Re: A proposal

[Keld J|rn Simonsen <keld@xxxxxxxx>]

About Erik Naggums greeting line negotiation: 
Yes that is a way of doing it.

As stated earlier we need more than 8859-1, and this can be done in
two ways:

1. using a universal codeset, in my mind that is 10646.

2. allowing a set of character sets, and namings for those.
   This could be the set of 8859 standards and also Japanese,
   Chinese and Korean 16-bit standards. Maybe some others
   which could be added along the way, eg the Indian ISCII code.
   RARE wants vendor specific character sets too, and anyway you will
   in many places need to convert locally to the character set
   of your display hardware.
 
   The negotiation could be done by "the third word of the greeting
   message" being "I-CAN" or other kinds of clever negotiation,
   and then a parameter telling the character set being available.

   If we do not employ 10646 and want to not loose information, we
   need mechanisms to encode information in an 8-bit character set
   from a different 8-bit character set. This could be the TEX-HEX
   encoding (which I still have no information on, could somebody
   please enlighten me?) or the mnemonic scheme that I have proposed.
   The 8-bit in 8-bit encoding could hopefully be the same as
   the 8-bit in 7-bit encoding, to reduce complexity!

Keld