[Tokkemon]

The format is just so underpowered for what music notation requires, unfortunately. Michael Good pulled off a miracle writing the standard, but it doesn't do enough and now we have to deal with those shortcomings by fixing thousands of files that were perfectly fine.

[ksr]

You're uniquely positioned to comment on this, of course. Are there other music notation formats that do a better job (MEI, MNX, etc.) ?

I've been working with MusicXML as a consumer (and occasional contributor) in my own apps for a few years now, and my pet theory is that the _problem domain_ is so complex that any format will fall short of the universe of use cases - or become so cumbersome that its application to a codebase will forever remain buggy / incomplete.

[Tokkemon]

MNX is probably good but they are taking an awful long time to get the job done. They ran into the same roadblocks as any format, there's too many competing needs and too many opinions. MIDI 2.0 has also had some really interesting talk about embedding notation data inside the wrapper to make it an "all-encompassing" music format.

I'm not as familiar with MEI but I don't believe it's any better for preserving music engraving data compared to the musical data itself.