Well, my understanding is that it's not that simple.
1) **THE** reason for the existence of the Arabic presentation forms
is almost entirely political. A few of the presentaion forms
apparently were in some legacy encodings (EBCDIC based, I believe),
but the wast majoroty of them are there for no technical reason at all.
2) As far as I know, normal practice in OT and AAT fonts supporting
the Arabic script is to support the Arabic characters that are
contextually shaped, but rarely to support any of the Arabic
presentation forms characters. Note that the glyph indices need
not have anything to do with the character codes in TT fonts.
3) Apparently experts on Arabic (which I'm very far from) agree
that the presentation forms encoded as characters does not even
nearly cover the needs for good typography for the Arabic script.
4) The very existence of the "Arabic presentation forms" block
entices suggestions of adding more Arabic ligatures for particular
uses. However, neither the UTC nor WG2 has any inclination
whatsoever (based on advice from experts on this) to encode
any more Arabic (or Latin for that matter) ligatures of any
kind. Liguture generation is a font issue, not a character
issue.
5) While the "Arabic presentation forms" block can be handy
for glyph indices for typographically less demanding
applications, it is still not sufficient (I'm told): more
glyphs (for ligatures) are sometimes needed even then.
And as you mention the handiness of point 5 is not available
for Indic scripts. So another solution must be found.
For OT and AAT fonts (both TT based) another solution is already
in place (and there the glyph indices are internal to the fonts,
no need for a general agreement on which indices the glyphs
are to have).
Kind regards
/kent k
-----Original Message-----
From: Karl Koehler
To: Karlsson Kent - keka
Cc: '[EMAIL PROTECTED]'
Sent: 2000-11-15 19:59
Subject: RE: Standards (Re: Arabic in fixed width fonts)
On Wed, 15 Nov 2000, Karlsson Kent - keka wrote:
> Relevant to this e-mail thread is also the general
> recommendation from Unicode consortium to NOT use
> Arabic presentation forms characters, which are there
> for purely political reasons.
The question is not if one should use them, but for what
they should be used.
Documents should not use them; they should use the relevant
sections starting at 0x620.
But glyphs for a font cannot be encoded this way, so an application
which wants to display arabic must use a different font encoding.
This is what the region of the presentation forms is used for.
I think this is legitimate, as it makes life simpler.
Otherwise one would have to declare that there is no such thing as a
unicode
font for arabic, and would have to agree upon a way of encoding fonts.
For the Devanagari script, this already has to be done; no ligature
forms are in unicode.
Karl
