RE: To submit or not to submit

From: John Hudson (tiro@tiro.com)
Date: Mon May 13 2002 - 14:55:02 EDT


At 04:03 5/13/2002, Marco Cimarosti wrote:

>On the basis of this precedent, and on the basis of the fact that sample
>glyphs on old copies Unicode book's will have a strong influence on font
>designers for years, it may be wise in this case to leave the old
>U+0643-like character alone and add a new U+06A9-like characters.

There are only a handful of Arabic fonts that support U+06AC, and the
developers I have heard from, directly and indirectly, are aware that the
form shown in the book is *rare* in Jawi text. Note that, as Amir pointed
out, the form is not incorrect, it is simply much less common than the
other form.

I wouldn't worry about the form of the glyphs to render this character.
This is a font issue, and at the end of the day, if a user doesn't like
what he sees he can go in search of a different font.

I would be much more concerned about the existence of text that uses U+06AC
for this Jawi character. The Unicode standard very clearly identifies this
character as 'Old Malay' and, as Ken pointed out, this is the intention of
the standard: that U+06AC be used for the Jawi ga. Do you really want to
turn around and tell developers that no, they should now start using a
different codepoint for this character? Fixing a couple of composite glyphs
in a font is much easier than worrying about whether U+06AC in a text
string might need to be cross-mapped to a new character.

>I would say that this would be a prudent choice in the case of an
>international script like Arabic: who can make an oath that no language's
>orthography ever used a letter like U+0643 with a dot above?

True enough, but that language was not 'Old Malay'. If you want to maintain
a character for U+0643 with dot above, I would recommend adding *this* as
the new character, and more clearly identify U+06AC as the preferred Jawi
form. Personally, I would wait until someone provided evidence of the use
of such a character, rather than making assumptions. Font developers
certainly won't thank you for encoding arbitrary characters requiring the
design of glyphs that will never be used: it is a waste of our time and of
our clients' money.

John Hudson

Tiro Typeworks www.tiro.com
Vancouver, BC tiro@tiro.com

If meaning is inherently public and rule-governed, then the
fact that I can't read 'Treasure Island' without visualising
Long John Silver as a one-legged version of my grandmother
is of interest only to my psychotherapist and myself.
                                                   Terry Eagleton



This archive was generated by hypermail 2.1.2 : Mon May 13 2002 - 15:30:57 EDT