[00:22:31] *** stuartdouglas has left #weld-dev
[00:25:40] *** stuartdouglas has joined #weld-dev
[00:37:54] *** alesj has quit IRC
[00:56:34] *** PeteRoyle has joined #weld-dev
[06:25:52] *** magesh has joined #weld-dev
[06:58:49] *** magesh has quit IRC
[07:01:18] *** magesh has joined #weld-dev
[07:13:47] *** oskutka has joined #weld-dev
[07:18:15] *** PeteRoyle has quit IRC
[07:35:18] *** alesj has joined #weld-dev
[07:49:23] *** PeteRoyle has joined #weld-dev
[08:01:02] *** mkouba has joined #weld-dev
[08:14:15] *** mbg has joined #weld-dev
[08:20:04] *** ge0ffrey has joined #weld-dev
[08:42:11] *** jharting has joined #weld-dev
[08:51:13] *** maschmid has joined #weld-dev
[08:51:39] *** emmanuel has joined #weld-dev
[08:58:42] *** PeteRoyle has quit IRC
[09:01:09] *** PeteRoyle has joined #weld-dev
[09:03:34] *** oskutka has quit IRC
[09:08:21] *** kevinpollet has joined #weld-dev
[09:08:51] *** PeteRoyle has quit IRC
[09:15:33] *** PeteRoyle has joined #weld-dev
[09:19:44] *** struberg has joined #weld-dev
[09:22:36] *** PeteRoyle has quit IRC
[09:24:04] *** PeteRoyle has joined #weld-dev
[09:24:17] *** mathieuancelin has joined #weld-dev
[09:40:19] *** PeteRoyle has quit IRC
[09:44:05] *** oskutka has joined #weld-dev
[09:52:51] *** PeteRoyle has joined #weld-dev
[10:07:37] *** PeteRoyle has quit IRC
[10:12:17] <stuartdouglas> alesj: I talked to the osgi guys, and got their branch sorted out with the re-formatting
[10:12:33] <stuartdouglas> They are just checking it now, and then I will push the reformatting changes
[10:12:40] <alesj> yeah, saw the email
[10:12:54] <alesj> but yeah, go ahead once they are done
[10:13:20] <alesj> mathieuancelin has done a really good job with the osgi stuff
[10:13:29] <alesj> at least from reading his judcon slides :-)
[10:13:48] <mathieuancelin> alesj: thanks
[10:14:07] <alesj> mathieuancelin: do you still need me to discuss something ? or did the emails explain it all?
[10:14:17] <mathieuancelin> stuartdouglas: for the record everything seems okay
[10:15:08] <mathieuancelin> stuartdouglas: I'm passing our test suite right now and it seems to work
[10:16:05] <mathieuancelin> alesj: just want to know what to do about @OSGiService and the embedded API
[10:16:18] <alesj> i'm now fine with both
[10:16:25] <alesj> i see the embedded api value
[10:16:32] <alesj> and also why @OSGiService is good
[10:16:50] <alesj> or, i think, time will tell
[10:16:59] <alesj> e.g. will the users like the @OSGiService or not
[10:17:08] <alesj> i could definitely live w/o it :-)
[10:17:16] <mathieuancelin> :)
[10:17:22] <alesj> but it's your call
[10:17:27] <alesj> as you are more into this
[10:17:43] <alesj> although, i wrote similar stuff from OSGi and MC :-)
[10:17:45] <mathieuancelin> yeah me too, but maybe it's too magical w/o it
[10:18:01] <alesj> and i didn't need any marker there
[10:18:03] <mathieuancelin> :)
[10:18:09] <alesj> hence, i still feel like we don't need it
[10:18:13] <mathieuancelin> ok
[10:18:30] <alesj> i agree, that osgi is pretty dynamic
[10:18:35] <alesj> and this could then cause problems
[10:18:42] <alesj> or, it could be too computation heavy
[10:18:47] <mathieuancelin> thats's my main concern
[10:18:53] <alesj> to check all injection points against all osgi services
[10:18:59] <mathieuancelin> yep
[10:19:35] *** PeteRoyle has joined #weld-dev
[10:19:44] <mathieuancelin> and as I said in my email
[10:19:52] <alesj> perhaps make it optional
[10:20:07] <alesj> e.g. little # service == easy to check them all
[10:20:07] <mathieuancelin> i'm not sure we can check at startup if services are available in the registry
[10:20:19] <alesj> yeah
[10:20:31] <alesj> hmmmm ...
