[00:00:00] I don't have it installed on my local wiki [00:00:07] Krenair: i know. i'm getting that error too, and i've been ignoring it :P [00:00:08] MatmaRex: Its not running at all and those were the only errors I was getting [00:01:30] (03CR) 10GWicke: Fix Parsoid transformation calls (032 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183375 (https://phabricator.wikimedia.org/T86046) (owner: 10Alex Monk) [00:01:54] MatmaRex: I do have ErrorLogging installed. Perhaps a problem with Parsoid then? [00:02:27] Negative24: no idea. want to file a bug? (see channel topic for info) [00:02:30] EventLogging, not ErrorLogging [00:02:32] i don't have time right now, sorry [00:03:54] MatmaRex: Sure. I just want to make sure its not a problem on my end. Parsoid is going through localhost, right? Port 8000 is blocked outbound and inbound. [00:04:27] Krenair: I left two small requests for a tweak inline, up to you whether you want to address them in this patch as they are just something I noticed along the way [00:04:43] not addressing in this patch [00:07:55] 3VisualEditor, VisualEditor-MediaWiki-Media: The "Apply Changes" button in Media Settings dialog for newly added images does not get enabled after applying/changing the link of the caption, when the cursor is at the beginning index - https://phabricator.wikimedia.org/T85845#961262 (10Mooeypoo) This is an issue w... [00:15:03] (03CR) 10GWicke: [C: 032] Fix Parsoid transformation calls [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183375 (https://phabricator.wikimedia.org/T86046) (owner: 10Alex Monk) [00:15:34] btw, is Roan down under? [00:16:58] according to his calendar he leaves in two days [00:17:11] it says he's in the netherlands at the moment [00:17:14] (03Merged) 10jenkins-bot: Fix Parsoid transformation calls [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183375 (https://phabricator.wikimedia.org/T86046) (owner: 10Alex Monk) [00:18:45] TimStarling: ah, thx! [00:21:51] 3VisualEditor, VisualEditor-MediaWiki: Edit source link broken after editing the page with VisualEditor - https://phabricator.wikimedia.org/T85552#961319 (10Etonkovidova) [00:31:13] (03PS33) 10Mooeypoo: [WIP] Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [00:32:25] Krenair: It seems as though Parsoid is working because Flow works fine (with store as html on). Even weirder, if I comment out $wgVisualEditorParsoidURL then it works. Is there a default set anywhere? [00:36:10] 3OOjs-UI, VisualEditor, VisualEditor-MediaWiki: Dialog section labels are way too big compared to the rest of the contents of a dialog - https://phabricator.wikimedia.org/T78037#961448 (10Etonkovidova) [00:39:07] 3VisualEditor, VisualEditor-MediaWiki: Update bits of VisualEditor still referring to Bugzilla - https://phabricator.wikimedia.org/T75733#961469 (10Etonkovidova) [00:42:28] (03CR) 10Alex Monk: "Trevor, Bartosz: Please see Ed's comment" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/182432 (https://phabricator.wikimedia.org/T85619) (owner: 10Alex Monk) [00:42:38] 3VisualEditor, VisualEditor-MediaWiki: [Regression wmf14] Response from parseFragment is blank content, meaning newly-inserted or -modified content is blank until saving - https://phabricator.wikimedia.org/T86046#961488 (10Ryasmeen) Verified the fix in Betalabs.Math function , gallery and template getting added... [00:44:02] 3VisualEditor, VisualEditor-MediaWiki: [Regression wmf14] Response from parseFragment is blank content, meaning newly-inserted or -modified content is blank until saving - https://phabricator.wikimedia.org/T86046#961496 (10Ryasmeen) 5Open>3Resolved [00:44:07] 3VisualEditor-EditingTools, VisualEditor: Changes to extension node attributes are ignored - https://phabricator.wikimedia.org/T85818#961499 (10Etonkovidova) [00:45:57] 3VisualEditor: [Regression pre-wmf14] While trying to open Re-use Reference dialog immediately after inserting a citation (not basic reference), throws the TypeError: Cannot read property 'length' of null and the newly added citation does not appear in the list - https://phabricator.wikimedia.org/T86048#961504 (1... [00:46:10] 3VisualEditor: [Regression pre-wmf14] While trying to open Re-use Reference dialog immediately after inserting a citation (not basic reference), throws the TypeError: Cannot read property 'length' of null and the newly added citation does not appear in the list - https://phabricator.wikimedia.org/T86048#961506 (1... [00:47:12] 3OOjs-UI, VisualEditor, VisualEditor-MediaWiki: Clash between rounded corners and text content is jarring for the category editor in VisualEditor when using the OOjs UI MediaWiki theme - https://phabricator.wikimedia.org/T78039#961515 (10Jay8g) Whatever you call it, (nearly) everything else in VE is using the st... [00:47:14] 3VisualEditor, VisualEditor-MediaWiki: VisualEditor: When I click the Save button, please take my cursor out of the main text and focus it in the Save dialog - https://phabricator.wikimedia.org/T75675#961516 (10Etonkovidova) [00:50:16] 3MediaWiki-Interface, MediaWiki-skins-Vector, VisualEditor, VisualEditor-MediaWiki: All links inside jQuery UI dialogs are black since update - https://phabricator.wikimedia.org/T85857#961526 (10Etonkovidova) [01:03:27] 3OOjs-UI, VisualEditor, VisualEditor-MediaWiki: Clash between rounded corners and text content is jarring for the category editor in VisualEditor when using the OOjs UI MediaWiki theme - https://phabricator.wikimedia.org/T78039#961589 (10matmarex) (The category selector really should be implemented using the kin... [01:08:32] (03CR) 10Bartosz Dziewoński: "Dunno, I fixed that bug a month ago. df4e5b4b" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/182432 (https://phabricator.wikimedia.org/T85619) (owner: 10Alex Monk) [01:13:29] (03CR) 10Divec: "The IME test breakage is a little tricky, but I'm working on it" (035 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/179223 (owner: 10Divec) [01:13:50] (03PS1) 10Jforrester: Fix Parsoid transformation calls [extensions/VisualEditor] (wmf/1.25wmf14) - 10https://gerrit.wikimedia.org/r/183409 (https://phabricator.wikimedia.org/T86046) [01:15:11] 3VisualEditor, VisualEditor-MediaWiki, VisualEditor-MediaWiki-References: [Regression pre-wmf14] While trying to open Re-use Reference dialog immediately after inserting a citation (not basic reference), throws the TypeError: Cannot read property 'length' of null and the newly added citation does not appear in the... [01:22:15] (03PS34) 10Mooeypoo: [WIP] Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [01:24:12] 3VisualEditor-EditingTools, VisualEditor: VisualEditor: When inserting an image, set its caption by default to be the Commons image description - https://phabricator.wikimedia.org/T53032#961643 (10Jdforrester-WMF) a:3Mooeypoo [01:29:51] (03PS4) 10Divec: Native backspace and delete [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/179223 [01:30:55] 3OOjs-UI, VisualEditor: MediaWiki theme: Toolbar popup widgets have no visible anchor - https://phabricator.wikimedia.org/T85998#961649 (10KHammerstein) May and I are working on this and will get back soon! [01:31:34] 3OOjs-UI, VisualEditor: MediaWiki theme: Toolbar border too strong - https://phabricator.wikimedia.org/T85974#961650 (10KHammerstein) Good point Ed, we're working on solutions for this too. [01:33:08] (03CR) 10jenkins-bot: [V: 04-1] Native backspace and delete [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/179223 (owner: 10Divec) [01:48:10] 3VisualEditor: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#961678 (10Negative24) 3NEW [02:08:23] 3VisualEditor, VisualEditor-MediaWiki-Media: In the VisualEditor media selector, show the name of each file so it can be identified - https://phabricator.wikimedia.org/T54459#961697 (10Jdforrester-WMF) [04:47:52] 3VisualEditor: There should be a "magazine" option in the VisualEditor citation dropdown - https://phabricator.wikimedia.org/T86117#961761 (10MarkTraceur) 3NEW [07:24:17] (03PS17) 10Paladox: Re add svg images [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/181786 (https://phabricator.wikimedia.org/T37342) [07:26:12] 3VisualEditor, Editing: "Future of Editing" plenary session at 2015 MediaWiki Developer Summit - https://phabricator.wikimedia.org/T85152#961954 (10Qgil) [07:33:35] 3OOjs-UI, UI-Standardization: Our new front-end platform – OOjs UI - https://phabricator.wikimedia.org/T86072#961985 (10Qgil) [10:59:28] 3MediaWiki-ResourceLoader: Errors in one module can affect other modules via $.ready - https://phabricator.wikimedia.org/T72772#962314 (10Krinkle) * This applies to any event handler, not just DOMContentLoaded. E.g. click events: http://jsfiddle.net/mj8yeLmy/. Most instances of domready-handlers should probably... [11:07:53] (03PS1) 10Prtksxna: MediaWiki Theme: Remove text-shadow on button [oojs/ui] - 10https://gerrit.wikimedia.org/r/183468 [11:14:41] 3MediaWiki-ResourceLoader: ResourceLoader: Support CoffeeScript in MediaWiki:Common.js (and such) - https://phabricator.wikimedia.org/T40236#962328 (10Krinkle) p:5Low>3Volunteer? Performance and scalability aside, I don't expect Wikimedia to adopt CoffeeScript. And providing it as an option to gadgets wouldn... [11:18:06] (03CR) 10Prtksxna: "Waiting for May to +1 this before adding more reviewers." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183468 (owner: 10Prtksxna) [11:25:38] (03PS1) 10Prtksxna: MediaWiki Theme: Fix height and vertical text alignment on buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183474 [11:28:45] (03CR) 10Prtksxna: "Waiting for May to +1 this before adding more reviewers." