[00:05:11] *** stansilvert has quit IRC
[00:16:09] *** aslak has quit IRC
[00:19:33] *** jclingan has quit IRC
[00:27:31] *** smcgowan has quit IRC
[00:28:16] <baileyje> dmlloyd: https://github.com/baileyje/jboss-as/commit/025bc37e3fea90416aa30d5dde3f143e540fdebb
[00:28:17] <jbossbot> git [jboss-as] 025bc37.. John E. Bailey [JBAS-8978] - Implement the initial @Asynchronous parsing
[00:28:17] <jbossbot> jira [JBAS-8978] Complete async EJB invocation [Open (Unresolved) Task, Major, John Bailey] https://issues.jboss.org/browse/JBAS-8978
[00:29:47] *** maxandersen has joined #jboss-as7
[00:29:47] *** ChanServ sets mode: +v maxandersen
[00:31:01] <Nihility> man i love it how we have so much of our stuff back in the main codebase
[00:31:12] <Nihility> its really easy to jump between subsystems now
[00:32:14] <dmlloyd> baileyje: is that against my branch?
[00:32:14] *** maxandersen has quit IRC
[00:32:20] <baileyje> dmlloyd: Yeah
[00:32:54] <dmlloyd> I guess we could probably put this in upstream
[00:33:02] <dmlloyd> now that there's something to it
[00:33:43] <dmlloyd> wow, you can have @Async at the class level?
[00:33:48] <baileyje> Yeah...
[00:33:50] <dmlloyd> weird
[00:34:00] <baileyje> I didn't know that either till I was checking out the spec
[00:34:24] <dmlloyd> I don't see any validation though.
[00:34:36] <dmlloyd> if a method is marked async and it doesn't return void or Future it's an error
[00:34:36] <baileyje> dmlloyd: On the method sigs?
[00:34:39] <dmlloyd> yeah
[00:34:49] <baileyje> Yeah.  I am adding it..
[00:35:20] <baileyje> I was more interested in having you look at the piping of the async through the desc, config, component entry
[00:35:32] <dmlloyd> oh oka
[00:35:32] <dmlloyd> y
[00:35:45] <dmlloyd> it seems reasonable
[00:35:54] <baileyje> THere is no way to validate the class level until later
[00:36:09] <dmlloyd> yeah that makes sense
[00:36:22] <baileyje> Maybe we can just validate them all when the config is being created from the desc
[00:36:54] <dmlloyd> yeah
[00:37:22] <baileyje> I will move it there and rebase for you again
[00:37:27] <Nihility> how is that supposed to work
[00:37:38] <Nihility> async at the class level, injection
[00:37:42] <Nihility> with a proxy?
[00:38:00] <dmlloyd> async at the class level presumably means, all business methods are async
[00:38:12] <dmlloyd> of course if any of them do not return void or Future then the class is in error
[00:38:22] <dmlloyd> presumably that means that injection has to be done into fields only
[00:38:54] <baileyje> " If @Asynchronous is applied at the class level, all business methods declared on that specific class are asynchronous."
[00:39:02] *** adietisheim has quit IRC
[00:40:53] <Nihility> interesting
[00:41:40] <dmlloyd> but it appears to apply just to a view, not to the bean methods themselves.
[00:41:59] <dmlloyd> which is just fine, and compatible with our arch, just interesting.
[00:43:03] <dmlloyd> hmmm, interceptors apply to the bean method right?  not the view method?
[00:43:08] <dmlloyd> @Interceptor I mean
[00:44:19] <dmlloyd> we can do it both ways, theoretically, but right now it's only implemented to operate on the bean class
[00:45:51] <stuartdouglas> Nihility: can you merge my master? It has some tests for the @Local thing, and some JAX-RS fixes and tests
[00:47:18] <dmlloyd> do you guys think that it's likely for someone to want to serialize a local interface proxy?  or just remote?
[00:49:06] <baileyje> dmlloyd:  https://github.com/baileyje/jboss-as/commit/e155687e66e1d9a459a90ad4b3f15f96ef06a592
[00:49:06] <jbossbot> git [jboss-as] e155687.. John E. Bailey [JBAS-8978] - Implement the initial @Asynchronous parsing
[00:49:07] <jbossbot> jira [JBAS-8978] Complete async EJB invocation [Open (Unresolved) Task, Major, John Bailey] https://issues.jboss.org/browse/JBAS-8978
[00:49:12] <baileyje> back int a bit
[00:50:43] <stuartdouglas> would it be possible to get a hudson job set up that runs the integration tests after every commit
[00:50:58] <stuartdouglas> and if they are broken emails whoever broke them?
[00:51:07] <stuartdouglas> Cause they seem to break fairly regularly
[00:55:57] <Nihility> dmlloyd: my thinking is that you would want to serialize local interfaces only if you are doing session replication
[00:56:29] *** alesj has left #jboss-as7
[00:56:36] <dmlloyd> we really need to come up with an EJB client lib soon
[00:56:49] <Nihility> dmlloyd: the use case would be a colocated pool of servers
[00:57:12] <Nihility> dmlloyd: and one goes down, you fall over to a replicated node
[00:58:04] <Nihility> dmlloyd: this of course implies that you need a session identifier and some kind of global registry thats shared on multiple nodes for the same deployment
[00:58:37] <dmlloyd> yeah, session replication
[00:58:50] <dmlloyd> it should be doable, I was just curious
[00:59:04] <dmlloyd> a local interface is really just a remote interface with a couple fewer pieces
[00:59:06] <Nihility> perhaps bstansberry can explain how it used to work, unless pferraro is around
[00:59:08] <dmlloyd> no pass-by-value for example
[00:59:22] <dmlloyd> all the hard stuff is in the EJB implementation part
[01:00:52] <Nihility> then there is also the passivation case of course
[01:02:05] <dmlloyd> yeah, that really applies to the bean instance though
[01:02:23] <bstansberry> in 5.x/6.x the proxy was serializable
[01:02:32] <bstansberry> i don't think it mattered local or remote
[01:04:11] <dmlloyd> okay this is all starting to make sense
[01:05:46] <dmlloyd> I've pulled the async stuff into my master, baileyje.  Tomorrow I think I'm going to make a couple of small changes that should solve opalka, wolfc, your, and smarlow's problems all at once...
[01:05:57] <dmlloyd> and I'm going to also javadoc it all very heavily...
[01:06:30] <dmlloyd> I'm quitting for the night but I'll be working late tomorrow without a doubt.  And probably saturday as well.
[01:06:44] <dmlloyd> by monday I want to have the component stuff and the error reporting stuff all sorted out...
[01:08:29] <dmlloyd> okay new rule
[01:08:40] <dmlloyd> no more renumbering phases by anyone except for the person tagging a release ;)
[01:08:54] * dmlloyd is tired of pointless conflicts in there...
[01:09:31] <Nihility> we could just have a perl script or something do that
[01:09:45] <Nihility> phase-cleanup :)
[01:12:30] <dmlloyd> okay I thought better of putting it in my master and instead put it in a new JBAS-8978_async branch (name changed because I am too easily confused :)
[01:12:34] <dmlloyd> and rebased
[01:12:39] <dmlloyd> will continue tomorrow
[01:12:41] <dmlloyd> adios!
[01:12:41] <jbossbot> jira [JBAS-8978] Complete async EJB invocation [Open (Unresolved) Task, Major, John Bailey] https://issues.jboss.org/browse/JBAS-8978
[01:12:52] <dmlloyd> well done jbossbot!
[01:12:57] <dmlloyd> a bit slow to the punch though!
[01:13:13] <stuartdouglas> Nihility: did you get a chance to look at my master?
[01:15:41] <Nihility> stuartdouglas: yes its building now
[01:17:27] *** jpearlin has quit IRC
[01:17:43] *** jpearlin has joined #jboss-as7
[01:20:55] *** pferraro has joined #jboss-as7
[01:20:55] *** ChanServ sets mode: +v pferraro
[01:32:15] *** jamezp has left #jboss-as7
[01:33:56] <stuartdouglas> So I just started looking at setting up env entries from web.xml
[01:34:23] <stuartdouglas> And the thought occurs that for components that don't have their own comp namespace
[01:34:45] <stuartdouglas> the jndi bindings should probably belong to the module, rather than to individual components
[01:35:57] <stuartdouglas> because the deployment descriptor can be used to set module wide bindings, that will override the bindings set by components
[01:36:33] <stuartdouglas> and it will also make it much easier to detect when two components try and bind incompatible resources to the same place
[01:37:39] <rmaucher> stuartdouglas: shouldn't the injection and instance manager code from AS 6 be ported ?
[01:37:46] <rmaucher> and then we're done ?
[01:38:10] <stuartdouglas> no, that is all being handled differently in AS7
[01:38:29] <stuartdouglas> all the annotation based injection etc is already done
[01:38:58] <stuartdouglas> but none of the deployment descriptor stuff has been implemented yet
[02:02:19] *** jwulf has quit IRC
[02:06:43] *** pferraro has quit IRC
[02:10:27] *** jwulf has joined #jboss-as7
[02:26:05] <bstansberry> Nihility: I've got a couple commits on https://github.com/bstansberry/jboss-as/commits/JBAS-9100
[02:26:14] <jbossbot> jira [JBAS-9100] Flesh out management representation of sockets [Open (Unresolved) Task, Major, Emanuel Muckenhuber] https://issues.jboss.org/browse/JBAS-9100
[02:37:12] <jbossbot> git [jboss-as] push master cc9bbd0.. Stuart Douglas Add bindings to java:module and java:app for EJB's with a single interface
[02:37:12] <jbossbot> git [jboss-as] push master fce3d62.. Stuart Douglas Add test that local interface are handled properly
[02:37:13] <jbossbot> git [jboss-as] push master 15ec78b.. Stuart Douglas Fix composite index bug
[02:37:13] <jbossbot> git [jboss-as] push master 8a2a377.. Stuart Douglas Add integration tests for JAX-RS 1.1 2.3.2
[02:37:13] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/a95e704...8a2a377
[02:37:57] <Nihility> bstansberry: they look good to me, did you want me to test them?
[02:38:26] <bstansberry> if you want; if not that's fine
[02:39:40] <bstansberry> time for some deployment scanning
[02:39:43] <Nihility> ok
[02:40:22] <bstansberry> are you going to test them?
[02:40:54] * bstansberry notes that the JBAS JIRA count has increased by nearly 100 in a week!
[02:41:37] <Nihility> yeah thats why i dont want to unarchive 6
[02:41:55] *** rmaucher has quit IRC
[02:42:02] <Nihility> yeah ill spot test
[02:42:06] <Nihility> give me a second
[02:42:22] <bstansberry> fixed-port doesn't work because of JBAS-9140
[02:42:24] <jbossbot> jira [JBAS-9140] Incorrect value for boolean parameter [Open (Unresolved) Bug, Major, Alexey Loubyansky] https://issues.jboss.org/browse/JBAS-9140
[02:42:34] <bstansberry> well, if you use HTTP it should work
[02:43:08] <Nihility> i keep forgetting that git shll allows me to tab complete branches
[02:55:19] <jbossbot> git [jboss-as] push master 70d53c5.. bstansberry at jboss dot com [JBAS-9139] Use the correct constants
[02:55:20] <jbossbot> jira [JBAS-9139] Eliminate dependency of domain-controller module on server module [Open (Unresolved) Task, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9139
[02:55:20] <jbossbot> git [jboss-as] push master 9f26805.. bstansberry at jboss dot com [JBAS-9100] Add socket write-attribute handlers
[02:55:22] <jbossbot> jira [JBAS-9100] Flesh out management representation of sockets [Open (Unresolved) Task, Major, Emanuel Muckenhuber] https://issues.jboss.org/browse/JBAS-9100
[02:55:22] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/8a2a377...9f26805
[02:55:28] <Nihility> it works
[02:55:29] <Nihility> :)
[02:55:33] <bstansberry> thanks
[03:09:28] *** JimMa has joined #jboss-as7
[03:17:37] *** asaldhan has left #jboss-as7
[03:17:58] <jbossbot> git [jboss-as] push master 6ed234d.. asoldano [JBAS-9138] Initial changes for moving to JBossWS-CXF 4.0.0.Alpha3: cxf update, xmlsec and wss4j modules' addition, further minor fixes
[03:18:02] <jbossbot> jira [JBAS-9138] Upgrade to JBossWS-CXF 4.0.0.Alpha3 [Open (Unresolved) Component  Upgrade, Major, Alessio Soldano] https://issues.jboss.org/browse/JBAS-9138
[03:18:02] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/9f26805...6ed234d
[03:23:38] <Nihility> ok we should be caught up on merging
[03:23:46] <Nihility> let me know if i missed anything
[03:54:36] *** jpearlin has quit IRC
[03:54:36] *** bstansberry has quit IRC
[03:54:38] *** miclorb_ has quit IRC
[03:54:38] *** Nihility has quit IRC
[03:57:44] *** jpearlin has joined #jboss-as7
[03:57:44] *** bstansberry has joined #jboss-as7
[03:57:44] *** calvino.freenode.net sets mode: +v bstansberry
[03:58:22] *** irooskov has quit IRC
[04:00:37] *** miclorb_ has joined #jboss-as7
[04:00:37] *** Nihility has joined #jboss-as7
[04:00:37] *** calvino.freenode.net sets mode: +v Nihility
[04:02:40] *** maxandersen has joined #jboss-as7
[04:02:40] *** ChanServ sets mode: +v maxandersen
[04:06:44] *** mbg has joined #jboss-as7
[04:06:44] *** ChanServ sets mode: +v mbg
[04:10:00] *** lgao has joined #jboss-as7
[04:12:50] *** jclingan has joined #jboss-as7
[04:15:44] <smarlow> dmlloyd:  I saw what you said about about making some changes tomorrow that will help me.  In what way?
[04:19:32] <smarlow> I'm working on some hacks for extended persistence context injection to deal with some ordering issues (injection happens before interceptors are triggered).  So, I'll probably move some of my injection logic (that looks for the current SFSB being invoked) to an interceptor that needs to run before the users postconstruct..
[04:22:57] <smarlow> I wonder if we could construct the ComponentInstance before injectors run in AbstractComponent and make the ComponentInstance available to injectors that need it like me.  :)
[04:23:16] *** bgeorges has joined #jboss-as7
[04:23:16] *** ChanServ sets mode: +v bgeorges
[04:23:54] <smarlow> so, injectors could get it via a AbstractComponent TLS variable
[04:34:20] *** fnasser has quit IRC
[04:39:41] *** maxandersen has quit IRC
[04:43:34] *** jpearlin has quit IRC
[04:44:13] * smarlow I'm hacking the TLS into JPA for now...
