Re: Use case comparison of REST & Web services

On Fri, Sep 20, 2002 at 11:35:23AM -0600, Champion, Mike wrote:
> OK, but there are some details that aren't described:  What's the exact
> syntax for the URI to specify the *specific* bits of information that my
> program needs (e.g., the team)?

In both cases I was assuming that the URI identified the team.

> What's the actual schema (broadly defined)
> of the result?

Follow xsi:schemaLocation, or the namespace as Ugo mentions.

This is the same for both cases.

>  How would my program find the specific information it needs
> in the response?

Also the same in both cases.

Perhaps that wasn't clear.  That XML document I described was being
returned in two different ways;

- because I have a {URI, WSDL}, and I know which WSDL port type to use, or
- because I have a URI, and I implicitly know to use GET

But in both cases, in my example, the same XML document is returned.

Just trying to compare apples-to-apples.

>  Humans can do the "really really late binding" to figure
> this out, but I still assert that this is typically bound up in application
> code ... hence the appeal of SOAP and WSDL to allow machines to do a later
> binding.
> 
> I don't particularly expect Mark to agree :-)

I'm shocked, shocked I tell ya! 8-)

> but I think this could be
> useful (perhaps in the Primer) to explain the rationale for SOAP and WSDL in
> the context of the Web As We Know It (which may or may not be RESTful, I'm
> agnostic).

I don't expect this conversation to yield much for the architecture
document, or for the primer for that matter.  I consider it a side-bar
that I'd be happy to take offline so we can talk about the architecture
document (did you see the comments I forwarded?)

I'd like the architecture document and primer to focus on SOAP 1.1 and
WSDL 1.1 for its first draft, and completely ignore REST/Web issues.
Let's let TimBL and TAG worry about that.

> > The Content-Type header in the response.  I think we can all agree
> > that application/xml or application/soap+xml will be common values.
> 
> Sure, but XML is only a *meta* syntax

Ok, wasn't sure what you meant by "format".  Check the namespace, in
that case.

> > If you wrote some code that knew about teams, games, scores, etc.. all
> > the standardized types that the sports industry decides upon 
> > (and their
> > corresponding XML schemas), then you could write a program which
> > navigated through that information as required.
> 
> OK, sure, in some ideal world where industries can agree on simple,
> interoperable formats for this kind of thing.  Somebody slap me around a bit
> if I'm totally wrong here, but I see WSDL in particular as a way around the
> dilemma of needing a standardized schema for some business process before
> integration can take place.  With WSDL, you don't need industry-wide
> standards, you just need to publish (or let partners discover) the WSDL
> description of how someone should send messages to YOU. 

Sure, but you're just trading one hard problem (schema agreement) for
another (API agreement).  And since I said I wouldn't mention RDF, I
can't get into why I believe that it solves the schema agreement
problem. 8-)  And AFAIK, there is no solution to the API agreement
problem.

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 14:56:03 UTC