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183474 (owner: 10Prtksxna) [11:47:00] (03CR) 10GOIII: [C: 031] Use mw.Api() instead of $.ajax() [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/183237 (owner: 10Gerrit Patch Uploader) [11:50:04] (03PS1) 10Prtksxna: MediaWiki Theme: Fix focus state for buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183479 [11:59:57] (03PS1) 10Prtksxna: MediaWiki Theme: Fix disabled state of buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183481 [12:16:47] (03PS1) 10Prtksxna: MediaWiki Theme: Fix overlap between hover and active states [oojs/ui] - 10https://gerrit.wikimedia.org/r/183482 [12:17:43] 3OOjs-UI, UI-Standardization: Our new front-end platform – OOjs UI - https://phabricator.wikimedia.org/T86072#962447 (10Qgil) p:5Triage>3Normal [12:22:45] 3MediaWiki-Page-editing: Cite button of RefToolbar not working - https://phabricator.wikimedia.org/T86039#962473 (10Aklapper) 5Open>3stalled [12:23:10] (03PS1) 10Esanders: Make @anchor-size a less variable and calculate borders from it [oojs/ui] - 10https://gerrit.wikimedia.org/r/183484 [12:33:09] 3VisualEditor: Ina 2 x 4 table, inserting column before/after gives JS error “Uncaught TypeError: Cannot read property 'isPlaceholder' of null” - https://phabricator.wikimedia.org/T86160#962507 (10Swainr) 3NEW [12:35:22] 3VisualEditor: Ina 2 x 4 table, inserting column before/after gives JS error “Uncaught TypeError: Cannot read property 'isPlaceholder' of null” - https://phabricator.wikimedia.org/T86160#962531 (10Swainr) Found in production too. [12:44:05] (03PS1) 10Esanders: Fix styling of MWCategoryWidget to match MW theme [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183486 (https://phabricator.wikimedia.org/T78039) [12:46:01] 3OOjs-UI, VisualEditor, VisualEditor-MediaWiki: Clash between rounded corners and text content is jarring for the category editor in VisualEditor when using the OOjs UI MediaWiki theme - https://phabricator.wikimedia.org/T78039#962546 (10Esanders) The rounded corners are just left over the from Apex skin where t... [13:06:01] 3VisualEditor: Ina 2 x 4 table, inserting column before/after gives JS error “Uncaught TypeError: Cannot read property 'isPlaceholder' of null” - https://phabricator.wikimedia.org/T86160#962575 (10Catrope) a:3Esanders [13:13:57] 3VisualEditor-Tables, VisualEditor: Ina 2 x 4 table, inserting column before/after gives JS error “Uncaught TypeError: Cannot read property 'isPlaceholder' of null” - https://phabricator.wikimedia.org/T86160#962595 (10Krenair) [13:15:15] 3VisualEditor, VisualEditor-MediaWiki: VisualEditor: Where VisualEditor is the primary editor, redlinks in read mode should go to veaction=edit not action=edit - https://phabricator.wikimedia.org/T55441#962601 (10Krinkle) I think the editor should initialise at a canonical url. Perhaps with a way to force a spec... [13:22:17] (03CR) 10Krinkle: [C: 032] Use mw.Api() instead of $.ajax() [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/183237 (owner: 10Gerrit Patch Uploader) [13:22:22] (03CR) 10Krinkle: "Thanks!" [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/183237 (owner: 10Gerrit Patch Uploader) [13:22:43] (03Merged) 10jenkins-bot: Use mw.Api() instead of $.ajax() [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/183237 (owner: 10Gerrit Patch Uploader) [13:25:34] (03PS2) 10Esanders: Make @anchor-size a less variable and calculate borders from it [oojs/ui] - 10https://gerrit.wikimedia.org/r/183484 [13:29:38] 3VisualEditor-Tables, VisualEditor: In a 2 x 4 table, inserting column before/after gives JS error “Uncaught TypeError: Cannot read property 'isPlaceholder' of null” - https://phabricator.wikimedia.org/T86160#962628 (10Esanders) [13:31:01] 3VisualEditor-Tables, VisualEditor: Merging cells to 'delete' entire rows or columns causes errors - https://phabricator.wikimedia.org/T86160#962507 (10Esanders) [13:34:09] 3VisualEditor-Tables, VisualEditor: Merging cells to 'delete' entire rows or columns causes errors - https://phabricator.wikimedia.org/T86160#962647 (10Esanders) The problem is that you are generating an invalid table by merging entire rows. The HTML you generate, with rowspan="3" but not actually spanning any r... [13:35:43] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962651 (10Lokal_Profil) 3NEW [13:41:57] 3VisualEditor: VisualEditor language screenshots come out with pixelated fonts - https://phabricator.wikimedia.org/T78648#962675 (10Amire80) This patch fixes the pixelation. Thank you. When it's merged and the new non-pixelated screenshots are uploaded to Commons, this bug can be marked as resolved. [13:43:26] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962678 (10Krenair) It sounds to me like you are reporting a feature as a bug...? Also, is that really an ogg (audio) file or has Phabricator broken it while uploading? [13:51:53] 3VisualEditor, VisualEditor-MediaWiki: VisualEditor: Where VisualEditor is the primary editor, redlinks in read mode should go to veaction=edit not action=edit - https://phabricator.wikimedia.org/T55441#962687 (10Ricordisamoa) I support not making assumptions about the editor, but I think making action=edit poin... [13:52:16] RoanKattouw, um wat https://en.wikipedia.org/wiki/CSCL_Globe?veaction=edit [13:52:24] where did all the CSS go? [13:55:40] Looks like Parsoid is stripping class and style attributes [13:58:43] Hmm some styling did disappear for me but not all of it [13:58:45] Looking [13:59:55] Yeah you're right [14:00:08] The style and class attribute is dropped on the [14:00:14] but not on the
s, those are still blue [14:00:31] I think it's the multi-part template [14:00:45] regular infobices are ok [14:01:09] I think it might be because the class&style attributes are template-generated [14:01:13] With {| synta [14:01:17] Let's see if I can reduce this [14:02:40] yeah {|{{Infobox ship begin}} [14:03:49] I thought that syntax was wrong but https://en.wikipedia.org/wiki/Template:Infobox_ship_begin#Repeating_sections [14:03:51] WHYYY! [14:05:24] Yeeaaah [14:05:48] {{infobox ship begin}} expands to: [14:05:49] class="infobox" style="width:315px;float:right;clear:right;border-spacing:2px;font-size:90%" [14:05:51] |+ [14:06:12] I'm just stepping on Parsoid mines everywhere: https://en.wikipedia.org/wiki/List_of_largest_container_ships?veaction=edit [14:06:46] SWEET that reproduces it [14:07:04] Ouch [14:07:13] OK I'm gonna file a bug for that CSCL Globe one [14:07:19] Because I have a minimal test case for it [14:08:37] Bhaha and it also dirty-diffs, classic [14:08:55] {|{{echo|1=bunch of garbage round-trips to {| {{echo|1=bunch of garbage [14:08:59] (added space between {| and {{) [14:09:23] https://phabricator.wikimedia.org/T86169#962729 [14:10:15] RoanKattouw, you'll want to modify that one ^^ [14:11:34] What the actual fuck [14:11:54] So tests/parse.js has a --normalize option that cleans up the output for test purposes, by removing data-parsoid etc [14:12:03] --normalize removes class but not style [14:12:09] --normalize=parsoid removes style but not class [14:13:05] 3VisualEditor-Tables, VisualEditor: Merging cells to 'delete' entire rows or columns causes errors - https://phabricator.wikimedia.org/T86160#962778 (10Esanders) This is actually conceptually not that straightforward, consider the following merge operation: [14:15:49] RoanKattouw, I haven't given it much though yet but if you have any on this ^^, how to check for row/col deletion when merging [14:15:57] *ideas on this [14:23:15] Oh geez [14:23:32] I see what's going on in your latest comment [14:23:39] A conceptual row disappears due to a merge [14:23:46] But how is that happening in the originally reported table? [14:24:00] It seems like that one still has all of its conceptual rows and columns? [14:24:31] Oh, I guess you've merged rows 2-4 to become row 2 [14:25:06] right [14:25:11] And browsers choke on this? [14:25:11] So I think the answer is: our model stores the full 4x4 matrix, with 'real' cells in the top left corner, and 'placeholder' cells where things have been merged in. [14:25:34] so if we check across all affect rows and colummns for a row or column filled entirely with placeholders, then it should be deleted [14:25:41] *affected [14:26:01] browsers choke and so does our code in some situations [14:26:33] 3VisualEditor-Tables, VisualEditor: Tables: If you select N rows / columns and click "insert above/below" or "… left/right", VE should insert N rows / columns, not 1 - https://phabricator.wikimedia.org/T75227#962812 (10matmarex) I realize that other text editors do this, but I've always found it to be annoying a... [14:26:34] Hmm yeah that sounds sensible [14:28:10] edsanders: Can I pick your brain about a problem of mine, BTW? [14:28:50] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962824 (10Krenair) [14:29:31] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962828 (10matmarex) [14:29:34] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962651 (10matmarex) >>! In T86164#962678, @Krenair wrote: > Also, is that really an ogg (audio) file or has Phabricator broken it while uploading? It's an ogg video fil... [14:30:08] MatmaRex, yeah I'm just looking into that now [14:30:12] sure [14:30:32] So, I'm rewriting some of InternalList's reference usage tracking in DocumentSet [14:30:44] ok [14:30:48] This is the firstNodes/keyedNodes etc stuff, tracking which citation is used by which