[04:45:49] <Nihility> smarlow: you still here
[04:46:07] <smarlow> Nihility:  yep, I'm here
[04:46:13] <Nihility> im seeing that in andy's deployment, enttity beans arent discovered
[04:47:16] <Nihility> about to turn trace logging on but was going to ask if you have seen that
[04:47:22] <smarlow> hmm, I fixed a bug earlier this week that didn't get pulled in yet.  If debug/trace logging is on, I get a NPE I think.  are
[04:47:59] <smarlow> if debug/trace is off, you shouldn't be hitting that.
[04:48:15] <smarlow> Its a two line fix if you want to apply it before turn trace on?
[04:48:20] * smarlow looking...
[04:49:46] <smarlow> when you say entity beans aren't discovered, do you mean they don't show up in the verbose hibernate logging about discovered entity beans?
[04:53:00] *** mbg has quit IRC
[04:53:27] <Nihility> i dont seem to get an npe
[04:53:36] <Nihility> well it doesnt log anything about them
[04:53:42] <Nihility> and hten when they get used
[04:53:43] <Nihility> i get
[04:53:48] <Nihility> "invalid entity"
[04:54:00] <Nihility> and the name queiries arent known either
[04:54:19] *** jwulf has quit IRC
[04:54:21] <Nihility> did your demo work?
[04:55:15] <smarlow> Nihility:   yes, its pretty basic but injects an entity manager and uses it against the entity to insert/select
[04:55:37] <Nihility> hmm
[04:55:44] <Nihility> what if i send you a copy of this
[04:55:53] <Nihility> could you see if you notice anything
[04:55:58] <Nihility> ill keep looking on my en
[04:56:00] <smarlow> sure, that works.  Do I need mysql?  probably not
[04:56:04] <Nihility> maybe its an issue of nested deployment
[04:56:07] <Nihility> i just trick it
[04:56:16] <Nihility> and give it an h2 datasource
[04:56:24] <smarlow> cool, I like that trick
[04:58:26] *** jclingan has quit IRC
[04:58:29] <Nihility> log on to dev01
[04:58:36] <Nihility> look for the zip in /tmp
[04:58:44] <Nihility> EJB3StressTestForAS7
[05:00:29] <smarlow> do I need a full name for dev01?
[05:00:59] <Nihility> dev01.qa.atl2.redhat.com
[05:01:10] <Nihility> do you have access otherwise i can email
[05:02:51] <smarlow> I probably don't, has the app changed since Andy emailed it originally?  I have his email still
[05:03:33] <Nihility> emailed
[05:03:39] <Nihility> make sure you use the latest checkout
[05:07:33] <smarlow> I just pulled the latest as7 source and am building.  Any special setup beyond just copying the ear in standalone/deploy (and maybe creating a marker file if needed)?
[05:09:32] <Nihility> you have to rename the h2d2 ds
[05:09:33] <Nihility> to be
[05:09:37] <Nihility> MySqlDS
[05:09:48] <Nihility> that should be it
[05:10:01] <Nihility> that has andy's eclipse project
[05:10:11] <Nihility> but it uses jboss tools
[05:10:22] <Nihility> to package the deployment
[05:10:39] <Nihility> Project -> Build Packages if you need it
[05:13:48] *** jwulf has joined #jboss-as7
[05:14:02] *** JimMa has quit IRC
[05:17:44] *** JimMa has joined #jboss-as7
[05:18:24] <smarlow> hmm, using org.hibernate.dialect.MySQL5InnoDBDialect, might need to change that for H2
[05:22:31] <Nihility> yeah
[05:22:48] <Nihility> it never got that far to use it though :)
[05:23:43] <smarlow> should I hit http://localhost:8080/Order?
[05:23:45] <smarlow> I got a blank screen but http://localhost:8080/OrderManagerWeb/ gave me something
[05:23:47] <smarlow> really, I didn't get any deploy errors in the console?
[05:24:07] <Nihility> http://localhost:8080/OrderManagerWeb/
[05:24:14] <Nihility> then fill in the form
[05:24:27] <Nihility> you will see a no named query error
[05:24:33] <Nihility> then you can also do
[05:24:46] <Nihility> http://localhost:8080/OrderManagerWeb/CreateData
[05:24:50] <Nihility> and you post
[05:24:53] <Nihility> and invalid entity
[05:25:16] <smarlow> I got http://fpaste.org/ADvN/
[05:25:37] <smarlow> " Named query not found: selectCustomerByPrimaryKey"
[05:25:47] <Nihility> yep
[05:25:54] <Nihility> thats defined on the entity
[05:26:44] <smarlow> pretty cool that we got that far :)
[05:27:00] *** jclingan has joined #jboss-as7
[05:27:14] <smarlow> I mean, we actually got an EntityManager from JPA
[05:27:27] <smarlow> I'll look at the error in hibernate
[05:27:29] <stuartdouglas> Does anyone know of the top of their head, if I have an injection-target-class element in a deployment descriptor
[05:27:38] <stuartdouglas> does the injection happen for sub classes as well?
[05:27:43] <stuartdouglas> I would assume yes
[05:28:19] <Nihility> hmm i dont recall anything like that
[05:29:23] <stuartdouglas> If you have @Resource on a superclass it must be honoured
[05:29:56] <stuartdouglas> but I can't seem to find anything either way about this
[05:30:59] <Nihility> yeah you think it would be equivalent
[05:36:21] <Nihility> i would assume it is identical
[05:36:25] <Nihility> to @resource
[05:36:44] <Nihility> it makes a lot of sense to support it
[05:36:46] <stuartdouglas> sounds good
[05:36:58] <Nihility> ok off to bed
[05:37:00] <Nihility> good night
[05:37:02] <stuartdouglas> at the moment we are not inspecting superclasses for @Resource
[05:37:07] <stuartdouglas> going to have to fix that
[05:37:08] <stuartdouglas> night
[05:37:13] <smarlow> Nihility:  night
[05:42:38] <smarlow> Nihility:  the app issue could be a persistence unit scoping issue
[05:42:52] * smarlow need to debug it
[06:08:46] <smarlow> Nihility:  We can probably work around it but I missed JPA 8.2.1.6.1 (auto discovering annotated classes in the root of the PU).  We need to scan for such classes with Entity, Embeddable, MappedSuperclass annotation
[06:12:10] <smarlow> Nihility:  JBAS-9141
[06:12:12] <jbossbot> jira [JBAS-9141] handle JPA 8.2.1.6.1 Annotated Classes in the Root of the Persistence Unit [Open (Unresolved) Sub-task, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9141
[06:17:07] *** smarlow has quit IRC
[06:26:52] *** bstansberry is now known as bstans_zzz
[06:58:18] *** bgeorges has quit IRC
[07:34:55] *** miclorb_ has quit IRC
[07:53:54] *** opalka has joined #jboss-as7
[07:53:54] *** opalka has joined #jboss-as7
[07:53:54] *** ChanServ sets mode: +v opalka
[07:54:11] <opalka> morning
[07:57:21] *** alesj has joined #jboss-as7
[08:14:58] *** AndyTaylor has joined #jboss-as7
[08:14:58] *** ChanServ sets mode: +v AndyTaylor
[08:38:51] *** AndyTaylor has quit IRC
[08:47:04] *** Jaikiran has joined #jboss-as7
[08:47:04] *** ChanServ sets mode: +v Jaikiran
[09:02:52] *** jclingan has quit IRC
[09:05:27] *** miclorb has joined #jboss-as7
[09:05:45] *** pilhuhn has joined #jboss-as7
[09:05:45] *** pilhuhn has joined #jboss-as7
[09:05:45] *** ChanServ sets mode: +v pilhuhn
[09:09:17] *** jcosta has joined #jboss-as7
[09:18:51] *** torben has joined #jboss-as7
[09:18:51] *** torben has joined #jboss-as7
[09:18:51] *** ChanServ sets mode: +v torben
[09:23:39] *** AndyTaylor has joined #jboss-as7
[09:23:39] *** ChanServ sets mode: +v AndyTaylor
[09:28:46] *** slaboure has joined #jboss-as7
[09:34:05] *** slaboure has quit IRC
[09:34:10] *** emuckenhuber has joined #jboss-as7
[09:34:10] *** emuckenhuber has joined #jboss-as7
[09:34:10] *** ChanServ sets mode: +v emuckenhuber
[09:40:34] *** wolfc has joined #jboss-as7
[09:40:34] *** ChanServ sets mode: +v wolfc
[09:56:21] *** lgao has quit IRC
[09:57:20] *** lgao has joined #jboss-as7
[10:01:18] *** stalep has quit IRC
[10:08:55] *** slaboure has joined #jboss-as7
[10:14:18] *** rmaucher has joined #jboss-as7
[10:14:52] *** adietisheim has joined #jboss-as7
[10:22:30] *** tdiesler has joined #jboss-as7
[10:22:30] *** ChanServ sets mode: +v tdiesler
[10:31:37] *** maeste has joined #jboss-as7
[10:31:38] *** ChanServ sets mode: +v maeste
[10:41:59] *** miclorb has quit IRC
[10:52:45] *** opalka has quit IRC
[11:02:41] *** tdiesler has quit IRC
[11:07:01] *** stalep has joined #jboss-as7
[11:12:53] *** pmuir has joined #jboss-as7
[11:12:53] *** ChanServ sets mode: +v pmuir
[11:21:10] *** ge0ffrey has joined #jboss-as7
[11:21:35] <ge0ffrey> hi guys,  i am working on drools/guvnor. I am trying to build a guvnor war that deploys to jboss AS 7 beta
[11:22:01] <ge0ffrey> if I take the war which deploys to jboss AS 6, I get this: Caused by: java.lang.IllegalArgumentException: Failed to determine resource binding value for BindingDescription{bindingName='org.jboss.seam.async.TimerServiceDispatcher/timerService', bindingType='javax.ejb.TimerService', referenceSourceDescription=org.jboss.as.ee.component.LazyBindingSourceDescription@8f5944}
[11:24:15] *** sannegrinovero has joined #jboss-as7
[11:24:16] *** sannegrinovero has quit IRC
[11:24:16] *** sannegrinovero has joined #jboss-as7
[11:24:16] *** ChanServ sets mode: +v sannegrinovero
[11:32:19] <Jaikiran> ge0ffrey: does it have @Resource TimerService on some field?
[11:34:25] <ge0ffrey> Jaikiran: not in our code
[11:34:38] <ge0ffrey> Jaikiran: but we use seam 2, so maybe that indirectly depends on it
[11:34:43] <Jaikiran> looks like seam has it org.jboss.seam.async.TimerServiceDispatcher
[11:34:45] <Jaikiran> yeah
[11:34:54] <ge0ffrey> I'll see if we use that
[11:34:55] <Jaikiran> currently TimerService isn't implemented in AS7
[11:35:16] *** miclorb_ has joined #jboss-as7
[11:35:39] <ge0ffrey> ah k, so no hope of getting it to work on 7 beta?
[11:35:50] <ge0ffrey> to bad, jboss as 7 starts up so fast
[11:37:32] <ge0ffrey> doesn't look like we use that TimerServiceDispatcher anywhere
[11:37:49] <ge0ffrey> might upgrading to seam 3 affect this?
[11:39:23] <Jaikiran> any references to jboss-seam.jar from within the application?
[11:39:53] <Jaikiran> i guess some annotation scanning of that jar is leading to it being identified as a class which needs @Resource injection
[11:41:54] *** darranl has joined #jboss-as7
[11:41:55] *** darranl has joined #jboss-as7
[11:41:55] *** ChanServ sets mode: +v darranl
[11:43:30] <ge0ffrey> we use seam and include the seam 2.1.0 jars in WEB-INF/lib
[11:44:59] <Jaikiran> right, that's triggering the annotation scanning
[11:45:09] <Jaikiran> I'm not too familiar with seam
[11:45:35] <Jaikiran> do you require annotation scanning in those jars in WEB-INF/lib?
[11:45:45] <ge0ffrey> yes, TimerServiceDispatcher is in jboss-seam-2.1.0..GA.jar
[11:46:21] <ge0ffrey> good question, it's likely, but not a 100% sure. How can I turn it off for that jar to test if it works?
[11:46:43] <Jaikiran> actually, i was thinking you could try turning it off for all jars in that war
[11:47:08] <Jaikiran> so setting metadata-complete = true in the <web-app> element web.xml file probably might help
[11:47:23] <Jaikiran> (assuming metadata-complete flag is currently handled in AS7)
[11:47:35] <ge0ffrey> doesn't hurt to try :)
[11:47:41] <Jaikiran> sure :)
[11:47:48] <rmaucher> scanning will be done anyway
[11:47:52] <ge0ffrey> the metadata-complete element is part of the 2.5 servlet spec?
[11:48:16] <ge0ffrey> k, no need to try then
[11:48:57] <ge0ffrey> Is there any info about which versions of seam certainly do and which certainly don't work on the beta? or is it to soon for that story?
[11:49:09] <ge0ffrey> soon=early
[11:49:28] <Jaikiran> we haven't seen many users try out seam or other user apps
[11:49:39] <Jaikiran> so it's not 100% known what might work and what might not
[11:49:46] <Jaikiran> currently we have (partial) EJB support
[11:50:00] <Jaikiran> a bit more web support (a lot actually)
[11:50:08] <Jaikiran> and I guess CDI
[11:50:17] <Jaikiran> but other than that, I don't know what else might work
[11:50:27] <ge0ffrey> if you have cdi, you're using weld?
[11:50:32] <ge0ffrey> so seam 3
[11:50:37] *** opalka has joined #jboss-as7
[11:50:37] *** opalka has joined #jboss-as7
[11:50:37] *** ChanServ sets mode: +v opalka
[11:50:58] <ge0ffrey> so we 'd better tackle the seam 3 story before we try to tackle the jboss as 7 story?
[11:51:50] <Jaikiran> yes, weld is currently supported http://relation.to/Bloggers/AS7BetaStreamBegins
[11:57:23] <ge0ffrey> Jaikiran: tnx for the info :) I am making specific jboss as 5 and 6 wars now and we'll make the 7 one after we're on seam 3 or jboss 7 is out (which ever comes first)
[12:18:28] *** tdiesler has joined #jboss-as7
[12:18:28] *** ChanServ sets mode: +v tdiesler
[12:22:42] *** frainone has joined #jboss-as7
[12:22:42] *** ChanServ sets mode: +v frainone
[12:31:30] *** JimMa has quit IRC
[12:33:19] *** frainone has quit IRC
[12:34:07] *** stansilvert has joined #jboss-as7
[12:39:28] *** aslak has joined #jboss-as7
[12:39:29] *** aslak has joined #jboss-as7
[12:39:29] *** ChanServ sets mode: +v aslak
[12:41:01] *** tdiesler has quit IRC
[12:42:41] *** tdiesler has joined #jboss-as7
[12:42:42] *** ChanServ sets mode: +v tdiesler
[12:42:53] *** miclorb_ has quit IRC
[12:47:47] *** maeste has quit IRC
[12:51:28] <Jaikiran> wolfc: quick question - is MDB working in upstream?
