[00:08:14] 10VisualEditor: Extension of the Visual Editor namespaces on the French Wikiversity - https://phabricator.wikimedia.org/T126561#2017326 (10JackPotte) 3NEW [00:18:22] 10VisualEditor: Visualeditor should follow my wpuseeditwarning settings in the "Review your changes" window - https://phabricator.wikimedia.org/T126562#2017343 (10Quiddity) 3NEW [00:18:36] 10VisualEditor: Visualeditor should follow my wpforceeditsummary settings in the "Review your changes" window - https://phabricator.wikimedia.org/T126562#2017350 (10Quiddity) [00:19:10] huh, feel like I already did ^ [00:38:58] 10VisualEditor, 10VisualEditor-MediaWiki, 6Editing-Analysis: Revise the Edit schema and its use - https://phabricator.wikimedia.org/T118063#2017425 (10Neil_P._Quinn_WMF) [00:45:12] 10VisualEditor, 10Phlogiston, 15User-JAufrecht: Experiment with a project burnup that shows open and closed separately - https://phabricator.wikimedia.org/T126565#2017446 (10JAufrecht) 3NEW a:3JAufrecht [00:45:59] 10VisualEditor, 10Phlogiston, 15User-JAufrecht: Experiment with a project burnup that shows open and closed separately - https://phabricator.wikimedia.org/T126565#2017446 (10JAufrecht) [00:46:01] 10VisualEditor, 10Phlogiston, 15User-JAufrecht, 7WorkType-NewFunctionality: Generate Burnup chart with maintenance shown below the axis - https://phabricator.wikimedia.org/T113270#2017460 (10JAufrecht) [00:49:08] 10VisualEditor, 10Phlogiston, 7Epic, 15User-JAufrecht, 7WorkType-NewFunctionality: Display all active projects in backlog and burnup form combined - https://phabricator.wikimedia.org/T113270#2017472 (10JAufrecht) [00:49:53] (03PS7) 10Jforrester: FocusableNode: watch for image loaded/layout changes for invisible icons [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/268437 (https://phabricator.wikimedia.org/T125767) (owner: 10DLynch) [00:50:00] (03CR) 10Jforrester: [C: 032] FocusableNode: watch for image loaded/layout changes for invisible icons [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/268437 (https://phabricator.wikimedia.org/T125767) (owner: 10DLynch) [00:51:16] (03CR) 10Jforrester: [C: 031] Use plain DOM in ve.ce.LinkAnnotation [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269546 (https://phabricator.wikimedia.org/T126348) (owner: 10Esanders) [00:52:41] 10VisualEditor: Visualeditor should follow my wpforceeditsummary settings in the "Review your changes" window - https://phabricator.wikimedia.org/T126562#2017489 (10Jdforrester-WMF) This seems to work as expected for me. Can you explain what isn't happening? [01:04:34] (03Merged) 10jenkins-bot: FocusableNode: watch for image loaded/layout changes for invisible icons [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/268437 (https://phabricator.wikimedia.org/T125767) (owner: 10DLynch) [01:11:13] RoanKattouw_away: https://gerrit.wikimedia.org/r/#/c/269546/ [01:12:14] RoanKattouw_away: https://gerrit.wikimedia.org/r/#/c/269449/ [01:12:54] (03CR) 10Catrope: [C: 032] Use plain DOM in ve.ce.LinkAnnotation [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269546 (https://phabricator.wikimedia.org/T126348) (owner: 10Esanders) [01:17:25] (03Merged) 10jenkins-bot: Use plain DOM in ve.ce.LinkAnnotation [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269546 (https://phabricator.wikimedia.org/T126348) (owner: 10Esanders) [01:20:24] (03PS2) 10Jforrester: Update VE core submodule to master (5ea4fdf) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269714 [01:21:37] (03CR) 10Catrope: [C: 04-1] [WIP] Subdocuments: Current uses of Document are contained in a DocumentSet (0310 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269449 (owner: 10DLynch) [01:21:54] 10VisualEditor, 10VisualEditor-DataModel, 5WMF-deploy-2016-02-16_(1.27.0-wmf.14): Can't unbold multiple table cells - https://phabricator.wikimedia.org/T125978#2017550 (10Jdforrester-WMF) 5Open>3Resolved [01:22:58] RoanKattouw_away: https://gerrit.wikimedia.org/r/269714 [01:23:49] (03CR) 10Catrope: [C: 032] Update VE core submodule to master (5ea4fdf) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269714 (owner: 10Jforrester) [01:54:01] 10VisualEditor: Visualeditor should follow my wpforceeditsummary settings in