Re: Music Notation on the Web - Last Call?

Hello,

> I must confess that I'm about to give up on this -- basically because I seem to be out there completely alone.  I have a pretty strong feeling that something really needs to be done, but I don't seem to be striking a lot of chords with people.  An analogy that occurs to me:  The W3C people had a tendency to think that the XML representation of RDF solved all problems -- except in practice just about nobody actually used it, or perhaps only used it as a last resort.  Now they have finally figured out that recognizing and standardizing Turtle (which is much more compressed, readable and easy to write for simple tasks) is a REALLY good idea.  You MusicXML guys seem to be saying that MusicXML can do anything anybody wants -- but for some reason it really doesn't seem to have much uptake.  For example, you note that there are 94 MusicXML scores in CPDL.  Well, there are links to 84 ABC scores.  Neither number is very impressive and they are all external links -- CPDL does not "natively" offer either.
> 
> If there are people lurking out there who would like to see something happen here it would be nice to hear from you.
> 

I identify myself as a lurker, but I would like to see something happening in that space. MusicXML is good, but it lacks two things IMHO: the ability of linking between musical scores, and the ability to deal with non-Western music. Your mention of RDF strikes a chord (hem...) here - I definitely think that a RDF-based framework for musical score would enable some very interesting use cases. For example:

* Linking musical scores to other music-related data (e.g. performance information, particular recordings, etc.) - it would be very nice to be able to link a particular phrase in a score to the part of a recording it corresponds to.
* Interlinking musical scores - it would be great to be able to link similar phrases in different scores.

I made a couple of related experiments a couple of years ago, trying to push musical scores in the Music Ontology (http://musicontology.com) framework. It is just a braindump though, nothing really serious yet. The main idea was to use an event-based model to describe score information, which would allow a fair degree of flexibility in the sort of musical events you want to represent.

http://motools.svn.sourceforge.net/viewvc/motools/so/rdf/so.n3?revision=372&view=markup

and an example of the first bars of Debussy's Syrinx:

http://motools.svn.sourceforge.net/viewvc/motools/so/examples/syrinx.n3?revision=372&view=markup

All this is based on the Event Ontology, which we use quite extensively at the BBC:

http://motools.sourceforge.net/event/event.html

Best,
y

> -----Original Message-----
> From: public-xg-audio-request@w3.org [mailto:public-xg-audio-request@w3.org] On Behalf Of Dan Brickley
> Sent: Tuesday, December 14, 2010 11:44 AM
> To: Michael Good
> Cc: public-xg-audio@w3.org
> Subject: Re: Music Notation on the Web
> 
> On Tue, Dec 14, 2010 at 6:27 PM, Michael Good <musicxml@gmail.com> wrote:
> > Hi Joe,
> >
> > You may well be right about this, but they are perceived issues even if not
> > real issues. I think it's best to be able to go to site owners and say
> > "we've fixed your problem" rather than saying "that's not really a problem."
> > Sometimes just the aesthetics of space inefficiency are enough to make it a
> > problem.
> >
> > The compressed file format offers many other advantages anyway. This
> > includes keeping linked/included images together with scores in a single
> > file, and offering a dedicated .mxl suffix rather than a generic .xml
> > suffix. The tradeoff is that it's a binary file rather than a text file,
> > albeit a very well-understood, standardized binary format (vanilla,
> > Java-compatible zip files).
> 
> For what it's worth, this was the design also taken by the W3C Widgets
> group, see Widget packaging spec, http://www.w3.org/TR/widgets/
> 
> I think they had some headaches figuring out how exactly to cite Zip
> from a formal W3C spec, but http://www.w3.org/TR/widgets/#zip-archive
> is the current text.
> 
> More than a few ebooks formats do the same I'm sure,
> http://en.wikipedia.org/wiki/Comparison_of_e-book_formats ...
> 
> cheers,
> 
> Dan
> 
>

Received on Wednesday, 15 December 2010 10:24:49 UTC