From: Chris Jacobs (chris.jacobs@freeler.nl)
Date: Mon Apr 04 2005 - 06:36:56 CST
> Well, yes and no. I remember being told repeatedly that it is quite
> permissible and conformant for an implementation to support only a
> subset of Unicode characters. But of course that only worked with a
> subset closed under the operations performed by the implementation. Now,
> suddenly, without any announcement or apparent consideration of the
> implications, the subset is no longer closed under normalisation.
>
> --
> Peter Kirk
> peter@qaya.org (personal)
> peterkirk@qaya.org (work)
> http://www.qaya.org/
It is for interpreting and rendering that it is allowed to support only a
subset. But e.g. for storing you have to support the whole set.
I don't see why the subset should be closed. Suppose I have a simple font
which can display ä but not ä, then I should still be able to normalize to
decomposed if I want to.
This archive was generated by hypermail 2.1.5 : Mon Apr 04 2005 - 06:40:24 CST