Vorbis - Dec 2001 - tags in comment field - why?

Well, my last question about this was in the middle of other stuff, but
it's a basic one: why put tag info in the comment field at all?  Why not
use the XML field that's already there?  It gives the flexibility needed
to do things like translated tags easily, and keeps tag info in the same
format as other data--potentially simplifying editors which read other
types of metadata.

The only reason I've come up with is to simplify reading tags for
embedded devices who don't want an XML parser.  A better solution for
this, I think, would be to define the tags as XML and then export the
data to the comment field, reducing it as necessary (ie. where there are
multiple languages, use one based on user preferences), and always
considering the XML version authoritative.

Another thing to bear in mind is that UTF-8 data alone isn't always enough
to render it correctly.  You need to know what language it's in, too,
when you have CJK text.  You can guess--chances are, if it's my file,
it's Japanese, not Chinese--and less intelligent renderers will just 
select a font and if it's the wrong one, display the wrong thing.
However, it'd be nice to allow it to be done right; to allow programs to
let me choose MS Gothic for Japanese and NSimSun for Chinese.  (IE allows
this.)

Why put tags in a separate, severely limited format?  ID3V1 did that
(at orders of magnitude worse, of course), and the switch to ID3V2 was
painful.  


-- 
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@xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.

On 2001-12-26, Glenn Maynard wrote:
> Well, my last question about this was in the middle of other stuff, but
> it's a basic one: why put tag info in the comment field at all?  Why
not
> use the XML field that's already there?  It gives the flexibility
needed
> to do things like translated tags easily, and keeps tag info in the same
> format as other data--potentially simplifying editors which read other
> types of metadata.
>
> The only reason I've come up with is to simplify reading tags for
> embedded devices who don't want an XML parser.  A better solution for
> this, I think, would be to define the tags as XML and then export the
> data to the comment field, reducing it as necessary (ie. where there are
> multiple languages, use one based on user preferences), and always
> considering the XML version authoritative.
>
> Another thing to bear in mind is that UTF-8 data alone isn't always
enough
> to render it correctly.  You need to know what language it's in, too,
> when you have CJK text.  You can guess--chances are, if it's my file,
> it's Japanese, not Chinese--and less intelligent renderers will just
> select a font and if it's the wrong one, display the wrong thing.
> However, it'd be nice to allow it to be done right; to allow programs
to
> let me choose MS Gothic for Japanese and NSimSun for Chinese.  (IE allows
> this.)
>
Erh...  Good point.  That's a question to Unicode, though.  Why did they
do it this way?  I thought there is single glyph per unicode character but
you know better...  What format can currently do better than unicode?  How
does IE do it - it checks the language of the file?  Then what about a
Japanese/Chinese mix?  Or does it check the LANG properties in HTML tags
delimiting parts of the text?
> Why put tags in a separate, severely limited format?  ID3V1 did that
> (at orders of magnitude worse, of course), and the switch to ID3V2 was
> painful.
>
ID3v2 didn't fix it.  It's still more limited than vorbis tags.  I
won't
speak for Monty on the original decision but using XML is overkill.


-- 
Beni Cherniavsky <cben@tx.technion.ac.il>
                 (also scben@t2 in Technion)
I like Common Lisp* more than Common Source
   and Open Source  more than Open Collector.
(*)Scheme is better than any scheme.

<p><p>--- >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@xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.