[09:02:57] <DaveMethvin> jaubourg timmywil gibson042 rwaldron mikesherov markelog gnarf m_gol and all interested parties
[09:03:09] <m_gol> present
[09:03:33] <DaveMethvin> hey m_gol
[09:03:37] <mikesherov> po
[09:03:48] <mikesherov> ^ hebrew for "here"
[09:03:54] <DaveMethvin> interesting article here, mentioned a Polish site doin it rong http://www.html5rocks.com/en/tutorials/speed/animated-gifs/
[09:03:58] <gibson042> po po
[09:04:14] <DaveMethvin> never thought that leaving an animated gif obscured would cause that much problem
[09:04:37] <markelog> hey-hey
[09:04:57] <DaveMethvin> hey markelog
[09:05:03] <jaubourg> hi all
[09:05:46] <DaveMethvin> pretty good turnout today!
[09:06:03] <DaveMethvin> so i'm actually going to be home this week
[09:06:11] <DaveMethvin> i want to get some stuff done!
[09:06:33] <m_gol> haha, nice, this is a pretty popular site here
[09:06:42] <m_gol> Polish eBay, everyone sells/buys there
[09:06:49] <DaveMethvin> so jaubourg since you're here
[09:07:09] <DaveMethvin> coupla new ajax tix having to do with ie11
[09:07:17] <jaubourg> yeah, just saw that
[09:07:33] <jaubourg> if it's just a matter of changing the test, then so be it
[09:07:33] <DaveMethvin> looks like we just need to tweak the detect for 1.x
[09:07:45] <DaveMethvin> but the other one was for 2.x where we took it out?
[09:08:40] <DaveMethvin> http://bugs.jquery.com/ticket/14424
[09:08:41] <jaubourg> I need to check, sorry I'm being chatted and messaged at the same time (crazy time)
[09:08:57] <DaveMethvin> sure, can you take a look at these this week?
[09:09:10] <jaubourg> yep
[09:09:37] <jaubourg> the isLocal stuff was tricky in the first place
[09:10:09] <DaveMethvin> yeah, if it can be solved an easier way we just need to tell people that
[09:10:23] <jaubourg> ok
[09:10:52] <DaveMethvin> like if it seems really specialized we could tell people to pass in the xhr factory
[09:12:02] <DaveMethvin> damn bug tracker ... too  slow!
[09:12:59] <m_gol> sorry guys, I didn't have enough time to put up this source map proposal in more detail; I re-posted what I wrote on our bug tracker here, though: https://groups.google.com/d/msg/mozilla.dev.js-sourcemap/HsImWQq9J8g/jaRrwfp85OwJ
[09:13:16] <m_gol> there's one issue with this proposal, though; namely the fact that there might be >1 source file which makes specifying them outside of the map file somewhat difficult
[09:13:37] <m_gol> anyway, at least a reference in the minified file could be easily ommited
[09:13:43] <DaveMethvin> m_gol oh, a single map can have multiple sources?
[09:13:53] <m_gol> sure, happens quite often
[09:14:04] <jaubourg> one script per page mantra
[09:14:08] <DaveMethvin> well not in jquery :)
[09:14:13] <m_gol> e.g. when compiling lots of Sass files to one resulting CSS
[09:15:21] <m_gol> it seems the file parameter will be optional so once Chrome removes the warning we might just point to the CDN instead of removing the reference
[09:16:05] <DaveMethvin> can we leave out the file name in the map and it will still work?
[09:16:17] <DaveMethvin> how does it know the original source name?
