ruote log_2010-10-21

2010-10-21 00:13:37 utc hassox that's not too huge
2010-10-21 00:13:46 utc hassox I'm sure I can work with that if it turns out to be an issue
2010-10-21 00:13:57 utc hassox atm it's not ;)
2010-10-21 00:14:30 utc hassox are you familiar with ruote-kit
2010-10-21 00:14:30 utc jmettraux it's ruby code, not engraved in stone
2010-10-21 00:14:30 utc hassox ?
2010-10-21 00:14:35 utc jmettraux a little bit yes
2010-10-21 00:15:11 utc hassox can I run multiple instances of that in my rack graph, and set each one to a different engine?
2010-10-21 00:15:30 utc hassox hrm.. it looks like no
2010-10-21 00:15:44 utc hassox s'ok... when I get to it, I can add that feature
2010-10-21 00:15:45 utc jmettraux RuoteKit.engine soloton
2010-10-21 00:15:51 utc jmettraux ok
2010-10-21 00:15:58 utc hassox dude
2010-10-21 00:16:02 utc hassox thanx heaps for all your help
2010-10-21 00:16:04 utc hassox I
2010-10-21 00:16:10 utc hassox I'm feeling so much better about this
2010-10-21 00:16:14 utc jmettraux maybe one engine will suffice
2010-10-21 00:16:24 utc jmettraux have fun with it, break it
2010-10-21 00:16:41 utc hassox :D
2010-10-21 00:17:03 utc hassox better go hang with the kids for a while
2010-10-21 00:17:06 utc hassox thanx dude
2010-10-21 00:17:13 utc jmettraux OK, enjoy !
2010-10-21 00:17:16 utc jmettraux you're welcome
2010-10-21 06:37:35 utc tosch_le hello jmettraux. forgot to ask: how was the oktoberfest?
2010-10-21 06:37:46 utc jmettraux hello !
2010-10-21 06:38:13 utc jmettraux was good : http://www.youtube.com/watch?v=mDgC6UTmUdE
2010-10-21 06:40:37 utc jmettraux 150 years of relation between Pruss... Germany and Japan
2010-10-21 06:41:55 utc tosch_le really looks like it was fun. but i don't like _her_: http://www.cornelia-pieper.de/
2010-10-21 06:42:45 utc tosch_le (she's from the same federal state as me, so i got to know her a bit)
2010-10-21 06:42:51 utc jmettraux her bites must be lethal
2010-10-21 06:43:00 utc tosch_le lol
2010-10-21 06:43:15 utc tosch_le anyway, thanks again for theboard.
2010-10-21 06:43:22 utc jmettraux it's not finished
2010-10-21 06:43:41 utc jmettraux and there's no design yet
2010-10-21 06:43:59 utc tosch_le do you think it's ok not including the worker by default? thinking about switching that, especially with passenger 3 out now
2010-10-21 06:44:36 utc jmettraux user lambda says : "I checked it out, but when I launch flows, nothing happens"
2010-10-21 06:45:16 utc tosch_le to true. i'll better change that, then.
2010-10-21 06:45:33 utc jmettraux should it fail at first try or later at first deploy time ?
2010-10-21 06:45:34 utc tosch_le s/to/too/
2010-10-21 06:45:49 utc tosch_le better at deploy time.
2010-10-21 06:45:52 utc jmettraux instant satisfaction, then education
2010-10-21 06:45:56 utc tosch_le :-)
2010-10-21 06:45:57 utc jmettraux maybe I'm wrong
2010-10-21 06:46:17 utc jmettraux thanks a lot for helping hassox out yesterday
2010-10-21 06:46:17 utc tosch_le no, there'll have to be some warning signs in any case
2010-10-21 06:46:23 utc tosch_le it was fun
2010-10-21 06:46:51 utc tosch_le should have pointed to the wiki page earlier, lbt helped out
2010-10-21 06:47:19 utc jmettraux (ouch, the wiki is still down)
2010-10-21 06:47:58 utc tosch_le ? http://github.com/jmettraux/ruote/wiki/Overview works for me
2010-10-21 06:48:43 utc jmettraux oh cool
2010-10-21 06:48:46 utc jmettraux it's back
2010-10-21 06:48:51 utc jmettraux great
2010-10-21 06:58:23 utc jmettraux tosch_le: the template thing is good stuff, we've got to highlight that in the doc of ruote-on-rails
2010-10-21 06:59:07 utc tosch_le oups. github is down
2010-10-21 07:00:03 utc tosch_le i thought i already highlighted that
2010-10-21 07:00:22 utc tosch_le anyway, i thought of putting that info in the readme of ruote-kit, too.
2010-10-21 07:00:43 utc jmettraux +1
2010-10-21 07:01:10 utc jmettraux or would the template belong to ruote-kit ?
