[00:09:30] *** ALR has quit IRC
[00:10:29] *** ALR has joined #jbosstesting
[00:10:59] *** ALR has quit IRC
[00:23:30] <jbossbot> git [descriptors] push master a25b987.. Lincoln Baxter, III Faces Project Stage now creates correct values in web.xml
[00:23:30] <jbossbot> git [descriptors] push master URL: http://github.com/shrinkwrap/descriptors/compare/9e71662...a25b987
[00:45:30] *** maeste is now known as maeste_zzz
[00:49:54] *** aslak has quit IRC
[00:54:56] *** michaelschuetz has quit IRC
[00:58:05] *** ALR has joined #jbosstesting
[01:00:07] *** lightguard_jp is now known as lightguard_jp_aw
[01:04:07] *** bleathem has quit IRC
[01:06:52] *** wolfc has quit IRC
[01:39:30] *** jdlee has quit IRC
[01:39:38] *** PeteRoyle has joined #jbosstesting
[01:40:38] *** jdlee has joined #jbosstesting
[01:40:38] *** jdlee has joined #jbosstesting
[01:40:38] *** ALR has quit IRC
[01:44:29] *** ALR has joined #jbosstesting
[02:17:06] *** johnament has joined #jbosstesting
[02:29:51] *** ALR has quit IRC
[02:47:56] *** ALR has joined #jbosstesting
[02:49:46] *** bleathem has joined #jbosstesting
[02:50:24] *** ldimaggi has joined #jbosstesting
[03:00:35] *** ALR has quit IRC
[03:13:09] *** stuartdouglas has quit IRC
[03:20:53] *** kenfinnigan has joined #jbosstesting
[03:51:45] *** johnament has quit IRC
[04:00:29] *** rruss has joined #jbosstesting
[04:19:00] *** kenfinnigan has quit IRC
[04:58:11] *** bleathem has quit IRC
[05:14:21] *** lightguard_jp_aw is now known as lightguard_jp
[05:16:47] *** ldimaggi has quit IRC
[05:55:06] *** bleathem has joined #jbosstesting
[06:14:22] *** rruss has quit IRC
[06:14:50] *** rruss has joined #jbosstesting
[06:34:44] *** lightguard_jp has quit IRC
[06:56:23] *** kpiwko has joined #jbosstesting
[07:00:37] *** OndraZizka has quit IRC
[07:00:59] *** OndraZizka has joined #jbosstesting
[07:01:10] *** michaelschuetz has joined #jbosstesting
[07:07:00] *** michaelschuetz has quit IRC
[07:27:18] *** lightguard_jp has joined #jbosstesting
[07:29:57] *** lightguard_jp has joined #jbosstesting
[07:32:34] *** lightguard_jp has joined #jbosstesting
[07:33:23] *** lightguard_jp has joined #jbosstesting
[07:33:33] *** lightguard_jp has quit IRC
[08:03:20] *** mgoldmann has joined #jbosstesting
[08:03:28] *** mgoldmann has quit IRC
[08:03:28] *** mgoldmann has joined #jbosstesting
[08:07:04] *** oskutka has joined #jbosstesting
[08:19:09] <PeteRoyle> Hi. Does Arquillian Alpha 5 / ShrinkWrap Alpha 12 support deploying descriptors or would I need a dev/snapshot version?
[08:19:24] <PeteRoyle> as per: http://docs.jboss.org/arquillian/reference/latest/en-US/html_single/#d0e1613
[08:20:09] <PeteRoyle> or more to the point, as per: http://docs.jboss.org/arquillian/reference/1.0.0.Alpha5/en-US/html/advanced.html#d0e1613
[08:23:02] <PeteRoyle> ok nm I think I got it
[08:23:09] *** jharting has joined #jbosstesting
[08:25:45] <PeteRoyle> Is it possible to just deploy an existing file as the descriptor?
[08:28:59] *** rruss has quit IRC
[08:40:49] *** wyer has joined #jbosstesting
[09:42:24] *** maeste_zzz is now known as maeste
[09:54:47] *** bleathem has quit IRC
[09:58:46] *** Royle has joined #jbosstesting
[09:59:20] *** michaelschuetz has joined #jbosstesting
[09:59:52] *** wolfc has joined #jbosstesting
[09:59:56] *** Royle has left #jbosstesting
[10:02:07] *** PeteRoyle has quit IRC
[10:03:07] *** jeand has joined #jbosstesting
[10:09:01] *** vtunka_wfh has joined #jbosstesting
[10:36:10] *** PeteRoyle has joined #jbosstesting
[10:42:23] *** alesj has joined #jbosstesting
[10:42:40] *** PeteRoyle has quit IRC
[10:55:44] *** Jaikiran has joined #jbosstesting
[12:07:10] *** wolfc has quit IRC
[12:07:21] *** wolfc has joined #jbosstesting
[12:22:00] *** aslak has joined #jbosstesting
[12:24:59] *** ALR has joined #jbosstesting
[12:33:38] *** aaddriancole has joined #jbosstesting
[12:33:43] <aaddriancole> aslak moving here
[12:33:50] <aaddriancole> ALR yo
[12:33:58] <aaddriancole> hey, so what do you think of this idea...
