Re: Issue with binding message references in the abstract component model from Sanjiva Weerawarana on 2003-09-21 (www-ws-desc@w3.org from September 2003)

From: Sanjiva Weerawarana <sanjiva@watson.ibm.com>
Date: Sun, 21 Sep 2003 13:51:30 +0600
To: "Roberto Chinnici" <Roberto.Chinnici@Sun.COM>, <www-ws-desc@w3.org>
Message-ID: <00e601c38015$2dbf0db0$f4c16720@lankabook2>

"Sanjiva Weerawarana" <sanjiva@watson.ibm.com> writes:
> 
> I know I don't understand the fault rules etc. in the MEP stuff
> yet, but why doesn't the name matter for faults too? In a complex
> MEP there can be faults going in different directions etc. and 
> in such cases it seems to be necessary to say which fault I'm
> talking about when I indicate the actual message contents.

However, I do agree that for the simple In-only and In-Out meps 
(and their corresponding reverse ones) the message references
are indeed irrelevant for faults.

What I'm concerned about is dropping them for the general case.

Sanjiva.

Received on Sunday, 21 September 2003 13:51:37 UTC