2010-10-21 07:02:27 utc tosch_le wondered about that, too. both places seem appropriate to me. in the end i thought it can be left as it is. after all, it's just a file/an url which may be referenced from anywhere
2010-10-21 07:02:48 utc jmettraux yes
2010-10-21 07:03:13 utc jmettraux but I have the impression it doesn't belong to ruote-on-rails
2010-10-21 07:03:21 utc jmettraux it's a step right after ruote-kit
2010-10-21 07:03:29 utc jmettraux it's a ruote-kit integration tool
2010-10-21 07:03:40 utc jmettraux ruote-on-rails is an example product
2010-10-21 07:03:50 utc jmettraux (product not in the commercial sense)
2010-10-21 07:04:28 utc jmettraux I'd put the template in ruote-kit
2010-10-21 07:04:30 utc tosch_le ruote-on-rails was about the question how to integrate ruote-kit into rails.
2010-10-21 07:04:39 utc jmettraux true
2010-10-21 07:04:46 utc jmettraux but it's more an example
2010-10-21 07:04:53 utc jmettraux a reference
2010-10-21 07:05:00 utc tosch_le but now it'll become more an example app – so you have a strong point
2010-10-21 07:05:01 utc jmettraux the template is a tool
2010-10-21 07:05:44 utc tosch_le i just don't like putting rails things into rk when lots of people won't need them there
2010-10-21 07:06:22 utc jmettraux it's under raw/ ;-)
2010-10-21 07:06:51 utc tosch_le that's just a pointer for github to deliver the file directly and not with all the fancy stuff around
2010-10-21 07:06:56 utc jmettraux it goes nicely with the "how to put this into rails" doc of rk
2010-10-21 07:07:25 utc jmettraux if we change stuff in rk, we might have to change the template in the same go
2010-10-21 07:07:32 utc jmettraux s/go/commit/g
2010-10-21 07:07:47 utc tosch_le and the same applies to ruote-on-rails
2010-10-21 07:08:02 utc jmettraux indirectly yes
2010-10-21 07:08:20 utc tosch_le basically, there's the same code inside the template and the ruote-on-rails example app
2010-10-21 07:08:22 utc jmettraux let's sleep a night on it
2010-10-21 07:08:38 utc jmettraux yes, since ruote-on-rails might be the produce of running the template
2010-10-21 07:08:52 utc tosch_le exactly
2010-10-21 07:09:06 utc tosch_le nobody will know it was the other way round ;-)
2010-10-21 07:09:13 utc jmettraux but the template is tightly bound to rk
2010-10-21 07:09:15 utc jmettraux indeed
2010-10-21 07:09:29 utc jmettraux sorry, playing the "Alles in Ordnung" game
2010-10-21 07:09:42 utc tosch_le it's ok.
2010-10-21 07:09:57 utc jmettraux ACTION loves those kind of discussions
2010-10-21 07:10:02 utc tosch_le i just don't like to have that file in the rk gem
2010-10-21 07:10:19 utc tosch_le to, it feels like it doesn't belong there
2010-10-21 07:10:43 utc jmettraux I'd make sure it doesn't get packaged
2010-10-21 07:11:58 utc tosch_le hmm. tell jeweler to ignore it? anyway, that seems to be a way ok for me.
2010-10-21 07:14:04 utc tosch_le another point: that's a template suitable for rails 3. rails 2 would need some changes (no changes to routes.rb, include rk as rack app). i'm not entirely sure we do need a template for rails 2, but if there is a need, we'll have to think about that. but as the template is ruby code, there'll be a way
2010-10-21 07:14:49 utc tosch_le we could even extend the template so that the worker is included in development and not in production. but that is for later…
2010-10-21 07:20:05 utc jmettraux ok
2010-10-21 07:20:12 utc jmettraux I wouldn't move for rails2
2010-10-21 07:21:00 utc tosch_le ok. i'll only move when there's a request for it.
2010-10-21 07:23:23 utc jmettraux great
2010-10-21 07:23:53 utc tosch_le everything sorted out :-)
2010-10-21 07:23:59 utc jmettraux (the more I think about it, the more I think the template belongs to rk)
2010-10-21 07:25:22 utc tosch_le btw, there are some more possibilities which could be fun within the template. asking the user which storage implementation he likes to use, for instance
2010-10-21 07:26:21 utc jmettraux Kenneth wanted to have ruote-KIT, it fell short when the storage variants multiplicated
2010-10-21 07:26:30 utc jmettraux this could bring the idea back
2010-10-21 07:26:34 utc jmettraux great
2010-10-21 07:27:03 utc jmettraux when we speak of the wolf...
2010-10-21 07:27:17 utc tosch_le and i'd like to include some easy (and ex[tendable|changeable]) way to restrict access to /_ruote
2010-10-21 07:27:28 utc jmettraux kennethkalmer: coffee ?
2010-10-21 07:27:41 utc kennethkalmer definitely!
