[21:08:11] is there an ooui demo online somewhere? [21:08:15] 3VisualEditor / 3Editing Tools: VisualEditor: Search results (for links, redirects, categories…) should select the exact match by default - 10https://bugzilla.wikimedia.org/69716 (10WhatamIdoing) 3NEW p:3Unprio s:3enhanc a:3None This is somewhat related to bug 62917. When you add an internal link,... [21:08:29] TrevorParscal: No, I keep trying to get Timo to build that [21:14:58] Quick question... is the CitationDialog currently being used somewhere? [21:15:09] I see ReferencedDialog appear, but not CitationDialog. [21:17:40] Yeah it should be used [21:17:45] See generateCitationFeatures in ViewPageTarget [21:17:57] Or the Cite dropdown on enwiki [21:24:55] (03CR) 10jenkins-bot: [V: 04-1] Update VE core submodule to 9582742c8de4b0af78cfdaefb80a328aa3602abf [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/150051 (owner: 10Trevor Parscal) [21:25:47] Now I have 'ooey' sound in my head every time I see ooui. [21:27:05] aharoni, can you remind me what you said about having localized? Was it tags you talked about? [21:27:44] you should change your nick to moouipoo [21:34:07] .... THAT is brilliant mvolz [21:34:48] \o/ [21:34:51] that's as clever as I could come up with for me [21:35:36] I'm totally keeping this nick for a whlie. [21:35:42] :D [21:42:59] 3VisualEditor / 3ContentEditable: ; not recognized as wikitext - 10https://bugzilla.wikimedia.org/69689 (10WhatamIdoing) [21:42:59] 3VisualEditor / 3Editing Tools: VisualEditor: Support for definition-lists - 10https://bugzilla.wikimedia.org/37938 (10WhatamIdoing) [22:16:43] RoanKattouw, if I want to check into stripping spaces from beginning of paragraphs (to reduce s) is this done in the toDataElement of the ve.dm.ParagraphNode ? [22:16:57] I'm still a bit confused about which part sets up what we send back to parsoid [22:17:57] Uhm [22:17:59] Maybe [22:18:06] I haven't really thought about how to do this yet [22:19:05] This may need to be implemented in ve.dm.Converter directly [22:19:12] But that would be ugly [22:19:29] But right now I don't think ParagraphNode's toDomElements has access to the right information? [22:19:31] I don't know [22:24:23] ParagraphNode doesn't have toDomElement [22:24:27] or toDataElement [22:28:56] Oh maybe there's a default implementation? [22:28:59] I forget how this works [22:30:07] Yeah, there's a default implementation down in dm.Model [22:30:13] Ed must have refactored this at some point [22:36:20] Hmm. I was looking at BranchNode [22:36:35] since ParagraphNode is extending ve.dm.BranchNode [22:37:24] Yeah [22:37:26] I was too [22:37:27] RoanKattouw, what does "significantWhitespace" mean btw, do you know? [22:37:32] I just used doc.wikimedia.org to track it down [22:37:35] ve.dm.Node.prototype.hasSignificantWhitespace [22:37:43] It turns out to be ParagraphNode -> BranchNode -> Node -> Model or something [22:37:45] "Check if the node has significant whitespace." [22:37:48] Oh, that [22:37:54] You know how
s are different?
[22:38:02] 	 They start with a space?
[22:38:08] 	 No, sorry
[22:38:10] 	 in wikitext?
