[09:01:26] <scott_gonzalez> Hey everyone!
[09:02:21] <rworth> hey
[09:03:16] <scott_gonzalez> rworth: Want to get us started with docs?
[09:03:16] <jzaefferer> hello
[09:03:46] <gnarf> howdy
[09:04:55] <rworth> sure
[09:05:18] <rworth> I need to take a look at effects docs
[09:05:27] <rworth> I guess some of them didn't make it over from mediawiki?
[09:05:40] <rworth> did I understand that ping the other day jzaefferer ?
[09:05:50] <jzaefferer> yeah, there's stuff missing
[09:06:22] <rworth> kswedberg is making good progress on api.jquery.com, we'll start to see some of those changes. And when the duplication is just too much (it almost is already) we'll figure out how to share code
[09:06:38] <rworth> I'll have the designers working on the api sites next week
[09:06:45] <rworth> really anxious to get them looking amazin
[09:06:57] <rworth> I think that's it from me atm
[09:07:56] <jzaefferer> at least we found a way to share templates between the api subsites
[09:08:08] <jzaefferer> even without the option to multi-inherit themes
[09:08:12] <rworth> yeah, I saw mention of that, that's great
[09:08:44] <rworth> I had looked into multi-inherit themes way back when we started on parent-child and saw it wouldn't work
[09:09:03] <jzaefferer> scott_gonzalez: where are we at with the MU install script? or resource redirects?
[09:09:07] <rworth> jzaefferer is there still housekeeping to be done in functions.php?
[09:09:39] <rworth> oh, forgot to mention, I got XIncludes working in api.jquery.com
[09:09:47] <scott_gonzalez> Supposedly it's essentially done, but nacin is never around for more than like 5 minutes at a time :-(
[09:09:49] <rworth> so that looks like a safe pattern to copy on other sites
[09:10:18] <rworth> it unfortunately requires two passes with xsltproc, but those aren't want takes so much time, it's the pygmentize that takes forever
[09:10:30] <gnarf> resource redirects?
[09:10:36] <scott_gonzalez> jzaefferer: I can work on an .htaccess redirect for resources today.
[09:10:46] <gnarf> scott_gonzalez: .htaccess wont work
[09:10:47] <rworth> jzaefferer where were you going to do xinclude next, I forget
[09:10:58] <rworth> ui-docs, right? for common stuff from widget factory?
[09:10:59] <scott_gonzalez> gnarf: We're going to double up on redirects.
[09:11:10] <nacin> sorry guys. today is release day. been swamped. will be around for at least 10 minutes at a time after it is out the door.
[09:11:10] <gnarf> scott_gonzalez: ?
[09:11:11] <scott_gonzalez> One in nginx for production and one in .htaccess for dev.
[09:11:16] <gnarf> aha
[09:11:21] <gnarf> i see now
[09:11:30] <gnarf> i use nginx for dev anyway
[09:11:33] <jzaefferer> rworth: widget-inherited methods/events/options
[09:11:44] <rworth> yeah, so take a look at notes in api.jquery.com
[09:11:45] <scott_gonzalez> gnarf: So you'll get to run the full production stack locally :-)
[09:11:51] <rworth> and let me know if you have any questions
[09:12:10] <scott_gonzalez> Requiring everyone to run nginx locally seemed excessive.
[09:12:15] <jzaefferer> gnarf: can you share the nginx config? I currently proxy Apache with nginx, which is rather stupid
[09:12:29] <scott_gonzalez> Especially since I'm assuming that the actual redirects will be setup through puppet.
[09:12:45] <scott_gonzalez> jzaefferer: nginx doesn't have the redirects yet.
[09:13:02] <gnarf> scott_gonzalez: if you can just write the block of nginx location statements
[09:13:07] <gnarf> i can include it in the vhost file
[09:13:10] <gnarf> and jzaefferer lemme find it
[09:13:37] <jzaefferer> scott_gonzalez: right, but even without redirects I could simplify my local setup quite a bit
[09:15:43] <gnarf> so - they moved my flight to another gate... semi-afk while i walk to it
[09:16:16] <scott_gonzalez> Ok, on to tabs...
[09:16:23] <scott_gonzalez> The ARIA updates finally landed.
[09:16:40] <scott_gonzalez> ARIA + keyboard handling is massive
[09:16:47] <scott_gonzalez> 700 bytes
[09:17:06] <scott_gonzalez> It also accounts for 1/3 of the tabs tests now.