2010-10-21 07:27:50 utc kennethkalmer tosch_le: sorry for not coming back yesterday
2010-10-21 07:27:52 utc jmettraux tosch_le: OK
2010-10-21 07:28:20 utc jmettraux ACTION emits coffee towards DE and ZA
2010-10-21 07:28:27 utc tosch_le kennethkalmer: it's ok, we sorted it out
2010-10-21 07:28:39 utc tosch_le ACTION receives coffee
2010-10-21 07:28:58 utc tosch_le i know you're a very busy person :-)
2010-10-21 07:29:11 utc kennethkalmer ACTION receives coffee, transfers to take-away cup :)
2010-10-21 07:29:27 utc kennethkalmer i'll go through the logs, have to run to meet the general manager
2010-10-21 07:29:31 utc kennethkalmer thanks guys, ttyl !
2010-10-21 10:13:39 utc hassox tosch_le: hey dude
2010-10-21 10:13:40 utc hassox you here?
2010-10-21 10:14:06 utc tosch_le yes,
2010-10-21 10:14:14 utc tosch_le but just for another five minutes
2010-10-21 10:14:20 utc tosch_le lunch time after that :-)
2010-10-21 10:14:21 utc hassox I'll make it quick ;)
2010-10-21 10:14:26 utc hassox :D
2010-10-21 10:14:30 utc hassox http://ruote.rubyforge.org/participants.html
2010-10-21 10:14:40 utc hassox I'm looking there to help me actually start coding the processes
2010-10-21 10:15:02 utc tosch_le ok
2010-10-21 10:15:04 utc hassox https://gist.github.com/78c69fc51d9915cd3a19
2010-10-21 10:15:10 utc hassox so that's what I've come up with
2010-10-21 10:15:31 utc hassox I'm not sure how the 'task' option is given to the participant though
2010-10-21 10:15:34 utc hassox or what it really means
2010-10-21 10:17:51 utc hassox does that "task" just come in as a field in the workitem?
2010-10-21 10:18:11 utc tosch_le yes, there's a field called 'params'
2010-10-21 10:18:28 utc tosch_le params is a hash, 'task' will be a key in there
2010-10-21 10:18:34 utc hassox oh right
2010-10-21 10:18:38 utc hassox so that could be anything
2010-10-21 10:18:49 utc tosch_le iirc there is a helper method, you may call workitem.params
2010-10-21 10:18:52 utc hassox task is just a nice wordy way of telling it what to do
2010-10-21 10:19:03 utc hassox so you'd have to case it in the consume method?
2010-10-21 10:19:14 utc tosch_le but don't be to sure about the params method
2010-10-21 10:19:22 utc tosch_le yes
2010-10-21 10:19:32 utc hassox cools
2010-10-21 10:19:47 utc tosch_le another way would be to use different participant implementations for different tasks
2010-10-21 10:19:57 utc hassox sure
2010-10-21 10:20:10 utc hassox what's the prefered way to do it?
2010-10-21 10:20:13 utc tosch_le imho better for "automatic" participants which doesn't share to much code
2010-10-21 10:20:25 utc hassox right
2010-10-21 10:20:37 utc tosch_le you could even use subclassed participants to share code
2010-10-21 10:20:44 utc tosch_le ok, have to leave now
2010-10-21 10:20:51 utc tosch_le will be back in an hour
2010-10-21 10:22:10 utc kennethk_ howzit hassox
2010-10-21 10:22:26 utc hassox tosch_le: thanx
2010-10-21 10:22:28 utc hassox hey kennethk_
2010-10-21 10:22:29 utc hassox :D
2010-10-21 10:22:30 utc hassox how are you mate
2010-10-21 10:22:40 utc kennethk_ great thanks, and yourself ?
2010-10-21 10:22:53 utc hassox yeah really good :D
2010-10-21 10:23:05 utc kennethkalmer awesome :)
2010-10-21 10:23:18 utc kennethkalmer tosch_le made a great suggestion, one of subclassing
2010-10-21 10:23:18 utc hassox just porting over my hand rolled stuff to ruote
2010-10-21 10:23:22 utc kennethkalmer +1
2010-10-21 10:23:33 utc hassox so
2010-10-21 10:23:39 utc hassox lemme tell you how I'm doing it
2010-10-21 10:23:57 utc kennethkalmer cool, i'll whip up a gist of home I do it in as well
2010-10-21 10:24:55 utc hassox so when an account is created,
2010-10-21 10:25:06 utc hassox I want to setup some default category like stuff
2010-10-21 10:25:09 utc hassox labels and such
2010-10-21 10:25:31 utc hassox and then also, I need to create a todo list for the account, and pre-populate it with actions that they should do when the account is created
2010-10-21 10:26:05 utc hassox as they move through the system, different todo items will be genereated for the account
2010-10-21 10:26:06 utc hassox but
2010-10-21 10:26:21 utc hassox there todo lists are a bit polymorphic... they're attached to lots of stuff
2010-10-21 10:26:44 utc hassox and when things get created, they get a list, and then when specific events occur, they get items generated and appeneded to the relveant list
2010-10-21 10:26:46 utc hassox does that all make sense?