[12:33:59] <ALR> aaddriancole: Yo man.
[12:34:03] *** dblevins has quit IRC
[12:34:12] <ALR> You scared off Blevins.
[12:34:26] <aaddriancole> heh
[12:34:29] <aaddriancole> that's how I roll
[12:34:42] <aaddriancole> ALR I'm gonna try to get up to JBW
[12:34:45] <ALR> aaddriancole: You w/ Manik?
[12:34:51] <aaddriancole> might have to sleep with manik
[12:34:52] <ALR> aaddriancole: That'd be great.
[12:35:08] <aaddriancole> yeah  love to see you guys
[12:35:15] <aaddriancole> just need to work out cheapest via
[12:35:18] <ALR> Suhweet.
[12:35:26] <ALR> aaddriancole: You're still based in SF?
[12:35:29] <aaddriancole> yea
[12:35:33] <aaddriancole> just moved to the mission
[12:35:36] <aaddriancole> or in the process
[12:35:46] <ALR> I like the Mission.
[12:36:03] <aaddriancole> yeah I'm across the street from zeitgeist
[12:36:10] <aaddriancole> the official bar of jclouds from now on
[12:36:12] <aaddriancole> :D
[12:36:32] <aaddriancole> so. on arq ... add a @Maven annotation to the arq createArchive methds
[12:36:49] <aaddriancole> with coordinates as params
[12:36:51] <ALR> Instead of @Deployment?
[12:36:55] <ALR> Yeah, that's hot.
[12:37:11] <ALR> ARQ could internally use the MavenResolver API to work out the details.
[12:37:11] <aaddriancole> yep... this will force either through instlal or deploy
[12:37:15] <aaddriancole> yep
[12:37:26] <aaddriancole> so in xml or whatever, you could change to deploy
[12:37:32] <aaddriancole> or whatev
[12:37:42] <aaddriancole> but the point is that you'd have a releaseable artifact
[12:37:52] <ALR> IMO we still have some work to do regarding defining N deployments and *when* they're deployed.
[12:38:00] <ALR> But I know aslak has some ideas there
[12:38:13] <aaddriancole> yeah. this is a good point
[12:38:15] <ALR> Though the @Maven thing may not help with the "current" artifact.
[12:38:23] <ALR> We still don't have a great solution for the "current"
[12:38:31] <ALR> Because "current" may not be built/installed!
[12:38:39] <aaddriancole> well, it could help in a way..
[12:38:42] <aaddriancole> here's one idea
[12:38:45] <aaddriancole> after tests pass
[12:38:49] <aaddriancole> it is published
[12:38:54] <aslak> ALR, well, current would be bind to integration-phase and import your builds output
[12:39:04] <ALR> aslak: integration-test is *before* install.
[12:39:16] <ALR> Yet after package.
[12:39:22] <aaddriancole> oh yeah.. well this is a good point
[12:39:30] <aslak> ALR, aa, for using the Resolver sure, but you can still import the packaged artifact
[12:39:34] <aaddriancole> I mean most projects are running maven so there's overlap on this thing
[12:40:09] <ALR> Alternatively, bind the test lifecycle to "validate" or "verify", whichever of those is the really late-running one after install.
[12:40:23] <ALR> But no one's gonna wanna put that in the Surefire config in the POM.
[12:40:27] <aaddriancole> heh
[12:40:43] <aaddriancole> maybe you can have a sneaky arq agent :)
[12:40:54] <ALR> aaddriancole: An aside, you been following AS7 at all?
[12:40:55] <aslak> true, but i'm not a fan of making arq your build tool either
[12:41:06] <ALR> aslak: +1.  Arq ain't a build tool.
[12:41:33] <aaddriancole> yeah. but I like the idea of getting tested components releasable
[12:41:39] <aaddriancole> even if it isn't a good idea :p
[12:41:54] <ALR> aaddriancole: IMO the testsuite would be in a separate module
[12:42:02] <ALR> Which would then depend upon the components you're looking to test
[12:42:25] <aaddriancole> ALR p.s. haven't been following too closely AS7, except mopping up aslak's puddle of drool related to how fast it is
[12:42:26] <ALR> So like the testsuite module would depend on the CDI or EJB module
[12:42:32] <ALR> And just deploy the @Maven thing.
[12:42:49] <aaddriancole> yeah that's a good way out
[12:42:55] <ALR> aaddriancole: It gives us what's shaping up to be the fastest EE runtime.  We're very excited.
[12:43:22] <ALR> So yeah, I'd say for best practices: Make an EJB module, a WAR module, whatever.  And the testsuite depends upon and deploys those.