MWReferenceNodes [14:31:37] I decided to simplify the data structure a bit so that all I have now is nodes = { name: [ node1, node2 ] } and nameOrder = [ name1, name2 ] [14:32:18] So for each name, I have an array with references to the nodes that use it, in order, and I also have an array with the order of the names (ordered by first occurrence, so name1 < name2 if nodes[name1][0].getOffset() < nodes[name2][0].getOffset() ) [14:32:25] MatmaRex, https://secure.phabricator.com/T6674 [14:32:36] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962831 (10Krenair) yep, here: https://secure.phabricator.com/T6674 [14:32:38] This is enough data to do everything that InternalList was doing before, reference numbering-wise [14:32:40] However [14:32:54] I'm using binary search insertion and deletion to keep the arrays ordered [14:33:03] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962833 (10Lokal_Profil) @krenair @matmarex : T75990 [14:33:06] Which is working very well (after introducing a ve.binarySearch helper function) [14:33:10] BUT [14:33:48] When MWReferenceNode#onUnroot is called and wants to deregister itself from these tracking arrays, node.getOffset() returns 0 because the node has already been unrooted [14:34:14] And even if you defer unrooting, node.parent is already null, and even if you defer unsetting node.parent, node.parent.children[] already doesn't contain node anymore [14:34:40] So it appears to be difficult to impossible to determine what a node's offset is (or more accurately at that point, used to be) in onUnroot [14:35:33] 3VisualEditor-Tables, VisualEditor: Highlighting table cells differ before and after entering VEedit - https://phabricator.wikimedia.org/T86164#962838 (10Lokal_Profil) >>! In T86164#962678, @Krenair wrote: > It sounds to me like you are reporting a feature as a bug...? It's a bug in that it is unexpected and in... [14:35:47] But in order to remove the node from the array (and possibly the name from the names array, if there are no more nodes left with that name), I need to find it, and in order to find it I need its offset (which binary search used to sort it) [14:35:54] So, I've considered several options [14:36:01] 1. Use linear search instead (meh) [14:36:54] 2. Come up with some sort of simple caching in .getOffset() (maybe based on completeHistory.length, like SurfaceFragment) and abuse that because it would likely survive detachment [14:37:56] 3. Instead of putting node references and strings into my arrays directly, put them in wrapper objects that explicitly store the offsets they were indexed with [14:38:25] 3b. Because keeping those offsets in sync sounds nasty, maybe have the wrapper objects perform the kind of magic that I was trying to get in #1 [14:38:41] 4a. Use SurfaceFragments 4b.??? 4c. Profit [14:38:56] Those are all my ideas and I don't really like any of them so far [14:41:07] I'm leaning towards 3b a bit I think, also because "a < b if a.getOffset() < b.getOffset()" glosses over the fact that a and b can now be in different documents (like if one of them is in a caption) [14:42:13] I have an idea for how to handle that, and I think 3b might provide a nice abstraction point for that [14:57:09] (My first thought was using something like (a.document.index, a.offset) < (b.document.index, b.offset) but that's also bad because that puts nested references after all top-level references, so what I think you actually want is something like (a.document.firstUseOffset, a.offset) < (b.document.firstUseOffset, b.offset) ) [15:03:38] Morning. [15:03:42] edsanders: Can I grab +2s from you on two wmf14 cherry-picks so I can schedule them for deployment? https://gerrit.wikimedia.org/r/#/c/183330/ and https://gerrit.wikimedia.org/r/#/c/183409/ [15:04:14] so my internet is currently down - but apparently it's just DNS because this is still working [15:04:25] edsanders: Odd. [15:06:00] wikitech still has a broken install of ve [15:07:08] RoanKattouw, so 1. how much time does binary search actually save you? is it premature optimisation? [15:07:44] and 2. what happens if we didn't remove the items on unroot, but just left them as dead references? [15:08:44] Krenair: Yeah, has for months. [15:11:40] 3VisualEditor: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#962963 (10Krenair) Cannot reproduce on beta.wmflabs.org. Where is that wiki you're testing on? [15:12:43] parsoid works there: http://parsoid-lb.eqiad.wikimedia.org/labswiki/LDAP?oldid=128750 [15:13:13] think it's just got the wrong parsoidurl config or something [15:14:07] (03CR) 10Krinkle: "Why was the @demo removed from ve.init.sa.Target? That was our only jsduck demo that ensured the toolbar worked properly and as showcase i" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177109 (owner: 10Esanders) [15:16:47] edsanders: I think for the nodes array it may well be premature optimization, because there are rarely going to be >5 reuses of the same reference [15:17:12] For the names array, though, I don't think it is, because it's not that strange to have 100s or reference names in use [15:17:25] (this is all namespaced per ref group, but still, Obama has like, what, 300 refs in the main group?) [15:17:41] sure, but a iterating over 100 things every once in a while isn't per se slow [15:17:48] provided it's a quick comparison [15:18:02] Yeah, maybe not [15:18:18] the error from it doesn't make sense though. it's trying to look the page name up as a hostname? [15:18:37] maybe VisualEditorParsoidURL and VisualEditorParsoidPrefix are unset somehow [15:18:55] For removals it may well be faster to iterate over the array and using a linear number of === operations to find the node, rather than using a logarithmic number of .getOffset() calls [15:18:57] Krenair: Is Wikitech pointed at a different URL for Parsoid? [15:19:00] Krenair: Or that. [15:19:28] 3operations, MediaWiki-Core-Team, MediaWiki-ResourceLoader: Bad cache stuck due to race condition with scap between different web servers - https://phabricator.wikimedia.org/T47877#963015 (10chasemp) p:5High>3Normal Seems like no meaningful update in a month, no assignee, and no direct actionable that I can... [15:19:33] === is probably way way faster than .getOffset() to the point where 1000 === ops are probably still faster than lg(1000)~=10 .getOffset() ops [15:20:43] Hmm and I only have this problem when removing, never when inserting, so that should work just fine [15:20:55] I don't think VE exposes those variables to the client though, will have to get a shell user to check :/ [15:23:06] edsanders: So it turns out that ship infobox bug was filed in Bugzilla in 2013 (!), and Subbu's just about finished writing a patch for it [15:23:16] Krenair: Both null according to eval on tin [15:23:17] :| [15:23:18] With a great commit message: https://gerrit.wikimedia.org/r/#/c/173834/ [15:23:19] nice [15:25:48] RoanKattouw: Indeed. [15:26:01] edsanders: Ping re. the merge please. [15:26:39] (03CR) 10Esanders: [C: 032] Work around $.parseHTML( '' ) returning null [extensions/VisualEditor] (wmf/1.25wmf14) - 10https://gerrit.wikimedia.org/r/183330 (https://phabricator.wikimedia.org/T86056) (owner: 10Jforrester) [15:26:55] (03CR) 10Esanders: [C: 032] Fix Parsoid transformation calls [extensions/VisualEditor] (wmf/1.25wmf14) - 10https://gerrit.wikimedia.org/r/183409 (https://phabricator.wikimedia.org/T86046) (owner: 10Jforrester) [15:27:12] Thanks edsanders. [15:28:57] (03Merged) 10jenkins-bot: Work around $.parseHTML( '' ) returning null [extensions/VisualEditor] (wmf/1.25wmf14) - 10https://gerrit.wikimedia.org/r/183330 (https://phabricator.wikimedia.org/T86056) (owner: 10Jforrester) [15:29:12] RoanKattouw: How do we load html into a Target? I'm adding the @demo back to sa.Target and it seems something exploded in the demo. There's a ton of logic. [15:29:14] (03Merged) 10jenkins-bot: Fix Parsoid transformation calls [extensions/VisualEditor] (wmf/1.25wmf14) - 10https://gerrit.wikimedia.org/r/183409 (https://phabricator.wikimedia.org/T86046) (owner: 10Jforrester) [15:29:29] Do callers now need to create their own surface and parse the html separate from the dom? [15:29:31] I'm confused. [15:29:32] Krinkle: Ask edsanders , he changed all that stuff with the multi-surface thing [15:29:40] There's no documentation [15:31:34] (03PS1) 10Krinkle: ve.init.sa.Target: Add jsduck @demo for simplest editor [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183508 [15:31:44] (03CR) 10Krinkle: "To be fixed in I1786ad222e8d08a14d50ef2bbaed1." