It’s not a big deal.
What I should have said was protocol operation vs. protocol sequence related
faults.
From: Gilbert Pilz
[mailto:Gilbert.Pilz@bea.com]
Sent: Tuesday, March 07, 2006
10:19 AM
To: Marc Goodner;
ws-rx@lists.oasis-open.org
Subject: RE: Issue i098 - This is
about protocol vs. sequence faults
I guess
I don't understand the opposition of "protocol vs. sequence". In my
mind an InvalidAcknowledgement is certainly a "protocol fault" (as
in, "Dude, you're not following the protocol 'cause you acknowledged
something I never sent you!"). It is also a "sequence-related
fault" in that the invalid acknowledgements included a <wsrm:Identifier>
that identified a valid sequence (otherwise the RMS would presumably have sent
an UnknownSequence fault).
All the
faults defined by WS-RM are "WS-RM protocol faults". Some of these
faults relate to a particular sequence. Some of these faults do not.
- gp
From: Marc
Goodner [mailto:mgoodner@microsoft.com]
Sent: Tuesday, March 07, 2006 9:11
AM
To: Gilbert Pilz;
ws-rx@lists.oasis-open.org
Subject: Issue i098 - This is
about protocol vs. sequence faults
The issue says “sequence” and
“non-sequence” faults, I would characterize it more as protocol vs.
sequence faults though.
I agree that the text referred to is garbled, I think the
editor may have already made a change to that text to try to improve it. If so
could one of the editors provide that to the list so we can discuss this with
the most recent language in the spec?
It seems to me that if this text is fixed, i.e. that it is
referring to protocol operation faults, that is probably sufficient. Are you
suggesting that more needs to be done? For example do you think that every
fault needs to be annotated in the spec as to whether it is a protocol or
sequence fault?
Marc Goodner
Technical Diplomat
Microsoft Corporation
Tel: (425) 703-1903
Blog: http://spaces.msn.com/mrgoodner/