[00:39:37] *** rmaucher has quit IRC
[00:40:14] *** bgeorges has quit IRC
[00:41:01] *** jpearlin has quit IRC
[00:44:43] *** jpearlin has joined #jboss-as7
[00:45:39] *** bgeorges has joined #jboss-as7
[00:49:09] <Nihility> smarlow: baileyje I think it's important that we differentiate between the jpa provider we care about (hibernate) and the third party stuff
[00:50:50] *** jclingan1 has quit IRC
[00:50:55] <Nihility> smarlow: baileyje in other words, just because portable providers have poor integration due ton spec limitations doesn't mean our code has to
[00:56:48] *** balunasj_mtg has quit IRC
[00:59:32] *** aslak has quit IRC
[01:06:47] *** mbg is now known as mbg|away
[01:09:19] *** torben has joined #jboss-as7
[01:09:20] *** ChanServ sets mode: +v torben
[01:10:26] *** torben has quit IRC
[01:15:49] *** jwulf has joined #jboss-as7
[01:26:59] *** pgier has quit IRC
[01:29:41] <ALR> Any Git experts in the room?
[01:29:54] *** jamezp has left #jboss-as7
[01:29:57] <ALR> I'm looking to import a folder from one repo into another.
[01:30:07] <ALR> While preserving history of course.
[01:30:17] <ALR> For instance I have Repo A.
[01:30:23] <ALR> Repo B has moduleX
[01:30:32] <ALR> I wanna put moduleX into Repo A.
[01:30:46] <ALR> I think "git fetch" is the route in.
[01:30:55] <ALR> Fetching into a branch and moving from there...
[01:31:00] <ALR> But am a bit uncertain. :)
[01:35:16] <dmlloyd> yeah fetch the original into a branch, check out the branch, move the original into the target path, and then merge it.  I think you'll pretty much be stuck with a merge commit though.
[01:35:24] <dmlloyd> but I think it's warranted in such a case.
[01:35:38] <dmlloyd> baileyje: do you know what the deal is with org.jboss.as.ee.component.AbstractComponentDescription#methodInterceptors vs org.jboss.as.ee.component.AbstractComponentDescription#methodInterceptorSet ?
[01:36:06] <ALR> Hmm http://st-on-it.blogspot.com/2010/01/how-to-move-folders-between-git.html
[01:36:21] <ALR> I can filter out everything
[01:36:39] <ALR> dmlloyd: I'll have ugly merge commits everywhere anyway
[01:36:42] <dmlloyd> ah I've never bothered to do that, always just keeping all history
[01:36:55] <ALR> I plan on squashing down afterwards before merging to master
[01:37:05] <ALR> Depending upon how difficult the rebase --i squash is
[01:37:09] <dmlloyd> well, the great thing about git is you can try it a few different times :)
[01:37:16] <ALR> Oh definitely.
[01:37:24] <ALR> Though gotta admit, SVN makes this stuff easier.
[01:37:44] <ALR> (Because in SVN a "repo" is a much wider concept)
[01:40:00] <stuartdouglas> dmlloyd: I think it was just so there was no looping through a list when figuring out if an interceptor was already applied to the method
[01:49:16] <dmlloyd> just curious why we'd keep the list then
[01:51:00] <stuartdouglas> the set is a String, the list is a InterceptorDescription
[01:57:31] <ALR> Preferred Git GUI?
[01:58:01] * ALR most of the time uses GitHub, but in this case would like to see what's up locally
[01:58:20] <stuartdouglas> I find GitX quite good on the map
[01:58:22] <stuartdouglas> mac
[01:58:28] <ALR> Yeah, Fedora.
[01:58:29] <ALR> :)
[01:58:36] <ALR> Everything's pretty on the Mac.
[01:59:54] <ALR> stuartdouglas: haha https://twitter.com/#!/stuartwdouglas/status/9351147936546817
[02:01:12] <stuartdouglas> was not good enough to justify spending 60 bucks on it
[02:02:20] <ALR> git-gui in the Fedora repo, so I'll try that
[02:02:33] <dmlloyd> git gui?  I don't understand the question :)
[02:02:54] <stuartdouglas> gitk also does the job, not pretty though
[02:03:06] <ALR> Aha
[02:03:08] <ALR> gitk
[02:03:11] <ALR> That's what I was looking for
[02:03:21] <ALR> It launches from git-gui
[02:03:26] <ALR> Perfect
[02:03:38] <ALR> I wanted a GUI to visualize my last commit
[02:11:44] <ALR> dmlloyd: Arquillian has the notion of a test enricher
[02:11:58] <ALR> Thomas made one for MSC, and put it into the Arquillian OSGi branch
[02:12:01] <ALR> Now I've moved that into AS
[02:12:14] <ALR> But I'm wondering if you think it's better served in the MSC tree.