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/177109 (owner: 10Esanders) [15:32:31] (03CR) 10Krinkle: [C: 04-1] "Currently broken. Demo is overcomplicated, I assume there's a simple way still to give VE html and have it be parsed into a document and r" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183508 (owner: 10Krinkle) [15:32:33] edsanders: ^ [15:34:27] Target used to be the entry point from where other classes can be discovered and a working editor can be created. There's no documetnation in its constructor, but maybe we it moved elsewhere? I'm happy to do it differently. [15:35:03] (03PS1) 10Esanders: Remove redundant rows/cols after cell merge [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183510 (https://phabricator.wikimedia.org/T86160) [15:35:20] ym [15:36:37] yes, target's are initialised without surfaces [15:36:49] Phabricator doesn't mention the branch in commit notifications? [15:36:58] as a target can have multiple surfaces [15:37:20] I know in this case that it was wmf/1.25wmf14, but that should be in the email... [15:37:50] probably - it was a breaking change [15:38:19] Krenair: It doesn't? [15:39:27] Krinkle, I don't think you've been able to create from HTML directly for a while [15:39:32] edsanders: multiple ui or ce surfaces? For the same dm document, or different ones? [15:39:44] multiple ui surfaces [15:39:51] with Flow in mind [15:40:32] the old target constructor took a dmDoc [15:40:34] James_F, nope [15:40:55] one could add a version of addSurface which takes html as an argument [15:41:01] edsanders: Right, so that one can have an inline add-reply area that constructs its own editor. But wouldn't those be separate targets? [15:41:04] Krenair: Oh, in the "this patch was added" hack? I don't think that ever did on Bugzilla either. [15:41:22] they only share the same platform. [15:41:23] no [15:41:27] not maniphest [15:41:48] Krinkle, targets have only ever been singletons on the page [15:41:55] and this way they can share a toolbar [15:42:21] assuming the same buttons/menus are wanted for both. [15:42:31] But OK [15:42:56] Target had a global instance (ve.init...target) , but afaik it wasn't meant to stay a singleton. [15:42:56] James_F, https://phabricator.wikimedia.org/P198 [15:43:10] Anyway, slicing it up at the surface level instead of target works just as well I suppose. It's just names. [15:43:19] certainly the way the MW hooks up it is [15:43:20] Krenair: Oh, that. [15:43:47] edsanders: Yes, but that was all up for refactoring at the time. I wrote a fair bit of it. Anyway,it's been taken into a different direction, I;m fine with that. [15:43:50] * James_F goes for breakfast. [15:44:36] edsanders: So what is the minimal code path to getting an editor working from a piece of input html? Could you extend that @demo block? I"ll find out the methods and documentation bits from there and write it up. [15:44:37] (I filed https://secure.phabricator.com/T6897 ) [15:44:57] Removing the block was a rather easy way out of that breaking change.. [15:44:58] It's as in demo.js [15:44:59] this.surface = this.target.addSurface( [15:44:59] ve.dm.converter.getModelFromDom( [15:44:59] ve.createDocumentFromHtml( pageHtml ), [15:44:59] this.target.$element.ownerDocument, [15:44:59] this.lang, [15:45:00] this.dir [15:45:02] ) [15:45:04] ); [15:45:06] this.target.setSurface( this.surface ); [15:45:38] the last line (setSurface) is unnecessary in a minimal case [15:45:49] There is also a platform.done() wrapper in demo.js [15:45:51] as that is called automatically if the surface list was previously empty [15:46:02] that's for message loading I think [15:47:02] so you can include that too [15:47:19] That's a lot of code for a simple surface. [15:47:51] Did we used to make assumptions for lang/dir? Or was that not supported? [15:47:57] I'm just curious where it all went [15:49:12] OK. The other arguments are optional in getModelFromDom(). [15:49:53] (03CR) 10Esanders: "Code looks good but some IME + RTL testing would be good." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/179223 (owner: 10Divec) [15:50:00] Krinkle: Yeah getModelFromDom() defaults to 'en', 'ltr' there I thin [15:50:07] (or at least indirectly defaults to those values) [15:50:58] (03CR) 10Esanders: "Maybe you could add a page full of strange unicode to the demo for that purpose?" [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/179223 (owner: 10Divec) [15:50:59] edsanders: Ahm, wait, this.target.$element.ownerDocument should surely be this.target.$element[0].ownerDocument ? [15:51:18] RoanKattouw, yup [15:51:29] Also the fact that you have to pass that manually is a code smell, since you're already calling a method on this.target. It would be nice to have a wrapper for that [15:51:52] yup [15:52:01] RoanKattouw: it's not [15:52:04] RoanKattouw: it's optoinal [15:52:24] Krinkle: But it defaults to the global document, probably? [15:52:36] RoanKattouw: it defaults to the document passed to it as first argument [15:52:45] Which, I suppose as soon as we finally kill this.$ and frame support and crap, that'll be fine [15:52:56] Krinkle: Huh wtf [15:52:59] Which is not the visual document... [15:53:01] What is the name of that paranmeter [15:53:09] doc, surfaceDoc [15:53:20] RoanKattouw: Wiat a second, Target#addSurface doesnt take Surface as parameter? [15:53:26] Where does the surface come from [15:53:44] createSurface [15:53:53] except that one already exists [15:54:10] Krinkle: Yeah targetDoc should NOT be set to the visual document, that's broken. If Ed copy-pasted that code then we are incredibly lucky, because he tried to pass the wrong document but actually passed undefined which triggers the default which happens to be the correct behavior [15:54:38] Krinkle: No, one doesn't already exist [15:54:44] it does [15:54:47] Target#createSurface [15:54:53] this.surface is the return value of addSurface() which is the return value of createSurface [15:54:59] getModelFromDom returns a ve.dm.Document [15:55:22] Target#addSurface( dmDoc, config ) -> Target#createSurface( dmDoc, config ), surfuaces.push() ..something.connect(); return; } [15:55:46] Yeah [15:55:56] I'd assume addSurface takes a Surface [15:56:01] e.g. createSurface() return [15:56:03] haha yeah [15:56:18] You would think that addSurface() would take a surface [15:56:22] I don't really like that it doesn't [15:56:30] But I think that's a naming problem rather than a type problem [15:56:41] Because the knowledge of how to create the surface is also in the target [15:56:47] RoanKattouw: So then there's just no way for it to have a different visual HTMLDocument other than the global one [15:56:50] or is it a config option [15:57:10] Krinkle: The visual document is managed elsewhere, not in getDomFromModel() [15:57:18] I get that [15:57:19] Remember, getDomFromModel() is a DM method and DM doesn't know about CE [15:57:28] So wherever the visual document is passed in, it's not there [15:57:33] RoanKattouw: But after addSurface() runs, the surface is done and already decided on a visual docuemnt [15:57:45] It's probably the ui.Surface's or the ce.Surface's .$element[0].ownerDocument [15:58:06] Krinkle: Well yes, because addSurface() also performs target.$element.append( surface.$element ) [15:58:11] so target and surface must be in the same document [15:58:19] And target is already attached, so its document is already set [15:58:47] addSurface -> createSurface - >DesktopSurface -> Surface [15:58:51] Nope, no document option anywhere [15:59:00] It's implicit in config.$ I would think [15:59:19] var surface = ve.init.sa.Target.super.prototype.addSurface.apply( this, arguments ); [15:59:20] this.$element.append( $( '< [15:59:20] Not used in that case, [15:59:23] $() [15:59:32] Anyway, don't need an iframe here [15:59:38] No we don't [15:59:48] But it should probably support that [15:59:52] I think to answer your question though I'd have to look in ce.Surface or ce.Document, let me see [16:00:23] Krinkle: Yeah see the ve.ce.Document constructor [16:00:41] new ve.ce.DocumentNode( model.getDocumentNode(), surface, { $: surface.$ } ) [16:00:47] That's how the view document is passed to CE [16:01:02] Because uiSurface is an OO ui Element [16:01:05] Right [16:01:15] otherwise config.$ would never end up as surface.$ [16:01:16] Yes, and so is ceSurface, and in fact so are all ce.*Node classes [16:01:26] The latter is obvious [16:01:30] but Surface being element is new [16:01:37] cool [16:02:12] I think ce.Surface and ui.Surface have been elements for a long time; but anyway, yeah, as long as you manage to pass the correct config.$ to ve.ce.DocumentNode, everything else follows clearly [16:02:33] Also, as you may know, Trevor and I plan to eliminate .$ completely (except for b/c) [16:03:43] (03PS2) 10Krinkle: ve.init.sa.Target: Add jsduck @demo for simplest editor [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183508 [16:04:05] RoanKattouw: So now ^ the toolbar, dm and ce work, but messages are broken. [16:04:27] http://i.imgur.com/6lZgbOF.png [16:04:37] I don't know what changed there [16:04:45] AFAIK messages are managed in Target, not Surface [16:04:51] I really don't want to add a promise callback around the example to wait for that to load... [16:04:57] Or Platform, or whatever [16:05:00] yes [16:05:01] platform [16:05:10] Yeah exactly, should not be the caller's responsibility in any case [16:05:14] the demo was changed in ed's breaking commit to be wrapped in a giant platform.done() [16:05:19] You should just be able to .addSurface() something [16:05:28] Hmm [16:05:42] Had we kept this jsduck demo up to date, it would've been obvious. [16:05:44] Maybe that's because the demo has its own mechanism for loading messages because it allows you to switch languages? [16:05:49] Which the base platform doesn't let you do [16:06:14] The take-away, I think, is that maybe the demo should have its own platform subclass [16:06:17] And maybe also its own target subclas [16:06:21] Rather than just being a JS file [16:06:24] Dogfooding etc [16:06:45] Krinkle: However I'm surprised that the sa Platform+Target don't have functional message loading, that's a problem [16:08:40] (03CR) 10Krinkle: [C: 04-1] "Messages broken. Roan says, this should be taken care of by Target internally, not by the caller." