Re: ly2musicxml / possible sponsoring

[Paul]

Hello Christoph,

On 01/27/2017 08:14 AM, Christoph Hagemann wrote:
> thanks for your interest. Please excuse my delayed reply, busy with
> real life...

No worries, same here on my side...

> Jan-Peter Voigt and Urs Liska replied off-list. They recommended
> enhancing frescobaldi's export (python-ly). Which way is more apt?

I'd say in the long run, and maybe even in the shorter run, the 
LilyPond/Scheme approach is more promising.  As I understand it 
python-ly works on LilyPond's text input format so you're basically 
rewriting a chunk of LilyPond to process that into the musical data 
"itself".  Whereas the LilyPond/Scheme approach starts from LilyPond 
having already created that data from the input format.  (Not sure what 
Urs has in mind with combining the two approaches.)

Also, as I understand it the GSOC project got to a state that was 
starting to be useful.  I believe it was in basic working order for 
non-polyphonic music, the big challenge facing it was how to handle 
polyphony / simultaneous music.  And I understand your piece is not 
polyphonic, right?  So that improves the prospects for a useful outcome.

I'm cc'ing David Garfinkle, as it was his project.  David, it would be 
great to hear your input on the state of things.

One question is the release of LilyPond 2.20 and the level of guile2 
support.  (There are significant performance slow-downs with guile2 and 
some other issues...)  If guile2 works well enough, albeit as a slower, 
likely optional thing, then that paves the way for musicxml output.

> There is no fixed deadline, but can't hold the publisher on standby
> forever. As a guideline: I'd like the export to be working within six
> month' time.

Ok, good to know.  I think that's at least within the realm of 
possibility, and worth further discussion.

Cheers,
-Paul