Keith Moore writes:
I would interpret the first rule as "every symbol has a unique
interpretation"...where "symbol" is not necessarily a byte...this rules out
things like ISO 646, in which the meaning of some code points differs from
place to place.
Yes, it leaves out ISO 646, but not the versions of ISO 646.
Each of these versions have a well-defined code and corresponding
character. E.g ASCII is well defined, the 2 Swedish versions are
well defined and so are the rest of the ISO 646 versions registered
with ECMA.
RC1345 contains wording on what is understood by a character set in
RFC 1345 sense, this was written to be in accordance with the Santa Fe
discussion.
keld