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183508 (owner: 10Krinkle) [16:08:58] I've gotta get back to the CI stack I was working on before I forget what I was doing. [16:09:22] I just noticed a demo was removed and as I suspected it caused a use case or principle to be missed. [16:09:30] Well, excellent catch [16:09:38] We did apparently break the i18n system in some way [16:25:22] 3VisualEditor-Tables, VisualEditor: Merging cells to 'delete' entire rows or columns causes errors - https://phabricator.wikimedia.org/T86160#963155 (10Jdforrester-WMF) p:5Triage>3High [16:25:29] 3VisualEditor-Tables, VisualEditor: Merging cells to 'delete' entire rows or columns causes errors - https://phabricator.wikimedia.org/T86160#962507 (10Jdforrester-WMF) p:5High>3Normal [16:27:21] 3VisualEditor, VisualEditor-MediaWiki-References: [Regression wmf13] Creating a citation without filling in a required field throws "Cannot read property 'createElement' of null" - https://phabricator.wikimedia.org/T78626#963158 (10Jdforrester-WMF) p:5Triage>3Normal [16:36:56] (03PS1) 10Jforrester: [WIP] Move contents of VisualEditor.php to extension.json [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183524 [16:46:00] James_F: Oooh fancy [16:46:12] RoanKattouw: More fancy if it worked. :-) [16:46:19] "Don't even look at this yet" [16:46:19] RoanKattouw: It landed in core last night. [16:46:22] RoanKattouw: Yeah. :-) [16:46:29] * RoanKattouw takes that to heart and goes back to preparing his talk [16:46:32] :-) [16:46:34] James_F: What landed in core last night? [16:46:44] RoanKattouw: legoktm's extension.json support. [16:47:14] Link? [16:47:51] gerrit.wikimedia.org?s=something [16:48:11] I can find it only as fast as you, except asking takes twice as long for each of us. :-) [16:48:23] https://gerrit.wikimedia.org/r/#/c/166705/ [16:49:15] :D [16:49:53] legoktm: Any documentation would be appreciated, BTW. [16:50:19] I'm writing up an email to send to wikitech about it [16:50:25] Cool. [16:52:07] Speaking of preparing my talk, Chrome apparently fixed a bug I reported but didn't tell me [16:52:20] I mean, no one ever responded to my report, so they probably never saw it to begin with [16:52:29] Dupe of another, maybe? [16:52:32] Krinkle: How can I close https://code.google.com/p/chromium/issues/detail?id=314291 as "works now"? [16:52:37] Or maybe a dupe of something else or something? [16:53:03] I tried logging in with both my @gmail.com and my @wikimedia.org accounts and I don't see a UI affordance for changing the status of the bug [17:03:26] 3MediaWiki-Page-editing: text input history/autocomplete doesn't work in IE + https - https://phabricator.wikimedia.org/T55636#963330 (10Aklapper) p:5Low>3Volunteer? [17:19:54] RoanKattouw_away: Only google CEO can edit issues once filed on google code issue trackers [17:20:03] Or a slight super set of that [17:20:05] I dunno [17:20:12] But not the author of the issue [17:20:31] James_F: I created a nice badge for you on test2wiki [17:20:38] Might help confused users fromthinking their on wikipedia [17:20:43] since test2wiki has enwiki logo [17:20:52] they're [17:20:53] For me? [17:20:59] Just for you :) [17:21:02] No, but for you. [17:21:13] I see you also ran into https://test2.wikipedia.org/wiki/User_talk:Pinky_Jones [17:23:49] RoanKattouw_away: The comment you left is sufficient. They'll close it at some point. [17:24:13] Krinkle: As quickly as they looked at an triaged the issue when I filed it? :P [17:24:56] 2-3 days yeah [17:25:17] Krinkle: All interaction I have ever gotten from anyone with a @chrome or @google address on bugs I have filed (excluding batch/automated processes or tagging) was a single "I can confirm this on my machine" comment [17:30:34] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=110691088 [17:30:34] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=110700919 [17:30:34] Corruption alert: visualeditor-needcheck on zhwiki: https://zh.wikipedia.org/?diff=33877424 [17:30:34] Corruption alert: visualeditor-needcheck on zhwiki: https://zh.wikipedia.org/?diff=33883777 [17:30:34] Corruption alert: visualeditor-needcheck on ukwiki: https://uk.wikipedia.org/?diff=15398417 [17:36:04] RoanKattouw_away: Yeah, I hear ya [17:36:09] RoanKattouw_away: They're not the most responsive. [17:36:13] Though things do happen. [17:36:15] Somehow [17:37:34] RoanKattouw_away: I've gotten success on quite a few issues though. [17:37:35] https://code.google.com/p/chromium/issues/detail?id=443203 [17:37:41] https://code.google.com/p/chromium/issues/detail?id=445794 [17:37:43] at least responses [17:37:48] from collaborators [17:43:13] 3VisualEditor: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#963582 (10Negative24) I've been working on the Performance instance on WMF Labs. You can see it [[ http://performance.wmflabs.org | here ]] (performance.wmflabs.org) but you pr... [17:46:48] I think in total about 30 bugs I've starred have been fixed. [18:22:35] 3MediaWiki-Page-editing: Cite button of RefToolbar not working - https://phabricator.wikimedia.org/T86039#963787 (10MKar) 5stalled>3Invalid [18:41:22] 3MediaWiki-Page-editing: text input history/autocomplete doesn't work in IE + https - https://phabricator.wikimedia.org/T55636#963824 (10chasemp) [18:44:19] huh, I could've sworn cy wiki used to have a customised special character inserter [18:44:48] https://cy.wikipedia.org/w/index.php?title=Mathemateg&action=edit [18:45:58] 'Nodau arbennig' is the standard wikieditor SCI [18:46:55] 3VisualEditor, VisualEditor-MediaWiki: Can't edit https://pl.wikimedia.org/ (Polish chapter wiki) using VE - https://phabricator.wikimedia.org/T86216#963851 (10matmarex) [18:47:29] ^ not nice [18:48:57] RoanKattouw, Krinkle|detached: would you have half an hour to(day|night) for IRC perf work triage? [18:49:10] Ahm, sure [18:49:39] ori: Also, what were you saying when the room hung up? "I think it would be really worthwhile to speak with uhm " - disconnect [18:50:32] Steven Walling about guided tours [18:50:34] ori: But can we do it soon? It's almost 8pm, I have to leave the house at 9am for my flight, want to go to sleep early and still need to pack [18:50:37] Ooooh [18:50:41] Yeah [18:51:10] sure, give me 5 minutes to gather my thoughts? [18:51:29] Actually when I was at the WMNL office I spoke to Sebastiaan (who was in SF the week before Christmas and filmed in the office) and he said he was toying with the idea of applying for a grant to improve the state of tutorial videos and teaching resources etc about editing [18:51:49] Also observing that the quality/presence of such materials is bad for wikitext but worse for VE [18:51:53] OK sure [18:52:34] shower thought: there's no 'I' in team, but there are both 'I' and 'ME' in IME. [18:55:47] ori: :-) [18:56:48] / There's no "I" in Timocil / / at least not where you think / [18:56:53] *Teamocil [18:56:57] (Freudian slip) [18:57:15] RoanKattouw: :-P [18:57:24] "There's no 'you' in 'team'; you're fired." [18:58:24] "If you can't act like a team you're all fired" [19:02:29] RoanKattouw, Krinkle: http://etherpad.wikimedia.org/p/VEQ3Perf [19:02:50] this is my 10,000-foot understanding of what the priorities are perf-wise, but I want to know what you think [19:03:59] ori: Not tonight. Probably monday, which is when I've arrived in SF. [19:04:15] Krinkle: k, no problem. [19:04:31] Send me an e-mail or invite, or I'll e-mail the link to myself now. [19:05:05] Krinkle: nominally related: I really liked [19:06:56] Just got notification that I was mentioned on village pump (tech) [19:07:12] "and the vertical-align: top; which was added by Ed g2s on August 11, 2005, is why all of the labels are top aligned." [19:07:33] lol [19:08:02] I will not justify the things I did 10 years ago - it is statistically likely I was wasted at the time [19:09:49] although I am impressed that stuff I added to common.css has survived that long [19:10:36] edsanders: :-) [19:11:03] ori: Cool. Will read further later. [19:11:10] ori: Curious, how did you discover that post? [19:11:32] HN [19:11:37] (hackers news) [19:12:40] RoanKattouw: cool, the current etherpad looks like a good overview of the current state of the world. That's all I was after for now, so maybe pack and get some sleep? [19:13:29] James_F: would it be fair to ask you to create tasks for the items that don't have one yet? [19:14:57] ori: I would also like to get input from the Parsoid team, at least for them to tell us what things they're planning that will reduce HTML output size and when those will happen [19:15:00] e.g. moving out data-parsoid [19:15:11] * ori nods [19:15:29] ori: Sure, happy to. [19:15:51] input in what form? is this something we should meet with the parsoid team to discuss, or would an e-mail thread be sufficient? [19:15:51] ori: Also, a note on timezones real quick [19:16:17] I am flying tomorrow and through most of the weekend. Starting Saturday afternoon PST I will be in New Zealand, which is 3 hours behind (but a day ahead) SF [19:16:30] So I will be on an SF-similar timezone but working Sunday - Thursday [19:16:42] (03PS1) 10Bartosz Dziewoński: PHP LabelElement: Actually allow non-plaintext labels [oojs/ui] - 10https://gerrit.wikimedia.org/r/183566 [19:16:47] Krinkle will be arriving in SF on Sunday afternoon I believe [19:16:59] So he will be in the office with you next week [19:17:05] MatmaRex: Only in PHP? mooeypoo was talking about needing this in JS. [19:17:09] I'm coming back over the MLK weekend, in time for the all-staff [19:17:16] James_F: we allow it in JS. [19:17:29] yeah, I'm probably going to make mistakes and be inadvertently insensitive to timezones initially, but feel free to slap me if I make