[00:32:31] *** aslak has quit IRC
[01:20:34] *** bfitzpat has quit IRC
[02:24:18] *** ldimaggi has joined #switchyard
[03:22:55] *** lanceball has joined #switchyard
[05:12:29] *** magesh has joined #switchyard
[05:17:27] *** tcunning has quit IRC
[05:35:30] *** ldimaggi has quit IRC
[05:38:57] *** magesh has left #switchyard
[06:04:08] *** dbevenius has quit IRC
[06:07:55] *** lanceball has quit IRC
[08:43:11] *** aslak has joined #switchyard
[08:47:39] *** dbevenius has joined #switchyard
[09:31:34] *** tfennelly has quit IRC
[09:43:35] *** aslak has quit IRC
[09:43:53] *** aslak has joined #switchyard
[09:57:23] *** tfennelly has joined #switchyard
[13:06:35] *** ldimaggi has joined #switchyard
[13:14:54] *** aslak has quit IRC
[13:15:38] *** aslak has joined #switchyard
[13:37:12] *** dbevenius has quit IRC
[13:45:53] *** tcunning has joined #switchyard
[14:18:25] *** tcunning has quit IRC
[14:42:07] *** lanceball has joined #switchyard
[15:01:06] *** dbevenius has joined #switchyard
[15:21:48] *** tcunning has joined #switchyard
[15:41:33] *** bfitzpat has joined #switchyard
[15:52:52] *** antollinim has joined #switchyard
[15:53:13] <tfennelly> tcunning: hey T... I have a few pull requests there
[15:53:19] <tfennelly> tcunning: sorry for being a pita
[15:53:30] <tcunning> tfennelly: i'll take them in a bit
[15:54:10] <tfennelly> tcunning: thanks
[15:56:01] <tfennelly> tcunning: there's 4 of them :( https://github.com/organizations/jboss-switchyard/dashboard/pulls
[16:01:33] <tcunning> tfennelly: wow.   i'm gonna have to queue up the ESB questions for you today! :)
[16:01:49] <tfennelly> tcunning: haha ;)
[16:01:59] <tfennelly> tcunning: bribery
[16:02:17] <tfennelly> tcunning: you're not a FIFA executive are you?
[16:02:27] <tfennelly> tcunning: hehehe
[16:03:51] <tcunning> tfennelly: World Cup Qatar makes total sense right?    No corruption there!
[16:04:26] <tfennelly> tcunning: for sure... totally above board... nobodies ass was greased up for that one
[16:24:15] *** tfennelly has quit IRC
[16:42:05] *** errantepiphany has joined #switchyard
[17:34:01] *** tfennelly has joined #switchyard
[17:50:35] *** GitHub113 has joined #switchyard
[17:50:35] <GitHub113> core: master Tom Fennelly * f0df6c3 (1 files in 1 dirs): SWITCHYARD-259: jboss-as6 deployer module has a dependency on a broken pom - http://bit.ly/lYB7wO
[17:50:35] *** GitHub113 has left #switchyard
[17:51:54] <tcunning> tfennelly: fix version for these is all 0.1, right?
[17:52:49] <tfennelly> tcunning: yeah... did I forget to set it on them?
[17:53:12] <tcunning> tfennelly: i got it, no big deal, just wanted to make sure
[17:54:14] <tfennelly> tcunning: which one did I miss?
[17:54:40] <tcunning> tfennelly: it was 259, but i set it when resolving it
[17:55:07] <tfennelly> tcunning: ah cool... thanks and sorry
[18:07:21] <errantepiphany> tfennelly: ping
[18:07:37] <tfennelly> errantepiphany: hey
[18:10:05] <errantepiphany> tfennelly: I have 2 cases: 1) switchyard managing a jbpm5 process (done), and 2) a jbpm5 "action" invoking a switchyard service (one hiccup). Now, they are running in the same VM, so I'd like to be able to do some sort of INVM invocation, but I don't see how I can properly get my hands on the proper domain, so that I can get a service reference, and thus create an exchange.
[18:10:43] <errantepiphany> tfennelly: I can see how I can CREATE a domain, but I don't wanna do that...
[18:11:26] <tfennelly> errantepiphany: how are the jbpm components bundled?