2010-10-21 10:27:13 utc kennethkalmer roughly yes
2010-10-21 10:27:16 utc hassox https://gist.github.com/78c69fc51d9915cd3a19 <--
2010-10-21 10:27:29 utc hassox or
2010-10-21 10:28:28 utc hassox https://gist.github.com/78c69fc51d9915cd3a19
2010-10-21 10:28:34 utc hassox perhaps something like that
2010-10-21 10:29:12 utc kennethkalmer it is making more sense now
2010-10-21 10:29:42 utc hassox so this would just be the new account process
2010-10-21 10:29:58 utc hassox there's all sorts of events where simliar processes need to take place
2010-10-21 10:30:04 utc hassox create a todo list, attach items
2010-10-21 10:31:40 utc kennethkalmer perfect fit :)
2010-10-21 10:31:54 utc kennethkalmer http://gist.github.com/638269
2010-10-21 10:32:08 utc kennethkalmer how I handle delegation to local participants
2010-10-21 10:32:15 utc kennethkalmer through subclassing
2010-10-21 10:32:15 utc hassox ohhh I like
2010-10-21 10:32:32 utc hassox that's tidy
2010-10-21 10:32:35 utc kennethkalmer daemon-kit also uses the :command field when having remote amqp participants
2010-10-21 10:32:48 utc kennethkalmer and each participant is isolated and focussed
2010-10-21 10:32:53 utc kennethkalmer let me expand quickly
2010-10-21 10:33:19 utc hassox ACTION <3's daemon-kit btw
2010-10-21 10:33:38 utc hassox ACTION installed an xmpp client on our office mac mini, that reads out the messages sent to it via 'say' 
2010-10-21 10:33:45 utc hassox the manager I don't think likes it )
2010-10-21 10:33:46 utc hassox ;)
2010-10-21 10:37:03 utc kennethkalmer :)
2010-10-21 10:37:12 utc kennethkalmer http://gist.github.com/638269
2010-10-21 10:37:21 utc kennethkalmer righto, that is something very close to what I currently do
2010-10-21 10:37:26 utc kennethkalmer works very neat
2010-10-21 10:37:56 utc kennethkalmer s/fail/failed/g
2010-10-21 10:38:40 utc hassox I love it
2010-10-21 10:38:41 utc codebeaker ruote still bends my brain :)
2010-10-21 10:38:45 utc hassox that's awesome
2010-10-21 10:38:49 utc kennethkalmer in a good way at least :)
2010-10-21 10:39:02 utc kennethkalmer it took me a while as well, and then the proverbial penny droped
2010-10-21 10:39:07 utc kennethkalmer ^dropped
2010-10-21 10:39:30 utc hassox https://gist.github.com/78c69fc51d9915cd3a19 <-- what do you think of that?
2010-10-21 10:40:30 utc codebeaker yeah, I exepct the same… I'm building a "recommendation" system - and I want that the recommendation goes throgugh a workflow… submitted, reviewed (by a user-moderator, or admin) then it interacts with the reputation system, and the notification system, and some other wheels…
2010-10-21 10:41:04 utc codebeaker state machine with lots of guards seemed like a good fit, but since I have ~5 kinds of recommendation, all with a same workflow, mixing in a generric `state` behaviour feels dirty
2010-10-21 10:41:29 utc kennethkalmer state machines easily grow out of control
2010-10-21 10:41:53 utc hassox yup
2010-10-21 10:42:04 utc hassox we have some crazy arse shit happening with some of ours
2010-10-21 10:42:27 utc kennethkalmer http://ruote.rubyforge.org/common_attributes.html#on_error
2010-10-21 10:42:56 utc kennethkalmer maybe the :break_if can be sent to a error handler rather
2010-10-21 10:43:00 utc kennethkalmer plenty ways to do it
2010-10-21 10:50:13 utc hassox kennethkalmer: when I want to spec a participant.. how do I instantiate one?
2010-10-21 10:50:39 utc kennethkalmer just vanilla instantiation
2010-10-21 10:50:48 utc hassox just with a new?
2010-10-21 10:50:52 utc hassox or new({})
2010-10-21 10:51:11 utc kennethkalmer let me check the code quickly
2010-10-21 10:51:50 utc kennethkalmer yep
2010-10-21 10:51:55 utc kennethkalmer plain new
2010-10-21 10:52:17 utc kennethkalmer then stub out reply_to_engine, or better yet, set a proper expectation on reply_to_engine()
2010-10-21 10:52:49 utc kennethkalmer so you can test for workitem changes, if any are made
2010-10-21 10:53:00 utc hassox erm..
