Skip to content

Instantly share code, notes, and snippets.

@jmikola
Created February 17, 2011 17:10
Show Gist options
  • Select an option

  • Save jmikola/832145 to your computer and use it in GitHub Desktop.

Select an option

Save jmikola/832145 to your computer and use it in GitHub Desktop.
Log of #symfony-dev meeting 20110217 (all times GMT-5)
Feb 17 11:00:13 <fabpot> ok, so let's start
Feb 17 11:00:34 <Seldaek> not many seem to be around :/
Feb 17 11:00:35 <fabpot> first topic is about the Symfony sub-namespaces
Feb 17 11:01:01 <fabpot> don't know why we need to talk about that as last time I did, everybody seem to be againts any change
Feb 17 11:01:01 * na8flush ([email protected]) has joined #symfony-dev
Feb 17 11:01:18 <Seldaek> bschussek: avalanche123 iampersistent jmikola|w jmikola johanness kriswallsmith lala ping
Feb 17 11:01:25 <jmikola|w> pong pong pong pong
Feb 17 11:01:28 <avalanche123> Seldaek lala pong
Feb 17 11:01:30 <jmikola|w> oh, meeting time :)
Feb 17 11:01:32 * aways|bnc Bye all
Feb 17 11:01:42 * kriswallsmith is here
Feb 17 11:01:50 <Seldaek> fabpot: I believe the thing is that bschussek proposed something new
Feb 17 11:01:51 <lsmith> fabpot: about to head out ..
Feb 17 11:01:59 <lsmith> but yeah .. there was a new proposal on the list
Feb 17 11:02:14 <fabpot> bschussek proposal is to move bundles under Framework
Feb 17 11:02:15 <Seldaek> http://groups.google.com/group/symfony-devs/msg/5e9b52a340557a26
Feb 17 11:02:25 <fabpot> and move up components under the main namespace
Feb 17 11:02:28 <Seldaek> yes
Feb 17 11:02:33 <Seldaek> you didn't answer afaik
Feb 17 11:02:42 <fabpot> Seldaek: I missed that one
Feb 17 11:02:45 <Seldaek> so that's the main question, is it a no go for you or not :)
Feb 17 11:02:56 <kriswallsmith> i like the current layout
Feb 17 11:03:25 <fabpot> but then, the components are not clearly stored under one place
Feb 17 11:03:27 * johnwards_ ([email protected]) has joined #symfony-dev
Feb 17 11:03:36 <fabpot> I think the current way is good enough
Feb 17 11:03:38 <Seldaek> sure, but if we move them to their own repo at some point anyway
Feb 17 11:03:47 <Seldaek> then the subnamespace would make less sense
Feb 17 11:03:48 <jmikola|w> Seldaek: with bernhard's proposal, would the best practice for other vendors be to have no Bundle/ in their bundle namespaces?
Feb 17 11:04:03 <Seldaek> jmikola|w: no that has nothing to do with vendors I think
Feb 17 11:04:16 <fabpot> the problem is that the "best practice" will not be followed by Symfony anyway
Feb 17 11:04:24 <Seldaek> yes, that's irrelevant
Feb 17 11:04:43 <jmikola|w> I don't mind Symfony having a component namespace... after all, those will become http://components.symfony-project.org/ right? :)
Feb 17 11:04:43 <fabpot> IIRC, we advocate Sensio\FooBundle
Feb 17 11:04:44 <kriswallsmith> on the other hand, if i wanted to use a component in a non-symfony project, Symfony\Yaml would be nicer than Symfony\Component\Yaml
Feb 17 11:04:53 <bschussek> kriswallsmith: +1
Feb 17 11:05:00 <Seldaek> yeah that's what I meant kriswallsmith
Feb 17 11:05:05 <bschussek> I think it makes sense, because the bundles are the "framework"
Feb 17 11:05:09 <Seldaek> if they're in their own repo/package, then the extra namespace becomes annoying
Feb 17 11:05:13 <bschussek> so Symfony\Framework is that, and Symfony\X is everything else
Feb 17 11:05:16 <fabpot> as discussed on the CMF ML, we will have a Symfony\CMF
Feb 17 11:05:18 <jmikola|w> bschussek: would we have Framework\FrameworkBundle? :)
Feb 17 11:05:39 <kriswallsmith> CoreBundle?
Feb 17 11:05:42 <bschussek> jmikola|w: for example. or CoreBundle, I don't mind
Feb 17 11:05:42 <fabpot> so, there is no clear difference between a component, the framework (a bunch of bundle), and the CMF
Feb 17 11:05:49 <fabpot> looks messy, no?
Feb 17 11:05:55 <Seldaek> fabpot: not really
Feb 17 11:05:56 <bschussek> I don't think so
Feb 17 11:06:02 <Seldaek> the framework is a component of sorts
Feb 17 11:06:10 <fabpot> Seldaek: not at all
Feb 17 11:06:11 <Seldaek> you can use it on its own, it just has lots of dependencies
Feb 17 11:06:16 <fabpot> the framework is definitely NOT a component
Feb 17 11:06:33 <fabpot> the "few dependencies" is part of the component definition
Feb 17 11:06:42 <Seldaek> yes of course it's special
Feb 17 11:06:53 <Seldaek> but I don't see it as bad to have it under the same namespace
Feb 17 11:07:02 <Seldaek> it's quite clear what Framework is..
