[00:01:22] <flipzagging>	ok
[00:01:40] <flipzagging>	i'll try to make it over there... maybe will cab it
[00:05:24] <TrevorParscal>	we're thinking of going somewhere nearby
[00:05:34] <TrevorParscal>	and then heading to my house to play video games and hack
[00:05:40] <TrevorParscal>	my wife is out of town until later tonight
[00:05:49] <TrevorParscal>	like 10pm
[00:05:55] <TrevorParscal>	you are welcome to come as well
[00:06:03] <TrevorParscal>	anyone is
[00:06:13] <TrevorParscal>	well, except for tolls
[00:06:14] <TrevorParscal>	no troll
[00:09:52] <enhydra>	bastique, suddenly
[00:11:09] <bastique>	enhydra, great, would you test the registration form using your scholarship ID?
[00:11:27] <bastique>	Which I can provide to you via PM as well as the link
[00:18:51] <thedj>	"De nieuwste aanpassingen zijn zeer gemakkelijk en vormen een ware verbetering. Het aanpassen gaat nu beter, het is meer overzichtelijk. Ga zo door! Ckdewaal"
[00:18:56] <thedj>	wow positive feedback
[00:21:26] <thedj>	"The latest changes are very easy and are a true improvement. Making changes is much simpler now, it is more well ordered. Keep going! Ckdewaal
[00:24:03] <flipzagging>	ok i'm going to cab it over to the wmf
[00:25:44] <flipzagging>	but first, delicious pills
[00:26:12] <TrevorParscal>	we're watiting for ya
[00:26:40] <flipzagging>	i can only walk slowly, this might limit yr choices
[00:36:12] <TrevorParscal>	i'm not too worried
[00:36:21] <TrevorParscal>	there's plenty of good places around
[04:50:01] <Superjane>	hi
[16:03:55] <pdhanda>	morning guillom , how goes the usability study?
[16:04:41] <guillom>	hi pdhanda; it's going quite well so far, our first user is a very experienced wikipedian, as demanding and obsessive as they come :)
[16:05:58] <pdhanda>	haha.
[16:06:07] <RoanKattouw>	Wait, you're doing usability tests with experienced Wikipedians and they *still* fail? :O
[16:06:09] <pdhanda>	i'll head there with my laptop at around 10pm
[16:08:22] <guillom>	RoanKattouw: they don't necessarily fail, but they have old habits
[16:09:09] <guillom>	RoanKattouw: his main criticism is about the search box, it's all your fault!
[16:09:13] <guillom>	your = Stanton
[16:09:16] <guillom>	:)
[16:09:43] <RoanKattouw>	hah
[16:12:34] <pdhanda>	guillom:  what is the address again, I'm sure you sent it in the email but I'm email-stupid today :/
[16:12:51] <{cary|working}>	pdhanda, cool, I'll go with you.
[16:12:59] <RoanKattouw>	250 Sutter, I think
[16:13:37] <guillom>	pdhanda: Fleischman Field Research, 250 Sutter street, 2nd floor
[16:13:41] <RoanKattouw>	Coming from Montgomery it'll be on the right-hand site of the street
[16:14:57] <guillom>	when you get there just tell them you're with the WMF
[16:15:09] <guillom>	the room is the one on the right when you get out of the elevator
[16:19:59] <pdhanda>	sweet! thanks
[16:46:11] <Otourly>	{cary|working}: Silly question: why it is this logo used on this page : http://www.wiktionary.org/ and no wiktionnaries use it ?
[16:48:16] <{cary|working}>	Otourly, the one the meta community voted on.
[16:49:50] *Otourly 	ask himself when we will really see this logo on a project
[16:50:24] *Otourly 	should try to make an SVG
[16:50:29] <Otourly>	a'
[16:50:52] <{cary|working}>	Otourly, I don't have time to say, "look, this is what the general Wiktionary community voted for.  You never voted for the old logo, and more of your members like the new one than the old one.  You need to stop impeding progress."
[16:51:58] <Otourly>	{cary|working}: I know the result of the meta vote, but this logo could be better
[16:52:39] <{cary|working}>	Otourly, indeed.
[16:52:43] <{cary|working}>	I don't much like it.
[16:59:28] <guillom>	participant looking at the Contributions page: "uuh, this is ugly; but wikipedia pages like this usually are"
[17:01:15] <pdhanda>	ok heading over now...
[17:01:25] <TrevorParscal>	is it cool if I roll? is there space?
[17:01:28] <TrevorParscal>	RoanKattouw: hi
[17:01:30] <RoanKattouw>	TrevorParscal: I'm here now
[17:01:33] <RoanKattouw>	Good morning
[17:01:38] <TrevorParscal>	hi
[17:03:01] <TrevorParscal>	ok
[17:03:24] <TrevorParscal>	so, we should probably identify parts of the system that aren't specified yet in the document
[17:04:03] <TrevorParscal>	maybe reorg the doc a bit
[17:04:10] <TrevorParscal>	but, it's not too critical that the doc is amazing
[17:04:27] <guillom>	TrevorParscal: feel free to drop by, there's plenty of space
[17:04:30] <guillom>	and, we have wifi
[17:04:45] <TrevorParscal>	mmmm
[17:04:52] <TrevorParscal>	maybe I will
[17:04:55] <TrevorParscal>	RoanKattouw: brb
[17:05:00] <TrevorParscal>	chaning locations
[17:19:51] <TrevorParscal>	RoanKattouw: ok
[17:19:56] <TrevorParscal>	i'm at the research facility now
[17:20:04] <TrevorParscal>	but we can continue our work
[17:20:08] <TrevorParscal>	I'm just observing for fun
[17:20:35] <RoanKattouw>	Yeah
[17:21:13] <TrevorParscal>	OK, so we need to figure out how CSS files get included
[17:21:29] <RoanKattouw>	Yes
[17:21:41] <RoanKattouw>	I wonder whether it's possible to do that dynamically in a reliable fashion
[17:22:04] <RoanKattouw>	I've tried adding <link> elements to the <head> in IE but that just doesn't seem to work
[17:22:06] <TrevorParscal>	the whole $( 'head' ).append( $( '<style>body { background-color: red; }</style>' ) );
[17:22:08] <RoanKattouw>	May have been just IE 7
[17:22:18] <TrevorParscal>	we should try that in diff browsers..