[18:11:52] <tfennelly> errantepiphany: the ServiceDomainManager is an MC Bean
[18:12:10] <tfennelly> errantepiphany: if you could get a ref to that you should be able to invoke any service in the container
[18:12:18] <errantepiphany> tfennelly: Well there's an activator where I get a serviceRef, but that's to ME (the jbpm5 service), not necessarily the one I wanna invoke.
[18:12:27] <tfennelly> errantepiphany: so if you had some sort of deployer
[18:13:04] <tfennelly> errantepiphany: do you have the code anywhere?
[18:13:11] <tfennelly> errantepiphany: that I can see
[18:13:17] <errantepiphany> tfennelly: yeah one sec
[18:15:14] <errantepiphany> tfennelly: http://goo.gl/0kXRB
[18:15:37] <errantepiphany> tfennelly: specifically http://goo.gl/s14DO
[18:15:52] <errantepiphany> tfennelly: the executeWorkItem is WRONG.
[18:16:26] <errantepiphany> tfennelly: since it uses the _serviceRef of the bpm component deployment, which is ITSELF. I need to pull out the parameters that say which domain, service qname, etc.
[18:16:42] <errantepiphany> tfennelly: which I CAN DO, but how I can look up the proper domain is where I'm stuck.
[18:17:34] <tfennelly> errantepiphany: you can get messages in, but not send them out
[18:18:08] <errantepiphany> tfennelly: I can process messages (get them "in"), yes, via: http://goo.gl/qQOBA
[18:18:24] <errantepiphany> tfennelly: I'm trying to send them out here: http://goo.gl/s14DO
[18:18:55] <tfennelly> errantepiphany: serviceRef is the target Service?
[18:20:17] *** GitHub183 has joined #switchyard
[18:20:17] <GitHub183> components: master Tom Fennelly * 124108a (6 files in 6 dirs): SWITCHYARD-260: pom parent path fixes - http://bit.ly/kYg9YU
[18:20:17] *** GitHub183 has left #switchyard
[18:20:33] <errantepiphany> tfennelly: the serviceRef in the constructor of http://goo.gl/s14DO is WRONG. It's actually the service of the BPM component impl itself. So that constructor goes away. I just need a way to lookup a DIFFERENT service inside executeWorkItem. I can get the proper parameters to tell me which one, and even which domain - by name, but how do I get a handle on that domain is the question.
[18:21:22] <tfennelly> errantepiphany: ok... so you don't know the target service ahead of time.. only when processing the message?
[18:22:16] <tfennelly> errantepiphany: what creates ServiceWorkItemHandler?
[18:22:29] <errantepiphany> tfennelly: unfortunately, the API of WorkItemHandler (the jbpm5/drools interface) only gives me the parameters in the executeWorkItem method. I can cache it after the first time, though.
[18:22:53] <tcunning> tfennelly: your spammer is back asking about groovy imports - did he ask the same on the smooks list?
[18:23:00] <tfennelly> errantepiphany: right... but you don't know before that hmmm
[18:23:07] <errantepiphany> tfennelly: the exchangehandler creates them in the start() method here: http://goo.gl/qQOBA
[18:23:19] <tfennelly> tcunning: lol... yeah.. and on the Freemarker list
[18:23:37] <errantepiphany> tfennelly: s/creates them/creates the ServiceWorkItemHandler/
[18:24:09] <errantepiphany> tfennelly: but see the serviceRef in the startMethod of the handler is ME, not whom I want to invoke.
[18:24:20] <tfennelly> errantepiphany: I should get the code and have a look dave... versus asking you stupid questions
[18:24:39] <errantepiphany> tfennelly: really you don't need the code. the problem can be boiled down to this:
[18:25:13] <tfennelly> errantepiphany: well I can understand the problem for sure... but I need to seee the structire of your code in orde to come up with a suggestion for you
[18:25:20] <errantepiphany> tfennelly: If I have code running inside the same VM as switchyard, and I know the NAME of the domain and NAME of the Service I want to invoke, how can I get a handle on a ServiceReference?
[18:25:34] <tfennelly> errantepiphany: yeah... I get all that
[18:26:03] <tfennelly> errantepiphany: now we need to figure out how to get the ServiceDomainManager to where you need it
[18:26:33] <tfennelly> errantepiphany: it's a container wide object
[18:26:43] <errantepiphany> tfennelly: You know what would be nice? If we exposed the Domain as a getter on ServiceReference, then the DomainManager as a getter of the Domain.