Feb 17 11:07:04 <Seldaek> same for CMF
Feb 17 11:07:15 <fabpot> the problem is not the CMF or Framework namespaces, but all the other ones
Feb 17 11:07:16 * johnwards has quit (Ping timeout: 240 seconds)
Feb 17 11:07:16 * johnwards_ is now known as johnwards
Feb 17 11:07:52 <Seldaek> well the notion of what a component is honestly is just as vague
Feb 17 11:07:59 <Seldaek> component means something to us because it has been defined
Feb 17 11:08:00 <fabpot> it's not
Feb 17 11:08:10 <Seldaek> if we define that those things in Symfony, except Framework, are components
Feb 17 11:08:20 <Seldaek> then it's the same.. it's just a matter of documenting it properly I think
Feb 17 11:08:36 <Seldaek> in another framework, Component could be something totally not reusable and standalone
Feb 17 11:08:43 <fabpot> let's put it another way: Why would you want to remove the Component\ sub-namespace?
Feb 17 11:08:44 <Seldaek> it's not a universal meaning, that's what I mean
Feb 17 11:09:12 * gordonslondon_ ([email protected]) has joined #symfony-dev
Feb 17 11:09:13 * gordonslondon has quit (Read error: Connection reset by peer)
Feb 17 11:09:13 * gordonslondon_ is now known as gordonslondon
Feb 17 11:09:14 <Seldaek> on the plus side it makes things shorter, and it has no real benefit imo
Feb 17 11:09:16 <bschussek> fabpot: for clarity, mainly. I like the short namespaces of Doctrine
Feb 17 11:09:22 <bschussek> but it's no must have
Feb 17 11:09:31 <jmikola|w> Seldaek: i think other projects using namespaces are more single purpose - while Symfony\ here is lumping a lot of independent things together... perhaps one of the few things to compare to would be ZF2, but those are essentially all components afaik
Feb 17 11:09:40 * weaverryan has quit (Ping timeout: 276 seconds)
Feb 17 11:09:43 <pgodel_work> you can always use "use namespace;" to shorten things
Feb 17 11:10:00 <fabpot> to be clear, I'm not opposed to the change. But we are now in a stabilization phase. So we only change things if there is a real advantage
Feb 17 11:10:20 <bschussek> what about a voting then?
Feb 17 11:10:25 <Seldaek> pgodel_work: well, you gotta write the long use statements.. :)
Feb 17 11:10:33 <kriswallsmith> can someone recap why the CMF is going into Symfony\CMF?
Feb 17 11:10:36 <pgodel_work> IDE autocompletion ;)
Feb 17 11:10:38 <fabpot> bschussek: I want pros and cons before
Feb 17 11:10:51 <Seldaek> kriswallsmith: please later, unless it's a pressing matter
Feb 17 11:11:01 <kriswallsmith> it will inform this decision
Feb 17 11:11:02 * weaverryan ([email protected]) has joined #symfony-dev
Feb 17 11:11:10 <fabpot> Seldaek: it is important to make the right decision
Feb 17 11:11:22 <fabpot> let's add another one: Symfony\Twig
Feb 17 11:11:26 <Seldaek> well, they will never be in the same repo I guess, so I'd say it's less important
Feb 17 11:11:32 <fabpot> that's something I'm thinking about
Feb 17 11:11:37 <jmikola|w> if CMF, as a project is going into the top namespace, i see value in keeping sf2 components isolated from that
Feb 17 11:11:48 <Seldaek> but the rationale behind the CMF is that it's endorsed by Symfony
Feb 17 11:11:50 <fabpot> jmikola|w: +1
Feb 17 11:11:57 <pgodel_work> jmikola|w: and there may be more things like that in the future
Feb 17 11:11:58 <Seldaek> and it is a CMFramework
Feb 17 11:12:10 <Seldaek> it's not really a product on its own
Feb 17 11:12:12 <jmikola|w> pgodel_work: yes, i considered Symfony\Twig, too
Feb 17 11:12:15 <kriswallsmith> why is CMF different than YAML?
Feb 17 11:12:16 <johanness> fabpot, what about different release cycles and dependencies between components?
Feb 17 11:12:29 <fabpot> johanness: What do you mean?
Feb 17 11:12:30 <johanness> that wouldn't work so well if they are all in the same repo no?
Feb 17 11:12:35 <Seldaek> kriswallsmith: it's not a component, it'll use many dependencies
Feb 17 11:12:48 <fabpot> johanness: there are in the same namespace, but can be in different repo
Feb 17 11:12:48 <pgodel_work> same namespace does not mean same repo
Feb 17 11:12:52 <kriswallsmith> Seldaek: i see
Feb 17 11:12:55 <fabpot> they*
Feb 17 11:13:08 <kriswallsmith> +1 for no change
Feb 17 11:13:09 <fabpot> 2 minutes left
Feb 17 11:13:19 <pgodel_work> +1 no change
Feb 17 11:13:22 <Stof> +& for no change too
Feb 17 11:13:24 <fabpot> +1 no change
Feb 17 11:13:27 <Seldaek> well, I'll live either way, but I would have liked the shortened ns
Feb 17 11:13:30 <Nenuial> +1 no change
Feb 17 11:13:33 <bschussek> +1 change :)
Feb 17 11:13:37 <fabpot> bschussek: ?!
