xml-dev - Re: scalable DTDs.

[Paul Tchistopolskii <paul@qub.com>]

----- Original Message ----- 
From: Marcus Carr <mrc@allette.com.au>

> Sorry, it's in UAT at the moment, so isn't available for viewing. Even if it
> was, all you'd see is a website with a whole lot of forms - there aren't any
> hints to the underlying data structures. It's only a conceptual design issue,
> or am I missing what you're looking for?
> 
> >  From your letter I don't even understand are you talking about DTDs or
> > Schema.
> 
> It shouldn't really matter, should it? It was designed against DTDs, but will
> soon be migrated to schemas.

Well ... 

1. I was under impression that expressive power of schemas is 
greater than expressive power of DTDs. If this is right - it think 
it *should* really matter.

2. My understading is that both schema and DTDs are *not* 
giving ideal support for 'base architecture' mentioned in original 
letter.

3.  (1) and (2) are of course realive. One can implement some 
base architecture in, say, m4. I mean - #include  this and that, 
#define this and that and call it 'scalable'.  Like people did in 
C - before C++. 

I guess scalable DTD's are based on defining / redefining entities, 
right?  It is like #defines juggling, right ? 

That's why I want to see actual DTDs. I'm wondering how convinient 
is this trick in the real-life. I think #define's juggling was not convinient 
for complex cases. That's why we don't need #defines in C++ .

Rgds.Paul.