[17:16:46] morning brion [17:17:05] mornin! [17:17:18] happy friday :-) [17:20:00] brion: watch/unwatch star is finally live in production [17:20:07] awwwwesome [17:20:23] *brion is at http://www.techcrunch.com/real-time-crunchup-sf/ [17:20:35] it was your original idea which we materialized [17:20:40] yay! [17:21:22] techcrunch sounds interesting [17:22:09] twitter coo is speaking, interesting [17:25:30] it's very puurty [17:25:33] I like it [17:25:39] it spins! [17:26:17] translation: oooh, shiny! [17:27:59] hi peeps [17:28:54] hi TrevorParscal [17:49:58] morning all [17:50:32] goooood morning [17:55:14] heading in now.. [18:41:34] brion: parul would like to use ideatorrent, but there is still an issue with authentication. what would we need to do to get openid going? [18:47:27] openid going for what? [18:47:29] in what mode? [18:53:25] werdna: openid on wikimedia, as a provider [18:53:32] openid provider? i think we could turn that on pretty easy; i've been playing with it for statusnet and fixed some regressions in the provider support recently [18:53:38] werdna: so that other non-mediawiki apps can use it [18:53:44] (i'd recommend not deploying the client mode yet though) [18:54:04] an openid provider would be nice [18:54:07] brion: yeah, just the provider portion is necessary for now [18:54:11] handy for the toolserver and such [18:54:20] (until we've confirmed it interacts right with sul) [18:54:22] *nod* [18:54:33] brion: but... doesn't each wikimedia wiki have its own user database? [18:54:50] quick warning: openid plugin requires grabbing the php-openid library also, currently it's not shipped with th eplugin in svn [18:55:36] Ryan_Lane: right... and there's the central sul one, and the openid users are kinda specially handled, so not sure what happens when you combine them ;) [18:55:49] should be no prob as provider, but i'd say still test it with centralauth on before deploying ;) [18:56:01] oh also -- the php-openid lib fails on php 5.3, but last i checked we're still running 5.2 so that's ok [18:56:02] brion: won't that make it difficult to make a provider? [18:56:24] Ryan_Lane: nope, you can just use your user page at any wiki as a provider url [18:56:31] ooohhhh ok. [18:56:44] awesome. that would be perfect [19:05:49] ugh. the openid plugin for drupal really isn't that nice. it's a little buggy too [19:21:54] TrevorParscal and nimish_g don't forget to sign-up for lunch on monday [19:22:10] i'll sign up for you, howiefung [19:24:12] thank you! [19:24:44] nkomura, do you have a link for the lunch? [19:25:03] it is in the email i just sent out [19:34:20] where is Roan? I forget, is he not working today? [19:38:49] brion: are parser tests running and updating CodeReview using SVN hooks? [19:39:53] yeah, there's a post-commit hook to ping the test server (on wikitech.wikimedia.org) [19:40:07] which then svn up's, runs the tests, and posts the results to mediawiki.org [19:41:18] brion: ok. cool. thanks. [19:44:03] TrevorParscal: roan was working earlier in his normal hours, but he has an family event to attend tonight [20:40:01] hi RoanKattouw [20:40:07] Hi [20:40:17] how did the celebration go? [20:41:57] It was nice [20:42:16] TrevorParscal: I was at a dinner for my grandparents' 50th anniversary [20:42:53] aweseom! [20:42:58] how did that go? [20:43:26] It was nice. We went to an "American Fun Restaurant" which, as you can imagine, was kind of wild-west like and had a buffet of mostly American food [20:43:29] RoanKattouw: you wanted to take next week off, is that still a plan? [20:43:45] Did I say that? [20:43:54] I don't even remember; I'll be working like usual next week [20:43:56] I'd love to know what their version of american food was like! [20:44:10] Well their meatloaf wasn't as good as Melissa's, honest [20:44:11] you wanted to adjust your hours, remember? [20:44:27] awww [20:44:27] so you wanted to take time off during thanksgiving and christmas [20:44:29] Yeah I have to do that at some point [20:45:05] but if you are planning to be available, that's great news for us as we need to crank up for citron [20:45:07] But we don't celebrate Thanksgiving so next week sounds kind of arbitrary; of course if you guys are all out, it makes sense [20:45:48] Yeah I'll probably just continue working blindly till I hit the limit, then relax a little bit and