[18:27:04] <tfennelly> errantepiphany: yeah... that'll never fly with Keith
[18:27:09] <errantepiphany> tfennelly: why?
[18:27:53] <tfennelly> errantepiphany: I just know it won't... exposing too much
[18:28:14] <errantepiphany> errantepiphany: what if we passed the Domain or DomainManager into the init() method of the ExchangeHandler?
[18:28:14] <tfennelly> errantepiphany: let me get the code and have a guuu
[18:28:18] <errantepiphany> tfennelly: okay.
[18:28:34] <errantepiphany> sorry - I mean Activator
[18:28:37] <errantepiphany> tfennelly: ^^
[18:29:05] <errantepiphany> tfennelly: public ExchangeHandler init(QName, Model, ServiceDomainManager) - in ExchangeHandler interface
[18:29:09] <errantepiphany> grrrr
[18:29:14] <errantepiphany> tfennelly: I mean Activator interface
[18:30:00] <tfennelly> errantepiphany: if it were me... I'd supply the ServiceDomain to all "internal" components (spi level components)
[18:30:09] <errantepiphany> tfennelly: +100
[18:30:15] <tfennelly> errantepiphany: I'm just fairly sure Keith would not go for that
[18:30:31] <errantepiphany> tfennelly: If it's SPI and not API, I don't see the problem...
[18:31:00] <tfennelly> errantepiphany: right.. but we're not making the decision really :)
[18:31:15] <tfennelly> errantepiphany: let me have a quick look Dave
[18:31:22] <tfennelly> errantepiphany: there might be a way around it
[18:31:24] <errantepiphany> tfennelly: okay, sorry and thanks.
[18:31:53] <tfennelly> errantepiphany: no worries man... just gimme a while... have to run home now and will look at it then... won't take me long
[18:32:35] *** tfennelly has quit IRC
[18:46:32] *** ldimaggi has quit IRC
[18:52:18] *** GitHub161 has joined #switchyard
[18:52:18] <GitHub161> quickstarts: master Tom Fennelly * cd661fd (5 files in 5 dirs): SWITCHYARD-260: pom parent path fixes - http://bit.ly/isFPNm
[18:52:18] *** GitHub161 has left #switchyard
[18:53:48] <tcunning> errantepiphany: i'm gonna run out and grab lunch, back in a half an hour
[18:57:10] <errantepiphany> tcunning: okee-dokee
[19:01:59] *** tfennelly has joined #switchyard
[19:02:00] *** tfennelly has joined #switchyard
[19:03:19] *** tfennell_ has joined #switchyard
[19:06:08] *** tfennelly has quit IRC
[19:21:16] <errantepiphany> tfennell_: FYI I just did a rebase and push -f to my origin of SWITCHYARD-174, just in case you pulled it down already.
[19:21:33] <errantepiphany> (in components)
[19:22:01] <tfennell_> errantepiphany: I got it about 5 mins ago.... should I pull again?
[19:22:13] <errantepiphany> tfennell_: Not really that necessary.
[19:22:21] <errantepiphany> tfennell_: for what you're looking at
[19:25:10] <tfennell_> errantepiphany: I don't think there's a way around this without injecting the ServiceDomain into the activator
[19:26:00] <errantepiphany> errantepiphany: yeah that's what I was thinking
[19:26:02] <tfennell_> errantepiphany: for now I'd implement a local hack so as to let Keith see what you're trying to do
[19:26:13] <tfennell_> errantepiphany: if this is for 0.2
[19:26:41] <tfennell_> errantepiphany: add a setDomainManager method to the Activator interface
[19:28:29] <errantepiphany> tfennell_: why not just pass it into the init(QName,Model) method of Activator, so it becomes init(QName,Model,ServiceDomainManager)?
[19:29:23] <tfennell_> errantepiphany: because that's called multiple times in the life of the Activator
[19:29:33] <tfennell_> errantepiphany: you just want it to be set once
[19:29:59] <errantepiphany> tfennell_: why is init called multiple times?
[19:30:22] <tfennell_> errantepiphany: so where the activators are created in the Deployment... iterate over it and set the domain manager
[19:30:31] <tfennell_> errantepiphany: once for each service
[19:30:57] <errantepiphany> tfennell_: So do it in Deployment.createActivators()?