[10:20:44] <alesj> perhaps still fill in a proxy
[10:20:48] <mathieuancelin> the services can comes later
[10:20:55] <alesj> which would die at execution time
[10:21:00] <mathieuancelin> yep
[10:21:06] <alesj> which is not what CDI is about ...
[10:21:13] <alesj> as you would like to validate things ...
[10:21:18] <mathieuancelin> yep
[10:21:25] <alesj> hence @OS makes sense
[10:22:05] <alesj> what about if we just checked at validation
[10:22:13] <alesj> if @OS is there, it could come in later
[10:22:18] <alesj> if @OS is not there, fail
[10:22:24] <mathieuancelin> yep, but we can't guarantee that a particular service will be available
[10:22:31] <alesj> true
[10:22:35] <alesj> but that's users fault
[10:22:39] <mathieuancelin> :)
[10:22:43] <mathieuancelin> totally
[10:22:49] <alesj> @OS makes him aware of that chance
[10:22:56] <alesj> no @OS ? fail
[10:22:56] <mathieuancelin> yep
[10:23:07] <alesj> + some call back would be useful
[10:23:17] <alesj> if that OSGi service goes away ? later on
[10:23:23] <mathieuancelin> what will be the behavior with no @OS ?
[10:23:25] <alesj> to at least notify the users
[10:23:33] <alesj> no @OS == plain CDI
[10:23:35] <mathieuancelin> it's a standard bean right ?
[10:23:39] <mathieuancelin> ok
[10:23:43] <alesj> no ...
[10:23:47] <mathieuancelin> that's how it works right now
[10:23:54] <alesj> i would still do injection from ServiceRegistry
[10:24:12] <alesj> i mean
[10:24:23] <alesj> if the only tthing that resolves that IP
[10:24:28] <alesj> is some OSGi service in SR
[10:24:35] <alesj> then just use it from SR
[10:24:48] <alesj> it's true that that service could go away any time
[10:24:54] <alesj> hence some callback ...
[10:24:56] <alesj> dunno how
[10:24:58] <alesj> :-)
[10:25:06] <stuartdouglas> I think I will push the code reformat later on tonight
[10:25:13] <alesj> stuartdouglas: ok
[10:25:24] <stuartdouglas> actually I could just push it now I suppose
[10:25:33] <mathieuancelin> stuartdouglas: ok
[10:26:12] <mathieuancelin> yeah but how can we differentiate standard CDI injection inside the bundle
[10:26:24] <mathieuancelin> and OSGi injection w/o @OS ?
[10:26:52] <mathieuancelin> if no @OS and no matching bean type then it should be a plain injection from OSGi SR ?
[10:27:24] <jbossbot> git [core] push master 472b728.. Stuart Douglas Re-format java files to jboss community style
[10:27:24] <jbossbot> git [core] push master 1131e1d.. Stuart Douglas Re-format pom.xml files
[10:27:25] <jbossbot> git [core] push master ee40d25.. Stuart Douglas WIP checkstyle plugin
[10:27:25] <jbossbot> git [core] push master 1acdccd.. Stuart Douglas Fix checkstyle errors
[10:27:25] <jbossbot> git [core] push master URL: http://github.com/weld/core/compare/1a60420...1acdccd
[10:27:45] <stuartdouglas> boom, done
[10:27:55] <mathieuancelin> stuartdouglas: :)
[10:28:16] <stuartdouglas> I want to reformat the API and SPI as well, but that can wait until we start on the next branch, 2.x or 1.2 or whatever
[10:28:21] <stuartdouglas> as it won't change before then
[10:28:43] <stuartdouglas> And I can't be bothered trying to sync it to the osgi changes, cause that was a PITA :-)
[10:29:30] *** PeteRoyle has quit IRC
[10:29:47] <alesj> :-)
[10:31:05] *** oskutka has quit IRC
[10:31:56] <jbossbot> git [core] push master 7bfbf40.. Stuart Douglas remove old eclipse profile
[10:31:56] <jbossbot> git [core] push master URL: http://github.com/weld/core/compare/1acdccd...7bfbf40
[10:32:42] <mathieuancelin> :)
[10:33:00] <stuartdouglas> I have sent round and email
[10:33:51] <alesj> where?