[22:38:14] 	 So, OK
[22:38:31] 	 Normally, if we get HTML like 

Foo

, we strip the whitespace at the edges [22:38:35] As well as whitespace between elements [22:38:43] Then we restore it on the way out [22:38:49] Excellent, that should solve the issue--- okay, we shouldn't restore it [22:38:55] This is because wikitext like "* Foo\n* Bar\n* Baz" is common [22:39:12] And it turns into something like
  • Foo
  • \n
  • Bar
  • \n
  • Baz
[22:39:33] What you don't want is for that to turn into a linmod that looks like
  • Foo

  • \n

    ... [22:39:42] Hmm [22:39:43] Instead, you want it to be
    • Foo
    • ... [22:39:50] when editing it at least [22:40:03] So, yes, this is exactly what you're looking for, but no, it's in the exact opposite direction [22:40:07] wait, wait... the \n problem is not a whitespace, it's an end of line. I'm confused. [22:40:14] \n is a whitespace character [22:40:20] It is?? [22:40:23] If we just let it sit there, it would be considered bare text and it would be wrapped in a paragraph [22:40:32] I thought whitespace is just whitespace [22:40:38] Yeah when you parse that you actually get a TextNode with a \n in it [22:40:44] I see [22:40:49] so it's not an actual 'space' space [22:41:20] It's not a U+0020 SPACE character no [22:41:36] But things like \n, \r and \t are also considered whitespace [22:41:37] I see. I thought that might help with stripping. [22:41:42] When I say whitespace I mean /\s/ [22:41:46] right. I didn't think about that. Yeah [22:41:51] Anyways [22:42:00] So under normal circumstances we strip inner whitespace [22:42:12] And that's fine because you don't want the space in "* Foo" to be editable [22:42:20] But for
       that would be a terrible ide
      [22:42:22] 	 a
      [22:42:33] 	 Right, I see what you're saying
      [22:42:38] 	 Because in a 
       every whitespace character matters, including the ones at the beginning, they affect the rendering
      [22:42:46] 	 and that's why we say that PreformattedNode "has significant whitespace"
      [22:42:49] 	 The name isn't great
      [22:42:54] 	 how to do you git review to a branch?
      [22:42:59] 	 But it's somewhat documented in ve.dm.Node.js line 122
      [22:43:04] 	 mvolzoid: git review branchname
      [22:43:11] 	 well that's straightforward :)
      [22:43:16] 	 ty
      [22:43:19] 	 mvolzoid: Before you do that though
      [22:43:26] 	 BUT WAIT
      [22:43:28] 	 Can I ask you some questions to make sure you're doing what you should be doing?
      [22:43:32] 	 THERE'S MORE
      [22:43:46] 	 sure, I'm basically dumping stuff to a branch on extension/Citoid though
      [22:43:50] 	 that doesn't work yet
      [22:43:53] 	 isn't it -d somewhere? (I use shortcuts)
      [22:44:04] 	 moouipoo: That's downloading *from* Gerrit
      [22:44:07] 	 just to like, say, hey I did this, even though it is not ready yet
      [22:44:08] 	 oh, right.
      [22:44:10] 	 :)
      [22:44:12] 	 mvolzoid: If you're just parking nonfunctional stuff, then that's simpler
      [22:44:18] 	 yeash
      [22:44:18] 	 Just submit it as you would a normal commit
      [22:44:29] 	 You can even make stacks of commits on top of each other and Gerrit will understand what that means
      [22:44:41] 	 Conventionally, you'd put [WIP] at the beginning of the commit summary to indicate Work In Progress
      [22:44:52] 	 e.g. https://gerrit.wikimedia.org/r/140244
      [22:44:58] 	 oh ok
      [22:45:40] 	 and while you're here, how do you add new patches to the same commit? 
      [22:46:03] 	 I've been just reverting and sending a new commit and no one cares because it's just me but I should probably not do that.
      [22:46:42] 	 You're saying you want to make a subtle (for some rough definition of subtle) change to a commit?
