[sc34wg3] Quietly Storming Out

Murray Altheim sc34wg3@isotopicmaps.org
Sun, 16 Nov 2003 21:51:13 +0000

Martin Bryan wrote:
> Having been away looking at the next generation of applications with the
> European Commission I have not been able to add my voice to the debate this
> week, otherwise I would have had time to get very hot under the collar by
> now :-)


I tend to also get very hot under the collar when people pile on
the existing discussion without adding anything substantive except
another "me too". As I said before, what is good for Topic Maps and
was is good for the world is beyond this small group's ability to
know. We all do know how bad arbitrary markup is; it's not "the
state of the art", it's simply a bad design. And citing "emergency
situations" only sounds like a politician on a stump. I don't need
to hear of eschatological scenarios, they're not remotely convincing.
If a city somewhere establishes emergency services that are reliant
on bad architectural ideas like arbitrary markup, who's to stop them?
But since you were the one to inject the scary scenario element in
this conversation, are you willing to be responsible for any *deaths*
due to XTM documents that can't be correctly processed because the
arbitrary markup wasn't interpretable? At least with PCDATA nothing
is lost, and with links you have at least the URI. If you can't
predict the markup, you have no guarantee that anything will work.
And you want to *standardize* that? It's a nonsequitor.

Almost all of the supposedly technical arguments I've heard apply to
specific applications of Topic Maps where there is no plan or even
possibility (legally or technically) for interchange. Jim has said
as much -- he's not even allowed to *talk* about some of his stuff,
much less share it outside of his domain. Eric has a proprietary
LexisNexis application in mind. Your "emergency situation" scenario
is certainly going to be a specialized, custom application (it sure
had better be if lives are at stake).

Everything I've heard so far has been about proprietary topic map
systems, not interchange between them. My argument for XTM has been
a conservative one (the normal approach to standardization, not the
one taken by the W3C) about preserving the possibility of interchange.
Nothing stops anyone from creating proprietary Topic Map applications
right now, and by exporting to XTM 1.0 there is a guaranteed level of
interchange. Yes, there is a loss. How could there not be between
LexisNexis and your "emergency" system?  What you're all arguing for
is simply the false belief that you'll be able to miraculously
interchange arbitrary markup between systems, a fool's errand if I
ever heard one. I'd certainly not want to be reliant on any system
that relied on that in an emergency:

   "...the location of the nuclear weapon is <element 'loc' undefined>..."

which points out how silly any such scenarios are -- who'd ever put
themselves in such a situation? Who'd be so irresponsible as to develop
a technology that would allow them to? What's surprising to me is that
you guys all know this. I should be preaching to the choir here. Isn't
this a standards group? (and yes, Martin, you are inconsistent. You
either argue for the ability to validate or you don't. We're not talking
about specific applications, XSLT, whatever, we're talking about
removing the validity guarantee in the general case. RELAX-NG isn't
going to buy you that if your schema has a deliberate hole in it.)

All you'll do is destroy XTM as a existing simple and safe interchange
format for Topic Maps. What you'll have with your new version will be
documents that will provide few if any guarantees of validity, and no
guarantees about interchange, as they'll be completely unpredictable
(i.e., useless) outside the context of their originating environment.
Every vendor and every installed TM application will have their own
proprietary way of doing things and nothing will be reliably
interchangeable. But you'll all be able to say you conform to the ISO

Put it simply: if you can provide a cogent argument that shows that
Eric's, Jim's, your "emergency situation" application could all
safely and unambiguously (i.e., miraculously) interchange arbitrary
markup, I'd be convinced. I note that the W3C activity for exchanging
XML fragments is currently inactive. Anything less than what Paul was
trying to accomplish with that will not work. With XTM 1.0 you can at
least interchange at the level that the TopicMaps.Org group had in
mind. I like to think that was a shared vision:  minimal, but
functional. You're going to lose that the moment you open up the
namespace muckgates.

I have no vote in ISO or ability to do anything other than express
my opinion, and apparently nobody is interested in that so there's
little point in me continuing. Since I've made my case repeatedly
and the vocal majority of the list members here seem predisposed to
make these changes to XTM, let me at this moment state my protest by
resigning any active involvement in this list.

Martin, I seem to remember you loudly storming out of the group because
you disagreed rather violently with the direction it was heading, so
perhaps it's appropriate I should leave with a message in response to
you. I've fought long and hard to keep this muck from infecting XTM, but
you all all seem pretty convinced it's where you're heading. I'll say
goodbye with a little less noise, and very sorry to see a good thing --
something into which I personally and professionally have invested an
enormous amount of time and energy -- ruined, turned into yet another
namespace muckfest.


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

      Q: So exactly how is Ahmad Chalabi different from Manuel Noriega?


      A: One speaks fluent Arabic, the other Spanish.

        "Noriega took refuge in the Vatican embassy, where US troops played
         hard rock music until Noriega surrendered on January 3, 1990."