- From: Sanjiva Weerawarana <sanjiva@watson.ibm.com>
- Date: Thu, 6 Feb 2003 22:39:19 +0600
- To: "Dale Moberg" <dmoberg@cyclonecommerce.com>, "FABLET Youenn" <youenn.fablet@crf.canon.fr>
- Cc: "Roberto Chinnici" <roberto.chinnici@sun.com>, <www-ws-desc@w3.org>, "Jean-Jacques Moreau" <jean-jacques.moreau@crf.canon.fr>
"Dale Moberg" <dmoberg@cyclonecommerce.com> writes:
>
> Gudge expressed "reservations" in a forceful way about re-inventing the
^^^^^^^^^^^^^^^^^
> cardinality apparatus within input,output, and fault.
I would've been disappointed with Gudge if he hadn't done precisely
that ;-).
I am ok with dropping those, but IMHO they're very useful. Yes, it
reinvents (i.e., copies) a useful concept, but IMO that's ok because
it brings in the key value of using XSD for that level without
bringing in all the complexity of XSD. As a result, it allows
me to say something like "I can handle 1 or more image/gif attachments."
> Also the wsdl-MEP combinatorics did not fit with the repeated input
> element. That would be easily correctable, though, if input, output (and
> fault?) were allowed to have content models with repeated "part" EIIs.
Yeah, there are several ways of addressing that problem.
> So, my scorecard now shows that the distaste for repeating the
> cardinality apparatus in attributes on input, etc as being the main
> obstacle to the Sanjiva compromise approach. Others may be tracking
> different statistics though.
Can others indicate what other problems were found please?
Thanks,
Sanjiva.
Received on Thursday, 6 February 2003 11:42:18 UTC