[08:59:46] hey all [08:59:59] hey, shhh yer early [09:00:02] okay [09:00:04] now its time [09:00:26] ok, let's get started [09:00:27] hey all y'all! [09:00:34] hi hi [09:00:36] danheberden here? [09:00:44] danheberden is on lion [09:01:23] $.when( danheberden ).then( infrastructure ); [09:01:25] ? [09:01:40] heh i think i'm alive and working [09:01:48] though i can't swipe back in chrome :( [09:01:50] infrastructure [09:01:54] ajpiano, any news on recruiting? [09:02:18] so clarkbox is totally rocking the view.jqueryui.com work [09:02:29] nice [09:02:37] last step is to get the url form and that branch listing on the home page [09:02:40] :) [09:02:41] and then it's good to go [09:02:52] and instead of proxying github evertytime [09:02:58] its just serving local content [09:03:03] yay [09:03:06] all thanks to clarkbox's smarts [09:03:15] recruiting... uhh. [09:03:18] you guys see that gnarf guy [09:03:19] :p [09:03:21] sounds like something we could also reuse elsewhere? [09:03:28] jzaefferer ya, just not sure where [09:03:35] like - ui is the place with demos [09:03:39] so wherever else needs that [09:03:40] ajpiano: am I ur success story? [09:03:48] i think so [09:03:48] :p [09:03:49] so, where's the code? [09:04:01] so what did ajpiano do when he was dev relations lead [09:04:05] its a separate branch right now, on jquery/view-jquery-org i think [09:04:08] .... "recruited gnarf" .... [09:04:10] jzaefferer: ^ [09:04:28] totally 404s [09:04:37] the index page for code, eddie said he was going to tackle it along with a big list of other things [09:04:53] but still think he's on vacatinon/in hiding [09:04:54] maybe this? https://github.com/jquery/web-view-jquery-org [09:05:00] jzaefferer that's the one [09:05:11] why is it on http://view.jquery.org/ ? [09:05:18] it's the same site [09:05:34] like - literally [09:05:48] it just [used to] serve different content depending on the server_name [09:06:00] but i think core doesn't particularly need one [09:06:01] alright [09:06:10] very nice :-) [09:06:14] but that's a discussion for non-ui [09:06:21] yep [09:06:26] can we make the repo public? [09:06:31] yeah, clarkbox has been super helpful [09:06:39] jzaefferer i'll just make sure there's nothing private in there [09:06:48] of course :p [09:06:48] github api key or something [09:06:57] MAH PRIVATES IN TEH REPO [09:07:07] ssh root [09:08:27] alright [09:08:41] so danheberden ajpiano, is there still a need to do recruiting for infrastructure? [09:08:58] yeah, we still need to talk about it together [09:09:09] i think there is always a need for it [09:09:22] okay, then talky talky, NEXT! [09:09:46] when you're as cool as ajpiano and I, its hard to find time for that! [09:09:55] dizzocs [09:09:59] scott_gonzalez, danheberden, are you still totally opposed to Wordpress? seems like starting with that for jquery-docs could get as closer to something that actually works much sooner [09:10:23] jzaefferer wordpress/deployment is hardly the problem [09:10:45] otherwise we have to tackle navigation and search from scratch, and we've got nothing for that so far [09:10:54] I think pros/cons of wordpress vs. completely custom would help make that decision [09:11:07] jzaefferer well we're tackling all of those issues with the learning site anyway [09:11:11] so why not reap those rewards [09:11:16] I'm pro WP [09:11:39] its extra infra overhead, harder to manage in my opinion [09:11:50] and we already have a system being built [09:11:56] I'd like to see us write a simple script that handles a github hook on a commit to docs content, dumps the content into the mysql table [09:12:07] so what's the status of navigation and search on the learning site? [09:12:25] you say we already have a system being built, I say we already have a system built [09:12:28] danheberden: i suspect it's because jzaefferer doesn't want to wait for us to cross that bridge.... [09:12:32] rworth a bad system built [09:12:41] rworth: that system is getting built no matter what we do with docs [09:12:45] is this due next week? [09:12:59] are we rushing to get docs deployed? [09:13:09] docs won't be deployed until 1.9 [09:13:16] no ones rushing, but that's also no reason to have no progress at all [09:13:32] jzaefferer i was under the impression we were still working the xslt's and formatting [09:13:34] not deployment [09:13:42] if the xml and stylesheets are a ok [09:13:48] then i'll start focusing on deployment [09:13:52] if we have no solution for deployment I'd rather have that in focus sooner [09:13:58] but we don't even have a schema finalized [09:14:00] the one doesn't block the other [09:14:13] agreed, they can work in parallel [09:14:14] jzaefferer no, but we're just adding more todos [09:14:40] instead of focusing on something and getting it done [09:14:47] we're not