In defense of:
The canonical form of any MIME text type MUST represent a line
break as a CRLF sequence. Similarly, any occurrence of CRLF
in text MUST represent a line break. Use of CR and LF outside
of line break sequences is also forbidden.
Ned Freed said:
The choice, then, is simple: Either you ban the use of stray CR and LF in text
or else you require agents to maintain a comprehensive list of all the
character sets and whether or not conversion to canonical form is possible
and/or necessary.
There's something interesting about this. We've been unable to control
the proliferation of zillions of charset registrations, so the idea of
'a comprehensive list of all the character sets' sounds daunting.
However, it isn't necessary that the agent know all of the charset
registrations if it were possible to determine SOLELY FROM THE CHARSET
NAME whether such the charset used something other than CR and LF as a
line break sequence, even with a stupid lexical trick like
charset=*unicode-1-1-ucs2.
I apologize; I hate to rehash something that you've discussed
endlessly without having also suffered through the hundreds and
hundreds of messages, but I think disallowing the simplest binary
representation of 16-bit charsets on-the-wire seems like a serious
restriction and worthy of just a little more mooting.