[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: RDDL and WSDL

• To: "Simon St.Laurent" <simonstl@s...>,"Xml-Dev@Lists. Xml. Org (E-mail)" <xml-dev@l...>
• Subject: RE: RDDL and WSDL
• From: "Chris Wilper" <cwilper@c...>
• Date: Thu, 30 Jan 2003 16:39:37 -0500
• Thread-index: AcLImNPg4PPhI43CSQuBu/ZpvNcmaAABTe8g
• Thread-topic: RDDL and WSDL

> Is answering Simon Fell's original question about roles [1] enough to
> deal with this, or is something more powerful needed?  

My answer to his question is: I wouldn't try to 
come up with a "nature" identifier indicating somehow what's 
inside a compound structure.  I'd just identify the containing 
structure(s) (http://schemas.xmlsoap.org/wsdl/) and let the client 
do more discovery work from there.  

I don't really like using a namespace identifier to denote 
structure, but it's pragmatic.  In fact, the piece that is 
missing, that I believe could make this resource discovery 
thing work in XML land, is a *way* to declare the supposed 
"form" or structure of a given piece of data inside a compound 
structure.  It's tempting to want to assume we can already 
infer the form by looking at an element's namespace uri...  
yet no one is (officially) willing to say that xml namespaces
are more than a collection of names... no semantics, no structure,
just tokens. If that's the case, I vote for an xml:struct 
element for declaring forms, ala:

<ns1:containing>
  <ns2:contained xml:struct="http://xyzstruct.com/v1">
    <ns2:assert>A QName does not assert structure</ns2:assert>
  </ns2:contained>
</ns1:containing>

- Chris

-----Original Message-----
From: Simon St.Laurent [mailto:simonstl@s...]
Sent: Thursday, January 30, 2003 2:51 PM
To: Xml-Dev@Lists. Xml. Org (E-mail)
Subject: RE:  RDDL and WSDL


cwilper@c... (Chris Wilper) writes:
>The concept is an API.  If it has a URI, great.  If it can be
>resolved, excellent.  If I get a choice as to what representation of
>the resource I want, even better. 
>
>I'm not sure RDDL is the right thing, as currently scoped
>as a "namespace document" -- but the idea of a resolved URI 
>providing an index of descriptive representations of
>a concept/resource ... is appealing.

Interesting, though I'll admit that I cringe any time people map XML
vocabularies directly to APIs.

Is answering Simon Fell's original question about roles [1] enough to
deal with this, or is something more powerful needed?  

I guess I can see RDDL pointing to "here's how to process this stuff"
via WSDL, but worry that WSDL itself specifies something larger than
RDDL, as most Web Services I've seen involve interactions between
multiple (and frequently more than 2) namespaces.

[1] - http://lists.xml.org/archives/xml-dev/200103/msg00171.html

-- 
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!
http://simonstl.com -- http://monasticxml.org

-----------------------------------------------------------------
The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
initiative of OASIS <http://www.oasis-open.org>

The list archives are at http://lists.xml.org/archives/xml-dev/

To subscribe or unsubscribe from this list use the subscription
manager: <http://lists.xml.org/ob/adm.pl>

• Prev by Date: RE: catch (was Re: RDDL and WSDL)
• Next by Date: Re: Roles and functions [was: many-to-many]
• Previous by thread: RE: RDDL and WSDL
• Next by thread: RE: RDDL and WSDL
• Index(es):
  • Date
  • Thread