making progress on the one thing, but at least that's somewhat solid; I'd like to have something in place that shows the entire setup can actually work [09:15:03] which will also help to convince kswedberg that we're on the right track [09:15:10] +1 [09:15:13] no matter what we are doing, the xml+xlst will be somewhere right? [09:15:15] and getting his help would be nice [09:15:21] which we're getting to/doing with the learning site [09:15:23] gnarf: yes, in github [09:15:39] the open question is how does it end up in prod [09:15:40] if people want to contribute to the learning site [09:15:43] so, what's the status of navigation and search on the learning site? [09:15:45] then we could get some more progress [09:16:21] I talked with Adam last week, I think he was like 'here are various options, we haven't tried anything' [09:16:25] jzaefferer navigation works via nanoc, but ajpiano was working on getting hierarchal nav to work [09:16:30] we still need to investigate search options [09:16:43] yes, that is still where i am [09:16:46] yeah, the biggest first step was rendering markdown into the pages [09:17:03] it's just adam and I on the actual build stuff i think [09:17:13] I've got someone interesting in helping on the QUnit site, which could potentially use the same stack as learning and api [09:17:19] and adam is the one doing everything [09:17:26] right now I dunno what to tell him except work on jquery-docs xml [09:17:32] i guess re: wp [09:17:40] i really, really want to get away from it [09:17:45] why? [09:17:46] if we can get this learning site type of deployment working [09:17:51] it'd be such less overhead on the servers [09:18:01] easier to manage via git [09:18:42] Adam, can we put the search options together somewhere and get some opinions on them? [09:19:05] navigation is static, I'm not worrying about that too much, but search isn't... [09:19:19] i think we just have to actually try implementing one [09:19:20] I feel like we're not yet 50% done with our transition to WP, which I think is a mighty fine system, esp. seeing where we came from. And it sounds like you're proposing we now move to a system that's in development [09:19:30] rather than just assessing things blindly [09:19:39] rworth you were all for this system [09:19:42] wtf changed? [09:19:50] and would this (eventually) be for everything? Caus a couple years ago we sat down and said "Yes, WP for pretty much everything" [09:20:02] I'm all for docs being in github [09:20:14] ajpiano: right now all I have is a skype log somewhere, would be good to write down the options (not on an etherpad) and share that at least with this group [09:20:21] rworth and everyone didn't like that idea via ui meetings [09:20:29] they said that "someone" said wp for everything [09:20:35] this isn't a blog [09:20:41] so why are we using wordpress [09:20:47] ++ [09:20:51] and not editing XML via WP, but there are a ton of other WP features you're now suggesting we walk away from [09:21:04] like the overhead of running shit through mysql [09:21:10] which crashes on all of these mt boxes [09:21:19] that's a data point I don't have [09:21:43] well, its built, right? [09:21:47] i mean - api is using it currently [09:21:49] so if we hit 1.9 [09:21:56] and we can't do the stataic deployment [09:22:01] we can easily deploy using wp, ya? [09:22:30] I think rworth and jzaefferer are looking for proof of that [09:22:33] git pull, import into mysql - that's a simple work around if wee need it [09:22:42] since we have a different structure than core [09:22:48] seems quite simple to me, so let's start there [09:22:53] this would also work for core [09:23:03] fine [09:23:20] btw, direct injection into the tables sounds scary to me [09:23:38] just seems like we're asking the busiest section of the team to working something we are going to try our best to throw away [09:24:03] we're going to try our best to throw away WP? [09:24:15] that's what's coming as brand new news to me, if I'm understanding right [09:24:21] we're going to try our best to throw away WP for API [09:24:33] we've talked about getting rid of WP in many ui meetings [09:24:36] docs sites in general right? [09:24:38] and everyone was on board [09:24:40] WP for authoring and managing the content of API, at a minimum, from my perspective [09:24:53] we were very clear on not using WP at all for API [09:25:30] so we're using wp to deploy content, not even edit it [09:25:33] discussing how to implement search is the first time that WP has been mentioned in months [09:25:34] why use git then? [09:25:47] what's not clear however is how our site functions without it [09:25:58] .html [09:26:11] rworth: what specifically isn't clear? [09:26:15] xslt translated into coolness? [09:26:50] search, navigation, hierarchy, tags, permalinks, slugs, caching, etc. [09:27:07] the