Re: Public Identifiers, and CATALOGS
Peter Murray-Rust wrote:
> Of course with careful client-side maintenance this restriction can be
> relaxed, but the likelihood of inexpert users breaking the integrity is
> considerable. A (?the) primary role of PUBLIC is thus to act as an integrity
> check at the client side, even if it's not used for resolution.
The word "manifest" was used earlier and that if nothing else is
worth the trouble. In our system where we depend on non-SGML
support files (stylesheets, wordlists, frame files, etc.), the
management and transmission of the multiple files has been and
is a problem unless one is careful. It requires knowledge of the
system and attention when packing up a document and moving it.
So, where a catalog preserves these relationships in a system
neutral form, it is a considerable help.
I think anyone who believes they can use XML or SGML without
some management help from tools or the language is being very
optimistic. I agree that some classes of simple well-formed
documents will stand on their own just as HTML does now without
stylesheets, but to get the advantages we assume the responsibilities.
This is why I replied to Paul Prescod on CTS that I don't think
the rush to XML will be fast: inertia is dependable. On the
other hand, where it solves problems of complex document
production, validation, management and persistence, it will be
adopted just as SGML has been where these same problems exist.
Steve Newcomb has it right: keep the requirements up front. They
should be user requirements, not just, Web system requirements.
But bear in mind, not one without the other. My problem with
Jon's catechism is that it does not seem to be both.