[00:01:53] ok, we have an idea [00:03:56] we add a data attribute to link tags generated by ResourceLoader on the server, which contains an array of modules that the link implements [00:03:56] 3VisualEditor / 3Data Model: VisualEditor: [Regression] “Uncaught Error: offset -1 out of bounds” when adding bullet/numbered list with link inspector open - 10https://bugzilla.wikimedia.org/72558#c3 (10etonkovidova) Verified the fix in betalbs and test2. [00:04:26] in RL on the client, we look for link tags with such attributes, parse the arrays and mark those dependencies as satisfied [00:04:34] (03CR) 10MaxSem: [C: 032] Followup 31bafa1: forgot to rename 'data' to 'args' for citation commands [extensions/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171451 (https://bugzilla.wikimedia.org/73001) (owner: 10Jforrester) [00:04:49] now you can have a JavaScript module depend on a module that was added using addModuleStyles() [00:05:03] and the dependency will not be loaded again [00:05:12] MatmaRex: your thoughts? [00:05:48] TrevorParscal: my thought is that Krinkl.e is going to veto that, based on things he vetoed in the past [00:06:22] Yeah, I will talk about this change on Friday in the FSG meeting [00:06:22] MatmaRex: Decisions are made by those who turn up. :-) [00:06:23] apart from that it looks reasonable, but if i were doing this, i'd go further and implement the support for CSS-only JS-independent modules [00:06:33] addModuleStyles() is a terrible hack that should have never been needed [00:06:41] Agreed [00:06:47] I agree that would be ideal [00:07:21] My initial suggestion was to have addModuleStyles() (or its future replacement) make a note of which modules were loaded through it, and output something like [00:07:21] I'm just suggesting that addModuleStyles not causing a module to be satisfied on the client is very evil already [00:07:33] Then Trevor ran with that and came up with the data attribute thing [00:07:57] the issue of course being that if someone uses addModuleStyles with a module that contains more than just styles [00:08:02] we can be in a bit of trouble [00:08:12] Yeah that's what I was thinking about [00:08:22] When I started saying "hmm no maybe this is more than 10 lines of code" [00:08:44] I think such things are already horrible and should be/have been -2'd [00:09:47] but, this would change the consequence of being an idiot from sending the styles to the client and adding them to the DOM 2x (possibly causing CSS issues, but probably not) to the JavaScript mysteriously never loading [00:10:42] 3VisualEditor / 3MediaWiki integration: VisualEditor: [Regression] Don't let the user insert the same category multiple times - 10https://bugzilla.wikimedia.org/72906#c5 (10etonkovidova) The fix is verified in betalabs and test2. [00:11:11] (03Merged) 10jenkins-bot: Followup 31bafa1: forgot to rename 'data' to 'args' for citation commands [extensions/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171451 (https://bugzilla.wikimedia.org/73001) (owner: 10Jforrester) [00:15:14] 3VisualEditor / 3Editing Tools: VisualEditor: “Uncaught Error: Unable to parse title” when entering "Talk:" as a template - 10https://bugzilla.wikimedia.org/71998#c8 (10etonkovidova) the fix is a part of 71249. [00:16:29] (03CR) 10Catrope: [C: 032] Disable review button while save dialog is pending [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171474 (https://bugzilla.wikimedia.org/65310) (owner: 10Alex Monk) [00:17:19] Krinkle|detached: Ping re https://gerrit.wikimedia.org/r/170874 ? [00:18:12] 3VisualEditor / 3ContentEditable: VisualEditor: Be able to merge table cells (setting colspan/rowspan as appropriate) - 10https://bugzilla.wikimedia.org/52179#c1 (10etonkovidova) The functionality for merging(un-merging) table cells is in test2 and betalabs. [00:20:56] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression pre-wmf7] Cite command not working (but cite-tranclusion does?) - 10https://bugzilla.wikimedia.org/73001#c5 (10James Forrester) 5PATC>3RESO/FIX Fix now live in production. [00:21:05] (cc etonkovidova.) [00:21:11] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf6] A Label appears momentarily, showing all "Use this Group" options, when Re-Using a Reference - 10https://bugzilla.wikimedia.org/72904#c7 (10etonkovidova) Verified in test2. [00:35:19] (03Merged) 10jenkins-bot: Disable review button while save dialog is pending [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171474 (https://bugzilla.wikimedia.org/65310) (owner: 10Alex Monk) [00:39:12] 3VisualEditor / 3MediaWiki integration: VisualEditor: The "Review your changes" button should be disabled when save dialog is busy - 10https://bugzilla.wikimedia.org/65310 (10Alex Monk) 5PATC>3RESO/FIX [00:41:27] 3VisualEditor / 3MediaWiki integration: VisualEditor: The "Review your changes" button should be disabled when save dialog is busy - 10https://bugzilla.wikimedia.org/65310 (10James Forrester) a:3Alex Monk [00:45:04] sucheta: http://jsfiddle.net/au1pxxzb/ [00:45:12] 3VisualEditor / 3Data Model: VisualEditor: Clicking in a table that contains templated rows causes JS error - 10https://bugzilla.wikimedia.org/72717#c3 (10etonkovidova) Verified with the Evil_table given in the link - the js error is not displayed anymore. [00:48:42] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression pre-wmf7] Cannot add any Citations or Basic reference in a page , throwing Typeerror " Cannot read property 'replace' of undefined " - 10https://bugzilla.wikimedia.org/72948#c7 (10etonkovidova) The fix is re-verified in betalabs and test2. [01:00:42] 3VisualEditor / 3ContentEditable: VisualEditor: I can't select multiple refs when going backwards from the keyboard in Firefox - 10https://bugzilla.wikimedia.org/72835#c4 (10etonkovidova) The fix is confirmed in test2 and betalabs. [01:24:57] (03PS1) 10SuchetaG: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) [01:42:16] (03CR) 10Catrope: [C: 04-1] "Looks fine, but breaks coloring of red links. But that should be done with LinkCache anyway." (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [01:45:52] (03CR) 10Catrope: "...or maybe don't move it, see inline" (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [02:02:12] (03CR) 10Catrope: "Link coloring needs to be fixed before merge." [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [02:06:23] (03PS13) 10Divec: Refactor SurfaceObserver pollOnceInternal [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/166188 [02:08:44] (03CR) 10jenkins-bot: [V: 04-1] Refactor SurfaceObserver pollOnceInternal [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/166188 (owner: 10Divec) [02:35:04] (03PS14) 10Divec: Refactor SurfaceObserver pollOnceInternal [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/166188 [02:35:20] (03PS1) 10Divec: WIP: Rerender ContentBranchNodes when the selection anchor leaves [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171494 [04:07:28] 3VisualEditor / 3ContentEditable: VisualEditor: Copying a template from one page to another: render is correct, but template is gone. - 10https://bugzilla.wikimedia.org/58419#c3 (10WhatamIdoing) 5NEW>3RESO/WOR I can no longer reproduce this in Firefox. It's been months since the last time I remember see... [04:44:43] (03PS1) 10Catrope: ve.utils.js updates [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171498 [04:44:45] (03PS1) 10Catrope: Add instrumentation for edit schema [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171499 [04:45:13] (03PS2) 10Catrope: ve.utils.js updates [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171498 [04:45:19] (03PS2) 10Catrope: Add instrumentation for edit schema [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171499 [04:46:24] (03CR) 10jenkins-bot: [V: 04-1] ve.utils.js updates [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171498 (owner: 10Catrope) [04:46:29] (03CR) 10jenkins-bot: [V: 04-1] Add instrumentation for edit schema [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171499 (owner: 10Catrope) [04:57:31] (03PS2) 10SuchetaG: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) [05:22:31] (03CR) 10Jforrester: [C: 04-1] ve.utils.js updates (032 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171498 (owner: 10Catrope) [08:29:55] 3VisualEditor / 3Language: VisualEditor: a line started with Japanese input disappear when I switch to and start non-Japanese (latin) input - 10https://bugzilla.wikimedia.org/72508#c2 (10miya) I first experienced the same bug around 2014-10-24. Environment:Windows 7 PC (Firefox 33.0.1, Chrome 38.0.2125.104 m... [09:33:30] (03CR) 10Zfilipin: "Chris, a couple of comments:" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171295 (owner: 10Cmcmahon) [11:33:07] (03PS8) 10Esanders: Support for generic file drop handlers [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/170907 [11:33:11] (03CR) 10Esanders: Support for generic file drop handlers (032 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/170907 (owner: 10Esanders) [11:54:56] (03CR) 10Krinkle: [C: 032] Tag v1.1.2 [oojs/core] - 10https://gerrit.wikimedia.org/r/171304 (owner: 10Jforrester) [11:59:40] (03Merged) 10jenkins-bot: Tag v1.1.2 [oojs/core] - 10https://gerrit.wikimedia.org/r/171304 (owner: 10Jforrester) [12:15:44] (03CR) 10Krinkle: [C: 04-1] Use hasOwnProperty instead of 'key in obj' (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171328 (owner: 10Esanders) [12:17:45] (03CR) 10Krinkle: [C: 031] Use hasOwnProperty instead of 'key in obj' [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171326 (owner: 10Esanders) [12:20:45] (03CR) 10Esanders: Use hasOwnProperty instead of 'key in obj' (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171328 (owner: 10Esanders) [12:23:27] (03PS1) 10Vikassy: Fixed failing browser tests [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171531 (https://bugzilla.wikimedia.org/73033) [12:26:07] (03PS2) 10Vikassy: [BrowserTest] Fixed failing language screenshot browser tests [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171531 (https://bugzilla.wikimedia.org/73033) [12:35:39] (03CR) 10Krinkle: [C: 032] Use hasOwnProperty instead of 'key in obj' (031 comment) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171328 (owner: 10Esanders) [12:35:51] (03CR) 10Krinkle: [C: 032] Use hasOwnProperty instead of 'key