[09:17:43] <jzaefferer> Oh well
[09:17:59] <scott_gonzalez> This morning I added grunt to jquery-color.
[09:18:03] <jzaefferer> Without ARIA and keyboard handling, you wouldn't need UI to begin with /o\
[09:18:19] <gnarf> thx scott_gonzalez - I'm gonna try to tackle this last modification I want to make to it on the plane today
[09:18:36] <scott_gonzalez> gnarf: cool, then release a beta?
[09:18:42] <scott_gonzalez> rc?
[09:19:38] <gnarf> Yes
[09:19:42] <rworth> yay
[09:19:52] <scott_gonzalez> We'll pull the release into UI.
[09:20:01] <scott_gonzalez> For now we'll just stick it right into effects core, like we've been doing.
[09:20:11] <gnarf> scott_gonzalez: sounds fair
[09:20:14] <scott_gonzalez> Eventually we'll just include it as a separate file.
[09:20:16] <gnarf> the "effects preamble" :)
[09:20:24] <scott_gonzalez> And it'll be in the download builder as it's own entry.
[09:20:36] <gnarf> cool
[09:20:52] <scott_gonzalez> We should be able to close out a bunch of UI tickets once that lands.
[09:21:06] <scott_gonzalez> And we'll pull in the stable release once that comes out (prior to 1.9).
[09:21:41] <gnarf> yup
[09:21:58] <jzaefferer> Sounds good
[09:22:06] <gnarf> i think color closes like 4-6 bugs in trac
[09:22:06] <jzaefferer> gnarf: did you see the swarm failures?
[09:22:27] <gnarf> jzaefferer: yesh - not really sure what the issue is there, trying to browserstack it but I really hate debugging tools in oldIE :)
[09:22:32] <jzaefferer> last build actually ran: http://swarm.jquery.org:8080/job/jQuery%20Color/22/console - http://swarm.jquery.org/job/7
[09:22:53] <gnarf> seems like its setting some css prop to "undefined" or something?
[09:22:59] <gnarf> more info soon
[09:23:00] <jzaefferer> animate: Setting CSS to empty string / inherit (1, 1, 2) - Died on test #2 undefined: Invalid property value.
[09:23:01] <jzaefferer> mhm
[09:23:19] <jzaefferer> at least both IE7 and IE6 have the same issue
[09:23:22] <gnarf> yeah
[09:23:22] <scott_gonzalez> Well, it's nice that Color is back in TestSwarm :-)
[09:23:39] <gnarf> IE7 debugger isn't much better than the IE6 one tho ;)
[09:23:46] <jzaefferer> Its also using the new rsync-script that Krinkle created (and I just improved)
[09:23:50] <gnarf> jzaefferer: out of curiosity
[09:23:57] <jzaefferer> so only one clone per job now, not two
[09:24:00] <gnarf> right
[09:24:03] <jzaefferer> should also speed up jobs
[09:24:12] <scott_gonzalez> Cool, does that mean we're closer to have -git builds working again?
[09:24:15] <gnarf> does that also mean we can submit a job from a local if we have the right keys in settings?
[09:24:40] <gnarf> not that its nessecary
[09:24:45] <jzaefferer> gnarf: not really, but you can clone a job and point it at your repo
[09:24:50] <gnarf> coo
[09:24:53] <jzaefferer> I did that earlier to test grunt-junit
[09:25:10] <gnarf> scott_gonzalez: yeah - we should be able to update the CDN via jenkins job now
[09:25:15] <jzaefferer> scott_gonzalez: should be, yeah, I don't yet know how the CDN deployment would look like
[09:25:25] <gnarf> jzaefferer: scp to code.origin.jquery.com
[09:25:32] <gnarf> and hit the "refresh" url
[09:25:36] <jzaefferer> can we just create another grunt task that uses keys installed on that server?
[09:25:39] <gnarf> danheberden can maybe help describe?
[09:25:50] <jzaefferer> gnarf: any existing scripts at hand?
[09:25:59] <jzaefferer> apart from the mobile one, which delegates to some php thingie...
[09:26:14] <gnarf> jzaefferer: not that I have or know of - danheberden has the deets
[09:26:32] <scott_gonzalez> I do it for releases.
