RE: pattern URIs

> Roberto Chinnici writes:
> Sanjiva Weerawarana wrote:
> >
> >>I still support your original position, hence I disagree with making
> >>this pattern normative. How about the compromise of having the
pattern
> >>Amy proposes in a non-normative appendix to the patterns spec? We
define
> >>it properly, assign it a URI, use it to elucidate the patterns
> >>framework, make it available for anybody to use it (if they have
> >>a binding for it, that is), but it's *not* normative.
> >
> > I think this is ok .. and basically this is all we can do for any
> > patttern. It is however normative to the point that if someone
> > uses that pattern URI then the semantics MUST be exactly as
specified.
> > Beyond that there's no real "normativeness" for pattern URIs. Some
> > patterns will of course get exercised in bindings we do (i.e., the
> > bindings will only be applicable to those patterns), but that's as
> > far as it goes.
> 
> I think we are in agreement. If a pattern is non-normative, then a
tool
> doesn't have to recognize it. (It's not even a MAY recognize. The
> normative spec doesn't even know about the pattern.) If the tool does
> recognize it, though, it must follow the appropriate specification,
like
> for any other extension.

I don't see the distinction here between behavior for 'normative'
patterns and 'non-normative' patterns. A WSDL processor is not required
to recognize any patterns, but if it does recognize one of the patterns
we define, it must follow the semantics as specified.

Are we going to start saying that a WSDL processor must recognize all
bindings? WSDL 1.1 didn't do that, and the community seems to have voted
with their implementations.

--Jeff

Received on Wednesday, 1 October 2003 22:54:08 UTC