[
Lists Home |
Date Index |
Thread Index
]
Simon St.Laurent wrote:
> jonathan@openhealth.org (Jonathan Borden) writes:
> >You are not the first to desire this. Indeed the concept of SGML
> >"Groves" was intended to allow a full fidelity representation via an
> >API of an SGML document (or any arbitrary subset as desired).
>
> I agree that I'm not the first to desire this, but at the same time
> Groves is a lousy comparison. I'm focusing on concrete characters, only
> then permitting abstraction, while Groves starts with the abstraction
> and lets you reach full fidelity only afterward.
>
Not all "groves" need to be overly abstract (but granted you are correct for
the most part). But for example <a href="http://www.openhealth.org/XSet/";
>XSet</a> is based _directly_ on the character matching productions of XML
1.0 -- whose terminals are all characters.
My question is how many folks need to care about these details? These
details seem to fall on the wrong side of the 80/20 divide. I am not sure
why I should get excited about attribute order, is there some application
(outside an editor) where this ought be important?
Jonathan
|
