Dear Behnam,
No, this is another story. The sad news is that there are multiple implementations of Unicode in Mac OS X. WebKit (The engine of Safari) has its own Unicode/Bidi engine. Cocoa has its own Unicode with no native Bidi with some ugly Carbon ATSUI patches bolted on and some ICU thrown in to get limited Bidi. Carbon uses an incomplete and degraded implementation of ATSUI which is a downgraded and crippled version of QuickDraw GX layout engine of system 7 days. That is not all. I really hope Apple will start to clean up this extremely ugly mess, otherwise they will be forced out of bidi markets for good. It is amazing how much worse their bidi text engine is compared to 12 years ago.
The problem is that each of these have their own bugs. Sometimes the bugs are a result of the same thing being applied twice because of API layering. This is the case with Safari. In some combinations of style sheet and page tags it tends to mirror a glyph twice which will result-in no mirroring which is wrong. Actually the workaround in such case is to use a buggy font which does not have a 'prop' table (like a PC font) and then it will work because it would not be mirrored by the normal mechanism and just WebKit's extra mirroring would create the correct result.
I really hope someone at an influential Apple position would listen to me.... It really frustrates me to see Apple (who once was a pioneer in bidi and was one of the key founders of Unicode) in its current sad position in bidi support. The problems are deep rooted and want a real effort and will in high management positions to solve.
- Hooman Mehr
On Jun 12, 2004, at 7:51 PM, Behnam wrote:
Short of missing something on the list, that would be me providing alternatives to Apple standard keyboards. But they are not "fix" of existing standards. In fact, they are not standard at all! But you are right. This is a minor issue and can be fixed. I can do it for Mac community but I rather ask Apple to do it in its original issue.
My concern is more to do with different approaches in dealing with mirroring characters.
The point being, it doesn't seem to be the way mirroring characters are mapped on MS keyboards. And most of the web-pages are typed by MS keyboards. Am I on the right track?
Behnam
On 12-Jun-04, at 10:54 AM, Hooman Mehr wrote:
Hi,
I checked it and can confirm that Apple's ISIRI 2901 keyboard has a bug in this regard. The Persian opening parenthesis in ISIRI 2901 is located on shit-0 and closing parenthesis on shift-9, but Apple's implementation have them reversed. This is a minor issue. The keyboard file is an XML file that can be easily edited with sys. admin. privileges. I think someone already posted information on a fixed and enhanced Persian Mac OS X keyboard on the list.
- Hooman Mehr
_______________________________________________ PersianComputing mailing list [EMAIL PROTECTED] http://lists.sharif.edu/mailman/listinfo/persiancomputing