      [22:47:00] 	 mvolzoid, I always have this open: https://www.mediawiki.org/wiki/Gerrit/Getting_started
      [22:47:11] 	 or, well, now it's more for reference every now and then
      [22:47:25] 	 but that might help, it has all the basic routine for sending stuff to gerrit
      [22:47:41] 	 The way you do that is basically git commit --amend -a    then   git review
      [22:47:48] 	 aha, thanks moouipoo and RoanKattouw
      [22:48:14] 	 Though I'm not sure I understand the final section in that page.. i think it was changed recently. What's up with the git review --download bit
      [22:48:24] 	 oh, that's the -d
      [22:48:26] 	 moouipoo: --download is the long form of -d
      [22:48:27] * moouipoo is silly
      [22:48:57] 	 I also never use git review --no-rebase
      [22:49:13] 	 Well that's because I've configured that to be the default in our repo I think
      [22:49:24] 	 interesting
      [22:49:26] 	 Yeah, you'll see that our .gitreview has defaultrebase=0
      [22:49:35] 	 By the default it does in fact rebase every time you submit anything
      [22:49:38] 	 Which is super annoying
      [22:49:41] 	 ha, so I should do that when I send to all other repos?
      [22:49:55] 	 It's somewhat sensible for submitting new things, but it's a pain when amending
      [22:50:01] 	 yeah
      [22:50:15] 	 No MW core also has defaultrebase turned off
      [22:50:23] 	 I wouldn't be surprised if all WMF repos did
      [22:50:29] 	 TemplateData seems to too
      [22:50:38] 	 i didn't use --no-rebase anywhere, and i never had a problem
      [22:50:54] 	 so either my rebases weren' terrible (and I didn't notice them) or all the repos I've used have it turned off
      [22:51:30] 	 mvolzoid: Hmm, it looks like https://gerrit.wikimedia.org/r/#/c/154962/ largely reverses https://gerrit.wikimedia.org/r/#/c/154961/ ?
      [22:51:42] 	 3VisualEditor / 3Editing Tools: VisualEditor:  Search results (for links, redirects,  categories…) should select the exact match by default - 10https://bugzilla.wikimedia.org/69716#c1 (10Derk-Jan Hartman) This is aggravated if the exact match is a redirect, because matching redirects are in a separate sectio...
      [22:51:46] 	 moouipoo: I think we made --no-rebase the default widely well before your time
      [22:52:06] 	 woo it failed.
      [22:52:10] 	 It started being a pain pretty early on, so Antoine submitted a patch to git-review to add that option
      [22:52:39] 	 RoanKattouw, that explains it. Still good to have it in the docs, I guess
      [22:52:42] 	 I just never used it
      [22:52:43] 	 (the option to turn it off repo-wide, that is, the flag was already there but having to remember to use it was a pain)
      [22:53:13] 	 Yeah
      [22:53:17] 	 mvolzoid, what failed?
      [22:53:27] 	 which stage in the process?
      [22:54:07] 	 oh, just jenkins. probably something silly, will fix
      [22:54:48] 	 RoanKattouw: no, it doesn't reverse things, also that's all a mess because I had commits to that branch but never pushed any of them. They should all be WIP :) very confusing
      [22:55:09] 	 there's like 3 in there that are all WIP. 
      [22:56:11] 	 mvolzoid: What I mean is, 154961 adds a pile of stuff, and 154962 removes the same pile of stuff
      [22:56:30] 	 (The huge citationParams object, apparently)
      [22:57:10] 	 (03CR) 10Catrope: ve.Filibuster: Scrupulous state logging (0315 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/152123 (owner: 10Divec)
      [22:58:07] 	 yeah there's a lot of junk in that last commit that isn't used at all. it's a huge mess.
      [22:58:11] 	 OK
      [22:58:26] 	 If you would like help cleaning that up (git-wise), I'd be happy to help
      [22:58:27] 	 but my internship literally ends in 2 minutes, just wanted it all up
      [22:58:31] 	 haha lol
      [22:59:04] 	 I will make it nice later. :)
      [22:59:44] 	 next step is actually to commit to Real Projects, which I promise won't look like that
      [23:00:09] 	 namely TemplateData and ve-mw
      [23:00:19] 	 Awesome
      [23:01:04] 	 I'm currently using changes to those on localhost but yeah, a mess, so I won't inflict that on ya'll
      [23:03:10] 	 wait, internship? You're opw?
      [23:03:46] * moouipoo is confused
      [23:07:27] 	 erm, yes. we met? at wikimania? :)
      [23:07:52] 	 wait I will fix my nick
      [23:07:57] 	 there!
