[00:02:35] *** bfitzpat has quit IRC
[00:11:54] *** babo has quit IRC
[00:22:09] *** lancebal_ has quit IRC
[00:26:26] *** aslak has quit IRC
[01:13:29] *** kcbabo has joined #switchyard
[01:59:47] *** ldimaggi has joined #switchyard
[02:21:21] <tfennelly> kcbabo: got the schemas to work in eclipse
[02:21:29] <kcbabo> tfennelly:  good man!
[02:21:36] <tfennelly> kcbabo: had to create a full blown plug in though
[02:21:41] <kcbabo> tfennelly:  yech
[02:21:42] <tfennelly> plugin
[02:21:49] <tfennelly> kcbabo: yeah
[02:22:20] <tfennelly> kcbabo: if that's the case we may as well just do it in the eclipse tooling
[02:23:01] <kcbabo> tfennelly:  well, this could just serve as an ingredient to that tooling distribution, right?
[02:23:16] <tfennelly> kcbabo: yeah.... we can reuse it for sure
[02:25:37] <kcbabo> tfennelly:  hey - refresh my memory
[02:25:42] <kcbabo> tfennelly:  having a bit of a brain fart here
[02:26:03] <tfennelly> kcbabo: well... my brain is fried by now
[02:26:08] <kcbabo> tfennelly:  I thought we decided that reference testing could be handled by using MockHandler
[02:26:32] <kcbabo> tfennelly:  instead of using the mock endpoint in camel
[02:26:59] <kcbabo> tfennelly:  tying to remember if that was right and also how the test code would register that service
[02:27:15] <kcbabo> tfennelly:  obviously not urgent, just came to mind as I'm working on generating service test code from forge
[02:27:22] <kcbabo> tfennelly:  which actually is kinda useful I decided
[02:27:36] <tfennelly> kcbabo: jays... I'm finding it hard to muster a brain fart even
[02:28:51] <kcbabo> tfennelly:  no worries ? we can discuss tomorrow
[02:28:59] <tfennelly> kcbabo: so refresh me... how are references registered?
[02:28:59] <kcbabo> tfennelly:  really not important at the moment
[02:29:14] <kcbabo> tfennelly:  in this case, it's not the reference that needs to be registered
[02:29:34] <kcbabo> tfennelly:  it's the service that you want to resolve a reference
[02:29:52] <kcbabo> tfennelly:  so the activator will lookup the service when processing references
[02:30:04] <tfennelly> kcbabo: ah right yeah
[02:30:07] <kcbabo> tfennelly:  if you are unit testing a CDI bean for example
[02:30:18] <kcbabo> tfennelly:  you want to mock out the service that will satisfy a reference
[02:30:32] <tfennelly> kcbabo: so you can use the test kit to register a MockHandler (or something else of you want) as the Service
[02:30:52] <kcbabo> tfennelly:  OK, didn't know if that method was available
[02:31:01] <kcbabo> tfennelly:  and I'm obviously too lazy to look
[02:31:02] <kcbabo> :-)
[02:31:31] <kcbabo> tfennelly:  shit ? registerInOnlyService()
[02:31:36] <kcbabo> tfennelly:  it's right there in front of my face
[02:31:37] <tfennelly> kcbabo: no it's there alright
[02:32:02] <kcbabo> tfennelly:  ahhh ? it's more satisfying to bother you late at night first ;-)
[02:32:10] <tfennelly> kcbabo: you can also get the ServiceDomain so you can do whatever you want on that
[02:32:15] <kcbabo> tfennelly:  speaking of which ? geez it's late for you
[02:32:56] <tfennelly> kcbabo: so we do the schema catalog as part of the main tooling then I think?
