<SIMON>
[... TimB and JonB snipped...]
>I'd like to see the syntax completed by the end of June - a much tighter
>timeframe than the usual deliberations. By keeping it simple and tightly
>focused, I think we stand a good chance of completing at least one round of
>this before the WG gets started. Hopefully the discussions (and perhaps even
>implementations) produced by this proposal will be of use to the XML WG. I
>hope strongly that this proposal will be seen as contributing, not competing.
</SIMON>
I agree with this schedule - I think it's achievable if we stick to
elements and attributes. I think we should avoid getting into complex
discussions of other possibilities at this stage.
Note that if we get a representation, then it may very well act as a
catalyst to add 'DTD' information to SAX - we shall have agreed what the
most important aspects are and we shall have kept them simple :-)
P.
BTW. I see a DTD as having a tree structure because I do not treat the
element in the contentSpec as equivalent to the element declaration. IOW with:
<!ELEMENT A (B, C)>
<!ELEMENT B (#PCDATA)>
<!ELEMENT C (#PCDATA)>
I see A's tree ending with the references to B and C. If we need to find
out what they are we link to B and C rather than including them in the
tree. In that way we don't recurse.