> Could we go through the bone of contention up front
> and just settle it: are grove and grove plan definitions of
> use to an object-oriented API design? Really just asking here,
> so this gets settled and doesn't crop up again and again.
I think so. It's very straightforward to translate class
definitions from a grove plan into class definitions in
(Java|C++|Modula-3|OO-language-of-choice). That's pretty
much what James Clark did for the generic SP interface.
That's a good approach for processing XML-as-XML. A slightly
different approach is useful when you want to add application
semantics, namely, to create a new class for each element
type of interest, (or better yet, for each architectural form).
That's sort of what SGMLS.pm, the original CoST, and (I think)
Peter M-R's CML tools do.
--Joe English
jenglish@crl.com