[01:43:09] *** aslak has quit IRC
[03:26:49] *** ldimaggi has joined #switchyard
[05:34:25] *** ldimaggi has quit IRC
[06:37:01] *** magesh has joined #switchyard
[07:14:54] *** magesh has quit IRC
[07:15:29] *** magesh has joined #switchyard
[07:53:10] *** magesh1 has joined #switchyard
[07:56:40] *** magesh has quit IRC
[09:49:54] *** magesh has joined #switchyard
[09:51:02] *** magesh1 has quit IRC
[10:49:32] *** aslak has joined #switchyard
[12:00:11] *** magesh has quit IRC
[12:15:06] *** magesh has joined #switchyard
[12:42:59] *** tfennelly has joined #switchyard
[12:43:21] *** tfennelly has quit IRC
[13:02:12] *** kcbabo has joined #switchyard
[13:32:11] <kcbabo> magesh:  hey magesh
[13:32:48] <magesh> kcbabo: Hi Keith, God Morning. Hope you had a great 4th of July!
[13:33:07] <kcbabo> magesh:  not too bad thanks ? hope you had a chance to rest and recover as well
[13:34:56] <kcbabo> magesh:  wanted to confirm something with you real quick re: AS7
[13:35:07] <magesh> kcbabo: Shoot
[13:36:10] <kcbabo> magesh:  IIRC, we establish a set of module dependencies for each deployment which includes the installed components
[13:36:13] <kcbabo> magesh:  is that right?
[13:36:35] <magesh> kcbabo: Yes
[13:37:14] <kcbabo> magesh:  in that case, any exported package in a module would be visible to the deployment, yeah?
[13:37:23] <magesh> kcbabo: Yes
[13:37:29] <kcbabo> magesh:  ok, that's it :-)
[13:37:37] <kcbabo> magesh:  easy questions after a three-day weekend
[13:37:45] <magesh> kcbabo: See my recent pull request :)
[13:37:47] <kcbabo> magesh:  it's about all I can handle ;-)
[13:37:56] <magesh> kcbabo: https://github.com/jboss-switchyard/core/pull/226
[13:37:58] <kcbabo> magesh:  ah, fantastic
[13:38:10] <kcbabo> magesh:  I was just talking to aamonten about this last week
[13:38:32] <magesh> kcbabo: Cool, I am way ahead of you guys ;)
[13:38:32] <kcbabo> magesh: he was looking for an area to help and I said that might be a good one
[13:38:36] <kcbabo> magesh:  but you're too fast!
[13:39:13] <kcbabo> magesh:  I'll push that this morning ? need to take care of a few other things first
[13:39:22] <kcbabo> magesh:  any scary issues or is it pretty much good to go?
[13:39:46] <magesh> kcbabo: See the comment about https://github.com/jboss-switchyard/release/pull/13
[13:39:54] <magesh> kcbabo: Not scary though
[13:40:53] <kcbabo> magesh:  ok, so it's just the hanging issue with a failed deployment then?
[13:41:07] <kcbabo> magesh:  is that a known issue with ARQ?
[13:41:39] <magesh> kcbabo: Not sure, just discovering as always
[13:41:45] <magesh> kcbabo: :)
[13:42:09] <kcbabo> magesh:  ok, I thought there was some sorta master timeout available
[13:42:57] <magesh> kcbabo: I will look into that. If any of you guys get that, please post it here or the forums
[13:42:59] <kcbabo> magesh:  main thing I'm wondering about is our hudson build ? I'm sure these tests are gonna fail at some point and if there's a hang, I'm wondering how we will (a) detect and (b) cleanup
[13:43:10] <kcbabo> magesh:  sure thing ? one step at a time
[13:43:14] <magesh> kcbabo: :)
[13:43:29] <kcbabo> magesh:  I'm certainly happy with the start :_0
[13:43:31] <kcbabo> :-)
[13:44:04] <magesh> kcbabo: I am happy now nobody can get past the build before they get it right for AS7, atleast the deployment ;)
[13:44:33] <kcbabo> magesh:  yeah, we totally needed that
[13:44:59] <magesh> kcbabo: I am sorry for not getting the dummy Test filled
[13:49:23] <kcbabo> magesh:  dummy test?
