Salaam again Mohammed, > And what about the small alif (8 in the proposal)? > I think it's now clear that it's like U+06E6.
The floating superscript alif is in the same category as the floating hamza (found in al-aakhirat in the Quran). Its visual behaviour is predictable from the characters adjacent to it. This hamza is graphemically the same as any other independent hamza. If you write aakhirat without "al" article preceding it, the hamza in the beginning of the word aakhirat is still the same hamza as the hamza in the word "al-aakhirat". Encoding this hamza different than any other regular chairless hamza is very illogical. Imagine you type the word aakhirat with a regular hamza in the beginning. And then you add Al in front of it. Now should you have to go and replace that hamza with a special code to trigger lam+hamza+alef visual behaviour found in the Quran? That would be contradictory to the principle of graphemic encoding. Similarly the special behaviour of superscript alif is triggered by a preceding fatha, therefore it would be against the principle of graphemic encoding to use a different code for superscript alif there than the superscript alif found in other places. What needs to be done is that a proposal should be submitted to Unicode to add a new "contexual behaviour" to superscript alif U+0670. The added contexual behaviour would be that when a fatha precedes a superscript alif, the referred special visual behaviour is triggered. Insha'Allah this would be one of the items in our joint proposal to Unicode. As to independent hamza U+0621, unfortunately it is too late now to add a new contexual behaviour for the special visual behaviour found in al-aakhirat. This is because U+0621 is alredy used in Farsi to cause non-joining behaviour for the letters adjacent to it, and adding a new contexual behaviour to U+0621 would break existing Farsi texts. Since Unicode has to be backwards-compatible, we cannot add this new contexual behaviour but we have to add a new chairless hamza character that is conditionally equivalent to U+0621. But for fatha+superscript_alef there isn't such a situation (as far as I know) so insha'Allah we should be able to add the new contexual behaviour to superscript alef without problems. > And what about 9,10,11 and 12? 9 and 10 are already covered by the proposed logical code to trigger sequential fathatan behaviour. 11 and 12 are already covered by item 6 in your proposal. Kind Regards, Mete _______________________________________________ General mailing list [EMAIL PROTECTED] http://lists.arabeyes.org/mailman/listinfo/general