[10:34:06] <alesj> ah, ok, got it
[10:35:11] *** PeteRoyle has joined #weld-dev
[10:41:51] *** PeteRoyle has quit IRC
[10:42:59] *** PeteRoyle has joined #weld-dev
[10:51:11] *** PeteRoyle has quit IRC
[10:52:54] *** oskutka has joined #weld-dev
[10:53:37] *** PeteRoyle has joined #weld-dev
[10:55:08] *** mathieuancelin has left #weld-dev
[10:55:20] *** mathieuancelin has joined #weld-dev
[11:25:57] <mathieuancelin> stuartdouglas: just pushed my new branches based on your work
[11:26:05] <mathieuancelin> stuartdouglas: it works :)
[11:26:16] <mathieuancelin> stuartdouglas: thank you again
[11:28:57] <alesj> cool!
[11:36:17] <mathieuancelin> alesj: what's the plan now ?
[11:37:07] <alesj> mathieuancelin: does your osgi branch pass all tests?
[11:37:23] <alesj> or, what's the deal with those context stuff you've been asking stuartdouglas?
[11:38:01] <mathieuancelin> all the tests pass on weld
[11:38:27] <mathieuancelin> because it uses the existing singletonproviders
[11:38:42] <alesj> ah, ok
[11:38:47] <mathieuancelin> everything should work as before
[11:38:56] <alesj> and how does one then setup your singleton provider?
[11:39:57] <mathieuancelin> SingletonProvider.initialize(new RegistrySingletonProvider());
[11:40:33] <mathieuancelin> the context stuff i asked stuartdouglas was for AS-7 integration
[11:40:52] <mathieuancelin> to use RegistrySingletonProvider
[11:41:11] <mathieuancelin> instead of AS-7 singleton provider based on TCCL
[11:41:39] <mathieuancelin> but right now I still have issues with the Java EE related stuff
[11:42:55] <mathieuancelin> the app server is supposed to pass deploymentId to the Java EE component like WeldListener, etc ...
[11:43:12] <mathieuancelin> so they can find the right weld container
[11:43:32] <mathieuancelin> stuart told me to use JBossWebMetadata
[11:43:36] <mathieuancelin> to pass it
[11:43:58] <mathieuancelin> but it doesn't work with my code
[11:47:04] <alesj> any idea why not?
[11:48:39] <mathieuancelin> no it's very strange
[11:49:02] <mathieuancelin> I even try to pass the value as an init param in the weld integration of as 7
[11:49:10] <mathieuancelin> and when weld starts
[11:49:16] <mathieuancelin> the param isn't found
[11:51:06] <mathieuancelin> same with JBossWebMetadata
[11:51:29] <mathieuancelin> and last week I had not much time to investigate
[11:52:56] <stuartdouglas> mathieuancelin: it may be an ordering issue, you need to make sure you set in the correct phase
[11:53:43] <mathieuancelin> i try to add it in WebIntegrationProcessor in the deploy method
[11:54:35] <stuartdouglas> hmm, that should work
[11:57:00] <mathieuancelin> I'll run the test again
[11:57:42] <mathieuancelin> but when I walk through the servletContext in weldListener
[11:58:05] <mathieuancelin> I can't see value that were in the JBossWebMetadata
[11:59:03] <stuartdouglas> mathieuancelin: is your code on github?
[12:02:43] <mathieuancelin> not yet for this part
[12:02:55] <mathieuancelin> i'm pushing it right now
[12:05:58] *** emmanuel has quit IRC
[12:09:21] <mathieuancelin> https://github.com/mathieuancelin/jboss-as/blob/registrySP/weld/src/main/java/org/jboss/as/weld/deployment/processors/WebIntegrationProcessor.java
[12:13:54] <stuartdouglas> mathieuancelin: That adds it as an init param of the filter, rather than as a context param
[12:14:07] <stuartdouglas> also it looks like you have been merging rather than rebasing
[12:14:22] <mathieuancelin> ?
[12:14:40] <mathieuancelin> yeah I tried two thing
[12:14:49] <mathieuancelin> CPF.getInitParam().add(deploymentId);
[12:14:58] <mathieuancelin> to see if I can get it from init params
[12:15:05] <mathieuancelin> and also
[12:15:27] <mathieuancelin> https://github.com/mathieuancelin/jboss-as/blob/registrySP/weld/src/main/java/org/jboss/as/weld/deployment/processors/WebIntegrationProcessor.java#L126
[12:15:35] <mathieuancelin> in metadata
[12:16:31] *** epbernard has joined #weld-dev
[12:16:31] *** epbernard is now known as emmanuel
[12:16:45] <stuartdouglas> you are you trying to get the param value?