SF-chauvinist scheduling suggestions. I'll try to be mindful so that doesn't happen. [19:17:32] you can set any jQuery object as the label [19:17:34] MatmaRex: Hmm. Moriel thought otherwise. [19:17:46] ori: I think an email thread is probably sufficient. Mainly I want to know timelines for data-parsoid moving and data-mw moving, and learn about any other things they might be doing that will affect output size [19:17:52] this was also supposed to work in PHP, but didn't [19:17:53] hmm [19:17:53] or that they could be doing that would reduce output size [19:18:03] James_F: i'll need reproduction steps in that case :) [19:18:21] ori: That's fine, I'll only be 3 hours off from SF time so it's hard to screw that up [19:18:34] Just know that I'll be pretty much totally absent tomorrow [19:19:51] sure. I think Damon plans to send a "VE all hands on deck" e-mail to wmfall that should grease the wheels for this sort of cross-team requests, so if you send that e-mail in a few days (if you don't mind) that would be good timing both organizationally and for you personally (by the sound of things) [19:20:10] * James_F nods. [19:21:03] OK, will do [19:23:35] James_F: we have an alarming lack of a "secure" icon in OOUI. [19:23:50] MatmaRex: Steal it from VE? [19:23:52] Hmm. [19:23:53] * MatmaRex draws a little lock [19:23:59] Actually, I think it got deleted a while ago. [19:24:11] MatmaRex: There's already several assets for this. Don't make another. :_) [19:24:16] i want one for the login form [19:24:23] So… take one? [19:24:31] existing assets, as far as i can tell, are not monochrome [19:24:35] * MatmaRex decolorizes a little lock [19:24:45] There is one somewhere. TrevorP|Away would know. [19:31:09] (03PS1) 10Anomie: Update ApiResult handling for mediawiki/core change I7b37295e [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183605 [19:31:18] James_F: speaking of TrevorP|Away, can you make him merge that little OOUI patch above? https://gerrit.wikimedia.org/r/183566 [19:31:41] MatmaRex: He may be nearer me than you, but he's still ~20km away. :-) [19:31:59] that slacker. ;) [19:33:06] 3MediaWiki-Page-editing: Cite button of RefToolbar not working - https://phabricator.wikimedia.org/T86039#964007 (10Aklapper) MKar led me through the steps on IRC and this looks a lot like T85489. Classing editing, please go to https://or.wikipedia.org/w/index.php?title=%E0%AC%AD%E0%AC%BE%E0%AC%B0%E0%AC%... [19:34:21] MatmaRex: Does https://gerrit.wikimedia.org/r/#/c/183390/ not need PHP->JS rennervation so you can do isValid() properly? [19:34:27] MatmaRex: Or is the plan for that to wait? [19:35:19] MatmaRex: Also, can you review the 6 things in https://gerrit.wikimedia.org/r/#/projects/oojs/ui,dashboards/default from Prateek and the latest from Ed so I can branch OOjs UI again? [19:36:08] James_F: i was planning to only minimally adjust the existing CSS and JS for login form for now, no OOUI client-side [19:36:12] i'm here [19:36:16] looking at patches [19:36:31] 3MediaWiki-Page-editing: Cite button of RefToolbar not working - https://phabricator.wikimedia.org/T86039#964019 (10Aklapper) So if you check the web browser's console, you see that the offending code line is: $('#cite-date-'+CiteTB.escStr(this.shortform)+'-'+field).live('click', CiteTB.fillAccessdate);... [19:36:32] MatmaRex: Is the plan to do that later? [19:36:51] (03CR) 10Trevor Parscal: [C: 032] PHP LabelElement: Actually allow non-plaintext labels [oojs/ui] - 10https://gerrit.wikimedia.org/r/183566 (owner: 10Bartosz Dziewoński) [19:36:53] MatmaRex: Thinking of how we can have the login form be in a dialog in the middle of an edit sanely. [19:37:02] James_F: IMO it would be best to wait for the login forms to be converted to HTMLForm, but that's not going to happen soon (unless i do it myself) [19:37:18] MatmaRex: That's fine. [19:37:31] MatmaRex: "Not soon" is OK. "Not ever" is what makes me sad. [19:38:37] James_F: please elaborate on 'that's fine'; do you want to wait with the login forms until we have the JS-PHP "integration" in OOUI done? [19:38:55] (03CR) 10Trevor Parscal: [C: 032] Make @anchor-size a less variable and calculate borders from it [oojs/ui] - 10https://gerrit.wikimedia.org/r/183484 (owner: 10Esanders) [19:39:03] (also yes, i can review the patches) [19:39:08] MatmaRex: Not touching the existing JS is fine. [19:39:13] MatmaRex: Sorry, I'm not being clear. [19:39:21] MatmaRex: Looks like Trevor's doing that already. [19:39:56] last time i asked, it was somewhere further down the to-do list [19:40:00] (03CR) 10Trevor Parscal: [C: 032] MediaWiki Theme: Fix overlap between hover and active states [oojs/ui] - 10https://gerrit.wikimedia.org/r/183482 (owner: 10Prtksxna) [19:40:12] MatmaRex: Kk. [19:40:22] (03CR) 10Trevor Parscal: [C: 032] MediaWiki Theme: Remove text-shadow on button [oojs/ui] - 10https://gerrit.wikimedia.org/r/183468 (owner: 10Prtksxna) [19:40:31] MatmaRex: Won't we need it for OO-ifying most pages and the skins, though? Or am I being pessimistic? [19:40:46] (03CR) 10Trevor Parscal: [C: 032] MediaWiki Theme: Fix disabled state of buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183481 (owner: 10Prtksxna) [19:41:37] James_F: okay, so: 1. we want the MW core login and create account forms to be build using PHP OOUI now-ish; 2. we don't need the JS enhancements to use OOUI right now; 3. we want to make them HTMLForms somewhere in the future, but definitely not now? [19:41:48] MatmaRex: Yes. [19:41:58] James_F: prooobably we'll need it. [19:42:01] MatmaRex: Right. [19:42:39] (03CR) 10Trevor Parscal: [C: 032] MediaWiki Theme: Fix focus state for buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183479 (owner: 10Prtksxna) [19:44:18] (03CR) 10Trevor Parscal: "This may cause problems when used in context. I need to test it in VisualEditor or other uses of this library before I can sign off on tha" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183474 (owner: 10Prtksxna) [19:44:40] huh, the images on the logic pages are not SVGs. [19:44:45] (03CR) 10Trevor Parscal: [C: 032] MediaWiki Theme: Add state change transition to checkbox [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:45:09] MatmaRex: Oh dear. [19:45:33] * MatmaRex files a bug [19:47:14] (03Merged) 10jenkins-bot: PHP LabelElement: Actually allow non-plaintext labels [oojs/ui] - 10https://gerrit.wikimedia.org/r/183566 (owner: 10Bartosz Dziewoński) [19:47:33] (03Merged) 10jenkins-bot: Make @anchor-size a less variable and calculate borders from it [oojs/ui] - 10https://gerrit.wikimedia.org/r/183484 (owner: 10Esanders) [19:47:35] (03Merged) 10jenkins-bot: MediaWiki Theme: Fix overlap between hover and active states [oojs/ui] - 10https://gerrit.wikimedia.org/r/183482 (owner: 10Prtksxna) [19:47:37] (03Merged) 10jenkins-bot: MediaWiki Theme: Remove text-shadow on button [oojs/ui] - 10https://gerrit.wikimedia.org/r/183468 (owner: 10Prtksxna) [19:47:39] (03Merged) 10jenkins-bot: MediaWiki Theme: Fix disabled state of buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183481 (owner: 10Prtksxna) [19:47:41] Whee. [19:48:12] (03Merged) 10jenkins-bot: MediaWiki Theme: Fix focus state for buttons [oojs/ui] - 10https://gerrit.wikimedia.org/r/183479 (owner: 10Prtksxna) [19:48:14] (03CR) 10jenkins-bot: [V: 04-1] MediaWiki Theme: Add state change transition to checkbox [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:48:23] so, I'm not so sure about this: https://gerrit.wikimedia.org/r/#/c/183474/ [19:48:30] Yeah. [19:48:39] we just finally fixed the hight of the buttons, and now they want them to be 0.1em larger - for what reason? [19:48:41] Isn't that what we had to remove to get the theme working in December? [19:48:44] "according to design spec" [19:48:52] The spec linked doesn't have numbers. [19:48:55] Just a PNG. [19:48:55] So… [19:49:00] yeah [19:49:03] it was just a drawing [19:49:10] (03PS2) 10Jforrester: MediaWiki Theme: Add state change transition to checkbox [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:49:13] we matched the proportions visually [19:49:32] (03CR) 10Jforrester: [C: 032] "…" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:49:52] https://phabricator.wikimedia.org/T86225 [19:49:55] (03CR) 10Trevor Parscal: [C: 04-1] "Also, the design spec linked to makes not case for specific sizes. We already are within 0.1em of the supposed target, so I don't see why " [oojs/ui] - 10https://gerrit.wikimedia.org/r/183474 (owner: 10Prtksxna) [19:50:47] (03CR) 10jenkins-bot: [V: 04-1] MediaWiki Theme: Add state change transition to checkbox [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:51:16] * James_F sighs. [19:51:23] What is schema.Edit? [19:52:04] 3VisualEditor: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964058 (10Krenair) I just tried it myself (logged out) and it loaded up fine. Strange error in console though. [19:52:48] Krenair: The core EventLogging schema for VE and soon every editor. Why? [19:53:30] do we depend on that? [19:54:03] performance.wmflabs.org (see task I just commented on) shows Uncaught Error: Unknown dependency: schema.Edit [19:54:16] Krenair: Yes, through ve.track. [19:54:24] Krenair: Well, not "depend" depend. [19:54:29] It doesn't show up on my wiki [19:54:46] But if you've got EventLogging and WikimediaEvents installed you'll have issues if you don't have that schema. [19:55:13] Okay, so it's a setup issue [19:55:34] Yeah. [19:55:46] Essentially, if they uninstall EventLogging it'll just be quiet. [19:56:10] I wonder if that is what's causing