RE: Closing issue X

> -----Original Message-----
> From: Mark Baker [mailto:distobj@acm.org]
> Sent: Friday, December 13, 2002 9:17 AM
> To: Newcomer, Eric
> Cc: www-ws-arch@w3.org
> Subject: Re: Closing issue X
> 
> 
> 
>  
> Problem; machine-to-machine integration over the Internet.
> Solution; services where each has an interface different than the
> others
> 
> That's broken, sorry.  I don't know any other way to say it.  I could
> perhaps be more sensitive with my wording, but the message 
> would remain the same.

Mark, you've argued this point here for months, and now you've taken it to
the TAG.  I strongly suggest that we 

a) Wait to see if the TAG agrees with you

b) Wait to see if Web services developers agree with you (after all, they
MAY honor the "uniform interface constraint" with SOAP 1.2 and the WSA doc
as it stands, you seem to be arguing that this should be a MUST).

c) Wait to see if Reality agrees with you.  You seem to be making some
strong predictions that Web services that ignore that uniform interface
constraint will in some sense fail, and those which honor it will succeed.
The world is (perhaps unwittingly) performing a giant experiment to test
this hypothesis.

d) In the meantime, make sure that the WSA document is clear that a range of
interface definition options are possible, and that best practice has yet to
be determined.  

Telling the people who have been doing business together for years and
"know" what interfaces are available that they have to adopt the Uniform
Resource Constraint is a non-starter. These folks are using SOAP as is,
often with much success.  WSDL helps automate things further.  On the other
hand, we could reasonably talk about how best to bootstrap Web service
interactions between parties that have little or no a priori knowledge of
one another.  That is a use case for Web services, clearly.  SOAP, WSDL, and
UDDI seem to be moving in a direction that would accomodate the Web more
cleanly to discover what services another offers and how to invoke them.  I
can easily imagine that some "best practice" language on how offering the
HTTP interfaces as a lowest common denominator can assist this bootstrapping
process, and certainly some concrete real-world examples of success stories
employing this approach, would find a happy home in the WSA document.  

We have resolved to focus on the document, and on clarifying the rationales
for various approaches and the situations in which one or the other is most
appropriate.  I invite you to help us incorportate your perspective into the
document in an appropriately qualified way rather than continuing to try to
persuade us that all perspectives besides your own are "broken."

Received on Friday, 13 December 2002 10:33:58 UTC