Log of #symfony-dev meeting 20110210 (all times GMT-5)

transcript.txt

Feb 10 11:02:52 <Stof>	lsmith: meeting ? or do you want to wait ?
Feb 10 11:03:10 <lsmith>	i guess we might as well start
Feb 10 11:03:16 <lsmith>	not sure if he is traveling back today
Feb 10 11:03:18 <lsmith>	or whatever
Feb 10 11:03:41 <lsmith>	then again .. most of the topics seem to require his presence
Feb 10 11:03:53 <Stof>	yes
Feb 10 11:04:05 <lsmith>	skip bootstrap.php in app_dev.php: http://bit.ly/htff3I
Feb 10 11:04:10 <lsmith>	maybe that can be discussed ..
Feb 10 11:04:16 <lsmith>	it should just be a small pull for the sandbox
Feb 10 11:04:27 <lsmith>	anyone else around?
Feb 10 11:04:37 <lsmith>	jmikola, kriswallsmith, Seldaek?
Feb 10 11:04:46 <Seldaek>	yup
Feb 10 11:05:11 <mvrhov>	what's interesting is that, I wasn't weven using bootstrap.php.. if it wasn't for your email I'd probably forget about it
Feb 10 11:05:11 <Seldaek>	bootstrap, I'm with you on that one
Feb 10 11:05:41 <Stof>	well, this would mean having a bootstrap_dev.php which require the autoloader and the autoload.php file
Feb 10 11:05:52 <lsmith>	yeah
Feb 10 11:05:53 <Stof>	instead of requiring the Symfony bootstrap file
Feb 10 11:06:04 <Stof>	but I think it is a good idea
Feb 10 11:06:20 <lsmith>	not sure why i didnt just make a pull request
Feb 10 11:06:48 <lsmith>	hmm rande is also not around
Feb 10 11:06:56 <lsmith>	johanness: ping
Feb 10 11:07:47 <jmikola|w>	lsmith: hi
Feb 10 11:07:54 <jmikola|w>	oh snap, meeting time
Feb 10 11:08:02 <jmikola|w>	i think lots of people are travelling 
Feb 10 11:08:07 <lsmith>	yeah
Feb 10 11:09:47 <lsmith>	well i am not sure if we are having a meeting at all
Feb 10 11:10:07 <jmikola|w>	i would suggest bumping to friday, but realized that's like 5pm for you european folks :)
Feb 10 11:10:18 <johnwards>	Beer O'clock
Feb 10 11:10:26 <fabian>	stzbr!
Feb 10 11:10:37 *	henrikbjorn has quit (Remote host closed the connection)
Feb 10 11:10:42 <Stof>	is there any subject which does not require fabpot's presence ?
Feb 10 11:10:49 <lsmith>	fabian: now you have to explain stzbr to the uninitiated :)
Feb 10 11:11:10 <lsmith>	well we could try to push a topic forward
Feb 10 11:11:17 <lsmith>	like Eager Response Creation: http://bit.ly/f3FskB
Feb 10 11:11:23 <lsmith>	which is still in the early thought stages
Feb 10 11:11:33 <lsmith>	but johanness doesnt seem to be around
Feb 10 11:11:47 *	notjosh (~notjosh@S0106001d60439f63.vc.shawcable.net) has joined #symfony-dev
Feb 10 11:12:32 <Stof>	well, if we need the presence of other peoples for all subjects, the meeting seems like being a fail today
Feb 10 11:12:51 <jmikola|w>	re: eager response creation, when a controller could have several exit paths, returning redirects, forwards or rendered templates, those would all capitalize on the same response object?
Feb 10 11:13:03 <lsmith>	jmikola|w: yes
Feb 10 11:13:30 *	Herzult (~Herzult@ip-22.net-89-2-70.rev.numericable.fr) has left #symfony-dev
Feb 10 11:13:31 <lsmith>	the main reason to do this is for listeners that happen before the controller
Feb 10 11:13:33 *	Herzult (~Herzult@ip-22.net-89-2-70.rev.numericable.fr) has joined #symfony-dev
Feb 10 11:13:40 <jmikola|w>	which need to alter the response
Feb 10 11:13:46 <lsmith>	yeah
Feb 10 11:14:12 <lsmith>	with scoping in place now ..