only thing WP provides there is search [09:27:18] all the non-author-using-a-wysiwig-editor-from-wp-admin features of WP that we can use even if we author an manage content in git [09:27:20] we have to define navigation, heirarchy, tags [09:27:30] there's no caching because it's all static [09:27:42] except that WP has themes and plugins, and lots of people know themes and plugins and they get tested and updated [09:27:49] and are easy to deploy and upgrade [09:27:57] but we don't need them [09:28:00] what theme? [09:28:03] we have our own custom written theme [09:28:13] right, and isn't it a WP theme? [09:28:15] no [09:28:20] it's raw HTML + CSS [09:28:25] isn't that what Darcy's working on, a WP theme? [09:28:25] it had to be MADE into a wordpress theme for the sake of old sites [09:28:31] got it [09:28:51] search is the only feature we're talking about [09:28:52] seems silly to use git if you want to use wp [09:29:14] we could just copy the api setup over, replace the templates and xslt [09:29:35] we want the docs source to be public [09:29:57] and having it tagged with code will be useful as well [09:29:57] maybe there's a wp plugin to show original content? [09:30:15] you can't send a pull request against WP [09:30:34] right now we have this conversation all the time: [09:30:43] community member: "how do I fix these incorrect docs?" [09:30:45] so we're using wordpress for search, url rewrites and themes? [09:30:49] I don't understand why it seems silly, WP does not mean blog, and why can't a WP site present content that originates in a db other than its own [09:30:53] and importing shit into it [09:30:58] us: "umm...we attempt to give you access to a broken MediaWiki install" [09:31:44] what I'm trying to tackle is how/whether we'll have two sep. sites. I'd rather have jqueryui.com as a WP site that can also host docs, so it can be as integrated as possible [09:31:44] injecting content into WP without an actual API seems really scary to me [09:31:58] I feel like it's almost guaranteed to fail [09:32:14] how so? It's just content in a single field in a single table in mysql? [09:32:21] no way [09:32:26] api.jqueryui.com / blog.jqueryui.com / jqueryui.com [09:32:26] let's table that discussion for now, focus on the xslt/xml stuff and helping the learn-site team set up a static site generator with search [09:32:34] there's crazy metadata on everything, isn't there? [09:32:37] then check if that can work for us [09:33:13] ajpiano, is anyone but you working on learn? [09:33:17] it's dan and i [09:33:34] but "it's dan and i" describes a lot of projects so [09:33:37] weren't there other contributors? [09:33:46] addy osmani is hepling as well [09:33:48] addy is working on content stuffs [09:33:51] but not really in technical stuff [09:34:00] ajpiano: fill me in on learn [09:34:06] i'll see if there's something I can add [09:34:18] do you know ruby [09:34:19] :p [09:34:25] gnarf, wanna explore options for searching? [09:34:37] or rather, test one or two... [09:34:46] not by memory (yet) but have been looking for a really good reason to learn it [09:34:56] hired! [09:35:20] i wish you had said "node" [09:35:21] instad [09:35:23] speaking of hired, your grid contract should be available for signing any day now [09:35:32] gnarf heh, such a hipster [09:36:35] ok, moving on... [09:36:36] I know over 15 programming languages - i'm confident that ruby wont throw me for a loop [09:36:48] I'll update the colorize demo [09:37:04] ajpiano: what's the status of widget factory docs? [09:37:15] and where are they? [09:37:18] because http://wiki.jqueryui.com/w/page/12138135/Widget-factory is way out of date [09:37:53] the same as it was the last time we checked- the first part of that page is relatively recent [09:38:17] just the properties and nuts-and-bolts API points need to be updated [09:38:38] ok [09:38:47] I'll do a run through of that page too [09:40:12] jzaefferer, kborchers: menu [09:40:35] I need to review the last pull request(s) [09:40:47] and the issues with the navigation menu [09:41:06] in general going well [09:41:14] Writing unit tests for keyboard nav [09:41:20] yay!