On Sun, 2002-04-28 at 01:24, John F Schlesinger wrote:
> "The relational database people, whatever their other sins, did an
> excellent job expressing things like normalization approaches."
>
> But you should remember that the reason for normalization was to
> eliminate update, insert and delete anomalies. Do XML documents have
> such a principle? If not, what is their equivalent principle? Does it
> even make sense to look for such a principle if you don't know what the
> mark-up is for?
Hmmm... I'm thinking that the relational people focused more explicitly
on these issues because the costs of answering those questions wrong was
much more drastic. Bad table design and interactions would make
relational databases downright impossible, especially early on.
With XML, any single program can pretty much duck those issues, and
interchange and reuse when it matters. XML's flexibility makes it
possible to take a number of different answers on the interchange angle,
and easy reuse doesn't seem to be an exciting enough problem for
developers to notice.
SOAP enveloped-messages, however much a mash I may find them, are still
a lot easy to reuse than the binary formats my wife uses for embroidery.
This gives me considerably less hope that developers will get their act
together. The funny thing is that I think Web developers (those who
touch markup, not just GUIs), with all their supposed laxness and
aesthetic concerns, would likely do a better job developing these
formats than developers who are fundamentally programmers and have
little patience for the niceties of how information is stored.
Or maybe it's just that I come from a background where people actually
work with markup rather than treating it as a black box.
--
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!
http://simonstl.com