2010-10-21 10:53:04 utc hassox oh
2010-10-21 10:53:13 utc hassox the work item gets passed to consume
2010-10-21 10:53:25 utc kennethkalmer yes, you build up the hash yourself
2010-10-21 10:53:29 utc hassox coosl
2010-10-21 10:53:33 utc kennethkalmer :)
2010-10-21 10:55:13 utc codebeaker sometimes you guys make no sense :-D
2010-10-21 10:55:35 utc kennethkalmer only in the beginning codebeaker, only in the beginning ;)
2010-10-21 10:55:56 utc codebeaker ^ does my proposed usecase make sense?
2010-10-21 10:55:58 utc hassox codebeaker: it's hurting my head too ;)
2010-10-21 10:56:08 utc hassox codebeaker: I think so
2010-10-21 10:56:25 utc codebeaker hassox: where you from dude, Europe - it's too early for you to be a Yankee ?
2010-10-21 10:57:04 utc hassox australia
2010-10-21 10:57:12 utc codebeaker haha, oopse :-D
2010-10-21 10:57:20 utc codebeaker strange upside down people :)
2010-10-21 10:57:23 utc kennethkalmer codebeaker: and you ?
2010-10-21 10:57:47 utc codebeaker I'm on a "expats" forum… helped a lass move from Austrlia (I'm English) to Germany… today in Syndey it was 18 degrees at it's coldest… here in Hamburg, it's snowing
2010-10-21 10:57:59 utc codebeaker kennethkalmer^
2010-10-21 10:58:29 utc kennethkalmer Johannesburg
2010-10-21 10:59:02 utc kennethkalmer also upside down :D
2010-10-21 10:59:07 utc codebeaker haha
2010-10-21 10:59:29 utc codebeaker former colleague of mine moved down there, took his 6-mill share options, and spent a year in SA surfing and chilling
2010-10-21 10:59:33 utc hassox codebeaker: haha
2010-10-21 10:59:34 utc codebeaker … I'm not jealous at all :(
2010-10-21 10:59:53 utc hassox indeed
2010-10-21 10:59:56 utc hassox that would be sweet
2010-10-21 11:00:01 utc hassox ACTION is not anywhere awesome like that
2010-10-21 11:00:49 utc kennethkalmer i'd also love to couch around here for a year :)
2010-10-21 11:01:21 utc kennethkalmer afk for lunch quickly
2010-10-21 11:05:21 utc tosch_le ACTION is back from lunch
2010-10-21 11:05:45 utc codebeaker yo
2010-10-21 11:06:01 utc tosch_le codebeaker: any questions?
2010-10-21 11:06:14 utc codebeaker not really, haven't had any time to rig into ruote :(
2010-10-21 11:06:26 utc codebeaker but I see that hassox is working with it, and it's nice to see his pain :D
2010-10-21 11:06:35 utc codebeaker and bookmark every gist and link that scrolls by :)
2010-10-21 11:06:45 utc hassox hehe
2010-10-21 11:06:54 utc hassox I'm having pain with bundler right now :(
2010-10-21 11:07:10 utc codebeaker haha, unlucky hassox :(
2010-10-21 11:07:27 utc tosch_le hassox only problem is that he tries to understand all the internals of ruote ;-)
2010-10-21 11:07:40 utc hassox heh
2010-10-21 11:07:48 utc hassox I need to know how it all fits together!
2010-10-21 11:07:50 utc tosch_le (of course not – imho it's the best way to learn it)
2010-10-21 11:08:15 utc tosch_le i like kenneth's participant implementation
2010-10-21 11:08:24 utc hassox yeah it's nice hey
2010-10-21 11:08:28 utc hassox ACTION is just trying to spec it
2010-10-21 11:08:42 utc tosch_le do it as kenneth said
2010-10-21 11:09:00 utc tosch_le whip up a helper method to generate a workitem hash and go for it
2010-10-21 11:09:20 utc tosch_le those hashes aren't that complicated…
2010-10-21 11:09:42 utc hassox ja
2010-10-21 11:09:52 utc hassox can't get my specs to run atm :(
2010-10-21 11:09:54 utc hassox bundler is sucking
2010-10-21 11:10:45 utc tosch_le hmm. didn't have any problems with bundler till now
2010-10-21 11:11:51 utc tosch_le ruote-kit had some issues with the context method of rspec, though
2010-10-21 11:12:12 utc hassox it just can't find the bin
2010-10-21 11:12:21 utc hassox ACTION has blown away the gems dir and will try again
2010-10-21 11:12:40 utc tosch_le are you using rvm?
2010-10-21 11:12:45 utc hassox ja
2010-10-21 11:13:39 utc tosch_le "rvm gemset export" and "rvm gemset import" may be helpful
2010-10-21 11:14:13 utc hassox I really need to spend some time getting rvm to work for me
2010-10-21 11:14:55 utc codebeaker RVM is a cake walk
2010-10-21 11:15:09 utc tosch_le if you've troubles, ping wayneeseguin, he's very helpful :-)
2010-10-21 11:15:23 utc codebeaker and a ruote fan :-D
2010-10-21 11:15:45 utc hassox bah
2010-10-21 11:15:52 utc hassox the spec command has changed to rspec
2010-10-21 11:15:58 utc hassox yeah he is
2010-10-21 11:15:59 utc codebeaker Rspec 2 ?
