[09:01:46] <DaveMethvin> timmywil gibson042 rwaldron gnarf m_gol orkel jaubourg mikesherov core meeting time!
[09:02:03] <timmywil> boosh
[09:02:24] <DaveMethvin> hey timmywil howya doon?
[09:02:37] <timmywil> doin good
[09:02:44] <timmywil> and doon good
[09:02:47] <orkel> .
[09:02:54] <DaveMethvin> thanks for cleaning up the new tix this morning
[09:03:00] <DaveMethvin> hey orkel
[09:03:01] <timmywil> np
[09:03:03] <orkel> hey!
[09:03:30] <DaveMethvin> so i think we're in pretty good shape for both 1.10 and 2.0.1
[09:03:33] <m_gol> present
[09:03:47] <DaveMethvin> marked a few tix that need to be wrapped up but i don't think they're hard ones
[09:04:14] <DaveMethvin> anything else you guys think should be in 1.10 or 2.01?
[09:04:49] <DaveMethvin> i keep meeting to verify with rwaldron that his perf fix for the .data() stuff  landed, don't think it did
[09:05:21] <m_gol> ha, right
[09:05:45] <rwaldron> DaveMethvin|away no, not yet
[09:05:56] <rwaldron> I've been tied up with mozilla/b2g stuff
[09:06:00] <rwaldron> sorry :(
[09:06:24] <DaveMethvin> definitely wanted that in 2.0.1 but i think it was pretty much ready wasnt it? I can just find your branch and land it i think rwaldron
[09:06:48] <m_gol> will this: http://jsperf.com/jquery-1-9-1-vs-2-0-0-beta/5 be faster in 2.x after this fix?
[09:07:17] <DaveMethvin> i seem to recall it was a lot better
[09:07:33] <DaveMethvin> honestly i think we should look at attaching the data to dom elements again in 2.x
[09:07:45] <DaveMethvin> it would save a lot of the cleanData costs
[09:07:49] <rwaldron> At the price of eliminating all of the Data rewrite maintainability wins
[09:08:29] <DaveMethvin> when i do profiler runs on real code, cleanData ends up near the top relatively often
[09:08:45] <DaveMethvin> that's the time when it's worthwhile to look at perf
[09:08:49] <m_gol> we can't use dataset bc of IE, right?
[09:08:53] <rwaldron> DaveMethvin the problem with attaching data to DOM elements is that now you're asking for two different code paths
[09:08:59] <DaveMethvin> m_gol right ie9
[09:09:13] <DaveMethvin> rwaldron for jquery 2.0?
[09:09:38] <rwaldron> you're forgetting the plain object use case
[09:09:56] <DaveMethvin> that should work tooo.. i think?
[09:10:15] <DaveMethvin> if someone loses all refs to the plain object, they'll lose the attached data
[09:10:21] <rwaldron> Sure, if you don't mind dumping our data properties onto user objects
[09:10:35] <m_gol> DaveMethvin: IE10 doesn't support it either :/
[09:10:38] <DaveMethvin> well the only complaint we got in the past was boris
[09:10:41] <rwaldron> and we can't guarantee non-enumerability thanks to old android webkit
[09:11:00] <rwaldron> so now we'll be doing this:
[09:11:07] <rwaldron> var mine = {};
[09:11:08] <DaveMethvin> plain objects behavior shouldn't be the tail that wags the dog
[09:11:23] <rwaldron> jQuery(mine).data({foo: "bar"})
[09:11:46] <rwaldron> mine { jQuery394093458039456094586094856: { user: {foo: "bar"}, priv: {...} } }
[09:11:51] <DaveMethvin> yep
[09:12:00] <rwaldron> has nothing to do with tails or dogs
[09:12:07] <rwaldron> it has to do with breaking code that does this:
[09:12:18] <rwaldron> for (var p in mine) { .... }
[09:12:44] <DaveMethvin> sure, i understand taht
[09:12:55] <DaveMethvin> but i also know that $(plain_object) is relatively rare
[09:13:00] <DaveMethvin> compared to $(anything else)
[09:13:23] <DaveMethvin> and we've actually changed that behavior at least 3 or 4 times in the past
[09:13:24] <DaveMethvin> already
[09:13:34] <DaveMethvin> so it's not like behavior on plain objs has been consistent
[09:13:52] <rwaldron> it will break our own tests and data() will become incompatible with 1.x
[09:13:55] <DaveMethvin> so anyway, on the perf regression here, is that code ready? i think it was
[09:14:33] <rwaldron> I haven't looked in a very long time
[09:14:44] <DaveMethvin> if 2.0 were faster i'd be good with it but it's slower and that doesn't seem good since this does turn out to be an issue for code that does a lot of dom manip
[09:14:53] <rwaldron> oh this
[09:14:54] <rwaldron> http://jsperf.com/jquery-1-9-1-vs-2-0-0-beta/7
[09:15:32] <DaveMethvin> i think that was it, or something similar
[09:15:40] <rwaldron> no, that's not anything
[09:15:43] <m_gol> well, here it's faster on Chrome
[09:15:48] <m_gol> but on /5 not
[09:15:49] <rwaldron> all i did was remove all the other operations
[09:16:02] <rwaldron> and reduced it to _just_ the data() operation from /5
[09:16:43] <m_gol> so we have other perf regressions then, outside of data
[09:17:07] <DaveMethvin> remember tho that removing an event calls data apis several times
[09:17:36] <rwaldron> DaveMethvin let me look for it and see how bit rotted it is
[09:17:40] <rwaldron> back in a few
[09:17:52] <DaveMethvin> ok if it seems fine just let me know and I can grab it and land it
[09:19:13] <DaveMethvin> so on 13789 any objections to that? Seems like an extra insurance policy but not really critical
[09:20:12] <m_gol> I'd be for landing it just in case
[09:20:48] <m_gol> at least until ES6 comes ;)
[09:20:58] <DaveMethvin> so, about 20 years :)
[09:21:02] <m_gol> :D
[09:21:35] <DaveMethvin> okay, so how about a ship date for 1.10 and 2.0.1?