[12:51:32] <Jaikiran> from a user p.o.v
[12:51:41] *** frainone has joined #jboss-as7
[12:51:42] *** ChanServ sets mode: +v frainone
[12:51:42] <wolfc> Jaikiran, to some extend
[12:51:53] <wolfc> There is a demo
[12:51:59] <Jaikiran> ok let me check
[12:52:03] *** hbraun has joined #jboss-as7
[12:52:07] <Jaikiran> i am planning to extend the xml dd support to MDBs too
[12:52:12] <Jaikiran> so just checking where it stands
[12:52:21] <hbraun> emuckenhuber: ping
[12:52:54] <emuckenhuber> hbraun: pong
[12:53:19] <hbraun> emuckenhuber: can you show me en axmple operation that add's a resource? I.e. a new serber -group?
[12:53:27] <hbraun> server-group
[12:54:41] <hbraun> well, maybe I should look at the test cases first
[12:55:38] <Jaikiran> hmm so it's not a typical JMS delivery
[12:55:41] <emuckenhuber> hbraun: well adding a server-group should be a simple op like: http://pastebin.com/Xcemvejg
[12:55:46] <Jaikiran> where you can specify the JMS destination
[12:56:04] <emuckenhuber> although that obviously just creates the server-group without any addtional configuration
[12:56:15] *** lgao has quit IRC
[12:56:28] <hbraun> yeah, that#s what I wonder about
[12:56:29] <emuckenhuber> hbraun: maybe the xml parsers help: https://github.com/emuckenhuber/jboss-as/blob/master/controller/src/main/java/org/jboss/as/controller/parsing/DomainXml.java
[12:56:38] *** lgao has joined #jboss-as7
[12:56:45] <emuckenhuber> at least we create those updates for startup
[12:57:31] <hbraun> tnx, that#s a good starting point
[13:00:28] *** maeste has joined #jboss-as7
[13:00:28] *** ChanServ sets mode: +v maeste
[13:06:37] *** frainone has quit IRC
[13:06:51] *** alesj has quit IRC
[13:07:14] *** smarlow has joined #jboss-as7
[13:07:14] *** ChanServ sets mode: +v smarlow
[13:08:06] *** maeste has quit IRC
[13:10:03] <smarlow> Nihility:  I sent you a workaround in email for JBAS-9141, the workaround is to list the entities in persistence.xml
[13:13:41] *** alesj has joined #jboss-as7
[13:14:18] *** jwulf has quit IRC
[13:16:09] <smarlow> dmlloyd:  one issue that I might have with SFSBCreateInterceptor is ensuring that is triggered before user interceptors.
[13:16:43] * smarlow as a system interceptor would be
[13:18:13] *** balunasj has joined #jboss-as7
[13:18:13] *** ChanServ sets mode: +v balunasj
[13:21:25] *** maeste has joined #jboss-as7
[13:21:25] *** ChanServ sets mode: +v maeste
[13:21:54] *** hbraun has quit IRC
[13:34:38] *** adietisheim has quit IRC
[13:35:59] <Jaikiran> wolfc: just a fyi - implicit no-interface view has been implemented in my local branch. will ask dmlloyd to push it upstream once he's here
[13:36:24] <wolfc> Jaikiran, I'm just fixing it as well :-)
[13:36:34] <wolfc> Let's compare notes
[13:36:43] <stansilvert> Hey guys, should I do pull requests form GitHub or just send mail to the list?  Do pull requests from GitHub forward to the list?
[13:37:00] *** pilhuhn is now known as pil-food
[13:37:24] <Jaikiran> wolfc: https://github.com/jaikiran/jboss-as/commit/df8a81298a24ce070edb6453d2d2f598d9f38feb
[13:37:25] <jbossbot> git [jboss-as] df8a812.. jaikiran Allow for implicit no-interface view on EJBs (EJB3.1 spec, section 4.9.8) and few other minor cosmetic changes
[13:38:49] <smarlow> stansilvert:  I like the GitHub ui for pull requests but we aren't using that (and they don't forward to the list).
[13:38:57] *** adietisheim has joined #jboss-as7
[13:39:31] <smarlow> stansilvert:  sending mail to jbossas-pull-requests at lists dot jboss.org is the official way currently
[13:41:22] <stansilvert> I saw this though: https://github.com/jbossas/jboss-as/pulls
[13:42:17] <wolfc> Jaikiran, I went for another approach, which resulted in less code.
[13:42:38] <Jaikiran> wolfc: it's there in some remote branch?
[13:42:40] <wolfc> Let me get it to pass the proper tests :-)
[13:43:23] <smarlow> stansilvert:  I used that also but ended up requesting a pull on irc (here)  :-)
[13:43:26] <Jaikiran> sure :)
[13:44:13] <stansilvert> Is there any reason NOT to use GitHub for pull requests?  Looks like a more organized way to do it.
[13:45:26] <wolfc> wow 18 seconds for the servlet injection test
[13:47:29] <wolfc> Jaikiran: https://github.com/wolfc/jboss-as/commits/default-local-view
[13:49:44] <Jaikiran> wolfc: that one won't work if the bean isn't annotated, would it?
[13:49:57] <Jaikiran> mine doesn't use jandex for that reason
[13:50:32] <wolfc> it works of description
[13:50:54] <wolfc> Oh you mean it doesn't do xml
[13:52:43] <Jaikiran> i see that BusinessViewAnnotationProcessor uses ClassInfo and other Jandex stuff to detect the absence of implements clause https://github.com/wolfc/jboss-as/commit/a9747e0cdec283d76303bd9843c07ac244b14e8a#L2R194
[13:52:44] <jbossbot> git [jboss-as] a9747e0.. Carlo de Wolf Implement implicit local view EJB 3.1 FR 4.9.7 & 4.9.8
[13:53:01] <Jaikiran> so if the bean isn't annotated with anything (not even @Stateless) then this won't work
[13:53:09] *** asoldano has joined #jboss-as7
[13:53:09] *** asoldano has joined #jboss-as7
[13:53:18] *** asoldano has left #jboss-as7
[13:53:20] *** asoldano has joined #jboss-as7
[13:53:20] *** ChanServ sets mode: +v asoldano
[13:54:29] <Jaikiran> or for that matter if the bean does have a @Stateless annotation and is marked for local/remote views via xml, then this DUP might end up wrongly assigning an implicit no-interface view
[13:54:31] <wolfc> Jaikiran, unit test says it does: https://github.com/wolfc/jboss-as/commit/a9747e0cdec283d76303bd9843c07ac244b14e8a#L4R96
[13:54:31] <jbossbot> git [jboss-as] a9747e0.. Carlo de Wolf Implement implicit local view EJB 3.1 FR 4.9.7 & 4.9.8
[13:54:53] <smarlow> cool, the InterceptorContext private data has the ComponentInstance
[13:54:56] <wolfc> Jaikiran, well it all started in the LocalView processor
[13:55:27] <opalka> Jaikiran, wolfc I see some EJB3 regression on AS7 upstream
[13:55:35] <Jaikiran> opalka: wher/
[13:55:37] <Jaikiran> where?
[13:55:56] <opalka> Jaikiran, wolfc I have @Stateless UserAccountService implements IUserAccountService
[13:56:11] <opalka> Jaikiran, wolfc @Remote interface IUserAccountService
[13:56:23] <opalka> Jaikiran, wolfc This doesn't expose remote view for me :(
[13:56:39] <Jaikiran> opalka: @Remote isn't implemented
[13:56:44] <Jaikiran> so not an regression :)
[13:56:48] <wolfc> Jaikiran, I think we should remove implicit interface from the BVAP and move it to yours
[13:57:03] <opalka> wolfc, Jaikiran weird thing is it used to work before ;)
[13:57:18] <wolfc> opalka, cool, bring that version back :-)
[13:57:21] <Jaikiran> wolfc: yeah, the rest of the BVAP does reduce code
[13:57:28] <opalka> wolfc, LOL
[13:57:37] <opalka> Jaikiran, wolfc OK, i'm excluding it ;)
[13:57:49] <opalka> wolfc, Jaikiran Is there a Jira for 'Implement @Remote' ?
[13:57:50] <Jaikiran> opalka: i guess IUserAccountService was being picked as (default) local
[13:57:58] <Jaikiran> which should still work upstream
[13:57:59] <opalka> Jaikiran, might be
[13:58:19] <opalka> Jaikiran, then it's not the case. it's failing now
[13:58:28] <wolfc> Jaikiran, opalka, no Stuarts patch excluded @Remote
[13:58:51] <opalka> wolfc, what U mean ? ^^^
[13:59:02] * opalka 's reviewing AS7 git history ...
[13:59:28] <wolfc> https://github.com/jbossas/jboss-as/commit/a95e70457588b3028db514b36f43f235458d1947
[13:59:29] <jbossbot> git [jboss-as] a95e704.. Stuart Douglas Update ejb local view annotation processor
[14:00:44] <opalka> wolfc, just found that. I see
[14:01:12] <wolfc> Jaikiran, refactor your INOIVP to ILVP, then I remove ILV from BVAP and merge your ILVP
[14:02:34] <wolfc> Jaikiran, and use AbstractAnnotationEJBProcessor. Strips another 5 lines of logic :-)
[14:03:17] <wolfc> Oh wait this is not an annotation processor
[14:03:24] <Jaikiran> right
[14:03:32] <Jaikiran> and won't work in the absence of ClassInfo :)
[14:04:50] * Jaikiran is refactoring and running tests
[14:06:23] <wolfc> opalka, did you find out how wsInvocation.getJavaMethod is constructed?
[14:06:37] <opalka> wolfc, I'll do it now ;)
[14:07:09] <Jaikiran> wolfc: done https://github.com/jaikiran/jboss-as/tree/ejb3-dd
[14:12:49] *** stansilvert has quit IRC
[14:15:30] *** torben is now known as torben|brb
[14:19:07] *** jpederse has joined #jboss-as7
[14:19:17] *** jpederse has quit IRC
[14:19:17] *** jpederse has joined #jboss-as7
[14:22:48] <smarlow> dmlloyd:  unless your planning to change my (interceptor) world, I have the first round of XPC working.  I would like to squash and get my changes merged in today
[14:24:21] *** lgao has quit IRC
[14:26:03] <smarlow> dmlloyd:  I hacked the PostConstruct ComponentLifecycles to trigger before user PostConstructs, so I should be good for now with that.
[14:26:32] *** AndyTaylor has quit IRC
[14:28:01] *** torben|brb is now known as torben
[14:28:24] *** bgeorges has joined #jboss-as7
[14:29:39] *** jhalliday has joined #jboss-as7
[14:29:48] *** pil-food is now known as pilhuhn
[14:38:16] *** jamezp has joined #jboss-as7
[14:57:29] *** mmoyses has joined #jboss-as7
[14:57:29] *** ChanServ sets mode: +v mmoyses
[14:59:39] *** fnasser has joined #jboss-as7
[15:02:52] *** frainone has joined #jboss-as7
[15:02:53] *** ChanServ sets mode: +v frainone
[15:04:47] *** sbalmos has joined #jboss-as7
[15:05:00] <opalka> wolfc, The Method is propagated to Invocation from CXF internals
[15:05:12] <opalka> wolfc, I'm preparing workspace locally so I can debug it deeply
[15:05:23] <opalka> wolfc, I didn't work with CXF for month or so :(
[15:06:23] *** frainone has quit IRC
[15:06:26] <wolfc> opalka, I think my previous reference is probably correct
[15:06:34] <opalka> wolfc, might be
[15:06:50] <opalka> wolfc, anyway this represents big integration problem, right?
[15:07:03] <wolfc> opalka, not big, but a bummer
[15:07:03] <opalka> wolfc, I mean the fact the method comes from CXF internals ...
[15:07:36] <opalka> wolfc, yes I know. But this is really hard to fix bummer :(
[15:07:37] <wolfc> opalka, it'll probably be fastest then to do a method.invoke instead of trying to translate it
[15:08:14] <Nihility> smarlow: oops! thats kind of important
[15:09:46] <opalka> wolfc, I need to debug it to see if it can be invoked directly ...
[15:09:55] <Nihility> smarlow: thanks for the workaround
[15:10:55] <wolfc> opalka, it can. We just need another sort of interface into Component
[15:11:16] <opalka> wolfc, what kind of interface U have in mind ?
[15:11:34] <wolfc> opalka, I mean API. Let's wait for dmlloyd to wake up.
[15:11:51] <opalka> wolfc, and how about EJB3 interceptors wrapping the invoked methods. They won't work or am I missing something ?
[15:12:18] <opalka> wolfc, or your idea of new API extension counts with that ? ;)
[15:12:22] <wolfc> opalka, method.invoke will probably be on the view proxy and that'll lead into the interceptor chain.
[15:15:07] <opalka> wolfc, it's becoming really interesting ;) I mean we need special kind of view for WS EJB endpoints and it have to behave even really special way ...
[15:16:06] <opalka> Nihility, ping
[15:16:10] <Nihility> opalka: pong
[15:16:15] <opalka> nickarls, hi Jason
[15:16:22] <opalka> Nihility, Hi jason
[15:16:25] *** tdiesler has quit IRC
[15:16:31] <opalka> Nihility, I have a question
[15:16:43] <Nihility> opalka: do you now? :)
[15:17:09] <opalka> Nihility, in AS6 days we used to have PortComponentLinkServlet in jbossws.war (bundle with JBossws console)
[15:17:27] <opalka> Nihility, from my understanding we want to avoid having system archives in deploy directory
[15:17:45] <opalka> Nihility, how to provide this PCLinkServlet to AS7 to meet our requirements?
[15:17:51] <Nihility> good question
[15:17:58] <opalka> Nihility, Or there's no other choice than providing system.war ?