      [23:08:12] 	 mvolz, yes, I know we've met
      [23:08:20] 	 oh, you didn't know I was opw?
      [23:08:24] 	 I even gave you a hug
      [23:08:29] 	 I was sure you work for the WMF!
      [23:08:34] 	 oh ha!
      [23:08:44] 	 :D
      [23:08:46] 	 Not that that was the reason for which I said a friendly hello
      [23:09:02] 	 :D
      [23:09:25] 	 Wait, you're finished with your opw, but you're going to stick around, I hope!
      [23:09:54] 	 of course! I have like, 3 whole users!
      [23:11:01] 	 :D
      [23:11:53] 	 I can't believe OPW and GSoC are already at an end for the year
      [23:12:22] 	 YOU can't believe it!
      [23:12:53] 	 I knew about the planning fallacy and yet I fell right into its gaping maw
      [23:13:12] 	 hah
      [23:13:29] 	 (tl;dr i have so much more work to do)
      [23:13:37] 	 that wasn't really very much shorter.
      [23:41:14] 	 3VisualEditor / 3Editing Tools: VisualEditor:  Please show the file name in the media dialog, and make it easy to copy it - 10https://bugzilla.wikimedia.org/69726 (10WhatamIdoing) 3NEW p:3Unprio s:3enhanc a:3None Steps to reproduce:  1.  Open a page that contains an image that has a really long or co...
      [23:41:44] 	 3VisualEditor / 3Editing Tools: VisualEditor:  Please show the file name in the media dialog, and make it easy to copy it - 10https://bugzilla.wikimedia.org/69726 (10WhatamIdoing)
      [23:41:44] 	 3VisualEditor / 3Editing Tools: VisualEditor: Support media (tracking) - 10https://bugzilla.wikimedia.org/37870 (10WhatamIdoing)
      [23:50:39] 	 (03CR) 10Catrope: [C: 04-1] Add an event to updating current values in scalable (037 comments) [VisualEditor/VisualEditor] - 10https://gerrit.wikimedia.org/r/153069 (owner: 10Mooeypoo)
      [23:50:46] 	 moouipoo: ---^^ some relatively minor stuff
      [23:51:27] 	 (03CR) 10Catrope: "I think we have "Insert XYZ" as the button title for every action that inserts something? James?" [extensions/VisualEditor] - 10https://gerrit.wikimedia.org/r/154871 (owner: 10Kaldari)
      [23:52:07] 	 (03CR) 10Catrope: [C: 032] Fixing formatting of Agora (mobile) dialogs [oojs/ui] - 10https://gerrit.wikimedia.org/r/154870 (https://bugzilla.wikimedia.org/69710) (owner: 10Kaldari)
      [23:52:42] 	 (03Merged) 10jenkins-bot: Fixing formatting of Agora (mobile) dialogs [oojs/ui] - 10https://gerrit.wikimedia.org/r/154870 (https://bugzilla.wikimedia.org/69710) (owner: 10Kaldari)
      [23:54:20] 	 RoanKattouw, the reason I'm checking the 'old' dimensions are valid AND new dimensions are valid before doing ve.compare is because I ran into issues with cases where my old dimensions is either undefined or has partial data
      [23:54:59] 	 The entire issue is that there's a problem with ve.compare, I think
      [23:55:16] 	 Well ve.compare() /should/ only break if you feed it undefined as a top-level thing to compare
      [23:55:32] 	 So if one of dimensions or oldDimensions is itself undefined, then it'll go boom
      [23:55:52] 	 yeah, for some reason on initialization it is
      [23:55:53] 	 But if there's something /inside/ there that's missing or undefined or whatever, that should not explode and in fact return false
      [23:55:56] 	 despite the fact I'm assigning ={} to it in the constructor
      [23:56:00] 	 but I'll check it again
      [23:56:05] 	 That sounds very weird
      [23:56:13] 	 I had a couple of issues that I then changed, I'll recheck