Re: [topicmapmail] Re: [topicmaps-comment] back to the lists

[Murray Altheim <m.altheim@open.ac.uk>]

Daniel L. Koger wrote:

[...]
> *Mary Nishikawa 
> 
>>I wonder which tool you used for the conversion. The XTM DTD is not a 
>>complex one and doing a simple W3C schema generation with  a tools such
>> as XML Spy with some tweaking shouldn't take 3 hours!
> 
> <SOAPBOX type="religious">
> I tried to get a schema that would open in XMLSpy, TurboXML(Tibco) and
> Microsoft Visual Studio .NET.  This was actually a surprisingly
> difficult (tedious) task.  I ended up going to Textpad, making the file
> as generic as possible and noting the diffs and failures in the 3
> environments.  One would hope that a schema based on a standard would
> open and function in a similar manner on multiple applications.  That
> is, after all, the intent of a large amount of the work we as a
> community have been developing.
> </SOAPBOX>
> 
> Of course, it could be operator error :)


Well, my soapbox on this occurred a few years ago when Michael
Sperberg-McQueen asked me to publicly review XSD, and my impression
at the time (which I still believe) is that I don't think it very
likely that two or more implementations of XSD could possibly agree
with each other. XSD at its core is spaghetti. I didn't advocate
using it to describe XTM as a markup language, as schema languages
have different purposes and DTDs suffice quite nicely in describing
markup (as opposed to constraining instances of it).

Furthermore, because of the nature of topic maps when serialized,
there's no guarantee that XSD would correctly constrain an XTM
document, in fact I think it won't. For example, a topic could
be spread out among a number of <topic> elements, there are
topics implied in a topic map that may not exist in the
serialization at all, there may be structures of associations
that duplicate each other, but not in syntax, etc.

At first I was a bit dismayed at having Steve and Michel go off
and create a non-XML processing model, ignorant at the time (or
just stubborn) of the fact that topic map processing doesn't
happen at a lexical/syntactic level, the level at which XSD
functions. Constraint of topic maps must occur likewise at a
higher level than XML.

Murray

......................................................................
Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK

     "In Las Vegas Mr Gates also demonstrated a prototype
      fridge magnet which can be programmed to receive traffic
      reports, sports results and advertisements from local
      restaurants using the same FM signal as the wristwatch."
                                  -- The Guardian, 10 Jan 2003.