Peter:

I was merely saying that it takes much more consideration than, "I'll
do it!". A group of people should get together, understand the
encoding, and discuss implementation details. This could all be done
on the Swiki. I'll leave it to a more experienced Swiker to set that
up.

-John
----- Original Message -----
From: Peter William Lount <peter at smalltalk.org>
To: Squeak Mail List <squeak at cs.uiuc.edu>; John Duncan
<jddst19+ at pitt.edu>
Sent: Wednesday, September 15, 1999 5:50 PM
Subject: Re: Unicode support


> Hi John,
>
> I agree that a UNICODE implementation must meet the specification.
> Unfortunately UNICODE does not currently represent all the languages
of the
> world. By using 32bit character object instances and seperate
> encoders/decoders for each "character set standard" you gain
tremendous
> flexibility in supporting all langages as they get an encoding.
>
> Furthermore, there can be many other notions associated with
"character
> objects" beyond their "space efficient" encodings.
>
> All the best,
>
> Peter William Lount
> peter at smalltalk.org
> http://www.smalltalk.org
>
>
> >I'm wondering about the space and time efficiency of this prospect.
It
> >intuitively seems much more reasonable for 256 characters than for
> >57,709 16-bit values (some characters, some not) as defined in
Unicode
> >3.0. I, of course, could be stuck in alarmist mode, so a good set
of
> >reasons would justify it to me.
>
> >ASCII is such a messy thing, there is nothing wrong with doing it
one
> >way or another so that it seems intuitive. But for Unicode, there
are
> >many special circumstances that are spelled out distinctly by the
> >standard and many of them are required. Look at the document on
> >collation, for an example. Doing Unicode is a non-trivial thing,
but
> >it will be beautiful if it is done exactly to spec, and all of
Squeak
> >is transformed to use it.
>
> >-John
>
>