do some volunteer work on other MW-related things as well [20:46:20] TrevorParscal: I took a stab at redesigning the link dialog status thingy per Hannes's design [20:46:22] we (team in sf) are off Thursday and Friday [20:46:26] I ran into a bug though were http://www.techcrunch.com/real-time-crunchup-sf/|http://www.techcrunch.com/real-time-crunchup-sf/ [20:46:28] Oops [20:46:50] I ran into a bug where keypresses in any dialog textbox don't result in a letter being added, but do trigger the AJAX stuff [20:46:58] nkomura: Isn't Thanksgiving on Wednesday? [20:48:05] thanksgiving is always on a thursday [20:48:24] Hm, then I don't know why I thought it was on my birthday this year [20:48:51] RoanKattouw: we are around till wednesday [20:49:00] and the following day is called black friday... i think something to do with the retail stores all going into the black (postiive cashflow) from all the christmas shopping [20:49:05] haha [20:49:17] OK so it's just Mon-Wed next week [20:49:25] right [20:49:39] and i want to move our meeting to Wedneday actually [20:49:53] would it for you RoanKattouw? [20:50:15] Yes, but I'll be away from 9:45 till 10:15ish [20:50:18] (PT) [20:50:51] you can join late [20:51:13] or we can start 10:30pm [20:51:13] Yeah, that should work [20:51:19] Either works for me [20:58:48] i sent code review request to tim yesterday [20:59:04] but the rev status hasn't changed [20:59:18] What did you request review for? [20:59:28] adam's revs [20:59:54] Ah yes, got the CC just now [21:00:02] just forwarded you the email as i cc'd rob instead of you :-( Thunderbird's auto fill often tricks me [21:02:33] nkomura: why a star for watch/unwatch? A star mlooks like "favotite" or "rate" to me. :S [21:03:32] fajro: we kicked around the ideas and staged for open for comments for a few months [21:04:11] and the star was most recommended approach by our team and also supported from the community [21:05:20] i wanted to wait until adam_miller's revs are deployed [21:05:48] but i think we should start updating prototypes as you all develop new stuff as early as monday [21:06:25] *nkomura goes to pick-up lunch [21:22:43] nkomura_away: Does that mean we go back to the old scheme of running bleeding edge code on the prototypes and the deployment candidate on the deployment wikis? [21:23:26] Also, for the Citron release, I would really prefer not to have a deployment branch and to just deploy trunk. We can do that now because trunk won't be polluted with features for the next release, because there isn't one [21:24:56] isn't there something after citron? [21:25:44] i like not having difference from what we prototype and deploy [21:25:58] nkomura_away: Me too, makes it much easier on the tech side too [21:26:05] And this time it's actually doable [21:26:08] werdna: Not that I know of [21:26:46] there's something scheduled for january or so, no? [21:26:54] if not, what do you guys do after december? :P [21:27:06] if we can stage to-be-Citron from early development through testing till the deployment, that'd be ideal [21:27:12] Exactly [21:27:40] werdna: Well I personally doubt Citron will be completely ready by January, but that aside, there's a difference between deploying Citron as an opt-in thing and making it default [21:27:55] The latter will probably trigger more bug reports and stuff [21:28:09] so you'll be doing maintenance and improvement after full release? [21:28:14] we do need environment where we can stage the language communities we are evaluating [21:28:31] It's a 20 month project which began in February, in theory you're only half way through the project [21:28:39] so I'm wondering what you get up to for the other half :) [21:29:08] 20 month project? [21:29:28] stanton usability is wrapping up in march as far as this project cycle goes [21:30:18] multi-media usability which is kicking off will last till October 2010 [21:30:34] werdna: Did you mean to ask which features will be in Citron, or what we'll be doing after March? (Answer to the latter: unknown) [21:30:54] the latter [21:31:02] nkomura: ah, I thought stanton was for 20 months [21:31:14] Right well yeah the project as it exists now dies in mid-April IIRC [21:31:28] (or the grant expires at that time, dunno) [21:32:04] i'm hopeful that project still lives regardless its form [21:32:17] yes, but obviously