[15:18:39] <opalka> Nihility, This is old JAX-RPC staff, but it's needed for TCK :(
[15:18:55] * opalka 's praying for days when JAX-RPC will be deprecated and removed from AS
[15:19:11] <Nihility> so you guys are ahead of the game
[15:19:12] <Nihility> :)
[15:19:28] <Nihility> i think we want jax-rpc off in our default profile
[15:19:41] <Nihility> but maybe it can be on
[15:19:52] <Nihility> so back to your question
[15:19:57] <Nihility> i see this as a common issue
[15:20:18] <Nihility> i think the solution is probably that the jax-rpc subsystem use some api in the web layer to create the servlet
[15:20:40] <Nihility> there is the servlet 3 create servlet api thing so there must be something there
[15:20:50] *** jfd has joined #jboss-as7
[15:20:50] *** ChanServ sets mode: +v jfd
[15:20:52] <Nihility> rmaucher: you have any thoughts on that
[15:20:54] <opalka> Nihility, OK, I know what U mean
[15:20:59] <opalka> Nihility, This should be doable
[15:21:10] <opalka> Nihility, we're doing the same for EJB3 beans ...
[15:21:48] <opalka> Nihility, So this is your proposal, dmlloyd any objections david ? ^^^ If now we'll go ahead this road
[15:22:00] <opalka> s/now/no/
[15:22:03] <smarlow> Nihility:  Yup, I'll get that fixed soon.
[15:22:24] <rmaucher> Nihility, I recommend creating the appropriate metadata instead
[15:22:46] <opalka> rmaucher, U mean creating WarMetaData and attaching it to DU ?
[15:22:50] <rmaucher> right
[15:23:04] <opalka> rmaucher, yep, this is the way I'm going to do it (the same like for EJB3) ...
[15:23:21] <rmaucher> I got good reviews about this way of doing things in AS 6 from the people who needed to add servlets/filters/whatever
[15:23:54] <Nihility> ok that works great too
[15:25:05] *** sbalmos has left #jboss-as7
[15:26:07] <Nihility> bstans_zzz: you see that hbraun is already sending in patches
[15:26:19] *** bstans_zzz is now known as bstansberry
[15:26:33] <bstansberry> lol; i wondered why no one was bugging me :)
[15:27:01] <bstansberry> didn't check my nick
[15:27:29] <bstansberry> yeah, haven't had a chance to check them yet, but I love patches
[15:27:34] <Nihility> WAKE UP BUELLER!
[15:27:40] <bstansberry> wow, 9:30 already
[15:27:46] <baileyje> bstansberry: Did you get a chance to push in the timeout fix?
[15:27:53] <baileyje> for deployments?
[15:28:26] <bstansberry> my JBAS-9063 stuff?
[15:28:28] <jbossbot> jira [JBAS-9063] Failing rar deployment is not cleaned up [Open (Unresolved) Bug, Major, Brian Stansberry] https://issues.jboss.org/browse/JBAS-9063
[15:28:46] <bstansberry> i pushed that
[15:30:38] <baileyje> Do you think it will affect JBAS-9051 and JBAS-9077?
[15:30:39] <jbossbot> jira [JBAS-9051] FilesystemDeploymentService hangs during deploy and doesn't pick any subsequent deployments [Open (Unresolved) Bug, Critical, Unassigned] https://issues.jboss.org/browse/JBAS-9051
[15:30:40] <jbossbot> jira [JBAS-9077] Deployment failure leads to a server hang [Open (Unresolved) Bug, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9077
[15:31:06] <baileyje> It seems like it should for 9051 to some degree at least.
[15:33:10] <jamezp> Nihility: Switching to woodstox I was successfully able to start a standalone instance on OpenJDK 1.6.0_17.
[15:34:02] <Nihility> i love it when things are that simple :)
[15:34:27] <Nihility> jamezp: if you are up for more fun, we have jndi stuff and brian has deployment stuff
[15:34:51] <jamezp> I'm always up for more. :-)
[15:36:18] <Nihility> opalka: are you guys cool with us switching to woodstox for the stax provider
[15:36:38] <Nihility> opalka: it fixes older openjdks
[15:36:53] <opalka> Nihility, We're fine with that change Jason
[15:37:02] <baileyje> Nihility: What are the current priorities for beta2? There are only a few unassigned Jiras at this point, and I suspect they may be related to something bstansberry has already pushed.
[15:37:04] <jamezp> I'll get this cleaned up and commit it. There is one spot I had to comment out that excluded woodstox. I want to double check why that was. JBAS-8874
[15:37:05] <jbossbot> jira [JBAS-8874] Woodstox dependency causes smoke test failures [Closed (Out of Date) Bug, Major, Thomas Diesler] https://issues.jboss.org/browse/JBAS-8874
[15:37:14] <Nihility> rmaucher, you have any objections
[15:37:15] <Nihility> ??
[15:37:21] <Nihility> oops should be one question mark
[15:37:48] <bstansberry> baileyje: sorry, I'm getting pinged elsewhere
[15:38:01] <rmaucher> any objection to what ?
[15:38:12] <bstansberry> no that work will not fix JBAS-9051
[15:38:13] <jbossbot> jira [JBAS-9051] FilesystemDeploymentService hangs during deploy and doesn't pick any subsequent deployments [Open (Unresolved) Bug, Critical, Unassigned] https://issues.jboss.org/browse/JBAS-9051
[15:38:40] <Nihility> switching the default stax provider to be woodstox, it fixes bad jdks, and is active (has more bugfixes)
[15:39:15] <baileyje> bstansberry: You want me to dig into that?
[15:39:37] <rmaucher> ah ok, nice :) the bugs were worked around, supposedly
[15:40:24] <Nihility> the biggest example of how you might be affected is that there was a spec defect where it did not document whether "" or null should be used on unspecified namespaces
[15:40:47] <bstansberry> baileyje: the concern with JBAS-9051 is you'll probably conflict with what I'm doing for the auto-deploy stuff
[15:40:48] <asoldano> Nihility, switching to woodstox might even be better for us
[15:40:48] <jbossbot> jira [JBAS-9051] FilesystemDeploymentService hangs during deploy and doesn't pick any subsequent deployments [Open (Unresolved) Bug, Critical, Unassigned] https://issues.jboss.org/browse/JBAS-9051
[15:41:00] <rmaucher> ok, good
[15:41:22] <baileyje> bstansberry: Ok.  I will hold off then.  I may have to start grabbing beta3 tasks pretty quick.
[15:42:14] <Nihility> ok we might end up having to pick a version of xerces to set everywhere at some point as well
[15:42:15] <bstansberry> baileyje: 9077 is different
[15:42:22] <Nihility> for now im defferring that
[15:42:39] <bstansberry> what I did on 9063 was to add a timeout to a Future.get()
[15:42:40] <Nihility> baileyje: i really want descriptors in beta2, and as much EE as we can get
[15:42:45] *** ALR has joined #jboss-as7
[15:42:46] *** ChanServ sets mode: +v ALR
[15:42:47] <bstansberry> that's in some utility code
[15:42:50] <bstansberry> for the demos
[15:43:09] <Nihility> baileyje: if you are out of tasks, i have something, want to fix handling of application.xml?
[15:43:24] <bstansberry> for the smoke tests, the Arquillian container is doing the same thing the demo utility was: future.get()
[15:43:27] <baileyje> Nihility: Sure.  What is needed for applicaiton.xml?
[15:43:40] <Nihility> baileyje: it needs to propogate all settings, like the web context
[15:43:41] <bstansberry> application.xml sound better :)
[15:43:49] <dmlloyd> good morning
[15:44:10] <jamezp> Morning.
[15:44:12] <Nihility> baileyje: also deployers are not using the context of the ear name which they should
[15:44:54] <Nihility> we have another critical task which is that JPA is not scanning for annotations! smarlow i think was working on it but he might want help
[15:44:54] <baileyje> So the descriptor is on the deployer, but the war deployments are not using it?
[15:45:14] *** pmuir has quit IRC
[15:45:21] <opalka> rmaucher, is Servlet 3x API, concretely part of the API that allows U dynamically register servlets via Java API working on AS7 ?
[15:45:32] <baileyje> Nihility: How is JPA not scanning for annotations?  Which annotations?
[15:45:41] <Nihility> @Entity etc
[15:45:45] <Nihility> it was not implemented
[15:46:09] <baileyje> ahh..  Entity beans..
[15:46:18] <Nihility> baileyje: no the apllication.xml stuff is just not processed by the ear deployer and its not passed off to other deployers
[15:47:07] *** pgier has joined #jboss-as7
[15:47:07] *** ChanServ sets mode: +v pgier
[15:47:17] <Nihility> baileyje: and the other issue is that like for example, the web deployer, when it generates a default context, it doesnt use the full ear path, which it should
[15:47:29] <Nihility> baileyje: so we need to come up with a generic solution to that
[15:48:13] <baileyje> what specifically do you mean by context, still need more coffee and that is a majorly over used term :)
[15:48:21] <Nihility> URL
[15:48:54] <Nihility> application.xml can specify the servlet context for a web deployment
[15:49:10] <baileyje> Nihility: The application.xml is parsed and attached to the deployment.  What additional processing is needed at the EAR level?
[15:49:41] <baileyje> The URL context info seems like that would be at the WAR level
[15:49:51] <baileyje> but maybe that is the part I am not picking up..
[15:50:32] <Nihility> http://fpaste.org/sVJV/
[15:50:36] <Nihility> line 17
[15:51:14] <baileyje> Yeah.  I get that.  But it would be the web deployments responsibility to use that information when setting up the context correct?
[15:51:38] <rmaucher> opalka, yes Servlet 3 has an API for that, but you need the webapp to be fully deployed
[15:51:56] <opalka> rmaucher, right
[15:51:58] <baileyje> Nihility: Can you send me that jar, or create a JIRA with it attached.  I would like a decent test case
[15:52:23] <opalka> rmaucher, neither Servlet 3 API nor special DU processor creating WarMD is applicable here
[15:52:32] <opalka> Nihility, rmaucher things are not that easy with dynamically creating web apps
[15:52:41] *** pmuir has joined #jboss-as7
[15:52:41] *** pmuir has joined #jboss-as7
[15:52:41] *** ChanServ sets mode: +v pmuir
[15:52:57] <opalka> Nihility, rmaucher we need something special here, probably some third API
[15:53:00] <Nihility> baileyje: possibly
[15:53:06] <opalka> Nihility, rmaucher DUProcessors operate on top of DU
[15:53:13] <baileyje> Nihility: If not, I will create a sample quick.
[15:53:14] <Nihility> baileyje: either we put the logic in an ear processor that passes it off
[15:53:23] <rmaucher> like a virtual deployment that is dynamically created ?
[15:53:24] <opalka> Nihility, rmaucher we need probably to register some special service in MSC
[15:53:33] <Nihility> baileyje: or the web deployer and other deployers are responsible to consider the info
[15:53:40] <opalka> Nihility, rmaucher that would be detected by some JBossWeb MSC listener
[15:53:45] <opalka> Nihility, rmaucher or something like that ?
[15:53:54] <baileyje> IT is much easier to look up the DU stack then down.
[15:54:11] *** torben has quit IRC
[15:54:20] <baileyje> I would vote the others consider it by looking at their parent to see if it has more info then them..
[15:54:33] <asoldano> rmaucher, opalka, Nihility : to me this looks like what is asked in JBAS-8956 ...
[15:54:34] <jbossbot> jira [JBAS-8956] Implement the HttpService ontop of JBossWeb [Open (Unresolved) Feature Request, Major, Emanuel Muckenhuber] https://issues.jboss.org/browse/JBAS-8956
[15:54:35] <Nihility> baileyje: well as an example the ear deployer could create an attachment WEB_CONTEXT_INFO
[15:54:35] <dmlloyd> you can however look down assuming it is properly coordinated (which the INSTALL phase is)
[15:54:37] <smarlow> Nihility, baileyje:  its JBAS-9141
[15:54:38] <baileyje> but I can certainly put something in place to set the context.
[15:54:38] <jbossbot> jira [JBAS-9141] handle JPA 8.2.1.6.1 Annotated Classes in the Root of the Persistence Unit [Open (Unresolved) Sub-task, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9141
[15:55:03] <Nihility> baileyje: im not saying it should be this way, just saying it should be thought about
[15:55:15] <smarlow> dmlloyd:  should I squash my interceptor changes with other jpa changes to be pulled?
[15:55:26] <rmaucher> asoldano, which I'm not going to do, it's a direct hack into the servlet container
[15:55:32] <baileyje> Nihility: Yeah.  I will just start looking into it.  I would be easy to attach it either way
[15:55:48] <rmaucher> you'd be better served using its API to embed directly in terms of functionality I think
[15:56:29] <asoldano> rmaucher, any example to look at for that?
[15:56:29] <opalka> rmaucher, how can we deploy servlet to JBossWeb today? Is there some API for that ?
[15:56:43] <Nihility> ah right so the issue is we have a subsystem
[15:56:47] <Nihility> which has now DU
[15:56:50] <Nihility> no DU
[15:56:57] <Nihility> and the subsystem wants to register a context
[15:57:03] <Nihility> it either has to create a DU
[15:57:07] <Nihility> or use an API
[15:57:14] <rmaucher> the web subsystem is for deploying EE webapps
[15:57:15] <asoldano> Nihility, right..
[15:57:59] <Nihility> asoldano, opalka the service is global for the whole container correct, its not based on a deployment?
[15:58:10] <asoldano> Nihility, right, it's global
[15:59:06] <rmaucher> so it's not designed to deploy any other sort of individual components at the moment
[15:59:25] <Nihility> asoldano: does it have resources, or is it just a servlet(s)?
[15:59:49] <opalka> Nihility, http://fpaste.org/ac9x/
[16:00:05] <ALR> Anyone seen the OSGi guys around?
[16:00:06] <Nihility> ok so its a pure servlet
[16:00:28] <opalka> ALR, yep, Thomas quit 10 minutes ago
[16:00:34] <ALR> opalka: Ah, was on a call.
[16:00:37] <ALR> Thanks.
[16:00:57] <baileyje> Nihility: Is there a JIRA for that web-context issue?
[16:01:03] <baileyje> If not I will create one.
[16:01:07] <Nihility> baileyje: no i was going to create one
[16:01:13] <baileyje> I will add one.
[16:01:14] <Nihility> baileyje: but then you were looking for stuff to do
[16:01:14] <Nihility> :)
[16:01:17] *** stalep has quit IRC
[16:01:25] <opalka> Nihility, rmaucher how about special folder like deployments/system ?
[16:01:35] <opalka> Nihility, rmaucher Tomcat uses the same approach
[16:01:41] *** jclingan has joined #jboss-as7
[16:01:53] <opalka> Nihility, rmaucher or we really want to avoid system deployments although they're required by the EE spec ?
