RE: REST and CRUD (was Re: WSA architectural concepts and relationsihips related to WS, SOA, and the Web

Well, something else we can agree to disagree on I guess, I'd say CRUD as a concept includes the equivalent of HTTP POST, since that's one way to accomplish a C.

Eric

-----Original Message-----
From: Baker, Mark 
Sent: Wednesday, May 07, 2003 2:51 PM
To: Narahari, Sateesh
Cc: www-ws-arch@w3.org
Subject: Re: REST and CRUD (was Re: WSA architectural concepts and
relationsihips related to WS, SOA, and the Web



On Wed, May 07, 2003 at 11:25:03AM -0600, Narahari, Sateesh wrote:
> >>Not really, no.  CRUD doesn't have the equivalent of HTTP POST which
> is the workhorse uniform semantic that enables 
> >>computation beyond data manipulation.  For example, POSTing a
> representation of an order to an order processor for the 
> >>purposes of ordering something, isn't representable in CRUD.
> 
> That's easy( much easier & elegant than REST). Create Order( order
> representation).
>
> If you are more concerned about the process of fulfilling the order,

Yes, that's what I was talking about.

> it
> can easily be achieved with triggers.

Which isn't CRUD.

Anyhow, that's beside the point.  CRUD semantics are not uniform and
therefore not REST.  I agree that CRUD, or tuple spaces, or any of
a number of coordination languages make for good comparisons of the
"Constrained interface" approach to SOAs.  But saying REST is CRUD
(for multiple interpretations of "CRUD" 8-) is incorrect.

I'll respond to Mike's 1.6.2 stuff shortly ...

MB
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Web architecture consulting, technical reports, evaluation & analysis

Received on Thursday, 8 May 2003 10:23:51 UTC