I was wrong about the duration of stanton [21:32:53] so back to prototype discussion [21:33:44] i'd like to discuss it among the team, but i think we want to reflect the work-in-progress going forwrad [21:35:03] Yeah that was my idea [21:35:29] I honestly don't see what else we'd want to do, as there are no intermediate releases and there's no point in stabilizing something that's not gonna be released [21:37:03] the approach we took for staging beta enhancements on prototypes were necessary steps [21:37:32] as deployment environment was not scaling to verify multi-language instances [21:37:58] we will have to the same for citron if there is a major gap to usability release and deployment branch [21:38:11] and start staging on test.wikipedia.org early too [21:40:35] Two problems there: [21:41:21] First, like I said, I would like not to have a deployment branch, but to just do everything in trunk. We used deployment branches previously to separate deployment candidate code from ongoing development, but in this case there will be no ongoing development for post-Citron stuff, as there is none [21:42:50] Second, staging on test.wikipedia.org scares me because 1) it's in the cluster (albeit a little bit isolated) and 2) every time someone does a largeish (more than 1 or 2 files) rollout of something, our code will be pushed to all servers. It would be disabled, but the code would still be there and could break stuff in (albeit trivial) ways [21:42:54] it is possible we have another release in march [21:43:06] which we call it D-something releasea [21:43:15] Yeah that's true [21:43:21] If we decide there's gonna be one, we should branch [21:44:03] But as long that decision hasn't been made, I'd like not to branch [21:44:35] let's discuss when we talk next wednesday [21:44:39] can it wait till then? [21:45:42] Sure [21:46:22] Is it OK to start running bleeding edge code on the prototypes once Adam's stuff goes through? I'll keep the deployment wikis around [21:47:07] yeah [21:47:49] as long as the code is intended for citron release [21:48:14] i would like to keep experimental features in sandboxes [21:48:40] as we set the expectation to the community that prototypes are where they can see upcoming features [21:54:07] Yeah [21:54:33] Off the top of my head I don't think we have any experimental non-Citron-bound features now or in the immediate future, right? [22:04:58] side-by-side preview is unlikely be part of citron [22:05:21] Why is that? [22:05:56] we need to make sure template forms and content folding are working [22:06:20] dialogues and NTOC require quite polishing work too especially with i-frame introduction [22:07:03] You mean like time/resource-wise then? [22:07:27] yes, and also completing our unfinished work [22:08:26] we cannot keep reaching out to new features when the features we are offering are not complete [22:09:15] we are basically looking at three weeks of development time [22:10:17] so we need to be selective about the features we can put our effort in [22:11:02] Rigth [22:11:54] In my (possibly biased) opinion, side-by-side preview is working nicely, so I'd advocate having Calcey (or ourselves) just taking a stab at it to see how stable it is [22:12:44] have you tested for RTL in IE family? [22:12:53] Nope :) [22:12:55] IE families I should say [22:13:13] I'll do that real quick tomorrow, see what happens [22:13:23] I think I tested it in RTL on Firefox [22:13:43] k, but the priority is link dialogue as you know [22:14:27] Yeah [22:23:46] TrevorParscal: will you comment to the following post? [22:23:47] http://usability.wikimedia.org/wiki/Talk:Prototype#Question_from_JS_developer [22:24:40] I'll do it [22:25:04] thanks RoanKattouw [22:36:52] *RoanKattouw is on a talkpage commenting spree [22:39:20] nkomura: Quick question: would you agree that a "trash can" or "death row" area where articles stay for a while before really being deleted is outside our project scope? [22:40:27] as far as the stanton usability project is concerned, yes [22:41:02] if it is a suggestion, you do not need to shoot them though [22:41:33] I just commented that, while it might be a good idea, we're not gonna implement it [22:43:02] RoanKattouw: please remember this wiki is likely to live beyond march [22:43:19] True [22:43:45] A "death row" or "trash can" is a policy