2010-10-21 11:16:00 utc codebeaker yeah…
2010-10-21 11:16:01 utc codebeaker :(
2010-10-21 11:16:03 utc hassox wayneeseguin: is very awesome
2010-10-21 11:16:12 utc tosch_le +1
2010-10-21 11:16:18 utc codebeaker Is it a bird, is it a plane… no it's wayneeseguin
2010-10-21 11:17:00 utc tosch_le ACTION is running "rake spec" most times
2010-10-21 11:18:21 utc codebeaker yeah, ^ I patched that in too - got tired of remembering which verson of tools I was using on any given project
2010-10-21 11:18:39 utc codebeaker also, if you have Rails 3, and RSpec2 autotest is working flawlessly
2010-10-21 11:19:41 utc hassox oh awesome
2010-10-21 11:33:43 utc codebeaker if I'm thinknig about modeling a workflow… a flow chart is as good a place as any, right ?
2010-10-21 11:34:49 utc codebeaker woah, hassox - you're using ruote to give your new users a workflow to get their accounts upto a usable state ?
2010-10-21 11:34:57 utc codebeaker I juse read your Gist
2010-10-21 11:35:37 utc tosch_le codebeaker: didn't quite understand your flow chart question
2010-10-21 11:36:34 utc codebeaker tosch_le: it was somehow rethorical :)
2010-10-21 11:36:56 utc codebeaker … just thinking about the actors in my scenario for using a workflow… and it lends itself to a flow char, naturally
2010-10-21 11:37:08 utc codebeaker since that's the only way I know to model a workflow
2010-10-21 11:38:15 utc tosch_le do you know ruote-fluo? it shows ruote's expression trees in a nice diagram (using canvas)
2010-10-21 11:39:08 utc codebeaker ohh, lvely :)
2010-10-21 11:39:45 utc tosch_le ruote-kit uses it
2010-10-21 11:39:58 utc codebeaker cool… I still need to dive into that
2010-10-21 11:40:03 utc hassox codebeaker: yup
2010-10-21 11:40:14 utc codebeaker quick question though what persistence backends are available?
2010-10-21 11:40:26 utc codebeaker and – hassox - is your plan insane, or is it working out ?
2010-10-21 11:40:33 utc hassox dunno ;)
2010-10-21 11:40:35 utc codebeaker … (brb in 5 – gotta play Fußball!))
2010-10-21 11:40:37 utc hassox I've just started
2010-10-21 11:41:41 utc tosch_le codebeaker: http://ruote.rubyforge.org/configuration.html#storage
2010-10-21 11:42:19 utc codebeaker ohh dudes - I suck :(
2010-10-21 11:42:24 utc codebeaker at reading documentation and everything
2010-10-21 11:46:00 utc hassox codebeaker: me too
2010-10-21 11:46:09 utc hassox if it wasn't for tosch_le I wouldn't be coding right now :(
2010-10-21 11:47:08 utc tosch_le *blush*
2010-10-21 11:48:05 utc codebeaker haha, tosch_le is drining on me if he's ever in Hamburg :-D
2010-10-21 11:49:03 utc hassox https://gist.github.com/ca7f9f9bd42d84187946 <<-- kennethkalmer's nice base class with specs if anyone wants them
2010-10-21 11:49:13 utc tosch_le codebeaker: sorry, dict.leo.org fails to translate 'drining' ;-)
2010-10-21 11:49:26 utc codebeaker drinking*
2010-10-21 11:49:34 utc codebeaker apparently I suck at spelling too
2010-10-21 11:49:53 utc codebeaker Mr NiceTeee :)
2010-10-21 11:49:58 utc codebeaker Tree* god dammit
2010-10-21 11:50:14 utc hassox haha
2010-10-21 11:50:18 utc tosch_le oh, got it
2010-10-21 11:50:31 utc codebeaker nice work hassox
2010-10-21 11:50:38 utc tosch_le having a lot of fun with spelling mistakes, too
2010-10-21 11:51:17 utc tosch_le hassox: you do know that you may fork gists in github?