the bug the author describes [19:56:21] although I can't reproduce it [19:57:26] (03Merged) 10jenkins-bot: MediaWiki Theme: Add state change transition to checkbox [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:59:37] (03CR) 10Krinkle: "recheck" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183170 (https://phabricator.wikimedia.org/T78372) (owner: 10Prtksxna) [19:59:41] James_F: I think we should revert to having MediaWiki-Vagrant load schema from metawiki. It was formerly the default, but Dan D. talked me out of it because it violates hyiegene for tests (you shouldn't rely on external resources). That is true, but I think user convenience and the principle of least astonishment trumps that consideration, in hindsight. [19:59:56] That was quick [20:00:16] James_F: I fixed it 1 second before the job came out of the queue. [20:00:29] ori: Yeah, that sounds reasonable. [20:00:32] Krinkle: :-) [20:00:39] it was already queued, hehe [20:00:55] the failure was for test, not gate pipeline. Though gate would've failed too [20:00:57] {{fixed}} [20:01:24] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964075 (10Jdforrester-WMF) That's a message from the #Analytics-EventLogging extension, because `ve.track` writes events to that. They don't have any im... [20:01:59] https://github.com/jscs-dev/node-jscs/issues/787 [20:10:06] (03PS1) 10Bartosz Dziewoński: Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 [20:11:41] (03CR) 10Jforrester: "Does this need to be provided in constructive/etc. flags?" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [20:12:07] (03CR) 10Bartosz Dziewoński: "I don't know, but I doubt it." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [20:12:23] MatmaRex: Isn't it normally gold? [20:13:06] James_F: that particular icon isn't. [20:13:12] it's only used in the login form [20:13:13] MatmaRex: Eh. OK. [20:13:16] in some obscure configurations [20:13:27] Anything else you need to get into OOjs UI this week? [20:13:48] https://gerrit.wikimedia.org/r/#/c/175568/ maybe? [20:14:58] https://phabricator.wikimedia.org/T78083 or https://phabricator.wikimedia.org/T76539 could do with work if you're idle. ;-) [20:15:21] 3VisualEditor, OOjs-UI: MediaWiki theme: Toolbar border too strong - https://phabricator.wikimedia.org/T85974#964102 (10Jdforrester-WMF) a:3KHammerstein [20:15:27] 3VisualEditor, OOjs-UI: MediaWiki theme: Toolbar popup widgets have no visible anchor - https://phabricator.wikimedia.org/T85998#964103 (10Jdforrester-WMF) a:3KHammerstein [20:15:33] 3VisualEditor, OOjs-UI: MediaWiki theme: Toolbar popup widgets have no visible anchor - https://phabricator.wikimedia.org/T85998#959169 (10Jdforrester-WMF) p:5High>3Low [20:15:37] 3VisualEditor, OOjs-UI: MediaWiki theme: Toolbar border too strong - https://phabricator.wikimedia.org/T85974#958556 (10Jdforrester-WMF) p:5Normal>3Low [20:15:51] 3VisualEditor, OOjs-UI: MediaWiki theme: Border radius used inconsistently - https://phabricator.wikimedia.org/T85973#964110 (10Jdforrester-WMF) a:5TrevorParscal>3None [20:16:13] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964113 (10Negative24) Can you post the messages here? I am getting some messages about unknown dependencies but I don't think it has anything to do with... [20:16:39] James_F: let me know what needs doing and i'll look at it. i was planning to add some missing secret sauce that'd make implementing the MW login form with OOUI widgets actually possible. [20:16:40] 3VisualEditor, OOjs-UI, VisualEditor-MediaWiki: Clash between rounded corners and text content is jarring for the category editor in VisualEditor when using the OOjs UI MediaWiki theme - https://phabricator.wikimedia.org/T78039#964115 (10Jdforrester-WMF) a:3Esanders [20:17:07] MatmaRex: That sounds most useful. [20:17:12] MatmaRex: Don't mind me. [20:18:06] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964124 (10Krenair) > Uncaught Error: Unknown dependency: schema.Edit was the error caused by EventLogging. No others, VE loads fine for me. [20:18:10] …alright. [20:21:19] hi [20:21:32] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964144 (10Negative24) I'm getting that as well. I was told that it is safe to ignore those messages. Do you think this could be a problem with Parsoid? [20:22:21] can I add html characters via VE search and replace? [20:23:38] 3VisualEditor, VisualEditor-MediaWiki-Templates: [Regression wmf14] Cannot add any template, throws error TypeError: Cannot read property 'length' of null - https://phabricator.wikimedia.org/T86056#964149 (10Ryasmeen) Verified the fix in test2 [20:24:29] i see it doesnt work, so I think I will create a task guys [20:24:44] 3VisualEditor-MediaWiki-Templates, VisualEditor: [Regression wmf14] Cannot add any template, throws error TypeError: Cannot read property 'length' of null - https://phabricator.wikimedia.org/T86056#964150 (10Ryasmeen) [20:25:09] 3VisualEditor, VisualEditor-MediaWiki, VisualEditor-MediaWiki-References: [Regression pre-wmf14] While trying to open Re-use Reference dialog immediately after inserting a citation (not basic reference), throws the TypeError: Cannot read property 'length' of null and the newly added citation does not appear in the... [20:25:50] 3VisualEditor, VisualEditor-MediaWiki: [Regression wmf14] Response from parseFragment is blank content, meaning newly-inserted or -modified content is blank until saving - https://phabricator.wikimedia.org/T86046#964154 (10Ryasmeen) [20:27:10] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964160 (10Jdforrester-WMF) >>! In T86112#964144, @Negative24 wrote: > I'm getting that as well. I was told that it is safe to ignore those messages. Do... [20:27:11] Juandev: What do you mean? [20:27:29] Juandev: Search and replace will replace any Unicode characters with other Unicode characters. [20:27:53] Juandev: Typing """ into VisualEditor doesn't work, exactly like "[[foo]]" doesn't work. [20:28:04] James_F: well I need to insert   [20:28:16] Juandev: OK. What OS are you on? [20:28:25] but maybe that might be set automatically by MW [20:28:29] Vista [20:28:54] Hmm. That makes it tricky. [20:29:08] There's a task to add NBSP to the special character inserter. [20:29:12] But it's not done yet. [20:29:57] James_F: but it can be done automatically by mw no, there are somehow clear rules when to use it and when not [20:30:28] Juandev: It can't be done automatically without breaking a few million pages, I'd guess. :-) [20:30:47] Juandev: And the "clear rules" only apply (at best) to one language at a time, probably. [20:31:25] well if you set different rules in different projects [20:31:42] yes, thats what I am talking about, set it for different languages [20:32:24] you can paste a unicode nbsp between two words [20:32:46] just copy from here: https://en.wikipedia.org/wiki/Non-breaking_space [20:32:55] edsanders: But Juandev's on Vista. [20:33:24] my commiserations, how does that affect the issue [20:33:25] edsanders: Oh, yeah, copy from there works. [20:33:39] edsanders: Vista doesn't have a good native special character inserter. [20:34:53] 3VisualEditor, Analytics-EventLogging: VisualEditor not loading unless edit page is refreshed with the cache cleared. - https://phabricator.wikimedia.org/T86112#964172 (10Negative24) I'll try with a different browser when I get home. Jdforrester, you tested it on the Performance wiki? [20:35:28] umm? let me test [20:36:20] I should also state that some browsers scripts colapses to me, when using that feature [20:36:48] the space probably won't work at the end of a paragraph [20:36:52] What browser scripts? What do you mean by "colapses"? [20:36:59] but if you place it between words it should be fine [20:38:01] well I should create a test page at first [20:38:21] Base64.js [20:38:36] That's… odd. [20:42:27] edsanders: it doesnt work to me, what exactly I should place there? [20:42:34] can you quote it? [20:43:19] huh? I was just suggesting pasting the nbsp character from that page [20:45:05] this one: >> << [20:46:04] 3VisualEditor: When the cursor is not active in the document do not activate the toolbar, as nothing happens on tapping on those options which is very confusing - https://phabricator.wikimedia.org/T86230#964179 (10Ryasmeen) 3NEW [20:46:09] edsanders: yes and it is "nbsp" or " " or U+00A0 or " "? [20:46:21] edsanders: That becomes a space when I copy it. [20:46:22] 3VisualEditor, VisualEditor-MediaWiki-Mobile: When the cursor is not active in the document do not activate the toolbar, as nothing happens on tapping on those options which is very confusing - https://phabricator.wikimedia.org/T86230#964179 (10Ryasmeen) [20:46:38] trying copying one from the javascript console [20:47:02] press F12, go to the console and type '\u00a0' [20:47:07] then copy the output [20:47:07] edsanders: it starts to be complicated [20:47:38] yes - typing a nbsp into a text editor is not a common use case [20:48:18] doesnt work to me F12 [20:48:23] (03PS35) 10Mooeypoo: [WIP] Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [20:48:36] ok, never mind [20:49:51] nuh dont work [21:10:06] 3VisualEditor-MediaWiki-Mobile, VisualEditor: In iOS Safari, all dialogs in mobile VE opening halfway or as empty on the first attempt of every edit session - https://phabricator.wikimedia.org/T86233#964244 (10Ryasmeen) 3NEW [21:13:14] (03PS18) 10Paladox: Re add svg images [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/181786 (https://phabricator.wikimedia.org/T37342) [21:13:35] (03CR) 10MarkTraceur: "Stop. Rebasing." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/181786 (https://phabricator.wikimedia.org/T37342) (owner: 10Paladox) [21:14:09] (03CR) 10Paladox: "Try the image in chrome." [extensions/WikiEditor] - 10https://gerrit.wikimedia.org/r/181786 (https://phabricator.wikimedia.org/T37342) (owner: 10Paladox) [21:17:46] 3MediaWiki-Page-editing: Cite button of RefToolbar not working - https://phabricator.wikimedia.org/T86039#964256 (10MKar) The problem is solved. Thanks to Aklapper (andre__) for his valuable help through IRC. [21:27:43] the cite options are not aligned properly on mobile [21:29:51] 3VisualEditor-MediaWiki-Mobile, VisualEditor: Citation dialog not opening in mobile VE for en.wiki - https://phabricator.wikimedia.org/T86077#964290 (10Krenair) I can reproduce this on en.m.wikipedia.beta.wmflabs.org [21:31:12] (03PS36) 10Mooeypoo: [WIP] Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [21:36:53] I refreshed and it showed differently :/ [21:39:25] 3VisualEditor-MediaWiki-Mobile, VisualEditor: Citation dialog not opening in mobile VE for en.wiki - https://phabricator.wikimedia.org/T86077#964320 (10Krenair) And now I can't reproduce it anywhere. [21:40:51] wtf [21:41:13] now when I type in the input boxes, random parts of the page start going blank [21:44:18] (03PS2) 10Trevor Parscal: Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [21:55:54] hmm. [21:56:10] will bad things happen if my OOjs mixins have their own mixins? [21:56:40] probably not [22:27:03] Krenair: I think , you are talking about input boxes for citation dialog going blank as you start typing right? [22:27:55] thats also reported as a separate issue: https://phabricator.wikimedia.org/T84914 [22:28:17] yeah something is really broken there [22:28:37] and the citation dialog not appearing , this happens only on the first attempt of every edit session [22:29:18] yeah! [22:32:38] ryasmeen, were you able to reproduce this on ff? [22:32:42] or only chrome? [22:33:30] oh I usually check mobile VE with safari on iOS device [22:34:56] Krenair:yeah i dont see this issue, with Chrome or Firefox [22:35:34] ok, I see it on chrome but not ff [22:42:12] Krenair: There is another one reported, where the dialogs are opening half way.Not sure if they are same issue or not :https://phabricator.wikimedia.org/T86233 [23:07:08] (03PS1) 10Bartosz Dziewoński: Synchronize @abstract class annotations between PHP and JS [oojs/ui] - 10https://gerrit.wikimedia.org/r/183716 [23:07:10] (03PS1) 10Bartosz Dziewoński: Don't test abstract classes [oojs/ui] - 10https://gerrit.wikimedia.org/r/183717 [23:07:12] (03PS1) 10Bartosz Dziewoński: Introduce and use TabIndexedElement [oojs/ui] - 10https://gerrit.wikimedia.org/r/183718 [23:10:26] (03PS3) 10Esanders: Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:10:31] (03CR) 10Jforrester: [C: 032] Synchronize @abstract class annotations between PHP and JS [oojs/ui] - 10https://gerrit.wikimedia.org/r/183716 (owner: 10Bartosz Dziewoński) [23:11:08] (03CR) 10Jforrester: [C: 031] Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:11:34] (03CR) 10Esanders: [C: 031] "Have centred, pixel snapped and scaled to match their respective themes." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:11:46] (03PS37) 10Mooeypoo: Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [23:11:49] that's three software engineers working on one ten-by-ten-pixel icon? we really should get a badge for being meticulous or something [23:11:58] (03CR) 10Jforrester: [C: 032] Don't test abstract classes [oojs/ui] - 10https://gerrit.wikimedia.org/r/183717 (owner: 10Bartosz Dziewoński) [23:12:15] (03CR) 10Jforrester: [C: 031] "OK from a Product perspective. Let's land this?" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) (owner: 10Mooeypoo) [23:12:31] :D [23:12:31] MatmaRex: I believe the badge is called "Editing Team". :-) [23:13:30] (03CR) 10jenkins-bot: [V: 04-1] Don't test abstract classes [oojs/ui] - 10https://gerrit.wikimedia.org/r/183717 (owner: 10Bartosz Dziewoński) [23:13:37] Eurgh. [23:14:34] (03CR) 10Jforrester: [C: 032] Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:14:55] hmph, that probably shouldn't go under my name [23:15:02] (03CR) 10Bartosz Dziewoński: [C: 04-2] "One sec" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:15:06] MatmaRex: Ha. [23:15:24] since the icon looks completely unlike what i had [23:15:24] MatmaRex: I'm /trying/ to get a release of OOjs UI for you like you asked. :-) [23:16:26] (03CR) 10jenkins-bot: [V: 04-1] Introduce and use TabIndexedElement [oojs/ui] - 10https://gerrit.wikimedia.org/r/183718 (owner: 10Bartosz Dziewoński) [23:16:39] (03PS4) 10Bartosz Dziewoński: Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 [23:16:56] (03CR) 10Bartosz Dziewoński: "Don't mind me!" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:17:15] (03Merged) 10jenkins-bot: Synchronize @abstract class annotations between PHP and JS [oojs/ui] - 10https://gerrit.wikimedia.org/r/183716 (owner: 10Bartosz Dziewoński) [23:17:58] James_F: heh [23:18:54] James_F: perhaps tomorrow, i still need to implement a bunch of stuff? [23:19:19] (we probably need some basic version of SelectWidgets in PHP, i'm afraid.) [23:19:28] MatmaRex: not to mention reply on :P [23:20:13] ori: done! [23:22:01] thanks [23:25:36] (03CR) 10Bartosz Dziewoński: [C: 032] "*bump*" [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:29:32] (03Merged) 10jenkins-bot: Add 'lock' icon [oojs/ui] - 10https://gerrit.wikimedia.org/r/183619 (owner: 10Bartosz Dziewoński) [23:32:05] MatmaRex: Sure, we can wait. [23:32:15] (03CR) 10Prtksxna: "Trevor, I was waiting for May to approve this. Should've -1'd it I guess." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183468 (owner: 10Prtksxna) [23:35:23] (03PS2) 10Jforrester: Don't test abstract classes [oojs/ui] - 10https://gerrit.wikimedia.org/r/183717 (owner: 10Bartosz Dziewoński) [23:35:29] (03PS2) 10Jforrester: Introduce and use TabIndexedElement [oojs/ui] - 10https://gerrit.wikimedia.org/r/183718 (owner: 10Bartosz Dziewoński) [23:36:07] (03CR) 10Bartosz Dziewoński: "I'd appreciate someone's opinion on the ButtonInputWidget curiosity." [oojs/ui] - 10https://gerrit.wikimedia.org/r/183718 (owner: 10Bartosz Dziewoński) [23:37:08] James_F: hmm, there are things in http://etherpad.wikimedia.org/p/OOjsUI-TODO . (and they've been there for a while.) [23:39:51] MatmaRex: Did we not move all of them to Phabricator? [23:39:57] no idea [23:40:11] i had the tab open for a few weeks and looked at it now [23:40:30] (03CR) 10jenkins-bot: [V: 04-1] Don't test abstract classes [oojs/ui] - 10https://gerrit.wikimedia.org/r/183717 (owner: 10Bartosz Dziewoński) [23:40:51] Oops. [23:41:18] huuuuh. [23:41:19] (03CR) 10jenkins-bot: [V: 04-1] Introduce and use TabIndexedElement [oojs/ui] - 10https://gerrit.wikimedia.org/r/183718 (owner: 10Bartosz Dziewoński) [23:41:25] (03PS38) 10Mooeypoo: Transform the search widget to show image details [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/161342 (https://phabricator.wikimedia.org/T78161) [23:41:30] since when do we have rubocop voting? [23:41:57] obviously that's going to cause the tests to fail [23:41:59] "bin/docparser.rb:18:1: C: Cyclomatic complexity for parse_file is too high. [61/60]" [23:42:04] "bin/docparser.rb:18:1: C: Method has too many lines. [164/162]" [23:42:13] i don't like smartass linters. [23:43:02] MatmaRex: Weeks ago [23:43:14] (03PS1) 10Mooeypoo: Add filename to the media dialog [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183729 (https://phabricator.wikimedia.org/T54459) [23:43:44] James_F: ugh [23:44:07] this is so stupid. we have a config file that hard-codes the values for these scripts, so that the tests don't fail. [23:45:29] 3VisualEditor-MediaWiki-Mobile, VisualEditor: The Edit Source mode for mobile VE does not load the contents of all sections in en.wiki - https://phabricator.wikimedia.org/T86265#964755 (10Ryasmeen) 3NEW [23:45:39] MatmaRex: :-( [23:46:12] 3VisualEditor-MediaWiki-Mobile, VisualEditor: The Edit Source mode for mobile VE does not load the contents of all sections in en.wiki - https://phabricator.wikimedia.org/T86265#964770 (10Ryasmeen) {F27048} [23:46:59] (03PS1) 10Jforrester: build: Bump copyright notice to 2015 [oojs/core] - 10https://gerrit.wikimedia.org/r/183731 [23:47:04] (03PS1) 10Jforrester: build: Bump copyright notices to 2015 [oojs/ui] - 10https://gerrit.wikimedia.org/r/183732 [23:47:55] (03PS1) 10Jforrester: build: Bump copyright notices to 2015 [unicodejs] - 10https://gerrit.wikimedia.org/r/183733 [23:50:02] (03PS1) 10Bartosz Dziewoński: TextInputWidget: Add support for 'autofocus' config option [oojs/ui] - 10https://gerrit.wikimedia.org/r/183735 [23:52:19] (03PS1) 10Jforrester: build: Bump copyright notices to 2015 [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183736 [23:52:36] (03CR) 10jenkins-bot: [V: 04-1] TextInputWidget: Add support for 'autofocus' config option [oojs/ui] - 10https://gerrit.wikimedia.org/r/183735 (owner: 10Bartosz Dziewoński) [23:54:16] (03PS1) 10Jforrester: build: Bump copyright notices to 2015 [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183740 [23:54:39] (03CR) 10jenkins-bot: [V: 04-1] build: Bump copyright notices to 2015 [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/183740 (owner: 10Jforrester) [23:55:25] (03CR) 10jenkins-bot: [V: 04-1] build: Bump copyright notices to 2015 [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/183736 (owner: 10Jforrester) [23:58:53] … really?