Feb 17 11:13:41 <mvrhov> +1 fo no change.
Feb 17 11:13:57 <bschussek> I'm alright with no change, but as for the voting :P
Feb 17 11:13:59 <Seldaek> bschussek: it's the liip office influence, you start having ideas that nobody agrees with
Feb 17 11:14:07 <Seldaek> it's okaay
Feb 17 11:14:09 <jmikola|w> in the interest of how close we are to stabilizing, +1 for no change (this seems more opinion than functional need)
Feb 17 11:14:10 <fabpot> Seldaek: hehe
Feb 17 11:14:22 <jmikola|w> Seldaek: Symfony\ViewBundle? :)
Feb 17 11:14:25 <fabpot> Seldaek: by the way, I'm the one who started this thread
Feb 17 11:14:38 <fabpot> jmikola|w: looks ugly o)
Feb 17 11:14:46 * kriswallsmith gavels
Feb 17 11:14:48 <Seldaek> well then next topic I guess :p
Feb 17 11:14:53 <fabpot> yep
Feb 17 11:15:04 <fabpot> what's missing for RC1?
Feb 17 11:15:17 <fabpot> bschussek: my first main concern is the form/validation component
Feb 17 11:15:24 <fabpot> 2 things: intl and the binding process
Feb 17 11:15:42 <bschussek> fabpot: intl is being worked on
Feb 17 11:15:52 <bschussek> progress can be seen at github.com/Infranology/symfony
Feb 17 11:16:03 <fabpot> bschussek: great, I will have a look then
Feb 17 11:16:06 <fabpot> any ETA?
Feb 17 11:16:14 <igorw> also igorw/symfony for the dateformatter
Feb 17 11:16:16 <kriswallsmith> i have work to do in assetic and AsseticBundle
Feb 17 11:16:27 <fabpot> locale branch?
Feb 17 11:16:34 <bschussek> numberformatter, I think
Feb 17 11:16:36 <igorw> numberformatter
Feb 17 11:16:40 <igorw> and intldateformatter on mine
Feb 17 11:16:50 <bschussek> \Locale and \NumberFormatter are already ported
Feb 17 11:16:59 <bschussek> igorw is working on \IntlDateFormatter
Feb 17 11:17:11 <bschussek> then there's \Collator missing, but I hope this'll be a quick one
Feb 17 11:17:12 <jmikola|w> kriswallsmith: assetic won't depend on bulat's imagine stuff for initial release, correct? not sure how much overlap there was on spriting
Feb 17 11:17:16 <igorw> bschussek: would be great if you could review what I have and tell me what else is needed
Feb 17 11:17:27 <kriswallsmith> jmikola|w: correct
Feb 17 11:17:36 <kriswallsmith> we may work together when it comes to image sprites
Feb 17 11:17:50 <fabpot> let's finish intl before
Feb 17 11:17:53 <bschussek> igorw: I'll do that tomorrow
Feb 17 11:18:10 <fabpot> bschussek: so, as I understand, you are confident about it, right?
Feb 17 11:18:11 <bschussek> fabpot: so if ecosta has some time, we might have a first intl candidate by the end of next week
Feb 17 11:18:26 <bschussek> yes
Feb 17 11:18:36 <fabpot> great
Feb 17 11:18:56 <bschussek> as for the binding process, I'll change that, but I also want to take Bulat's, Jon's and Lukas' feedback into account
Feb 17 11:19:02 <bschussek> more on that next week
Feb 17 11:19:06 <fabpot> great
Feb 17 11:19:19 <fabpot> anything special on assetic?
Feb 17 11:19:27 <fabpot> kriswallsmith: what about the documentation for the Symfony book?
Feb 17 11:19:45 <kriswallsmith> i'm tracking issues here https://github.com/kriswallsmith/symfony/issues
Feb 17 11:20:06 <kriswallsmith> there are issues with how assets configured in templates are cached
Feb 17 11:20:23 <kriswallsmith> i'll probably write a custom cache mechanism in assetic for that
Feb 17 11:20:25 <Seldaek> I really hope I find time to play with assetic next week
Feb 17 11:20:31 <Seldaek> I'm very interested in the topic
Feb 17 11:20:35 <fabpot> Seldaek: same for me
Feb 17 11:20:48 * johnkary ([email protected]) has joined #symfony-dev
Feb 17 11:20:56 <fabpot> kriswallsmith: Will you find some time to write a small chapter on assetic for the book?
Feb 17 11:20:58 <kriswallsmith> i also need to write the loader that extracts assets from php templates
Feb 17 11:21:21 <igorw> who is responsible for spriting stuff? one of the phpBB guys was doing something with that lately, he might be able to help
Feb 17 11:21:22 <kriswallsmith> i need to solve the caching issue before i document it
Feb 17 11:21:31 <fabpot> kriswallsmith: ok
Feb 17 11:21:34 * krymen ([email protected]) has joined #symfony-dev
Feb 17 11:21:36 <kriswallsmith> once that's in place i'll be confident it's stable enough
Feb 17 11:21:55 <pgodel_work> talking about phpBB, is the bundle dependency going in RC1 ?
