Re: Request to relax PCDATA content model restriction
I would also like to second Eve's call for a relaxation of 45
((#PCDATA | a | b | c)* )
Mulberry also has many models that look like this, in fact our "house
style" recommends this over the alternative (#PCDATA | a | b | c)*.
But my objections are purely aesthetic, I like the symmetry and closure
of knowing that element content will always be enclosed in parenthesis or
(in the larger case of SGML) will be a keyword. Were I to write a
program that parsed element declarations, I would find it simpler to
depend on that fact.
Let's here from the other side of the coin. Simplicity and beauty are
not as powerful as technical arguments. Why NOT go with Eve on this one?