[vdr] Re: [PATCH] fix segfaults because of channels with umlauts

[Nicolas Huillard <nhuillard@ghs.fr>]

Klaus Schmidinger wrote:

> "Dr. Werner Fink" wrote:
>
> > On Mon, Jun 14, 2004 at 01:47:31PM +0200, Klaus Schmidinger wrote:
> >
> > > "Jörg Knitter" wrote:
> > >
> > > > Klaus Schmidinger wrote:
> > > >
> > > > > If I ever install a Linux distro that has UTF-8 as default, I'll probably
> > > > > do the same ;-)
> > > > If I see it correctly, SuSE 9.1 also seems to use UTF-8 - at least this is
> > > > set per default in the Language Settings of YaST.
> > > It would seem so - and I find that a major PITA!
> > Please explain _why_ do you think so?  What is wrong
> > on text like this one:
> >
> > Arabic اÙ?سÙ?اÙ? عÙ?ÙSÙfÙ?
> > Greek (Î*λληνικά) Î?ειά ÏfαÏ,
> > Hebrew ש×?Ö¸×?×*Ö¹×?
> > Japanese (æ-¥æ?¬èª?) ã??ã,?ã?«ã?¡ã?¯, ｺï¾?ï¾?ï¾?ï¾S
> > Chinese (中æ-?,æ?®é?sè¯?,æ±?语) ä½ å¥½
> > Cantonese (ç²µèª?,廣æ?±è©±) æ-©æ?¨, ä½ å¥½
> >
> > ... no problem to read this within an UTF-8 locale ;^)
> >
> >        Werner
>
> Wow, I didn't know you speak that many languages.
>
> I for one need only German and English.
As a developper, you need to take care of that for your users...
I think most problems with UTF-8 come from the fact that many 
developpers think it's too complicated, not usefull, etc. and then use 
UTF-8 in a half manner, even with wrong code, etc. That lead to 
effectively complicated interoperation.
WRT VDR, the implications are :
* if EPG or channel names come with no character coding specification, 
it won't be easy at to do any character transocding...
* the great localisation effort of VDR will bring all sorts of encoding 
problems : there will actually be files in VDR that will have many 
encodings at once
* taking UTF-8 into account can be as simple as warning the user it's 
not supported, and set a well-defined LOCALE in the runvdr script : that 
way, the main vdr program won't have to transcode anything...

NH