Re: controlling digits substitution in IE/FF (Arabic/Hindi/Decimal)

On 5/14/2009 1:49 AM, Jeroen Ruigrok van der Werven wrote:
> -On [20090514 09:59], Richard Ishida ( wrote:
>> I guess the bigger question here is "Should it be possible to represent
>> European digit characters with different glyphs (that correspond to other
>> Unicode characters)" ?
Ideally not.

However, certain 8-bit character sets did not have any Arabic-Hindi 
digits, and therefore, when mapping or displaying legacy data from these 
character sets, one must allow for both interpretations of the ASCII digits.

Sometimes, you have Unicode data, but it's impossible to tell whether 
any ASCII digits in the text truly correspond to ASCII digits, or 
whether some where converted from a legacy set with ambiguous usage - 
without being properly resolved.

Similar situations exist for other types of legacy from other regions. 
(Character 0x5C comes to mind...).

As I said, not ideal, but sometimes a practical necessity.

That said, where platform-specific support exists to handle 
legacy-related issues, it would of course be ideal if all 
implementations for those platforms supported them equally.


Received on Thursday, 14 May 2009 13:36:44 UTC