in obj' [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171326 (owner: 10Esanders) [12:37:06] (03Merged) 10jenkins-bot: Use hasOwnProperty instead of 'key in obj' [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171328 (owner: 10Esanders) [12:37:55] (03Merged) 10jenkins-bot: Use hasOwnProperty instead of 'key in obj' [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171326 (owner: 10Esanders) [16:07:12] TrevorP|Away: cool (re OOjs/UI PHP in core). can you point me at a gerrit # so I can keep tabs on progress? [19:33:35] [13oojs.github.io] 15jdforrester pushed 1 new commit to 06master: 02https://github.com/oojs/oojs.github.io/commit/a6e414d41381d05d90b72be759c6c508614d8180 [19:33:35] 13oojs.github.io/06master 14a6e414d 15James D. Forrester: Publish v1.1.1 and v1.1.2 releases from npm [19:33:35] [13oojs.github.io] 15jdforrester created 06names (+1 new commit): 02https://github.com/oojs/oojs.github.io/commit/8dd38bbbf419 [19:33:35] 13oojs.github.io/06names 148dd38bb 15James D. Forrester: Fix file names from typo [19:33:35] [13oojs.github.io] 15jdforrester pushed 1 new commit to 06master: 02https://github.com/oojs/oojs.github.io/commit/3742065afc84d0bf22834e7c30e394cb6c379897 [19:33:35] 13oojs.github.io/06master 143742065 15James Forrester: Merge pull request #5 from oojs/names... [19:33:35] Issue with oojs event emitters if y'all have a sec [19:33:35] I have an object created that is an emitter [19:33:35] And it emits something from the constructor [19:33:35] Which gets lost 'cause the object is still getting constructed [19:33:35] Is there a way to queue events so I don't have to split the method in two or anything? [19:33:35] I mean, I can do, but this seems like something that should be possible [19:33:36] marktraceur: how does it get 'lost'? got a code example? [19:33:36] MatmaRex: Well, not lost, just not listened to, because there's nothing to listen to it yet [19:33:36] Here, something like this: new Thing().connect( this, { 'event', [ 'emit', 'event' ] } ); [19:33:36] hmm [19:33:36] If I emit( 'event' ) in the Thing constructor, then whatever's connecting won't hear it [19:33:36] yeah, got it now [19:33:36] At least I think that's what's happening [19:33:36] But I don't see anything else wrong [19:33:42] you could probably subclass EventEmitter to do this easily enough, but it seems messy [19:33:42] you'd need to keep track of all events forever, in case something else connects later [19:33:43] sounds like you need a different "information-passing model"? promises or something [19:33:44] Maybe so [19:33:44] Or maybe I'll just split the method to not be stupid. That's fine too. [19:33:44] MW core has something that acts exactly like you want, mw.track [19:33:44] Ooh, really [19:33:44] which uses http://api.jquery.com/jQuery.Callbacks/ [19:33:44] it's in mediawiki.js [19:33:44] so, do yours the same way? or i might be misunderstanding what you really want :) [19:33:46] I'll figure it out...the word "analytics" in the docs make me think it's not what I need, but maybe [19:33:46] yeah, no [19:33:46] you probably don't want to *use* mw.track, just look how it works and reimplement something like it :D [19:33:46] Ahhh. [19:33:46] Maybe [19:34:00] TrevorParscal, let me know when you have a moment. My issue is CSS now. I'm adding a $placeholder (similar to what you do in your code) and code-wise it responds well (floats left/right when it should) but the CSS doesn't work. According to a few articles I've ready in StackOverflow, in order to get a div to float near another div, it should be float:right or float:left.. but my div is always to the right [19:34:00] ok, 3 min [19:34:00] I also tried prepending vs appending (without float) and that didn't work either. [19:34:00] TrevorParscal, just putting it up there for when you have time. [19:34:00] TrevorParscal, also, for reference, see the code here https://github.com/mooeypoo/ooui-reorder/blob/master/ReorderWidget.js#L86 [19:34:08] RoanKattouw: https://gerrit.wikimedia.org/r/#/q/If0634147a2a01e10c66acc3c8f335d3d15cf7180,n,z [19:34:11] FYI, the Gerrit and Bugzilla bots aren't here because all of labs is broken right now [19:34:11] "hardware failure" [19:34:11] ok, mooeypoo, sorry that was much more than 3min [19:34:11] mooeypoo: so, lets just talk CSS for a sec [19:34:11] what does this look like? Are you reordering a rows, columns, a grid? [19:34:12] the demo looks like columns [19:34:12] what about wrapping? [19:34:12] TrevorParscal, what do you mean ? [19:34:12] wrapping? [19:34:12] RoanKattouw: And also OOjs UI code can't merge… [19:34:12] RoanKattouw: Whoops. :-) [19:34:12] well, if you make the browser narrow [19:34:12] your floated items wrap [19:34:12] TrevorParscal, yeah, but each item is wrapped with a div that, ideally, shouldn't wrap. If the items themselves wrap, then the spacing will happen either before or after the item [19:34:12] shouldn't affect multiple rows like that. [19:34:13] I'm talking about when you have 20 items, you can fit only 5 per line, you end up with 4 lines of items - not talking about text wrapping within the items [19:34:13] my point is only that if you are doing floating, you are