I'm not a developer (as most of you know well!) but I wish to echo Ned's
concern, which has also been troubling me as I've followed this discussion,
about the delays that this issue necessarily causes in the deployment of
MIME, which I personally am not convinced are warranted by the advantages
proposed.
I believe that MIME is needed NOW, amd I would hate to see us spend so
much time refining it that it doesn't get implemented. It would really be
tremendously useful to have its present capabilities widely available
today.
Jim