[13:49:42] <magesh> kcbabo: https://github.com/jboss-switchyard/release/pull/13/files#diff-4
[13:50:26] <magesh> kcbabo: Just tests the deployment, but doesn't run an actual end-to-end test
[13:50:45] <kcbabo> magesh:  ok, so deployment failure causes the test to fail, but we are not asserting an outcome?
[13:51:39] <kcbabo> magesh:  that's what I would expect with a number of those quickstarts as verifying external to the unit test in the QS is difficult
[13:51:51] <magesh> kcbabo: Test to hang exactly. Not right now, I would need to help from the individual contributors of that QS or need time to add on my own
[13:52:23] <magesh> kcbabo: Oh we are in the wavelength okay :)
[13:52:29] <kcbabo> magesh:  hehe
[13:52:42] <kcbabo> magesh:  I see that orders test does assert the outcome, so that's good
[13:53:02] <magesh> kcbabo: That was a copy from AS6, thanks to TomF
[13:54:31] <magesh> kcbabo: We can still move the AS6 test to its own sub-folder and move common classes to somewhere else, all are minor issues for me
[13:54:49] <magesh> kcbabo: Which we can handle later
[14:15:09] *** antollinim_afk has joined #switchyard
[14:19:45] *** antollinim_afk is now known as antollinim
[14:22:44] *** PeteRoyle has quit IRC
[14:23:05] <antollinim> Hey Keith, how are you? I have implemented SWITCHYARD-322. I would like you to take a look at it before doing the pull request.
[14:23:28] <antollinim> kcbabo: do you want me to commit it and you take it from my repository?
[14:25:05] <kcbabo> antollinim:  yes, best way to do it is to push to a topic branch in your repo and then send the linl
[14:25:07] <kcbabo> link
[14:25:30] <antollinim> kcbabo: good, I'll do it
[14:43:14] *** PeteRoyle has joined #switchyard
[14:52:41] <antollinim> kcbabo: https://github.com/antollinim/components/tree/SWITCHYARD-322
[14:52:57] <kcbabo> antollinim:  k, thanks mari
[14:52:58] <kcbabo> o
[14:53:05] <kcbabo> antollinim:  I'll have a look in a bit
[14:53:20] <antollinim> kcbabo: perfect, I'll grab a coffee now
[14:53:40] *** ldimaggi has joined #switchyard
[15:00:19] *** bfitzpat has joined #switchyard
[15:16:34] *** magesh has left #switchyard
[15:16:59] <kcbabo> antollinim:  I wonder if it would be better to use an empty string for the default on name?
[15:17:38] <kcbabo> antollinim:  this has the added benefit of explicitly excluding the empty string as something the user can enter
[15:18:32] <antollinim> kcababo: I didn't get what you mean with the benefit
[15:19:30] <antollinim> kcbabo: any string for the default value is actually the same because, ones the model has scanned it the value dissapears from it
[15:20:14] <kcbabo> antollinim:  is this valid?
[15:20:14] 
[15:20:42] <antollinim> kcbabo: I did not try that, but it sould be valid
[15:21:02] <kcbabo> antollinim:  in that case, you are creating a service with no name
[15:21:07] <kcbabo> antollinim:  that doesn't seem valid to me
[15:21:21] <antollinim> antollinim: I mean "should" as it is implemented, but maybe that is not a good idea to get it allowed
[15:21:33] <kcbabo> antollinim:  right :-)
[15:21:39] <antollinim> kcbabo: I understand what you mean
[15:22:03] <kcbabo> antollinim:  now, we really should have a check for this later down the line in the deployer
[15:22:05] <antollinim> kcbabo: so, it does not matter where the empty value cames from, the scanner will take that as a NULL
[15:22:13] <kcbabo> antollinim:  but I think we can kill two birds with one stone here
[15:22:47] <antollinim> kcbabo: so, good, the null value will be ""
[15:23:03] <kcbabo> antollinim:  correct - or stating it differently, empty string is equivalent to not specifying a service name
[15:23:49] <antollinim> kcbabo: any other thing?