doing a grid, and if you are doing a grid, there are problems with grid drag and drop that you should acknowledge and solve for (as my code does) [19:34:15] if you are not intending it to be 2d (you only want horizontal or vertical arrangement and reordering) than you should avoid using floats, instead use text-flow (display:inline-block, white-space:nowrap;) [19:34:15] James_F: Ugh, right [19:34:15] RoanKattouw: Not the end of the world. [19:34:16] you will notice that in the grid list code I showed you, items never moved between rows during dragging (because it's really confusing to the user), only after dropping [19:34:16] Hmm [19:34:17] TrevorParscal, okay, so can I do it without floating? [19:34:17] the placeholder just pushes items to the sides within the row [19:34:17] TrevorParscal, I've tried doing it with margins [19:34:17] you have to actually lay it out in a grid, if it's 2d, or you will never have enough control [19:34:18] TrevorParscal, aye, I saw it, and I also saw that you're calculating "top" value of the drag, I thought for rows [19:34:18] but if you don't need 2d, if you only need 1d, then you can simplify it a lot [19:34:18] But -- yeah, exactly -- I thought it would just be simpler if I add spacing left/right to the items [19:34:18] And it seems to work, just has annoying movement-related bugs [19:34:18] You have 2 choices [19:34:19] either do a full on grid system and lay items out into row divs and all that (which is what gridList does) or don't make space to indicate the drop point [19:34:19] Hmm [19:34:19] you can instead make the placeholder position: absolute [19:34:19] make it a vertical line [19:34:19] that shows the insertion point you will drop into [19:34:19] ^^ that's what I was trying to do [19:34:19] like how drag and drop works for text [19:34:19] append or prepend it to items [19:34:20] no, don't modify the group's items by inserting DOM nodes [19:34:21] oh. hm. [19:34:21] have a drop point overlay on top of the items and position it to make it LOOK like it's between the items [19:34:21] I guess if you wanted to do an effect where the images to either side kind of move off to the side to make a little space, you could still do that with an absolutely-positioned placeholder, you could just change the padding/margin of the neighboring items [19:34:21] (but also that's the icing on the cake, not that important :) ) [19:34:21] RoanKattouw: http://demo.collabkit.org/ [19:34:21] so, I'll move its position [19:34:22] TrevorParscal: Let me guess, that's exactly what you're doing, plus a CSS transition? :D [19:34:22] I was going for a particular effect, I think it's a nice effect, but it imposes a LOT of restrictions [19:34:22] RoanKattouw: dude, gridList does all sorts of magic man [19:34:22] Oh yeah [19:34:22] TrevorParscal, to be honest, I got a bit lost with how you do that part in your code, mostly because of the $outside.add( $left ).add( $right ) [19:34:22] That one makes a LOT of space [19:34:22] I got confused with those. [19:34:22] I was thinking of a few pixels [19:34:22] Especially snce it seemed you then add padding to the items on top of that ? [19:34:23] I'm not saying you have to mimic the effect in gridList [19:34:23] I think it's very restrictive [19:34:23] if you want to make a general purpose drag and drop ordering mixin for GroupElement [19:34:23] * mooeypoo nods [19:34:23] you probably should use a cursor-like insertion point that overlays a line at the drop point [19:34:23] Yeah that's why I tried to simplify as much as possible. [19:34:23] and doesn't move anything around at all during drag [19:34:23] Hmmm [19:34:23] Yeah for a generic thing that sounds good [19:34:23] If you wanted to support effects you could add a class to the neighbors [19:34:23] TrevorParscal, I seeeeeee.... so not messing around with the shifting items to the sides -- just putting a line/cursor-like thing in between where you need it [19:34:23] that will work. [19:34:24] okay cool, I'll work on that. [19:34:24] and have the caller style those classes if it wants to [19:34:24] RoanKattouw, I did add classes to the neighbors, I get a dancing-around effect during drag [19:34:24] :( [19:34:24] :S [19:34:24] mooeypoo: Also yeah when he says "in between", what that really means is "absolutely positioned so it *looks like* it's in between" [19:34:25] He said that already but just making sure you got that [19:34:25] $outside = $outside.add($right).add($left) is just building a jquery selection to include things on the left and right of the drag point [19:34:25] it's just confusing to read [19:34:25] sorry [19:34:25] jquery add is always quite akward to use I think [19:34:25] $outside = $outside.add($right).add($left) is roughly equivalent to $outside = [ $outside, $right, $left ] [19:34:27] * mooeypoo nods [19:34:27] TrevorParscal, yeah, I just got confused about *what* those items are. The $left and $right are defined in the method as $( [] ) [19:34:27] so it seemed like you're adding an empty something before and empty something after and then selecting it? [19:34:27] anywyas, it's no longer relevant, I think.. I'll go with your suggestion [19:34:27] mooeypoo: