W3C home > Mailing lists > Public > public-ws-semann@w3.org > September 2006

what I would say to the WS-Mex folks

From: Eric Prud'hommeaux <eric@w3.org>
Date: Tue, 26 Sep 2006 15:14:27 +0200
To: SAWSDL public list <public-ws-semann@w3.org>
Message-ID: <20060926131427.GB5352@w3.org>

On Mon, Sep 25, 2006 at 06:38:25PM +0200, Jacek Kopecky wrote:
> ----------------------------------------
> 6. Relation of SAWSDL to WS-MEX
>  - Email from Eric [1]
>    - Seems like there is little relation
>  - How should we (the WG) handle comments on WS-MEX from Eric?
> 
> [1] http://lists.w3.org/Archives/Public/public-ws-semann/2006Sep/0014


3. Metadata Resources
"The representaiton of a metadata resource MAY be retrieved" does not
tell me what the server MUST do.

4. Web Services Metadata

Does @Dialect trump the metadata object? Broken down into the three
possible children of mex:Metadata, this question is phrased "Does the
semantics associated with mex;Metadata/@Dialect override the
semantics of"...:

... the endpoint-reference found at
    mex;Metadata/mex:MetadataSection/mex:MetadataReference? (Note: I
    don't know what in a Get response specifies the semantics.)

... the mimetype of the URI found at mex;Metadata/mex:MetadataSection/
    mex:Location/text()

... the "one child element" found at
    mex;Metadata/mex:MetadataSection/node()[node()!=mex:MetadataSection
    					AND node()!=mex:Location]?

Is each "distinct unit of metadata" in /mex:Metadata specifically
monotonic? That is, is the presence of one metadata unit specifically
prohibited from changing or retracting the interpretation of another
metadata unit? I propose "yes".

What is the use of /mex:Metadata/mex:MetadataSection/@Identifier given
that "the interpretation of identifier is Dialect-specific"? I believe
that any generic use of this attributte either imposes semantics on
the value of the attribute, or the use of it may be unfortunately
mislead by some Dialect. If there is not generic use of the attribute,
it should lie in the Dialect-specific content of the MetadataSection.

The references to "distinct unit of metadata" appear to be
self-referential. For example, "/max:Metadata ... contains one
Metadata Section child for each distinct unit of metadata" appears, in
the absense of some guidance like "for the purposes of this spec,
a metadata units is ..." in the Introduction  2.

5. WS-Transfer Get
How do the @Dialect and @Identifier end up in the ws addr?

6. Metadata in Endpoint References
In the first example, could the <wsp:Policy/> be embedded in the
<mex:Metadata>? Would that have the same semantics?

8. Security
 1 implies that the communication should, as a default, be secured.
Adding "when necessary" or something like that would clarify that any
securing should be down with ws-security.



MAY and MUST look like RFC2119 keywords, but RECOMMENDED comes from
someplace else.

There is not Conformance section. Error results of failed requeests
are not specified.
-- 
-eric

home-office: +1.617.395.1213 (usually 900-2300 CET)
	    +33.1.45.35.62.14
cell:       +33.6.73.84.87.26

(eric@w3.org)
Feel free to forward this message to any list for any purpose other than
email address distribution.
Received on Tuesday, 26 September 2006 13:13:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 17 April 2012 12:14:27 GMT