On Thu, 16 Jan 2003 07:45:55 -0800
Dave Crocker <dcrocker(_at_)brandenburg(_dot_)com> wrote:
Keith,
Although well intentioned, you are pursuing a question that is
fundamentally irrelevant to an IETF technical effort.
That's okay, ietf-822 is no longer an IETF working group, hasn't been one for
around ten years, and the discussion here doesn't have to strictly stay within
the boundaries of IETF technical efforts.
The IETF does not produce technical specifications that use
heuristics.
Tell that to the ftpext working group!
Even assuming that your statement were true, and that your statement were
relevant to this discussion -- IETF working groups do need to evaluate the
feasibility of proposals, and part of that is understanding the impact on the
installed base. For better or worse, and in spite of language in the standards
prohibiting such behavior, extensive use of unencoded non-ASCII text in message
headers has been observed.
One thing that we observed in introducing MIME is that there can be
considerable expectation and/or demand from users that nonstandard practices
continue to work (at least, as well as they ever "worked"), and that failure to
consider that can delay adoption of new standards, even when those standards
are technically superior. So it would seem prudent that any proposed
extensions to the message format to support a wider range of charcters than
US-ASCII be evaluated according to whether they would adversely impact
nonstandard practices that appeared to "work", even for a limited community.
Anyway, at this stage I'm just asking a question - not making any kind of
proposal to be attacked, nor even presuming such a proposal. I hope that
someone here knows the answer, and I hope that the answer will be interesting
to others in this discussion.
--
Keith Moore http://www.cs.utk.edu/~moore/
27 February 1933 11 September 2001