John, On May 17, 2010, at 11:29 PM, John Mettraux wrote:
> > On Mon, May 17, 2010 at 11:05:23PM -0400, Cappelaere Patrice wrote: >> >> I did read the example... in the case of ruote_fluo...If you take that >> definition and put it in a file,,, I pretty much described what happens, I >> thought. > > Interesting, for me, when I put it into ruote-fluo, I get : > > http://ruote.s3.amazonaws.com/alpha_bravo.png > > Oh wait, let me guess, you put "engine.launch(pdef)" part with it too ? nope! > > >> The point was to show that you may want to expose more of what is happening >> at the workflow level rather than hide it in a participant (which would not >> be obvious to the casual reader). This is why you want to use workflows >> after all. > > What is not obvious about > > sequence do > calculate_next_trajectory > emit_trajectory > end What's not obvious is what's inside of those participants. I might as well hide the whole thing as a participant and what I have is a simple asynchronous RPC call to a web processing service. > > Do you remember that ruote is totally asynchronous ? Why do you want to slap > around synchronous calculations right in the middle of a process definition ? > > Programming is about building "higher" abstraction levels. > Oh yes, I remember that. But sequences are done in step manner. Yes I know what Programming is about and agree with your statement. Programming is also about communication of intent for the customers that will have to work with the system. Sometimes they may want to know what is going on. >> Why couldn't I use a workflow to design an algorithm and the many processing >> steps? >> This would be really cool.... soooo close!. > > Lean the tool, use participants. That's really a low blow! You were probably not having a good day? Pat. -- 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
