W3C home > Mailing lists > Public > www-ws-desc@w3.org > April 2005

Re: Contradictions regarding transitivity of wsdl:import

From: Ryan Betts <rbetts@gweep.net>
Date: Tue, 19 Apr 2005 14:37:02 -0400
To: www-ws-desc@w3.org
Message-ID: <20050419183702.GA80551@sidehack.sat.gweep.net>

>From Roberto Chinnici's mail on Tue, Apr 19, 2005 at 10:35:57AM -0700:

> Besides the confusion caused by the proposed name change, I find option
> 2 in Arthur's email a lot more appealing. Since David Booth called it
> "a considerably clearer and more straightforward way to go", I would
> suggest that examine it more carefully. My impression is that by
> aligning the description component with the intuitive concept of a
> description document, it will make things easier to grok for users.
> The issues around component equivalence don't seem unsolvable.

As a user attempting to grok, I think section 2. Component Model [1]
makes it clear that there is a single component model that contains
all components including those from external infosets.

Creating a component named "component model" would be quite confusing.

Specifying the component model as a directed graph of top level
components AND providing rules to serialize that graph into WSDL
documents / XML Infosets (which [1] explicitely avoids doing?)  and
which Mr. Gudgin implies is feasible [2] would be a very useful
change, even if that serialization was only an algorithm to walk the
directed graph.

Then those of us trying to understand and use import, include and
extend can work at the component model knowing an interoperable
serialization exists.

[1] http://www.w3.org/TR/wsdl20/#component_model
[2] http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0092.html

Ryan Betts
rbetts@datapower.com
Received on Tuesday, 19 April 2005 18:37:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:35 GMT