Re: [Synedit-devel] Multibyte-issues

[<ma...@ar...>]

>And if we decide to use UTF-8, (it seems) DBCS will be almost integrally
reused.
>(That said, DBCS is a pain because no one can test it!)
I think WideString(UTF-16) would be better than UTF-8 because Windows (al=
so
9X has some APIs) and CLX supports it natively and we could remove MBCS/D=
BCS
handling.
As I already often said: I did partial conversion and this worked quite w=
ell
and didn't show the problems there are with MBCS. (You can switch on Chin=
ese
in Windows XP and test this way).

For all important characters, handling would be again one index =3D one
character. (except perhaps for those strange things that only have a dire=
ct
representation in utf-32 but are definitley out of the scope of SynEdit, =
not
even mentioning there is merely a font that can represent those extra
characters (these characters are mainly from dead languages, no not latin
:-), and some very specific ones).
UTF-8 should IMHO only be used for data transfer but not for data
processing.

>So, now, what do think?
>Do "2" right now, or put this the list of 'known issues' do "2" after th=
e
release?
>(No, you don't want "1" ;-)
I would suggest we wait until we got your trademark registered and put th=
is
in the known issues. Hard tabs are still noted as a feature in the readme
:-)
Also I fear that this will take some time and/or show up some new problem=
s.

cheers, Ma=EBl.