On Jan 24, 2011, at 9:28 AM, Matthias Melcher wrote:
> 
> On 24.01.2011, at 17:50, Michael Sweet wrote:
> 
>> Unicode isn't 16-bits - you need to allow for 20-bits.
> 
> He allows for 20 bits, but only caches the character width of the first 65536 
> characters. Maybe a hashing scheme could be used?

The current scheme is already hashed (bits 8-15 to select a bin of 256 
characters) so I'm not sure how we'd measurably improve on that. However, my 
point is just that if you are going to do this, do it right and handle 20 bits 
to cover the range of Unicode characters supported by the OS.  Otherwise we are 
going to get bugs...

________________________________________
Michael Sweet, Easy Software Products

_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev

Reply via email to