RE: SOAP Modules and Features (LC18, LC29b)

Hi Glen,

Thank you for following up on LC18.

> I just reread the text you point to, and I do 
> think we could profitably add some text which describes 
> the fact that Features in fact are typically resolved/
> implemented by bindings or Modules, and that Modules can
> satisfy abstract Feature requirements.  I think this might
> help with Marc's concern as well, without going so far as
> to generate a syntactic connection between the two.  I'll 
> volunteer to write this (i.e. switch my ACTION from writing 
> this email to writing some resolution text :)).

I am assuming that this new text will span all three parts of the spec. I am
looking forward to read it.

I like to comment on one of your statements,

"By virtue of the fact that you understand the SOAP Module URI, you
understand which, if any, abstract features the module implements."

You are making an assumption that the SOAP Module spec author:

- will specify an URI for a WSDL Abstract Feature
- will declare the WSDL Abstract Feature that their SOAP Module realizes

If so, I expect the WSDL spec to stipulate these requirements for SOAP
Module spec authors. Right?

Regards,
Asir S Vedamuthu
asirv at webmethods dot com
http://www.webmethods.com/

-----Original Message-----
From: Glen Daniels [mailto:gdaniels@sonicsoftware.com] 
Sent: Thursday, October 21, 2004 3:01 AM
To: hadley@sun.com; Asir Vedamuthu
Cc: WS Description List
Subject: SOAP Modules and Features (LC18, LC29b)



Hi Marc, Asir:

You both raise issues regarding the lack of explicit relationship
between SOAP Modules and Features as defined by WSDL.  Asir asks whether
we should have text which describes that Modules can implement Features,
and Marc asks whether there should be some (syntactic?) reference from
Modules declared in a binding/endpoint to Features declared elsewhere.

Marc:  We discussed this at the F2F in Toronto, and decided that an
explicit syntax for this kind of thing isn't necessary.  By virtue of
the fact that you understand the SOAP Module URI, you understand which,
if any, abstract features the module implements.  And if you don't
understand a given (non-required) SOAP Module, you can't use it to
confirm implementation of any features at all, since you don't know what
it is and won't be using it.

The only real use I could see for having an explicit syntax for
describing which Modules implement which Features would be to allow WSDL
tools which did NOT in fact understand the Modules/Features in question
to decide if the WSDL seemed valid or not.  This doesn't seem all that
useful to me in the first place, and in the second, it provides an
opportunity for a mismatch between the claims of the WSDL and the claims
of the specifications.  The specs (and the code written to implement
them) should always be the final arbiter.  Therefore I think we should
take no action with respect to LC29b.

Asir: Although we had originally rolled 18 in with 29b, I just reread
the text you point to, and I do think we could profitably add some text
which describes the fact that Features in fact are typically
resolved/implemented by bindings or Modules, and that Modules can
satisfy abstract Feature requirements.  I think this might help with
Marc's concern as well, without going so far as to generate a syntactic
connection between the two.  I'll volunteer to write this (i.e. switch
my ACTION from writing this email to writing some resolution text :)).

Seem reasonable?

--Glen

Received on Thursday, 21 October 2004 14:47:29 UTC