[09:26:47] <scott_gonzalez> I can work with jzaefferer on it.
[09:27:25] <gnarf> okay - my only possible issue is giving keys to code.origin to jenkins, but I think we are pretty safe there...
[09:28:08] <gnarf> im not even sure the "new" build will need that
[09:28:16] <gnarf> as I think code.origin will be hosted on the same server as jenkins
[09:28:21] <scott_gonzalez> gnarf: It's just SSH keys, right?
[09:28:25] <gnarf> scott_gonzalez: aye
[09:28:35] <gnarf> scott_gonzalez: make a new pair and install em
[09:28:57] <scott_gonzalez> Can you get that setup and then jzaefferer and I will work on fixing -git.
[09:29:07] <jzaefferer> Sounds good :-)
[09:29:42] <jzaefferer> We can either do separate jobs like mobile has, or mix with regular builds to only deploy stable results
[09:29:51] <jzaefferer> The latter is what we really should aim for
[09:29:54] <scott_gonzalez> Only deploying stable sounds good.
[09:30:19] <gnarf> eh - maybe make a jquery-edge too then
[09:30:21] <gnarf> for the fails
[09:30:51] <gnarf> cuz i'd hate a unit failure not propagating -git and being confusing
[09:31:06] <gnarf> perhaps -git and -git-stable
[09:31:20] <scott_gonzalez> I guess it also depends on what stable means.
[09:31:26] <jzaefferer> yeah..
[09:31:42] <scott_gonzalez> Stable could mean that grunt finished successfully.
[09:31:58] <jzaefferer> I mean, we already have a lot less random errors, and I want that to get still a lot better, not worse. Making -git builds depend on that should add a little more motivation
[09:32:06] <jzaefferer> Even core is now getting close to a green build :-)
[09:32:27] <jzaefferer> yeah, we can also make the -git deploy before running testswarm, after verifying with phantom
[09:32:41] <jzaefferer> That's the sanity check and might be enough
[09:34:55] <scott_gonzalez> Ok, moving on...
[09:34:59] <scott_gonzalez> _bind() was renamed to _on()
[09:35:15] <gnarf> YAY!
[09:35:20] <scott_gonzalez> We're still working on the uuid support to deal with multiple instances binding to elements outside of the widget.
[09:35:33] <scott_gonzalez> And I'll continue to look into _off() at the same time.
[09:35:35] <gnarf> did you test the theory at all?
[09:36:01] <scott_gonzalez> Yup. I ran the theory though my head and all the tests passed :-P
[09:36:33] <rworth> than he checked his head into a gist
[09:37:05] <jzaefferer> btw. grunt-contrib has a zip task now, I'll give that a spin to get rid of our custom stuff
[09:37:08] <gnarf> scott_gonzalez: :)
[09:37:26] <gnarf> scott_gonzalez: myheadjs also passed the units
[09:37:31] <gnarf> so it must pass
[09:37:46] <jzaefferer> git push head master
[09:38:37] <scott_gonzalez> :-)
[09:39:02] <rworth> git reset --hard head
[09:39:36] <jzaefferer> scott_gonzalez: for uuid, for the most part it should be enough to also unbind from element without the id
[09:39:38] <scott_gonzalez> So this morning I realized that having tooltip track the mouse is a bit convoluted.
[09:39:52] <jzaefferer> not sure if that would miss anything...
[09:40:08] <scott_gonzalez> jzaefferer: Without the id? You mean no namespace?
[09:40:14] <gnarf> scott_gonzalez: a bit more info there?
[09:40:19] <scott_gonzalez> Or do you mean two unbinds?
[09:40:21] <jzaefferer> only the namepsace, but not the uuid
[09:40:22] <jzaefferer> two!
[09:40:24] <jzaefferer> my standalone plugin had a track option
[09:40:35] <scott_gonzalez> I think we should add an option.
[09:40:40] <scott_gonzalez> gnarf: Yeah, one sec.
[09:41:47] <scott_gonzalez> https://github.com/jquery/jquery-ui/commit/a47f5dc4f21762126ff460af9f49dffeb36e4086
[09:41:48] <socialhapy> ★ Commit on jquery-ui by Scott González (7h, 34m ago): Tooltip mouse tracking demo: Only track the mouse when the tooltip is opened on hover. Don't let the positioning of one tooltip affect other tooltips.
