Re: [wsrf] question on WS-BaseFaults

[Steve Graham <sggraham@us.ibm.com>]

Interesting.  Re-reading section
3 with your concern in mind does indicate that there is some ambiguity
here.  Although there is no clear text that states it, one can interpret
section 3 to state: any WSDL fault on any WSDL operation associated with
a WS-Resource MUST follow the rules outlined in the section.

I think it is an inhibitor to composability
to make this strong a normative requirement on base faults.

I would propose that we log an issue
against base faults and propose that the text in section 3 be relaxed to
a SHOULD.

sgg

++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, IBM Software Group, Web services and SOA
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++

"Springer, Ian P." <ian.springer@hp.com>

02/17/2005 04:26 PM

To	<wsrf@lists.oasis-open.org>
cc
Subject	[wsrf] question on WS-BaseFaults

For implementation purposes, I'd like to know whether
all faults thrown
by operations in a WS-Resource WSDL are required to be base faults (i.e.
types that extend BaseFault), or whether other forms of soap faults may
also be used. After reading the Nov 04 draft of the WSRF-BF spec, I am
still unsure. Section 3 (Use of Base Faults in WSDL) seems like it might
be suggesting that all faults must be base faults, but it's not clear.
Can someone please clarify?

Thank you.
Ian