Thanks Maciej, Given that, I think I'll weigh in and say I support the decoupling of PID and process name ( because as Maciej mentioned, there isn't anything here that forces it ). That and the fact that the use case "Process Reuse" seems pretty compeling to me ( we don't force process renaming, we don't force WS-Addressing, we ship a simple partner repository solution ).
Lance On 8/25/06, Maciej Szefler <[EMAIL PROTECTED]> wrote:
Lance, The "name" attribute on the <process> element is the PID. There is currently an optional <type> element under the process element that does the decoupling. If the <type> element is present it is used to specify the BPEL definition's QName. If it is omitted then the definition's QName is assumed to be the same as the name/PID. -Maciej On 8/25/06, Lance Waterman <[EMAIL PROTECTED]> wrote: > > Maciej, > > I would like to get a bit of clarification from you. I have been looking > at > the current implantation ( again, just to get an understanding of where we > are vs where we are going ) and as I currently see it the deployment > descriptor contains the PID in the form of the name attribute on the > <process> element - is this correct? If so, moving forward are you > suggesting that another element/attribute might be added to the deployment > descriptor that decouples the process name from its PID? > > Lance > >