[15:24:26] <kcbabo> antollinim:  right now, the example shows a QName being passed in as the value ("{urn:bleh}CustomServiceAnnotationService")
[15:24:40] <kcbabo> antollinim:  but we are not parsing that as a QName in the scanner
[15:24:45] <kcbabo> antollinim:  just as a string
[15:25:00] <antollinim> kcbabo: yes, just a string
[15:25:06] <kcbabo> antollinim:  to be honest, I'm not sure we're prepared to support a separate namespace being used as the service name
[15:25:16] *** tcunning has joined #switchyard
[15:25:30] <kcbabo> antollinim:  so my suggestion would be to change the test annotation to just be a simple string (no namespace URI)
[15:25:40] <kcbabo> antollinim:  minor detail, but it will help avoid confusion
[15:25:46] <kcbabo> antollinim:  other than that, it looks good to go
[15:25:49] <antollinim> kcbabo: you are right
[15:41:13] *** lanceball has joined #switchyard
[15:46:12] *** lanceball has quit IRC
[15:51:42] *** lanceball has joined #switchyard
[16:04:58] <antollinim> kcbabo: I added a service with an empty name and the test checks the empty name is not there now
[16:05:32] <kcbabo> antollinim:  cool, thanks
[16:06:00] <antollinim> kcbabo: I want more :)
[16:07:06] <kcbabo> antollinim:  that's the spirit!  :-)
[16:07:24] <antollinim> kcbabo: whenever you have time hint me on what else is needed, no rush
[16:07:34] <kcbabo> antollinim:  okie dokie
[16:10:37] *** rcernich has joined #switchyard
[16:20:22] *** ldimaggi has quit IRC
[16:27:34] *** errantepiphany has joined #switchyard
[16:50:25] *** ldimaggi has joined #switchyard
[17:16:07] *** tfennelly has joined #switchyard
[17:16:24] <tfennelly> kcbabo: hey Keith
[17:16:44] <kcbabo> tfennelly:  hey tom!
[17:17:24] <tfennelly> kcbabo: hey... so something has changed in the AS re how it handles dependencies
[17:17:56] <kcbabo> tfennelly:  yeah, did you see my comments on the JIRA?
[17:18:57] *** rcernich is now known as rcernich_away
[17:18:58] <tfennelly> kcbabo: I did
[17:19:22] <kcbabo> tfennelly:  so something definitely changed - my guess is they locked down some of the dependency mapping
[17:19:34] <kcbabo> tfennelly:  either that, or there's a subtle difference in how we are configuring our deployer
[17:19:41] <kcbabo> tfennelly:  but I didn't see anything obvious in there
[17:20:24] <tfennelly> kcbabo: rightio
[17:21:17] <tfennelly> kcbabo: so maybe wrapping the saaj type is the best thing to do?
[17:21:55] <tfennelly> kcbabo: and as a general rule... we should try keeping types going into Message as something we can "get" easily
[17:22:04] <tfennelly> kcbabo: just like you were saying actually
[17:22:23] <kcbabo> tfennelly:  yeah, that makes sense to me
[17:22:29] <kcbabo> tfennelly:  I'll start a dev forum thread on it
[17:22:46] <kcbabo> tfennelly:  good news is that I don't think we need to do anything super nasty in our plumbing
[17:22:49] <tfennelly> kcbabo: cool
[17:23:01] <tfennelly> kcbabo: right
[17:23:08] <kcbabo> tfennelly:  just be sensible about the types we use and document how to work with other types if required
[17:23:17] <tfennelly> kcbabo: so would you mind commiting that change?
[17:23:35] <tfennelly> kcbabo: you have that already, right?
[17:23:45] <kcbabo> tfennelly:  the change for the module definition?
[17:24:07] <tfennelly> kcbabo: no... wrapping the SAAJ Element in a DOMSource
[17:24:14] <tfennelly> kcbabo: wasn't that what you did?
[17:24:26] <kcbabo> tfennelly:  I did both actually
[17:24:38] <tfennelly> kcbabo: thought we said we'd do that Vs dicking about with the dependencies?