[02:33:09] <kcbabo> tfennelly:  yeah, probs
[02:33:27] <tfennelly> kcbabo: might have one last little play in the morning and see can I get it
[02:33:37] <tfennelly> kcbabo: when not tired
[02:33:44] <kcbabo> tfennelly:  sure thing ? get some sleep, man
[02:33:48] <tfennelly> kcbabo: ok... outa here... chat later
[02:33:56] *** tfennelly has quit IRC
[05:41:27] *** dbevenius has joined #switchyard
[05:44:35] *** ldimaggi has quit IRC
[05:46:52] *** kcbabo has quit IRC
[06:05:35] *** dbevenius has quit IRC
[06:26:40] *** magesh has joined #switchyard
[08:01:27] *** aslak has joined #switchyard
[08:01:28] *** aslak has quit IRC
[08:01:28] *** aslak has joined #switchyard
[08:14:51] *** tcunning has quit IRC
[08:30:55] <wyer> magesh, morning
[08:32:39] <magesh> wyer: Hi Justin
[08:33:14] <wyer> magesh, everything works great with your dependency fixes
[08:33:34] <wyer> magesh, just testing on AS 6 the quickstart demo failed to deploy
[08:33:39] <magesh> wyer: Okay cool
[08:33:55] <magesh> wyer: What error you get on AS6
[08:34:35] <wyer> magesh, not sure, arquillian said the deployment failed in the unit test error log, starting AS 6 and deploying manually now to see why
[08:35:39] <magesh> wyer: Okay
[08:36:00] <magesh> wyer: Might be the bpm files are not configured
[08:36:04] <wyer> magesh, looks like switchyard-component-bpm-0.2.0-SNAPSHOT.jar is missing from the deploy dir
[08:36:16] <wyer> magesh, yeah :)
[08:36:32] <magesh> wyer: Correct, we did not add that in AS6 release
[08:47:32] *** dbevenius has joined #switchyard
[08:47:43] *** wyer has left #switchyard
[08:47:49] *** wyer has joined #switchyard
[09:12:16] <wyer> magesh, just working through dependencies on the as6 side, got to do some other things now, will complete it later
[09:16:52] <magesh> wyer: okay
[10:22:05] *** aslak has quit IRC
[10:22:24] *** aslak has joined #switchyard
[10:22:28] *** aslak has quit IRC
[10:22:28] *** aslak has joined #switchyard
[10:28:24] *** aslak has quit IRC
[10:28:44] *** aslak has joined #switchyard
[11:05:16] *** rbalent has joined #switchyard
[11:37:45] *** tfennelly has joined #switchyard
[12:09:29] *** GitHub139 has joined #switchyard
[12:09:29] <GitHub139> [core] tfennelly created SWITCHYARD-289 (+1 new commit): https://github.com/jboss-switchyard/core/compare/d408e00^...d408e00
[12:09:29] <GitHub139> [core/SWITCHYARD-289] SWITCHYARD-289 Create service unit tests from forge - Keith Babo
[12:09:29] *** GitHub139 has left #switchyard
[12:11:53] *** GitHub14 has joined #switchyard
[12:11:53] <GitHub14> [core] tfennelly pushed 1 new commit to master: https://github.com/jboss-switchyard/core/commit/d408e00c41190b53db51f00619c334d23c0f4cb0
[12:11:53] <GitHub14> [core/master] SWITCHYARD-289 Create service unit tests from forge - Keith Babo
[12:11:53] *** GitHub14 has left #switchyard
[12:12:57] *** GitHub105 has joined #switchyard
[12:12:57] <GitHub105> [components] tfennelly pushed 1 new commit to master: https://github.com/jboss-switchyard/components/commit/b89d6e41bf40709be7c1a17594bc634e32996a65
[12:12:57] <GitHub105> [components/master] SWITCHYARD-289 Create service unit tests from forge - Keith Babo
[12:12:57] *** GitHub105 has left #switchyard
[12:57:49] *** dbevenius has quit IRC
[12:58:38] *** tcunning has joined #switchyard
[13:03:33] *** kcbabo has joined #switchyard
[13:04:33] <tfennelly> kcbabo: I pushed those changes for you but couldn't update JIRA because it's offline
[13:05:18] <tfennelly> kcbabo: actually... it's back now :)
[13:05:31] <kcbabo> tfennelly:  thanks
[13:05:53] <tfennelly> kcbabo: nps
[13:12:21] <wyer> magesh, in some kind of classpath hell with AS6 tests
[13:12:37] <wyer> hibernate is very unhappy
[13:12:41] <magesh> wyer: Do you want me to look at it?
[13:13:04] <magesh> wyer: I haven't done anything with AS6 locally
[13:13:08] <tfennelly> kcbabo: so want me to look at tomcat deployment?
[13:13:29] <wyer> magesh, yeah i might let me if i can spot duplicate hibernates on the CP first
[13:13:45] <tfennelly> kcbabo: well... really webapp deployment I guess
[13:14:12] <magesh> wyer: Let me add the bpm component and see what happens
[13:14:36] <wyer> magesh, the component is fine
[13:14:55] <magesh> wyer: Then the test is failing?
[13:14:59] <wyer> magesh, its the actual test that is foobar ;) it can't start the HT server
[13:15:08] <wyer> cos of hibernate wackiness
[13:15:34] <wyer> magesh, think i spotted it hibernate-annotations 3.4 is a dependency as well as hibernate 3.6
[13:16:16] <magesh> wyer: You mean bpm component needs those?
[13:16:25] <wyer> magesh, need to find where these are been added into the dependency hierarchy
[13:16:51] <magesh> wyer: Probably a transitive deps from drools-persistence-jpa
[13:16:54] <wyer> magesh, jbpm-human-task needs hibernate but for some reason old hibernate deps are on the CP
[13:17:24] <kcbabo> tfennelly:  yeah, I think it's worth a quick look to see what a stand-alone bootstrap would look like
[13:17:53] <kcbabo> tfennelly:  then that same thing could potentially be used inside a context filter or something to bootstrap inside a webapp
[13:18:04] <tfennelly> kcbabo: okidoki
[13:18:44] <tfennelly> kcbabo: right .. a context listener or something like that
[13:18:55] <kcbabo> wyer: are you able to deploy and test the bpm app outside of the test?