[17:22:18] *TrevorParscal 	checks
[17:22:38] <TrevorParscal>	we suspect IE 6 and 7 won't cope with it...
[17:22:39] <TrevorParscal>	right?
[17:23:53] <RoanKattouw>	Yes
[17:24:09] <TrevorParscal>	testing now
[17:25:26] <TrevorParscal>	worked in IE6
[17:26:02] <RoanKattouw>	:O
[17:26:06] <RoanKattouw>	Well yeah
[17:26:09] <RoanKattouw>	You're using hardcoded <style>
[17:26:14] <RoanKattouw>	What about <link href=""> ?
[17:26:38] <TrevorParscal>	IE7 is OK
[17:26:45] <{cary|working}>	no laughing
[17:26:45] <TrevorParscal>	hmm
[17:26:50] <{cary|working}>	NO LAUGHING
[17:26:51] <TrevorParscal>	i will test that too
[17:27:03] <guillom>	{cary|working}: I prefer to be laughing than crying
[17:29:58] <TrevorParscal>	worked in IE7
[17:30:40] <RoanKattouw>	Well
[17:30:49] <RoanKattouw>	How do you test whether it worked?
[17:31:19] <TrevorParscal>	worked in IE6
[17:31:25] <TrevorParscal>	will show you
[17:34:13] <guillom>	jorm: see http://www.gotomedia.com/wikimedia/ for the links
[17:38:04] <TrevorParscal>	RoanKattouw: hmm
[17:38:13] <TrevorParscal>	it was working, but Im messing with it and it's not...
[17:38:19] *TrevorParscal 	is discovering
[17:38:57] <RoanKattouw>	I was gonna say
[17:39:03] <TrevorParscal>	it was working
[17:39:04] <RoanKattouw>	I'm sure this failed on me in IE7 at least
[17:39:17] <RoanKattouw>	I mean the element will show up in the <head> just fine
[17:39:22] <TrevorParscal>	but there must have been some strange combo of conditions
[17:39:23] <RoanKattouw>	But the styles just won't get applied
[17:43:39] <TrevorParscal>	ok, my example is flawed, even FF won't apply the styles
[17:48:27] <mdale>	RoanKattouw:  you have to use a special entry method
[17:48:46] <mdale>	see addStyleSheet in mwEmbed.js
[17:49:25] <mdale>	I mean addStyleString
[17:49:56] <mdale>	works across all the browsers that I have tested .. I go down to ie6
[17:51:03] <RoanKattouw>	Does it add a <style> tag with content instead of a <link> tag?
[17:51:30] <mdale>	yep.
[17:51:43] <mdale>	All the tricks for transformations and contextual absolutting of url if resource loader is on a different domain fixing relative paths etc.  exist within the existing implementation.
[17:52:46] <mdale>	( this is how all the interface components are delivered in a single pay load ) ... this is of course only for dynamically loaded modules.. the normal css is put in buckets so it works without js.
[17:54:15] <mdale>	and it gets mapped to the raw files in raw file mode ( of course )
[17:55:08] <mdale>	... I get the feeling people have not looked at what is there
[17:55:48] <RoanKattouw>	It's true that I haven't looked at JS@
[17:55:48] <RoanKattouw>	*JS2
[17:55:50] <mdale>	or that they are unaware that we tested these things across many browsers etc.
[17:56:12] <RoanKattouw>	Thing is, JS2 is a rather large body of code
[17:56:15] <mdale>	yea
[17:56:20] <mdale>	it does a lot of things
[17:56:34] <RoanKattouw>	So instead of diving into a large body of code I don't know, I prefer to just ask the guy who wrote it
[17:56:43] <mdale>	I am happy to help people navigate it...
[18:00:04] <mdale>	From what I gather from the current spec is developing ... The only potential major architectural  difference is direct resource name loading vs  php dependency mapping .. most other differences are naming or api conventions.
[18:00:26] <RoanKattouw>	We don't intend to map dependencies in PHP
[18:00:31] <mdale>	all the concepts seem more or less aligned.
[18:02:01] <pdhanda>	seems like the one at 12:45 will be interesting (non tech user)
[18:02:15] <pdhanda>	^ was about the UX test
[18:02:17] <mdale>	 It would take a day or two to rename things to "resource" instead script / class... and then another day or two to move the resource lists into php... And you could arbitrary chose the api ... ie its really only a few lines to change if you want to use $wgOut->addNamedResource( $resourceName, array( 'path'=>{path} ) .. instead of $wgNamedResources['resourceName'] = {resourcePath} ....
[18:02:40] <mdale>	or what have you
[18:11:56] <mdale>	the main advantage of doing the rename and api update is that it will keep all the mwEmbed modules and the UploadWizard working.... instead of a mismatch of features that need to be slowly ported over... but its of course I understand the desire to do-a-new ... but I worry when I see you guys circling around things wholly unaware that the issue has been addressed, tested and used for some time in the existing system. ... you should at least be
[18:12:41] <RoanKattouw>	Well right now we plan on first spec'ing this thing out, then looking at what's out there to see if we can use it
[18:12:46] <mdale>	oky
[18:13:17] <RoanKattouw>	So we may borrow code or interfaces or whatever from JS2 but that's the next stage
[18:19:02] <{cary|working}>	jorm, did you see a little light?
[18:19:16] <jorm>	i did.  that's what i was looking back at.
[18:19:28] <jorm>	it's reflecting in the glass.
[18:19:32] <{cary|working}>	i'm doing archive footage
[18:27:48] <RoanKattouw>	TrevorParscal, howief: I'm introducing a preference for SimpleSearch. Does "Enable enhanced search suggestions (Vector skin only)" sound good?
