[vorbis] TAG Standard - ENSEMBLE/PERFORMER tags

[Glenn Maynard]

On Tue, Jan 08, 2002 at 09:14:38PM -0800, Dan Hollis wrote:
> > Oggenc uses iconv to convert local encodings to UTF-8. If other
> > encoders follow suit, then this mess will probably not happen.
> 
> That's what they said about ID3!

Neither ID3 allowed specifying the language, either.

> Everyone ignored it and just used local encodings instead. Noone uses 
> UTF-8, they all use EUC-TW, BIG5, SJIS, etc. Any program which wants to 
> display these has to guess (as does kterm, netscape, etc). And unless you 
> have got encoding tagging like RFC2047, you've got a real mess.

(And if you do, you have lots of encodings, which is also a mess.)

Is there any valid reason Japanese people don't like Unicode?  As far as
I can tell, as long as the text is labelled with the language, you've
got all you need, and nobody's come forward and given an example to
counter this.

If someone did, I'd bring it up elsewhere with people more familiar with
Unicode.  As is, I've asked about this elsewhere, but I only did so about
an hour ago; there hasn't been time for many responses.  (One response,
from hpa, saying that most of the complaints are Japanese, only a few are
Chinese and that he's never heard complaints from Koreans.)

> I'll say it again - if you *force* UTF-8 as the only encoding allowed, 
> everyone will just ignore you. Period. Just like everyone's ignoring me 
> now :-(

We're not ignoring each other, we're discussing it.

> > Admittedly, I was a little dismayed to see that iso-8859-1 and
> > iso-8859-2 were the only two local encodings supported. It wasn't
> > too difficult for me to get euc-jp in there, though... but that
> > is certainly not a task which all users are capable of doing. I
> > should probably offer up some patches that squeeze more non-euro
> > encodings in...

It needs to be able to deal with *any* encoding the local system might
have.  Why does it need to implement any encodings directly?  (Well,
it's reasonable to implement UTF-8 directly, in case the local system
doesn't support it, for convenience, etc.) It's extremely important
that the utilities be just as functional in any locale the user might be
using.

> And JIS/SJIS/EUC-TW/BIG5/KOI8/etc... ;P

(like these)

> But really, if everyone is so hog wild about UTF-8, then 
> oggenc/vorbiscomment/etc had better be the poster boy for handling local 
> encodings -> UTF-8.

Er, all you need is iconv.  It's easy to make an application that runs
in the local locale (ie. SJIS) and handles files in UTF-8.  Someone with
LC_CTYPE=ja_JP.ISO-2022-JP should be able to run the encoder or tagger
with JIS text and the tag should be written properly; there's no question
about that.

Likewise, any library that reads Ogg tags (especially a reference one!)
should probably take and return all data in LC_CTYPE, not UTF-8.  This
is also extremely important: if you say "all data passed to these functions
must be UTF-8", people will indeed ignore it and pass it locale-encoded
data.  (It's not the app's job to do that kind of conversion, either; if a
library wants everything internally and in its own data formats to be in
UTF-8, it should deal with converting the data automatically.)


-- 
Glenn Maynard

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.