But this rush to incorporate ideas that were un-thought-of and
un-heard-of only a month ago is a bit scary.
I agree with this. However I think that parameterized character sets are
definitely in this category.
Content-type: TEXT/ISO-2022; charset=US-ASCII,JIS
That, of course, provides another little problem for the proposed
syntax.
Maybe
Content-type: TEXT/ISO-2022; charset=(US-ASCII,JIS)
Yuck! In the context of "Content-type: text" we need to know what
algorithm we use to convert the sequence of octets to a sequence of
glyphs. I call that algorithm a Character set, but if we want to
call it something else I don't care. I suggest one _unparameterized_
name per algorithm. Anything more is new country. That algorithm
name used to be the subtype of text, then we moved it to the
charset parameter of type text/plain. Now we want to make the
charset parameter be something that isn't an algorithm. We have
squeezed out the one bit of information that is absolutely essential
for the mail reader to do its job. When various people agreed to
moving the subtype information to the charset parameter they didn't
agree to this much change in the underlying structure.
To make ISO-2022 into an algorithm you have to specify various extra
bits of information: what modes are supported and what the initial mode
is. I once again suggest that we have an algorithm called ISO-2022-J
which is ISO-2022 with JIS and US-ASCII and starting in US-ASCII mode.
This is a well defined algorithm or the Japanese couldn't be using it
so successfully. I don't care if you call it a character set or not, but
it has to appear in the Content-type line in the same position as all
the other well defined algorithms (which may or may not be character
sets) like US-ASCII, ISO-8859-1 and MNEMONIC.
Bob Smart