A few clarifications of points Erik asked about in this example:
<excerpt><bold>Excerpts from internet.ietf-822: 4-Feb-92 richtext <lt>nl>
model conside.. Erik Naggum(_at_)naggum(_dot_)no (2255)</bold></excerpt><nl>
<nl>
1. The truncated subject, truncated address, and parenthesized number
are all Andrew-isms, and not relevant to richtext.
2. The "<lt>nl>" should have been obvious -- the real subject line had
a literal "<" which was replaced by "<lt>" in the richtext rendition.
3. The semantics of "<nl>" inside excerpts are the same as anywhere
else -- they cause line breaks. The place where it gets hairy is when a
justifying-text mail composer (e.g. Andrew) excerpts mail in
non-justified form (e.g. most plain text mail). While the resulting
excerpt might LOOK better without the hard newlines, this is just plain
not correct -- you really HAVE to assume that the newlines in the plain
text mail you're excerpting are hard newlines, or you're going to mess
up a lot of things. I really don't see any way around this, and it will
apply to any richtext-like scheme whenever you try to excerpt plain text
with hard newlines inside a mail composer that is using justified text
(soft newlines). I don't think this is an artifact of richtext, but
rather of any attempt to excerpt hard newlines in justified text.
4. As I've said before, the " </excerpt><excerpt>" stuff is an Andrew
artificat that I think I've mostly managed to get rid of now. The
details are kind of fascinating, actually, but irrelevant to this
discussion.
The idea of machine-readable attribution is a very intriguing one -- I'd
file in in the category of "neat ideas of things to do with richtext
that grow inevitably out of its use". I think this list will grow
substantially with the real-world use of richtext, but not without it.
I'd LOVE to see this feature included in a future richtext, but again I
think delaying the current one to include it (or virtually any other
feature) will work against the wider adoption of ANY richer text
mechanism. -- Nathaniel