[09:22:01] <DaveMethvin> i think they're both pretty much done unless someone has stuff that need to go in
[09:22:13] <m_gol> I'm interested in what other perf regressions could be there in http://jsperf.com/jquery-1-9-1-vs-2-0-0-beta/5
[09:22:22] <gibson042> did we get a "go" on the line-height fix?
[09:22:28] <m_gol> since http://jsperf.com/jquery-1-9-1-vs-2-0-0-beta/7 shows .data() isn't the most significant one, apparently
[09:22:29] <gibson042> (for 1.10)
[09:22:34] <DaveMethvin> yeah, that looked good to me gibson042
[09:23:18] <DaveMethvin> m_gol could you do some poking around there to see if you can figure out the perf regression?
[09:23:29] <m_gol> yup, will do
[09:23:35] <m_gol> just started ;)
[09:24:22] <DaveMethvin> i'll be traveling early next week, so for me I'd rather do the releases either late this week or late next
[09:24:35] <DaveMethvin> if we have everything wrapped up then this thursday would be fine with me
[09:24:45] <DaveMethvin> otherwise we could wait a week
[09:25:35] <m_gol> seems OK for me unless this 1.x/2.x comparison shows us a major perf regression that's not so easy to fix
[09:25:53] <m_gol> it'd be good to be able to send a clear message that 2.0.1 is faster than 1.10
[09:25:58] <DaveMethvin> yeah and we can decide later this week on that based on the results
[09:26:00] <m_gol> or not slower
[09:26:12] <m_gol> without introducing delaying code into 1.x :P
[09:26:15] <DaveMethvin> right, if they're basically equal that's fine
[09:26:18] <DaveMethvin> lol
[09:26:50] <DaveMethvin> "Ticket 12345, add more slow to 1.10 (assigned, m_gol)"
[09:27:50] <DaveMethvin> everything seems in good shape for now, anybody have anything else?
[09:28:22] <gibson042> I don't see anything else in Trac
[09:28:38] <DaveMethvin> yep, looks pretty clean
[09:29:00] <DaveMethvin> which reminds me, we should come up with a complete wish list for 1.11/2.1 soon
[09:29:23] <m_gol> is there any draft somewhere?
[09:29:26] <DaveMethvin> i don't know if i'll make the monday meeting  next week, but perhaps the following week we can talk about a few possiblities
[09:29:33] <DaveMethvin> none other than tickets
[09:29:44] <rwaldron> DaveMethvin https://github.com/jquery/jquery/pull/1268
[09:29:45] <DaveMethvin> i wanted to add the ability to let build leave out promise/callbacks
[09:29:52] <rwaldron> rebased
[09:30:00] <DaveMethvin> awesome thanks rwaldron
[09:30:16] <rwaldron> it could be even faster if you agreed to one thing
[09:30:21] <DaveMethvin> wazzat
[09:30:28] <rwaldron> a small addition to events
[09:31:10] <rwaldron> keep track of how many properties are in  `priv { events: {} }`
[09:31:29] <DaveMethvin> oh just a counter of event names in there?
[09:31:32] <rwaldron> yes
[09:31:36] <rwaldron> pretty minor
[09:31:40] <DaveMethvin> doesn't it disappear when it reaches 0?
[09:31:56] <rwaldron> naw, leave it at 0
[09:32:15] <rwaldron> It will save a call to Object.keys(), which i'm using to determine if a given path even needs to be taken
[09:32:38] <DaveMethvin> i'll take a look
[09:32:40] <rwaldron> there is _alot_ of conditional execution there now
[09:32:45] <gibson042> oh, I did find one: http://bugs.jquery.com/ticket/13835 ought to make 1.10 too
[09:32:57] <rwaldron> DaveMethvin I have to get back to moz/b2g
[09:33:03] <DaveMethvin> np thanks rwaldron
[09:33:06] <rwaldron> but the PR is there for you to land and tweak
[09:33:13] <gibson042> rwaldron mind if I grab the hasClass one?
[09:33:17] <rwaldron> that's the one that i got the most benefit out of
[09:33:34] <rwaldron> gibson042 please do
[09:33:35] <DaveMethvin> gibson042 that one is already on the agenda, agreed
[09:34:40] <DaveMethvin> ok awesome
[09:34:48] <DaveMethvin> so anything else?
[09:35:49] <DaveMethvin> jQuery 2.1/1.11 changes :Add tickets for issues you think should be addressed
[09:36:10] <DaveMethvin> we can discuss in two weeks
[09:36:17] <m_gol> k
[09:36:31] <DaveMethvin> alrighty then, lets move it over to -dev
[09:36:36] <DaveMethvin> thanks guys!