[02:12:15] <ALR> Thoughts?
[02:13:25] <dmlloyd> what would this do?
[02:13:35] <ALR> What does the project do?
[02:13:43] <ALR> The test enricher?
[02:13:50] <dmlloyd> the enricher
[02:14:03] <dmlloyd> we have a fairly comprehensive test suite for MSC as it is
[02:14:08] <ALR> "Enrichment" of a test is the process of doing things like injection.
[02:14:12] <Nihility> i could use some riches
[02:14:19] <ALR> Nono, this isn't to help test MSC
[02:14:26] <ALR> It's to make tests MSC-aware
[02:14:48] <ALR> For instance the CDI enricher handles @Inject stuff into tests.
[02:15:04] <ALR> (For CDI beans)
[02:17:40] <dmlloyd> sounds like what we'd normally do in a setup method
[02:18:05] <ALR> Arquillian is death to @Before
[02:18:10] <stuartdouglas> so does it let you inject services into tests?
[02:18:14] <ALR> Yes.
[02:18:19] <ALR> Exactly.
[02:18:35] <ALR> Or, that's it's intended goal.  I've no idea if the current thing works.
[02:18:44] <ALR> But it's a good starting point, so I'm moving it.
[02:19:07] <ALR> (Also because I'm retiring the jbosgi ARQ fork)
[02:19:18] <ALR> So everything is moving into some more appropriate upstream component.
[02:19:25] *** bstansberry has quit IRC
[02:19:37] <dmlloyd> for now putting it into AS is fine, until we know if it's useful and what it really provides
[02:19:42] <ALR> The question is if we want to bind this bit to the MSC release cycle (hence introducing a dependency upon ARQ) or leave it in AS.
[02:19:58] <ALR> That's fine by me.
[02:20:14] <ALR> A better definition:
[02:20:15] *** fnasser has quit IRC
[02:20:30] <ALR> This is the thing that makes ARQ MSC-aware when it comes to bolstering tests with injection.
[02:22:10] *** jclingan has joined #jboss-as7
[02:23:49] *** bstansberry has joined #jboss-as7
[02:23:50] *** ChanServ sets mode: +v bstansberry
[02:27:47] *** jclingan1 has joined #jboss-as7
[02:30:57] *** jclingan has quit IRC
[02:38:31] *** jclingan1 has quit IRC
[02:42:47] *** jclingan has joined #jboss-as7
[02:51:26] *** ccrouch has quit IRC
[02:51:59] *** emuckenhuber has quit IRC
[03:03:31] *** jclingan has quit IRC
[03:04:21] *** bgeorges has quit IRC
[03:05:28] *** bgeorges has joined #jboss-as7
[03:10:13] <smarlow> Nihility:  the question in my mind is, should we change Hibernate to work with VFS URLs?  Or should we improve the URL that we are passing to the Persistence Provider to work with the current Hibernate codebase?
[03:14:38] *** ALR has left #jboss-as7
[03:14:41] *** ALR has joined #jboss-as7
[03:14:41] *** ChanServ sets mode: +v ALR
[03:16:54] *** miclorb has joined #jboss-as7
[03:49:07] *** lgao has joined #jboss-as7
[03:52:29] *** stelios_RH_Athen has joined #jboss-as7
[03:54:59] *** stelios_RH_GR has quit IRC
[04:05:16] *** bgeorges has quit IRC
[04:07:37] *** lgao has quit IRC
[04:46:16] *** jpearlin has left #jboss-as7
[04:52:38] <Nihility> smarlow: how do you mean improve
[04:53:03] <smarlow> hmm
[04:54:21] <smarlow> I didn't mean improve the actual URL but deal with getting the VirtualFile to work with Hibernate by making a copy (yech) and passing a File based url
[04:55:06] <smarlow> on the flip side, we could support only hibernate 4.0 and make sure hibernate 4.0 works with the Virtualfile based url
[04:55:19] *** miclorb has quit IRC
[04:55:50] <smarlow> and we could do the same for a new  3.something Hibernate release, if one is planned
[04:57:23] <Nihility> how did the work on 6
[04:57:29] <Nihility> baileyje: you around?
[04:57:41] <smarlow> In AS6 we didn't use the JPA spi and instead used a native Hibernate data structure, let me find it...
[04:58:18] <Nihility> ah can we just do that then
[05:03:40] <smarlow> I need to refresh my memory on how org.hibernate.ejb.packaging.PersistenceMetadata got handled.  It just has a Set of jar file names (Strings) passed in it looks like.