there are some things I had to do with adding CSS classes to things to make both touch and drag events work, with hardware accelerated transformations (modulating left and top is SLOW, using transform2D is FAST) [19:34:27] (03PS1) 10Esanders: Refactor out show/hide loading spinner [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171564 [19:34:27] (03PS2) 10Esanders: Refactor out show/hide loading spinner [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171564 [19:34:28] I see [19:34:28] edsanders: What's an "out show/hide loading spinner"? [19:34:32] (03CR) 10jenkins-bot: [V: 04-1] Add 'indeterminate' state to progress bar widget [oojs/ui] - 10https://gerrit.wikimedia.org/r/171548 (owner: 10Esanders) [19:35:07] (03CR) 10Catrope: [C: 032] Update OOjs UI to v0.1.0-pre (12b66051ff) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171614 (owner: 10Jforrester) [19:35:13] (03PS4) 10Catrope: Add instrumentation for edit schema [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171499 [19:35:15] Whee [19:35:15] labs is back [19:35:17] (03CR) 10jenkins-bot: [V: 04-1] Add instrumentation for edit schema [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171499 (owner: 10Catrope) [19:35:30] *showing/hiding of [19:35:42] (03CR) 10Catrope: [C: 032] Update OOjs to v1.1.2 [oojs/ui] - 10https://gerrit.wikimedia.org/r/171609 (owner: 10Jforrester) [19:36:28] (03Merged) 10jenkins-bot: Update OOjs UI to v0.1.0-pre (12b66051ff) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171614 (owner: 10Jforrester) [19:36:38] * James_F nods. [19:37:53] (03Merged) 10jenkins-bot: Update OOjs to v1.1.2 [oojs/ui] - 10https://gerrit.wikimedia.org/r/171609 (owner: 10Jforrester) [19:37:57] mooeypoo: yeah, I think that the making space for the drop point is a special and neat effect, but it's going to cause a lot of complications, whereas if you make a generic system that can be thrown into vertical or horizontal mode, all you need to know is which edge of the item you are dragging over to render the insertion marker at [19:38:25] and then render the insertion marker using absolute positioning [19:38:56] * mooeypoo nods [19:39:02] TrevorParscal, it will only make complications if we're against the items dancing around all over the screen when you're trying to rearrange them..... [19:39:04] and when you drop just use addItems( [ draggedItem ], droppedItemIndex ) [19:39:04] :D [19:39:08] * mooeypoo nods [19:39:24] (03PS3) 10Jforrester: Add 'indeterminate' state to progress bar widget [oojs/ui] - 10https://gerrit.wikimedia.org/r/171548 (owner: 10Esanders) [19:39:28] the reorder dance, while entertaining, is only of value if you add music [19:39:35] Indeed [19:39:41] We should totally do that. On auto play. [19:39:44] So unless you add music, I will not +2 [19:40:00] * mooeypoo looks for the Flight of the Bumblebee [19:40:26] mooeypoo: Given the catastrophic UX issues, maybe Sarabande #2 instead? [19:40:27] ;-) [19:40:33] haha [19:41:30] James_F, that will work if the user drags slowly. When they start chasing objects in the screen, we can switch to the Bumblebee. [19:41:42] ... this could make a great game, actually. Zynga, beware! [19:41:54] * James_F puts his head in his hands. [19:43:10] You'll start getting notices on Facebook: Your friend Moriel managed rearrange MediaWiki Categories! Beat her record! [19:43:17] * mooeypoo will stop while she's ahead and go get something to eat [19:44:02] :-) [19:48:40] (03PS3) 10SuchetaG: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) [19:55:05] (03CR) 10Catrope: [C: 04-1] Matching pages from suggestions in link dialog now actually work as links (033 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [20:21:00] running the script manually... [20:21:29] Corruption alert: visualeditor-needcheck on cswiki: https://cs.wikipedia.org/?diff=11986980 [20:21:29] Corruption alert: visualeditor-needcheck on lawiki: https://la.wikipedia.org/?diff=2804002 [20:21:29] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=108874556 [20:21:29] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=108878185 [20:21:30] Corruption alert: visualeditor-needcheck on frwiki: https://fr.wikipedia.org/?diff=108897582 [20:21:30] Corruption alert: visualeditor-needcheck on ruwiki: https://ru.wikipedia.org/?diff=66634978 [20:21:31] Corruption alert: visualeditor-needcheck on itwiki: https://it.wikipedia.org/?diff=69025731 [20:21:31] Corruption alert: visualeditor-needcheck on itwiki: https://it.wikipedia.org/?diff=69025751 [20:21:31] Corruption alert: visualeditor-needcheck on itwiki: https://it.wikipedia.org/?diff=69025895 [20:21:32] Corruption alert: visualeditor-needcheck on itwiki: https://it.wikipedia.org/?diff=69025947 [20:25:38] (03PS4) 10SuchetaG: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) [20:25:56] That cswiki entry - I've seen a bunch fixing links broken in that weird way... [20:26:14] TrevorParscal: so, i'm currently working on putting more of my image response code from directly in load.php to ResourceLoader::respond(). does that make sense to you? [20:26:19] Krinkle|detached: ^ [20:26:28] although... https://cs.wikipedia.org/w/index.php?title=Renegade&diff=11986977&oldid=10790955 - it's causing them as well? [20:32:25] 3VisualEditor / 3Language: VisualEditor: a line started with Japanese input disappear when I switch to and start non-Japanese (latin) input - 10https://bugzilla.wikimedia.org/72508#c3 (10etonkovidova) the similar issue is bug 72619. [20:35:42] MatmaRex: If anyone has an objection to it, I'm sure we can move it to a class that load.php instantiates [20:36:28] TrevorParscal: i just wonder how much do we want to separate the image returning stuff from the "real" ResourceLoader code [20:36:49] but keeping them together seems to solve some caching problems for us magically, and such [20:41:11] 3VisualEditor / 3Editing Tools: VisualEditor: Tool to set/unset text as highlighted () - 10https://bugzilla.wikimedia.org/64376#c9 (10Elitre) I had seen that, but given that it didn't mention the word "text" and that elsewhere you were so strongly against the idea, I had to assume it was actually about... [20:42:58] MatmaRex: I think using RL to server the rendered images is good [20:43:24] 3VisualEditor / 3Mobile: VisualEditor Mobile: [Regression wmf4] Link suggestions are appearing in a broken way inside Link Inspector in mobile - 10https://bugzilla.wikimedia.org/72402 (10James Forrester) p:5Normal>3Low [20:43:27] the concerns I have are more to do with "are we getting the API right, are these URL params a mistake?" [20:43:40] TrevorParscal: one thing i'm already finding unpleasant is how we assume that we can always prepend a /* */ comment to the response, but that can be worked around [20:45:32] the load.php "api" looks great to me… we definitely need to pass all this information (module, image, variant, format) and don't need to pass anything else [20:46:09] we do want to identify images by the '(module, image)' tuple, not just 'image', right? [20:48:16] Oh, right, /* [20:48:28] Yeah I did that that way because that syntax works in JS and CSS [20:48:32] * RoanKattouw apologizes to MatmaRex [21:02:00] MatmaRex: Yes, that makes sense (image response code in load.php) [21:02:02] well, in RL::respond() [21:02:26] MatmaRex: since they're relatively expense to compute, and because it's cool, I'd like to consider using a unified syntax instead of arbitrarily ordered query parameters (something like image=modulenamevariantetc. [21:02:32] but that's just an idea, we'll see [21:03:41] Krinkle: that'll be easy to change [21:04:13] (03CR) 10Catrope: [C: 04-1] "One last thing, sorry for not noticing this earlier :(" (032 comments) [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [21:06:27] (03CR) 10Catrope: [C: 04-1] WIP: Rerender ContentBranchNodes when the selection anchor leaves (032 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171494 (owner: 10Divec) [21:06:57] (03CR) 10Catrope: "This looks cool. I wonder if this simple logic could replace a lot of the more complex there-can-only-be-one logic in unicorn management." [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171494 (owner: 10Divec) [21:20:37] (03PS2) 10Divec: WIP: Rerender ContentBranchNodes when the selection anchor leaves [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171494 [21:22:45] MatmaRex: I think so, yes [21:22:51] we do want to use that in the API [21:23:16] keeps module-specific images (rather than them being in the global scope on the client as they are currently) possible [21:23:26] do we want to rethink that btw? [21:23:27] TrevorParscal, my placeholder div is not showing.. is it supposed to be display: block; ? [21:23:59] mooeypoo: It should be position: absolute; and in the DOM it should be after the thing that contains all the draggable things [21:24:08] Or inside of it at the end [21:24:12] position: absolute; display: block; width: 5px; height: 20px ; [21:24:20] I'll work the height to be dynamic when i get this to show [21:24:22] Are top and left set? [21:24:28] let me verify [21:24:29] background-color: red; ? [21:24:36] TrevorParscal, *testing* [21:24:43] ... yes [21:24:54] I'm just saying, div's need some sort of background or border or something to be visible [21:25:06] TrevorParscal, oh, I thought you were cirticizing my choice of testing color [21:25:11] no no no [21:25:12] which is criticize-worthy btw [21:25:24] He was suggesting you use that for testing [21:25:36] background-color: red; is also my go-to for "where the hell is this thing" [21:25:44] okay, found the issue... left was NaN [21:25:44] fixed [21:25:46] now it shows [21:25:50] nice [21:25:53] NaNpx! [21:26:13] Someone should make that an actual position [21:26:47] OMG [21:26:48] > $foo.css( 'velocity' ); [21:26:48] OMG IT WORKS [21:26:49] 2e-8 [21:26:51] * mooeypoo stares [21:26:52] pinch me [21:26:56] someone pinch me [21:26:57] > $foo.css( 'position' ) [21:26:59] NaN [21:27:02] Quantum CSS? [21:27:14] RoanKattouw: Heisenberg CSS. [21:27:17] haha [21:38:28] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] The cursor stays active on CE even after opening the dialog for Basic Reference/Citation/Reference list/Page Settings - 10https://bugzilla.wikimedia.org/73097 (10Rummana Yasmeen) 3NEW p:3Unprio s:3normal a:3None Steps to reproduce: 