xml-dev - Re: Options in XML 1.0

[Rick JELLIFFE <ricko@geotempo.com>]

Jonathan Borden wrote:

> But my concern is that the proposed XML Schema spec is much more
> complex than XML 1.0 itself including DTDs. So, I don't grok how suggesting
> that we drop DTDs is going to solve any problem in this regards.
 
Because writing parsers is fun?

Cheers
Rick Jelliffe

P.S. There have been at least three members of the XML Schemas WG who
have severely criticized it at various times: me, Murata-san, and now
Jonathan.  I view it like a bushfire: back in Australia the firefighters
don't try to stop a large fire heading towards a town, they steer it in
a least harmful direction.  And the results of the fire is often good. 
I am happy to have XML Schemas as a brilliant language for specifying
automatically-generated interfaces to Java and to provide fodder for
query optimisers: database vendors need something like that to make and
sell high-transaction-rate fine-grain query systems (of course, it is
perhaps an underemphasized point that XML lends itself to coarse queries
which are transformed by some subsequent light-weight system, and that
this is an appealing architecture that to some extent obviates the need
for highly optimised fine-grain queries in the first place!)