[13:19:19] <wyer> magesh, no 3.6 is from drools
[13:19:19] <wyer> org.switchyard:switchyard-test and org.jboss.jbossas:jboss-as-depchain
[13:19:24] <wyer> no that is netbeans lying to me
[13:19:55] <magesh> wyer: Let me do a build locally
[13:20:05] <wyer> kcbabo, will do that now and let you know
[13:20:26] <kcbabo> wyer:  the idea behind the demo apps is that they require a bit more user understanding/involvement to run
[13:20:46] <kcbabo> wyer:  in other words, they can be a bit more complicated then the normal quickstarts
[13:21:18] <kcbabo> wyer:  that helpdesk app has a ton of stuff in it to kick the workflow when human tasks are encountered
[13:21:40] <kcbabo> wyer:  so if that's complicating things, we can create a simpler bpm quickstart and use that for automated testing
[13:21:55] <kcbabo> wyer:  but that helpdesk app should still be able to deploy and run on AS7
[13:22:02] <wyer> kcbabo, yep and it works just not within the as6 dist tests ( works with as 7 ) i have a test case to kick off all that stuff
[13:22:07] <kcbabo> wyer:  just a matter of whether it's a PITA to test
[13:22:51] <kcbabo> wyer:  oh geez - just realized you were talking about AS6 and not AS7
[13:23:00] <wyer> kcbabo, so the current status is that the release will test as 7 with the helpdesk app and all the human task goodness but as6 not so much ( will test it manually now )
[13:25:34] <wyer> kcbabo, works like a charm let me see if i can dump the hibernate 3.6 dependencies some how
[13:28:31] <magesh> wyer: Can you share your jboss-as6/pom.xml
[13:28:47] <wyer> kcbabo, can i remove jboss-as-depchain dependency ?
[13:29:21] <wyer> kcbabo, and add the actual deps ofc
[13:29:26] <wyer> magesh, sure one sec
[13:29:29] <kcbabo> wyer:  I have no idea - magesh and tfennelly are the best to ask on that front
[13:29:33] *** dbevenius has joined #switchyard
[13:29:35] <kcbabo> wyer:  I'm not familiar with that dependency
[13:29:43] <wyer> magesh, comments ?
[13:30:50] <tfennelly> wyer: depchain dependency... refresh my mem... what is that?
[13:33:02] <wyer> tfennelly, its inside release/test/pom.xml i think it was a lazy way to get certain as6 deps for logging from tests maybe
[13:33:27] <wyer> magesh, https://gist.github.com/1122411
[13:33:51] <magesh> wyer: Thanks
[13:34:05] <wyer> magesh, everything is working by the way, the human task stuff just can't start when run from maven for the as6 tests
[13:34:58] <wyer> tfennelly, do you mind if i remove that dep and add the actual required deps so that all the tests pass ?
[13:35:53] <tfennelly> wyer: I think that would be fine
[13:36:14] <tfennelly> wyer: might be worth doing a little bit of analysis to see what the difference is
[13:36:55] <magesh> wyer: building...
[13:37:38] <wyer> magesh, let me push my SWITCHYARD-336 branches
[13:38:16] <magesh> wyer: Okay cool, I needed the test anyhow
[13:44:10] <wyer> magesh, you can pull my SWITCHYARD-336 branch on my release repo
[13:44:50] <magesh> wyer: Okay
[13:51:08] <magesh> wyer: I get org.apache.http.impl.client.DefaultHttpClient - I/O exception (org.apache.http.NoHttpResponseException) caught when processing request: The target server failed to respond
[13:52:02] <wyer> magesh, hrm did the app server start correctly ?
[13:52:28] <magesh> wyer: Server started.
[13:52:53] <magesh> wyer: Looks like the test passed too
[13:53:22] <magesh> wyer: Oh wait it printed *** Could not start task server: class org.hibernate.cfg.ExtendedMappings has interface org.hibernate.cfg.Mappings as super class ***
[13:54:08] <wyer> magesh, yeah it passes but not really (need to make sure it fails but lets solve the problem first)
[13:54:53] <wyer> magesh, the problem is that jboss-as-depchain adds a million dependencies which are not really needed once of those is hibernate-3.6 and its conflicting with the jbpm hibernate deps
[13:57:41] <magesh> wyer: hibernate-3.6 is the one bundled with AS6
[13:58:09] <wyer> magesh, yes but its not a dependency of anything under release/test
[13:58:32] <magesh> wyer: What version jbpm uses?
[13:58:41] <wyer> magesh, 3.4 i think
[13:59:05] <wyer> magesh, only the human-task jbpm module uses it
[13:59:49] <wyer> magesh, but we're adding those jboss 6 deps to the release/test module which has nothing to do with jboss 6 (other than deploying the tests to it) so we don't need those deps
[14:00:23] <magesh> wyer: I guess Arquillian picks it up from the AS6 directories. Did you remove the as6 depchain and does it work then?