[16:02:03] <Nihility> rmaucher: this  seems like a common concern, osgi for example wants to be able to dynamically register a Context, in the meantime i suppose we could just create DUs
[16:02:15] <smarlow> Nihility, baileyje:  hmm, I'll have to debug as it looks like I should of covered JBAS-9141 already (with a call to PersistenceUnitInfo
[16:02:15] <smarlow> .setPersistenceUnitRootUrl)
[16:02:17] <jbossbot> jira [JBAS-9141] handle JPA 8.2.1.6.1 Annotated Classes in the Root of the Persistence Unit [Open (Unresolved) Sub-task, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9141
[16:03:17] <opalka> Nihility, How can we create DUs using AS7 API and force DU chain execute upon it ?
[16:03:34] <rmaucher> Nihility, well, I have to stick to webapps as the components I deploy, they provide the Servlet's environment and API features
[16:04:14] <rmaucher> opalka, Tomcat does not have any "system" deployments, it uses regular webapps
[16:04:46] <opalka> rmaucher, yes I know, I just called them System, because they're not user deployments
[16:05:26] *** mbg has joined #jboss-as7
[16:05:26] *** ChanServ sets mode: +v mbg
[16:05:29] <jbossbot> git [jboss-as] push master efe5e82.. Heiko Braun Fix JBAS-9142
[16:05:29] <rmaucher> all the "non user" deployments like the manager webapps are still in the user "webapps" folder
[16:05:29] <jbossbot> jira [JBAS-9142] Support "include-runtime" parameter on "read-children-resources" operation [Open (Unresolved) Feature Request, Major, Heiko Braun] https://issues.jboss.org/browse/JBAS-9142
[16:05:30] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/6ed234d...efe5e82
[16:06:07] <baileyje> opalka: A DU is really just a service.  Take a look at AbstractDeploymentUnitService
[16:06:30] * opalka 's having a look ...
[16:07:07] <baileyje> The only issue is it would have an issue if there is no deployment content in the repository.
[16:07:08] <rmaucher> anyway, a Servlet has access to webapp features, even if you use the OSGI thingies, so it's really a webapp with one servlet
[16:07:36] <baileyje> but that may need to be fixed.
[16:08:09] <baileyje> opalka: Does the deployment have an archive or something representing it?
[16:08:40] <opalka> baileyje, yes, it's jbossws.war in AS6
[16:08:48] <opalka> baileyje, U can ignore ContextServlet defined there
[16:08:53] <Nihility> thats an implementation detail though
[16:09:06] <opalka> Nihility, it definitely is
[16:09:21] <Nihility> you could for example use something other than servlet to fill the same need, its  not a generic EE thing, its just something that needs to handle web requests on a url
[16:09:24] <rmaucher> I think you can make it more "virtual" though
[16:09:42] *** balunasj has quit IRC
[16:09:57] *** stansilvert has joined #jboss-as7
[16:09:59] <Nihility> it sounds though like generating a DU is the quickest path forward will reuse the same exissting web logic etc
[16:11:05] <opalka> Nihility, baileyje I will give it a try
[16:11:14] <opalka> baileyje, what is the suggested way of doing it right?
[16:11:28] <rmaucher> Nihility, if you can do that, it's a very nice capability
[16:12:00] <baileyje> opalka: It would be nice if you created a RootDeploymentUnit and attached some metadata and just added the DU service.
[16:12:01] <opalka> Nihility, baileyje I'm afraid of class loading issues in that approach
[16:12:32] <baileyje> That would basically run the deployment chain against it.  It would be nice to not have an archive if all the stuff could just be pre-attached by the subsystem.
[16:12:37] <opalka> Nihility, baileyje to be more precise how modules will hook here ...
[16:13:09] <opalka> baileyje, what U mean with ' stuff could just be pre-attached by the subsystem' ?
[16:13:12] <rmaucher> do you mind if I add the xsd and wsdl mime-mapping to the default list ? they seem legitimate additions
[16:13:27] <Nihility> no not at all
[16:13:29] *** jclingan has quit IRC
[16:13:43] <baileyje> opalka: Yeah.  Like what module deps it needs, maybe some default web metadata, etc
[16:13:57] <Nihility> opalka: you can actually supply that
[16:14:06] <baileyje> but that all depends on what is in the jbossws.war
[16:14:10] <Nihility> opalka: we might want to create a wrapper API at some point to make this easier
[16:14:20] <opalka> baileyje, there is nothing, just web DD ;)
[16:14:21] <Nihility> opalka: so i imagine you want to reuse the subsystem cl
[16:14:42] <opalka> baileyje, and it depends on JBossWS modules that are not provided in that web archive ...
[16:14:47] <baileyje> opalka: They you should be able to just create the JBossWebMetaData and attach it to the DeploymentUnit
[16:14:56] <baileyje> as well as some module deps
[16:15:32] <opalka> baileyje, Nihility OK, I need to try it first so I can grab more information to discuss ...
[16:15:58] <Nihility> opalka: yeah we may need to help, its actually a good way thought to test the system
[16:16:11] <rmaucher> JBossWebMetaData is the "final" metadata, so it's a good idea
[16:16:23] <baileyje> You will also want to attach a deployment root as well.
[16:16:29] <rmaucher> it can be used to configure some servlet container components if needed
[16:16:54] <baileyje> Since the root requires a VFS virtual file, it can be completely virtual
[16:17:29] <baileyje> but, I will stop babbling, and let you just start trying it.  :)
[16:17:42] <rmaucher> if there's a bug with that procedure, I can look at it ;)
[16:17:48] <asoldano> baileyje, Nihility at the end of the day, this would be a programmatic way of deploying to AS7
[16:17:53] <asoldano> right?
[16:17:58] <baileyje> correct
[16:18:06] <Nihility> right you could pretty much do whatever you wanted
[16:18:12] <baileyje> Which we could put a nice wrapper API on it.
[16:18:31] <baileyje> The whole deployment system is built under the assumption deployments can come from anywhere.
[16:19:31] <smarlow> baileyje:  I need to drill in the HIbernate Entity scanner but it looks like they don't like vfs:/content/OrderManagerApp.ear/OrderManagerEJB.jar/
[16:19:52] <opalka> wolfc, baileyje, Nihility, asoldano OK, this is what I'm going to work on until the EJB3 integration issues are automagically resolved in the mean time (right carlo) ;)
[16:19:54] <baileyje> smarlow: Can we swap it out and write a nice jandex version?
[16:20:17] <baileyje> opalka: Ping me with any questions you have along the way
[16:20:25] <opalka> baileyje, sure
[16:20:29] <asoldano> opalka, baileyje, Nihility thanks
[16:20:31] <wolfc> opalka  ?
[16:20:44] <Nihility> rmaucher: it could be useful though to have something similar to the JDK's HttpHandler web server api, for basic cases such as this (where someone doesn't need full EE capabilities, just the ability to accept requests on a context)
[16:20:46] <opalka> baileyje, I'm also thinking about changing my working times to be in sync with US folks
[16:20:59] <Nihility> rmaucher: so please think about that long term :)
[16:21:13] <opalka> wolfc, I just said that I'm going to work on different integration pieces with hope EJB3 WS integration will improve in the mean time ...
[16:21:29] <rmaucher> ok, but it would be like another "web" subsystem probably
[16:21:37] <Nihility> rmaucher: yeah that would make sense
[16:21:55] <wolfc> opalka, well https://github.com/wolfc/jboss-as/commits/ejb3-ws brings in Singleton support
[16:22:13] <wolfc> apart from that we have no explicit failures
[16:22:22] <smarlow> baileyje:  I'm not sure what it would take but we could suggest that for their 4.x release timeframe.
[16:22:25] <opalka> wolfc, yes, I hope it will get upstream soon
[16:22:34] *** asaldhan has joined #jboss-as7
[16:22:34] *** ChanServ sets mode: +v asaldhan
[16:22:43] <rmaucher> Nihility, ok
[16:23:30] <smarlow> baileyje:  do you want to visit #hibernate-dev with me?  I could suggest it and you could maybe answer a few questions?
[16:23:35] <asoldano> rmaucher, Nihility: let me know if something comes out from that long term effort; currently the Endpoint.publish API relies on the httpserver stuff, but I could probably move it upon this new system
[16:23:47] <bstansberry> JBAS-9100
[16:24:17] <bstansberry> sorry, jbossbot, seems JIRA doesn't work for you either ;)
[16:24:47] <wolfc> opalka, https://github.com/wolfc/jboss-as/commits/default-local-view should take care of the @Remote regression, but it needs cleanup and integration with Jaikirans branch
[16:24:58] <opalka> asoldano, Endpoint.publish() shouldn't rely on it anymore
[16:25:06] <opalka> asoldano, I removed it from AS6 months ago
[16:25:18] *** pmuir has quit IRC
[16:25:20] <opalka> asoldano, Endpoint.publish() is meant for client usage only
[16:25:54] <opalka> wolfc, Jaikiran Yes, I reviewed both branches and both seems to incorporate further EJB3 WS fixes
[16:25:59] *** pmuir has joined #jboss-as7
[16:26:00] *** pmuir has joined #jboss-as7
[16:26:00] *** ChanServ sets mode: +v pmuir
[16:26:00] <asoldano> opalka, I'm referring to the cxf side and when running a client in j2ee env...
[16:26:02] <wolfc> dmlloyd, ping
[16:26:09] <dmlloyd> phone, one sec
[16:26:11] <opalka> wolfc, Jaikiran I hope your joint effort will take it to upstream soon too
[16:26:17] <asoldano> opalka, anyway not an issue at all atm
[16:26:26] <dmlloyd> well... more like 90 minutes :)
[16:28:18] <wolfc> opalka, I'll rebase ejb3-ws in the mean time. So you'll see a shift.
[16:28:58] <opalka> wolfc, ok, thanks
[16:32:20] *** baileyje has quit IRC
[16:33:08] *** baileyje has joined #jboss-as7
[16:33:08] *** ChanServ sets mode: +v baileyje
[16:36:04] *** mmoyses has quit IRC
[16:36:33] <smarlow> Nihility:  I have to look closer but it looks like JBAS-9141 is a gray area bug :-)  I actually covered the functionality but Hibernate doesn't get any entities out of vfs:/content/OrderManagerApp.ear/OrderManagerEJB.jar from their scanner
[16:36:34] <jbossbot> jira [JBAS-9141] Redirected to: https://issues.jboss.org/JiraLockedError
[16:36:56] <dmlloyd> heh
[16:37:35] <wolfc> I've an occasional hang in ServletInjectionTestCase
[16:38:35] *** ge0ffrey has quit IRC
[16:45:46] <baileyje> Any idea why I can't access a WAR on my local running server?
[16:45:46] *** stelios has joined #jboss-as7
[16:46:06] <wolfc> baileyje, I have the same problem
[16:46:36] <baileyje> yet, somehow the demos work.
[16:46:39] <wolfc> but it's not fully reproducible
[16:47:25] <wolfc> baileyje, can you save your logs and msc services dump and try again?
[16:49:05] <smarlow> baileyje:  could you take a quick look at PersistenceUnitParseProcessor.getPersistenceUnitURL() and let me know if I could get a non-vfs URL out of that code instead of vfs:/content/OrderManagerApp.ear/OrderManagerEJB.jar/?  That might address JBAS-9141
[16:49:06] <jbossbot> jira [JBAS-9141] handle JPA 8.2.1.6.1 Annotated Classes in the Root of the Persistence Unit [Coding In Progress (Unresolved) Sub-task, Major, Scott Marlow] https://issues.jboss.org/browse/JBAS-9141
[16:50:13] <baileyje> smarlow: In AS6 we had to hack it to force the contents to be fully temp filed out and then use a hack to get the URL.
[16:50:23] <wolfc> smarlow, there are multiple ways to provide a proper URL
[16:50:36] <baileyje> We may need some hack, until the hibernate annotation scanning is plugable.
[16:51:20] <wolfc> hmm I would have to dig deep in memory to remember what kind of urls are acceptable my JPA impls
[16:51:21] <smarlow> I'm likely to have difficulties with other persistence providers as well
[16:51:30] <wolfc> s/my/by/
[16:51:37] <smarlow> wolfc:  I remember that one
[16:52:08] <smarlow> wolfc:  at least I think I remember the code that generated the URL, I think is in the Hibernate jar
[16:52:09] <wolfc> from the top of my head a file: is accepted, so is a JarInputStream
[16:52:34] <wolfc> the problem lies in other JPA impls
[16:52:35] *** ge0ffrey has joined #jboss-as7
[16:55:41] <baileyje> smarlow: http://anonsvn.jboss.org/repos/jbossas/projects/jpa/trunk/vfs3/src/main/java/org/jboss/jpa/deployment/vfs3/PersistenceUnitDeploymentImpl.java
[16:56:18] <wolfc> http://download.oracle.com/javaee/6/api/javax/persistence/spi/PersistenceUnitInfo.html#getPersistenceUnitRootUrl%28%29
[16:59:36] <smarlow> baileyje:  that looks like its also getting a VFS url
[17:00:02] <baileyje> smarlow: I am still looking for the other workaround.
[17:00:13] * smarlow thanks! :)
[17:00:53] <Nihility> baileyje: what do you mean by wars notworking?
[17:01:17] <baileyje> Nihility: I am not able to get to anything on port 8080 on my machine.
[17:01:41] <smarlow> Does anyone have the link for create jandex jiras?
[17:01:43] <rmaucher> pull request: most of the web configuration: https://github.com/rmaucher/jboss-as/commit/0cf196054ef500a6011b6bebc10d443822a1a1a3 https://github.com/rmaucher/jboss-as/commit/1f843110c23ddb82f3c2bce9a99b03d7e6632a26 https://github.com/rmaucher/jboss-as/commit/9a92f8e3bfb50769cf2aba84169db87d20121948 https://github.com/rmaucher/jboss-as/commit/108acd213dbb1f2283831f8fcf5cccc14fb26f68
[17:01:44] <jbossbot> git [jboss-as] 0cf1960.. Rémy Maucherat Add access-log and rewrite parsing, and implement access logging.
[17:01:45] <jbossbot> git [jboss-as] 1f84311.. Rémy Maucherat More configuration, but no SSL yet.
[17:01:49] <jbossbot> git [jboss-as] 9a92f8e.. Rémy Maucherat Finish the main web configuration with SSL. Untested ...
[17:01:49] <jbossbot> git [jboss-as] 108acd2.. Rémy Maucherat New web build and use the new APIs to implement the missing features.
