(Minor aside: The Xerox local ATK style "code" sounds like what I use
"Example" for, but that's an Andrew discussion, not an RFC-XXXX one.)
I agree with Bill that people tend to use semantic styles such as
"quotation" instead of formatting styles such as "italic". Certainly
richmail is easily extended to include them. In fact, I thought about
including them in the initial list of defined styles, but was trying to
keep that list short. Perhaps I made it too short.
However, as a previous message of mine implies, richmail can still be
used to pass these things through even without the user selecting
"italic/indent". For example, you could imagine an Andrew user choosing
"Quotation" just like he does now. If Andrew is modified to write
things out using richmail, it could handle this in any of several ways:
1. Redundancy, e.g. "%quotation(%italic(%indent(whatever)))", with the
latter two being interpreted as redundant by Andrew and the first being
ignored by most richmail viewers.
2. Embedded Andrew commands, e.g.
"%invisible(environsubstitution=quotation, next
2)%italic(%indent(whatever))". Here the invisible comment is ignored by
regular richmail readers, who use italic & indent. But when read by
Andrew (presumably having been alerted to the presence of embedded
Andrew information by an appropriate invisible comment at the
beginning), it would substitute the "quotation" style for the italic &
indent styles.
3. Devil-may-care, e.g. "%quotation(whatever)", knowing that most
richmail readers will see "whatever" as plain text.
Actually, the more I think about it, the more I think that adding a few
predefined "semantic" environments to richmail might be a very good
idea....