[14:01:17] <wyer> magesh, i removed the dep-chain and it blew up with a jboss logger CDNFE but that probably the only or one of only a few real deps
[14:01:56] <wyer> there is no need for every jboss 6 library to be a dependency
[14:02:24] <magesh> wyer: In anycase if the bpm component uses a different version than what is shipped with AS6 and it is not compatible then we need to talk to kcbabo and see what he suggests.
[14:02:56] <wyer> magesh, bpm component does not use human tasks
[14:03:39] <wyer> magesh, at least not yet
[14:03:48] <magesh> wyer: So it is only the test, can't we remove human tasks from the test then? kcbabo?
[14:03:58] <wyer> magesh, its only the test using it
[14:04:46] <magesh> wyer: Do the manual testing pass?
[14:05:08] <wyer> magesh, yes
[14:05:55] <magesh> wyer: I will let Keith decide on this. Sorry man you had to go through such an ordeal for just adding test
[14:06:31] <wyer> magesh, no probs
[14:06:47] <wyer> kcbabo, tfennelly verdicts ?
[14:07:03] <magesh> wyer: What if AS6 users want to automate their app and try using human tasks will it then fail in AS6?
[14:17:34] <wyer> magesh, it will be fine we don't package those dependencies
[14:18:21] <wyer> magesh, the problem solely exists with the test because we have two conflicting transitive dependencies at runtime
[14:19:39] <wyer> magesh, so human-tasks would use the jboss 6 version on hibernate if someone did that
[14:19:44] <wyer> *of
[14:22:53] <magesh> wyer: Can we override the AS6 versions of hibernate in the pom dependency for test?
[14:23:34] <wyer> magesh, we don't need the dependencies
[14:24:47] <magesh> wyer: Hmm, I am all out of ideas then :)
[14:25:09] <wyer> magesh, what i am saying is that we can strip out the as 6 deps from the test module
[14:25:42] <wyer> magesh, just add the deps that are really needed, instead of napalming dependencies in for no real reason
[14:25:57] <magesh> wyer: But then that would be like testing AS6 with our own dependencies, but may not run with an actual AS6 + switchyard combo
[14:27:17] <wyer> magesh, no the test code runs outside of AS6 only the helpdeskdemo is running inside AS6
[14:31:20] <magesh> wyer: Exactly that is what I meant. The app helpdeskdemo using hibernate-3.6 libraries inside AS6 right? And the test code wants to use 3.4, am I right?
[14:31:46] <wyer> magesh, no the helpdeskdemo does not use hibernate at all
[14:32:02] <magesh> wyer: via bpm?
[14:32:13] <wyer> magesh, nope
[14:32:29] <wyer> magesh, the human task stuff happens in the helpdeskdemo unit test
[14:32:43] <wyer> magesh, its not part of the app itself
[14:33:21] <magesh> wyer: I am thinking way too tangetly :)
[14:33:37] <wyer> magesh, hehe
[14:34:29] <magesh> wyer: I will take a walk now
[14:34:57] <wyer> magesh enjoy it!
[14:52:48] *** antollinim has joined #switchyard
[14:55:04] *** ldimaggi has joined #switchyard
[14:55:16] *** magesh has left #switchyard
[14:55:42] *** ldimaggi has quit IRC
[14:55:47] *** ldimaggi has joined #switchyard
[14:56:26] *** mriet has joined #switchyard
[15:35:07] <ldimaggi> kcbabo, ping
[15:38:24] *** lanceball has joined #switchyard
[15:44:58] *** bfitzpat has joined #switchyard
[17:10:08] *** dbevenius has quit IRC
[17:13:33] *** GitHub50 has joined #switchyard
[17:13:33] <GitHub50> [core] kcbabo pushed 1 new commit to master: https://github.com/jboss-switchyard/core/commit/ebfc6070ae80404d5e063764d8e0681be2531305
[17:13:33] <GitHub50> [core/master] SWITCHYARD-358 "Create a HornetQ test MixIn" - Daniel Bevenius
[17:13:33] *** GitHub50 has left #switchyard
[17:28:43] *** mriet has quit IRC
[17:58:22] *** bfitzpat is now known as bfitzpat_drappt
[18:26:19] *** rbalent has quit IRC
[18:39:06] *** ldimaggi has quit IRC
[19:15:43] <antollinim> kcbabo: hi Keith, I want to take a Jira. Is this on ok?: https://issues.jboss.org/browse/SWITCHYARD-282
[19:17:37] <kcbabo> antollinim:  sure, although I'm not sure that one is really our problem
[19:18:05] <antollinim> kcbabo: you prefer another one?
