Vorbis dev - Apr 2008 - base64 ALBUMART vorbiscomment (was Re: [ogg-dev] Ogg/Spots and Ogg/MNG)

[Cc: changed to vorbis-dev]

On 12/04/2008, Ivo Emanuel Gon?alves <justivo at gmail.com>
wrote:
> On Fri, Apr 11, 2008 at 10:11 PM, Conrad Parker <conrad at
metadecks.org> wrote:
>  >  No, vorbiscomments are meant to be human-readable.
>
>
> But I CAN read base64!
:-)
>  It's a win-win situation, with the added bonus that it wouldn't
>  require encapsulating images in a single stream Vorbis file.  After
>  all, Vorbis multiplexed with images is likely to break
>  non-libvorbis-based decoders since those don't expect muxed streams.
>  I know, because I have been testing around Vorbis+Skeleton files.
I thought that the only decoders that have problems with such files
are those based on earlier libvorbisfile, before Tahseen and Monty
fixed it. Those based on other libraries (GStreamer, OggDSF, xine-lib,
oggz+fishsound, etc.) have no problems with playing just the audio
track.

If you've found other players/libs with problems, we should fix them ...
>  Rules are not set in stone.  We are supposed to bend them a little
>  when it's necessary and, truth be told, this is a very small
>  concession to please an audience.
If we do go down this path, let's do some testing with existing
hardware and software players :-)

cheers,

Conrad.

> I thought that the only decoders that have problems with such files
> are those based on earlier libvorbisfile, before Tahseen and Monty
> fixed it. Those based on other libraries (GStreamer, OggDSF, xine-lib,
> oggz+fishsound, etc.) have no problems with playing just the audio
> track.
Those you describe + MPlayer don't have a problem, but VLC for
instance does.  For some reason, single stream Vorbis files don't call
the Ogg demuxer used on Theora files.

VLC is the player the FSF recommends to play Ogg files (through their
PlayOgg project), which means many are using it.  That without
counting the distros that haven't updated to libvorbis 1.2.0, the HW
players, and companies that create their own closed-source decoder
implementation (I know very few, but still...).

My point, for those who are just now following the discussion on this
list, is that if we could avoid the risk of breaking players by having
users stick images in Vorbis files for something so superflous, it
would be for the best.  Thus the base64 thing.
> If you've found other players/libs with problems, we should fix them
...
I have already submitted a bug report for the VLC issue.

Another implementation which I suspect to be broken is the Helix
player, but I have to confess I haven't installed it yet.  I'm
unconsciously avoiding it after suffering for days debugging the
broken (Helix) player in the XO laptop (the OLPC thingie).  To be
fair, it wasn't Helix's fault, but still...
> If we do go down this path, let's do some testing with existing
> hardware and software players :-)
I would have it no other way.

Thanks for being open to the idea.

-Ivo