-----------------------------
Application message id: 25954140202991/12295 X400
Grade of Delivery: Normal
-----------------------------
VMSmail To information: MUVAXA::MRGATE::"mci
_mta::*emsinternet::*mbx1ietf-822(a)dimacs.rutgers.edu::su=Re:"
Sender's personal name: John C Klensin
Nathaniel,
While I certainly sympathize with those who don't like looking
at <nl>, I would vote, strongly, for leaving well enough alone.
The SGML compatability, or near-compatability, is a significant
issue for certain of us; if you abandon it then I think we want
to go back and review how much of the other clutter it introduces
can be removed.
A comment of Walt's reinforces this general point:
If you don't want to be SGML, then don't use their
delimiters and make it look oh-so-close.
Also note that, for those of us who are getting very worried
about the size/efficiency issues in all of this, if mail crosses
a gateway into an environment that does fixed-length records, an
N+1 newlines convention translates into a lot of added records
and bandwidth requirement. Probably minor compared to things
MIME has done to them already, but it adds up.
I vote, strongly, for the following options in descending
order of preference:
(i) If everyone is extremely comfortable with what we have,
leave it alone.
(ii) If a single additional suggestion emerges from anywhere,
including the author, that people take seriously, take Richtext
out of the MIME document and start working on a separate RFC.
It is appreciably less necessary to have richtext bound to MIME
than it is to bind character sets and their representations.
Mnemonic has been taken out over the strong objections of Keld
and others, and, whatever its strengths and weaknesses, there is
more field experience with it than with richtext in its present
form. More generally, if this WG can be thought of as being made
of mostly of mail experts, richtext--as it has evolved--should
really be carefully examined by a group that contains a high
preponderance of text and representation experts. As far as I
can tell, we have one or three, including Nathaniel, but that is
perhaps not enough for something this important. So, if anyone
has serious doubts, or proposals for last-minute fixes, let's
separate it. If we are still debating issues this basic, it
just isn't ready.
--john
klensin(_at_)infoods(_dot_)mit(_dot_)edu