[18:28:06] <TrevorParscal>	thank you
[18:28:11] <TrevorParscal>	we've been meaning to do that
[18:28:18] <TrevorParscal>	also, please put the PS3 fix out there
[18:28:25] <jorm>	"The media wizard fills me with dread."
[18:28:33] <TrevorParscal>	if not already deployed
[18:28:47] <howief>	RoanKattouw: which section of preferences will this appear?
[18:29:00] <RoanKattouw>	Search
[18:29:07] <RoanKattouw>	Below "Disable AJAX suggestions"
[18:29:30] <RoanKattouw>	TrevorParscal: I think that may have been deployed already. You can check that in CodeReview
[18:29:37] <TrevorParscal>	k
[18:29:40] <RoanKattouw>	Go to the revision and see if it has an MFT rev linked to it
[18:30:34] <TrevorParscal>	67576 and 67577
[18:31:06] <TrevorParscal>	this user is afraid of the word "Wizard"
[18:31:20] <TrevorParscal>	she thinks some animated wizard might pop up and never go away
[18:31:23] <TrevorParscal>	this is great
[18:31:58] <RoanKattouw>	Like the Microsoft Word animated clip? :P
[18:34:11] <{cary|working}>	yes, just like that
[18:34:22] <RoanKattouw>	howief: So does that text sound good?
[18:34:37] <howief>	where in preferences is this going?
[18:34:42] <jorm>	exactly like that.  like clippy.
[18:35:57] <RoanKattouw>	howief: In the search section, directly beneath the "Disable AJAX suggestions" checkbxo
[18:36:04] <RoanKattouw>	Checking that checkbox will also disable SimpleSearch BTW
[18:36:19] <TrevorParscal>	maybe it should be checked by default
[18:36:24] <TrevorParscal>	and be Enable *
[18:36:31] <TrevorParscal>	howief: thoughts?
[18:36:58] <howief>	the wording could be a little more clear
[18:37:15] <howief>	i don't think "enable enhanced search suggestions" means anything to anyone other than us
[18:37:21] <howief>	but i can't think of anything better right now
[18:37:26] <howief>	i think people understand "search suggestions"
[18:37:32] <howief>	but we have that in monobook as well
[18:38:14] <howief>	we can probably come up with something better, but it'll take time
[18:38:21] <howief>	so let's just use what roan proposed
[18:38:47] <RoanKattouw>	It says (Vector skin only)
[18:38:59] <howief>	yes
[18:39:06] <RoanKattouw>	So people wanting to disable this thing will probably get it
[18:39:36] <howief>	i was considering using just the term "search suggestions" (removing "enhanced") for the feature, but since monobook has search suggestions, that doesn't work
[18:39:39] <howief>	true
[18:39:49] <howief>	so a minor nit
[18:40:14] <howief>	the first three items on "display options" (hits/lines/context) all apply to the search results page
[18:40:52] <howief>	and "disable ajax/enhanced search suggestion" apply to the type-ahead behavior, which in many cases helps users avoid having to go the search results page
[18:41:19] <RoanKattouw>	Yes
[18:41:24] <RoanKattouw>	We may want to restructure that
[18:41:27] <howief>	yeah
[18:41:29] <RoanKattouw>	But I'd like to not do that right now
[18:41:41] <howief>	conceptually, they can be bucketed in two distinct categories
[18:41:47] <howief>	sure
[18:41:53] <howief>	just wanted to point that out
[18:42:09] <howief>	but yeah, let's go ahead with what you have
[19:53:56] <{cary|working}>	me hungry
[20:01:58] <TrevorParscal>	{cary|working}: did you forget the / or are you just talking like the abominable snow man for fun?
[20:02:12] <{cary|working}>	the latter
[20:02:21] <TrevorParscal>	awesome!
[20:02:43] <{cary|working}>	ME HONGRY
[20:04:20] <mdale>	man its raining hard here... I hope the Internet does not go out...
[20:04:48] <{cary|working}>	are you in mexico again
[20:04:53] <mdale>	...yes
[20:24:53] <RoanKattouw>	Hey mdale can I ask you another question about JS2?
[20:25:02] <mdale>	Yea
[20:25:14] <RoanKattouw>	We're talking about using the use-the-highest-mtime-as-style-version approach that JS2 apparently uses
[20:25:32] <RoanKattouw>	But these mtimes are known on the server, but need to somehow be known in JS as well in order to construct the right URLs
[20:25:35] <RoanKattouw>	How is that handled?
[20:26:05] <mdale>	Its generated at page render time right now
[20:26:57] <RoanKattouw>	OK so how does that work exactly? Do you just generate URLs like loader.php?giveme=stuffIWant&ts=12345 on the server?
[20:27:04] <RoanKattouw>	How does that work for loading dynamic stuff
[20:27:13] <mdale>	right...it does not work so well with dynamic stuff presently
[20:27:29] <mdale>	but in practice I just purge my cache folder on svn updates
[20:27:38] <mdale>	but it should be smarter than that
[20:27:49] <RoanKattouw>	OK
[20:28:31] <mdale>	like every hour or so check the filemodinfed time for any resources it severing or if you get a must-revalidated header then it should probably check file modified times there too.
[20:28:52] <mdale>	in fact we can probably mostly shift away from unique request id's if the system is smart about dealing out 304's
[20:29:04] <RoanKattouw>	Yeah I was thinking that
[20:29:12] <RoanKattouw>	Maybe we don't need style versions if there's 304s around anyway
[20:29:19] <RoanKattouw>	Another approach I'm thinking about
[20:29:22] <mdale>	and if our entry point runs at .0005ms that does not hurt ;)
[20:29:31] <RoanKattouw>	Is to just calculate the maximum mtime of absolutely everything out there
[20:29:36] <RoanKattouw>	And put that in a JS var
[20:29:54] <mdale>	yea.. that is kind of costly for every page render
[20:30:23] <mdale>	On svn update setups you could read the svn folder info
[20:30:45] <RoanKattouw>	Yeah it is a motherload of mtime reads
[20:31:11] <RoanKattouw>	How do you do that?