the "Review your changes" window for section edits - https://phabricator.wikimedia.org/T126562#2017604 (10Quiddity) [01:56:11] 10VisualEditor: Visualeditor should follow my wpforceeditsummary settings in the "Review your changes" window for section edits - https://phabricator.wikimedia.org/T126562#2017624 (10Quiddity) >>! In T126562#2017489, @Jdforrester-WMF wrote: > This seems to work as expected for me. Can you explain what isn't happ... [02:12:38] (03PS1) 10Jforrester: ve.dm.ElementLinearData.static.compareElements: Add a couple of missed test cases [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269889 [02:12:40] (03PS1) 10Jforrester: ve.dm.ElementLinearData.static.getRelativeOffset: Add a missed exception-throwing test case [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269890 [02:13:09] (03PS2) 10Jforrester: ve.dm.ElementLinearData#getRelativeOffset: Add a missed exception-throwing test case [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269890 [02:13:37] (03CR) 10Jforrester: "This doesn't work. Roan and I couldn't work out why." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269890 (owner: 10Jforrester) [02:18:36] (03CR) 10jenkins-bot: [V: 04-1] ve.dm.ElementLinearData#getRelativeOffset: Add a missed exception-throwing test case [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269890 (owner: 10Jforrester) [02:22:20] (03PS2) 10Jforrester: Remove code for varying MWLinkInspectorTool title [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269875 (owner: 10Esanders) [02:39:07] (03PS2) 10Alex Monk: Try to make MW->VE switching failures break less things [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269589 (https://phabricator.wikimedia.org/T125580) [02:48:37] 10VisualEditor, 10VisualEditor-MediaWiki: In single edit tab mode, when you load VE directly by browsing to action=edit, and it fails to load, bad things happen - https://phabricator.wikimedia.org/T125998#2017701 (10Krenair) So, I think we can do this: When we detect failure to load on action=edit but there's... [04:09:07] 10VisualEditor, 10VisualEditor-MediaWiki-Mobile, 6Editing-Department, 10MobileFrontend, and 5 others: Mobile wikitext editing broken when user isn't in Schema:Edit bucket - https://phabricator.wikimedia.org/T126497#2017786 (10phuedx) >>! In T126497#2016762, @Neil_P._Quinn_WMF wrote: > Ay, this is my fault;... [11:12:26] 10VisualEditor, 10MediaWiki-Vagrant: RESTBase load failed: Service Unavailable - https://phabricator.wikimedia.org/T126595#2018206 (10zeljkofilipin) 3NEW [11:14:12] 10VisualEditor, 10MediaWiki-Vagrant: RESTBase load failed: Service Unavailable - https://phabricator.wikimedia.org/T126595#2018213 (10zeljkofilipin) [12:13:03] (03PS1) 10Zfilipin: [BrowserTest] Jenkins runs only passing Cucumber scenarios [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269952 (https://phabricator.wikimedia.org/T94162) [12:19:30] (03CR) 10Zfilipin: "At the moment it runs only two scenarios, but they are passing." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269952 (https://phabricator.wikimedia.org/T94162) (owner: 10Zfilipin) [12:41:29] edsanders, replied. I think we should move forward with the tag VE editor. We should remove the "mode" picker, and ignore the whole issue for maplinks and mapdata for now. [13:11:51] yurik, sure [13:12:36] edsanders, do you think you can poke at the geojson editor? it would be a killer feature from the start :) [13:13:14] i already have some of it done, but its not working fully yet :) [13:15:13] you have a patch? [13:15:24] I see it [13:18:56] edsanders, i think we should have a well defined featurecollection in VE, where all the geojson will be added, instead of adding it to the root. This way we can target the editor to that layer [13:19:46] (the root of the map control's features) [13:27:59] yurik, one needs to find a proper list of countries using imperial for map distances [13:28:12] the UK is officially metric but uses imperial for most distances [13:28:33] e.g. personal height is in feet and inches, and road signs/speeds are in miles [13:29:04] edsanders, i suspect we don't need it that much anyway - i copied it from the github.io code. I think its used for displaying