[19:32:29] <tfennell_> errantepiphany: right
[19:32:30] <errantepiphany> tfennell_: I wonder if I should just create a sub-interface of Activator, called DomainAwareActivator that adds that method, and only if an Activator is an instanceof DomainAwareActivator do I cast and call it.
[19:32:47] <tfennell_> errantepiphany: nah... think of this as a temp thing
[19:33:04] <tfennell_> errantepiphany: just to allow Keith see it before it becomes permanent
[19:33:11] <errantepiphany> tfennell_: okee-dokee. thanks.
[19:33:26] <tfennell_> errantepiphany: we can change it later once he agrees/disagrees to it
[19:33:40] <tfennell_> errantepiphany: so just something simple for now
[19:33:49] <errantepiphany> tfennell_: and if he disagrees, he's gonna have to give me another option
[19:34:29] <tfennell_> errantepiphany: exactly :)
[19:34:38] <tfennell_> errantepiphany: gota run here
[19:37:14] <errantepiphany> tfennell_: ok, thanks again.
[20:50:27] *** blixen has joined #switchyard
[20:51:19] *** blixen has left #switchyard
[21:11:48] *** blixen has joined #switchyard
[21:12:28] *** blixen has left #switchyard
[21:25:21] *** dbevenius has quit IRC
[21:29:00] *** stfkbf has joined #switchyard
[22:06:35] <tcunning> tfennell_: i just saw your tweet - i still have time to get SWITCHYARD-78 in tonight, right?
[22:10:49] *** lanceball has quit IRC
[22:13:19] <tfennell_> tcunning: sure thing
[22:14:09] <tfennell_> tcunning: I built and deployed it to maven from EC2
[22:14:44] <tfennell_> tcunning: it uploaded the 187Mb of the AS6 distro in about 15 seconds lol
[22:14:58] <tcunning> nice!
[22:15:09] <tfennell_> tcunning: must be some pipe
[22:17:57] <tfennell_> tcunning: hey T... what's the status of the Arquillian testing you looked at some time ago?
[22:19:10] <tcunning> as far as i know, i'm still blocked by the AS6 deployer from doing embedded testing.
[22:20:19] <tfennell_> tcunning: okidoki... I'll look at creating an automated test or two the old fashioned way then
[22:20:31] <tfennell_> tcunning: thanks
[22:21:04] <tcunning> tfennell_: i think one of quickstart branches has it doing testing on a local AS6
[22:21:13] <tcunning> oops -- one of my quickstart branches
[22:40:32] *** lanceball has joined #switchyard
[22:40:34] <tfennell_> tcunning: cool... if you can locate the branch maybe send me the name
[22:44:39] <tcunning> tfennell_: i'll look around for it tonight - i can help you with that tmmrw
[22:44:56] <tfennell_> tcunning: thanks
[22:45:16] <tcunning> i'm trying to finish off 78 and i need to get something done for kurt
[22:45:25] <tfennell_> tcunning: well hopefully the release distros will make it easier
[22:45:54] <tfennell_> tcunning: cool... don't worry if you can't to it
[22:46:07] <tcunning> i'll ping you first thing tmmrw
[22:55:47] *** antollinim has quit IRC
[23:05:35] <tcunning> tfennell_: so i'm replacing all the runtimeexceptions  - can i just assume that if we are catching a runtimeexception, we mean to be catching this new SwitchYardException i'm replacing them with?
[23:06:39] <tfennell_> tcunning: hmmm... I think maybe leave the catches within our code for now?
[23:06:48] <tfennell_> tcunning: they won't effect anyone outside
[23:06:59] <tfennell_> tcunning: yeah?
[23:07:36] <tfennell_> tcunning: I think as long as we're throwing a typed (even though it's unchecked) exception, we're moving in the right direction
[23:07:47] <tcunning> tfennell_: sounds good to me.     i'm going to be extending runtimeexception anyway, so they'll still be caught
[23:07:48] <tfennell_> tcunning: outside code can catch that
[23:08:00] <tfennell_> tcunning: exactly
[23:08:23] <tfennell_> tcunning: we might need to analyse the catches in more detail
[23:09:53] <tfennell_> I'm off guys
[23:10:00] <tcunning> tfennell_: have a good one
[23:10:05] <tfennell_> tcunning, errantepiphany: chat tomorrow
[23:10:24] *** tfennell_ has quit IRC
[23:39:15] *** errantepiphany has left #switchyard
[23:54:04] *** stfkbf has left #switchyard





top