[12:17:46] <mathieuancelin> https://github.com/mathieuancelin/core/blob/weld-osgi-checkstyle/impl/src/main/java/org/jboss/weld/servlet/WeldListener.java#L142
[12:18:08] <mathieuancelin> maybe i'm totally wrong about how it should work
[12:18:36] <stuartdouglas> mathieuancelin: try ServletContext.getInitParameter()
[12:19:10] <stuartdouglas> rather than getAttribute
[12:21:05] <stuartdouglas> mathieuancelin: And change it to how you first had it: https://github.com/mathieuancelin/jboss-as/commit/c11eb221cea97b540dfb8bf5fe08f8ef7ec8ab99#L0R144
[12:21:05] <jbossbot> git [jboss-as] c11eb22.. Mathieu ANCELIN Add integration with web layer
[12:22:18] <mathieuancelin> ok
[12:29:39] *** magesh1 has joined #weld-dev
[12:33:03] *** magesh has quit IRC
[12:39:32] <mathieuancelin> wow
[12:39:37] <mathieuancelin> seems to work :)
[12:39:47] <mathieuancelin> stuartdouglas: thanks
[12:39:59] <stuartdouglas> np
[12:40:22] <mathieuancelin> just a last question
[12:40:33] <mathieuancelin> if I want to do the same with ejb
[12:40:40] <mathieuancelin> and invocationContext
[12:41:07] <mathieuancelin> Jsr299BindingsInterceptor
[12:41:25] <mathieuancelin> or DelegatingInterceptorInvocationContext
[12:41:42] <mathieuancelin> ?
[12:43:35] <stuartdouglas> Jsr299BindingsInterceptor
[12:44:13] <stuartdouglas> you should be able to just pass it in from the factory constructor
[12:44:40] <mathieuancelin> ok
[12:44:45] <mathieuancelin> I'll try that
[12:45:00] <mathieuancelin> thank you again
[12:45:05] <stuartdouglas> no problem
[12:58:58] *** mkouba has quit IRC
[13:13:53] *** mkouba has joined #weld-dev
[13:25:50] *** pmuir has joined #weld-dev
[13:25:50] *** pmuir has quit IRC
[13:25:50] *** pmuir has joined #weld-dev
[13:39:18] *** rmartinelli has joined #weld-dev
[13:42:25] *** alesj has quit IRC
[13:50:11] *** alesj has joined #weld-dev
[15:23:12] *** magesh1 is now known as magesh
[16:02:33] *** oskutka has quit IRC
[16:17:53] *** mbg has quit IRC
[16:17:58] *** mbg has joined #weld-dev
[16:22:15] *** jharting has quit IRC
[17:02:15] *** mkouba has quit IRC
[17:03:32] *** rruss has joined #weld-dev
[17:57:29] *** mathieuancelin has quit IRC
[17:59:20] *** kevinpollet has quit IRC
[18:02:23] *** alesj has quit IRC
[18:04:10] *** aslak has joined #weld-dev
[18:04:10] *** aslak has joined #weld-dev
[18:10:14] *** kevinpollet has joined #weld-dev
[18:12:31] *** mathieuancelin has joined #weld-dev
[18:15:02] *** kevinpollet has quit IRC
[18:16:13] *** magesh has quit IRC
[18:23:29] *** oskutka has joined #weld-dev
[18:25:01] *** mathieuancelin has quit IRC
[18:25:34] *** mathieuancelin has joined #weld-dev
[18:46:35] *** kevinpollet has joined #weld-dev
[18:47:17] *** maschmid has quit IRC
[18:48:02] *** ge0ffrey has quit IRC
[18:52:32] *** oskutka has quit IRC
[19:23:07] *** cbrock has joined #weld-dev
[19:27:48] *** cbrock has quit IRC
[19:27:57] *** cbrock has joined #weld-dev
[19:48:59] *** emmanuel has quit IRC
[20:04:41] *** epbernard has joined #weld-dev
[20:04:41] *** epbernard is now known as emmanuel
[20:05:04] *** emmanuel has quit IRC
[20:06:34] *** kevinpollet has quit IRC
[20:30:56] *** alesj has joined #weld-dev
[20:42:55] *** kevinpollet has joined #weld-dev
[22:08:08] *** pmuir has quit IRC
[22:47:26] *** jbossbot has quit IRC
[22:47:31] *** dmlloyd has quit IRC
[23:42:14] *** jbossbot has joined #weld-dev
[23:43:44] *** dmlloyd has joined #weld-dev





top