[19:19:21] <kcbabo> antollinim:  I'm not sure there is anything in the 0.2 bucket that jumps out
[19:19:42] <kcbabo> antollinim:  one thing we haven't done yet is unit testing for our forge commands
[19:19:49] <kcbabo> antollinim:  that would take some research though
[19:19:59] <kcbabo> antollinim:  look at how lincoln does it in the forge codebase
[19:20:16] <kcbabo> antollinim:  then come back and apply the same approach to our commands
[19:20:32] <antollinim> kcbabo: what do you mean by "commands"?
[19:21:43] <kcbabo> antollinim:  we have a set of commands in forge that allow you to do things like "create a bean service"
[19:21:50] <kcbabo> antollinim:  are you familiar with forge?
[19:22:15] <antollinim> kcbabo: yes, but just as a forge user
[19:22:52] <antollinim> kcbabo: in which project are those commands?
[19:23:09] <kcbabo> antollinim:  core/tools/forge and components/tools/forge
[19:23:34] <antollinim> kcbabo: good, so I need to create tests for the commands... thanks
[19:24:05] <kcbabo> antollinim:  I'm not exactly sure how well we can test
[19:24:11] <kcbabo> antollinim:  but it's bugging me that we have no tests there
[19:24:29] <antollinim> kcbabo: that is the research part :)
[19:24:35] <kcbabo> antollinim:  which means that if something changes, I normally manually test the commands to make sure they still work
[19:24:41] <kcbabo> antollinim:  which is pretty stinky
[19:24:52] <kcbabo> antollinim:  right, I think the first part is just research - to see what's possible
[19:25:09] <kcbabo> antollinim:  we need to start doing this for sure, it's just a matter of someone spending some time on figuring out how it can be done
[19:25:23] <kcbabo> antollinim:  if anything pops up for 0.2 in the meantime, I will definitely give you a heads up
[19:26:03] <antollinim> kcbabo: may I know how you test this manually? can you give me an example of a command you run and the expected output?
[19:28:17] <kcbabo> antollinim:  the best way to learn about that is to read the code and see what it does
[19:28:26] <kcbabo> antollinim:  so each component will have it's own plugin
[19:28:44] <antollinim> kcbabo: good.. thanks!
[19:28:44] <kcbabo> antollinim:  and in that plugin, each command method is annotated with @Command
[19:29:01] <kcbabo> antollinim:  so if you look at what that command does, the test would have to assert that it really happened
[19:29:28] <kcbabo> antollinim:  so for example : the soap plugin has a "soap-binding bind-service" command
[19:29:40] <antollinim> kcbabo: I think I got it
[19:29:46] <kcbabo> antollinim:  if you run that command against a service, you should be able to assert in the test that the configuration was created
[19:30:03] <kcbabo> antollinim:  and if you pass a bad service name, you should be able to assert that it failed
[19:34:00] <kcbabo> antollinim:  after all this, I will understand if you prefer to take SWITCHYARD-282 :-)
[19:42:04] <antollinim> kcbabo: haha
[19:42:18] <kcbabo> antollinim:  sorry, requirements are a bit rough
[19:43:19] <antollinim> kcbabo: don't worry, I'm good with this
[19:44:01] <kcbabo> antollinim:  okie dokie
[19:57:20] *** errantepiphany has joined #switchyard
[20:04:37] *** antollinim has left #switchyard
[20:23:00] <kcbabo> wyer:  yt?
[20:23:11] <wyer> kcbabo, hi
[20:23:11] <kcbabo> wyer:  I'm OK with just including that test for AS7
[20:23:34] <kcbabo> wyer:  I will create a separate BPM test which is much smaller in scope that we can use for both AS6 and AS7
[20:23:52] <wyer> kcbabo, okay
[20:24:03] <kcbabo> wyer:  sorry I missed your question before
[20:24:08] <kcbabo> wyer:  had a morning meeting
[20:24:41] <wyer> kcbabo, was busy trying to sort it out now but i like your plan better
[20:25:04] <kcbabo> wyer:  cool, sounds like that works for everyone :-)
[20:25:22] <kcbabo> wyer:  did magesh talk to you about the version of mvel?
[20:25:24] <wyer> kcbabo, AS6 is all kinds of unhappy if arquillian starts it without that dep
[20:25:44] <wyer> kcbabo, nope, does it need to change ?
[20:25:59] <kcbabo> wyer:  yeah, it's listed as beta3 in your current pull request
[20:27:17] <kcbabo> wyer:  looks like jbpm 5.1 released with org.mvel:mvel2:2.1.0.drools2
[20:27:26] <wyer> kcbabo, i can change it too the drools one
[20:27:32] <wyer> kcbabo, yeah :)
[20:27:45] <kcbabo> wyer:  let's try that GAV with spaces:
[20:27:45] <kcbabo> org.mvel  :  mvel2  :  2.1.0.drools2
[20:27:56] <kcbabo> wyer:  basically the version should be 2.1.0.drools2
[20:28:05] <wyer> kcbabo, roger will update that
[20:28:10] <kcbabo> wyer:  thanks very much
[20:28:19] <wyer> kcbabo, will be done in a few min
[20:28:28] <kcbabo> wyer:  fantastic!