Feb 17 11:21:56 <fabpot> next thing to finish before RC1 is the conversion of the DIC extension configuration
Feb 17 11:22:12 <fabpot> pgodel_work: hopefully, but that will be though I think
Feb 17 11:22:14 <kriswallsmith> anyone want to do that for AsseticBundle?
Feb 17 11:22:16 <Stof> I think 2 docs are needed: the doc about Symfony integration and the doc of the standalone library
Feb 17 11:22:21 <weaverryan> I'm working on some tweaks to the "Definition" class right now per the extensions
Feb 17 11:22:46 <fabpot> TwigBundle should also be migrated; this one should be easy. Anyone?
Feb 17 11:22:50 <Stof> kriswallsmith: I can convert AsseticBundle
Feb 17 11:22:56 <kriswallsmith> Stof: thanks
Feb 17 11:22:59 <weaverryan> it's still somewhat difficult to use, because we're so "friendly" to different formats - something can be null, true, an array etc
Feb 17 11:23:16 * na8flush ([email protected]) has left #symfony-dev
Feb 17 11:23:27 <kriswallsmith> please send assetic-related pull requests through kriswallsmith/symfony
Feb 17 11:23:41 <Stof> ok
Feb 17 11:24:00 <jmikola|w> fabpot: what remains in Twig Bundle?
Feb 17 11:24:04 <jmikola|w> to the config builder?
Feb 17 11:24:18 * jonwage ([email protected]) has joined #symfony-dev
Feb 17 11:24:28 <jmikola|w> I can do that today if so
Feb 17 11:24:59 <fabpot> jmikola|w: that would be great
Feb 17 11:25:51 <weaverryan> I'm still on the ODM - it's mostly finished locally - will love some feedback soon from the guys using it in prod :)
Feb 17 11:25:57 <jmikola|w> weaverryan: and please ping me when you submit a pull for Def changes, i'd be interested in tracking those
Feb 17 11:26:16 <jmikola|w> weaverryan: can accommodate you on ODM feedback, too :)
Feb 17 11:26:27 <weaverryan> will-do thanks
Feb 17 11:27:22 <Stof> DoctrineBundle needs to be converted too
Feb 17 11:27:26 <Seldaek> fabpot: next?
Feb 17 11:27:57 <Stof> I can work on it this weekend
Feb 17 11:28:14 <weaverryan> Stof: that's very-much related to the mongodb one - so let's coordinate somewhat
Feb 17 11:28:18 <fabpot> next topic: bootstrap.php in dev env
Feb 17 11:28:19 <fabpot> http://groups.google.com/group/symfony-devs/browse_thread/thread/5418d43e7ffdebd9
Feb 17 11:28:22 <fabpot> https://github.com/symfony/symfony-sandbox/pull/30
Feb 17 11:28:24 <weaverryan> would love to have your thoughts on it anyways
Feb 17 11:29:07 <Seldaek> fabpot: yes, so we talked a bit last time but you weren't there
Feb 17 11:29:25 <jmikola|w> fabpot: how is bootstrap currently generated?
Feb 17 11:29:29 <kriswallsmith> do you mean in debug?
Feb 17 11:29:30 <Seldaek> basically it's about not using the bootstrap.php file in dev, because it makes it quite annoying to debug
Feb 17 11:29:39 <fabpot> it's not generated
Feb 17 11:29:43 <fabpot> it's included in the framework
Feb 17 11:29:44 * kertz_ ([email protected]) has joined #symfony-dev
Feb 17 11:29:53 <jmikola|w> oh, it's not the result of addCompiledClass() ?
Feb 17 11:29:58 <Stof> jmikola|w: no
Feb 17 11:30:03 <avalanche123> +1 on not using bootstrap.php it in debug=true
Feb 17 11:30:07 <Seldaek> jmikola|w: nah that's another bootstrap :)
Feb 17 11:30:17 <Seldaek> it's the bootstrapped bootstrap
Feb 17 11:30:19 <johnwards> As a n00b to deving with Symfony2 it is a pain for getting my head around the codebase
Feb 17 11:30:27 <Stof> addCompiledClass create a classXXXX.php file in your cache
Feb 17 11:30:28 <johnwards> with bootstrap in dev
Feb 17 11:30:45 <Stof> johnwards: even for non-noob :)
Feb 17 11:30:47 <jmikola|w> Stof: oh, with the special loader?
Feb 17 11:31:02 <Seldaek> so.. if nobody has any argument against, we agree on removing bootstrap from the sandbox?
Feb 17 11:31:08 <Stof> jmikola|w: yes, this file is generated by ClassCollectionLoader
Feb 17 11:31:10 <fabpot> I want all the env to work the same way as much as possible
Feb 17 11:31:13 <fabpot> so, I'm )1
Feb 17 11:31:17 <fabpot> -1
Feb 17 11:31:26 <Seldaek> but what about generating it on the fly vs having it compiled?
Feb 17 11:31:29 <fabpot> and we can probably document how to de-activate it
Feb 17 11:31:31 <Seldaek> because then, envs could be the same
Feb 17 11:31:40 <johnwards> fabpot has a big point here. that is a big pain in sf1
Feb 17 11:31:40 <fabpot> Seldaek: not possible
Feb 17 11:31:51 <Seldaek> fabpot: yeah not completely
Feb 17 11:32:00 <fabpot> the bootstrap file contains things that are needed very early
Feb 17 11:32:04 <fabpot> so, it cannot be generated
Feb 17 11:32:10 <jmikola|w> chicken and egg problem? :)
Feb 17 11:32:13 <Seldaek> well, if you load the autoloader
Feb 17 11:32:16 <Seldaek> I guess it can?
