Re: RFC: new vertical layout engine

[Neil Puttock]

Hi Joe,

Here's an attachment with three examples for you to test.

`L'Islejoyeuse.ly' is the file which triggers the assertion failure;
compiled with optimisations, there's a particularly nasty collision on
page nine between the third and fourth systems, seemingly triggered by
a combination of cross-staff beams and a hairpin.

`sonata_no5.ly' isn't as bad as I remember it being with earlier
versions of the new code, but there are still systems spilling off the
bottom of pages one and sixteen (using fixed distances).

Rather strangely, the example I was thinking of in relation to weird
beaming errors triggering spacing changes seems to have righted itself
both with master and dev/jneeman, so instead I've added an orchestral
score (`prelude.ly') which veers from very tight to loose and spilling
off the page.  There are many broken beams in the harp parts,
associated with weird stretching of all the staves (see page twenty).

Regards,
Neil

test-files.tar.bz2
Description: BZip2 compressed data