W3C home > Mailing lists > Public > ietf-charsets@w3.org > April to June 2001

Re: Registration of new charset: UTF-32

From: Keld Jørn Simonsen <keld@dkuug.dk>
Date: Sun, 20 May 2001 19:15:03 +0200
To: Asmus Freytag <asmusf@ix.netcom.com>
Cc: Keld Jørn Simonsen <keld@dkuug.dk>, Misha Wolf <Misha.Wolf@reuters.com>, Mark Davis <mark@macchiato.com>, ietf-charsets@iana.org, w3c-i18n-ig@w3.org
Message-id: <20010520191503.B31107@rap.rap.dk>
On Sat, May 19, 2001 at 06:28:35PM -0700, Asmus Freytag wrote:
> At 06:48 PM 5/14/01 +0200, Keld Jørn Simonsen wrote:
> >On Sun, May 13, 2001 at 11:34:24AM -0700, Asmus Freytag wrote:
> > > At 08:05 PM 5/11/01 +0100, Misha Wolf wrote:
> > >
> > >
> > > >Has anyone looked to see how this ties in with:
> > > >   Extensible Markup Language (XML) 1.0 (Second Edition)
> > > >   Autodetection of Character Encodings (Non-Normative)
> > > >   http://www.w3.org/TR/REC-xml#sec-guessing
> > > >
> > > >Misha
> > > >
> > >
> > > I took a quick look. The section already talks about 4-byte codes.
> > > Replace UCS-4 by UTF-32 in that section and it would seem to cover it.
> >
> >I think that the w3c specs should rather refer the 10646 specs,
> >and thus keep the reference to UCS-4.
> 
> We deliberately introduced the term UTF-32 since this regularizes
> the notation for everyone. Little is to be gained by using a mixed
> notation in that section using UTF-8, UTF-16 and UCS-4 together.
> 
> If you would like to fix the 10646 spec, you could propose that the
> term UTF-32 is formally added there as well. As it stands, 10646
> has an unfortunate asymmetry in notation that is cumbersome to use
> for the non-specialist.

You really should not do this. UCS-4 is the canonical representation of
10646. UTF-32 would be misleading, as the UCS-4 is not a transformation format,
but the "real thing".

Kind regards
Keld
Received on Friday, 25 May 2001 20:24:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 5 June 2006 15:10:51 GMT