1.Op... [21:38:56] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] The cursor stays active on CE even after opening the dialog for Basic Reference/Citation/Reference list/Page Settings - 10https://bugzilla.wikimedia.org/73097 (10Rummana Yasmeen) 5NEW>3ASSI a:3Roan Kattouw [21:41:12] 3MediaWiki extensions / 3TemplateData: TemplateData: Creating TemplateData blocks should be direction neutral - 10https://bugzilla.wikimedia.org/50436#c19 (10James Forrester) 5ASSI>3RESO/FIX This is now effectively "solved" through general availability of the GUI to edit TemplateData. [21:46:20] (03PS5) 10SuchetaG: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) [21:46:51] (03CR) 10Catrope: [C: 032] Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [21:49:08] (03Merged) 10jenkins-bot: Matching pages from suggestions in link dialog now actually work as links [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/171488 (https://bugzilla.wikimedia.org/51205) (owner: 10SuchetaG) [21:50:26] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression] Link inspector closes itself immediately in Internet Explorer - 10https://bugzilla.wikimedia.org/72392#c9 (10etonkovidova) IE11: Using a shortcut Ctrl+K to insert a link consistently gives the following message twice : DOM7009: Unable to decode imag... [21:50:57] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression ?] While opening the "Citation needed" dialog the "Add more information" link is appearing twice in the dialog - 10https://bugzilla.wikimedia.org/73099 (10Rummana Yasmeen) 3NEW p:3Unprio s:3normal a:3None Created attachment 17063 --> https:/... [21:52:54] 3VisualEditor / 3Editing Tools: VisualEditor: Matching pages list in link dialog should support right-click / ctrl+click to open page in a new window/tab - 10https://bugzilla.wikimedia.org/51205 (10James Forrester) 5PATC>3RESO/FIX [22:08:38] (03PS1) 10Catrope: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) [22:08:51] edsanders: [22:08:53] edsanders: https://gerrit.wikimedia.org/r/#/c/171734/ [22:09:28] James_F: That one --^^ will need a cherry-pick because it's a wmf7 regression [22:09:53] (03PS4) 10Esanders: Add 'indeterminate' state to progress bar widget [oojs/ui] - 10https://gerrit.wikimedia.org/r/171548 [22:10:56] RoanKattouw, why do we focus the document node on a null selection? [22:11:21] edsanders: When .focus() is called, presumably the surface should be focused? [22:12:00] I'm open to doing this differently, but I don't think that a call to .focus() not focusing anything is sensible behavior [22:12:12] I checked if onModelSelect( NullSelection ) blurs the surface but it doesn't seem to [22:12:13] it should do selectFirstContentOffset if selection is null [22:12:22] OK that's sensible [22:12:26] I'll do that [22:12:30] but until the model has been updated, the view shold stay blurred [22:12:35] +u [22:13:10] Yeah that's fair [22:14:19] although even that seems a bit wrong [22:14:37] the point of view.focus() is to reapply the selection from the model [22:15:56] so really the caller should just be calling model.ensureNonNullSelection(); view.focus(); [22:16:01] or something much better named [22:16:29] in most instances the first part won't be necessary (e.g. returning from a dialog) [22:16:45] (03PS1) 10SuchetaG: Murdering the config object rel from OptionWidget [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 [22:16:53] and that will only be used on surface(widget) load [22:17:11] RoanKattouw, ping, in case you're writing code [22:17:25] Hmmmmm yeah [22:17:26] I don't know [22:17:42] I feel like .focus() on a NullSelection should be helpful rather than do nothing [22:18:24] possibly - that's analogous to the DOM [22:18:27] In any case this logic should be in ce.Surface, not duplicated in all callers [22:18:32] (not that that's a good guide) [22:18:35] haha [22:19:09] I can live with that then [22:19:54] documentation should be clearer either way [22:21:36] (03CR) 10Catrope: [C: 04-1] Murdering the config object rel from OptionWidget (032 comments) [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 (owner: 10SuchetaG) [22:21:40] OK [22:21:44] Yeah I'll document what it does [22:22:54] (03PS2) 10Catrope: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) [22:23:29] (03CR) 10Esanders: "I'm going with the slider, as it's most consistent with the determinate mode, doesn't make you sick, and makes our theme look less like th" [oojs/ui] - 10https://gerrit.wikimedia.org/r/171548 (owner: 10Esanders) [22:24:56] edsanders: Amended ---^^ [22:33:09] 3VisualEditor / 3ContentEditable: VisualEditor: Tables - applying headings format to table cells will split the table - 10https://bugzilla.wikimedia.org/72990#c4 (10etonkovidova) Verified in betalabs. [22:36:27] RoanKattouw_away, I think you can return on line 555 [22:40:24] 3VisualEditor / 3ContentEditable: VisualEditor: handling preformatted text in tables - 10https://bugzilla.wikimedia.org/73100 (10etonkovidova) 3NEW p:3Unprio s:3normal a:3None Created attachment 17064 --> https://bugzilla.wikimedia.org/attachment.cgi?id=17064&action=edit Preformatted text in tables... [22:40:52] (03PS5) 10Esanders: Add 'indeterminate' state to progress bar widget [oojs/ui] - 10https://gerrit.wikimedia.org/r/171548 [22:41:24] (03CR) 10Esanders: [C: 04-1] Fix non-linear selection handling in ce.Surface#focus (031 comment) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) (owner: 10Catrope) [22:44:56] edsanders: Good point, fixing [22:45:13] (03PS3) 10Catrope: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) [22:45:14] I mean - test it too :) [22:45:26] edsanders: Also as part of testing this I noticed that pasting a table into a reference explodes [22:45:40] ha [22:45:42] Ouch. [22:46:10] Paste data was something like

