Flowr presentation at BPMinna
There were 30 people, rather young, seeming to come from different backgrounds, looking forward to the presentation and to the following beers.
(Disclaimer : my level of japanese is low, it’s very probable that I mistook some of the information given during this presentation)
Flowr is a Ruby workflow engine implemented on top of a FSM (finite state machine). Its context seems to be among blogs and wikis with its first use case is about publication workflows.
States and transitions are modeled via Ruby classes (
Transition). States are rather minimal, the logic seems completely located in the transitions.
Nakatani-san showed us examples where the transition action contained calls to an AtomPub library (I’m wondering if he’s using Atom-Tools).
Flowr seems to be an interesting tool to interact with Atom / RESTful web resources.
Here is a sample Flowr process definition (found at http://flowr.31tools.com) :
# flowr DSL の例(部分)
class SampleStateChart ‘http://sample.com/blog/’
class NotOwnedState < State name "所有者無し" end class OwnedState < State name "所有者有り" end class ThrowDiceTransition 0 '食べ物' else '参加してください' end end def action(context) # さいころを振る context.properties.dice = rand(6)+1 end end end [/sourcecode] I had a lot of questions, about persistence, process versioning and so and so. I reserved them for after the presentation. But unfortunately, there were so many other interesting discussions at the restaurant that I couldn't get the answers. Next time. Nakatani-san will probably release soon Flowr on Rubyforge under a BSD or a MIT license.
During the dinner/beer party that followed, I had the pleasure to chat with various people with interests in BPM, workflow and Ruby.
At first I discussed with Wakizaka-san who translated the jBPM user guide to Japanese. He’s also one of the main organizers of this BPMinna group (I’m looking forward the next meeting).
Then I had the pleasure to meet Buri‘s Habu-san and Buri’s author Makotan-san (Buri is a japanese open source workflow management system which uses xpdl process definitions).
It’s good to meet another workflow/BPM system creator, we seem to share the same challenges. Makotan-san promised me a demonstration of Buri at the next meeting.
As you can see, it was quite a gathering of workflow/BPM aficionados.
Katoh-san is the one of the two authors of AP4R, I had met him very briefly at the RubyKaigi 2007. Friday, we had an interesting discussion with Makotan-san, Katoh-san asked whether messaging was important to workflow/BPM systems, we had to agree on that, though I would love to discuss the subject further with them.
I chatted with other people, people not yet using BPM tools but “watching the thing closely”. I wished I could discuss with more people… As I’ve said, I had to escape (last train) before having the opportunity to exchange views with Nakatani-san.
Many thanks to Wakizaka-san and friends for this excellent event !