[20:31:24] <RoanKattouw>	Also a number of mtime() calls on page render?
[20:32:21] <mdale>	the "optimal" setup is you have single dynamic "manifest" that you must re-validate every hour or so... and it has the file-time of everything and you set expire forever so clients just get files once ... long initial setup but fast forever after that ;)
[20:32:32] <RoanKattouw>	That is smart, yes
[20:33:14] <mdale>	that is how some of the mobile html5 stuff works but the manifest one-ups that and just gives you diff locations.
[20:33:34] <mdale>	and supports combined packages
[20:33:41] <mdale>	but we can leave that for js3 ;)
[20:33:49] <RoanKattouw>	Wait so the manifest is something clients retrieve?
[20:33:53] <mdale>	yea
[20:33:57] <RoanKattouw>	I was thinking of it as an internal file regenerated by a cronjob
[20:34:13] <mdale>	nah.. because your page could be cached while you need to update a given script
[20:34:24] <RoanKattouw>	Right yeah
[20:34:40] <RoanKattouw>	OTOH that is Yet Another URL To Load
[20:34:45] <RoanKattouw>	And we're all about reducing those
[20:34:56] <mdale>	ResourceLoader->getURIDparam() is what your looking for
[20:35:04] <RoanKattouw>	Besides we plan on expiring everything in an hour anyway
[20:35:19] <RoanKattouw>	In what dir+file?
[20:35:25] <mdale>	... OTOOH if you had such a system you would only have to 304 a single file instead of a bunch of different stuff
[20:36:17] <RoanKattouw>	Well OTOH
[20:36:23] <RoanKattouw>	If everything expires after an hour
[20:36:28] <RoanKattouw>	Do we even /need/ style versions?
[20:37:14] <RoanKattouw>	This is something I'd like to pick Tim's brain about as well when he gets up
[20:37:38] <mdale>	( we would not have to expire things after an hour if you had such a system )
[20:38:22] <mdale>	because the state of every resource would have a unique URL identifier for all time.
[20:40:45] <RoanKattouw>	Well I'd like to not have a URL per resource , but combine stuff
[20:41:00] <RoanKattouw>	Expiring stuff in an hour is cheap with Squid in front
[20:42:24] <RoanKattouw>	The advantage of a short expiry is you're pushing JS updates fast
[20:42:44] <RoanKattouw>	And Squid is very smart about 304s
[20:43:19] <RoanKattouw>	You can tell Squid to cache a 304 response for 10 minutes or something
[20:43:43] <RoanKattouw>	Then if someone else hits the same URL within 10 minutes, they'll get the cached 304 and the script loader isn't even fired up
[20:44:07] <RoanKattouw>	Could expire those per hour too but then it could theoretically take 2 hours before a client gets the new vesion
[20:44:08] <mdale>	right... the other system is only to minimize 304 requests or minimize total request in general...
[20:44:21] <RoanKattouw>	Indeed
[20:44:42] <mdale>	But you sacrifice that in initial start up time.
[20:44:45] <RoanKattouw>	So it's the tradeoff of minimizing requests vs. easy updates
[20:45:11] <RoanKattouw>	Easy updates means more requests, but minimizing requests means going through style version hell to update stuff and even then it won't work properly
[20:45:51] <mdale>	right... having squid giving us practically free 304s makes it easy to lean that direction
[20:46:04] <RoanKattouw>	Yes
[20:46:10] <RoanKattouw>	Of course that does result in more bandwidhh
[20:46:17] <RoanKattouw>	usage
[20:46:36] <mdale>	right but 304 requests are virtually free
[20:46:47] <RoanKattouw>	304s are lightweight, only a few hundred bytes each way
[20:47:02] <RoanKattouw>	But millions upon millions of people are gonna fire those every hour
[20:48:56] <RoanKattouw>	OTOH we're hopefully saving bandwidth with aggressive combining and minifying
[20:49:02] <RoanKattouw>	Hey Trevor
[20:49:06] <mdale>	yea... you could greatly reduce the number of 304s by having a resource manifest file that lists the versions the top 50 or so resource-loader grouped requests.
[20:49:12] <TrevorParscal>	hi
[20:49:21] <mdale>	Hi TrevorParscal
[20:49:30] <RoanKattouw>	mdale: ... and increase the expiry time you mean?
[20:49:35] <TrevorParscal>	we still talking versions?
[20:49:39] <RoanKattouw>	TrevorParscal: So I'm still split how to handle versioning
[20:49:45] <RoanKattouw>	And I really wanna pick Tim's brain about this
[20:49:47] <mdale>	yea you could have expire time of a month
[20:50:05] <TrevorParscal>	seems like version numbers are great for "we just pushed out something major"
[20:50:16] <mdale>	and an expire time of 10 minutes on your resource manifest and more efficiently more quickly push out updates.
[20:50:20] <RoanKattouw>	Because he understands the ramifications in terms of bandwidth usage etc. better than us I think
[20:50:24] <TrevorParscal>	and no version with short exp. is fine (and much simpler) for we patched some tiny bug
[20:50:41] <RoanKattouw>	Yeah so the downside of short exps is we're generating lots of 304s
[20:50:50] <RoanKattouw>	Which is not hard on the servers because of Squid and stuff
[20:50:55] <RoanKattouw>	But those things do use bandwidth
[20:51:12] <RoanKattouw>	Small 304 response * millions of users * every hour = probably quite a bit of bandwidth
[20:51:13] <TrevorParscal>	seems like the amount of moving parts to do version numbers is something that needs a lot of justification
[20:51:17] <mdale>	but it does delay page render time by adding round trips
[20:51:20] <TrevorParscal>	it's hardly the path of least resistance
[20:51:24] <RoanKattouw>	Yeah so
[20:51:43] <RoanKattouw>	From an ease of use point of view, short expiries definitely win over style versions
[20:52:35] <RoanKattouw>	But I want to have Tim around because he has a deep understanding of any negative impact on bandwidth usage and the like
[20:53:08] <TrevorParscal>	where is he at?
