Re: Remember: namespaces break DTD's

keshlam@us.ibm.com wrote:

> >To me, there is a compelling argument for  associating descriptions of
> some
> >flavor, be it DTDs or schemata, with namespaces: namespaces as they stand
> >break DTD's.
>
> Completely separate issue.
>
> The fact that namespaces and DTD validation are (at best) imperfectly
> compatable is certainly an annoyance. But to perform namespace-aware
> validation you need a namespace-aware validation language, and DTDs just
> ain't it.

That last sentence is a reasonable paraphrase and elaboration of ``namespaces
break DTDs''.

I think there was another path the namespace spec could have gone, but it
would have required extensions that aren't in SGML.   If the entity
declaration that introduces a DTD into a document could specify a namespace
prefix, then everything in that entity declaration, aside from names that are
already prefixed, could implicitly have that prefix attached to it.

My reading of the situation -- and if yours is different, I'd like to hear it
-- is that in essence the W3C XML community sees DTD's as a historical
anachronism.  They have to be supported because there are a lot of them
around, but schemata are the way the future is going.  For that reason,
people in W3C are not willing to devote any energy to repairing anything that
might be wrong with the use of DTDs in XML, including the problems of using
them in conjunction with namespaces.

Paul Abrahams

Received on Monday, 22 May 2000 11:06:25 UTC