Re: [ft-devel] A review of ttfauthint by designer of Siri

[suzuki toshiya]

Werner LEMBERG wrote (2012/02/09 13:40):
> > I was testing Arabic webfonts on XP the other day and it was so
> > pathetic.  It's just, you can't get good rendering.  Period.
>
> Hehe.  Arabic is not supported (yet).

Indeed. The autohinting of FreeType2 uses per-script algorithm
to modify the outline. Unfortunately, no Arabic typography experts
had given the detailed comment describing how current autohinting
result should be improved, so the Arabic autohinting algorithm
is not separately modularized. An expert of Indic script had ever
found that CJK ideograph autohinting algorithm is better for some
Indic scripts (maybe the script with hanging baseline rule), so
the algorithm for Indic script is separated. I think the comments
from Arabic typography experts, with the comparison of unhinted,
current autohinting, and expected result, will be highly welcomed.

BTW, is it hopeless to ask for a comment in English from the experts
of Arabic typography? When I participate Internationalization &
Unicode Conference, often Thomas Milo (DecoType) make interesting
presentations, but nobody else make the presentation about Arabic
typography (although there are often the presentation by bidi experts).
Also, in ISO/IEC JTC1/SC34, when an expert from IBM had ever participated
(late 80s - early 90s), the typeface classification for Arabic fonts
(in ISO/IEC 9541) were made, but when he left, nobody could maintain
it. I'm afraid that the development of digital typography for Arabic
scripts is quite advanced and the knowledge in early 90s are already
obsoleted. Sorry for too off-topic note.


Regards,
mpsuzuki