[20:53:37] <RoanKattouw>	Asleep
[20:53:41] <RoanKattouw>	I guess
[20:53:43] <RoanKattouw>	It's 6:54
[20:53:45] <TrevorParscal>	right
[20:53:58] <TrevorParscal>	he needs to wake up already - gosh! :))
[20:55:27] <mdale>	Its a lot of complexity for little gain. ...if you have to check the 304 for 4 or 5 (grouped resources) .. instead of a 304 for a single resource manifest.. its not that much different for most moder browsers that parallelize requests.
[20:55:48] <mdale>	modern browsers I mean
[20:55:54] <RoanKattouw>	It doesn't actually add complexity
[20:55:57] <RoanKattouw>	You don't check a 304
[20:56:09] <RoanKattouw>	The JS just fires an AJAX request
[20:56:10] <mdale>	I mean the resource manifest adds complexity
[20:56:15] <RoanKattouw>	Ah right
[20:56:17] <RoanKattouw>	Yes
[20:56:23] <TrevorParscal>	and why would you be checking for all sort of different resources individually?
[20:56:27] <RoanKattouw>	304s are easier because the browser handles them transparently
[20:56:37] <RoanKattouw>	He's talking about groups of resources I think
[20:56:43] <TrevorParscal>	RoanKattouw: pause the convo
[20:56:47] <mdale>	I am talking about groups of resources
[20:57:02] <TrevorParscal>	what's the state of blacklisting IE for CollapsibleNav
[20:57:08] <TrevorParscal>	on the cluster
[20:57:12] <TrevorParscal>	erik says it's not deployed
[20:57:13] <RoanKattouw>	Ah lemme see
[20:57:15] <RoanKattouw>	Is that in any revision?
[20:57:21] <TrevorParscal>	let me find it
[20:57:41] <RoanKattouw>	It says IE >= 7 now
[20:57:49] <TrevorParscal>	hmm
[20:57:52] <TrevorParscal>	that should be working
[20:58:07] <TrevorParscal>	let me consult with erik quickly
[20:58:10] <TrevorParscal>	continue convo
[20:58:24] <RoanKattouw>	So yes 304s are totally the more convenient way
[20:58:40] <RoanKattouw>	But I want to consult with Tim to see how rational my bandwidth fears are
[21:00:14] <TrevorParscal>	pause
[21:00:25] <TrevorParscal>	RoanKattouw: erik's comp is accessing bits.wikimedia.org
[21:00:36] <TrevorParscal>	and getting a version of collapsiblenav without the IE blacklist
[21:00:42] <TrevorParscal>	I saw it myself in the code
[21:00:46] <TrevorParscal>	ideas?
[21:00:55] <RoanKattouw>	Lemme see if it's even in deployment
[21:00:59] <TrevorParscal>	k
[21:01:28] <RoanKattouw>	Ah there we go
[21:01:31] <RoanKattouw>	It was never deployed
[21:01:35] <TrevorParscal>	ok
[21:01:40] <TrevorParscal>	yes
[21:01:42] <TrevorParscal>	deploy that now
[21:01:44] <TrevorParscal>	please
[21:02:11] <TrevorParscal>	what's the bug in IE 6 anyways? it seems to work..
[21:02:37] <RoanKattouw>	Oh there we go
[21:02:40] <RoanKattouw>	Way to go Trevor
[21:02:53] <RoanKattouw>	You refactored all that browser checking stuff and snuck in the msie check in that rev
[21:02:59] <TrevorParscal>	ha ha
[21:03:01] <TrevorParscal>	sorry
[21:03:13] <RoanKattouw>	So that explains how I missed it in my inventory of stuff to deploy
[21:03:17] <RoanKattouw>	I'll just hack it into deployment
[21:03:18] <TrevorParscal>	surely
[21:03:21] <TrevorParscal>	thanks!
[21:03:31] <RoanKattouw>	Is it really broken in IE6 though?
[21:05:10] <TrevorParscal>	there must have been a reason we blocked it
[21:05:24] <RoanKattouw>	OK I'll just do it then
[21:05:26] *TrevorParscal 	can't recall
[21:05:41] <TrevorParscal>	if nothing else, we can blacklist it to punish IE6 users!
[21:05:43] <TrevorParscal>	mu ha ha ha
[21:06:20] <TrevorParscal>	i think there's a bug for it
[21:06:29] <TrevorParscal>	there's something going on there...
[21:08:30] *TrevorParscal 	begins working on chart to show how resource loader works
[21:08:47] <RoanKattouw>	ooh nice
[21:09:04] <RoanKattouw>	I think we've pretty much covered everything except the versioning strategy, right>
[21:09:35] <RoanKattouw>	TrevorParscal: IE blacklisting deployed
[21:09:44] <TrevorParscal>	thanks
[21:09:50] <TrevorParscal>	yes
[21:09:54] <RoanKattouw>	Due to Squid... yadda yadda yadda
[21:10:07] <TrevorParscal>	so, versioning needs more work, and consultation with Tim
[21:10:15] <TrevorParscal>	can you please do the communication with him
[21:11:06] <RoanKattouw>	Yes of coures
[21:11:14] <RoanKattouw>	You work on your charts
[21:11:18] <RoanKattouw>	We both do what we do best
[21:15:13] <TrevorParscal>	:)
[21:19:35] <EcceNux>	There is a syntax error in http://bits.wikimedia.org/w/extensions/UsabilityInitiative/Vector/Vector.combined.min.js?282fs
[21:19:47] <EcceNux>	'msie':[['>='.7]]
[21:19:52] <EcceNux>	'msie':[['>=',.7]]
[21:20:01] <EcceNux>	coma is missing
[21:20:17] <RoanKattouw>	Ouch
[21:20:19] <RoanKattouw>	Thanks
[21:22:13] <RoanKattouw>	EcceNux: OK should be fixed now, thanks for reporting
[21:24:45] <EcceNux>	No problem, and thanks. I thought I messed up something :-).