[09:17:33] <m_gol> it doesn't have to ;)
[09:17:50] <m_gol> since it all starts with the minified file asking for a map, the reference is already there
[09:18:12] <DaveMethvin> there's the min, the map, and the original uncompressed source file
[09:18:19] <DaveMethvin> the min references the map
[09:18:40] <DaveMethvin> how do we find the original js
[09:18:49] <m_gol> ah, this original
[09:18:55] <m_gol> yeah, that's the problem
[09:18:57] <DaveMethvin> lol
[09:19:09] <DaveMethvin> damn infinite loops
[09:19:22] <m_gol> since there can be many of the sources, specifying all of them on the script tags might not be the best
[09:19:37] <DaveMethvin> any solution that asks users to edit their html will fail
[09:19:40] <m_gol> but 99% of our problems are caused by the minified file asking for an incorrect map file
[09:20:00] <m_gol> so I guess once the map file is reached, the probability of the failure lowers a little
[09:20:06] <m_gol> users don't have to edit their HTMLs
[09:20:20] <m_gol> if they don't, they just wouldn't get the map and that's it
[09:20:23] <m_gol> but at least no error
[09:20:44] <m_gol> if one specifically wants a map, then they can put the reference in the script tag and it should work
[09:20:58] <m_gol> I think it's much better
[09:20:59] <DaveMethvin> we can put a complete CDN reference to the map but i don't think that fixes things unless the browsers make changes
[09:21:15] <m_gol> you mean, Chrome throwing warnings on the console?
[09:21:33] <m_gol> bc apart from that from what people said on the group, browsers are already ignoring the presence of this path
[09:21:46] <m_gol> I mean, the path, apparently not the presence though :P
[09:21:56] <m_gol> "we don't use it but you have to specify it!"
[09:22:12] <DaveMethvin> so do they use an implicit naming algorithm to find the source file?
[09:22:22] <DaveMethvin> maybe i need to go back and look at this agian
[09:22:24] <DaveMethvin> again
[09:23:38] <DaveMethvin> okay
[09:23:56] <DaveMethvin> and I quote from our map file
[09:23:57] <DaveMethvin> {"version":3,"file":"jquery.min.js","sources":["jquery.js"],"names":
[09:24:20] <DaveMethvin> that is for the map referenced by jquery.min.js
[09:24:26] <DaveMethvin> so we can make file: ""
[09:24:32] <DaveMethvin> but sources has to be there?
[09:24:37] <m_gol> yes
[09:24:39] <DaveMethvin> what is it relative to?
[09:24:42] <DaveMethvin> the map?
[09:24:58] <m_gol> sourceRoot
[09:25:07] <m_gol> I think it's the map file location by default
[09:25:12] <m_gol> unless you specify it
[09:25:23] <DaveMethvin> oh joy
[09:26:32] <DaveMethvin> but
[09:26:40] <DaveMethvin> we still haven't solved the renaming problem
[09:26:53] <m_gol> I wanted to have the sources property out as well but I see the problem when you have 100 files compiled to 1
[09:27:24] <DaveMethvin> i guess our map comment would say http://code.jquery.com/jquery-1.10.2.min.map
[09:27:24] <m_gol> at this point I don't see a simple way to solve the renaming problem and to still allow for maps with multiple sources which is crucial
[09:27:31] <m_gol> yep
[09:27:39] <DaveMethvin> that is, you could rename the min file all you wanted
[09:27:57] <DaveMethvin> it would still point back to the numbered map and js on our cdn
[09:28:19] <m_gol> but, as I pointed out, if the reference to the map was just left out from the file, then we: 1) get consistent hashes; 2) if sb copies just the minified file, they don't get the map but they don't get 404 as well and everyone is happy
[09:29:05] <DaveMethvin> you mean left out of the min file entirely? yes, at the expense of sourcemap support, which i am warming to
[09:29:06] <m_gol> DaveMethvin: I don't have a good idea of how to allow safe renaming together with supporting multiple sources in the map file, do you?
[09:29:15] <DaveMethvin> no
[09:29:30] <DaveMethvin> the spec just doesn't work well for such things
[09:29:53] <DaveMethvin> for our case, the easiest thing would be if users could associate a map manually
[09:30:52] <m_gol> sure, but that use case is a job for dev tools in browsers, not for the soruce map spec
[09:31:25] <DaveMethvin> i am still wondering whether we get enough benefit to leave this comment in ... what does everyone else think?
[09:31:56] <DaveMethvin> so should it just be the abs path to the cdn map?