2010-10-21 11:51:29 utc hassox good point
2010-10-21 11:51:31 utc hassox I didn't think of that
2010-10-21 11:51:37 utc hassox damn all these modern conveniences
2010-10-21 11:51:54 utc tosch_le ACTION is grinning broadly
2010-10-21 11:51:58 utc hassox forking
2010-10-21 11:52:37 utc tosch_le codebeaker: call around if you're in leipzig one day. it's not very likely that i'll be in hamburg the next months … or years ;-(
2010-10-21 11:54:20 utc codebeaker gladly :) … everyone warned me against visiting the east :-D
2010-10-21 11:54:31 utc tosch_le are there any javascript cracks here? is there any way to make "my_obj1 < my_obj2" using a comparison method i define? how i love ruby for that…
2010-10-21 11:54:41 utc codebeaker I lived with someone from Chemnitz - he was such an asshat, other than that - I only know Berlinners
2010-10-21 11:54:45 utc hassox http://gist.github.com/638362 <-- forked
2010-10-21 11:55:08 utc hassox tosch_le: no I don't think so
2010-10-21 11:55:25 utc hassox I don't think you can make < a function
2010-10-21 11:55:27 utc hassox hrmm
2010-10-21 11:55:38 utc tosch_le hassox: nice work on the specs
2010-10-21 11:56:00 utc hassox yeah you can't do foo.< = function in js
2010-10-21 11:56:18 utc hassox tosch_le: one of the few places I think mocks and stubs are warrented ;)
2010-10-21 11:56:44 utc codebeaker the same rings true for Hamburg
2010-10-21 12:01:14 utc tosch_le hassox: too bad
2010-10-21 12:01:38 utc hassox ja
2010-10-21 12:01:42 utc hassox that would be pretty ace
2010-10-21 12:01:42 utc tosch_le foo.greaterThan(baz), then
2010-10-21 12:01:53 utc hassox or foo.gt
2010-10-21 12:01:55 utc hassox foo.gte
2010-10-21 12:03:44 utc hassox thanx for all the help again lads
2010-10-21 12:03:45 utc hassox :D
2010-10-21 12:03:59 utc tosch_le it was a pleasure, as always
2010-10-21 12:23:17 utc jmettraux added xaop to the users list : http://ruote.rubyforge.org/users.html
2010-10-21 12:46:51 utc codebeaker urgh, I hate state machines :-D
2010-10-21 12:46:57 utc codebeaker anyone used Transitions?
2010-10-21 12:48:33 utc jmettraux it's the latest one isn'it it (ex-rails) ?
2010-10-21 12:48:58 utc codebeaker yeah, that's why I picked it
2010-10-21 12:49:44 utc codebeaker i'm not quite clear on how I should be using it - but I tried to do something liek this: event :run { transition :from => :pending, :to => :running, :on_transition => :do_work }
2010-10-21 12:50:15 utc codebeaker but if do_work fails… because my finalisation, and failure transitions require the job is running, the transition from pending to fiailed/finished is invalid
2010-10-21 12:50:23 utc jmettraux ouch, I was studying it for a blogpost, but the documentation is terse, the specs may be helpful
2010-10-21 12:50:32 utc codebeaker yeah, maybe
2010-10-21 12:52:18 utc codebeaker maybe I'm making a mistake putting :do_work on the transition… but I thought that's how you were supposed to do!
2010-10-21 12:53:37 utc jmettraux maybe state :running, :enter => :do_work
2010-10-21 13:23:27 utc jmettraux tosch_le: maybe we shouldn't have a .rvmrc for ruote-on-rails, wdyt ?
2010-10-21 13:23:42 utc jmettraux (one that is in git anyway)
2010-10-21 13:24:26 utc tosch_le +1
2010-10-21 13:24:51 utc tosch_le though my rvm ask if i do trust the .rvmrc
2010-10-21 13:25:02 utc jmettraux just once ?
2010-10-21 13:25:03 utc tosch_le s/ask/asks/
2010-10-21 13:25:07 utc tosch_le yes
2010-10-21 13:25:27 utc jmettraux may I push ?
2010-10-21 13:25:35 utc tosch_le of course yes
2010-10-21 13:27:07 utc jmettraux done, thanks !
2010-10-21 13:29:05 utc tosch_le pulled locally, thanks
2010-10-21 14:07:53 utc jmettraux ok, leaving, thanks a ton !
2010-10-21 14:08:11 utc tosch_le bye!
2010-10-21 14:09:16 utc jmettraux ciao !
2010-10-21 14:29:19 utc tosch_le i'm leaving, too. bye!
2010-10-21 23:25:52 utc jmettraux j_dana: hello, welcome to #ruote
2010-10-21 23:31:15 utc j_dana jmettraux: thanks
2010-10-21 23:32:59 utc hassox morning
2010-10-21 23:33:05 utc jmettraux hassox: morning
2010-10-21 23:33:12 utc hassox hey jmettraux
2010-10-21 23:33:14 utc hassox how are you mate
2010-10-21 23:33:20 utc jmettraux fine and you ?
2010-10-21 23:33:24 utc hassox pretty good
2010-10-21 23:33:31 utc hassox ACTION is on the train going to a full weekend of hacking
2010-10-21 23:33:37 utc jmettraux ouch
2010-10-21 23:33:38 utc hassox and I forgot my glasses :9
2010-10-21 23:33:47 utc jmettraux ouch x 2
2010-10-21 23:34:08 utc hassox ja
2010-10-21 23:34:09 utc hassox twill hurt I'm sure
2010-10-21 23:34:38 utc jmettraux glossy screen, respiration fog, ...