Feb 17 11:32:22 * rooster has quit (Remote host closed the connection)
Feb 17 11:32:25 <Seldaek> that's how it was before no?
Feb 17 11:32:35 <jmikola|w> fabpot: so when these core classes change (if at all), you have to update the code in two places?
Feb 17 11:32:42 * kertz has quit (Ping timeout: 260 seconds)
Feb 17 11:32:43 * kertz_ is now known as kertz
Feb 17 11:32:57 <Seldaek> jmikola|w: there is a script to generate it
Feb 17 11:33:01 <Seldaek> it's just not done on the fly
Feb 17 11:33:07 <fabpot> jmikola|w: there is a script that need to be called to regenerate them
Feb 17 11:33:27 <Seldaek> so .. -1?
Feb 17 11:33:29 <Seldaek> +1?
Feb 17 11:33:30 <mvrhov> you can live just fine without bootstra.php. I was until lsmith mentioned it last week that he'd like to exclude it from debug
Feb 17 11:33:32 <jmikola|w> and having bootstrap.php just require the things it currently includes inline would defeat the optimization? it would avoid autoloading at least
Feb 17 11:33:49 <Stof> Seldaek: no. The only change was adding the Autoloader to the bootstrap file. It was not generated on the fly before
Feb 17 11:34:20 <avalanche123> is bootstrap.php even necessary?
Feb 17 11:34:28 <fabpot> avalanche123: yes, performance
Feb 17 11:34:34 <avalanche123> were there any benchmarks?
Feb 17 11:34:41 <avalanche123> I assume with performance you'd use APC
Feb 17 11:34:43 <johnwards> I'm not voting, but I think having the envs the same is very important. Then having it well documented on how to disable it for doing debug
Feb 17 11:34:43 <fabpot> avalanche123: yes, plenty
Feb 17 11:34:48 <fabpot> avalanche123: even with APC
Feb 17 11:34:52 <avalanche123> fabpot, I see
Feb 17 11:35:07 <fabpot> for a simple hello world, the difference can be up to 30%
Feb 17 11:35:09 <fabpot> with APC
Feb 17 11:35:24 <avalanche123> fabpot, gotcha, I understand
Feb 17 11:35:35 <Seldaek> well then let's move on to next topic, those that want can still disable it. ack?
Feb 17 11:35:47 <fabpot> yes, I will close the pull request then
Feb 17 11:35:57 <avalanche123> agree, that doesn't seem like a bc breaking change anyway
Feb 17 11:36:01 <fabpot> and perhaps write a some tuto on how to disable it
Feb 17 11:36:07 <bschussek> short comment
Feb 17 11:36:19 <bschussek> your statement about that 30% in hello world, is that relevant?
Feb 17 11:36:26 <fabpot> bschussek: no
Feb 17 11:36:33 <avalanche123> bschussek, that's the game we need to play
Feb 17 11:36:40 <fabpot> except for all the blog posts that try to benchmark frameworks
Feb 17 11:36:46 <avalanche123> yup
Feb 17 11:36:48 <fabpot> and I don't want to suffer like we did with symfony1
Feb 17 11:36:51 <bschussek> yes sure, but we're talking about dev env right?
Feb 17 11:37:07 <fabpot> bschussek: all the envs should work the same
Feb 17 11:37:17 <fabpot> that's one of the big pbe in symfony1
Feb 17 11:37:19 <bschussek> ok
Feb 17 11:37:23 <fabpot> things work in dev and not in prod
Feb 17 11:37:52 <avalanche123> next?
Feb 17 11:38:09 <fabpot> yes
Feb 17 11:38:19 <fabpot> cache warming
Feb 17 11:38:34 * ce_afk is now known as cescalante
Feb 17 11:38:39 <fabpot> I'm aware that the current way the cache warming works is not ideal
Feb 17 11:38:46 <fabpot> so, I will work on it next week
Feb 17 11:38:52 <fabpot> I don't see a need to talk about it now
Feb 17 11:39:06 <Seldaek> nope, there's https://github.com/fabpot/symfony/pull/618
Feb 17 11:39:07 <avalanche123> cool
Feb 17 11:39:18 <Seldaek> just needs an answer:)
Feb 17 11:39:32 <kriswallsmith> fabpot: we should discuss before I write the Assetic cache stuff
Feb 17 11:39:40 <fabpot> kriswallsmith: ok
Feb 17 11:39:50 * unknownbliss has quit (Ping timeout: 240 seconds)
Feb 17 11:39:51 * igorkliks ([email protected]) has joined #symfony-dev
Feb 17 11:40:00 <fabpot> Seldaek: I work as much as I can but everyday is only 24 hours ;)
Feb 17 11:40:23 <Seldaek> fabpot: sure no problem, just pointing it out in case it got lost
Feb 17 11:40:37 <Seldaek> so, johanness are you there?
