[MOBY-dev] JAX-WS BioMoby API (2nd round).

[Paul Gordon]

As I see it, it would be nice to use standard WSDL stuff as much as 
possible, but we will always need to do some manipulation in the 
standard stack because you simply can't express everything we need in 
XML Schema.  That's what Dmitry's running into. That being said, getting 
rid of Axis would be great, and doable now that Eddie defined a WSDL for 
Moby Central (which Dmitry's using).  I'd like to see an implementation 
of the Central interface that doesn't use Axis, for purely selfish 
reasons...it adds a lot of weight to the Seahawk applet :-)  Also, I get 
funny ClassFormatErrors when trying to use Axis in Java SE6 (version 6 
is much more pedantic about class file structure), but that could be 
overcome I suppose by recompiling the Axis from source...

Mark Wilkinson wrote:
>
>>> My vision is that all Moby Datatypes must have a schema representation.
>> As far as I can recall, this is not possible
>
>
> You are correct, Paul.  It is not possible to define the structure of 
> MOBY Objects in XML schema (for some of the same reasons that it is 
> impossible to define RDF-XML documents in XML Schema)
>
> There's a belief "out there" that all XML documents can be represented 
> in XML schema, and I've been attacked by audience members during 
> presentations and workshops when I say this isn't true... but it isn't 
> true!!
>
> The Moby XML is "defined" by the Moby Object ontology, much as RDF-XML 
> can be "defined by" an OWL ontology (please don't attack me for saying 
> it that way - I KNOW that this is backwards from the way OWL and RDF 
> actually work!).  The resulting flexibility in document structure 
> cannot be handled by XML Schema.
>
> It sucks for us, since we want to use WSDL... but it's a painful reality.
>
> Mark
>
> _______________________________________________
> MOBY-dev mailing list
> MOBY-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/moby-dev
>
> !DSPAM:60005,478e2823109501033230397!
>
>
>