Why MusicXML export?

This page should list/discuss use cases for MusicXML export.
Who needs it and what for?
Why should LilyPond implement it (or why is it a minor issue)?

As a start I will present a few arguments (Urs Liska):

• Currently using LilyPond is a one-way street with the only exit being the graphical output formats (i.e. PDF, EPS or PNG)
  It isn't polite behaviour for a program not to allow exchange with other programs and this fact negatively affects LilyPond's acceptance in the music business.

• There are relatively few music publishers who accept delivery of scores in PDF format because they want to be able to edit the scores even after the initial publication.

• If one happens to be able to work for one of them this also implies that one is responsible for tuning the scores to publication quality oneself. Which is something not everybody (or better: rather few) can provide.

• Among the majority of publishers who insist on editable files there are very few who accept LilyPond sources as delivery medium:

  • They have tested and reliable workflows based on other (proprietary) software
  • They have their 'house styles', based on their existing workflows
  • Their staff is familiar with other software
  • They need a guarantee that they will also later have people being able to work on the scores

• We won't be (easily or at all) able to convince a publishing house to change their usual working style.
  So this effectively means that currently anybody intending to work for such a publishing house is unable to do it with LilyPond.

• I am absolutely convinced that LilyPond's text based approach offers unique advantages for workflows in preparing musical editions. In order to endorse this conviction in the editorial world it's crucial to offer a possibility to prepare scores with LilyPond that will nevertheless end up with a publisher working with other software.
  The faint hope in this is that if more editors start chosing to prepare their scores with LilyPond it might someday become evident that it is obsolete to have an extra step with other software in between.

Further remarks:

David Kastrup:
The question is what MusicXML would be a selling point for.

1. you can export your existing scores to LilyPond

2. you will be able to export your LilyPond scores to other applications
3. if convert-ly does not manage a conversion between LilyPond versions, maybe exporting and reimporting MusicXML will
4. you can run a MusicXML business with LilyPond doing the hard work in the background

Which of those points might be important for whom?
And who of those would be able to provide enough funds to pay for external work (work that would not otherwise be happening)?

Urs Liska:
ad 3): I can't really imagine a MusicXML im-/exporter will be more successful than convert-ly ...
ad 2): This is the main point of my arguments in the list at the top of the page
But exporting LilyPond scores may also be useful for other target groups: People who are interested preparing scores with LilyPond because of its versioning and other structural potentials who would like to benefit of other programs' capabilities that LilyPond won't offer in the foreseeable future.

• Producing better MIDI/Audio files

• Using advanced arrangement functions (automatic orchestration/piano reduction etc.)

Continue reading:
Requirements and development ideas