ruote tmp/log_2013-03-19.html

2013-03-19 19:59:52 utc ludovic_ hello there!
2013-03-19 20:01:15 utc ludovic_ has anyone been using ruote-amqp >= 2.3.0 ?
2013-03-19 20:05:43 utc ludovic_ I've just made my first contribution on ruote! I've patched the doc :).
2013-03-19 21:21:34 utc jmettraux ludovic_: hello, thanks for the pulling request, getting it in
2013-03-19 21:44:17 utc ludovic_ my pleasure :)
2013-03-19 21:44:52 utc jmettraux I'm replying to your email, stay tuned
2013-03-19 21:45:10 utc ludovic_ oh ok thanks
2013-03-19 21:45:45 utc ludovic_ btw do you have any idea about what's wrong with this code: https://gist.github.com/lgalabru/5200349
2013-03-19 21:45:55 utc jmettraux after the email
2013-03-19 21:46:12 utc ludovic_ ok
2013-03-19 21:48:41 utc jmettraux ludovic_: your code looks fine
2013-03-19 21:48:58 utc ludovic_ thanks, but it doesn't work :(
2013-03-19 21:49:29 utc jmettraux sorry, but I'm a programmer, not a medium, what does "it does not work" mean?
2013-03-19 21:49:36 utc ludovic_ lol
2013-03-19 21:49:38 utc jmettraux what are the symptoms?
2013-03-19 21:49:48 utc ludovic_ my process is timing out
2013-03-19 21:49:57 utc jmettraux ok
2013-03-19 21:50:46 utc ludovic_ whereas with this gist: https://gist.github.com/lgalabru/5200379, it works
2013-03-19 21:51:27 utc ludovic_ (almost the same, I'm just using AlertParticipant instead of Participant)
2013-03-19 21:51:50 utc jmettraux seems your participant does nothing (not even printing) with the workitem
2013-03-19 21:52:46 utc jmettraux email answered
2013-03-19 21:53:04 utc jmettraux ok, looking at your first gist
2013-03-19 21:54:25 utc jmettraux so you're starting a process that is looking at the "work1" queue
2013-03-19 21:54:32 utc jmettraux and it times out after a while
2013-03-19 21:54:47 utc jmettraux the default timeout is, iirc, 60 seconds,
2013-03-19 21:55:08 utc jmettraux the likely explanation is that your are not placing a valid workitem on the work1 queue within those 60 seconds
2013-03-19 21:55:51 utc jmettraux the script waits for the ruote process to hit an error or terminate, but nothing happens and it times out
2013-03-19 21:57:32 utc ludovic_ thanks i'm investigating
2013-03-19 21:57:55 utc jmettraux the alert participant is very dumb, it'll just pick the amqp message and place it in the workitem
2013-03-19 21:58:35 utc jmettraux https://github.com/kennethkalmer/ruote-amqp/blob/master/lib/ruote/amqp/alert_participant.rb#L128-L135
2013-03-19 22:04:32 utc ludovic_ so I don't need to subclass it (at least to do my tests) ?