[20:32:39] <wyer> kcbabo, just squashing these commits
[20:36:55] *** bfitzpat_drappt is now known as bfitzpat
[20:42:15] *** survery has joined #switchyard
[20:42:21] *** survery has left #switchyard
[20:46:45] *** dbevenius has joined #switchyard
[21:20:42] <kcbabo> wyer:  we good to go?  I see you updated core and just saw the pull for release.
[21:21:08] <wyer> kcbabo, yep
[21:21:22] <wyer> kcbabo, core, components and release
[21:22:05] <kcbabo> wyer:  awesome ? you da man
[21:22:12] <kcbabo> wyer:  I'll push them in a sec
[21:24:03] <wyer> kcbabo, hooray :)
[21:24:13] <kcbabo> long road ;-)
[21:24:21] <kcbabo> sorry about the mess with AS6
[21:24:40] <wyer> kcbabo, its cool sorry it took so long work was crazy
[21:24:58] <wyer> kcbabo, but i start a new job for a redhat advanced partner here in SA next month!
[21:26:23] <kcbabo> wyer:  sweet
[21:26:36] <wyer> kcbabo, they're a linux shop but redhat is pushing them to move into the jboss space
[21:26:45] <kcbabo> wyer:  and no worries on the timing ? we were happy to pick up from where you left off, but turned out we didn't have to
[21:26:59] <kcbabo> wyer:  nice
[21:27:07] <kcbabo> wyer:  jboss is your friend
[21:27:51] <wyer> kcbabo, i'll sign a company CLA once i start there i think so will have a bit more time to contribute
[21:28:15] <wyer> kcbabo, amen its also great fun to contribute
[21:28:58] <wyer> kcbabo, getting back to doing what you love
[21:29:25] <wyer> none of the drama heh
[21:30:23] <wyer> kcbabo, let me know if all the tests go through
[21:30:41] <kcbabo> wyer:  will do ? just finishing up my own JIRA right now
[21:30:46] <wyer> i expect they will
[21:31:09] <wyer> just now i did something stupid while merging from magesh
[21:32:37] <wyer> kcbabo, i'm going to get rolling with https://issues.jboss.org/browse/SWITCHYARD-359
[21:32:58] <kcbabo> wyer:  dynamite
[21:33:10] <kcbabo> wyer:  should be a smoother road with that one
[21:33:16] <kcbabo> wyer:  now that you've paved it ;-)
[21:33:24] <wyer> kcbabo, yeah much smoother ;)
[21:37:39] <errantepiphany> kcbabo: Should the namespaceURI of the ServiceDomain's QName always match the targetNamespace of the <composite> (CompositeModel)?
[21:38:32] <kcbabo> errantepiphany:  yes
[21:38:46] <errantepiphany> kcbabo: Are you sure? Or is it only that way now since we only have one app per domain OTM?
[21:39:02] <kcbabo> errantepiphany:  ah, it actually does depend
[21:39:21] <kcbabo> errantepiphany:  right now we support the concept of one domain per app which delegates up to a root domain
[21:39:40] *** dbevenius has quit IRC
[21:39:41] <errantepiphany> kcbabo: but it's not always gonna be like that?
[21:39:49] <kcbabo> errantepiphany:  what we don't support at the moment is creating domains individually and then assigning applications to them
[21:40:12] <errantepiphany> kcbabo: well if that's on the roadmap, then I need to be careful of that. So.... is it?
[21:40:18] <kcbabo> errantepiphany:  right, eventually you should be able to create a domain and then reference that domain from an app
[21:40:42] <kcbabo> errantepiphany:  I'm currently on the fence about whether service domain name should be a QName at all
[21:40:51] <errantepiphany> kcbabo: ok
[21:40:58] <errantepiphany> kcbabo: you told me what I needed to know, thx.
[21:41:09] <kcbabo> errantepiphany:  right on ? that was easy
[21:42:16] <errantepiphany> kcbabo: makes my life a little harder, but I'm glad it was easy for you (sarcasm - LOL)
[21:42:56] <kcbabo> errantepiphany:  I would hope for now it would be easier
[21:43:12] <errantepiphany> kcbabo: no worries
[21:43:16] <kcbabo> errantepiphany:  root level domain has no namespace like it does today
[21:43:31] <kcbabo> errantepiphany:  only other type of domain is the application domain
[21:43:49] <kcbabo> errantepiphany:  and what are we using for the name for that today?
[21:44:24] <errantepiphany> kcbabo: I don't know right off the bat... I'm just looking at the API.
