[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: Come On, DTD, Come On! Thoughts on DSDL Part 9

• To: 'Marcus Carr' <mcarr@a...>, xml-dev@l...
• Subject: RE: Come On, DTD, Come On! Thoughts on DSDL Part 9
• From: "Bullard, Claude L (Len)" <clbullar@i...>
• Date: Fri, 14 Jun 2002 08:30:15 -0500

Total agreement.  You may remember that we tried to 
get rid of them in the original XML WG/SIG and had 
to contend with the issue of DTD maintenance and 
the baffling books that suggest that treating them 
as a conceptual design tool is a good thing for 
very large DTDs.  It was countered that perhaps 
very large DTDs are the real problem and that 
those advocating DTD design methods as a verisimilitude 
of object-oriented design needed to do more 
object-oriented design so they would understand 
it better as software design and not a handy metaphor.

len

From: Marcus Carr [mailto:mcarr@a...]

The only reason I'd have for keeping them is to allow a DTD to be broken into
separate files. As long as that was catered for somehow, I'd be happy to see them
die.

I could do without the ability to group element and attribute declarations too -
like PEs, I think they make DTDs more difficult to read, particularly if you're
looking for a particular element declaration. I'd rather search for the string
"<!ELEMENT foo" than sift through all the places that foo gets mentioned in the
DTD.

Terseness in DTD redesign is of minimal importance...;-)

• Prev by Date: Re: Re: [dsdl-comment] Re: DSDL: use cases: namespace declaration notation
• Next by Date: Re: DSDL part 9: new namespace declarations not needed as part of DTD syntax?
• Previous by thread: Re: [dsdl-comment] Re: Re: Re: DSDL: use cases: namespace
• Next by thread: RE: Re: [dsdl-comment] Re: DSDL: use cases: namespace declaration notation
• Index(es):
  • Date
  • Thread