[Ffmpeg-devel] AVOptions

Måns Rullgård mru
Wed Sep 14 11:17:58 CEST 2005


Fran?ois Revol said:
>> "Fran?ois Revol" <revol at free.fr> writes:
>>
>> >> "Dario Andrade" <dario at ip.tv> writes:
>> >>
>> >> >> The point was that you can't use strcasecmp, because setlocale
>> > > > > can
>> >> >> > break it.
>> >> >>
>> >> >> It could, but AFAIK all commonly used encodings are ASCII
>> >> > > compatible.
>> >> >> Things like UTF-16 are rarely used.
>> >> >
>> >> > UTF8 is becoming pretty popular in Java.
>> >>
>> >> UTF-8 is becoming popular in many places.  Java uses utf-16
>> >> internally, well in line with all the other absurd things it does.
>> >
>> > BeOS has been using UTF-8 by default everywhere for years...
>>
>> That's of course good.  I've also been using utf-8 for quite a while
>> in Linux.  The problem is when people insist on sending mail (usually
>> with ms outlook) with the wrong encoding specified in the headers.
>> Similarly with web pages, ID3 tags, etc.
>
> Well, for web & mail there is an encoding information so you can always
> convert it.

Provided the encoding information in the headers is correct, yes.  I often
get mails purporting to be encoded with iso-8859-1, while in reality they
something entirely different.  Likewise with web pages.

> The problem is with thing like ID3 which doesn't tell which encoding it
> uses  (AFAIK).

ID3v2 has encoding information.  It supports iso-8859-1, utf-8 and utf-16.
Players like winamp seem to ignore this and assume iso-8859-1.  ID3v1
encodings are random, usually being the traditional encoding of the region
where the file originated.

> Same with shoutcasts servers... most of the time it's
> iso-latin-1 because they run on european/us NT servers, but I already
> saw other encoding, and no way of guessing them (except with
> statistical methods like Emacs does when opening a file but that's overly
> complex and not 100% safe).

At least emacs tends to get it right.  Firefox almost exclusively gets it
wrong.  It keeps insisting that pages using iso-8859-1 are in some far
eastern encoding it happens to have seen recently.

-- 
M?ns Rullg?rd
mru at inprovide.com





More information about the ffmpeg-devel mailing list