[21:44:43] <tfennelly> kcbabo, errantepiphany: so it depends on whether the domain/app model stays or goes I guess
[21:45:16] <kcbabo> errantepiphany:  for the jira you are working on now, there should not be difficult changes in this area
[21:45:27] <kcbabo> errantepiphany:  so if you come across something that doesn't line up right, we should talk
[21:45:46] <kcbabo> errantepiphany:  we don't have anything that relies on service domain names to be namespace-qualified today
[21:46:01] <kcbabo> errantepiphany:  and as tfennelly referenced above, we are reviewing that model and there will be changes in 0.3
[21:46:15] <kcbabo> errantepiphany:  so the path of least resistance here is best
[21:46:32] <kcbabo> tfennelly:  I think that model might stick around with possibly an intervening layer
[21:46:46] <kcbabo> tfennelly:  root ?> independent domain ?> application domain
[21:47:09] <kcbabo> tfennelly:  where the independent domain is optional, so you could have
[21:47:14] <kcbabo> tfennelly:  root ?> application domain
[21:47:19] <kcbabo> tfennelly:  which is what we have today
[21:47:24] <tfennelly> kcbabo: I think we need something that reps an "application"... at the moment that's really being done by domain
[21:47:38] <tfennelly> kcbabo: right... exactly
[21:47:42] <errantepiphany> kcbabo: it has proven more difficult than you might think, considering the lifecycle of Weld, and the fact that at the point I wanted the namespace, I have no handle on the deployment nor the switchyard model. I'm gonna try to post-process the registered bean service names in the activator's init, and see how that goes.
[21:48:31] <kcbabo> errantepiphany:  perhaps we should discuss this a bit more right now, cause I'm a bit unclear on why it's not straightforward
[21:48:51] <tfennelly> errantepiphany: in reality... the names on those bean services are not namespaced at that point ... so they prob really should not be QNames in the BeanMetaData thingie
[21:48:54] <kcbabo> errantepiphany:  there is no need to have the namespace in the CDI extension which discovers all the beans
[21:49:14] <kcbabo> tfennelly:  right, they are just local names
[21:49:22] <errantepiphany> tfennelly, kcbabo: yes, I've figured that out.
[21:49:24] <tfennelly> kcbabo: you don't have it at that point
[21:49:40] <kcbabo> tfennelly:  that's what I'm saying
[21:49:43] <errantepiphany> kcbabo: No, I don't wanna talk right now. I'm halfway through a coding change and I don't wanna do a context-switch.
[21:50:10] <tfennelly> errantepiphany, kcbabo: I'm thinking that bean metadata shouldjust have the localnames on them... the target namespace is added later to create the QName
[21:50:26] <errantepiphany> tfennelly, kcbabo: Yes, I'm making that change now.
[21:50:31] <kcbabo> tfennelly:  I agree 100%
[21:51:01] <errantepiphany> tfennelly: This means that not only does the beanmetadata have only localNames, but so do the servicedescriptors, right?
[21:51:09] <tfennelly> errantepiphany, kcbabo: the QName type on that object is confusing the picture a bit
[21:51:38] <tfennelly> errantepiphany: ehhh... yep, that would be the case
[21:51:40] <errantepiphany> tfennelly: lemme re-phrase
[21:51:53] <errantepiphany> tfennelly: the clientproxybeans as well as the servicedescriptors within the metadata.
[21:51:56] <tfennelly> errantepiphany: yer OK... I follow... I think that would be right Dave
[21:52:28] <tfennelly> errantepiphany: none of that stuff should have QNames... coz they're not QNames
[21:52:49] <tfennelly> errantepiphany: or at best... they're namespaceless QNames
[21:53:00] <errantepiphany> tfennelly: yeah, well I'd prefer Strings in that case. ;)
[21:53:22] <tfennelly> errantepiphany: would ya :)
[21:53:32] <tfennelly> errantepiphany: and some salt and pepper?
[21:54:10] <tfennelly> errantepiphany: anyway... yeah.... those should be changed to use String instead of QNames
[21:54:54] <errantepiphany> tfennelly: making that change now. thx.
[21:55:11] <tfennelly> errantepiphany: String for the app/domain name at the application scope.... merge the 2 strings together to get the fully qualified bean and service names
[21:55:38] <tfennelly> errantepiphany: and that merging can happen in the activator I think
[21:56:02] <tfennelly> errantepiphany: kinda feels wrong doing it there, but it gets messy doing it elsewhere I think
[21:56:10] <errantepiphany> tfennelly: agreed.
[21:56:19] <tfennelly> errantepiphany: could do it there to start anyway... get it working
[21:56:48] <tfennelly> errantepiphany: then circle back on it and see can we merge them in a more appropriiate place
[21:57:00] <errantepiphany> tfennelly: yup
[21:57:14] <tfennelly> errantepiphany: that's like something that should happen in the deployers
[21:57:28] <tfennelly> errantepiphany: but then you get into AS specific fixes
[21:57:50] <tfennelly> errantepiphany: which would obviously turn into a maint PITA
[21:59:52] <wyer> oh what is the reason for supporting AS6 and 7 ?