[09:42:08] <scott_gonzalez> You have to be careful to only respond to tooltip that open because of hover.
[09:42:17] <scott_gonzalez> And you have to be careful not to modify the underlying options.
[09:42:35] <scott_gonzalez> I'm guessing that pretty much everyone will get this wrong.
[09:42:43] <scott_gonzalez> Unless they copy the code straight from the demo.
[09:43:04] <gnarf> hrm
[09:43:35] <scott_gonzalez> You also have to manually mange the mousemove evenets.
[09:43:51] <scott_gonzalez> We left it out because we thought it was easy.
[09:43:58] <scott_gonzalez> Clearly it's cumbersome and tricky.
[09:44:02] <jzaefferer> yeah, apparently its not
[09:44:10] <jzaefferer> how do you handle the position-of property?
[09:44:21] <jzaefferer> always set it to the event when tracking is enabled?
[09:44:29] <gnarf> position of: can use "event" ? thats pretty awesome i never knew
[09:45:03] <scott_gonzalez> If tracking is enabled and the tooltip opened because of a mouse event, then we position against the event and setup a mousemove handler.
[09:45:33] <scott_gonzalez> So it gets tricky, because even with tracking turned on, we may not track.
[09:46:10] <gnarf> may not? meaning on "focus" tips?
[09:46:48] <scott_gonzalez> yeah
[09:46:59] <scott_gonzalez> Or from .tooltip( "open" )
[09:47:14] <jzaefferer> And if the user explicityl sets `of`, we'll leave it at that?
[09:47:31] <jzaefferer> -butchered word
[09:47:39] <jzaefferer> explicitly?
[09:47:54] <scott_gonzalez> jzaefferer: yeah
[09:48:04] <scott_gonzalez> hmm....
[09:48:12] <scott_gonzalez> not sure
[09:48:36] <scott_gonzalez> if you set `position.of` and `tracking` which is more important?
[09:48:39] <scott_gonzalez> Probably tracking.
[09:48:52] <scott_gonzalez> Assuming it's from mouse movement.
[09:49:36] <jzaefferer> so far we always use what the user sets
[09:49:46] <jzaefferer> but we have no idea why anyone would ever do that
[09:49:51] <scott_gonzalez> Yes, but so far we don't have trakcing :-P
[09:50:04] <jzaefferer> right
[09:52:11] <jzaefferer> so let's add a `track` option that will set `position.of` when its a mouse event
[09:52:18] <jzaefferer> just a boolean?
[09:52:23] <scott_gonzalez> Yeah, just a boolean.
[09:52:34] <scott_gonzalez> It does more than just set position.of, it also does the tracking :-P
[09:52:50] <scott_gonzalez> Which is calling .position() for every mousemove
[09:52:51] <jzaefferer> yeah :-)
[09:53:16] <scott_gonzalez> Ok, last thing on the agenda.
[09:53:34] <scott_gonzalez> We split the 1.10 release into 1.10 and 1.11.
[09:53:54] <scott_gonzalez> This will allow us to get a release that contains the updated dialog out before Drupal hits their code freeze for Drupal 8.
[09:54:20] <scott_gonzalez> The updated roadmap is available at http://wiki.jqueryui.com/Roadmap
[09:54:41] <gnarf> sounds good
[09:55:39] <gnarf> im gonna kill the laptop b4 i run out of juice for the plane
[09:55:43] <scott_gonzalez> Anyone have anything else?
[09:55:58] <kborchers> can we call Menu done?
[09:56:50] <scott_gonzalez> kborchers: I haven't had a chance to do an in-depth review, but I'm comfortable putting a beta prior to that.
[09:57:03] <scott_gonzalez> The goal is to have a beta out for the conf.
[09:57:30] <kborchers> ok ... also if there is anything i can help with for 1.9 let me know
[09:58:32] <scott_gonzalez> There are a few biggish changes that we should talk about.
[09:58:48] <scott_gonzalez> Like event name prefixes
[09:58:53] <scott_gonzalez> dialogopen -> dialog:open
[10:00:44] <kborchers> scott_gonzalez: was that to me or in general?
[10:00:51] <scott_gonzalez> both?
[10:00:55] <kborchers> :)
[10:01:13] <scott_gonzalez> Ok, my next meeting is starting.
[10:01:17] <scott_gonzalez> Let's continue in -dev.
[10:01:22] <kborchers> ok