Re: (Repeat) Decision: C.4 (Predefined entities)
% From: firstname.lastname@example.org (David G. Durand)
% I agree with this, and would strongly support overridable predefined
% Of course, that means that a DTD-less instance with redefined entities
% will produce different results. I expect people could learn to live with
% this, since those who create XML DTDs will have to learn how, and that
% learning will lilkely include the details about overriding. I am sure that
% the issue of DTD-free processing is what drove the committee to the
% un-redefinability condition.
The whole issue here seems to have been caught up with attitudes towards HTML.
I don't think HTML is even particularly relevant to the main question, which
Is the simplification inherent in defining a set of entities worth the loss of
flexibility for this version of XML?
The advantage of the simplification is knowing that applications won't do
something weird with your data if you don't supply a DTD. It's knowing that
ä is always a with an umlaut, etc.
The other side is the flexibility that SGML users are used to having, and that
hopefully XML users will soon get used to having. This would come at the cost
of some bandwidth to transmit the redefinitions, and some complication to the
We should make such a decision based on what we think will be good for XML,
not because of, or in spite of, anything that MS or NS may or may not do in
the HTML world.
I personally think having a base set is a good idea; I would probably prefer
to use other entity names to avoid clashes and possible misunderstandings;
but others who already have large quantities of data that they want to put up
on the Web with as little fuss as possible need to think about whether the
trade-off is worthwhile. I do; others may not.
Lauren Wood, SoftQuad, Inc.