[17:24:49] <kcbabo> tfennelly:  but  I can commit the Element change, but there are some other things associated with that
[17:24:50] <tfennelly> kcbabo: ah right
[17:25:05] <kcbabo> tfennelly:  need to change the tests for SOAP component
[17:25:17] <kcbabo> tfennelly:  and the assert method in the composer specifically requires Node
[17:25:19] <tfennelly> kcbabo: so if you have that and can commit it then I can do the other stuff... that OK?
[17:25:46] <tfennelly> kcbabo: ah right... gotcha
[17:25:56] <kcbabo> tfennelly: why don't we do this ? I will commit the simple fix to include the SAAJ impl class for now
[17:26:03] <kcbabo> tfennelly:  and then I will file a JIRA to clean it up proper
[17:26:12] <tfennelly> kcbabo: okidoki... cool
[17:26:18] <kcbabo> tfennelly:  and magesh can handle the changes to the assert method, the tests, etc.
[17:26:28] <tfennelly> kcbabo: I might not have much time on this for the rest of the week
[17:26:35] <kcbabo> tfennelly:  grand ?
[17:26:42] <tfennelly> kcbabo: small window of opportunity today :)
[17:26:48] <kcbabo> tfennelly:  give me 15 minutes
[17:26:52] <tfennelly> kcbabo: so if we're quick
[17:26:54] <kcbabo> tfennelly:  okie dokie :-)
[17:27:27] <kcbabo> tfennelly:  tell you what - you can just take the updated module definition and use it in your environment
[17:27:35] <kcbabo> tfennelly:  and then issue the pull requests
[17:28:01] <kcbabo> tfennelly:  I will make sure the module update gets committed before your stuff is pushed
[17:30:10] *** errantepiphany is now known as ee_drappt
[17:32:03] <kcbabo> tfennelly:  ok, on second thought, might as well just fix it right the first time
[17:32:22] <kcbabo> tfennelly:  so I'm gonna send you the module definition for now, then talk to Magesh about switching to DOMSource
[17:32:32] <kcbabo> tfennelly:  both approaches should be equivalent from your POV
[17:59:21] *** aamonten has joined #switchyard
[18:24:37] <tfennelly> kcbabo: okidoki... back... got the module.xml... thanks
[18:25:20] <kcbabo> tfennelly:  great ? I'm working on the JIRA now
[18:25:31] <kcbabo> tfennelly:  did you add the assert logic to the SOAP component?
[18:25:41] <kcbabo> tfennelly:  trying to get a feel for how important that is :-)
[18:26:24] <tfennelly> kcbabo: sorry, not sure what you mean... did I add the assert logic?
[18:26:52] <kcbabo> tfennelly:  sorry, the SOAP component has a method called assertComposedMessageOK()
[18:27:38] <tfennelly> kcbabo: lemme check that
[18:29:39] <tfennelly> kcbabo: yeah... I think I added that method
[18:29:58] <tfennelly> kcbabo: but it would be from before we had the transformation built into the Message
[18:30:11] <kcbabo> tfennelly:  right, so that's what I was trying to figure out
[18:30:21] <kcbabo> tfennelly:  when that logic came in and whether it's relevant any longer
[18:30:29] <kcbabo> tfennelly:  what do you think?
[18:30:32] <tfennelly> kcbabo: so I don't think the first part is valid anymore... checking for a Node payload
[18:30:33] *** bobmcw_ is now known as bobmcw
[18:30:34] *** bobmcw has joined #switchyard
[18:30:58] <tfennelly> kcbabo: yeah...
