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

Re: Re:DTD Design

• To: <xml-dev@l...>
• Subject: Re: Re:DTD Design
• From: "Rick Jelliffe" <ricko@a...>
• Date: Thu, 21 Nov 2002 17:13:34 +1100
• References: <5.1.0.14.0.20021120175432.0280cd80@m...> <5.1.0.14.0.20021120183758.02811898@m...>

 From: "Sean McGrath" <sean.mcgrath@p...>
 
> My memory is fuzzy on this. I believe you had to declare the elements like so:
> 
> <!ELEMENT sect 1 - o (...)>
> <!ELEMENT sect 2 - o (...)>
> 
> But that if you had in the instance:
> <sect1>
> <sect>
> </sect1>
> 
> The inner element was inferred to be a sect2...or something like that!
> 
> Rick?
 
Don't ask me. I can understand how this would have been useful 
from 1986 to 1991 when your applications mightn't have had access to 
the ancestor stack,  but almost all processing systems (Xpath, CSS 
selectors, OmniMark, etc) from the earliest 90s did provide context information.
When I started in the late 80s, our little custom-LISP processor for SGML
kept a stack, so I escaped ever seeing RANK. 

I think Sean's DTD is correct.

Cheers
Rick 

• References:
  • Re:DTD Design
    • From: Sean McGrath <sean.mcgrath@p...>
  • Re: Re:DTD Design
    • From: Sean McGrath <sean.mcgrath@p...>

• Prev by Date: Re: [ANN] Lucid XML Data Manager new release
• Next by Date: Re: canonical xml
• Previous by thread: Re: Re:DTD Design
• Next by thread: HI
• Index(es):
  • Date
  • Thread