[12:43:36] <aaddriancole> sounds good to me
[12:43:44] <ALR> In that sense the ARQ test would mimick the *real* thing you're gonna deploy
[12:43:50] <ALR> Not a manual ShrinkWrap deployment
[12:44:02] <ALR> (Which I still love for its explicit nature and simplicity)
[12:44:04] <aaddriancole> yeah well this could be the next layer as well
[12:44:14] <aaddriancole> I mean you could extend the integration test classes
[12:44:17] <aaddriancole> and override somehow
[12:44:28] <aaddriancole> instead of the shrink ones, use maven
[12:44:39] <ALR> Why bother extending them
[12:44:40] <ALR> ?
[12:44:59] <aaddriancole> well, I suppose the point would be being able to rerun the regression tests
[12:45:00] <ALR> When you can just move the int-tests into another module?
[12:45:11] <aaddriancole> oh all of them?
[12:45:19] <aaddriancole> well... maybe you're right
[12:45:19] <ALR> I suppose so.
[12:45:22] <ALR> Yeah.
[12:45:27] <aaddriancole> just that shrink is still more nimble
[12:45:28] <ALR> Unit tests in the component modules.
[12:45:36] <ALR> ARQ integration tests in a separate module.
[12:45:43] <ALR> And that becomes our recommendation for best practices.
[12:45:53] <ALR> For testing real build artifacts
[12:45:53] <aaddriancole> you don't have to build dependency graph to test minor bits if you keep some shrink in
[12:45:58] <ALR> Otherwise assemble on your own
[12:46:01] <ALR> "Microdeployment"
[12:46:04] <aaddriancole> not sure yet if the combination is too complicated
[12:46:07] <ALR> And you can test those wherever you want.
[12:46:22] <ALR> Right, minor bits == microdeployments.
[12:46:24] <aaddriancole> unit being microdeployment?
[12:46:26] <aaddriancole> yeah
[12:46:29] <ALR> No big bang testing.
[12:46:39] <ALR> Well, microdeployment is still integration.
[12:46:42] <ALR> For instance:
[12:46:48] <ALR> Your EAR is the big kahuna.
[12:46:59] <ALR> The EAR may have web, EJB, CDI, etc....lots of stuff.
[12:47:10] <ALR> But microdeployments, maybe you just test the EJB on its own.
[12:47:27] <ALR> I have some slides on this.
[12:47:31] <ALR> You're a musician, right
[12:47:32] <ALR> ?
[12:47:35] <aaddriancole> :)
[12:47:37] <aslak> the question on how to run your arq tests on your final deployment has come up, as in you have a bunch of microdeployments that run fine, but you also want to run the same ones on the final deployment. i hvae some thought on it, but not 100%
[12:47:54] <ALR> Microdeployments: Test that your phaser pedal, your tuner is working.
[12:48:06] <ALR> Full integration: Test that your whole pedalboard is wired up to the amp.
[12:48:36] <ALR> aslak: I think this is the missing gap for the "final" deployment case.
[12:48:42] <aaddriancole> yeap
[12:48:44] <ALR> 1) Make a separate module
[12:48:52] <ALR> 2) Define deployment by resolving through Maven.
[12:48:57] <ALR> Giving you the "real" stuff.
[12:49:03] <ALR> As generated by the build system.
[12:49:52] <aslak> sure, but using the same tests
[12:50:12] <aaddriancole> well this is the crux of the talk.. should this run the same tests again?
[12:50:13] <ALR> Where does @Maven go?
[12:50:23] <aslak> it can be done by having a separate module, extending a base test that defines the test assertons and in each module define the deployment etc, but it's not very pretty
[12:50:23] <ALR> Not on a method, I wouldn't think.
[12:50:28] <ALR> On the test class?
[12:50:45] <aaddriancole> well, this is a tricky annotation isn't it
[12:50:47] <ALR> @Deployments({@Maven(gav1), @Maven(gav2)})
[12:50:59] <aaddriancole> oh
[12:51:26] <aslak> would be on the Class i think. seems silly to make a method just for the annotation
[12:51:34] <ALR> In this sense, no @Deployment static method in the test class, it'd just be an annotation of deployments upon the test class
[12:51:38] <ALR> Right.
[12:51:48] <ALR> @Deployments(array)
[12:51:55] <aaddriancole> yeah I mean ARQ could make some assumptions right?
[12:51:58] <ALR> @Maven is a fully-declarative deployment
[12:52:05] <aaddriancole> if @Deployments on class, ignore static
[12:52:11] <ALR> And actually...
[12:52:21] <ALR> @Maven should have @Deployment as a meta-annotation in it.
[12:52:36] <ALR> So that users could put in ANY user-defined special annotation
[12:52:47] <ALR> So long as it was itself @Deployment
[12:52:49] <aslak> it's no biggy to impl. the extension point is there. you can override the whole @Deployment method handling using the DeploymentScenarioGenerator.
[12:52:50] <aaddriancole> makes sense.. I mean we may need to have @OSGi at some point
[12:53:13] <ALR> @Deployment
[12:53:13] <ALR> @interface OSGi{}
[12:53:17] <ALR> Same deal.