2013-03-19 22:04:46 utc jmettraux no
2013-03-19 22:06:54 utc jmettraux and emitted another response email where I try to clarify why I brought ruote-amqp to the table (reminder: as a model)
2013-03-19 22:07:38 utc jmettraux for the irc archive, link to the ongoing mail thread: https://groups.google.com/forum/?fromgroups=#!topic/openwferu-users/ooRSpBivW14
2013-03-19 22:08:27 utc ludovic_ ok so I just don't understand, might be stupid :). why is this one (https://gist.github.com/lgalabru/5200379) working and this one (https://gist.github.com/lgalabru/5200349) timing out
2013-03-19 22:09:06 utc ludovic_ I'm flushing my redis storage between the 2 tests so there's no configuration left
2013-03-19 22:09:40 utc jmettraux the "working" example, just places a message on the amqp queue and is immediately done
2013-03-19 22:10:23 utc jmettraux the script is done, "ite, missa est"
2013-03-19 22:10:26 utc ludovic_ well my amqp participant also responds
2013-03-19 22:10:40 utc ludovic_ not the case in the second case
2013-03-19 22:11:05 utc jmettraux as said above, in your not working case, your waiting for a message that seems to never come
2013-03-19 22:11:40 utc jmettraux you're waiting
2013-03-19 22:13:10 utc jmettraux the Ruote::Amqp::Participant pushes message to an amqp exchange
2013-03-19 22:13:29 utc jmettraux while the Ruote::Amqp::AlertParticipant subscribes to a queue for incoming message
2013-03-19 22:13:33 utc jmettraux messages
2013-03-19 22:14:37 utc jmettraux if you place no message on the queue the AlertParticipant is subscribed to, it will never reply back to the ruote engine (unless you set a timeout)
2013-03-19 22:15:01 utc jmettraux in the case of your script, the symptom will be the wait_for() timing out
2013-03-19 22:15:25 utc ludovic_ ok understood
2013-03-19 22:17:32 utc ludovic_ ok got it
2013-03-19 22:17:52 utc ludovic_ it has been 2 long days on ruote
2013-03-19 22:17:54 utc ludovic_ :)
2013-03-19 22:17:56 utc ludovic_ sorry
2013-03-19 22:18:27 utc jmettraux no worries
2013-03-19 22:22:19 utc ludovic_ cool everything's starting to work, thank you John
2013-03-19 22:23:08 utc jmettraux you're welcome, now I can really start to worry
2013-03-19 22:23:22 utc ludovic_ bug report: the ruote flou in route web interface doesn't draw my full pdef
2013-03-19 22:23:30 utc ludovic_ ruote fluo**
2013-03-19 22:23:34 utc jmettraux ok
2013-03-19 22:23:51 utc ludovic_ I'll look why later
2013-03-19 22:24:07 utc ludovic_ a bientôt :)
2013-03-19 22:24:15 utc jmettraux https://github.com/jmettraux/ruote-fluo/issues
2013-03-19 22:24:18 utc jmettraux a bientot!
2013-03-19 22:47:35 utc Dirklectisch jmettraux: Hi John! It's been a while. How's life on the other side of the globe? :-)
2013-03-19 23:42:28 utc jmettraux Dirklectisch: hello, a while indeed! Doing fine, how about you?
2013-03-19 23:44:01 utc jmettraux sorry, was having breakfast
2013-03-19 23:44:04 utc Dirklectisch jmettraux: Everything is great actually! Doing some midnight hacking on a Textmate Clojure bundle.
2013-03-19 23:44:28 utc Dirklectisch No problem at all :-)
2013-03-19 23:44:41 utc jmettraux :-) good to read!
2013-03-19 23:45:01 utc jmettraux is Clojure good fun?
2013-03-19 23:45:02 utc Dirklectisch How is Ruote been progressing lately?
2013-03-19 23:45:40 utc Dirklectisch yep I love Clojure. Def. my favorite programming thing at the moment.
2013-03-19 23:46:28 utc jmettraux a bit stalled, progressing slowly, I'm a few bug reports behind
2013-03-19 23:47:18 utc Dirklectisch Have you been able to apply it in your job already?
2013-03-19 23:47:44 utc jmettraux well, I got a job applying it
2013-03-19 23:47:48 utc Dirklectisch If I remember correctly you hoped you would be able to in your (back then) new job right?
2013-03-19 23:48:02 utc jmettraux but my latest job is VB dev
2013-03-19 23:48:31 utc Dirklectisch wow really, what are you working on?
2013-03-19 23:48:54 utc Dirklectisch it's been ages since I touched VB.
2013-03-19 23:49:02 utc jmettraux but it's with my brother as the boss and I have equity, so family business and good everything
2013-03-19 23:49:20 utc Dirklectisch sounds good
2013-03-19 23:49:28 utc jmettraux doing some .net mvc, webapps for insurance brokers and the like
2013-03-19 23:49:29 utc Dirklectisch what kind of products are you making in VB?
