Re: WSA diffs from REST

On Thu, Sep 19, 2002 at 06:20:33PM -0600, Champion, Mike wrote:
> I for one am more interested in building bridges betweeen the "Web" and
> "Web Services" worlds than in highlighting their differences.

I would say that those are not mutually exclusive.  On the contrary,
I consider doing a comparison between them as being a necessary
precursor to building bridges.  Otherwise, how do you know what
you're building bridges between?

>   As 
> far as I can see, "the Web" is rich enough to support both  RESTful
> and  distributed object design patterns.

IMO, REST is "distributed objects done right", where "right" is in the
context of Internet-scale systems (multiple administrative domains,
unreliable network, etc..). 8-)

>  It's helpful to note, 
> as Mark is doing, how these design patterns differ and
> perhaps to suggest where one or the other is most useful.  It's not
> EVEN REMOTELY helpful to imply that one is more Web-ically correct than the
> other at this stage of the game.  We've had this debate, the issues
> are out in the open, let's not go there again.  

In what I wrote, I never suggested that.  I was just examining
SOAP+WSDL and how its architectural constraints compared with REST's.
No judgment was made about being "Web-ically correct".

Does anybody here believe that the WSA is constrained to be stateless?
Or constrained to have a uniform interface?  Or constrained to indicate
cacheability implicitly or explicitly?  If so, please provide a
reference.  If not, what's the problem with writing this down?  Or, what's
to be gained by not writing it down?  I don't understand.

> Yup. If the TAG wants to give us authoritative guidance that "the Web" ==
> "REST" and web services must live within the constraints of REST, so 
> be it.  Until that (PROFOUNDLY unlikely, IMHO, event), I would much prefer
> to
> discuss where HTTP and URIs and XML *do* fit into the reference architecture
> we've sketched out.  Let's learn from the Web as it exists and let the TAG
> worry about the extent to which REST is an authoritative description of 
> the Web architecture. 

Roy and I have described how this works on several occasions.  It
concerns me that misunderstandings about this important issue continue
despite those explanations.  The best summation is, IMO;

"REST is an architectural style that models system behavior for
network-based applications.  When an application on the Web is
implemented according to that style, it inherits the interconnectivity
characteristics already present in the Web.  REST's purpose is
to describe the characteristics of the Web such that they can be
used to maximum advantage -- it does not need to define them.
REST isn't supposed to be a baseball bat; it is supposed to be a
guide to understanding the design trade-offs, why they were made,
and what properties we lose when an implementation violates them."
 -- http://lists.w3.org/Archives/Public/www-tag/2002Apr/0303

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com

Received on Friday, 20 September 2002 00:06:22 UTC