On Thu, 26 Apr 2001, Marco Cimarosti wrote:

> As a second thought, using codes higher than 0x0FFFFFFF is even safer,
> because it also accounts for the fact that, theoretically, ISO 10646 uses 31
> bits.

But this was the Unicode mailing list ;)

> Of course, all this is only possible for applications that use 32-bit
> quantities to handle Unicode characters.

Surrogating the noncharacters in the FDD0..FDEF range works for internally
16-bit apps.

--roozbeh


Reply via email to