Re: [wsbpel] Removing "compensating a process as a whole"

[Assaf Arkin <arkin@intalio.com>]

Is it fair to assume that the specification is intended to foster 
interoperability between systems? Can we do that and still leave a lot 
of things unspecified?

Satish Thatte wrote:

>Consider also the BA protocol in WS-Transaction which shows how the compensation of a subordinate instance would be invoked by a controlling instance, if we were to specify such a subordinate-controller relationship in BPEL, effectively "remoting" a scope as a separate instance (without implicit state sharing).  The BA protocol messages could be handled by implicit event handlers that maintain the BA protocol semantics.
>  
>
>I am not saying yet that we should actually do this, just throwing out some interesting possibilities to think about.
>  
>
It's a very interesting possibility. At least I'm interested in that 
possibility ;-)

The current spec seems to suggest in one place that it could be done 
(the appendix), and it seems to suggest in another place that it is 
permitted (implementation specific manner). Yet it does not define how 
it would be done in an interoperable fashion, leaving open the question 
of how would you go about using WS-TX, if you so choose to, and use it 
in an interoperable manner. Again, while most of the grunt work should 
be left up to the WS-TX spec, there's a few requirements placed on BPEL 
to actually clarify how things should be done.

So right now BPEL is in a limbo state where as an interoperable 
specification it does not allow you to use WS-TX (I can point out to 
specific issues, if you are interested), the only way to go about doing 
it is with something similar to the pattern described by Yuzo.

Is there any intent for BPEL to provide an interoperable way to use 
WS-TX (or similar protocol) to invoke compensation of a process as per 
its abstract definition?

arkin

> 
>
>Satish
>  
>