Feb 17 11:40:42 <fabpot> Seldaek: nothing is lost if it's on github
Feb 17 11:41:05 <fabpot> next topic is eager Response creation
Feb 17 11:41:08 <Seldaek> http://groups.google.com/group/symfony-devs/browse_thread/thread/0d1e05558bcbe275/7cdc3836e9de79b9#7cdc3836e9de79b9
Feb 17 11:41:16 * tecbot has quit ()
Feb 17 11:41:24 <Seldaek> you were already -1 on the ml though iirc
Feb 17 11:41:31 <fabpot> my answer is very clear: if we change that, it's not Symfony2 anymore, so -1
Feb 17 11:41:45 <Seldaek> so it's not happening so skip? :)
Feb 17 11:41:53 <kriswallsmith> sounds like a veto :)
Feb 17 11:41:57 <Seldaek> or someone wants to try and convince fabpot ? :p
Feb 17 11:41:58 <avalanche123> I agree with fabpot
Feb 17 11:42:02 <avalanche123> +1 on not doing that
Feb 17 11:42:02 <weaverryan> +1 to move on
Feb 17 11:42:30 <fabpot> route placeholders vs. GET parameters
Feb 17 11:42:30 <johanness> just as a note, i'm also not really convinced after doing a sample implementation
Feb 17 11:42:43 <Seldaek> http://groups.google.com/group/symfony-devs/browse_thread/thread/8d32595215d65b44
Feb 17 11:42:47 <fabpot> you will hate me, but I'm -1 also
Feb 17 11:42:51 <Seldaek> really
Feb 17 11:42:55 <bschussek> could you explain?
Feb 17 11:42:57 <Seldaek> I think this one actually makes sense
Feb 17 11:43:05 <Seldaek> for flexibility's sake
Feb 17 11:43:21 <fabpot> if you want that, that's possible by registering a listener
Feb 17 11:43:38 <avalanche123> you could still get the parameters from Request::$attributes, no?
Feb 17 11:43:42 <Seldaek> yeah but we know we can do anything we want in our fork.. that's not the point :)
Feb 17 11:43:45 <fabpot> I don't want that as parameters coming from the request must be isolated: POST, GET, PATH INFO, ...
Feb 17 11:44:15 <Seldaek> avalanche123: the problem is just that your controller has to know where it comes from
Feb 17 11:44:19 * ckwalsh (~ckwalsh@phpbb/developer/ckwalsh) has joined #symfony-dev
Feb 17 11:44:28 <avalanche123> Seldaek, make action that doesn't take any parameters
Feb 17 11:44:32 <avalanche123> and use Reqeust::get()
Feb 17 11:44:34 <avalanche123> :)
Feb 17 11:44:35 <Seldaek> fabpot: routing params are kind of get params in hiding though, prettified GET params
Feb 17 11:44:36 <jmikola|w> Seldaek: but a listener could move them to attributes, no?
Feb 17 11:45:09 <bschussek> jmikola|w: I think Seldaek is talking about a potential trap for other users
Feb 17 11:45:20 <bschussek> you think you can easily change the routing, but you can't
Feb 17 11:45:20 <fabpot> Seldaek: and what do you do when you have /foo/:id and the URL is /foo/12?id=foobar
Feb 17 11:45:34 <Seldaek> fabpot: route wins
Feb 17 11:45:39 <Seldaek> otherwise it's security issue
Feb 17 11:45:45 <Seldaek> because the route has the requirements
Feb 17 11:45:50 <Seldaek> query params don't
Feb 17 11:46:14 <jmikola|w> aren't those requirements a contract for the controller, though? (assuming your controller is only accessed through a given route)
Feb 17 11:46:17 <Seldaek> I'm not saying we have to advertise this as best practice, but I think it's nice if it's possible natively
Feb 17 11:46:21 * gordonslondon has quit (Quit: gordonslondon)
Feb 17 11:46:42 <avalanche123> fabpot maybe when we implement the full route template spec, that would be possible?
Feb 17 11:46:54 <avalanche123> as then you could specify requirements on GET parameters as well
Feb 17 11:47:00 <fabpot> avalanche123: I'm not sure how this is related
Feb 17 11:47:28 <avalanche123> well then get parameters can be validate and I don't see harm in injecting them automatically
Feb 17 11:47:35 * rande has quit (Quit: Leaving.)
Feb 17 11:47:44 <avalanche123> but since they are not validate now, it might lead to problems
Feb 17 11:47:45 * kertz has quit (Ping timeout: 246 seconds)
Feb 17 11:47:48 <kriswallsmith> avalanche123: does the spec support that?
Feb 17 11:47:49 <avalanche123> *validated
Feb 17 11:47:58 <fabpot> avalanche123: ok, let's discuss that issue when we have support for the template spec then
Feb 17 11:48:02 <Seldaek> avalanche123: by default params match .+?, so it's not like we prevent injection of anything
Feb 17 11:48:04 * rande ([email protected]) has joined #symfony-dev
Feb 17 11:48:04 <avalanche123> kriswallsmith spec let's you define get paramters for a route
Feb 17 11:48:15 <avalanche123> and routing lets you define validation (requirements)
Feb 17 11:48:17 <avalanche123> kriswallsmith ^
Feb 17 11:48:33 * rande has quit (Client Quit)
Feb 17 11:48:42 <avalanche123> Seldaek, but you could with attributes, not with get parameters
Feb 17 11:48:44 <kriswallsmith> avalanche123: if it's in the spec, that seems like the way to go
Feb 17 11:48:55 <bschussek> so when we'd have this, we could inject GET parameters into the actions right?