the distances when editing [13:29:30] but yes, if needed, we might have to do the whole proper list of locales [13:38:15] edsanders, do you want to rename it to and remove the mode tag, or should I? [13:38:48] go for it [13:39:08] have you fixed the decimal point issue btw, because VE will output many decimal places [13:39:27] edsanders, oh, thx for reminding [13:46:27] edsanders, do you remember the bug with the rounding? I can't reproduce it [13:47:07] https://maps.wikimedia.org/img/osm-intl,15,37.8040,-122.4098,1000x1000.jpeg [13:47:11] https://phabricator.wikimedia.org/T125304 [13:50:31] ah, its not the extension, its the static image. Thx [14:03:21] edsanders, i use node_modules/.bin/browserify lib/mapbox-main.js -o lib/mapbox-lib.js to generate mapbox-lib [14:56:10] edsanders, ready? https://gerrit.wikimedia.org/r/#/c/269966 [15:15:09] edsanders, should we use this leaflet.fullscreen plugin, or should we implement something mediaviewer-like? The goal is to open the map full screen in some cases, or to offer a "make fullscreen" button on the existing map [15:15:10] https://github.com/Leaflet/Leaflet.fullscreen [15:57:48] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269875 (owner: 10Esanders) [16:04:04] (03CR) 10Jforrester: [C: 032] "Fine. :-(" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269875 (owner: 10Esanders) [16:08:36] (03PS1) 10Divec: Document annotation content container usage [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269988 (https://phabricator.wikimedia.org/T121351) [16:09:50] (03CR) 10Jforrester: [C: 032] Document annotation content container usage [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269988 (https://phabricator.wikimedia.org/T121351) (owner: 10Divec) [16:17:25] (03Merged) 10jenkins-bot: Remove code for varying MWLinkInspectorTool title [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269875 (owner: 10Esanders) [16:18:23] 10VisualEditor, 10MediaWiki-Vagrant: RESTBase load failed: Service Unavailable - https://phabricator.wikimedia.org/T126595#2018842 (10mobrovac) 5Open>3Invalid a:3mobrovac This was resolved yesterday in {T126421}. You need to `git pull && vagrant provision && vagrant git-update`. [16:28:23] (03Merged) 10jenkins-bot: Document annotation content container usage [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/269988 (https://phabricator.wikimedia.org/T121351) (owner: 10Divec) [17:03:12] 10VisualEditor, 10VisualEditor-CopyPaste: Mis-nested lists won't paste, making people sad - https://phabricator.wikimedia.org/T126625#2019060 (10Jdforrester-WMF) 3NEW [17:03:13] edsanders|away, i'm having some issues with the kartotherian static service. It might take some time to fix it [17:16:39] (03CR) 10DLynch: [C: 032] After KeyDown, check unicorns before polling the surfaceObserver [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/268305 (https://phabricator.wikimedia.org/T123430) (owner: 10Divec) [17:26:00] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/268854 (https://phabricator.wikimedia.org/T55613) (owner: 10Alex Monk) [17:36:31] (03Merged) 10jenkins-bot: After KeyDown, check unicorns before polling the surfaceObserver [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/268305 (https://phabricator.wikimedia.org/T123430) (owner: 10Divec) [17:38:48] edsanders|away: Thoughts on https://phabricator.wikimedia.org/T126625 ? [17:44:50] (03PS1) 10Jforrester: Update VE core submodule to master (55dd20e) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/270013 [17:49:02] (03CR) 10Trevor Parscal: [C: 032] Update VE core submodule to master (55dd20e) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/270013 (owner: 10Jforrester) [18:07:40] 10VisualEditor, 10VisualEditor-CopyPaste: Mis-nested lists won't paste, making people sad - https://phabricator.wikimedia.org/T126625#2019311 (10Esanders) [18:07:42] 10VisualEditor, 10VisualEditor-CopyPaste: Cannot paste nested list from Google Docs - https://phabricator.wikimedia.org/T98100#2019312 (10Esanders) [18:10:10] (03Merged) 10jenkins-bot: Update VE core submodule to master (55dd20e) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/270013 (owner: 10Jforrester) [18:18:31] needing review on