[18:31:10] <tfennelly> kcbabo: I don't think that Node type check is valid anyway
[18:32:15] <tfennelly> kcbabo: looks like the rest is OK and should be irrelevant to this issue anyway
[18:33:22] *** bfitzpat has quit IRC
[18:33:35] <kcbabo> tfennelly:  yeah, it's just that it's gonna require a node or conversion to a node for the rest of that logic to work
[18:33:40] *** bfitzpat has joined #switchyard
[18:34:13] <tfennelly> kcbabo: ah of course
[18:34:16] <tfennelly> kcbabo: hmmm
[18:35:04] <kcbabo> tfennelly:  don't sweat it for now
[18:35:19] <kcbabo> tfennelly:  DOMSource can easily retrieve a Node, so we'll use that for the time being
[18:35:46] <tfennelly> kcbabo: exactly
[18:35:46] <kcbabo> tfennelly:  the validation logic can be evaluated as a separate review to set the bounds for what you can and cannot do in a composer
[18:35:53] <tfennelly> kcbabo: was just looking at that
[18:38:07] <kcbabo> tfennelly:  btw, the test app for SWITCHYARD-298 is not mine
[18:38:13] <kcbabo> tfennelly:  jiri wrote that one
[18:38:40] <kcbabo> tfennelly:  so don't feel like you have to include the whole thing as part of the test
[18:38:43] <tfennelly> kcbabo: ah OK.... oops
[18:39:12] <tfennelly> kcbabo: well I might skip that test for now because of the scanner issue... I'll sort it out later
[18:39:16] <kcbabo> tfennelly:  there are some strange elements to that app, so you could even just break the first part of it out (SOAP->Bean) and use that
[18:39:20] <kcbabo> tfennelly:  ah right, that works for me
[18:39:49] <kcbabo> tfennelly:  not sure what the deal was with the scanner
[18:39:58] <kcbabo> tfennelly:  works for classes in the test folder in other projects
[18:40:05] <tfennelly> kcbabo: I think it's a maven issue
[18:40:08] <kcbabo> tfennelly:  component-bean as an example
[18:40:20] <tfennelly> kcbabo: yeah... not sure
[18:40:45] <tfennelly> kcbabo: I stepped into it though and the scanner is getting an empty list of urls to scan
[18:40:52] <tfennelly> kcbabo: so is finding nothing
[18:43:07] <kcbabo> tfennelly:  hrmm
[18:48:14] <kcbabo> back in a bit
[18:48:16] *** kcbabo is now known as babo_afk
[18:50:40] <antollinim> tfenelly: there is a hint in the scanner that says: If running this test inside your IDE... you need to set the cwd to be the  root of the bean module... (I guess you are aware of that since you are the author of the scanner ;)) Just sharing that in case this can help
[18:58:56] *** rcernich_away is now known as rcernich
[19:00:06] *** ee_drappt is now known as errantepiphany
[19:42:13] <tfennelly> antollinim: hey Alejandro
[19:42:29] <tfennelly> antollinim: no that was running it in maven directly... not in the IDE
[19:42:37] <tfennelly> antollinim: thanks for the tip though
[19:43:00] <antollinim> tfennelly: sure, I pointed that just in case
[19:43:31] <tfennelly> antollinim: yep... no harm in double checking... thanks for that !!
[19:44:40] <antollinim> tfenelly: I was working with the scanner yesterday and something called my attention
[19:45:14] <antollinim> tfenelly: if a test had a path like this "./target/test-classes" it was working ok
[19:45:18] <tfennelly> antollinim: thanks man
[19:45:31] <antollinim> tfenny: if I got deeper down the path it did not work
[19:45:32] <tfennelly> antollinim: right
[19:48:00] *** aslak has quit IRC
[19:51:08] <tfennelly> babo_afk: I issued pull requests for that stuff (SWITCHYARD-318).... hope it covers it... have to run now anyway
[19:51:15] *** tfennelly has quit IRC
[19:55:23] <babo_afk> nick kcbabo
[20:45:10] <babo_afk> hehe ? guess I missed the slash there
[20:45:16] *** babo_afk is now known as kcbabo
[21:32:28] <tcunning> kcbabo errantepiphany rcernich - have you guys ever seen a situation where you ask ivy for a JAR and it gives you a .pom instead?
[21:32:40] <errantepiphany> tcunning: nope. :(
[21:32:47] <kcbabo> tcunning:  I no know ivy
[21:32:51] <rcernich> tcunning: sorry
[22:11:29] *** bfitzpat is now known as bfitzpat_afk
[22:32:46] *** kcbabo has quit IRC
[22:35:11] *** aslak has joined #switchyard
[22:41:46] *** bfitzpat_afk is now known as bfitzpat
[22:45:20] *** magesh has joined #switchyard
[22:53:22] *** ldimaggi has quit IRC
[23:02:21] *** antollinim has quit IRC
[23:13:30] *** magesh has quit IRC
[23:42:41] *** babo has joined #switchyard





top