Feb 17 11:49:02 <fabpot> anyway, this is something we might be able to add for a 2.1, there is no pressure to have it in 2.0 as BC will be kept
Feb 17 11:49:10 <avalanche123> fabpot +1
Feb 17 11:49:21 <kriswallsmith> I would love to be able to a ?page={page} to my routes
Feb 17 11:49:29 <kriswallsmith> *add
Feb 17 11:49:36 <Seldaek> bschussek: we could already imo, if we consider that /{id} is just /?id=foo beautified (which it is), {id} is a query parameter imo
Feb 17 11:49:42 <avalanche123> kriswallsmith yeah, that's what I mean
Feb 17 11:49:43 <fabpot> as of today, we need to concentrate on things that breaks BC
Feb 17 11:49:50 <bschussek> ok
Feb 17 11:49:51 <Seldaek> so we could treat them similarly, and allow injecting any of them in the action
Feb 17 11:49:54 <avalanche123> so -1
Feb 17 11:50:09 <bschussek> Seldaek: what about implementing that? we can still decide whether and when to merge it
Feb 17 11:50:13 <kriswallsmith> +1 for 2.1
Feb 17 11:50:33 <Seldaek> bschussek: if fabpot is somewhat open to it, sure
Feb 17 11:50:56 <Seldaek> I don't see the reason not to allow GET params to be injected in actions really
Feb 17 11:51:01 <fabpot> Seldaek: I'm somewhat open to it, but not for 2.0
Feb 17 11:51:18 <kriswallsmith> what's next?
Feb 17 11:51:29 <fabpot> we're done
Feb 17 11:51:42 <Seldaek> well, there's 2 left
Feb 17 11:51:48 <fabpot> Seldaek: but no vote on them
Feb 17 11:51:58 <fabpot> generate human readable docs out of bundle Configuration class
Feb 17 11:52:00 <fabpot> is the next one
Feb 17 11:52:04 <kriswallsmith> I'm going to send a pull req for subtree-merging Zend\Log
Feb 17 11:52:06 <fabpot> but there is nothing to discuss
Feb 17 11:52:17 <Seldaek> yes I don't think so either
Feb 17 11:52:21 <bschussek> maybe there's someone here wanting to implement it? :)
Feb 17 11:52:25 <weaverryan> I'm hoping to get to this - but actually getting the merging done in all extensions is more important
Feb 17 11:52:31 <Seldaek> fabpot: and what about that https://github.com/fabpot/symfony/pull/561
Feb 17 11:53:05 <kriswallsmith> is it possible to generate an xsd from a config definition?
Feb 17 11:53:11 <Seldaek> fabpot: is that the "one more thing"? is that why you don't answer? !! :p
Feb 17 11:53:17 <fabpot> lol
Feb 17 11:53:18 <Seldaek> damn secrets
Feb 17 11:53:19 <kriswallsmith> theoretically, at least?
Feb 17 11:53:21 <jmikola|w> kriswallsmith: perhaps for simple things
Feb 17 11:53:29 <jmikola|w> even with doc generation, the issue is closures :)
Feb 17 11:53:48 <bschussek> I'd like to throw in a random topic then
Feb 17 11:53:49 <johanness> kriswallsmith, theoretically it's possible
Feb 17 11:53:54 <bschussek> bug tracking
Feb 17 11:53:54 <fabpot> As these helpers use intl, and as we don't have the replacement yet, I don't see how we will be able to include them in 2.0
Feb 17 11:54:07 <kriswallsmith> it would be nice, to avoid duplication
Feb 17 11:54:08 <fabpot> bschussek: I'm working on installing Jira for 2.0
Feb 17 11:54:15 <bschussek> fabpot: awesome
Feb 17 11:54:20 <fabpot> should be available for the Symfony Live conf
Feb 17 11:54:26 <jmikola|w> for hackday?
Feb 17 11:54:28 <bschussek> cool
Feb 17 11:54:31 <Seldaek> I guess it's not worse than trac :(
Feb 17 11:54:43 <fabpot> Seldaek: any other viable option?
Feb 17 11:54:45 <bschussek> Seldaek: why? you've quite some experience with it right?
Feb 17 11:54:56 <Seldaek> fabpot: no I guess not, they all suck to some extent
Feb 17 11:55:06 <fabpot> Seldaek: correct
Feb 17 11:55:06 <Seldaek> but yeah I've gotten quite experienced at handling the 39302 options of jira
Feb 17 11:55:09 <Seldaek> if you need help
Feb 17 11:55:09 <jmikola|w> from my experience with jira, it just takes some initial time to define workflows and such
Feb 17 11:55:13 <fabpot> the concept of a bug tracker sucks
Feb 17 11:55:32 <bschussek> I'm quite satisfied with pivotal tracker, but that might be too small for our purpose
Feb 17 11:55:33 <johnwards> Githup issue tracker not suitable (i'm interested in why not)
Feb 17 11:55:38 <kriswallsmith> we're using target process at OpenSky
Feb 17 11:55:39 <fabpot> Seldaek: I will probably ask some help then, thanks
Feb 17 11:55:49 <pgodel_work> fabpot: someone was asking if ESI was supported in the symfony http cache
Feb 17 11:55:59 <fabpot> pgodel_work: of course it is
Feb 17 11:56:09 <jmikola|w> kriswallsmith: please no target process :)
Feb 17 11:56:14 <pgodel_work> that's what I thought, he was having some troubles
Feb 17 11:56:19 <kriswallsmith> lol
Feb 17 11:56:20 <Seldaek> fabpot: sure, once configured honestly it's usable. But the problem is most people use it with defaults and there is just too much craziness
Feb 17 11:56:51 <fabpot> Seldaek: how do you find the configuration of the Doctrine Jira instance?
