- From: Peter Furniss <peter.furniss@choreology.com>
- Date: Tue, 16 May 2006 23:17:39 +0100
- To: "'Gary Brown'" <gary@pi4tech.com>, "'Steve Ross-Talbot'" <steve@pi4tech.com>
- Cc: "'Nickolas Kavantzas'" <nickolas.kavantzas@oracle.com>, "'WS-Choreography List'" <public-ws-chor@w3.org>
I'm slightly surprised that I'm still on this list, but for what it is worth, I agree with Gary - a failure in finalize should be handled by the parent. Peter > -----Original Message----- > From: public-ws-chor-request@w3.org > [mailto:public-ws-chor-request@w3.org] On Behalf Of Gary Brown > Sent: 16 May 2006 10:13 > To: Steve Ross-Talbot > Cc: Nickolas Kavantzas; 'WS-Choreography List' > Subject: Re: Exception handling for the actions in a finalizerBlock > > > > Hi, > > My take on this issue is that the exception handlers for a > choreography > are intended to handle faults that occur during the scope of > the normal > activities within that choreography. Therefore once the > choreography has > reached the "successfully completed" state, those exception > handlers are > no longer relevant. > > The finalizer blocks are intended to perform the role of a commit or > rollback in terms of the work performed by its associated > sub-choreography, and although the finalizer blocks activities are > performed in the scope of the sub-choreography being > finalized, so that > it has access to the state associated with that > sub-choreography, it is > actually performed under the direct control of the parent > choreography. > > Therefore I think my preference is that a failure during the finalize > should be handled by the parent choreography - i.e. so the exception > handlers focus on handling failures in the normal flow of the > choreography in which it is associated, and the failure to finalize a > performed sub-choreo, is the responsibility of the parent > choreography > that both performs and then attempts to finalize the sub-choreo. > > If this position seems reasonable to other members, then we > just need to > ensure that the spec is consistent, which can be handled at the same > time as making minor changes resulting from the implementation stage. > > Regards > Gary > > > Steve Ross-Talbot wrote: > > > > Nick, > > > > can you comment on this email that we received. > > > > > > > http://lists.w3.org/Archives/Public/public-ws-> chor-comments/2006Apr/00 > > 11.html > > > > > > Cheers > > > > Steve T > > > > > > > > > > > >
Received on Tuesday, 16 May 2006 22:20:49 UTC