Re: Another REST challenge

Francis McCabe wrote:
> 
> Here is a second use case, which I think also represents a challenge to
> REST. In this case, not so much that you could not do it is REST, 

Wait, do you accept that you could "do" your other case in REST?

>...
> The point behind this use case is that it is in H's interest to be able
> to offer timely deals to P in order to maximize its potential. However,
> it is fairly unlikely that H can predict (at design time) all the
> possible offers and even more unlikely that P can be built to ask for
> ski packages, even though there IS a possibility that P could make use
> of such an offer.

As you describe it, your problem sounds like an artificial intelligence
problem quite unrelated to issues of protocol. Insofar as REST
discipline is highly compatible with the Semantic Web technologies (and
was explicitly designed to support them), I would look for your solution
there. If you can clarify the semantic issues involved in having two
computers negotiate about "ski trips" then REST could easily handle the
protocol issues. i.e. if you come up with a DAML schema that makes all
of the questions and answers *possible* then I can show you how to
compose the questions and answers into dynamically assembled HTTP
resources.

I would guess that REST+SemWeb technologies probably stand a much better
chance of tackling the issues than other approaches (CORBA, Messaging
Software, EDI, etc.)

-- 
Come discuss XML and REST web services at:
  Open Source Conference: July 22-26, 2002, conferences.oreillynet.com
  Extreme Markup: Aug 4-9, 2002,  www.extrememarkup.com/extreme/

Received on Wednesday, 17 July 2002 19:14:56 UTC