[17:02:06] <baileyje> Haven't been able to in weeks, but just assumed I was using some bogus url
[17:02:55] <Nihility> baileyje: try 127.0.0.1
[17:03:03] <baileyje> That is what I am working with
[17:03:05] <Nihility> the web layer was binding to 0.0.0.0 always
[17:03:13] <Nihility> now it pays attention to the socket stuff
[17:03:33] <Nihility> not that it will do a blank page for missing contexts
[17:03:50] <Nihility> baileyje: did you want a copy of this deployment
[17:04:06] <baileyje> Nihility: I put together a test one..
[17:04:12] <Nihility> oh ok
[17:04:15] <Nihility> thats better
[17:04:21] <wolfc> baileyje, do you get anything or a blank page?
[17:04:40] <baileyje> blank page no matter what I try
[17:04:50] <wolfc> ah okay, I get no reply
[17:05:17] <rmaucher> Nihility, yes, it does use the address now, it's part of the web configuration work
[17:06:07] <Nihility> baileyje: if you deploy the simple war, it does work, just use the context in the log message
[17:06:22] <Nihility> baileyje: also keep in mind that if there is a dep problem the deployment listener doesnt log anything
[17:06:44] <baileyje> Nihility: I am using the context returned by the log message
[17:07:34] <wolfc> baileyje, do you have an url pattern configured?
[17:07:55] <Nihility> baileyje: try the simple deployment
[17:08:29] <baileyje> It doesn't get much simpler than what I have.  :)
[17:09:24] <smarlow> dmlloyd:  you mentioned making some changes that would help me, last night.  Is that going to impact the changes that I want to merge in (XPC/SFSBlifecycle interceptor/SFSB invocation interceptor)?  If not, I'd like to get a merge in
[17:09:33] *** jclingan has joined #jboss-as7
[17:09:48] <rmaucher> for me the test EAR didn't deploy correctly
[17:10:09] *** maeste has quit IRC
[17:10:56] <Nihility> rmaucher: yeah i got that all working now it just uses the wrong context
[17:11:36] <Nihility> baileyje: right im just ssaying try something different to find out if its local to whatever you are doing
[17:11:51] *** mmoyses has joined #jboss-as7
[17:11:51] *** ChanServ sets mode: +v mmoyses
[17:12:20] <rmaucher> it's normal, it is set only from jboss-web.xml right now
[17:12:31] <Nihility> right
[17:14:57] <baileyje> Yeah.  I can't get the demo to work, right now
[17:15:19] <smarlow> Nihility:  Is there a issue tracking system for jandex?  The Hibernate guys were asking
[17:15:44] <baileyje> I will track it down though
[17:15:47] <asaldhan> ALR: ping
[17:16:02] <asaldhan> ALR: for arq what is the difference b/w managed and embedded?
[17:18:50] <ALR> asaldhan: Yo
[17:18:58] <ALR> asaldhan: Embedded brings up in the same VM
[17:19:13] <asaldhan> ALR: I am still struggling to get AS7 working with JPDA
[17:19:13] <ALR> Managed means ARQ is responsible for starting/stopping the server in a new process
[17:19:37] <asaldhan> ALR:
[17:19:37] <asaldhan> @RunWith(Arquillian.class)
[17:19:37] <asaldhan> @Run(RunModeType.AS_CLIENT)
[17:19:44] <asaldhan> ALR: those are my two annotations.
[17:19:51] <asaldhan> ALR: what is it running under now?
[17:19:56] <ALR> asaldhan: OK, so that means that ARQ won't alter the deployment
[17:20:21] <ALR> When ARQ alters the deployment, it does so such that you run the test from inside the container.
[17:20:26] <asaldhan> ALR: but when I run the test, I dont see the breakpoints in AS7
[17:20:40] <ALR> Managed vs. Embedded has to do with which ARQ Container adaptor you have on your ClassPath
[17:21:33] <ALR> asaldhan: Where's your config again?  I don't see you posting on jboss-dev-as7
[17:21:50] <asaldhan> ALR: I created a thread.   AS7 and JPDA
[17:21:59] <asaldhan> ALR: u run eclipse?
[17:22:25] <ALR> Right I remember the Thread
[17:22:33] <ALR> Maybe my Thunderbird search just sucks
[17:22:35] <ALR> One sec
[17:23:25] *** mmoyses is now known as mmoyses_
[17:23:29] <ALR> Found it
[17:24:07] <ALR> asaldhan: Right, so ARQ doesn't do anything special to start the server.
[17:24:14] <ALR> It just uses the config it's given
[17:24:19] <ALR> asaldhan: Where's your test?
[17:24:23] <ALR> Somewhere Githubby?
[17:24:24] <asaldhan> ALR: all I am doing is try to run a junit test with a war being deployed in AS7 and want to break
[17:24:30] <asaldhan> ALR: not checked in yet.
[17:24:43] <asaldhan> ALR: maybe I am setting the wrong break points. :)
[17:24:48] * asaldhan will check that first.
[17:25:07] <ALR> asaldhan: Also "ps aux |grep jboss" to see what AS is launching with
[17:25:28] <ALR> I have very little info to direct you w/ ATM :)
[17:25:56] <ALR> Need to see your config project; if you push it somewhere I can have a look-see.
[17:25:58] <asaldhan> ALR: no problem.  I see the war getting deployed in AS7.  so something is working.  I just need to see the JPDA working.  will be back in 5mins.
[17:26:04] <ALR> Ah O
[17:26:06] <ALR> OK
[17:26:10] <ALR> Hit me back
[17:26:32] <smarlow> asaldhan:  mvn install -Dmaven.surefire.debug="-Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=y,address=8787" worked for me
[17:26:55] <ALR> That sets JPDA on your TEST
[17:26:56] <asaldhan> smarlow: will try that.  but I am trying in eclipse.
[17:27:07] <ALR> Which is fine if you're running AS in Embedded.
[17:27:18] <ALR> But if you're running it managed, AS comes as a new process
[17:27:37] <ALR> And that process is the one that needs JPDA options
[17:27:50] <smarlow> stuartdouglas: responded on that email thread with ^ and also for managed:  -Djboss.options="-Xdebug -Xnoagent -Djava.compiler=NONE -Xrunjdwp:transport=dt_socket,address=5005,server=y,suspend=y"
[17:28:03] <asaldhan> smarlow: right.  ALR: all cool
[17:28:08] * asaldhan will shut up
[17:28:23] * asaldhan was setting bad breakpoints
[17:28:26] <smarlow> I added the comments to my testcase for easy access :)
[17:28:26] <baileyje> Nihility: Yeah.  There is something going on.  If I change the context path name to match what is in the EAR, it doesn't work.  There must be another piece that I am missing.
[17:28:29] <ALR> Hehe
[17:28:36] * asaldhan also missed ALR due to devnexus
[17:28:55] <ALR> asaldhan: Awww, that's sweet, sugar.
[17:29:20] <baileyje> Nihility: Nevermind, I might have it.
[17:29:39] <asaldhan> ALR: missed u because I could have asked about Arq here. :)
[17:29:57] <ALR> asaldhan: #jbosstesting
[17:30:49] <Nihility> smarlow: sorry for the delay
[17:30:54] <Nihility> smarlow: there is not currently
[17:31:04] <Nihility> smarlow: i could request one get made
[17:33:46] <smarlow> Nihility:  their first issue request was for a release of jandex :)
[17:36:23] <opalka> baileyje, ping
[17:36:37] <baileyje> opalka: Hey
[17:37:12] <opalka> baileyje, I have one question regarding JBAS-9082 - namely this commit - https://github.com/jbossas/jboss-as/commit/97652baeb4c46bc092fc009a420442602878d807
[17:37:13] <jbossbot> jira [JBAS-9082] Incorrect handling of runtime failures in composite operations if rollback is not enabled [Resolved (Done) Bug, Major, John Bailey] https://issues.jboss.org/browse/JBAS-9082
[17:37:13] <jbossbot> git [jboss-as] 97652ba.. John E. Bailey [JBAS-9082] - Correctly handle composite operation failures when rollback is not enabled
[17:37:29] <opalka> baileyje, I have a test case that tries to deploy incorrect web application
[17:37:39] <opalka> baileyje, the application deployment have to fail
[17:37:49] <opalka> baileyje, after your changes it deploys even if it fails
[17:38:02] <opalka> baileyje, why this semantic change?
[17:38:39] * smarlow tried to run the testsuite/integration but I'm getting a CNFE on org.jboss.as.testsuite.integration.jpa.epcpropagation.EPCPropagationTestCase
[17:38:51] <baileyje> opalka: How is it doing the deployment?
[17:38:58] *** torben has joined #jboss-as7
[17:38:59] * smarlow must need to inject more classes...
[17:39:07] *** ChanServ sets mode: +v torben
[17:39:10] <baileyje> My guess is the deployment should have the rollback runtime set to true
[17:39:10] <opalka> baileyje, using our RemoteDeployer
[17:39:35] <opalka> baileyje, there's some flag on deployment API I have to set ?
[17:39:43] <baileyje> bstansberry: There?
[17:40:27] <bstansberry> yes
[17:41:18] <baileyje> bstansberry: When we disabled rollback for changes with out the runtime rollback flag...
[17:41:19] <bstansberry> opalka: are you using ServerDeploymentManager?
[17:41:27] <baileyje> yeah, he is
[17:41:28] <opalka> bstansberry, org.jboss.as.webservices.deployer.RemoteDeployer
[17:41:35] <bstansberry> 1 sec
[17:42:54] *** torben has quit IRC
[17:43:16] <bstansberry> opalka: final DeploymentPlanBuilder builder = deploymentManager.newDeploymentPlan().withRollback().add(archiveURL).andDeploy();
[17:43:27] <bstansberry> note the addition of "withRollback()"
[17:43:55] <opalka> bstansberry, can U apply it locally and push upstream?
[17:44:10] * opalka wants to avoid pull request procedure
[17:44:36] <bstansberry> ok
[17:44:40] <opalka> bstansberry, thanks
[17:50:26] <asaldhan> rmaucher: ping
[17:50:38] <Nihility> smarlow:  they don't have to have a dep on jandex if they dont want to
[17:50:51] <Nihility> smarlow: there are some alternative strategies we can do
[17:51:05] <Nihility> smarlow: as an example, we can just give them a last of classes that they can analyze later
[17:51:11] <rmaucher> asaldhan, yes
[17:51:18] <smarlow> Nihility:  I think they wanted to use it standalone and integrated with AS
[17:51:50] <asaldhan> rmaucher: I have a webapp that is FORM secured.  But I enabled break points in both AuthenticatorBase and RealmBase.  None of them are hit.
[17:51:56] <asaldhan> rmaucher: sorry
[17:52:03] <asaldhan> rmaucher: s/realmbase/formauthenticator
[17:52:14] <rmaucher> bad luck ;)
[17:52:22] <asaldhan> rmaucher: is there any valve that is available for all web apps so I can trace
[17:52:38] <jamezp> Nihility: Just an FYI. I've got merge conflicts trying to rebase with the change to woodstox. I'll try to get these resolved tonight and pushed to my master.
[17:52:42] <asaldhan> rmaucher: or some class where I can start in the request pipeline
[17:53:34] *** pmuir has quit IRC
[17:54:33] <jbossbot> git [jboss-as] push master 6449383.. bstansberry at jboss dot com Rollback failed deployments
[17:54:33] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/efe5e82...6449383
[17:54:39] <bstansberry> opalka: ^^^
[17:54:47] <opalka> bstansberry, thanks Brian
[17:54:47] <rmaucher> well, it's either the form authenticator is there or it is not there
[17:55:14] *** hbraun has joined #jboss-as7
[17:55:14] <asaldhan> rmaucher: anything before that?
[17:55:15] *** hbraun has quit IRC
[17:55:28] <asaldhan> rmaucher: I am trying to figure out why the authenticator pipeline is not invoked
[17:56:00] <rmaucher> all valves that are in the pipeline are always invoked
[17:56:48] <asaldhan> rmaucher: top of ur head- do u remember the class that inserts the authenticator?
[17:56:59] <asaldhan> rmaucher: into the pipeline
[17:58:09] <rmaucher> I don't see the point of looking into that, this works
[17:59:27] <asaldhan> rmaucher: I am trying to see if the FormAuthenticator is introduced or not.  If not, why?  I am looking for where that decision is made in JBossWeb
[17:59:47] <asaldhan> rmaucher: it probably is a metadata issue.
[17:59:59] <rmaucher> can you stop adding my name all the time ?
[18:00:21] <asaldhan> rmaucher: WTF? u r JBossWeb. Correct?
[18:00:47] <rmaucher> it makes things beep all the time ...
[18:01:22] <asaldhan> then configure your chat settings or turn off your speakers.
[18:02:56] *** emuckenhuber has quit IRC
[18:03:26] <asaldhan> I have asked you for 5 full minutes to give me pointers in JBossWeb where the authenticator decision is made.
[18:03:45] <opalka> bstansberry, ok, we made little progress. However there's still one unexpected exception - http://fpaste.org/i1TM/
[18:04:19] * opalka have no idea what that exception means
[18:04:36] *** jcosta has quit IRC
[18:04:52] <rmaucher> I'm testing, it's easier
[18:05:08] *** pilhuhn is now known as pil-dinner
[18:05:29] <asaldhan> I can figure it out too. It will save me 20 minutes if you f$cking remember the classname from the top of your head
[18:06:32] <bstansberry> opalka: it means it's trying to undeploy something that isn't registered as deployed
[18:06:44] <rmaucher> so I don't see FormAuthenticator, so it's a problem
[18:06:51] <opalka> bstansberry, aha, then this will be fault on my side
[18:07:04] * opalka have an idea what might be wrong ...
[18:07:21] <asaldhan> I know that the authenticator is not getting introduced.  All I am asking is where in JBossWeb is that decision made?
[18:07:22] *** mmoyses_ is now known as mmoyses
[18:07:22] <bstansberry> good :-)
[18:07:44] *** asoldano has quit IRC
[18:07:58] <asaldhan> if you cannot remember some of your own JBossWeb classes, what is the point?
[18:08:16] <rmaucher> yes, I wonder ...
[18:08:25] <Nihility> dude calm down its all ok
[18:08:29] <Nihility> :)
[18:08:29] *** bstansberry is now known as bstans_afk
[18:08:38] <asaldhan> Nihility: he needs a tampon
[18:08:39] <dmlloyd> or I could just start banning people who are unable to be civil
[18:09:33] <rmaucher> so I'll debug the issue and fix it, since the authenticator is not there, as per my testing, it's not going to work
[18:09:56] <rmaucher> I don't understand why it doesn't work, apparently same code as in AS 6; bad luck
[18:11:15] <rmaucher> ok, do you have security constraints defined somewhere ?