[12:53:24] <aaddriancole> aslak I'm not surprised you have a means of handling this cleanly :)
[12:53:40] <ALR> Oh yeah, Aslak's got all sorts of prototype stuff in the wings all over.
[12:53:55] <aslak> why OSGI?
[12:54:04] <aaddriancole> right.  so I think it is useful for example of the design to have both maven and osgi
[12:54:09] <ALR> And it was kpiwko who gave us Maven Resolution to ShrinkWrap
[12:54:28] <aaddriancole> ex. bundle repository vs maven
[12:54:39] <aaddriancole> not terribly important.. just an example
[12:54:50] <aslak> i thought the bundle repo used maven ?
[12:55:10] * ALR is so sick of modular ClassLoading.  Bring back the Unified See-All, Know-All, Conflict-All CLs!
[12:55:12] <aslak> maybe that was just the Felix OG part..
[12:55:28] <kpiwko> ALR: quite lengthy discussion, it will take me some time to go through
[12:55:37] <ALR> kpiwko: I was just shouting you out.
[12:55:47] <ALR> For your contributions to MavenResolver
[12:55:54] <aaddriancole> aslak all I know is that OSGi is more complicated :)
[12:56:20] <aaddriancole> I know that there are some guys who have to go a couple places since there aren't osgified versions of various things in maven
[12:56:22] <ALR> kpiwko: We're talking about using something like @Maven(gav="org.jboss:jboss-something:1.0.0") to define deployments
[12:57:27] <aaddriancole> conceding I'm really not qualified to spell OSGi
[12:57:41] <aslak> @ArchiveRepository(cord = "ba:ba", resolv = Maven-.class|OSGi.class)
[12:58:03] <ALR> Until users demand OSGi support, I'm punting on it.
[12:58:05] <ALR> Here's why:
[12:58:10] <ALR> OSGi is for server implementors.
[12:58:23] <ALR> It's supposed to be totally abstracted away from user deployments.
[12:58:37] *** dblevins has joined #jbosstesting
[12:58:39] <ALR> It's in place just to support concurrent artifacts of different versions without conflict
[12:58:46] <ALR> And most users aren't digging into the internals.
[12:58:52] <aslak> ALR, we do have users requesting it tho
[12:58:56] <ALR> CRAP
[12:59:00] <ALR> Why oh why?
[12:59:11] <aslak> ALR, well, define user
[12:59:12] <ALR> Thomas Diesler is not a user, BTW. :P
[12:59:30] <aslak> ALR, a Arq user is not nessesarly a EE user
[12:59:42] <ALR> (He's our JBoss OSGi lead, for anyone reading along)
[12:59:53] <aslak> ALR, others as well
[13:00:07] <ALR> aslak: Right.  What users do we have who are interested in deploying and testing specifically to OSGi containers?
[13:00:23] <aslak> Thomas
[13:00:23] <ALR> ie. What users are digging into things like BundleContext?
[13:00:25] <aslak> :P
[13:00:29] <ALR> Hahaha
[13:00:35] <wyer> lol
[13:00:40] <aslak> ALR, we have one or two on the forums asking about it
[13:00:52] <aslak> ALR, don't remember names
[13:01:01] <ALR> My opinion is bigger fish to fry ATM
[13:01:08] <ALR> Though I don't discount its use.
[13:01:14] <ALR> Just as a matter of priorities and resources.
[13:01:20] <ALR> @Maven seems awesome and useful
[13:01:25] <ALR> And easy to implement, to boot
[13:01:32] <aslak> ALR, sure, we're closing the loop on EE now. but arq is 'anything', so OSGi is definetly on the table
[13:01:38] <ALR> Yep, agree.
[13:02:02] <ALR> When we get a mass of users looking to use ARQ to test their OSGi containers, then I think that makes sense.
[13:02:03] <aaddriancole> I brought up OSGi mainly as the cloudsoft guys who are looking at arq now are a pure osgi play
[13:02:07] <ALR> Or they can contribute it.
[13:02:20] <aaddriancole> granted they are more server-authors
[13:02:21] <aslak> with the next split out of impl-base, arq is even just a integration layer between the testframeworks.. no container/deployments involved
[13:02:48] <aaddriancole> maven's a really good first start anyway
[13:02:55] <ALR> I think so.
[13:02:56] <aaddriancole> starts to solidify this side of things
[13:03:03] <aslak> so it can be used for e.g. Ajaocado or Byteman to have standalone 'runners', but not having to worry about writing junit/testng/rspec/spock/easyb integrations
[13:03:04] <ALR> aaddriancole: Fork the repo and get us a patch. :P
[13:03:24] <ALR> Oh, BTW, who do I smack for "Ajocado"?
[13:03:32] <ALR> English, people.  I speak English!
[13:03:49] <aslak> ALR, it's a combination of Ajax and Avocado
[13:03:49] <aaddriancole> ALR maybe so :) I need it for jclouds anyway
[13:03:59] <aaddriancole> well, want it
[13:04:08] <aaddriancole> but want it like in a jonesing way
[13:04:11] <ALR> Asynchronous fatty vegetables?