thing that should be suggested to the wiki communities. It's not a technical thing and it's not in our project scope, so although it may be a good idea, this project is not gonna implement it. --Catrope 22:41, 20 November 2009 (UTC) [22:43:47] you can say the next two release are not focusing such a future [22:43:58] i saw it [22:44:23] i don't think people have different perception for what this project means [22:44:56] it will be even more vague once multi-projects start living in one wiki [22:45:44] i suggest your end your comment after the first sentence [22:46:39] How about clarifying that "this project" means the Stanton project? [22:47:11] Or saying it's outside the scope of the usability projects currently running [22:47:20] you can say that, stanton usability project [22:47:42] stanton grants multiple projects for WMF [22:48:09] Done [22:48:11] i've got to get back to my blogging, i will be idle for a while [22:48:23] Oh yeah? What other projects? (Go blog first if you like, no rush) [22:48:23] nkomura_idle [22:49:15] hardware purchase, outreach and now general operatons [22:49:38] these programs are announced through foundation's blog [22:53:05] whats the schedule for ford btw? [22:53:28] I thought it had started like months ago....? [22:53:39] the project is postponed [22:53:51] guillaume is doing some preparation work [22:54:05] to-be-hired developer is likely to start early december [22:54:35] the project is expected to take off in december and concluded in october 2010 [23:07:58] RoanKattouw: excellent response to http://usability.wikimedia.org/wiki/Talk:Prototype#Question_from_JS_developer [23:08:00] thank you [23:08:06] Sure [23:08:27] i saw your commit comment about my really long line [23:08:45] that line is a bit of dark magic [23:08:49] did you read it? [23:13:15] No, not really, will do [23:13:31] we are finding strange bugs in it... [23:13:48] opera is being retarded as usual [23:14:43] It binds some events on the main document to a handler inside the iframe? Why? [23:15:03] (Also, all dialog textboxes are freezing up for me on HEAD, can you reproduce that?) [23:16:03] in IE8 they are.... [23:16:18] and, we do that because that's the only way to get the events to propagate [23:16:29] we tried more... normal methods [23:16:32] Oh wait no, you bind on the iframe's document [23:16:41] Heh what else did you try? [23:17:02] (BTW, the iframe seems to have whitespace collapsing issues, which is kinda weird considering stuff's in a

)
[23:17:20] 	trying to do a bind from the main window, selecting the iframe docment
[23:17:25] 	basically the reverse approach
[23:17:38] 	I removed the pre...
[23:17:43] 	in head
[23:17:53] 	Ah that explains :P yeah I actually remember advocating thta
[23:18:00] 	yes
[23:18:07] 	That means my text insertion methods need to be smarter
[23:18:17] 	And IE-compatible, I'll work on that next week
[23:18:23] *RoanKattouw 	makes TODO list
[23:18:28] 	if it's causing too much trouble, we could just automatically recreate the pre anytime it's deleted and move the cursor inside of it...
[23:18:34] 	so on page blank you don't loose it
[23:18:54] 	but the issue of the browser allowing people to add things outside the pre is a problem still
[23:18:58] 	*would be
[23:19:28] 	Yeah let's just lose the pre
[23:19:56] 	i'm trying to work out some strange IE bugs
[23:20:12] 	did you get a copy ordered yet?
[23:20:45] 	Yeah I'm all set up
[23:20:52] 	Of course I didn't order one (shh)
[23:21:24] 	VirtualBox turned out to be the missing links (thanks Andrew)
[23:22:50] 	we can at least get you a license so you don't end up like this guy http://itvibe.com/news/4060/
[23:22:59] 	unless you already have one
[23:23:58] 	No, I don't have a license. Feel free to buy me one, but at least I'm not /selling/ copies
[23:24:35] 	OTOH I got Vista+license when I bought my laptop, but I threw it out and installed Ubunty
[23:24:37] 	*u
[23:24:54] 	TrevorParscal: My TODO list for this weekend and beyond, suggestions welcome http://pastebin.org/55644
[23:25:37] 	looks solid
[23:27:25] 	TrevorParscal: http://usability.wikimedia.org/wiki/Talk:Prototype#Location_of_the_search_box
[23:44:02] 	TrevorParscal: http://usability.wikimedia.org/wiki/Toolbar#Custom_buttons_and_icons --> this guy has a good point
[23:45:27] *RoanKattouw 	--> bed
[23:47:04] 	*roan = &bed;