On 06/11/2007 11:27 AM, Ludwig Nussel wrote:
> Klaus Schmidinger wrote:
>> On 06/11/2007 10:08 AM, Oleg Roitburd wrote:
>>> Am Sonntag, 10. Juni 2007 15:45 schrieb Klaus Schmidinger:
>>>> VDR developer version 1.5.3 is now available at
>>>>     ftp://ftp.cadsoft.de/vdr/Developer/vdr-1.5.3.tar.bz2
>>> Now I'm running this version.
>>> And I would say: Converting of EPG to UTF-8 don't work
>>> My system has  DVB-T card and I'm receiving only German "ÖRP" 
>>> In schedule menuitem I can't see any umlaut or "ß" 
>>> $LANG is setting to en_US.UTF-8
>> Please check whether the log says that the codeset is "known"
>> (should be right after VDR's startup message).
>> You may need to set LANG=en_US.utf8 (without the '-').
>> Maybe the strings in libsi/si.c should contain the extra '-'s,
>> not sure what's best.
> $ LC_ALL=en_US.utf8 locale collate-codeset
> UTF-8
> so the canonical spelling is "UTF-8". As already suggested using
> nl_langinfo() instead of manually parsing enviroment variables is
> probably the better choice though.

Thanks, I'll change it then.

>>> For my stable version 1.4.x I'm using UTF patch from Alexander Riedel. As I 
>>> remember he has inserted extra field for channel codepage. And this works. 
>> Using such an extra field is not the way to go, because there is no
>> standard way to fill it in (or is there?).
> ACK, considering that libsi can automatically determine the encoding
> for individual strings it seems to be awkward having to specify
> encodings per channel in channels.conf.

The latest version of libsi does use the character encodings specified
in the SI data for each string. Unfortunately, though, a certain German
pay tv broadcaster doesn't adhere to the standard and sends the data in
iso8859, but without marking the strings accordingly. So VDR assumes
that they are ISO6937 (which is the default if no encoding is specified)
and therefore converts the umlauts to garbage.
But that's their fault, not VDR's...


vdr mailing list

Reply via email to