Another spin-off discussion. The scheme used on
<http://wiki.xiph.org/index.php/Metadata> was to
packetize XML (split it up into packets at the
root level, each of which was self contained
(the terminology has temporarily abandoned me),
and have the bos packet as the opening <?xml
&c. stanza, the eos a </xml>.
It is probably more sensible to include a magic
number identifying it as a metadata stream
and probably a version number. I wouldn't
anticipate that version number being bumped
in the near future as it would mostly guarantee
simply:
1. An XML metadata stream.
2. That some form of the currently debated
metadata format would be present.
Anything I've missed?
This only leaves arguing over the magic number.
I like MOG (Metadata in OGg), but it's far too
short...
--
imalone
I suggest using the solution that CMML has come to use. The XML file is essentially the same as an unencapsulated physical bitstream. Then there is a mapping into a logical bitstream, where some of the default information - in particular the XML header - are split off and put into the bos packet - nothing really needs to go into the eos packet. There's also a magic number and a version number. Also, use the granulepos scheme that we defined for CMML pages- you're going to make your lives easier. CMML has had several years of fine tuning of the means to encapsulate. Don't underestimate the problems - it's very easy to get the impression that XML is simple to handle, while it's actually quite hard to get right. So much for technical feedback. Now I just ask myself the question as to what "Ogg Metadata" really is - since that wiki page has no actual technical specification on it at all... Silvia. On 9/12/07, Ian Malone <ibmalone@gmail.com> wrote:> Another spin-off discussion. The scheme used on > <http://wiki.xiph.org/index.php/Metadata> was to > packetize XML (split it up into packets at the > root level, each of which was self contained > (the terminology has temporarily abandoned me), > and have the bos packet as the opening <?xml > &c. stanza, the eos a </xml>. > > It is probably more sensible to include a magic > number identifying it as a metadata stream > and probably a version number. I wouldn't > anticipate that version number being bumped > in the near future as it would mostly guarantee > simply: > 1. An XML metadata stream. > 2. That some form of the currently debated > metadata format would be present. > > Anything I've missed? > > This only leaves arguing over the magic number. > I like MOG (Metadata in OGg), but it's far too > short... > > -- > imalone > _______________________________________________ > ogg-dev mailing list > ogg-dev@xiph.org > http://lists.xiph.org/mailman/listinfo/ogg-dev >
On Wednesday 12. September 2007 09:49:22 Silvia Pfeiffer wrote:> Now I just ask myself the question as to what "Ogg Metadata" really is > - since that wiki page has no actual technical specification on it at > all...Metadata is basically what RDF does: Describes a resource. In the case of Ogg files; metadata would describe the media [resource] in the Ogg container. Notably a description can be many things. Technical aspects (such as quality, aspect ratio, bitrate, software used to create the file, ...) are machine-metadata whilst media information (such as artist, composer, director, record label, musicans, ...) would be human-metadata. -- Daniel Aleksandersen