[vorbis] TAG Standard - ENSEMBLE/PERFORMER tags

[Glenn Maynard]

On Sun, Jan 06, 2002 at 10:34:41PM -0800, Dan Hollis wrote:
> > True.  How about this compromise?  We recommend people use
> > UTF-8 where they can, but where that leads to ambiguity (as with Asian
> > languages), they use the RFC2047 scheme?  I believe RFC2047 is upwards
> > compatible with UTF-8 anyway, so you can even use both at once if
> > the fancy takes you...
> 
> Yes, I think RFC2047 should be allowed in the tags. Please don't 
> unilaterally force UTF-8 on everyone.

Please don't unilaterally force RFC2047 on everyone.

That is, remember that if it's allowed at all, then every player out
there will have to support it, including embedded ones.  You need a way
to convert charsets; I have a copy of iconv.dll, a standalone library
to do this.  It's about 600k--that's huge for embedded devices.

In practice, this would probably never be used; editors would most
likely revert everything back to UTF-8.  However, every player would
still *have* to support it.  This isn't like an extra tag, which can
simply be ignored if an implementor doesn't want to support it.

There's no way of telling if "=?ISO-8859-1?Q?hello?=" is RFC2047 or just
a tag that says "=?ISO-8859-1?Q?hello?=".  So, any real tag that could
be misinterpreted as encoded in this way would have to be encoded in
this way.  (Mail clients do this.)  Yuck.

Changing encodings is not an appropriate way to indicate language.

A key benefit of the way the tags are now is they're completely
human-readable.  You can just dump them to stdout and it'll all make
perfect sense.  That all goes flying out the window if you introduce
other encodings or RFC2047.

RFC2047 is a gross hack.  Before there was UTF-8, you *needed* multiple
encodings because every language had its own.  That is no longer true,
so the hack is no longer necessary.

RFC2047 is a *mail* standard.  MIME.  Realize that it's not a
general-purpose encoding; it's not intended as such and it wouldn't be
very good at it.

You lose one of the greatest benefits of UTF-8: getting rid of all of
those other encodings.

This would be an incompatible change; it's been made clear by many that
we don't want to make the basic tag structure change incompatibly.

Note that this is nothing new--the tags have been UTF-8 for a long time.
In the spec, not the tag type proposal.


-- 
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.