[13:04:28] <ALR> Jonesing?  Meaning you want it ... now?
[13:04:30] <aslak> something like that.. :)
[13:04:43] <aslak> ALR, Lukas btw
[13:04:49] <ALR> aslak: I know :)
[13:05:04] <aslak> gotta run do some shopping.. bbl
[13:05:07] <ALR> aslak: He can get in line behind the creators of "Teeid".
[13:05:11] <ALR> Cool, lata
[13:06:03] <aslak> ALR, kpiwko, you guys up for a meeting in 3 hours ro so?
[13:06:07] <aaddriancole> ALR + aslak I need to catch up on current arq design anyway
[13:06:23] <aaddriancole> I'll try to get out to JBW and immerse myself
[13:06:27] <ALR> aslak: Yeah, I'll try to stay awake.
[13:06:30] <kpiwko> aslak: yes
[13:06:50] <aslak> cool
[13:07:06] <kpiwko> ALR: apart from Ajax+Avocado Ajocado name was motivated by Mockito
[13:09:14] <ALR> kpiwko: All I'm saying is...spelling and I generally go "WTF".  No idea how to pronounce it.  It's not a word. :)
[13:09:49] *** ldimaggi has joined #jbosstesting
[13:10:45] <kpiwko> ALR: well, we have 2 pronunciations here ... hope one of them sounds the same as native would pronounce :)
[13:17:23] <ALR> kpiwko: Someday I'll get to Brno to hear it from ya.
[13:18:23] <nickarls> The framework could have been called Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch after the town in Wales
[13:18:25] *** dblevins has quit IRC
[13:18:42] <wyer> nickarls: so true
[13:18:50] <wyer> now i wanna read duncton wood :/
[13:18:57] <ALR> nickarls: That's the capital of Slovenia.
[13:19:02] *** dblevins has joined #jbosstesting
[13:19:19] <wyer> beer \0/
[13:20:29] <kpiwko> still Ajacodo is much better then RichFaces-Selenium
[13:23:06] <ALR> aaddriancole: You have a JIRA lined up yet for this @Maven enhancement?
[13:25:08] <aaddriancole> sure.. will do
[13:26:25] <kpiwko> aaddriancole: you can assign it to me, I'll try to prototype it during weekend
[13:28:55] <ALR> wyer: Wanna do some SW work?
[13:29:18] <wyer> SW been ?
[13:29:29] <ALR> ShrinkWrap
[13:29:50] <wyer> sure can take a look
[13:30:35] <wyer> it is awesome sauce after all
[13:30:48] <ALR> wyer: :
[13:30:49] <ALR> SHRINKWRAP-247
[13:30:51] <jbossbot> jira [SHRINKWRAP-247] Support recursive addition of web resources [Open (Unresolved) Feature Request, Major, Karel Piwko] https://issues.jboss.org/browse/SHRINKWRAP-247
[13:31:05] <ALR> wyer: Wanna see if this one is in fact already done by nature of the linked issues?
[13:31:47] <kpiwko> ALR: wyer: seems similatr to work Davide had done
[13:32:17] <ALR> kpiwko: Yes.
[13:32:29] <ALR> Though I'm not 100% sure if the use case here is fully covered
[13:32:37] <ALR> I already pushed Davide's stuff upstream
[13:33:59] <aaddriancole> https://issues.jboss.org/browse/ARQ-417
[13:34:00] <jbossbot> jira [ARQ-417] Support maven deployments [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/ARQ-417
[13:34:20] <aaddriancole> kpiwko cool thanks!
[13:35:14] <ALR> Added our chat to the issue
[13:35:28] <wyer> ALR: will check it out after work
[13:35:53] <wyer> gmt +2 so 3 hours left
[13:35:57] <ALR> wyer: Sweet.  That'd be great.  Don't be afraid to dig into whatever you see as well.
[13:36:07] <ALR> wyer: Or if some other issue catches your eye.
[13:36:28] <ALR> wyer: Contribution guide: http://community.jboss.org/wiki/ShrinkWrapDevelopmentandContribution
[13:38:28] <wyer> cool
[13:39:21] <ALR> And that, my friends, is how you cook up free labor.
[13:39:23] *** dblevins has quit IRC
[13:39:29] <wyer> true story
[13:39:35] <kpiwko> ALR: any plans to add checkstyle plugin to sw?
[13:40:00] <wyer> ALR: gotta give something back esp to jboss
[13:40:01] *** dblevins has joined #jbosstesting
[13:40:05] <wyer> i use so much of it
[13:40:09] <wyer> except the app server
[13:40:10] <ALR> kpiwko: If we can package the proper formatters and cleanup stuff for Eclipse in the project (like AS7 does), then I'm not opposed.
[13:40:14] <wyer> maybe AS 7 will excite me
[13:40:17] <wyer> hope so
[13:40:29] <ALR> wyer: AS7 is amazeballs.  For real.
[13:40:56] <kpiwko> ALR: I'll ask Lukas, afaik he had prepared it for Ajocado already
[13:41:05] <ALR> kpiwko: Coolio.
[13:41:18] <ALR> Also could steal all the configs from AS7, though I don't particularly like that style.
[13:41:41] <ALR> And the notion of a whole-tree format to get the thing started is not so attractive.
[13:41:57] <ALR> As that'll create the conflicts we're looking to prevent
[13:41:59] <kpiwko> well AS uses different formatting but sooner or later we have probably to align
[13:42:48] <kpiwko> i'll be missing 3 space tabs :,-( :)
[13:46:21] <wyer> 3 space tabs ftw
[13:50:45] *** rruss has joined #jbosstesting
[13:50:52] *** rruss has quit IRC
[13:54:18] *** timte has joined #jbosstesting
[13:59:28] *** dblevins has quit IRC
[14:00:01] *** dblevins has joined #jbosstesting
[14:08:43] <timte> aslak: In Arquillian alpha5 I get "java.lang.IllegalStateException: No valid security context for the caller identity" from a context.getCallerPrincipal() call
[14:08:58] <timte> aslak: the same code works in alpha3 and alpha5
[14:09:15] <timte> I meant alpha3 and alpha4
[14:09:25] <ALR> timte: Possible you're not running the test in the container?
[14:09:38] <ALR> Some of how that works changed in the ARQ API in Alpha5.
[14:09:55] <ALR> The SecurityContext is associated w/ the Thread
[14:10:05] <ALR> timte: test case in pastebin?
[14:10:25] <timte> I have this on the test class
[14:10:26] <timte> @RunWith(Arquillian.class)
[14:10:26] <timte> @RunAsClient
[14:10:56] <ALR> timte: So @RunAsClient means you don't want the test to run in-container on the server
[14:11:17] <timte> right
[14:11:31] <ALR> And @Deployment(testable=false) too ?
[14:11:52] <timte> uhm, no
[14:12:13] <ALR> That bit will ensure the test case and runner isn't bundled up with the deployment.
[14:12:29] <ALR> But more importantly, what is the test doing, what is your stack trace?
[14:12:48] <ALR> If you have a client-side test looking for a security context which exists on the server, how could you expect that to work? :)
[14:13:34] *** dblevins has quit IRC
[14:14:28] <timte> ALR: the tests just sends http requests and the security context code is in the deployed war file
[14:15:19] <ALR> timte: Again, need to see some code and traces
[14:15:38] <timte> ALR: I'll try the testable=false trick first
[14:15:48] <ALR> timte: I don't think that one is your problem.
[14:21:25] <wyer> timte: you would have to be sending some http auth along with your request at least
[14:23:18] <ALR> Right, but that would have been true in Alpha3 and Alpha4 too
[14:24:18] <wyer> yeah i misread a bit there
[14:24:51] <wyer> more stack trace
[14:24:53] <wyer> :/
[14:27:28] *** dblevins has joined #jbosstesting
[14:40:46] <timte> http://pastebin.com/Dz5fMRXN
[14:40:57] <timte> ALR: testable=false didn't help
[14:41:04] <ALR> I know :)
[14:41:24] <ALR> timte: And the code?
[14:42:14] <ALR> This doesn't look like an ARQ issue.
[14:42:58] <timte> ALR: how can you tell?
[14:43:48] *** vtunka_wfh has quit IRC
[14:44:23] <ALR> timte: Because you're missing a SC from inside the server.
[14:44:30] <ALR> ARQ just does the deployment.
[14:44:34] <ALR> And starts the server.
[14:44:54] <ALR> Somewhere along the line you're not authenticating or trying to get at something that's not there.
[14:46:42] <timte> ALR: The app works fine and I am authenticated when using the app regulary. And I get authenticated with alpha3 and alpha4
[14:47:15] <ALR> timte: And your authentication code is...
[14:47:16] <ALR> ?
[14:47:47] <timte> ALR: JBoss does the authentication for me, I'm just passing along Authorization in the http request header
[14:49:08] <ALR> Great.  Arquillian has nothing to do with that.
[14:49:43] <timte> if Arquillian has nothing to do with it there wouldn't be a difference when using alpha4 or alpha5  :/
[14:49:46] <ALR> ARQ will manage server lifecycle and deployment.  That's about it.  So I'm trying to get to the root cause of the missing link here.
[14:50:41] <ALR> timte: I don't necessarily believe you when you say nothing else has changed.  :)  Barring some evidence to the contrary.  With the limited info I've seen (ie. no code), looks like a plain missing SC from inside your webapp.
[14:50:57] <timte> SC?
[14:51:34] <ALR> SecurityContext
[14:51:51] <ALR> ie. It's not on the executing Thread
[14:52:01] <ALR> So either something hasn't authorized, or hasn't propagated.