2010-10-21 23:38:32 utc hassox it's not going to be plesant
2010-10-21 23:38:32 utc hassox I might have to go home for a few hours tomorrow
2010-10-21 23:38:37 utc hassox grab them and come back down
2010-10-21 23:38:37 utc hassox see how it goes today
2010-10-21 23:39:18 utc jmettraux please do your best
2010-10-21 23:42:02 utc hassox jmettraux: given an engine... how do I get process definition 'foo' from it
2010-10-21 23:42:02 utc hassox ?
2010-10-21 23:42:10 utc jmettraux you can't
2010-10-21 23:42:21 utc hassox so
2010-10-21 23:42:48 utc jmettraux the philosophy is engine.launch("http://process.definition.server/my/freaking/process_definition.rb")
2010-10-21 23:42:56 utc hassox how do I get hold of my process definitions?
2010-10-21 23:43:02 utc hassox do I need to track them myself?
2010-10-21 23:43:06 utc jmettraux yes
2010-10-21 23:43:25 utc jmettraux with a webserver, a document management system, a cgi script, whatever
2010-10-21 23:43:33 utc jmettraux or just inside of your ruby code
2010-10-21 23:43:42 utc hassox ok
2010-10-21 23:43:48 utc hassox so... why name a process definition?
2010-10-21 23:43:59 utc hassox Ruote.process_definition :name => "/account/new" do
2010-10-21 23:43:59 utc jmettraux (for recursion)
2010-10-21 23:44:39 utc jmettraux let me check if I expose the name as a variable (I should)
2010-10-21 23:45:40 utc jmettraux p engine.process(wfid).definition_name # => "/account/new"
2010-10-21 23:45:46 utc jmettraux but a name is optional
2010-10-21 23:50:46 utc hassox jmettraux: so
2010-10-21 23:51:34 utc hassox if I store the definitions off in a variable, can I launch them multiple times with different args?
2010-10-21 23:51:34 utc hassox or do you have to define it with the arguments you're going to use?
2010-10-21 23:51:54 utc jmettraux it depends on the arguments
2010-10-21 23:52:17 utc jmettraux engine.launch(process_definition, { workitem fields initial hash }, { process variables hash })
2010-10-21 23:52:25 utc jmettraux where the two hashes are optional
2010-10-21 23:52:51 utc jmettraux for now you only need to remember engine.launch(process_definition, { fields })
2010-10-21 23:53:25 utc jmettraux the pdef in a variable will work great
2010-10-21 23:54:00 utc jmettraux the launch method is smart enough to figure out if it's a process definition or a URI to a process definition
2010-10-21 23:54:36 utc jmettraux some people have things like engine.launch(generate_process_definition(args))
2010-10-21 23:54:48 utc hassox that's what I was wondering
2010-10-21 23:55:00 utc hassox if it should be a static thing, or a generate the definition each tiem
2010-10-21 23:58:33 utc jmettraux it depends on what level of fluidity you want to achieve
2010-10-21 23:58:52 utc jmettraux ruote only cares about your process at launch time
2010-10-21 23:59:18 utc hassox right
2010-10-21 23:59:23 utc jmettraux (hence the "no ruote does no process definition collection management for you")
2010-10-21 23:59:38 utc hassox I just want to be able to hand it an account or a 'foo'
2010-10-22 00:00:01 utc jmettraux engine.launch(account_process_def, { 'account_id' => 'foo' })
2010-10-22 00:00:20 utc jmettraux will make sure the workitem has a field "account_id" with the value "foo"
2010-10-22 00:00:29 utc hassox that's enough :)
2010-10-22 00:00:36 utc fixr hassox: I wrote a very simple definition routine in rails to store and manage them on a table, and then launch processes with them. Take a look at the create method, it's the only special one which does definition validation: http://www.pastie.org/1239557 ;)
2010-10-22 00:00:41 utc jmettraux (sorry for the long explanation)
2010-10-22 00:00:52 utc hassox guys this is awesome
2010-10-22 00:00:55 utc jmettraux fixr: hello, many thanks
2010-10-22 00:01:00 utc hassox you guys are so helpful :)
2010-10-22 00:01:05 utc hassox thanx fixr
2010-10-22 00:01:30 utc fixr hassox: for the rest, with a simple scaffold will do
2010-10-22 00:01:41 utc fixr jmettraux: hey! no prob :)
2010-10-22 00:02:48 utc fixr jmettraux: I'm back from a 2 week vacation off work. I haven't forgotten about ruote, y'know hehe
2010-10-22 00:06:44 utc jmettraux :)
2010-10-22 00:09:38 utc hassox jmettraux: is there a way to launch the process and then say
2010-10-22 00:09:46 utc hassox engine.wait_for(wfid)
2010-10-22 00:09:55 utc jmettraux yes
2010-10-22 00:10:01 utc jmettraux just like you said
2010-10-22 00:10:22 utc hassox sweet :D
2010-10-22 00:10:44 utc jmettraux it will block until the process instance terminates or triggers an error