Re: [topicmaps-comment] RE: OASIS vs W3C

[Murray Altheim <murray.altheim@sun.com>]

"Tsao, Scott" wrote:
> 
> From my perspective ("personal knothole") I would like to see these
> types of collaboration:
> 
> (1) W3C/ISO
>     Harmonization of core standards for RDF and XTM, between W3C 
>     RDF Core WG and ISO 13250 WG.

This is the one I've had a problem with all along. Either there is
a misunderstanding (from not just you but many people) about the
essential nature of RDF and XTM or people are simply trying to 
get rid of one or the other. They're apples and oranges, no, they're
apples and meatball pie, or apples and eyeglasses. What harmonization
can there be between RDF (graphs-in-XML) and XTM (subject-based
mapping technology)? Certainly it's possible to create an RDF-based
syntax for XTM, but you could also create any number of other syntax
representations. There could be a binary TM standard for all we know
(which might be valuable for passing around very large or pre-processed
topic maps).

There is no essential advantage in having XTM be RDF-based, and given 
that even hard-core RDF fanatics don't much like its syntax, this 
doesn't make a lot of sense. XTM's syntax was designed from the 
ground up to represent topic map semantics in XML markup. You can't 
improve on that (in XML). Any advantages in RDF tools are mitigated 
by the lack of RDF tools vs. XML tools and frameworks, ie., there's 
*loads* of XML processing software out there. Besides, XTM processing 
requirements mean that specialized engines must be used, and these 
engines must import XTM syntax (and probably ISO 13250 syntax as well),
so adding some RDF representation would add only significant 
complexity (and you'd need an RDF schema validator just to be sure
the document made syntactic sense). Every production-quality XML 
parser out there can already validate XTM documents, so what's to
gain?

I keep hearing this argument but it just never makes sense to me.
The only "harmonization" I see is perhaps a document describing
how XTM could map RDF content (as according to a *specific* RDF
Schema, not RDF in general which is impossible), or vice-versa. 
We can already do this but a formalization might be valuable. 
This is pretty simple stuff, and could possibly be considered
within the scope of the OASIS Published Subject TC.

> (2) W3C/OASIS
>     Joint development of architectural framework and reference models 
>     for semantic-web applications, starting with Tim B-L's diagram 
>     (one layer at a time)
>     (http://www.w3.org/2000/Talks/1206-xml2k-tbl/slide10-0.html),
>     between W3C Semantic Web CG (including WebOnt WG) and the applicable
>     OASIS TC's.

I don't know of any OASIS TCs that are applicable. This also precludes
the idea that there is one global architecture, which would assume 
that there is one global community of interest. It's a windmill, IMO.
I don't see it's pragmatically or politically possible to harmonize
the CG, XTM, IEEE, ISO (and probably a half dozen other) "semantic
web" related activities under one roof. The Semantic Web conference
last month certainly told me one thing: that there's so much work
out there in related, overlapping, but varied communities, that we
can hold conferences to gain cross-polination benefits, but its 
probably *impossible* to have everyone work from a common data model
or architecture. I'm also not so certain that Tim BL's diagram is
really the place to start. There's many years of research from people
whose lives have been devoted on this subject that would take a much
higher priority, were I running the show... (thank God I'm not) 

> (3) OASIS/ISO
>     Joint development of "usage patterns" and "best practices" for 
>     topic-map applications (one pattern at a time), between OASIS 
>     TopicMaps.Org member section (including the applicable OASIS 
>     TC's) and ISO 13250 WG.

Now this sounds like an excellent idea, and one that was part of the
TopicMaps.Org work (we had a subgroup devoted to use cases). Earlier
in the summer I began to compile a set of Best Practices documents
but realized that (a) it's too much for one person, and (b) it really
*should* be a community activity, probably backed up by a topic map,
ie., online. If the OASIS TopicMaps.Org member section doesn't itself
take on this task, it could be handled by a subcommittee/TC.

Murray

...........................................................................
Murray Altheim, SGML/XML Grease Monkey  <mailto:murray.altheim&#64;sun.com>
XML Technology Center
Sun Microsystems, 1601 Willow Rd., MS UMPK17-102, Menlo Park, CA 94025

     i am going to see if i cannot reform insects in general
     i have constituted myself a missionary extraordinary 
     and minister plenipotentiary and entomological to bring
     idealism to the little struggling brothers -- archy (1927)