Re: Inheritance/defaulting of attributes

Paul Prescod (papresco@technologist.com)
Fri, 10 Oct 1997 09:58:20 -0400


akirkpatrick@ims-global.com wrote:
> Re the question below from Rick: as I understood Paul, if an element
> inherited a content model, it could define a more restrictive model for
> itself but still had to be valid according to the parent model (see
> below).

That's close, but not quite correct. According to my terminology, why
you've described is subtyping. If you *inherit* a content model then you
get that content model -- unchanged. We could invent straightforward
mechanisms to allow you to inherit content model parts but I'm not sure
yet if it is worth the extra syntax. Especially if we were to move to
structured attributes. Then today's complex content models could be
broken up into small components and the benefits of inheriting content
model parts would be reduced.

Paul Prescod