Serious question:
Like it or not, there is an existing population of "8-bit sendmail"
sites out there. Has anyone with more experience than I (>epsilon) given
thought to the question of how we migrate that population to MIME and
RFC-ZZZZ? Harald Alvestrand's "stupid semantics question" makes me
wonder if there might be a gateway solution that would induce such
populations back into the fold.
Thanks,
-andy kegel(_at_)zk3(_dot_)dec(_dot_)com
On a Sunet (Swedish University Network) technical reference group meeting
this week we discussed MIME and if and how to convert to it.
We agreed on recomending MIME as soon as it is feasible. As most sites
nowadays just install what is delivered by the manufacturer, it is very
important that these have a suitable MIME version. We will write a
document on what features are necessary in our environment to make
MIME fly.
The problem we see is that if some groups start to send MIME mail to
others that don't have MIME, this will make them very angry and will
most probably start a movement against MIME. This is because most users
don't like all the headers they receive today. With MIME they will get
even more headers. And worse, instead of plain swedish characters they
will see a lot of hex codes (if quoted printable encoding) or other
substitutions (if using mnemonic encoding). Compared to what they have
today, they will regard this as a big change to the worse. And the
8-bit systems would continue to grow as fast as they are today, and
MIME would never take off.
So what we need is a stepwise introduction. The first phase is to
distribute mail handlers that are MIME aware on reception, but does not
normally send MIME mail. When such versions have penetrated the market
sufficiently, the second phase with full MIME support as default also
on sending can be used (or better, use a configuration parameter for
what is sent).
To promote the switch from todays mail handlers to the firts phase
MIME handlers, we would like an extra feature not usually available
today. We would like them to have a way to easily switch presentation
of non-MIME mail using two different character sets. One set for 7-bit
national characters and the other for 8859-1. That way we would be able
to read plain ASCII (subset of 8859), "swedish ascii" and 8859-1-mail,
which cover 99.9% of current use.
Well, that is in short what will be in the document. Personally, I also
think that to be able to start phase 2 it is important to use mnemonic
encoding. The hopefully less than 10% minority that for different reasons
can not get MIME support will cry out loud when phase 2 is started. As
mnemonic is the least ugly to read for the common user, the volume will
be lowest that way.
Jan Engvald, Lund University Computing Center
________________________________________________________________________
Address: Box 783 E-mail:
Jan(_dot_)Engvald(_at_)ldc(_dot_)lu(_dot_)se
S-220 07 LUND Earn/Bitnet: xjeldc(_at_)seldc52
SWEDEN (Span/Hepnet: Sweden::Gemini::xjeldc)
Office: Soelvegatan 18 VAXPSI: psi%2403732202020::xjeldc
Telephone: +46 46 107458 (X.400: C=se; A=""; P=Sunet; O=lu;
Telefax: +46 46 138225 OU=ldc; S=Engvald; G=Jan)
Telex: 33533 LUNIVER S