- From: Andrew Berry <andyb@whyanbeel.net>
- Date: Wed, 16 Jul 2003 23:52:10 +1000
- To: public-ws-chor@w3.org
Jeff Lansing <jeff@polexis.com> wrote: > This example is interesting, but it seems like a lot of machinery to > describe what is essentially just connecting to something like a > GetStockQuote service. > > What if you try to do something just a little harder, such as pulling > (or pushing) the results of a StockQuote service through a > CurrencyConverter service? > > Does the amount of pi-calculus machinery then become so cumbersome, in > that case, that it starts to loose its intuitive appeal? I certainly found this to be the case when attempting to formalise work on a coordination language. There were a variety of issues, but mostly because pi-calculus is designed for describing the behaviour of a concurrent program and not the cooperative behaviour of autonomous distributed participants. As noted in my first post (http://lists.w3.org/Archives/Public/public-ws-chor/2003Jul/0065.html) the solution was to develop a formalism reflecting the higher-level semantic concepts. I did consider building that formalism on top of the pi-calculus or other process algebra, but found other limitations as well (global state assumption, interleaved concurrency model, difficult to model locality). Ciao, AndyB
Received on Wednesday, 16 July 2003 09:50:29 UTC