[09:32:14] <m_gol> I say with current state of things if Chrome stops throwing warnings about the missing file property we can just omit it and point to the CDN; if they don't remove the warning we should remove the map
[09:32:30] <m_gol> we should leave it for custom compilation, though
[09:32:42] <m_gol> since in such a case problems are not highly probable
[09:32:45] <gibson042> I'm on the fence, but inclined to agree with m_gol
[09:33:08] <m_gol> so no map in release scripts, but maps in regular grunt compilation; that's what I think
[09:33:21] <m_gol> cc paul_irish
[09:34:33] <DaveMethvin> okay, anyone want to tackle this?
[09:34:45] <DaveMethvin> m_gol?
[09:35:00] <m_gol> sure
[09:35:08] <m_gol> when do we plan to release?
[09:35:13] <m_gol> any new betas planned etc.?
[09:35:19] <DaveMethvin> definitely at least one beta
[09:35:39] <DaveMethvin> we  need to get these tickets closed
[09:35:49] <m_gol> ok, so we're not in a hurry; I'd like to give DevTools guys some time to remove the warning, then linking to the CDN will be enough
[09:36:04] <DaveMethvin> yeah i agree it would be better to have chrome not whine about this
[09:36:11] <m_gol> I'll be pinging paul_irish about that ;)
[09:36:33] <DaveMethvin> i'll put your  name next to it
[09:36:35] <DaveMethvin> :)
[09:36:46] <DaveMethvin> okay, back up the list
[09:36:54] <DaveMethvin> bower/grunt ... what was the outcome on that
[09:37:11] <DaveMethvin> i think after the meeting 2 weeks ago it sounded like we were not going to do a local install
[09:37:31] <markelog> I figure out the way how we can install bower dependencies without global package and do not break any conventions
[09:37:34] <DaveMethvin> but i don't know if anyone followed up with danheberden or others aobut it
[09:37:50] <markelog> i thought timmywil has pinged him?
[09:38:18] <markelog> anyway, we could use a grunt task to install bower dependencies
[09:38:26] <markelog> https://github.com/yatskevich/grunt-bower-task
[09:38:58] <markelog> if we use this task bower package is not required to be installed globally
[09:39:26] <markelog> this task let us even dispose from ugly bower components paths
[09:39:31] <markelog>  like this one
[09:39:35] <markelog> "../bower_components/qunit/qunit/qunit.css"
[09:39:44] <DaveMethvin> hmm seems promising
[09:39:45] <markelog> timmywil, scott_gonzales: thoughts?
[09:40:21] <DaveMethvin> markelog seems like timmywil is afk but can you discuss with him this week?
[09:40:29] <markelog> sure thing
[09:40:47] <markelog> current PR could be closed i suppose
[09:41:06] <DaveMethvin> yeah sounds like we're not going that direction
[09:42:33] <DaveMethvin> gibson042 are you still okay looking at the Migrate fails?
[09:42:55] <gibson042> https://github.com/jquery/jquery-migrate/pull/52 should fix them
[09:42:55] <b-ot> gibson042: Pull request #52 by gibson042 (1w 6d ago): Fix #51: New init method for subclasses
[09:43:13] <DaveMethvin> oh yeah, i almost kinda remember that
[09:43:17] <DaveMethvin> i'll review and land
[09:43:54] <gibson042> I just wanted another set of eyes since it messes with $.fn.init
[09:44:34] <DaveMethvin> yeah, definitely not to be trifled with! :)
[09:45:04] <DaveMethvin> alrighty then, any other issues?
[09:45:49] <DaveMethvin> let's get all these wrapped up and see where we are next Monday, I think we should be in good shape and maybe do another beta then
[09:46:24] <markelog> one other thing
[09:46:32] <markelog> https://github.com/jquery/jquery/pull/1396
[09:46:33] <b-ot> markelog: Pull request #1396 by markelog (2w 16h ago): Fix #12757 - use jscs to check the style guide
[09:46:37] <markelog> is it ready?
[09:47:03] <DaveMethvin> i need to review that again
[09:47:14] <markelog> okay!
[09:47:15] <DaveMethvin> i'll do that and comment on the pull
[09:47:43] <DaveMethvin> in the next beta we need to break something to be sure people are testing
[09:48:23] <DaveMethvin> alrighty i guess that does it
[09:48:31] <DaveMethvin> see you in -dev