[21:59:59] <wyer> just wondering
[22:00:21] <errantepiphany> tfennelly: Hmmm... in SimpleCDIDeployment (which is supposedly only used in tests), in the deployServicesAndProxies method, there is a domain.registerService(serviceName, handler, serviceInterface) call. At this point, we only have a localName per the serviceDescriptor. Can I just register it with the namespaceURI of the domain, just since this is test code?
[22:01:44] <tfennelly> errantepiphany: maybe just use a hardcoded NS?
[22:02:06] <tfennelly> errantepiphany: are we not saying that the domain NS is not hte appropriate NS?
[22:03:08] <errantepiphany> yeah but this is for tests...
[22:03:22] <errantepiphany> I'm guessing they'll all be in the same domain.
[22:05:09] <tfennelly> errantepiphany: yeah... I'm just thinking for maint reasons... if someone sees the code... us using the domain NS... might cause confusion... that's why I'm suggesting using a hardcoded NS (which is fine coz it's just for testing)
[22:05:35] <tfennelly> errantepiphany: anyway... either will work fine
[22:05:45] <tfennelly> errantepiphany: take yer pick :)
[22:07:04] <tfennelly> aslak: hey Aslak... did they change the Shrinkwrap API... replace the "addResource" method with something else?
[22:10:27] <kcbabo> tfennelly:  is Smooks going to release within the next two weeks? :-)
[22:10:39] <kcbabo> tfennelly:  looking at the SNAPSHOT jira that's assigned to you
[22:11:44] <kcbabo> tfennelly:  it's currently at 1.5-SNAPSHOT
[22:12:43] <kcbabo> tfennelly:  assuming 1.5 will not arrive for 0.2, we should just push the jira to 0.3
[22:13:14] <kcbabo> although we really should try to get off a snapshot version in our release at some point ;-)
[22:18:39] <kcbabo> wyer:  still around?
[22:21:13] <wyer> kcbabo, yup
[22:21:21] <kcbabo> wyer:  just commented on the release pull
[22:21:35] <kcbabo> wyer:  what's up with 7.1.0 alpha SNAPSHOT?
[22:22:07] <wyer> kcbabo, hrm argh thats wrong
[22:25:18] <wyer> kcbabo, sorry could swear i changed that back, crap internet and no 7.0.0.Final in my local repo at some point so i made a plan ;)
[22:25:45] <wyer> kcbabo, just retesting release and then will push a change
[22:25:57] <kcbabo> wyer:  no worries - if you want to push the update to the version, I can test it here too
[22:26:01] <kcbabo> wyer:  up to you
[22:30:57] <aslak> tfennelly, at some point all of them were changed to addAsXXX
[22:31:36] <wyer> kcbabo, tested, pushing it now
[22:31:50] <kcbabo> wyer:  great
[22:32:11] <wyer> kcbabo, think i undid it the change back to 7.0.0.Final when i merged from magesh
[22:34:18] <wyer> kcbabo, you can pull
[22:40:32] <tfennelly> aslak: ah okidoki... thanks Aslak... I see it now... interesting change
[22:40:46] <tfennelly> kcbabo: really good question
[22:40:56] <tfennelly> kcbabo: no... that's not going to happen
[22:41:05] <tfennelly> kcbabo: don't have time to work on it these days\
[22:41:29] <tfennelly> kcbabo: so yeah... I think push that
[22:41:40] <tfennelly> kcbabo: I'm the one holding up that release at this stage
[22:41:47] <aslak> tfennelly, they were changed basically because there was some confusion to what addResource meant, are we adding a resource form the classloader to the archive, or are we adding it as a resource..  so addAsX it became. :)
[22:41:52] <aslak> they/there
[22:42:02] <aslak> no, they.. heh
[22:43:52] <tfennelly> aslak: I getcha
[22:44:02] <tfennelly> aslak: makes sense
[22:52:35] <wyer> kcbabo, i am off let me know if there are any problems
[22:55:26] <kcbabo> wyer:  will do
[23:00:42] <tfennelly> aslak: hey Aslak... using embedded tomcat
[23:00:44] <kcbabo> wyer:  just an FYI, I'm gonna squash that extra commit
[23:01:27] <tfennelly> aslak: I get a NoClassDef error
[23:01:46] <tfennelly> aslak: for some richfaces classes
[23:01:49] <tfennelly> aslak: javax.faces.FacesException
[23:02:09] <tfennelly> aslak: or... JSF I should say.... does that makes sense?
[23:02:17] <tfennelly> aslak: is there a hard dep?
[23:04:12] <tfennelly> aslak: ah... maybe I see my booboo :)
[23:06:16] *** aslak has quit IRC
[23:06:44] <tfennelly> aslak: sorry Aslak... I had jsf-impl on the classpath... test is deploying for me now... sorry for the BS :)
[23:11:42] *** kcbabo has quit IRC
[23:32:42] *** errantepiphany has left #switchyard





top