[21:54:03] <bastique>	howief,
[21:54:19] <bastique>	which column to choose from?  <http://translatewiki.net/wiki/User:GerardM/usability-stats>
[21:55:32] <howief>	bastique: i think the best one would be the last one - wikieditor
[21:55:41] <howief>	that's prob the one we're going to be basing the 80% decision on
[21:55:51] <bastique>	okay
[21:56:03] <howief>	let me know if the mail is confusing
[21:56:14] <howief>	there's a lot of info to communicate
[21:56:33] <bastique>	I'm getting clarification here
[21:56:40] <bastique>	I'm rewriting it where necessary.
[21:56:45] <howief>	great
[21:56:46] <RoanKattouw>	All those groups are important
[21:56:50] <RoanKattouw>	I can explain what each group means
[21:57:03] <bastique>	RoanKattouw, don't you go complicataming things
[21:57:06] <RoanKattouw>	But they basically correspond to different features or sets of features
[21:57:24] <howief>	which ones are absolutely necessary for a reasonable experience
[21:57:38] <bastique>	I'm surprised at Danish
[21:57:44] <howief>	i know those are some pretty ambiguous terms
[21:57:57] <howief>	but i don't think we need everything to be at 80%
[22:02:11] <RoanKattouw>	Well OptIn link and EditWarning are rather essential
[22:02:21] <RoanKattouw>	Ah
[22:02:26] <RoanKattouw>	Actually, OptIn link is irrelevant
[22:02:38] <RoanKattouw>	It's PrefSwitch that's important but that one's missing from this table
[22:05:08] <howief>	is it buried within any of the other groupings?
[22:08:56] *kibble 	points out that "Usability Initiative" is the root group.
[22:09:08] <kibble>	So if they have 100% there, they'll have 100% everywhere else.
[22:09:45] <howief>	kibble: how would a 75% in "uability iniative" be calculated/
[22:10:36] <kibble>	Hmm?
[22:10:50] <kibble>	It looks at the total number of untranslated messages divided by the total number of messages.
[22:10:59] <kibble>	http://translatewiki.net/wiki/Special:Translate
[22:11:17] <kibble>	Search for the "usability initiative" part.
[22:13:18] <^demon>	You know this would all be a lot easier if we didn't have message files and just hardcoded everything. Faster too ;-)
[22:14:27] <kibble>	Yeah, it would be soo much easier if everyone spoke English.
[22:14:47] <^demon>	Not even that. Whoever's committing can write that part in their native language.
[22:15:26] <^demon>	MediaWiki will become really multilingual and you'll have to learn multiple languages to use it :)
[22:17:37] <kibble>	howief, you know, it would help if you guys communicated that we were gearing up for Phase IV. :-)
[22:20:25] <howief>	kibble: this is us communicating :)
[22:20:37] <howief>	no seriously - this is just something we discussed this past tuesday
[22:21:06] <howief>	but we do have to figure out a way to get the general message out
[22:21:58] <^demon>	Phase IV?
[22:22:08] <howief>	yes, phase iv
[22:22:11] <kibble>	[[usability:Releases/Default Switch]].
[22:23:05] <^demon>	Oh, ah ok.
[22:23:14] <^demon>	I was thinking /trunk/phase3/.
[22:23:19] <^demon>	And I was like wait WHAT?
[22:23:38] <^demon>	 /trunk/phase4/?
[22:26:51] <kibble>	Yeah, sorry, we have issues communicating, as you can see. ;-)
[22:27:56] <kibble>	howief, I think that's where the disconnect is.  Communicating isn't just about saying what you're going to do, it's about talking about involving people in the discussions that could have valuable feedback.
[22:28:31] <howief>	yes
[22:28:36] <kibble>	Like just cloning a CentralNotice isn't the best idea, because then you have a bunch of translations left over that you aren't going to use.  And trying to decipher a status template can be tough, so asking for help is good. :-)
[22:29:13] <howief>	this next phase is particularly challenging
[22:29:30] <howief>	and to be honest i don't think we've figured out the best way to get the message out/get people involved
[22:29:35] <howief>	with the previous releases
[22:29:42] <howief>	there were only a handful of projects
[22:30:02] <howief>	but this time, there are a lot more
[22:30:33] <kibble>	Indeed, which is why I'm sure you guys could use help planning. :-)
[22:30:44] <howief>	are you voluteering?
[22:31:33] <kibble>	I didn't think I had to explicitly "volunteer", to be honest. :P
[22:32:04] <howief>	:)
[22:32:09] <howief>	well we could definitely use the help
[22:32:24] <kibble>	But yes, I helped coordinate the translation for Phase III, and Az1568_ and I are happy to help continue with Phase IV.
[22:32:52] <Az1568_>	Yep, we are.
[22:33:00] <howief>	great!
[22:35:22] *^demon 	appoints kibble to be the UX team's relationship manager.
[22:36:58] <kibble>	howief, could you start by telling us what you're planning?  I think we know already, but it's hard to know if we know everything or not.
[22:38:01] <howief>	(sorry - was talking to trevor about some translation issues)
[22:38:34] <howief>	so the current plan for phase iv is to switch all wikipedias above a certain translation threshold by end of june
[22:39:41] <howief>	then phase v would be to switch all projects that don't have any blocking issues (e.g., extensions breaking)
[22:39:57] <howief>	phase v would happen at the end of july
[22:40:18] <^demon>	MediaWiki 1.17 will ship with vector as default :)
[22:40:42] <kibble>	howief, do you know the threshold yet?
[22:40:57] <kibble>	80% for all messages wouldn't be tough.