[18:11:30] *** alesj has quit IRC
[18:11:40] <Nihility> dont worry about asaldhan, he is still pissed chicago lost :)
[18:11:43] <asaldhan> dmlloyd: Look through the chat. I have been asking him politely. If he wants to just say "can you test it". I will do it.  But he has not answered any question
[18:11:47] <asaldhan> Nihility: :)
[18:12:32] <asaldhan> Nihility: also u ditched eclipse.;)
[18:12:37] <Nihility> haha
[18:13:00] <asaldhan> rmaucher: I have added some @DeclareRoles and there is a web.xml
[18:13:28] <asaldhan> rmaucher: anyway, carry on with your friday. I am sorry if I spoilt your Friday.
[18:13:33] <rmaucher> declare roles is not enough, you need a constraint
[18:13:43] <asaldhan> rmaucher: I have the constraints in web.xml
[18:14:47] <asaldhan> rmaucher: any carry on. I will figure it out.
[18:15:00] <rmaucher> this works fine for me then
[18:15:28] <rmaucher> however, no constraint, no valve, I had forgotten that optimization
[18:15:49] <asaldhan> Nihility: hope we still have a few eclipse users in as7-dev
[18:16:04] <asaldhan> Nihility: balance is tilting toward intellij
[18:16:27] <Nihility> asaldhan: if they ever implement maven integration outside of m2e I might reconsider
[18:16:57] <asaldhan> Nihility: it should be ok as many of as7 users will be on intellij too.
[18:17:04] <asaldhan> Nihility: also devs/contributors.
[18:17:19] <Nihility> i think bstans_afk is still using eclipse
[18:17:39] <rmaucher> I'm going to stick with Fedora Eclipse
[18:18:18] <opalka> bstans_afk, please cherry pick https://github.com/ropalka/jboss-as/commit/f469bef4e359fa3b318a1b7f367dc000902aff58
[18:18:19] <jbossbot> git [jboss-as] f469bef.. Richard Opalka do not undeploy what's not deployed
[18:20:23] *** fnasser has quit IRC
[18:20:45] *** opalka has quit IRC
[18:21:06] *** fnasser has joined #jboss-as7
[18:22:13] <asaldhan> rmaucher: thanks for sticking with eclipse.  I will dig around and email you if I find anything not working.  Have a good weekend.
[18:23:10] *** bstans_afk is now known as bstansberry
[18:23:43] <rmaucher> stack is looking like this: http://pastebin.com/Amdf8HpJ
[18:24:19] <rmaucher> can you paste your web.xml there ?
[18:24:29] *** Jaikiran has quit IRC
[18:25:40] <asaldhan> let me paste it for you. 1 sec
[18:26:38] *** emuckenhuber has joined #jboss-as7
[18:26:38] *** ChanServ sets mode: +v emuckenhuber
[18:27:08] <asaldhan> http://pastebin.com/aukqHZu4
[18:27:53] <rmaucher> the TCK tests for FORM which don't actually use the realm work ok too: https://hudson.qa.jboss.com/hudson/view/TCK6-AS7-WEB/job/tck6-as7-servlet_web-profile/lastCompletedBuild/testReport/unknownTestSuite.0/packageTests/com.sun.ts.tests.servlet.spec.security.secform/
[18:28:25] <asaldhan> for me the difficulty has been I have not been breakpoint into  jbossweb dependency classes but only into  org.jboss.as.web classes
[18:28:27] <rmaucher> there's no security-constraint in the web.xml: no constraint, no valve
[18:28:46] * asaldhan whacks himself
[18:28:58] <asaldhan> it is in the servlet
[18:29:15] <asaldhan> @ServletSecurity(@HttpConstraint(rolesAllowed = { "gooduser" }))
[18:29:46] <jbossbot> git [jboss-as] push master e62a90e.. Richard Opalka do not undeploy what's not deployed
[18:29:46] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/6449383...e62a90e
[18:29:51] <bstansberry> opalka: ^^^
[18:30:13] <rmaucher> and the servlet needs to have a mapping
[18:30:52] <asaldhan> have that in web.xml
[18:31:04] <Nihility> bstansberry: btw stumbled across an npe the other day, something like touching an isdeploy for a deployment that didnt exist, i meant to fix it but didnt get to it. not asking you to fix it, just letting you know in case someone complains
[18:31:11] <asaldhan> the annotations and the web.xml do not get merged?
[18:31:35] <rmaucher> ok, and if you want to test and not have problems, put the constraint in web.xml then; new features = possible new bugs
[18:31:46] <asaldhan> will do.
[18:32:01] <asaldhan> I will create JIRA issues once I put in a few tests in as7 testsuite
[18:32:03] <rmaucher> so I might have a bug with the @ServletSecurity translation then, will need to review
[18:32:07] * smarlow will squash my branch ask nicely for a pull in about ten minutes :)
[18:32:07] <bstansberry> Nihility: ok; if you see it again send me a stack trace
[18:32:15] <bstansberry> since i'll be in that code
[18:32:45] <asaldhan> Nihility: dude.  currently I am following the Cricket World Cup.  It is like the superbowl of cricket. :)
[18:32:51] <asaldhan> Nihility: still in playoffs.
[18:34:38] <rmaucher> you can send a war, but that was like, the worst possible test; you're also replacing the default servlet (which is supposed to work, but might cause problems with it not being done that often ...)
[18:35:03] <asaldhan> I will add the tests into the testsuite.
[18:35:17] <asaldhan> i need to add tests for form,basic and all the new servlet3 features.
[18:35:38] <Nihility> i thought the spec has restrictions on merging web.xml and servlet annotations
[18:35:42] <rmaucher> I am against adding edge cases in the testsuite
[18:36:07] <rmaucher> the spec has rules when merging, it's pretty complex
[18:36:30] <Nihility> i agree with not having a bunch of edge cases, or single scenarios in a testsuite
[18:36:32] <asaldhan> I need to use web.xml because no annotations for specifying login-config.  only the programmatic login api available in s3
[18:36:36] <rmaucher> so I don't know right now how that stuff is going to work
[18:37:05] <asaldhan> sure. let me start with the classic test cases. then once those are working.
[18:37:14] <asaldhan> we can validate the s3 usecases
[18:37:32] <Nihility> then you end up with thousands of tests that dont necesarily give you better coverage.
[18:38:06] <asaldhan> but we do need tests for pre servlet3, correct? Nihility
[18:38:14] <Nihility> sure totally
[18:38:28] <asaldhan> Nihility: at this time, as7 testsuite has zero web sec tests
[18:38:53] <Nihility> i just mean i agree that our tests should be organized by functional area
[18:39:21] <asaldhan> Nihility: basically the web security tests will have one test for FORM - one for BASIC etc  and then maybe one or two for s3 features.
[18:39:34] <smarlow> baileyje:  I was just looking at VirtualFile.getPhysicalFile(), I'll try using that I think.
[18:39:37] <asaldhan> Nihility: more like unit tests.
[18:39:51] <asaldhan> Nihility: since Remy tests against the tck also, we should have good coverage
[18:39:53] <Nihility> we can also have demos if you like
[18:40:21] <asaldhan> Nihility: at this time, I want to introduce tests for basic functionality.  then when we have sufficient coverage, we can add demo apps.
[18:40:39] <baileyje> smarlow: The only problem is if you are dealing with an archive file, not all the files will be available from that root, unless they have been accessed..  Which will be true if you are past the annotation indexing processor.
[18:42:36] <smarlow> baileyje:  I'm currently accessing the VFS in the PARSE phase, I could move that to the INSTALL phase
[18:43:00] <baileyje> smarlow: Just needs to be after the jandex phase
[18:43:16] <smarlow> I think thats very early, checking
[18:44:41] <baileyje> Nihility: https://github.com/baileyje/jboss-as/commit/0c06d92a814cdb09e7b07703cc08a27e7fc2cc7f
[18:44:42] <jbossbot> git [jboss-as] 0c06d92.. John E. Bailey [JBS-9143] - Properly use the web context-root defined in application.xml
[18:46:30] <smarlow> baileyje:  the STRUCTURE_ANNOTATION_INDEX is earlier than me, is that == jandex phase?
[18:46:39] <baileyje> yeah
[18:47:00] <rmaucher> so where is my test war ? resolving servlet security is complex, and I did rewrite the annotation processing code not too long ago
[18:47:00] *** mbg is now known as mbg|away
[18:48:52] <asaldhan> if I add the constraint to web.xml,  I should have the test working for now.
[18:49:09] <rmaucher> yes, but send me the test war
[18:49:20] <asaldhan> is the annotation processing code in web module?
[18:49:38] <asaldhan> I will send you the war later in the day.
[18:50:53] <rmaucher> yes, but you don't need to look into it
[18:52:30] <asaldhan> sure
[18:53:54] *** jclingan has quit IRC
[18:54:42] <Nihility> rmaucher: IIRC when a context-root was not provided in an ear's application.xml, we would set the context-root to be ear-module-name/war-module-name (to avoid conflicts). The specs seem to leave it open. Do you have an opinion on what we should do in this case?
[18:55:25] <rmaucher> ah it's possible to do that, sure
[18:56:03] <baileyje> Nihility: Yeah.  I just updated to to use the context root if provided.
[18:56:12] <baileyje> Not the case where it isn't..
[18:56:15] <Nihility> baileyje: yeah your patch made me wonder about that
[18:56:44] <baileyje> I will add that as well, but they are really separate things
[18:57:19] <asaldhan> Nihility: rmaucher: but are we allowing customization is maybe jboss-web.xml for the context root?
[18:58:01] <baileyje> asaldhan: yeah.  That is there as well.  It would be overriden by the EAR if defined in application.xml
[18:58:01] <Nihility> baileyje: yeah im not sure who should set it in that scenario
[18:58:03] <asaldhan> Nihility: do we still do the jboss-app.xml customization?
[18:58:48] <asaldhan> Nihility: baileyje:  maybe a strategy for merging may be needed that the user can configure.  But it is going to be complex.
[18:58:51] *** stelios_RH_GR has joined #jboss-as7
[18:59:03] <Nihility> baileyje: my thinking is that the web deployer should do that based on it's nesting which we have to be sure to provide
[18:59:09] <asaldhan> Nihility: web security itself is like one maze that is hard to conquer
[19:00:13] <baileyje> Nihility: So it should add the ear portion in as well?  Is that what you man by "...  deployer should do that..."
[19:00:57] <asaldhan> rmaucher: when I add the constraint to web.xml, I see the breakpoint in FormAuthenticator.  Let me prepare a war for you with my previous issue.
[19:01:09] *** stelios has quit IRC
[19:02:36] <rmaucher> my testing so far indicates @ServletSecurity works fine though, even without a constraint in web.xml
[19:03:01] <asaldhan> it works because there is no security involved
[19:03:14] <asaldhan> there is no authenticator injected.  It is a security hole.
[19:03:33] <asaldhan> but test case has one success and one 403 case and I was struggling with the 403 case.
[19:03:43] *** slaboure has quit IRC
[19:05:53] *** ge0ffrey has quit IRC
[19:08:20] <smarlow> baileyje:  the contents of the VirtualFile.getPhysicalFile() is empty.  Does that mean that jandex didn't scan it?
[19:08:51] <baileyje> hmm.  I guess for jars it may stay empty..
[19:09:08] <baileyje> since it only needs input streams which it gets from the JarFile
[19:12:01] <rmaucher> dmlloyd, can you do the pull I requested earlier ?
[19:13:41] <smarlow> baileyje:  seems like we either need to hack it in JPA code (if thats possible) or in VFS.  Another hack idea for the JPA side, would be to read the contents and create a file system copy somewhere, return a url to that.
[19:14:02] <rmaucher> ok, but about the breakpoint, that's probably not a good test; at this point, the interesting item is if there's an AuthenticatorBase on the stack
[19:14:10] <dmlloyd> on the phone again...
[19:14:28] <rmaucher> then the quality of the translation to a constraint can be checked
[19:14:28] <smarlow> baileyje: or change the VFS url to work with the Hibernate scanner code
[19:15:45] <smarlow> baileyje:  maybe a URL wrapper that just makes it work
[19:17:01] <asaldhan> I may have some issues with my test.  I am using the latest httpclient code to do form auth and there seems to be some issue there.
[19:18:58] <rmaucher>  I don't care, send me the war; if there's a bug, it's probably because you map your servlet as the default servlet, and it gets no mapping instead
[19:19:18] *** stelios_RH_Athen has joined #jboss-as7
[19:22:06] *** stelios_RH_GR has quit IRC
[19:25:36] <Nihility> baileyje: nah i think it makes sense for the ear deployer to set application.xml it reads that data
[19:25:44] <Nihility> baileyje: the default is another matter
[19:25:50] <Nihility> baileyje: ill merge your patch sec
[19:26:28] *** stelios_RH_GR has joined #jboss-as7
[19:26:33] <Nihility> im also merging stansilvert's
[19:26:55] <Nihility> dmlloyd: you have remy's or should i do that?
[19:27:16] <dmlloyd> I won't get to it for about another 15-20 so go ahead
[19:28:51] *** stelios_RH_Athen has quit IRC
[19:34:48] *** mbg|away is now known as mbg
[19:35:08] <asaldhan> ALR: hey. do I get to see the war archive someplace when using Arq?
[19:35:12] <rmaucher> asaldhan: I think there are issues with JBossWebRealm
[19:35:21] <asaldhan> rmaucher: probably
[19:35:31] <asaldhan> rmaucher: its a pretty basic realm right now.
[19:35:46] <rmaucher> authenticate will call the superclass which will call getPassword, which will return null ?
[19:36:16] <asaldhan> so JBWRealm missing that method?
[19:36:20] <asaldhan> getPassword?
