Re: Dubray paper comments + questions

John Dart wrote:
> I think a lot of this may be necessary, but (to revisit an earlier 
> issue raised on the list) I'm not sure that the requirements for 
> choreography necessitate a full-blown workflow modelling language. 
> In fact, I have some concern that the existing proposals may be 
> overly complex for modelling useful types of WS interaction 
> (especially in a B2B context), while being less than adequate 
> for more general worklow purposes, in which not everything 
> is directly in service of a WS message exchange. 
> Which is why the proposal to separate specification of message 
> flow from control flow was attractive, at least IMO.

I'm with you.  That's also why I favor separating the external
conversations (e.g. B2B) from internal workflows.
I don't even think procedural workflow is an appropriate model
for external conversations.

-Bob Haugen

Received on Wednesday, 26 February 2003 19:08:24 UTC