Feb 17 11:57:04 <Seldaek> fabpot: it's quite an outdated version
Feb 17 11:57:11 <fabpot> hehe
Feb 17 11:57:12 <Seldaek> new one is slightly nicer
Feb 17 11:57:21 <bschussek> what about GitHubs issue tracking? any experience with that?
Feb 17 11:57:32 <jmikola|w> pgodel_work: that was jseverson from exercise.com (ornicar could probably chime in on that, too)
Feb 17 11:57:35 * notjosh has quit (Quit: notjosh)
Feb 17 11:57:43 <Seldaek> fabpot: anyway just ping me when you have it running and we can talk
Feb 17 11:57:46 <pgodel_work> jmikola|w: yup
Feb 17 11:57:47 <jmikola|w> bschussek: i think github's is very lightweight
Feb 17 11:58:01 <bschussek> jmikola|w: in a sense of too little features?
Feb 17 11:58:04 <jmikola|w> probably not suitable for such a large project, vs other packages that support github integration
Feb 17 11:58:05 <Seldaek> bschussek: github is really not scaling too well for such a big project
Feb 17 11:58:12 <bschussek> ok
Feb 17 11:58:27 <jmikola|w> i think jira with fisheye plugin or something could interpret commit id's pointed at github
Feb 17 11:58:42 <igorw> there's also lighthouse
Feb 17 11:58:42 * dustinwhittle ([email protected]) has joined #symfony-dev
Feb 17 11:59:00 <Seldaek> jmikola|w: eww fisheye that's going too far for me:p
Feb 17 11:59:18 <kriswallsmith> igorw: a Portland company :)
Feb 17 11:59:18 <pgodel_work> Seldaek: I know doctrine uses it (I think)
Feb 17 11:59:36 <Seldaek> pgodel_work: well, it's just the slowest code browser in the universe, but besides that I'm sure it's great:p
Feb 17 11:59:44 <pgodel_work> lol
Feb 17 12:01:22 <naderman> pgodel_work: working on it as we speak ;-) (dependency stuff)
Feb 17 12:01:33 * inspiran has quit (Ping timeout: 250 seconds)
Feb 17 12:01:45 * bschussek ([email protected]) has left #symfony-dev
Feb 17 12:01:46 <fabpot> naderman: Perhaps you can say a bit more about the current status? and what still need to be done?
Feb 17 12:02:09 <avalanche123> naderman, are you working with everzet?
Feb 17 12:02:16 <naderman> well I have the rule generation working, I'm on the solving part now, I'll push something later tonight or tomorrow
Feb 17 12:02:37 * ornicar ([email protected]) has joined #symfony-dev
Feb 17 12:02:43 <fabpot> naderman: great! Are you still confident about having something for Symfony Live?
Feb 17 12:02:45 <naderman> which should be able to calculate the dependencies, what we then need is actual implementations of repositories/packages
Feb 17 12:02:49 <naderman> fabpot: yes
Feb 17 12:02:57 <fabpot> cool
Feb 17 12:03:10 <fabpot> naderman: sorry about the pressure
Feb 17 12:03:20 <naderman> no that's alright
Feb 17 12:03:24 <naderman> I would have wanted to work more on it
Feb 17 12:03:30 <Seldaek> naderman: if I can help with that let me know, instead of doing logging stuff that nobody wants you know:P though maybe you need logging in the depackager ;)
Feb 17 12:03:33 <naderman> but the last few days after I got back were a bit crazy
Feb 17 12:04:07 <naderman> heh, I'm sure it'll require plenty of work/help once there is a foundation ;-)
Feb 17 12:04:16 <pgodel_work> yup
Feb 17 12:04:23 <Seldaek> aye, don't be shy to share it though
Feb 17 12:04:31 <naderman> not at all
Feb 17 12:05:13 <Seldaek> well I don't see no repo! :)
Feb 17 12:05:21 <fabpot> Seldaek: there is a lot of ground work to do before being able to share -- naderman is porting an existing library (not aPython one this time, but a C one)
Feb 17 12:05:24 * webPragmatist has quit (Quit: Leaving.)
Feb 17 12:05:49 <Seldaek> fabpot: yeah I heard, but anyway I'm just very tired and being annoying.. guess I'll head home
Feb 17 12:05:50 <fabpot> and so, it does not make sense to share things before having ported lots of code
Feb 17 12:06:12 <fabpot> Seldaek: hmmm, I really need to work on Jira to get you some work then
Feb 17 12:06:32 <naderman> heh jira is a lot of fun to set up ...
Feb 17 12:07:08 <naderman> well if you want to adjust the workflow at least
Feb 17 12:07:17 <pgodel_work> you must have a very special meaning of "fun" :P
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment