Re: UTF-8N?

From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Mon Jun 26 2000 - 22:26:52 EDT


At 05:29 AM 6/23/00 -0800, Peter_Constable@sil.org wrote:
> >Yes. The Unicode Standard will deprecate the use of U+FFEF (Note: not
>U+FFFE)
> >as a zero-width non-breaking space (despite its formal name).
> >
> >And U+FFEF should *only* be used as a byte order mark and/or signature.
>(That
> >is already ambiguous and trouble enough -- without tossing in the
>orthogonal
> >issue of the need for a non-breaking zero-width space.)

In my book, this is one of the most egregious recent mistakes by the UTC.
Especially in light of the fact that the use of FEFF as ZWNBSP is supported
today in widely distributed software - who's going to convert all the data,
and what will happen when "programs will feel free to toss out U+FEFF" as
someone recently suggested.

While well-intentioned, it is essentially 'moving' a character, by moving
the semantics to a new character code.

To balance the scale of my previous remarks: An even more egregious mistake
was made by WG2 by not accepting U+FEFF as a byte order mark, but insisting
on a then dubious ZWNBSP name and semantics.

Such is life.

A./



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:04 EDT