[22:41:28] <howief>	we're thinking either 80% of all messages or 80% of wikieditor messages
[22:42:16] <howief>	from looking at the translatewiki report, if a language is 80% of wikieditor, it's probably at at least that level for the other messages as well
[22:42:35] <howief>	http://translatewiki.net/wiki/User:GerardM/usability-stats
[22:43:40] <howief>	to help with the translation effort
[22:43:57] <howief>	we were planning on running notices on the various projects, asking for help with the translation
[22:44:07] <howief>	which would direct translators to translatewiki
[22:44:26] <kibble>	Directing them to the landing page? http://meta.wikimedia.org/wiki/2010_Wikimedia_design_and_feature_change
[22:44:28] <howief>	the idea here is to get as many projects above the threshold as possible
[22:44:50] <howief>	yes, and form there, they would be directed to the translatewiki
[22:45:01] <kibble>	Yeah, I already cleaned up that page a bit.
[22:45:13] <howief>	awesome
[22:45:44] <kibble>	Do you have a target date yet?
[22:45:56] <howief>	and as you can see from the page, we're also asking for help with identifying bugs and for their general feedback
[22:46:01] <howief>	the target date hasn't been set
[22:46:03] <howief>	but for now
[22:46:17] <howief>	let's just assume end of June, say June 30
[22:46:43] <howief>	i'm not sure how aggressive it is, but we can use that date until our efforts tell us that it's unreasonable
[22:47:15] <howief>	there hasn't been a discussion of whether to use ambassadors
[22:47:40] <howief>	i'm sure we woudln't turn them away
[22:47:57] <howief>	but no one's decided whether to make an effort to recruit them as we did in phase iii
[22:48:35] *kibble 	nods.
[22:48:59] <howief>	that's pretty much what's been discussed so far
[22:49:17] <howief>	i actually have to run in 5 minutes
[22:49:23] <kibble>	Okay, thanks, I'll cleanup the stuff with that info.
[22:49:25] <howief>	is this something we can continue tomorrow
[22:49:35] *guillom 	huggles kibble
[22:49:42] <howief>	great
[22:49:50] <kibble>	I'm leaving at about 1pm tomorrow for vacation, which is 10am your time, so maybe, but I doubt it. :P
[22:49:56] *kibble 	embrasse guillom.
[22:50:02] <howief>	one thing we could really use help with is, as you mention, getting the word out
[22:50:23] <TrevorParscal>	RoanKattouw: you still alive?
[22:50:26] <RoanKattouw>	Yes I am
[22:50:35] <TrevorParscal>	how you feeling?
[22:50:40] <howief>	will you be able to help after coming back from vacation?
[22:50:48] <RoanKattouw>	I'm feeling good, no reason not to :)
[22:50:59] <RoanKattouw>	I heard I passed the software engineering course today
[22:51:05] <TrevorParscal>	wooo hooo
[22:51:07] *kibble 	^5 RoanKattouw 
[22:51:13] <RoanKattouw>	You know, the one for which I forsaked my WMF work for two months
[22:51:21] <TrevorParscal>	yes
[22:51:25] <TrevorParscal>	i know of that one
[22:51:27] <TrevorParscal>	:)
[22:51:28] *RoanKattouw 	wonders if "forsaked" is right
[22:51:39] <TrevorParscal>	deprioritized
[22:51:47] <RoanKattouw>	Yeah
[22:52:20] <RoanKattouw>	Got a C, too, which, when converted to our grades, is equal to the GPA I need to keep up with for cum laude graduation so yay
[22:52:47] <RoanKattouw>	Their grade mapping is weird though
[22:52:52] <RoanKattouw>	They map C to 80/100
[22:53:43] <flipzagging>	I found someone else who orders countries by ISO code rather than name
[22:53:54] <flipzagging>	http://creativecommons.org/choose/
[22:54:01] <flipzagging>	see the jurisdiction menu
[22:54:51] <RoanKattouw>	Either way I'm getting sleepy now but still crossing my fingers hoping Tim will pop in before I go to sleep
[22:56:11] <kibble>	howief, phase V would be all other projects or just all other Wikipedias?
[22:58:58] <TrevorParscal>	RoanKattouw: added diagram
[22:59:25] <RoanKattouw>	Cool lemme look
[22:59:47] <RoanKattouw>	"which are use the the"
[23:00:03] <RoanKattouw>	Now who needs sleep :P
[23:00:33] <TrevorParscal>	which are use the?
[23:00:37] <TrevorParscal>	oh
[23:00:38] <TrevorParscal>	ha
[23:02:31] <TrevorParscal>	k
[23:02:41] <TrevorParscal>	fixed the bit of language
[23:03:16] *TrevorParscal 	sees more typos
[23:03:24] <RoanKattouw>	Yeah there were more
[23:03:33] <TrevorParscal>	Resources loader
[23:03:39] <TrevorParscal>	got that one
[23:03:41] <TrevorParscal>	any others?
[23:03:55] <RoanKattouw>	Resrouces loader
[23:03:59] <TrevorParscal>	yeah
[23:04:50] <TrevorParscal>	ok
[23:04:55] <TrevorParscal>	should be good now
[23:05:11] <RoanKattouw>	OK I guess Proxy = Squid and Cache = browser cache?
[23:05:17] <TrevorParscal>	yes
[23:05:54] <TrevorParscal>	so, the versioning stuff
[23:06:03] <TrevorParscal>	what do we need to do to get that specified?
[23:06:21] <RoanKattouw>	Talk to Tim, basically
[23:06:48] <TrevorParscal>	you have some overlap between you waking up and him heading to bed?
[23:06:58] <RoanKattouw>	Well in theory
[23:07:05] <RoanKattouw>	Andrew's awake
[23:07:18] <TrevorParscal>	i see that
[23:07:21] <RoanKattouw>	Of course Tim will be there when I wake up tomorrow
[23:07:27] <TrevorParscal>	right
[23:07:29] <RoanKattouw>	It'll be his start of the evening, approximately
[23:07:49] <werdna>	indeed
[23:08:26] <TrevorParscal>	werdna: thoughts on script loader stuffs?
