Re: Proposal re REST and Arch doc

On Tue, Aug 20, 2002 at 01:15:21PM -0600, Champion, Mike wrote:
> I think it's controversial because it's not at all clear what the "Web
> Architecture" is.  The most casual reading of the TAG list should disabuse
> anyone who thinks that TimBL and Roy Fielding have it clear in their minds,
> agree with each other, etc.  I'd prefer to factor out the commonalities, to
> try to bridge the gaps between Web/REST and OMA/distributed objects.

There's most definitely strong differences of opinion about some fringe
(though important) issues between the "gurus".  So let's avoid those,
and just focus on the straightforward ones.  For example, I suggest it
would be easy for us to agree to say something like this;

  "REST emphasizes generic connector semantics, whereas the Web
   Services Architecture supports both generic and service-specific
   connector semantics."

(at least that's what our requirements doc currently suggests)

> Maybe some non-normative Note or Appendix on how what we ultimately come up
> with differs from REST would be useful. But it has to, IMHO, be called
> "REST" not "THE Web Architecture."  Reasonable people can disagree on how
> closely REST theory matches Web practice.

Sure, that's fine with me.

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               [email protected]
http://www.markbaker.ca        http://www.idokorro.com