On Tue, Jun 23, 2009 at 8:43 PM, Peter Dalmaris <[email protected]>wrote:
> > Hello, Hi Peter > we are preparing to start work on an application that will heavily > depend on a workflow engine. This engine will need to handle a dozen > workflows, around 50 users, route workitems through email and web. Congrats on choosing a WFE and not a state machine. > I have been looking into Ruote for a while now, but finally got > serious a couple of days ago. Since then, I have installed ruote-web2, > created a simple workflow definition and installed it, added users and > groups, ran the workflow; it all seems to work fine. > > However, I am not convinced yet that Ruote is the way to go. I would > be very interested to hear relevant stories from the members of this > group. Descriptions of production applications build with Ruote, > successes and failures, arguments for or against this solution, words > of wisdom etc. Be prepared to work. This is my only advice. John Mettraux is an awesome lead dev, very willing to help, but expects you to be helpful too, which I totally respect and has made me earn his respect in a big way... There is no spoon feeding in these parts. Now, here is the long and the short of it from my side, take John's reply into perspective as well. * It took me several months to understand how ruote works, purely because I was inexperienced in BPM/WFE. * I couldn't care less about BPEL, BP.., Visio or any other 'standards for business'... * IT IS NOT A STATE MACHINE (but can be _reduced_ to act like one) * Write your own participant/listener, using others as a reference, it is a great learning curve. * Use ruote-rest, or upcoming ruote-kit, if you're ever planning on running more than one Rails instance * Be prepared for a lot of manual testing * Live in #ruote, it's a cozy little channel * Use sunscreen As for production use, we went live last week on a Rails + ruote-rest system. We have well over 200 human participants and about 30 autonomous participants (running on 9 different servers). The engine is running of a YAML filestore, and is holding up brilliantly. It has not been without issues, but 99% of the issues is a result of our own code and implementation, not the engine's. John noted that ruote 2.0 will "scale" better, and ruote-kit will be built to leverage these improvements. > I suppose any responses could be compiled into a Ruote "showcase". > It is extremely difficult to compile a "showcase" without users giving away their business models. John and myself have been chatting on building some interesting workflows as samples, allowing us to show off various parts of the engine. As for a showcase, you can examine our company websites ( http://www.inx.co.za & http://www.ispinabox.co.za) to get an idea of our wholesale ISP business, and everything we do is now powered by ruote... My best advice is to do your homework, I don't wanna scare you off the project. I love it, it is an amazing piece of work. 1 year on I'm still learning new things about it every single day. My processes are getting leaner and faster, my participants are improving constantly and I bask in the rays of greatness. If ruote fits, wear it, you'll regret it at first and love it a few weeks later! Best -- Kenneth Kalmer [email protected] http://opensourcery.co.za @kennethkalmer --~--~---------~--~----~------------~-------~--~----~ you received this message because you are subscribed to the "ruote users" group. to post : send email to [email protected] to unsubscribe : send email to [email protected] more options : http://groups.google.com/group/openwferu-users?hl=en -~----------~----~----~----~------~----~------~--~---
