Dear John,
Thanks for your response. Response responses below.
> -----Original Message-----
> From: John Cowan [mailto:cowan@locke.ccil.org]
> Sent: Wednesday, July 14, 1999 8:52 AM
> To: Unicode List
> Subject: Re: Arabic - Alef Maqsurah
>
>
> Reynolds, Gregg wrote:
>
> > Unicode codepoint U+0649 is defined as "alef maqsurah", and the
> > representative glyph shows a dotless ya.
>
> It is clear that the glyphs are merely representative,
> and that other forms are not only possible but may be preferred.
> U+0123 shows only one of three possible glyphs for this character,
> and the "representative" glyph is in fact dis-preferred.
>
Clear in what way? Alternate forms of the same abstract shape, or alternate
shapes altogether? Dotless ya is not "representative" of alef maqsura but
one of two representations. (Note that this is not a contextual shaping
issue.) Also, U+0123 states explicitly that there are "3 glyph variants";
no such note on U+0649. And I'm not sure exactly what to make on the U+0123
note; 3 variants on the same shape (metafont?) or 3 distinct abstract
shapes. For alef maqsura its the latter.
>
> But because ISO character sets depend on the stability of names (in
> ISO theory, characters in different sets are identical iff their names
> are identical), changing them is effectively impossible, and inferring
> character properties from the names (as we have been reminded
> recently)
> is fraught with peril.
>
Ok, I can live with that. I wonder if there isn't some way to add, rather
than change. Effectively this would mean an errata sheet. I.e., in
addition to the official, cast in iron name (maksura), each character has a
(usually null) alias property which could be used for corrected names. As
for inferring properties from names, I would think that if we can't do that
it's a sign that we have the wrong abstractions; but I suppose it's
something we have to live with in most cases.
Sincerely,
Gregg
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:48 EDT