[vorbis] TAG Standard - ENSEMBLE/PERFORMER tags

[Jonathan Walther]

On Mon, Jan 07, 2002 at 04:53:05PM -0500, Glenn Maynard wrote:
>I've already said that there's a very high cost.  Every encoder and
>player that decodes tags at this level *must* deal with it; it's not
>optional.

It's no harder to deal with RFC2047 than it is to deal with UTF-8.
Whereas UTF-8 characters that are not in the font will probably map to a
character such as '@', RFC2047 encodings in character sets not
recognized by the encoder will just map to the '?' character.

>What if I have text in a tag that looks like an RFC2047 string?  (Don't
>say "that probably won't happen"; that would be extremely poor design.)

Supporting encoding in the US-ASCII format is not hard.

>If that happens, I have to encode the string with RFC2047, so it'll be
>decoded back; otherwise, things will try to decode it and odd things
>happen.

Yes, that is correct.  Is that a problem for you?

>This means regular, all-UTF-8 tags can have RFC2047, so every player
>must be able to decode it.

And I've said that it is trivial to support.  I thought you had done
a decent modicum of coding before, RevZoot.  I'm having to reconsider.

>I wouldn't argue much against most additions that really could be skipped
>completely by players and editors, if it was too much of a burden.  This
>isn't one of them.

Au contraire.  A player that doesn't want to support RFC2047 can skip
right over such tags.

>Both of these reasons are true, and you're the only one claiming they're
>"bogus".  Lots of people agree the tags should be human-readable; you
>(and maybe Dan) appear to be the the only ones who think this is

It was already discovered that its not possible for tags to be human
readable in all circumstances.  The current proposal tries to make them
readable for the largest number of people.  You probably can't tell the
difference between Chinese glyphs and a string of ????, so what shows up
in your player software doesn't matter.  On the other hand a Chinese
person will probably have things set up so that Korean glyphs show up as
Korean, and Chinese as Chinese.

>UTF-8 glyphs that aren't in the local font can be displayed
>intelligently, either as a placeholder font entry, a "best-fit", or by
>simply using another font.

Have you actually tried this?

Jonathan

-------------- next part --------------
A non-text attachment was scrubbed...
Name: part
Type: application/pgp-signature
Size: 797 bytes
Desc: not available
Url : http://lists.xiph.org/pipermail/vorbis/attachments/20020107/8e0dc112/part.pgp