I like Nathaniel's proposal!
Let me see if we can reach some closure on this issue by offering a proposal:
1. Nothing changes in the draft standard version, thus avoiding the
issue of setting back the clock.
2. We write a VERY SHORT informational RFC advising people on the
"reasonable" way to handle MIME-Versions other than 1.0. It has no
standardization status, but is simply advisory.
3. In the full standard version of MIME, which I guess will probably
come out in early 1994, we change the paragraph about MIME-version to
either refer to the above-mentioned RFC for advice or reproduce the
The end result of all this will be that "1.0" will remain the only
string defined even by the FULL standard version of MIME, but the final
standard will include some reasonable advice about how to handle other
versions. That way, if a major MIME revision is needed a decade down
the line, there will at least be SOME chance that the MIME-version field
can be used to flag it, on the assumption that by then the Final
Standard version will be the only one that matters.
Does this sound reasonable to people? If so, I'll try drafting the
aforementioned RFC... -- Nathaniel