On 11/22/07, Sanjiva Weerawarana <[EMAIL PROTECTED]> wrote:
> Mark Baker wrote:
> >
> > It's not a myth, RESTful systems do not need WSDL.  It doesn't need
> > any description languages in fact.  It needs forms: information sent
>
> I think WADL folks, Sam Ruby and Leonard Richardson all disagree with you
> on that.

Well, all I can say is that I've built a whole lot of Web based
systems, and I've never needed it nor anything like it.  That doesn't
mean you can't *use* it, as Sam and Leonard did, but I think it's
strong evidence that it's not *needed*.

> > I don't see what security & caching have to do with scalability.  Are
> > you talking about REST or HTTPS?
>
> Caching is a key part of the scalability architecture of HTTP. However,
> that only works for HTTP and not HTTPS .. right?

Sure.  But that's only because SSL/TLS doesn't (AFAIK) permit trusted
intermediaries to be integrated into the message path.  So your
criticism is really of the current state of specifications rather than
of REST itself by the looks of it.

> > I don't follow the "Lie: RESTful services have a uniform interface"
> > argument, but it's prima facie incorrect because they do by
> > definition.
>
> Did you look at the next page? Note I said RESTful *services* .. the URIs
> of the service are most certainly part of the *interface* of the service.
> Thereby, every RESTful *service* has its own custom interface. WADL
> appears to agree 100% as that's what WADL is describing.

The URIs aren't part of the interface, they're an argument, just as
"MSFT" would be an argument to getStockQuote().

Remember, the interface isn't to individual resources, it's to the
software component, i.e. the Web server.

> I hope we don't drop down to a REST vs. WS-* religious war again on this!

Me too.

Mark.
-- 
Mark Baker.  Ottawa, Ontario, CANADA.         http://www.markbaker.ca
Coactus; Web-inspired integration strategies  http://www.coactus.com

Reply via email to