As a member of the 822ext working group, I would like to query you
about EUNet's reasons for insisting that MNemonic be cited in the
MIME specification. I am copying the mailing list, since the question
of character set support in MIME is a topic that has been troublesome
in a number of different discussions.
Let me review my own sense of MNemonic's role and status:
The world has a painfully large number of character sets, some of which
are well-established and some of which are emerging in specification
and/or use. The 822ext working group, with specific encouragement from
the IETF Steering Group, limited its explicit references to character
sets that have extremely well-established international use. This
resulted in built-in citations only to ASCII and ISO 8859-[1-9], I
believe.
However, the document specifically provides for support of additional
character sets, through separate registration with the Internet
Assigned Numbers Authority (IANA). Hence, anyone may register the
name of character set, given a specification for it, and anyone may
also pursue standardization of the specification, separately.
Along with EUNet's vigorous support for MNemonic, there is other
vigorous support for 10646, for Unicode, and for ISO 2022. We have
chosen to have MIME make explicit reference only to the minimum of
extremely well-established and heavily-used specifications, since
the current flurry of character-set related standards activities has
not yet produced a clear winner.
Hence, it seems to me that MNemonic should be pursued through the
standards process, if you wish, but on its own merits, rather than
being tied to an electronic mail format standard. Further, direct
use of MNemonic, within MIME messages, is not prevented by the MIME
specifications. The IETF process usually involves the (painful)
identification and separation of those items with (almost) universal
agreement, from those that are more controversial and in need of
specific focus and discussion. The entire topic of character sets
has proved to be just such a difficult area.
So, again, I would appreciate understanding EUNet's reasons for
insisting the MNemonic be tied to MIME.
Thank you for voicing your position and, in advance, for pursuing the
topic further.
Dave Crocker