Object-oriented serialization (Was Re: Some questions)

[Dan Brickley]

On Mon, 6 Dec 1999, Andrew Layman wrote:

> Dan Brickley wrote:
> > I believe it will be possible to annotate XML schemas with
> > information
> > for mapping into (generic or domain specific) application datamodels
> > such as RDF. I don't think it is right to expect the hard-pressed XML
> > Schema group to define all these mappings within that working group.
> 
> I agree.  There are probably many ways to express mappings. One candidate is
> shown at the end of the "Schemas NG" paper.  See
> http://www.lindamann.com/xml/XML%20Schemas%20NG%20Guide%20HTML.htm, and look
> for the section titled "Mapping to Other Data Models."

This is interesting work, though it's unclear quite how it fits in with
the Canonical Format / Serializing Graphs paper. The 'Mapping to Other
Data Models' section of 'Schemas NG' shows one strategy for annotating
schemas to support directed labelled graph interchange in XML. It would
be good to see these two strategies drawn together in a single document
describing objects'n'properties DLG serialization strategies for XML
applications. By drawn together I mean having a common documented
model for the DLG representation rather than informal prose.

It is clear by now that the RDF 1.0 Syntax doesn't cut it as the One
True Graph Serialization for all XML applications. I don't think anybody
expected otherwise, but we now have general consensus [eg. 1] that a
more broadly usable DLG exchange syntax is needed by RDF apps. 
We have two proposals already floated on the RDF Interest
Group for alternate DLG-interchange syntaxes [2, 3] and their aims seem
to be basically the same as [4,5]: DLG interchange in XML.
It is also clear that a lot of (RDF-agnostic) XML data interchange apps
want to ship directed labelled graphs around using non-stilted XML
syntaxes. I've argued elsewhere [7] that these graphs will often want to
use URIs for edge types, node identifiers and node types in all but
tightly-coupled closed environments. 

My hope is that XML-DEV and the RDF Interest Group[6] will come up with
implementation-led proposals for XML DLG-interchange that both 
complement the XML Schema work (for mapping-based proposals) and fit
with colloquial (ie. mainstream) XML conventions (for serialisation syntaxes).
  
There's a bunch of interest in an improved syntax for RDF graph
serialization, and growing interest in more general XML DLG interchange
strategies layered on top of XML + XML Schemas. I have a hard time
thinking of these as different problems, hence my wish that the DLG
model mentioned in the schemas NG and canonical papers be documented a
bit more formally to aid comparison with similar proposals for a better 
RDF syntax...

Dan



Refs:
[1] http://www.w3.org/TR/1999/NOTE-schema-arch-19991007 (s3.8)
[2] http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0066.html
[3] http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0100.html
[4] http://www.biztalk.org/Resources/canonical.asp
[5] http://www.lindamann.com/xml/XML%20Schemas%20NG%20Guide%20HTML.htm#_ftn4
[6] http://www.w3.org/RDF/Interest/
[7] http://www.lists.ic.ac.uk/hypermail/xml-dev/xml-dev-Dec-1999/0121.html


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)