation code went live to Wikipedias with wmf10 yesterday, and Dan just built a simple strawman dashboard for it http://edit-reportcard.wmflabs.org/# [22:30:23] good stuff [22:30:45] T76736 [22:31:32] RoanKattouw, I don't feel like taking doc.baseURI and removing the query string and encodeURIComponent(wgPageName) from it [22:31:54] Basically I'm asking to get information I could find on SUL or central auth via the api about another user's lang code or tz. [22:31:57] TimStarling: There's timing stuff in that data set as well, we just have to write some more SQL to surface it [22:31:59] :) [22:32:31] Krenair: So nowadays doc.baseURI should no longer contain wgPageName [22:33:07] Instead it should be http://en.wikipedia.org/wiki/ or something [22:33:08] the HTMLDocument passed to getTargetDataFromHref? [22:33:27] Yes that exactly [22:33:32] Checking [22:33:48] you sure? [22:34:37] yup, we deployed our base href changes [22:34:38] Krenair: http://parsoid-lb.eqiad.wikimedia.org/enwiki/VisualEditor?oldid=632727197 , inspect and look at the tag [22:34:51] i can give you the gerrit patchset for the change if you are interested [22:34:54] [22:35:09] RoanKattouw: did you get around to trying out https://gerrit.wikimedia.org/r/#/c/177501/ btw? [22:35:19] Not yet [22:35:24] I have to prepare for a meeting I have soon [22:35:26] ok no worries [22:36:01] I logged doc.baseURI and got path/to/mw/wiki/Main_Page?veaction=edit [22:36:13] Ugh [22:36:31] That sounds like getTargetDataFromHref is being passed the wrong document? [22:36:36] oh [22:36:39] one moment [22:37:22] I had an old copy of parsoid checked out [22:37:44] YOU PEOPLE ROCK thank you for search and replace [22:37:47] Yeah it looks like ve.dm.MWInternalLinkAnnotation.static.toDataElement invokes it correctly but ve.ce.MWTransclusionNode.prototype.getRenderedDomElements may not [22:38:01] aharoni: Thank you :) it was all edsanders|away 's work [22:38:14] aharoni: :-) [22:38:22] aharoni: edsanders|away is amazing, yes. [22:38:32] Krenair: The code around transclusion nodes and GeneratedContentNode is a complete mess with respect to documents anywy [22:38:33] My first commit as a WMF contractor was a little fix to WikiEditor's search and replace ;) [22:38:39] Reviewed by RoanKattouw [22:38:42] Almost nothing in that code uses the correct document [22:39:06] This includes passing things that have been created in the wrong document, converting things to the wrong document at the wrong time, etc etc [22:39:32] It's super confusing so I haven't yet taken the time to sit down and chart the situation exactly, but I do know it's insane and needs to be rethought from scratch [22:40:03] aharoni, My first contribution to MediaWiki was... never mind :) [22:51:41] (03PS1) 10Trevor Parscal: Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 [22:53:31] nick mooeypoo [22:53:35] fail. [22:57:03] (03PS1) 10Trevor Parscal: Remove more lies from documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177682 [23:00:42] mooeypoo: food is good. [23:00:50] it's essential to your survival. [23:01:00] (03PS2) 10Bartosz Dziewoński: Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:01:07] (03CR) 10Bartosz Dziewoński: [C: 032] Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:01:12] aharoni, indeed [23:01:13] (03CR) 10Bartosz Dziewoński: [C: 032] Remove more lies from documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177682 (owner: 10Trevor Parscal) [23:03:10] (03PS5) 10Sn1per: Citoid: Implement BibTeX format [services/citoid] - 10https://gerrit.wikimedia.org/r/177478 [23:03:55] TrevorParscal: BTW there is a pile of code awaiting review in oojs-ui, would be nice if some of that could get reviewed [23:04:09] I have a meeting now but after that I will review the image module thing [23:04:53] 3VisualEditor-ContentEditable: VisualEditor Search and replace has weird display in RTL wikis - https://phabricator.wikimedia.org/T76807 (10Amire80) 3NEW p:3Triage [23:04:57] my image module? what image module? [23:05:18] mooeypoo: Bartosz's [23:05:21] 3VisualEditor-ContentEditable: VisualEditor Search and replace has weird display in RTL wikis - https://phabricator.wikimedia.org/T76807#820016 (10Amire80) [23:05:30] To do with icons and ResourceLoader [23:05:32] 3MediaWiki-extensions-WikiEditor: wikiEditor "replace" borked - https://phabricator.wikimedia.org/T76750#820031 (10Aklapper) [23:05:34] to be fair it also involves RTL ;) [23:06:24] Everything involves RTL [23:06:31] * mooeypoo muahahas [23:06:34] cscott, I updated parsoid to the latest master but requests to it never complete [23:06:49] (03CR) 10jenkins-bot: [V: 04-1] Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:08:14] * aharoni can't even figure out where is this highlighting in the DOM [23:08:32] (the highlighting that is broken in RTL, I mean) [23:10:55] (03Merged) 10jenkins-bot: Remove more lies from documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177682 (owner: 10Trevor Parscal) [23:32:03] 3OOjs-UI: Issues with dialog sizing on first open - https://phabricator.wikimedia.org/T72061#820060 (10Etonkovidova) [23:33:18] 3OOjs-UI: Issues with dialog sizing on first open - https://phabricator.wikimedia.org/T72061#820061 (10Etonkovidova) [23:55:47] RoanKattouw: so... you probably know, but just naively, what do you expect the return type to be for a method called isValid() ? [23:55:58] TrevorParscal: bool? [23:55:58] (03PS3) 10Bartosz Dziewoński: Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:56:05] (03CR) 10Bartosz Dziewoński: [C: 032] Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:56:11] TrevorParscal: Please tell me it's bool? [23:56:17] James_F: indeed, however, of course, our TextInputWidget returns a promise [23:56:21] Oh. [23:56:26] Yes [23:56:31] I dislike that interface [23:56:33] and it's nice, because then we can have server-side validation [23:56:42] isValidP [23:56:59] cscott: None of your Parsoid coding confusion here, please. ;-) [23:57:07] so, how could we rename the method to better suggest the return type? (without resorting to hungarian notation, as cscott suggests) [23:57:15] This was especially painful when I discovered that there were other classes that implemented an isValid() method that returned a bool, and one of them was in fact a subclass of TextInputWidget and that caused major problems [23:57:18] Krenair: parsoid update qs in #-parsoid ? [23:57:25] TrevorParscal: getValidityPromise() [23:57:26] ok [23:58:18] I think that methods that return promises should be named in a way that suggests they are doing some work, not just returning a cached value [23:58:40] is and get, for instance, lead you to believe that you are obtaining the value directly [23:58:55] suffixing the method with Promise or P seems quite messy [23:59:07] if the method were validate() [23:59:35] TrevorParscal: I'm sure there are other methods that are named getSomething() that return a promise; we may want to look into fixing those too [23:59:39] it seems like it would make more sense [23:59:49] (03Merged) 10jenkins-bot: Fix lies in documentation [oojs/ui] - 10https://gerrit.wikimedia.org/r/177681 (owner: 10Trevor Parscal) [23:59:52] RoanKattouw: agreed