xml:id RE: [sc34wg3] Compact syntax requirement question

[Mason, James David (MXM)]

To reply to a few comments from Robert Barta:

When I'm speaking of syntax-directed editors, I mean something like =
ArborText
or XMetaL, which can be cusomized as an editing environment so that the =
end
user never sees chicken lips or the other syntactic mess that we put =
into
XML. And so I'm not talking about the raw-XML mode of XMetaL or XMLSpy =
or
oXygen.

Robert says that CTM-like things are for "engineers out there who very =
easily
pick up languages". I have no complaint with that. But if topic maps are =
ever
to get outside the geek community of people who like learning languages =
and
messing with the neat internals of things, then we've got to get beyond
picking up languages. (I like picking up languages. My doctorate is =
basically
in comparative and historical linguistics. I took every Germanic =
language my
graduate school offered. But I consider myself a wierdo: just ask my =
wife,
who was dreaming in Old Norse by the time she finished typing my
dissertation.)

The end users, if we're successful, are going to be people who don't =
even
know that there is such a thing as a topic map. For these users, custom
editors make lots of sense. I think BrainBank, built from OKS, is a good
example of the sort of things the majority of creators of TMs will need. =
None
of my clients has ever seen XTM data (except perhaps as a horror-show =
slide
in a PowerPoint presentation); they always approach the TM through a =
custom
interface (though I admit mine are not nearly so slick as BrainBank).

I currently edit XTM in oXygen, which I was glad to pay for because it's =
a
good IDE for XSLT. It's fine for me; I've been looking at raw markup of =
one
sort or another nearly half my life. (I actually generate much of my XTM
through XSLT, and I need to look at the raw results to make sure I've =
done
the right things. You know the drill.) I'd never wish that on the folks =
for
whom I'm developing TMs. They'd never use it. But they'd never write LTM =
in
Notepad, either. Too much syntax to remember.

The overwhelming experience of the SGML/XML industry, coming out of more =
than
20 years' experience, is that tools like ArborText or FrameMaker are =
critical
to success. I've used practically all of them, starting with the =
Datalogics
editor that came to market even before SGML was finalized (I wrote my =
report
to ISO announcing that this committee had completed the formal =
requirements
for finalization in the Datalogics tool back in 1985, thanks to Pam =
Gennusa).

Let's face it. We're building these things for ourselves, and they're
proliferating because we have fun doing it. We need to think very hard =
about
how many of them are pushed for ISO standards. I had my doubts about =
RELAX NG
Compact, but it came in from another SDO with which we cooperate, so we =
might
as well take it. It can be argued that TMs aren't really an XML =
application:
Michel can tell you that I had my doubts about taking them in, too. They
wound up in SC34 because they looked interesting, there were some people
ready to take them on, and they were supposed to be something made out =
of
HyTime, which was supposed to be related to SGML, which we were =
responsible
for. (And the origin of HyTime is another story, which Newcomb can tell =
you
about, but it's another case of a hobby becoming a standard because it =
was
fun to do, and I did fight very hard to get that one into the =
committee.)

I keep a very skeptical eye on all the things we're devloping in WG3. =
I'm
particularly skeptical of claims that gobs of people are developing TMs =
in
LTM, etc. Sure, people are doing it, but they're not enough of a =
community to
justify the existence of SC34. I know some of the people, and most of =
the
ones I know are techno-nuts, including my colleague who taught his
12-year-old geek son to do a TM of his Pokemon cards in LTM. If we've =
really
got to have a new syntacic structure for technical reasons, then let's =
do it.
But let's avoid saying we're doing it for the end users. They won't buy =
it.

Jim Mason