Re: Proposed Infoset Addendum to SOAP Messages with Attachments

On Wednesday, Mar 26, 2003, at 12:17 US/Eastern, Martin Gudgin wrote:
>>
>> (ii) the current formulation misses an important facet of included
>> attachments: their identity - i'll try to illustrate with a
>> (contrived) example.
> <SNIP>example</SNIP>
>>
>> I suspect that a constrained use of the swa:Representation
>> element might help with this problem - that is if there is
>> only 1 representation of a given resource in the MIME package
>> the swa:Representation/@href can be used. Alternatively an
>> application specific attachment reference mechanism could be
>> used but that kind of defeats the purpose.
>
> Certainly I was thinking that where identity was important
> swa:Representation would be used.
>
Do you have any specific thoughts on how this could be done, 
particularly in respect to solving the following:

> I think the root of this problem is that the semantic of the 
> xbinc:Include/@href is value rather than reference based so multiple 
> references to a single attachment result in multiple logical copies of 
> the content in the XML infoset rather than multiple references.

I think that solving the value vs reference problem is going to be key 
to having a coherent and rigorous model.

Regards,
Marc.

--
Marc Hadley <marc.hadley@sun.com>
Web Technologies and Standards, Sun Microsystems.

Received on Monday, 31 March 2003 13:41:19 UTC