[14:52:24] <ALR> Unless you're relying upon client SC propagation into the HTTP request into the server
[14:52:42] <ALR> Where the client has authenticated and you're looking for that to make it inside the server via HTTP
[14:53:58] <ALR> Case in point: In your stacktrace grep on "arquillian". :P
[14:54:33] <ALR> ARQ is just the testrunner here.  Once you're in the server the SC's gotta be propagated in somehow.
[14:54:39] <ALR> And it's not ARQ's job to do that...
[14:54:40] <timte> I change Alpha3 to Alpha5 in pom.xml and I changed addLibrary to addAsLibrary, addWebResource to addAsWebResource and addManifestResource to addAsManifestResource
[14:54:59] <ALR> timte: Hmm, what could be happening:
[14:55:14] <ALR> We changed the mappings for addAsWebResource to go to a different target
[14:55:27] <ALR> If you have a file you're expecting to be in one place....it may be somewhere else.
[14:55:35] <ALR> Perhaps a web.xml or similar
[14:56:01] <ALR> archive.toString(true) will show you what's in the archive you're sending off
[14:56:14] <ALR> And if a file is in the wrong place, that'd mess up your auth for sure.
[14:56:27] <timte> .addAsWebResource(new File("/Users/timterlegard/Project/xlibris-arkiv-ng/xlibris-all/src/main/webapp/WEB-INF/web.xml"))
[14:56:46] <wyer> its addAsWebInfResource() now is it not ?
[14:56:49] <ALR> timte: archive.toString(true) .. let's see that
[14:56:59] <ALR> Mhmm.
[14:57:02] <wyer> from those tickets u had me reading ALR ?
[14:57:07] <ALR> web resource is now the web root.
[14:57:11] <wyer> thats what i have gleened so far
[14:57:27] <ALR> timte: I'll bet you have web.xml in the root of your WAR
[14:57:31] <wyer> that would definitely break you security since your security realm is no longer there
[14:57:37] <wyer> if web.xml is in the root
[14:57:45] <ALR> wyer: Exactly what I'm sayin'
[14:57:51] <timte> right
[14:58:07] <timte> it's in the root
[14:58:11] <ALR> Bingo.
[14:58:29] <ALR> In the future....paste code. :P
[14:58:34] <ALR> We'll skip the guesswork.
[14:58:58] <ALR> So in this case don't be mad at Arquillian.  You can be mad at ShrinkWrap for changing APIs and contracts in the Alpha series.
[14:59:07] <wyer> those bastards
[14:59:08] <ALR> And ultimately be mad at me for that decision.
[14:59:13] <wyer> :P
[14:59:14] <timte> :)
[14:59:23] <wyer> can't stick to an API duting alpha :P
[14:59:28] <wyer> *during
[14:59:31] <ALR> We're locking soon
[14:59:46] <wyer> i managed so its okay :)
[14:59:47] <ALR> I'm a big fan of changing Alpha APIs as much as is warranted
[14:59:55] <timte> sure
[14:59:58] <wyer> now changing to glassfish 3.1
[15:00:05] <ALR> To get it right for API lockdown in Beta
[15:00:07] <wyer> that killed all my testy goodness
[15:00:16] <wyer> still have to sort it out at some point
[15:00:40] <wyer> ALR: definitely changing API's before CR is all good imo
[15:01:31] <ALR> wyer: Yeah.  People who say you only have one chance to get an API right I think are lying.
[15:01:43] <ALR> You just need to be clear about what users can expect for compatibility
[15:01:49] <ALR> No one gets an API right the first time.
[15:02:28] <wyer> yeah, i think having a userbase in alpha who are willing to go along with the changes lets you get it right
[15:02:32] <timte> so how do I get the web.xml in the right directory?
[15:02:41] <ALR> addAsWebInfResource
[15:02:51] <wyer> designing a system before it has users is like designing a battle plan before hand
[15:05:53] <ALR> I'd rather have a battle plan beforehand.
[15:06:10] <ALR> Imagine if they just sent troops to Normandy and were like, "Allright guys....wing it."
[15:06:36] <timte> BUILD SUCCESS
[15:06:50] <ALR> Damn I'm good.
[15:06:50] <timte> yay, thanks
[15:06:56] <ALR> No problem.
[15:07:11] <ALR> Now go talk about how awesome ARQ and SW are to everyone you know.
[15:07:23] <timte> sure  :)
[15:08:50] *** wolfc has quit IRC
[15:11:27] <wyer> ALR: for the guys on the beach i think it pretty much was "wing it" :P
[15:18:03] *** rruss has joined #jbosstesting
[15:18:33] <aslak> timte, for adding the web.xml you can ofcourse use the setWebXML as well..
[15:20:50] <ALR> Ah yeah, I was assuming like a jboss-web.xml thing due to this being an auth issue
[15:20:53] <aslak> ALR, you have some guide lines here: http://community.jboss.org/message/600298#600298
[15:21:02] <aslak> aa right..
[15:21:40] <ALR> aslak: You mean, do I have some guidelines?