2013-03-19 23:49:34 utc Dirklectisch ah ok
2013-03-19 23:49:51 utc Dirklectisch cool :-)
2013-03-19 23:50:24 utc Dirklectisch I'm still working freelance, trying to convince one of my customers to start using Ruote.
2013-03-19 23:50:45 utc jmettraux you should write a Clojure equivalent ;-) with only the features you need
2013-03-19 23:51:27 utc Dirklectisch They have a Rails app that is for the most part concerned with workflowy stuff.
2013-03-19 23:51:29 utc jmettraux maybe using Jbpm or Activiti from Clojure is fun
2013-03-19 23:52:03 utc Dirklectisch hmm yeah I could do that, if a suitable project comes along.
2013-03-19 23:52:28 utc jmettraux maybe a simple state machine stashed on a business object is sufficient
2013-03-19 23:53:30 utc jmettraux sorry, love to play "against ruote"
2013-03-19 23:53:46 utc Dirklectisch haha I know
2013-03-19 23:54:06 utc Dirklectisch but hey there are some use cases for it, must be ;-)
2013-03-19 23:55:08 utc Dirklectisch I think this particular app could benefit from something more than statemachines on business objects.
2013-03-19 23:55:23 utc jmettraux ok
2013-03-19 23:56:54 utc Dirklectisch What about expressing workflows in a logic like language? Might be fun to dabble with Clojure's core.logic.
2013-03-19 23:57:46 utc Dirklectisch I think I read a paper on something similar a while back.
2013-03-19 23:57:50 utc jmettraux that'd be "rules" ?
2013-03-19 23:58:16 utc Dirklectisch yeah Prolog style.
2013-03-19 23:58:53 utc jmettraux :-) I'd love to have the time to tinker with Rete on Ruby, others already did
2013-03-19 23:59:08 utc jmettraux in Clojure, that must be super easy
2013-03-19 23:59:38 utc jmettraux the Jbpm-Drools guys have an optimized Rete implementation
2013-03-20 00:00:06 utc jmettraux iirc, the latest Jbpm is based on Drools (the Drools guy took over Jbpm it seems)
2013-03-20 00:00:16 utc jmettraux but you should just hack something
2013-03-20 00:02:07 utc jmettraux how is the Clojure moodscape? Are people welcoming overlapping innovation?
2013-03-20 00:04:43 utc Dirklectisch Well considering it is such a young language, people are still working on implementing the basic libs.
2013-03-20 00:05:27 utc Dirklectisch That being said, I feel there is definitely a very innovative moodscape to pretty much everything.
2013-03-20 00:06:37 utc Dirklectisch Pretty much everything that comes out is really a step forward from what has been done before.
2013-03-20 00:08:00 utc Dirklectisch Take for example something as simple as the Clojure REPL. They build it is such a way that the client connects over sockets to a server REPL. You can then plug in all sorts of middleware.
2013-03-20 00:09:15 utc Dirklectisch This enable all sorts of cool applications without much hassle, wanna build an integration with whatever you're environment you use? No problem, just connect, plug in some features.
2013-03-20 00:09:52 utc Dirklectisch Wanna do some remote administration? No problem, just connect to the REPL of you production environment.
2013-03-20 00:09:52 utc jmettraux sounds like an OS
2013-03-20 00:10:16 utc jmettraux looks a pleasure to work with
2013-03-20 00:10:42 utc Dirklectisch Yeah, this is just one small example. I feel like they are re-thinking lots of things.
2013-03-20 00:11:06 utc jmettraux have to restart for security update, I'll be back soon
2013-03-20 00:11:19 utc Dirklectisch so yeah to cut this very long answer short, definitely an innovative mindscape.
2013-03-20 00:11:25 utc Dirklectisch (y)
2013-03-20 00:11:34 utc jmettraux :-)