Re: General policy

Harald Tveit Alvestrand (harald.t.alvestrand@delab.sintef.no)
Tue, 03 Aug 1993 10:26:31 +0200


Date: Tue, 03 Aug 1993 10:26:31 +0200
From: Harald Tveit Alvestrand <harald.t.alvestrand@delab.sintef.no>
Subject: RE: General policy
To: luc@opus.spc.nl
Cc: ietf-charsets@INNOSOFT.COM
Message-id: <"10892*/I=t/G=harald/S=alvestrand/OU=delab/O=sintef/PRMD=uninett/ADMD=/C=no/"@MHS>

When talking about labelling, I was thinking of mechanisms outside
the bytestream itself, unlike ISO 2022.

That means that when you start the bytestream, you expect to see one
character set/encoding method, and when you go to the middle of the bytestream,
you still expect to see the same charset/method.

The MIME "charset=" construct is the paramount example of such a labelling.
It could probably be easily added as part of TELNET negotiation, and so on
for other protocols.

                         Harald A

--Boundary (ID uEbHHWxWEwCKT9wM3evJ5w)