~ [09:41:28] kborchers: using simulate? [09:41:34] Yes [09:41:36] yay [09:41:37] Selectmenu: Talked with Felix last week, he should be working on the remaining open issues [09:42:15] Datepicker: What is says there, update from Hans should come this week [09:42:54] is that on existing or branch? [09:43:02] datepicker branch [09:43:25] Tooltip: I actually reviewed Chris' feedback and posted a reply, mostly asking for clarifiaction, didn't hear back [09:45:09] scott_gonzalez, got an estimate for the 1.6.2 support? any news on 'boolean' ARIA attributes? [09:45:26] yeah, we figured out boolean ARIA attributes are actually strings [09:45:29] and should go through .attr() [09:45:38] I started making some progress on that at the hackathon [09:45:54] scott_gonzalez: anything I can do to help on that front? [09:45:54] still some more work to do, but I don't think we're stuck anymore [09:46:04] good [09:46:18] gnarf: checkout https://github.com/jquery/jquery-ui/tree/core-1.6.1 [09:46:28] you actually have to manually put 1.6.x in there :-P [09:46:38] I've just been pasting the contents of 1.6.2 into the jquery-1.5.1 file for testing [09:46:45] k [09:46:48] as for the Testing point, where do we need to test animations? [09:46:50] we need to test all widgets in all browsers [09:46:53] btw [09:47:07] how do you guys feel about dropping the -1.5.1 part of that filename eventually [09:47:19] gnarf: if you want to work on that, let's communicate on what we're working on so we don't waste time double testing [09:47:26] didn't we answer that a couple weeks back? [09:47:40] rworth: let's just find the ticket [09:47:41] maybe? if you did it slipped my mind [09:47:45] rworth, just write down your answer in public and link to it [09:47:47] we need an FAQ section on jqueryui.com [09:47:49] gnarf: I think we did last time [09:48:01] saying "we already answered that" is not helpful [09:48:20] if you did, i forgot the resolutions, sorry to double ask [09:48:33] np [09:48:35] http://bugs.jqueryui.com/ticket/7171 [09:48:54] there it is [09:49:03] got it about 2sec after scott_gonzalez [09:49:18] ha [09:49:46] again, any input on testing animations? gnarf? [09:50:01] otherwise I'll just drop that from the agenda, been pulling that along for some time [09:50:05] jzaefferer: i've been trying to get the actual effects tested [09:50:17] and making sure things like margins [09:50:19] position [09:50:20] etc [09:50:21] display [09:50:24] show correctly [09:50:27] all thats there already [09:50:33] but i don't know how detailed we want to get [09:50:39] with testing options / api /etc [09:50:46] or how detailed we can get.... [09:51:22] gnarf: I think the context here is testing animations in widgets [09:51:35] like ensuring that using effects for dialog open work properly [09:51:47] you could make a test effect [09:52:02] using the $.effects.effect.test = function() [09:52:09] animate using "test" [09:52:24] make sure it is animating the right elements [09:52:24] ? [09:52:38] is that animation hooking in widget core stuff? [09:52:38] could you put an example together, e.g. extend the tooltip testsuite with that? [09:52:43] yeah, let's figure out what specifically we'd want to test for various components and then figure out what we need to test them [09:53:07] see if tooltip works with Widget._show/_hide [09:53:16] aye [09:53:21] i'll look into it this week [09:53:52] ok, last item [09:54:08] jzaefferer: any news from Felix on download builder [09:54:34] talked to Felix yesterday, he started working on it again just then, but is basically busy all the time with other stuff [09:54:37] quick afk - be back for grid [09:55:00] though he considered supporting incremental updates, told him that's not important right now [09:55:06] what exactly is he working on? something non-Make? [09:55:10] probably worth waiting another week [09:55:18] node-based build file to replace ant [09:55:35] oh, this is download builder, sorry I was thinking Make+node not Ant [09:56:25] well, he's actually working on just the build file [09:56:34] but with the goal of using that for download builder, too [09:56:44] and is he doing that without Make? [09:56:52] I think so [09:57:04] does it Make a difference? /o\ [09:57:22] is he inventing his own build file/build system? Node doesn't even do that, they use WAF [09:57:57] so what, node can't build itself, therefore using something else [09:58:10] which, btw, is Python-based and as generic (programming language-wise) as Make is [09:58:44] I'm saying the fact that we're using Node should not preclude the use of a non-node build system, since (afaik) one does not exist [09:59:05] that is, I don't know of a node-specific build system [09:59:18] why do we need an exsiting build system? we don't actually compile anything [09:59:19] actually there's Jake [09:59:24] which isn't that great [09:59:28] certainly if one is being worked on, it's far behind Make and WAF [09:59:40] just saying let's not reinvent the wheel and NIH [09:59:41] rworth: actually i think rwaldron has already build a "plugin builder" in node [09:59:51] I mean, even building this from scratch wouldn't be any more time consuming that building something using an existing tool [09:59:52] nope [09:59:55] not me [09:59:58] hrm [10:00:01] someone was doing that [10:00:02] ... [10:00:36] the things we need doing are the things build systems do, copy files, run commands, run commands on each file, move