RE: Mapping Specs to the Architecture

Mapping Specs to the Architecture
  -----Original Message-----
  From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On
Behalf Of Newcomer, Eric
  Sent: Tuesday, March 11, 2003 7:31 PM
  To: Champion, Mike; www-ws-arch@w3.org
  Subject: RE: Mapping Specs to the Architecture


  I think we need to identify specs that fit the architecture.  I'm not sure
Frank's diagram is the right one for this, since it's focused on concepts
and relationships more than functional areas like choreography, security, or
transactions.

  I can see the problem - in some areas, such as transactions, we have
multiple specs that are sort of "competing" and we might be seen to be
taking sides, or trying to endorse "winners."

  I definitely agree with David. As a vendor I have a preference for one of
these specs, I would like to support just that one spce, and I hope it
becomes the winner in this caucus race. But can the WSA elect the winner and
if so on what ground?

  Mentioning some specs in the WSA is equivalent to electing or at least
promoting a winner. As we've all learned there are three things we can count
on: death, taxes and more WS specs.

  If we mention the two that exist today but not the three that exist
tomorrow that would create a clear preference by act of exclusion. And while
the W3C has the priviledge to promote specs it works on, many of these specs
are developed by other bodies, and some have the luck of being adopted by
mere act of publication..

  So I agrue that unless we can reach a concensus, as is the case with WSDL
and SOAP, we should not mention any spec in the WSA document. Leave it open
for now. But it goes without saying that we should start reaching concensus
on as many specs as possible.

  arkin

  But we have talked about specific technologies consistently in the context
of "examples" that support the abstractions.  I think this fits, since it's
impossible to generalize without specifics from which to draw conclusions,
and it's also not appropriate to document functional areas without concrete
instantiations of them.

  My recommendation is to come up with a stack diagram, along the lines of
Martin's suggestion during the F2F, and map example specs to it.  We are not
in a position to enforce conformance, or to pass judgement about the
suitability of specs for the purpose, but we can at least classify them and
identify their place in the stack.

  Eric
    -----Original Message-----
    From: Champion, Mike
    Sent: Tuesday, March 11, 2003 8:29 PM
    To: www-ws-arch@w3.org
    Subject: RE: Mapping Specs to the Architecture



      -----Original Message-----
      From: David Orchard [mailto:dorchard@bea.com]
      Sent: Tuesday, March 11, 2003 8:03 PM
      To: 'Champion, Mike'; www-ws-arch@w3.org
      Subject: RE: Mapping Specs to the Architecture


      II ask again, what would the point be? Is the ws-arch to provide
educational material, ala conferences/books?  There's a big difference
between doing an architecture for education reasons vs doing an architecture
for describing properties/constraints.

    I guess I see Dave and Mike M.'s point EVEN better now.  I wouldn't
object if we did some "education" work along the lines that TimBL suggested,
but clearly the point is to describe properties/constraints and let the
mapping to specs be left as an exercise for the reader.