Feb 10 11:14:21 <Seldaek>	it might also fix the fact that right now, the response is in the  prototype scope
Feb 10 11:14:22 <lsmith>	it doesnt seem such a bad idea anymore
Feb 10 11:14:27 <Seldaek>	which honestly is .. I don't know what
Feb 10 11:14:38 <jmikola|w>	Seldaek: that's just using the container as a respone factory :)
Feb 10 11:14:42 <lsmith>	Seldaek: protoype means that you get a fresh instance everytime
Feb 10 11:14:51 <jmikola|w>	per the mailing list, i think the big objection was if request contained a response, is that true?
Feb 10 11:14:59 <Seldaek>	yeah ok, but it should be in the request scope imo
Feb 10 11:15:16 <lsmith>	Seldaek: thats essentially what johanness was proposing
Feb 10 11:15:20 <lsmith>	if its in the request scope
Feb 10 11:15:30 <lsmith>	listeners could make an instance and set stuff on the response
Feb 10 11:15:40 <lsmith>	and when the controller then gets the response
Feb 10 11:15:43 <jmikola|w>	the only thing i can think of if we share a response object in that scope, we might want a method to essentially reset the response
Feb 10 11:15:45 <lsmith>	it would already have stuff set on it
Feb 10 11:16:05 <Stof>	jmikola|w: why resetting the response ?
Feb 10 11:16:31 <jmikola|w>	just thinking how the redirect() convenience methods both sets the location header and content for a meta html redirect
Feb 10 11:16:55 <lsmith>	jmikola|w: sure a reset() might be ok ..
Feb 10 11:17:00 <jmikola|w>	maybe someone has a listener that will set the location header but forget to erase or reset the content
Feb 10 11:17:20 <jmikola|w>	wouldn't really matter if the response code was a 3xx, of course - just thinking aloud
Feb 10 11:17:23 <lsmith>	for the rare cases that you think you need to start over
Feb 10 11:17:53 <lsmith>	do we see any real use case for multiple response instances inside a single (sub)request?
Feb 10 11:17:55 *	gordonslondon has quit (Read error: Connection reset by peer)
Feb 10 11:17:58 <jmikola|w>	i think initially, it would make it easy for people to port over code that currently constructs new responses
Feb 10 11:18:03 *	gordonslondon (~julesbous@ble59-5-82-233-204-65.fbx.proxad.net) has joined #symfony-dev
Feb 10 11:18:33 *	gordonslondon has quit (Read error: No route to host)
Feb 10 11:18:35 <jmikola|w>	ultimately, a request is only going to result in a single response, whether it's sub or master - i can't think of a reason to have multiple responses
Feb 10 11:18:37 *	gordonslondon (~julesbous@ble59-5-82-233-204-65.fbx.proxad.net) has joined #symfony-dev
Feb 10 11:18:57 <Stof>	I can't neither
Feb 10 11:19:09 <lsmith>	so then its settled :)
Feb 10 11:19:27 <jmikola|w>	there's no pull request for this yet, correct? just ML discussion
Feb 10 11:19:28 <lsmith>	also for the rare case where for some obscure reason you think you do need multiple response's
Feb 10 11:19:40 *	rooster has quit (Remote host closed the connection)
Feb 10 11:19:41 <lsmith>	you can just inherit the default service and set a different scope
Feb 10 11:19:58 <lsmith>	jmikola|w: the pull request is insanley trivial .. change the scope
Feb 10 11:20:09 <lsmith>	well then a couple of listeners could be simplified
Feb 10 11:23:27 <johnwards>	Meeting over?
Feb 10 11:23:39 <lsmith>	johnwards: ask away :)
Feb 10 11:23:52 <unknownbliss>	never started did it?
Feb 10 11:24:12 <Stof>	unknownbliss: well, we discussed two of the points
Feb 10 11:24:26 <Stof>	but most devs are not there today