As argued elsewhere, I think it makes more sense to use the STYLE
attribute and reserve CLASS for data-type distinctions, but the argument
seems otherwise sensible.
If we want to encourage the use of stylesheets for this kind of markup,
perhaps we should define CSS to have a collection of built-in styles for
the most common markup - things like subscript, superscript,
line-through, underlined, bigger, smaller, bold, italic, etc. - so
that authors can use them without having to add definitions for them
to their own STYLE element (or even to have a STYLE element if they
don't otherwise need it). Any UA supporting CSS would be required to
provide a default STYLE element containing a standard set of
definitions; a document-provided STYLE element would cascade over the
default, overriding definitions for the built-in styles, if desired.
Coupled with this, I suggest that we modify the STYLE element syntax to
allow using style names in right-hand sides, indicating that all the
properties associated with that named style should be included. This
would allow, for instance:
[CLASS=doctitle] { underlined }
[STYLE=C-keyword] { bold; font-family: helvetica ]
This allows the actual properties associated with, for instance, all
elements whose style includes underlining the text, to be hanged by
changing the definition of the underlined STYLE. For instance, a
stylesheet for a medium not supporting underlining could do something
like
[STYLE-underlined] { background-color: red }
and affect every style rule using the style underlined, rather than
individually overriding every such style rule, as would be required with
thte mechanism in the current draft.
scott
-- scott preece motorola/mcg urbana design center 1101 e. university, urbana, il 61801 phone: 217-384-8589 fax: 217-384-8550 internet mail: preece@urbana.mcd.mot.com