[05:04:40] <smarlow> so, maybe there is some hack I can take from AS6...  Will hunt that down
[05:22:08] <smarlow> the relevant code is not much different in as6
[05:23:15] <smarlow> its using the same VFS api calls, I wonder if something changed in VFS
[05:24:46] <smarlow> VirtualFile.toURL
[05:38:14] <smarlow> hmm, nothing has changed in vfs.  Maybe its our use of it that has changed
[05:39:19] <Nihility> that url is designed to handle directory scans even
[05:40:30] <smarlow> maybe i can get more answers stepping through the hibernate native scanner that seems to have trouble with it
[05:44:08] <smarlow> Hibernate version is 3.6.1.Final for both, VFS is the same.  Its gotta be me :)
[05:44:27] <smarlow> but the as7 code is based on the as6 code, its very simillar
[05:51:04] <jbossbot> git [jboss-dmr] push master 39829c8.. Jason T. Greene Let objects convert to properties
[05:51:04] <jbossbot> git [jboss-dmr] push master URL: http://github.com/jbossas/jboss-dmr/compare/41f5a26...39829c8
[05:53:20] <jbossbot> git [jboss-as] push master 07d3ffc.. Jason T. Greene Let objects be properties
[05:53:20] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/7c18132...07d3ffc
[05:55:02] <smarlow> JarVistorFactory.getVisitor checks the URL protocol for "jar" and "file"  why not check for "vfs"
[06:01:26] *** Nihility has quit IRC
[06:14:09] *** bstansberry has quit IRC
[06:24:00] *** stelios_RH_GR has joined #jboss-as7
[06:24:01] *** stelios_RH_Athen has quit IRC
[06:25:57] *** smarlow has quit IRC
[06:39:01] *** mbg|away is now known as mbg
[07:05:49] *** mbg has quit IRC
[08:28:10] *** stelios has joined #jboss-as7
[08:31:16] *** stelios_RH_GR has quit IRC
[08:31:16] *** stelios has quit IRC
[09:53:10] *** pil-dinner has joined #jboss-as7
[11:23:58] *** emuckenhuber has joined #jboss-as7
[11:24:02] *** emuckenhuber has quit IRC
[11:24:02] *** emuckenhuber has joined #jboss-as7
[11:24:02] *** ChanServ sets mode: +v emuckenhuber
[11:24:32] *** emuckenhuber has quit IRC
[12:00:09] *** aslak has joined #jboss-as7
[12:00:09] *** ChanServ sets mode: +v aslak
[12:55:31] *** aslak has quit IRC
[13:37:02] *** bstansberry has joined #jboss-as7
[13:37:02] *** ChanServ sets mode: +v bstansberry
[14:16:23] *** aslak has joined #jboss-as7
[14:16:23] *** aslak has joined #jboss-as7
[14:16:23] *** ChanServ sets mode: +v aslak
[14:26:27] *** pil-dinner has quit IRC
[14:31:02] *** pferraro has joined #jboss-as7
[14:31:02] *** ChanServ sets mode: +v pferraro
[14:31:06] *** pferraro has quit IRC
[14:45:37] *** jwulf has quit IRC
[14:54:26] *** smarlow has joined #jboss-as7
[14:54:26] *** ChanServ sets mode: +v smarlow
[15:08:57] *** mbg has joined #jboss-as7
[15:09:05] *** ChanServ sets mode: +v mbg
[15:10:00] *** mbg has quit IRC
[15:17:53] *** mbg has joined #jboss-as7
[15:17:53] *** ChanServ sets mode: +v mbg
[15:37:23] *** mbg has quit IRC
[16:14:28] *** ALR has quit IRC
[16:15:46] <smarlow> baileyje:  I think I found the missing ingredient in as7 JPA, I need to bring in HackTLScanner from AS6...
[16:24:21] <smarlow> we might actually be able to use jandex to do the annotation scanning after all :)
[16:24:38] * smarlow will hack on this tonight...
[16:28:54] <smarlow> we just have to set "hibernate.ejb.resource_scanner" to our annotation scanner...
[18:29:42] *** pferraro has joined #jboss-as7
[18:29:42] *** ChanServ sets mode: +v pferraro
[18:29:44] *** pferraro has quit IRC
[23:12:10] *** ALR has joined #jboss-as7
[23:12:10] *** ChanServ sets mode: +v ALR
[23:18:31] <stuartdouglas> dmlloyd: I have an initial implementation of parsing env-entry from web.xml at https://github.com/stuartwdouglas/jboss-as/tree/env-entry
[23:18:42] <stuartdouglas> it is also in my master as lots of little commits
[23:19:00] <stuartdouglas> It still needs a lot of work, but I wanted to check that you are happy with the approach
[23:19:44] <stuartdouglas> the biggest change is that now java:module and java:app bindings are 'owned' by the module / app rather than individual components
[23:20:44] <stuartdouglas> If you are happy with this approach I will try and implement all the deployment descriptor jndi binding stuff today
[23:59:23] *** smarlow has quit IRC

   March 26, 2011  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >