(People, please clean up the recipient lists, I do know a "group-reply" is easy
to use, but only one address is really needed:
ietf-822(_at_)dimacs(_dot_)rutgers(_dot_)edu(_dot_))
Mr. Satoshi KINOSHITA wrote:
....
I think this kind of interpretation is not good for both
iso-2022-* and MIME charset labeling.
A little smoother and reasonable consensus of rfc822 is necessary
so that we can use iso-2022-* or B and Q encoding as a valid body of rfc822.
I understand that rfc822 is not enough for I14Y(interoperability I mean ;-)
and we need some other methods.
But if rfc822 allows only strict ASCII,
two candidates, charset labeling and iso-2022-* profiling, will violate
rfc822.
One (to us "MIME-lovers") granted detail appears to be that
MIME AUGMENTS RFC822 by relaxing some rules on it.
(and creating a bunch of its own)
So yes, MIME is NOT rfc-822, but that should be obvious..
The RFC-822 does not say much about the message content,
and its interpretation, thus there is this profileration of
the various formats of "X-YYYYY:" -headers which may be present
on all email, or then may not.
Now what I would like to see instead of this heated discussion
is some acceptance of for the idea that also our esteemed Japanese
collegues would start using MIME labeling so that we others would
get SOME idea what the incoming message is all about.
(I do not mean we could learn to read it by that.. I wish it were
so simple..)
I do agree that "charset=" attribute is a misnomer in case
of ISO-2022, but if the solution in RFC-1554 doesn't work,
what would be better ? The ways of the "old days" ?
(Ok, RFC-1554 it has informational status, but such does not
mean it should not be followed. -- See RFC-1537 which I keep
referring to people having problems with their DNS..)
Our environments don't usually support ISO-2022-* natively, but
we may have external programs to serve us (like metamail).
To drive such programs automatically, at least I would prefer to
have MIME-header telling to my MUA that it is time to check
what the terminal can do, and possibly start converter.
Without those headers I would need an AI system -- perhaps I
can act it myself, but not all of my users are experienced
enough to recognize various formats.
Satoshi KINOSHITA
/Matti Aarnio <mea(_at_)nic(_dot_)funet(_dot_)fi>
PS: Please, DO edit your headers when you follow up to the list..