[15:21:49] <ALR> aslak: Descriptors needs a Design overhaul.
[15:21:59] <ALR> aslak: And then it's a simple matter of porting in the DSLs
[15:22:05] <ALR> But first really, it needs proper design.
[15:22:11] <aslak> ALR, i mean, would you be please reply with the descriptors guide lines to our happy new contributor.. :)
[15:22:34] <ALR> aslak: Saying what exactly?  The project is a mess and a prototype?
[15:22:35] <ALR> :P
[15:22:38] <ALR> Of course I'll reply.
[15:22:54] <aslak> ALR, i think he's mostly interested in, git, fork, jira stuff etc.. oh it all works, so he can contribute
[15:23:02] <ALR> Oh, like:
[15:23:11] <ALR> http://community.jboss.org/wiki/ShrinkWrapDevelopmentandContribution
[15:23:30] <aslak> there you go.. :)
[15:23:57] <wyer> i've had both arquillian and shrink wrap checked out for a while ;)
[15:24:14] <wyer> the guidelines are good tho
[15:24:17] <wyer> needed those
[15:24:37] <aslak> need to get one some new guidelines for arq.. it's all changed/chaging
[15:25:00] <wyer> well just working on SW for now
[15:25:53] <aslak> .. lunch..
[15:31:22] *** maeste has quit IRC
[15:33:11] *** wolfc has joined #jbosstesting
[15:35:06] *** PeteRoyle has joined #jbosstesting
[15:39:18] *** kpiwko has quit IRC
[16:00:54] *** bleathem has joined #jbosstesting
[16:06:24] *** timte has quit IRC
[16:24:53] <aslak> ALR, seems like we lost Karel
[16:25:07] <ALR> Maybe
[16:43:45] *** oskutka has quit IRC
[16:52:25] *** jharting has quit IRC
[17:03:04] <aslak> ALR, putting in a request to eng-ops to switch Arq over to the Git pull request workflow in jira, want me to add sw and sw-desc while i'm atit?
[17:03:17] <ALR> aslak: Yes, please
[17:03:20] <ALR> I've been meaning to do that.
[17:06:43] <aslak> ALR, yea me too.. my Jira Admin (Shane) has been busy with Seam3 lately.. figured i'd bug eng-ops instead.. :)
[17:06:55] <ALR> aslak: Replied.
[17:06:56] <ALR> Thanks.
[17:21:11] *** kpiwko has joined #jbosstesting
[17:21:49] <kpiwko> aslak: ALR: I'm here
[17:22:17] * ALR too
[17:23:11] <aslak> heya
[17:23:36] <aslak> ALR, got a number?
[17:24:00] <ALR> Lemme see.
[17:24:34] *** michaelschuetz has quit IRC
[17:25:45] <ALR> aslak, kpiwko: Hmm, the AS leader line isn't working.  One sec.
[17:28:49] <ALR> aslak, kpiwko: 2222272630
[17:40:35] *** ldimaggi has quit IRC
[17:57:23] *** PeteRoyle has quit IRC
[17:57:56] *** wyer has quit IRC
[18:09:53] *** alesj has quit IRC
[18:21:53] *** kpiwko has quit IRC
[18:39:10] *** ALR has quit IRC
[19:15:03] *** ldimaggi has joined #jbosstesting
[19:22:19] *** wolfc has quit IRC
[19:28:26] *** wolfc has joined #jbosstesting
[19:44:19] *** Jaikiran is now known as Jaikiran|bbl
[19:53:30] *** mgoldmann has quit IRC
[19:53:48] *** mgoldmann has joined #jbosstesting
[19:54:22] *** michaelschuetz has joined #jbosstesting
[19:56:05] *** wyer has joined #jbosstesting
[20:03:47] *** michaelschuetz has quit IRC
[20:12:30] *** michaelschuetz has joined #jbosstesting
[20:26:29] *** michaelschuetz has quit IRC
[20:44:57] *** OndraZizka has quit IRC
[20:48:19] *** rruss has quit IRC
[20:48:29] *** rruss has joined #jbosstesting
[21:16:26] *** timte has joined #jbosstesting
[21:17:27] *** OndraZizka has joined #jbosstesting
[21:27:36] *** mgoldmann has quit IRC
[22:06:09] *** Jaikiran|bbl is now known as Jaikiran
[22:08:56] *** timte has quit IRC
[22:11:00] *** Jaikiran has quit IRC
[22:16:26] *** maeste has joined #jbosstesting
[22:27:59] *** michaelschuetz has joined #jbosstesting
[22:34:59] *** timte has joined #jbosstesting
[22:42:50] *** jeand has quit IRC
[22:56:27] *** timte has quit IRC
[23:04:31] *** wolfc has quit IRC
[23:11:11] *** lightguard_jp has joined #jbosstesting
[23:36:54] *** jdlee has quit IRC
[23:47:41] *** jdlee has joined #jbosstesting
[23:47:41] *** jdlee has joined #jbosstesting





top