Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!wuarchive!usc!apple!ames!haven!mimsy!mojo!buzzard
From: buzzard@eng.umd.edu (Sean Barrett)
Newsgroups: comp.sys.amiga.advocacy
Subject: Re: Announcement--new "Unicode" standard
Keywords: the kissable breasts of Yong-Mi (t.m. xanthian lust)
Message-ID: <1991Feb25.154148.29542@eng.umd.edu>
Date: 25 Feb 91 15:41:48 GMT
References: <1991Feb24.180148.21954@ux1.cso.uiuc.edu> <1991Feb25.044130.8588@zorch.SF-Bay.ORG> <1991Feb25.080721.1628@zorch.SF-Bay.ORG>
Sender: news@eng.umd.edu (C-News)
Organization: Fraternity of Avian Deists
Lines: 28

I think we have gotten far off from the Amiga here. *sigh*

So xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) says:
>
> Classic_-_Concepts@cup.portal.com writes a description of Unicode
> thad@public.BTR.COM (Thaddeus P. Floryan) points out it takes 16 bits
>
>Actually, 16 bits is insufficient; a comprehensive Chinese dictionary of
>ideographs contains over 100,000 individual ones, and that is just one
>language's alphabet.

If the goal is portably *transmitting* characters, you can use 16 or 32
bits freely, making sure that the bit patterns you use don't have any
"naughty" 8/7-bit ASCII values in them.  If you don't need to be able
to intermix or auto-detect what language something is, this is sufficient.

> Ethan (deleted the ref, oops) is sure the people making the standards
>    aren't dumb, but that they'll get no support for 16 bit.
>
>Actually, the ASCII standard has extension mechanisms for ever larger
>byte sizes; the use of 7 bit ASCII is trememdously parochial of the US.

A trivial extension to ASCII (isn't this *obvious*?!!) if you need 32768
new characters is to throw out parity, leave ASCII 0-127 alone, and
on ASCII 128-255 fetch the next character and combine them, giving
15 bits of new characters (or you could fetch 3 and have 31 bits, or
you could have 128-191 fetch one character for 14 bits, 192-255 fetch
3 characters more for 30 bits).