https://gerrit.wikimedia.org/r/#/c/261995/, anyone? [18:21:59] ferdbold: I added edsanders to it for you. :-) [18:22:51] James_F, Google suck? [18:23:48] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 3 others: Provide an API flag to suppress auto-generated - https://phabricator.wikimedia.org/T113331#2019501 (10Esanders) Well given the initial approach was declined I was suggesting that we at least have a way... [18:23:57] edsanders: Also my point about some faulty contents shouldn't abort the entire paste. [18:24:11] What do you think? [18:25:17] yes - the converter should handle it [18:26:10] but I suspect it's going to be hard [18:26:11] which is why I've avoided it [18:27:03] that code should be under warranty, did we not make RoanKattouw_away sign a support contract when we gave to Collaboration? [18:27:21] * James_F grins. [18:27:42] He's such a slacker. Just look at him giving Kemayo thoughts last night. Tsk indeed. [18:28:09] For shame. [18:29:01] 10VisualEditor: Preformatted text and Enter key - https://phabricator.wikimedia.org/T126637#2019548 (10maiden_taiwan) 3NEW [18:30:37] 10Cite, 5Patch-For-Review, 3Reading-Web-Sprint-65-Game of Phones: Save references in page_props and cache - https://phabricator.wikimedia.org/T125329#2019563 (10Jdlrobson) @kaldari could you list some URLS for API requests you are worried about? We could set up a script that hits the API before and after the... [18:30:52] ah I thought I already added Ed, guess not [18:31:01] James_F ^ [18:31:24] * James_F nods. [18:33:29] edsanders, the static maps has been fixed and deployed [18:33:54] 10VisualEditor, 10VisualEditor-ContentEditable, 5WMF-deploy-2016-02-16_(1.27.0-wmf.14): Cannot move cursor left past nail at the end of styled text unless using the mouse to move the cursor first - https://phabricator.wikimedia.org/T123430#2019585 (10Jdforrester-WMF) 5Open>3Resolved [18:33:54] edsanders, i tried the editor on vem3 - looks great, but it erases existing layers :) [18:40:06] edsanders|away: What is your converter problem? [18:47:58] 10VisualEditor, 10VisualEditor-MediaWiki: Occasional "docserver-http: HTTP 404" when trying to save an edit made in the visual editor - https://phabricator.wikimedia.org/T123708#2019668 (10Elitre) I know these errors happen in the v.e., but I don't think it's its fault. [18:51:55] 10Cite, 5Patch-For-Review, 3Reading-Web-Sprint-65-Game of Phones: Save references in page_props and cache - https://phabricator.wikimedia.org/T125329#2019689 (10jcrespo) @kaldari Let me give you some light on the infrastructure side of things. I cannot for the most part say "this is ok" or "this is not ok" w... [18:52:08] (03CR) 10Jforrester: [C: 031] "Nice start. Some suggestions/issues:" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/169936 (https://phabricator.wikimedia.org/T124734) (owner: 10Alex Monk) [18:52:45] RoanKattouw,
  • Foo
    • Bar
[18:53:04] Ugh [18:53:05] Right, that one [18:53:08] https://phabricator.wikimedia.org/T98100 [18:54:48] Hmm so yeah, it's illegal to have a list inside of a list [18:54:55] I'm unsure how to best deal with this [18:55:33] Unicycle down to the GooglePlex and steal their hacky sacks? [18:56:14] Most probably, what we want is for "
  • Foo
    • Bar
" to produce FooBar right? [18:56:17] Because making the nested list a sibling of the paragraph containing Foo is also an issue I think [18:56:48] yes - it should be a nested list [18:57:12] (03CR) 10Jforrester: [C: 04-1] "* You can switch to an over-ride but not back" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/268854 (https://phabricator.wikimedia.org/T55613) (owner: 10Alex Monk) [18:58:14] (03CR) 10Jforrester: [C: 031] Try to make MW->VE switching failures break less things [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/269589 (https://phabricator.wikimedia.org/T125580) (owner: 10Alex Monk) [18:58:31] So IIRC this comment https://phabricator.wikimedia.org/T98100#1284584 is actually wrong [18:58:55] RoanKattouw: No, I don't think that's how the structure is meant. [18:59:00] James_F, RoanKattouw it would also be nice if we could raise the upstream issue with someone who pays attention: https://productforums.google.com/forum/#!category-topic/docs/documents/report-an-issue/desktop--other-please-specify/chrome-browser/RAwAhaEb0H8 [18:59:27] RoanKattouw: It certainly doesn't get created that way in GDocs/GMail, and doesn't render that way in Chrome/Firefox/etc. [18:59:44] Oh hah no you're right [19:00:03] RoanKattouw: The sub-list should be appended to the immediately parent LI (or have one inserted if it doesn't exist). [19:00:03] I just created a nested list in VE standalone, and I do get a list that's a sibling of a paragraph [19:00:03] I stand corrected [19:00:05] Yes, this is the VE output:

  • Foo

    • bar

[19:00:22] that's what makes it so evil [19:00:28] Yeah cause you can't just wrap it [19:00:54] OK, so the way I see it, we have two options, and they boil down to the same thing [19:01:15] We have to move the sublist into the preceding LI, either in a pre-conversion DOM transformation or during the conversion process itself [19:01:37] In either case, it's probably difficult to round-trip that structure, and we wish not to round-trip it [19:01:55] i.e. if you load a nested list with the wrong format into VE, it gets dirty-diffed into the right format [19:02:07] *and we probably wish [19:02:35] RoanKattouw: Dirty-diff for VE-core; VE-MW wouldn't receive such input, right? [19:02:52] Not from Parsoid, no [19:03:19] Only from paste, where conversion round-tripping is basically irrelevant [19:03:23] Yeah. [19:05:26] One could do $('ul>ul,ul>ol,ol>ul,ol>ol') on paste HTML [19:05:46] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 3 others: Provide an API flag to suppress auto-generated - https://phabricator.wikimedia.org/T113331#2019748 (10ssastry) >>! In T113331#2019501, @Esanders wrote: > Well given the initial approach was declined I w... [19:18:52] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019802 (10cscott) Can we bikeshed on a proper name? "stx: html" isn't really intuitive for an externally-vis... [19:21:12] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019808 (10Jdforrester-WMF) The values are `html` and `wt`? Is it an observation or a command to Parsoid? If t... [19:23:01] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019815 (10cscott) ``` ``` Of course one or the other of those (th... [19:23:32] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019824 (10ssastry) >>! In T100856#2019808, @Jdforrester-WMF wrote: > The values are `html` and `wt`? Is it an... [19:24:11] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 3 others: Provide an API flag to suppress auto-generated - https://phabricator.wikimedia.org/T113331#2019826 (10cscott) I'm suggesting `data-mw='{"auto":"true"}'` (or `data-mw='{"synthetic":"true"}'`?) as the not... [19:24:18] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019827 (10Jdforrester-WMF) Yeah, though the serialisation format might (?!) differ between tags, e.g. ` 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019836 (10ssastry) I am going to veto the 'serializeAs' attribute. Let us continue to use declarations. [19:30:01] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019844 (10cscott) @ssastry oh, agreed. I just spelled it as an attribute rather than embedded within `data-m... [19:32:09] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019851 (10Jdforrester-WMF) > This is also output from the wt2html pass. Yes, but in the context of "on round... [19:34:52] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019862 (10cscott) >>! In T100856#2019851, @Jdforrester-WMF wrote: >> This is also output from the wt2html pas... [19:50:31] (Sorry, got distracted) [19:50:35] [11:05] edsanders One could do $('ul>ul,ul>ol,ol>ul,ol>ol') on paste HTML [19:50:42] Yeah I was thinking something like that [19:51:12] annoying that css doesn't do (ol|ul)>(ol|ul) [19:51:13] Then .each(), and append to .previousSibling or if that's null prepend to parentNode [19:51:40] I'll do insert in place [19:51:47] for
  • Foo
      Bar [19:51:57] (Unfortunately that's one of the cases where the passing-null behavior in DOM methods doesn't magically do what you want) [19:52:06] Huh? [19:52:11] I don't think VE can deal with HTML like that [19:52:44] I mean [19:52:57]
      • Foo
          • ...