...

, pasting into a SurfaceWidget that was "empty" (only contained a wrapper paragraph) [22:46:27] Some nasty exception where it's trying to call .getClonedElement() on the document node [22:46:37] (03CR) 10Esanders: [C: 032] Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) (owner: 10Catrope) [22:46:48] Thanks [22:46:54] That's going straight to our deployment list [22:47:13] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] The cursor stays active on CE even after opening the dialog for Basic Reference/Citation/Reference list/Page Settings - 10https://bugzilla.wikimedia.org/73097 (10James Forrester) 5PATC>3RESO/FIX p:5Unprio>3High s:5normal>3major [22:48:40] (03Merged) 10jenkins-bot: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171734 (https://bugzilla.wikimedia.org/73097) (owner: 10Catrope) [22:49:43] (03PS1) 10Catrope: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171757 (https://bugzilla.wikimedia.org/73097) [22:50:27] (03CR) 10Catrope: [C: 032] Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171757 (https://bugzilla.wikimedia.org/73097) (owner: 10Catrope) [22:51:26] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] While opening the "Citation needed" dialog the "Add more information" link is appearing twice in the dialog - 10https://bugzilla.wikimedia.org/73099 (10Rummana Yasmeen) 5NEW>3ASSI a:3Roan Kattouw [22:52:09] (03CR) 10Catrope: "(Apart from the commit message nitpicks this is good to go)" [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 (owner: 10SuchetaG) [22:52:10] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] While opening the "Citation needed" dialog for the first time in a session, the "Add more information" link is appearing twice in the dialog - 10https://bugzilla.wikimedia.org/73099 (10Rummana Yasmeen) [22:52:22] (03Merged) 10jenkins-bot: Fix non-linear selection handling in ce.Surface#focus [VisualEditor/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171757 (https://bugzilla.wikimedia.org/73097) (owner: 10Catrope) [22:53:48] (03PS1) 10Catrope: Update VisualEditor core for cherry-pick [extensions/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171759 [22:55:36] (03CR) 10Catrope: [C: 032 V: 032] Update VisualEditor core for cherry-pick [extensions/VisualEditor] (wmf/1.25wmf7) - 10https://gerrit.wikimedia.org/r/171759 (owner: 10Catrope) [22:55:55] RoanKattouw: https://gerrit.wikimedia.org/r/#/c/171613/ [23:00:50] Guys, I played with the table editor [23:00:52] I love it! [23:01:05] InezK: Good to hear. :-0 [23:03:58] 3VisualEditor / 3Editing Tools: VisualEditor: [Regression wmf7] The cursor stays active on CE even after opening the dialog for Basic Reference/Citation/Reference list/Page Settings - 10https://bugzilla.wikimedia.org/73097#c5 (10James Forrester) 5PATC>3RESO/FIX Going out in 16:00 SWAT. [23:04:34] (03PS2) 10Trevor Parscal: Murdering OptionWidget's `rel` config option [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 (owner: 10SuchetaG) [23:04:42] (03CR) 10Trevor Parscal: [C: 032] Murdering OptionWidget's `rel` config option [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 (owner: 10SuchetaG) [23:07:11] (03Merged) 10jenkins-bot: Murdering OptionWidget's `rel` config option [oojs/ui] - 10https://gerrit.wikimedia.org/r/171740 (owner: 10SuchetaG) [23:34:09] TrevorParscal: you can probably start showing off ResourceLoaderImageModule :) just don't use it in production yet :D https://gerrit.wikimedia.org/r/#/c/165922/ [23:36:11] TrevorParscal: what's the use case for having per-"type" variants defined in the configuration? can we kill that? (i copied this approach from ooui) [23:37:38] TrevorParscal: also, we probably will need to add 'type' to the image identification tuple… i overlooked this. i don't like this "type" business at all [23:49:54] um, it's good for one reason [23:50:21] you can specify that all the images being listed should have a default set of variants (useful for stuff like invert) [23:51:00] and if you don't have per-type variants defined, you end up inverting things you might not want to [23:51:41] also, I will look - but really thank you for your work on this [23:54:36] (03PS9) 10Esanders: Support for generic file drop handlers [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/170907 [23:54:38] (03PS1) 10Esanders: Generic progress dialog for drag and drop uploads [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171770 [23:56:38] (03CR) 10jenkins-bot: [V: 04-1] Generic progress dialog for drag and drop uploads [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/171770 (owner: 10Esanders)