files, delete files, zip files, etc. [10:00:48] so we should be using one of those systems, not building one [10:01:14] node has that type of stuff pretty much built in [10:01:25] but, that wheel isn't as sexy! [10:01:30] I want a red bike shed! [10:01:46] ok, grid mtg time [10:01:56] (I'm all for a fun build system) [10:02:15] are we doing the grid meeting here or on skype today? [10:02:30] it's on the schedule as an irc mtg [10:02:35] i hope not skype, i'll sound like a robot 42 minutes later [10:02:48] irc is good; just checking since it moved to skype last week. [10:03:06] that happens at times when its 2-3 people at their discretion [10:03:13] the agenda just says whatever it had there for a long time [10:03:24] considering the small group, I'd rather go Skype [10:03:40] do we know the group size? Who's all here? [10:03:49] me [10:03:55] me [10:04:06] me (nothing to say, just observing) [10:04:18] here [10:04:51] borismoore: do you know if Brad will be here? [10:04:52] hey chrisBannon, independent of this meeting now, can you take another look at the Tooltip wiki page and provide some more details? [10:05:08] sure [10:05:11] Not sure. I think he may not be able to make it today [10:05:15] ok [10:05:22] From what he said last night... [10:05:59] jzaefferer: do you want to give us an update on editing? [10:06:25] there are a few improvements on the main demo: http://view.jqueryui.com/grid/grid-editing/grid.html [10:07:16] I spend some time on the focus handling of the inline editor, by now I think I should drop the blur handler and instead bind to click events outside - current approach is messy and unreliable [10:07:39] I've started migrating to jsrender, requires changes in more then just editing files, so took longer [10:07:48] still only local [10:08:14] but that was already helpful, want to discuss the plugin wrapper of jsrender with Boris [10:08:30] e.g. I think a method that returns a jQuery object, containing the rendered template, is very useful [10:08:39] as long as its a helper, not the required way to go [10:08:53] I just pushed the updated JsViews implementation to https://github.com/BorisMoore/jsviews, and also updated the Grid demos here http://borismoore.github.com/jquery-ui/grid-spf-observable/index.html https://github.com/BorisMoore/jquery-ui/tree/grid/grid-spf-observable to use it. [10:09:06] JsViews gives you the wrapper. [10:09:09] speaking of template, also want to discuss my {{wrap}} strawman, either use that or drop {{wrap}} completely [10:09:36] I'm not sold on jsviews as-is, so would help if we could move that over to jsrender [10:09:43] as that just needs polishing [10:10:09] I need to look at having a wrapper in JsRender, but it should not confuse the issue of JsRender being pure string concatenation [10:11:00] I think we should review JsViews, together, which we have still not done, so that we can then figure out about Grid needs for data link (or not) [10:11:09] What about using template for edit mode [10:11:11] jsrender needs some more API update anyway - the current mix of loading templates and rendering them, as nice as it is in the default case, is a bit messy [10:11:18] So ;users can provide editors [10:11:24] What about using template for edit mode [10:11:37] so users can provide editors (I meant) [10:12:10] that won't change the usage of jsrender, so I'd like to tackle that first [10:12:22] Yes, I'll look at it, and the scenarios [10:12:57] Well let's follow up on a Skype call. Difficult on IRC to explain the pros and cons in detail, or discuss the design implications [10:13:05] are we definitely using an actual input replacement for editing, or has there been any consideration given to using contenteditable? [10:13:31] ajpiano, spinner and autocomplete are based on input [10:13:57] ah, true [10:13:58] I haven't yet considered contenteditable, we don't have anything for that right now [10:13:59] do we know about a11y for contenteditable? [10:14:11] jzaefferer: i have some work i can show you [10:14:36] but it does make me uneasy, it's definitely not *as* well supported as doing input replacement [10:14:49] and i certainly don't know aobut that, rworth [10:15:16] Jennison managed to get his name on the ARIA hacakathon attendees list, so it can't be that bad [10:15:31] but the spinner and autocomplete and datepicker is a relevant thing to consider [10:15:31] though editing hrefs obviously wasn't working well [10:16:48] anything else? chrisBannon, vmitnick, anything you'd like to know about? [10:17:20] we are going to start eval'ing datasource for our grid use [10:17:33] can you point me to the right repo/source to use? [10:17:38] chrisBannon: just use a script tag, you don't have to eval it [10:17:41] :p [10:17:47] http://wiki.jqueryui.com/w/page/34246941/Grid#StatusandTasks [10:17:57]