      • Bar
      would have to be mapped to
      • Foo
          ...
      • Bar
      right? [19:53:07] Not
      • Foo
          ...
        Bar
      [19:53:22]
    • Foo
      • Bar
        • [19:54:18] oh wait - that's still append to previous sibling [19:54:35] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 2 others: Migrate some semantic information from data-parsoid to data-mw - https://phabricator.wikimedia.org/T100856#2019898 (10ssastry) Note that this patch, whenever it is written and deployed, could result in a complete re-r... [19:54:50] Yes [19:55:15] https://imgur.com/a/zxbtn [19:59:51] of course this fix only works if the HTML is broken in a specific way [20:00:04] Yeah :/ [20:00:08] And it's specific to lists [20:00:24] I think it's probably easier to do this as an HTML transformation than trying to do it in the converter as it goes along [20:01:04] But it'd be good to 1) figure out if there are other cases we need to tackle, and 2) fix the fact that this crashes the converter [20:02:24] You could set up any two node types such that B is only allowed inside of A (and B is the only thing that can be inside of A), but anything is allowed to be in B, and then you'd presumably get the exact same issue [20:02:33] if you nest an A directly inside of an A [20:02:55] Here A=list and B=listItem but you could define different types and the same exception should happen [20:03:21] It should probably be dealt with more gracefully by alienating the nested A, rather than throwing an exception while trying to find a place where the A can go [20:04:14] Hmm maybe nested definition lists do the same thing? [20:04:51] RoanKattouw: Nested DLs are worse, if anything. [20:14:43] (03PS1) 10Esanders: Fix structure of broken nested lists from Google Docs [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/270041 (https://phabricator.wikimedia.org/T98100) [20:15:46] 10VisualEditor, 10VisualEditor-CopyPaste, 10VisualEditor-Tables: If the selection is multiple cells on paste and the paste value is a single cell or a non-cell selection (e.g. a string of text or a template), paste the value into each cell, not just the first... - https://phabricator.wikimedia.org/T126649#2019954 [20:17:14] (03CR) 10Catrope: [C: 04-1] Fix structure of broken nested lists from Google Docs (031 comment) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/270041 (https://phabricator.wikimedia.org/T98100) (owner: 10Esanders) [20:17:41] I'm also thinking about how multiple nested occurrences of ul>ul would be affected, but I think that works fine [20:17:54] I don't think fixing up one will break the way the loop interacts with the other [20:18:01] yeah - it should get them in order [20:18:17] and the selector picks up the child list [20:19:26] Yup [20:19:44] And DOM stuff is live, so relocating the parent doesn't affect how we fix the child [20:22:03] so when there is no previous sibling [20:22:19] um [20:22:41] this.parentNode.insertBefore( null, this ); I think ? [20:22:54] Or this, null, I never remember the order of params [20:23:05] No, sorry, that's wrong, that appends [20:23:14] that's a different problem [20:23:29] this.parentNode.insertBefore( this, this.parentNode.firstChild ); [20:23:29] such a list is not even semantically well defined [20:23:45] it's missing an LI [20:23:45] Hmm right [20:23:55] Yeah, so I guess we have to decide what we want to do [20:24:01] we can insert a new LI I guess [20:24:05] in Chrome it will render with two dots [20:24:16] Google docs I guess does it's own list rendering [20:24:21] Wait yeah ignore everything I said about insertBefore, it's stupid and doesn't do anything [20:24:34] It probably just needs to be wrapped in an LI [20:25:29] So these three cases: [20:25:30]