[19:36:34] *** AndyTaylor has joined #jboss-as7
[19:36:34] *** ChanServ sets mode: +v AndyTaylor
[19:36:34] <smarlow> dmlloyd:  I squashed my changes into https://github.com/scottmarlow/jboss-as/commit/02ef117be2b030efcec01cd887264401640ecc1c, could you please look at merging it in when you free up.  Thanks
[19:36:36] <jbossbot> git [jboss-as] 02ef117.. Scott Marlow JBAS-9081 support for injecting PU into Hibernate session factory, needs to be tested (will probably need more changes but we need a test case to see),forked org.jboss.ejb3.test.epcpropagation.unit.EPCPropagationTestCase, ignore unwrap exception,remove unused XPCResolver interface,JBAS-8961 restructure use of persistence provider details,remove logging rather than add its classes to classpath, epcpro
[19:36:37] <rmaucher> it has it since it's abstract, but it does return null
[19:36:37] <jbossbot> jira [JBAS-9081] Inject Persistence Unit into Hibernate session factory [Coding In Progress (Unresolved) Sub-task, Minor, Scott Marlow] https://issues.jboss.org/browse/JBAS-9081
[19:36:38] <jbossbot> jira [JBAS-8961] restructure use of JPA persistence provider specific details [Coding In Progress (Unresolved) Sub-task, Major, Scott Marlow] https://issues.jboss.org/browse/JBAS-8961
[19:37:15] <asaldhan> Nihility: sorry to be nth time people asking about this.  any strategies for TCCL?
[19:37:42] <asaldhan> Nihility: the thing is if I put in two properties file inside a war.  what would be the right way for CL to locate them?
[19:38:07] <Nihility> asaldhan: yes the basic rule is TCCL is the deployment classloader, load user classes from there. Always load your classes from your defining classloader
[19:39:16] <Nihility> asaldhan:you probably don't want to reparse them on every call? maybe it should be analyzed in a deployer?
[19:39:49] <jbossbot> git [jboss-as] push master 96eb6f6.. Stan Silvert No longer need to add JSF ConfigureListener to shared web metadata.  It will now be read from jsf_core.tld.
[19:39:49] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/e62a90e...96eb6f6
[19:40:41] <rmaucher> ah ok, stan said his listener from shared metadata didn't work, I guess it's fine after all then
[19:41:01] <asaldhan> rmaucher: will take a look at the realm today.
[19:41:31] <Nihility> https://issues.jboss.org/browse/JANDEX
[19:41:33] <Nihility> wohhoo
[19:41:47] <rmaucher> just realized that we've shipped a beta 1 of a webserver without access logging or security ;)
[19:42:18] <Nihility> no one cares about security!
[19:42:19] <Nihility> hahahahahah
[19:42:20] <rmaucher> so it needs to be there for beta 2 I think
[19:42:23] <rmaucher> yeah
[19:42:45] <rmaucher> besides, access logs can fill up a server HDD which cause crashes -> baaaad
[19:42:53] <Nihility> totally
[19:42:58] * Nihility watches asaldhan sweat
[19:43:10] *** AndyTaylor has quit IRC
[19:43:26] <jbossbot> git [jboss-as] push master 5efe39a.. John E. Bailey [JBS-9143] - Properly use the web context-root defined in application.xml
[19:43:27] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/96eb6f6...5efe39a
[19:44:03] * asaldhan is sweating anyway due to feverish conditions (under the weather)
[19:46:10] <rmaucher> it's possible my commits will conflict a bit with that commit, we'll see
[19:46:23] <asaldhan> Nihility: when I deploy a war to AS7,  do I see it unexploded someplace?
[19:48:05] *** darranl has quit IRC
[19:49:31] <rmaucher> uh oh, and we also can't configure any non localhost bind address if I read Scott's email correctly
[19:53:23] <asaldhan> ALR: ping
[19:53:26] <rmaucher> ouf, it works
[19:53:44] *** jhalliday has quit IRC
[19:55:30] *** balunasj has joined #jboss-as7
[19:55:30] *** balunasj has joined #jboss-as7
[19:55:30] *** ChanServ sets mode: +v balunasj
[20:04:17] <bstansberry> jamezp: I created that deployment JIRA I mentioned yesterday: JBAS-9147
[20:04:18] <jbossbot> jira [JBAS-9147] Deployment-related OperationHandler improvements [Open (Unresolved) Task, Major, Unassigned] https://issues.jboss.org/browse/JBAS-9147
[20:05:09] <jamezp> bstansberry: Cool, taking a look now.
[20:06:31] <jamezp> bstandsberry: Looks good, should I assign it to myself?
[20:07:50] *** mbg1 has joined #jboss-as7
[20:09:11] *** adietisheim has quit IRC
[20:15:57] *** sannegrinovero has quit IRC
[20:16:32] *** jpederse has quit IRC
[20:21:15] *** balunasj is now known as balunasj_mtg
[20:29:11] <bstansberry> jamezp: sorry, didn't see your reply. Yes, if you want it, it's yours :)
[20:32:40] <smarlow> baileyje:  copying the VirtualFile contents into a regular file, seemed to help (not that we want to do that).
[20:43:12] <asaldhan> Nihility: rmaucher:  below
[20:43:14] <asaldhan> ============
[20:43:24] <asaldhan> Running org.jboss.as.testsuite.integration.websecurity.WebSecurityFORMTestCase
[20:43:24] <asaldhan> Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 1.531 sec
[20:43:25] <asaldhan> ===============
[20:44:14] * asaldhan is going to check hist test case in
[20:45:17] <baileyje> smarlow: Did the hibernate guys say when they might allow extension to allow other options?
[20:46:51] *** stansilvert has quit IRC
[20:47:13] <rmaucher> and the tests are ?
[20:47:17] <asaldhan> rmaucher: https://github.com/anilsaldhana/jboss-as/commit/0efdc673a45d6d172559183243a99e0927e33e65
[20:47:18] <jbossbot> git [jboss-as] 0efdc67.. Anil Saldhana add a web FORM authentication test case
[20:47:48] <asaldhan> rmaucher: can you have a secondary glance at my test case and opine about the httpclient code?
[20:48:13] * asaldhan based that code on the httpclient 4.x sample form login code
[20:48:29] <rmaucher> where is the thing which did not work ?
[20:48:49] <rmaucher> I tested replacing the default servlet, and it does appear to be fine as well
[20:49:12] <smarlow> baileyje:  I didn't ask that but its not just hibernate either, its part of the JPA api that I have to give out a URL.  Other persistence providers could also trip over the same issue
[20:49:45] <baileyje> smarlow: Is there anything in the spec that says we need to pass them file:// URLS
[20:49:52] <asaldhan> rmaucher: it was an issue with my test case. the sample code from httpclient webpage is inadequate.  I used the firefox live http header to get the proper interaction
[20:50:07] <smarlow> baileyje:  nope, not at all, just a URL
[20:50:11] <asaldhan> rmaucher: at this time, I do not see any issues.
[20:50:21] <rmaucher> ok
[20:50:35] <dmlloyd> okay rmaucher, smarlow have patches for me, anyone else?
[20:50:35] <baileyje> but the issue is you can't assume a URL can be turned into a File object
[20:50:45] <asaldhan> dmlloyd: mine too
[20:50:56] <asaldhan> dmlloyd: https://github.com/anilsaldhana/jboss-as/commit/0efdc673a45d6d172559183243a99e0927e33e65
[20:50:57] <jbossbot> git [jboss-as] 0efdc67.. Anil Saldhana add a web FORM authentication test case
[20:51:02] <dmlloyd> what branch, asaldhan?
[20:51:17] <dmlloyd> you can get a commit ID from a branch but you can't get a branch from a commit ID :)
[20:51:22] <smarlow> baileyje:  http://download.oracle.com/javaee/6/api/javax/persistence/spi/PersistenceUnitInfo.html#getPersistenceUnitRootUrl%28%29
[20:51:30] <asaldhan> dmlloyd: I am still learning.
[20:51:43] <dmlloyd> asaldhan, we all are.
[20:51:45] <asaldhan> dmlloyd: https://github.com/anilsaldhana/jboss-as
[20:51:53] <smarlow> baileyje:  "The URL will either be a file: URL referring to a jar file or referring to a directory that contains an exploded jar file, or some other URL from which an InputStream in jar format can be obtained."
[20:51:55] <asaldhan> dmlloyd: that is my master.  what is the branch?
[20:52:03] <dmlloyd> the branch is "master".
[20:52:10] <baileyje> smarlow: Then your hack is the only thing we can do.
[20:52:14] <baileyje> for now
[20:52:15] <dmlloyd> github leaves off the /tree/master if it's your default branch.
[20:52:16] <asaldhan> dmlloyd: thank you. Yes, Master. :)
[20:52:19] <dmlloyd> ha
[20:52:47] <rmaucher> I did rebase my web config patches: https://github.com/rmaucher/jboss-as/commits/master can someone pull them ?
[20:52:49] <asaldhan> rmaucher: remy, you still associated with the httpclient project?
[20:52:58] <rmaucher> no, I wrote the initial code
[20:53:06] <rmaucher> like, 10 years ago
[20:53:07] <smarlow> baileyje:  that is fine, I'll try to create the files inServerEnvironment.getServerTempDir()
[20:53:40] <asaldhan> dmlloyd: tell me if I have bungled up anything.  this is my second pull request in history
[20:53:43] * smarlow I mean in ServerEnvironment.getServerTempdir folder
[20:54:14] <baileyje> smarlow: Are they allowed to load classes based on that URL?
[20:54:15] <asaldhan> rmaucher: when I was using eclipse and repeated war deployment with arquilian, I was seeing
[20:54:23] <dmlloyd> rmaucher: I've got them.
[20:54:38] <asaldhan> ======
[20:54:38] <asaldhan> )
[20:54:39] <asaldhan> 14:20:23,793 ERROR [org.jboss.as.controller] (pool-2-thread-2) operation ("add") failed - address: ([("deployment" => "web-secure.war")]): java.lang.IllegalStateException: Resource at address [("deployment" => "web-secure.war")] already exists
[20:54:40] <asaldhan> ===========
[20:54:43] <asaldhan> rmaucher: ^
[20:54:59] * asaldhan had to clean the vfs and work folders in tmp
[20:55:19] * asaldhan is not sure about war undeployment with arquilian yet
[20:56:00] <asaldhan> http://pastebin.com/cf33iE0P
[20:56:11] <asaldhan> here is the proper exception if you want it
[20:56:15] <dmlloyd> smarlow: reminder - 4 space indent.  not a show-stopper, just be advised
[20:57:08] <smarlow> dmlloyd:  weird, was that in a bunch of files or just one?  (I'm checking my intellij settings...)
[20:57:08] * asaldhan says with the getPassword method returning null,  JbossWebRealm->authenticate was still seeing the right password from FormAuthenticator
[20:57:35] <dmlloyd> smarlow: the test files primarily
[21:00:26] *** jclingan has joined #jboss-as7
[21:01:17] <smarlow> baileyje:  we pass in a classloader and temp classloader, that is supposed to be used for Classloading.  The idea of passing that URL in, is so they can look for annotations in classes.  So, they could load the classes in that sense, through the URL.
[21:02:18] <asaldhan> rmaucher: did you see http://pastebin.com/cf33iE0P    I had to clean up  /tmp/vfs and /tmp/work often.    I use arquilian in test case and I do not yet know how to undeploy in test case.
[21:02:23] <smarlow> baileyje:  the spec doesn't say that they shouldn't load entity classes (for application use) that way, but they also don't say that the provider can.
[21:02:38] * smarlow too many negatives :)
[21:03:13] <smarlow> baileyje:  I don't think the spec lets them load classes, in the sense that you probably meant, through the URL
[21:03:30] <rmaucher> looks like some dual deployment
[21:03:35] <baileyje> My only concern is if they loaded from both places it woudl be bad.
[21:03:48] <asaldhan> will ask the arquilian guys if I am doing something wrong in the test case.
[21:03:53] <rmaucher> but I don't know the controller code ;)
[21:04:15] <asaldhan> if arquilian does not issue a undeploy command, then there is no cleanup.
[21:05:32] *** jclingan1 has joined #jboss-as7
[21:06:33] <jbossbot> git [jboss-as] push master 807bdad.. Rémy Maucherat Add access-log and rewrite parsing, and implement access logging.
[21:06:33] <jbossbot> git [jboss-as] push master 2dadb50.. Rémy Maucherat More configuration, but no SSL yet.
[21:06:33] <jbossbot> git [jboss-as] push master 7635428.. Rémy Maucherat Finish the main web configuration with SSL. Untested ...
[21:06:33] <jbossbot> git [jboss-as] push master 4f8b85b.. Rémy Maucherat New web build and use the new APIs to implement the missing features.
[21:06:34] <jbossbot> git [jboss-as] push master de72bd9.. Scott Marlow JBAS-9081 support for injecting PU into Hibernate session factory, needs to be tested (will probably need more changes but we need a test case to see),forked org.jboss.ejb3.test.epcpropagation.unit.EPCPropagationTestCase, ignore unwrap exception,remove unused XPCResolver interface,JBAS-8961 restructure use of persistence provider details,remove logging rather than add its classes t
[21:06:34] <jbossbot> jira [JBAS-9081] Inject Persistence Unit into Hibernate session factory [Coding In Progress (Unresolved) Sub-task, Minor, Scott Marlow] https://issues.jboss.org/browse/JBAS-9081
[21:06:35] <jbossbot> jira [JBAS-8961] restructure use of JPA persistence provider specific details [Coding In Progress (Unresolved) Sub-task, Major, Scott Marlow] https://issues.jboss.org/browse/JBAS-8961
[21:06:35] <jbossbot> git [jboss-as] push master 7c18132.. Anil Saldhana add a web FORM authentication test case
[21:06:36] <jbossbot> git [jboss-as] push master URL: http://github.com/jbossas/jboss-as/compare/5efe39a...7c18132
[21:08:48] *** jclingan has quit IRC
[21:10:33] <rmaucher> thanks :)
[21:11:27] *** mbg1 has quit IRC
[21:15:54] <asaldhan> http://community.jboss.org/docs/DOC-16664
[21:16:06] <asaldhan> AS7 countrymen, ^
[21:16:46] *** stelios_RH_Athen has joined #jboss-as7
[21:17:44] *** stelios_RH_GR has quit IRC
[21:29:05] *** stelios_RH_GR has joined #jboss-as7
[21:32:04] *** stelios_RH_Athen has quit IRC
[21:32:04] *** jamezp has quit IRC
[21:32:40] *** jfd has quit IRC
[21:34:27] *** jamezp has joined #jboss-as7
[21:57:16] *** mmoyses has quit IRC
[22:20:40] *** smarlow has quit IRC
[22:21:15] *** smarlow has joined #jboss-as7
[22:21:15] *** ChanServ sets mode: +v smarlow
[22:53:31] *** bgeorges has quit IRC
[22:57:39] *** bgeorges has joined #jboss-as7
[23:10:26] *** asaldhan has left #jboss-as7
[23:22:06] *** pil-dinner has quit IRC
[23:35:22] *** jpearlin has joined #jboss-as7
[23:54:43] *** wolfc has quit IRC

   March 25, 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 | >