Re: Sketch for broken beams with consistent slopes (issue 4961041)

[mtsolo]

Hey all,

This is now in a state where people can run regtests on it and comment
upon it.

There are a few things in the regtests that I still need to work on:

-) my feathered beam regtest has more vertical spacing between
systems...I still need to figure out why
-) i think the change in spanner-break-overshoot is kinda cool, but i
need to figure out a way to scrub the warning message.
-) beam-slope-stemlet, perennial thorn in my side, shows a change that i
need to scrub.

What I'd appreciate from people at this stage of the game are four
things:

1) Regtest running to confirm that my checks are valid.
2) General comments on design (properties I could add to make it more
tweakable, etc.).
3) A critical eye at dividing by staff_space.  You'll see that in
beam-quanting.cc, there are several Real values that are divided by
staff_space.  I've avoided doing this in all the new stuff I've added,
and it seems to have had no effect, but if someone could create a good
test file I could use to check the effects of staff space on beam
positions, I'd be much obliged.  Or, alternatively, for someone who gets
staff_space better than I do, if they spot any obvious problems, please
let me know.
4) Figuring out cases that break this code.  I'm sure that crazy broken
cross staff beams lead to odd output, but I'll deal with esoteric
cross-staff stuff in another patch if I need to.  Mostly, please try to
think of odd cases that may cause my work to segfault or return bad
beams.

Thank you all!

Cheers,
MS

http://codereview.appspot.com/4961041/