        • Foo
        [20:25:31]
          • Foo
        [20:25:31]
          • Foo
        [20:25:36] Render as: [20:25:42] * Foo [20:25:47] o Foo [20:25:56] * o Foo [20:26:42] #2 is totally invalid but produces the rendering the user wanted [20:27:05] BOOOOOOOOOOO [20:30:09] Hmm [20:30:09] I'm personally OK with remapping #2 to #3 [20:30:26] How does it render in FF? [20:36:24] so I did the wrapping and tested it with an example from CE sandbox [20:36:40] and it triggers the code and and does the extra wrap [20:36:40] then I wrote a test case [20:36:55] but apparently if you pass it through execCommand the double list converts to a single list... [20:38:16] (rendering is the same in FF) [20:45:42] (03PS2) 10Esanders: Fix structure of broken nested lists from Google Docs [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/270041 (https://phabricator.wikimedia.org/T98100) [21:04:25] 10VisualEditor, 6Collaboration-Team-Backlog, 10Flow, 6Language-Engineering, and 3 others: Provide an API flag to suppress auto-generated - https://phabricator.wikimedia.org/T113331#2020264 (10Esanders) @ssastry exactly [22:18:11] 10VisualEditor, 10Graph, 10SyntaxHighlight: "Unknown dependency `ext.visualEditor.mwcore`" when running QUnit tests - https://phabricator.wikimedia.org/T126681#2020653 (10jhobs) 3NEW [22:21:20] 10VisualEditor, 10Graph, 10SyntaxHighlight: "Unknown dependency `ext.visualEditor.mwcore`" when running QUnit tests - https://phabricator.wikimedia.org/T126681#2020679 (10Jdforrester-WMF) Both those extensions have code that depends on VE. CI is configured to respect this dependency, but I guess you mean it'... [22:24:44] 10VisualEditor, 10VisualEditor-ContentEditable: Cursor-right and cursor-left keys do not work correctly after inserting italics or bold text - https://phabricator.wikimedia.org/T126683#2020729 (10AxelBoldt) 3NEW [22:48:25] (03CR) 10Catrope: [C: 032] Fix structure of broken nested lists from Google Docs [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/270041 (https://phabricator.wikimedia.org/T98100) (owner: 10Esanders) [22:54:29] (03Merged) 10jenkins-bot: Fix structure of broken nested lists from Google Docs [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/270041 (https://phabricator.wikimedia.org/T98100) (owner: 10Esanders) [23:11:42] edsanders: Did you see https://gerrit.wikimedia.org/r/#/c/269890/ ? Roan and I couldn't work out whether the code for which I was trying to write a test was indeed broken or not. [23:12:47] More specifically, we couldn't figure out how to write a test that causes that exception to be thrown [23:14:55] 10VisualEditor, 10Graph, 10SyntaxHighlight: "Unknown dependency `ext.visualEditor.mwcore`" when running QUnit tests - https://phabricator.wikimedia.org/T126681#2020967 (10jhobs) I logged the task because yesterday I was not getting this error and today I was after not enabling/disabling any new extensions, s... [23:36:05] 10Citoid, 10VisualEditor, 10VisualEditor-MediaWiki-References, 10Internet-Archive, 10The-Wikipedia-Library: On URL submission, look up the archived page in the Internet Archive's index and add to the return data - https://phabricator.wikimedia.org/T115224#2021108 (10kaldari) Are there any plans to work o...