OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

wsbpel message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: RE: [wsbpel] process coordination ideas - counter-proposal


Thanks for the second version - helps considerably.  I'm also looking
forward to our time in Melbourne :-). Couple of thoughts in the mean

I agree one can get the same effect with application-specific messages -
and indeed that the work actually done in each circumstance would be the
same.  However, it does mean that the invoking process has to know what
operations are on the subordinate process that variously produce cancel,
confirm and "compensate" [1]. If we could delegate to a coordination
mechanism, we can increase the encapsulation provided by the subordinate
process.  (though certainly there will be cases where some element of
application information creeps back in again on the c* messages)

As I've been arguing elsewhere, I believe the various coordination
specifications overlap very closely when the focus narrows to a
two-party relationship. The differences mostly come down to what
everyone else involved is doing.

See you on Tuesday.


(apologies to anyone getting this twice - for some reason my mailer
decided to try sending it through the wrong 
account and OASIS bounced it)

> -----Original Message-----
> From: Satish Thatte [mailto:satisht@microsoft.com] 
> Sent: 07 December 2003 22:19
> To: Haugen Robert; wsbpel@lists.oasis-open.org
> Subject: RE: [wsbpel] process coordination ideas - counter-proposal
> Robert,
> Yes, the discussion promises to be interesting ;-)  
> I have serious doubts about the practical *detailed* plug 
> compatibility of the various coordination layer 
> specifications and their usage modes, although at some 
> conceptual level there is obviously very substantial overlap.  
> Regarding your question about "what if the goods are 
> shipped?"  relative to my example -- note that 
> "ConfirmAndShip" is the confirm handler operation and thus 
> the scope for cancel ends before the possibility of 
> compensate/confirm arises.
> I realize I had provided no such explanations in my PPT so 
> attached herewith is an update with an additional explanation 
> slide and a few other clarifications suggested by others.
> Your alternative pattern looks interesting.  Would like to 
> discuss face to face.  I am hopeful that it can also be 
> captured by existing features.
> Satish
> -----Original Message-----
> From: Haugen Robert [mailto:Robert.Haugen@choreology.com] 
> Sent: Sunday, December 07, 2003 12:16 PM
> To: wsbpel@lists.oasis-open.org
> Subject: RE: [wsbpel] process coordination ideas - counter-proposal
> Satish and all,
> I took the liberty of using your slides as a format for a 
> counter-proposal. 
> Hope to have an interesting discussion...
> -Bob Haugen

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]