[23:08:30] <RoanKattouw>	OK what we're not mentioning at this point
[23:08:34] <werdna>	TrevorParscal: not sucking would be a brilliant start
[23:08:46] <TrevorParscal>	RoanKattouw: yes?
[23:08:48] <RoanKattouw>	Oh wait
[23:08:56] <RoanKattouw>	I was gonna talk about parameterization of modules
[23:09:01] <RoanKattouw>	But that only makes sense for a bucket setup
[23:09:04] <RoanKattouw>	Which we don't use
[23:09:11] <RoanKattouw>	So Tim's story about parameterization is moot
[23:09:19] <TrevorParscal>	right
[23:09:24] <RoanKattouw>	We'd just put that stuff in separate modules
[23:09:29] <TrevorParscal>	agreed
[23:09:39] <TrevorParscal>	buckets are a layer of abstraction I don't see as needed
[23:10:08] <RoanKattouw>	Yeah
[23:24:36] <TrevorParscal>	RoanKattouw: before you go to bed
[23:24:45] <TrevorParscal>	take a look at my versioning section
[23:24:53] <TrevorParscal>	amd I missing something important?
[23:25:29] <TrevorParscal>	consider adding pros/cons to either idea
[23:34:06] <TrevorParscal>	RoanKattouw: seems like there must be more positives for option 1
[23:34:32] <RoanKattouw>	"Old HTML can load new resources, resources need be at least temporarily backwards compatible whenever possible (more complex to deploy changes)"
[23:34:36] <RoanKattouw>	This applies to option 2 also
[23:34:42] <TrevorParscal>	really?
[23:34:47] <TrevorParscal>	oh
[23:34:50] <TrevorParscal>	yeah, i know
[23:34:57] <RoanKattouw>	Well in practice it doesn't happen very much, especially with Varnish
[23:35:00] <TrevorParscal>	it says "Old HTML probably loads old resources"
[23:35:02] <RoanKattouw>	But it can in theory
[23:35:18] <TrevorParscal>	adjusting
[23:37:21] <TrevorParscal>	k
[23:38:01] <TrevorParscal>	so, I'm not planning on sharing this with everyone (read Ti, Mi, Ni) until we think it's done
[23:38:19] <TrevorParscal>	but obviously it's very close except for this section, which needs more input
[23:38:29] <TrevorParscal>	so, do you think we should invite them as viewers?
[23:40:11] <RoanKattouw>	Yes, I think it's ready for that
[23:40:16] <TrevorParscal>	ok
[23:40:24] <RoanKattouw>	Where viewers != editors, right?
[23:40:36] <RoanKattouw>	I'd like feedback in a separate place, not inline or in the form of changes
[23:44:22] <TrevorParscal>	yes
[23:44:37] <TrevorParscal>	correct, the only people with editing access are you, I, danese and erik
[23:45:11] <TrevorParscal>	they have access because I obey them, and I doubt they will start changing crap and adding comments
[23:45:23] <TrevorParscal>	but i don't think they care about editing
[23:45:39] <TrevorParscal>	i can lock them out of it if you think we should..
[23:45:52] <TrevorParscal>	it's not a big deal to them I think - but they might fix typos :)
[23:46:00] <RoanKattouw>	Nah don't worry about Danese and Erik
[23:46:05] <TrevorParscal>	right
[23:46:21] <RoanKattouw>	Locking your bosses out looks bad
[23:46:24] <TrevorParscal>	werdna: are you interested in reviewing the resource loader specification?
[23:46:28] <TrevorParscal>	RoanKattouw: agreed
[23:47:10] <TrevorParscal>	we're looking for input from people who know the space well, you easily make that list with your work in the core and with LQT
[23:48:18] <werdna>	TrevorParscal: certainly
[23:48:23] <werdna>	send it my way and I'll take a look
[23:48:40] <TrevorParscal>	your epstone acct?
[23:48:46] <TrevorParscal>	or do you have a wmf email?
[23:48:55] <werdna>	agarrett@wikimedia.org -- but it all goes to the same inbox
[23:48:55] <TrevorParscal>	it's a google doc
[23:48:58] <werdna>	so pick whatever you want
[23:49:12] *TrevorParscal 	chooses WMF
[23:50:33] <TrevorParscal>	werdna: you haz an invite
[23:51:08] <werdna>	so i haz
[23:51:19] <werdna>	TrevorParscal: btw I should start figuring out what I'm going to do with you guys next week
[23:51:58] <TrevorParscal>	well, we have a new UI guy, and he's going to talk to you about taking a pass at the LQT UI
[23:52:08] <TrevorParscal>	you are welcome to help Roan and I with the resource loader
[23:52:19] <TrevorParscal>	and I think everyone wants your hot body for 100 other projects...
[23:52:24] <TrevorParscal>	so, you shall be busy
[23:53:00] <werdna>	"projects"
[23:53:05] <werdna>	Let me guess, Philippe? :)
[23:53:11] <TrevorParscal>	dude, you know it
[23:53:30] <guillom>	haha
[23:54:09] <TrevorParscal>	I think you will have a bunch of meetings at first, and then be expected to sit quietly and crank out MW code at a dizzying pace while ingesting copious quantities of caffine
[23:55:07] <TrevorParscal>	I'm hoping to have some people over to my house to hack and play video games after work on one of the days my wife is at her mom's...
[23:55:35] <TrevorParscal>	but we have video games at the office now
[23:55:50] <TrevorParscal>	so, you can also just not work... that's always cool
[23:55:53] <RoanKattouw>	I'm up for that, but then I guess we won't have many if any other remote folks over at that time
[23:56:27] <TrevorParscal>	RoanKattouw: chad might end up coming at a similar time as you
[23:56:40] <RoanKattouw>	Oh nice
[23:56:49] <RoanKattouw>	Adam already said he had to be in Pittsburgh at that time