Ned,
Thanks for the detailed response. My sense is that the upshot of all
of this is that those of us who are (or were) leaning toward trying to
tag an "interchange Postscript" separately should wait for the
implementation and use experience to come in and then start writing, and
those who think that is impossible or useless should just sit back and
watch the fun.
To put this differently, I think that in this and a few earlier
messages you (and others) have made a strong and convincing case that
what is appropriately called and tagged as "postscript" is a messy,
although moderately-well-specified, business with uncomfortable
interchange implications in at least some cases. So be it: if I want
maximum interchangability, it may not be what I want to send and my
making that decision or not is not a MIME problem. That also implies
that some internet-agreed-upon minimal common interchange postscript is
a future piece of work, requiring a definition and a different name and
that it probably would be actually more misleading to label it
"postscript" than applying that name to the current, common practice,
situation.
In no event has a good case been made that anything is wrong with the
MIME spec, certainly nothing that couldn't be fixed at the next stage if
experience uncovers anything unexpected.
I wonder whether ISO SPDL, when it arrives, will change any of this
mess :-)
john