volleyball issue 1336

[Graham Percival]

This is an attempt to collaboratively solve issue 1336.  (that said,
if anybody can jump in and solve it instantly with a patch, go right
ahead!  if this issue is solved, we'll just pick another one)
http://code.google.com/p/lilypond/issues/detail?id=1336

There is a theory about why this happens, from Neil:
> It looks to me like the MultiMeasureRest has invalid bounds when cloned, so 
> the cloned object is NULL when the MultiMeasureRestNumber is aligned (the 
> parent object no longer exists)

Can anybody think of a way to test this theory?  i.e. I'm not asking
how to stop this from happening.  I'm not even asking for solid
evidence that this is, in fact, happening.  I'm just asking how we
could *begin* the *attempt* to gather such evidence.

The answer might be something like:
- "cloning is done by lily/grob-clone.cc"
- "hey, I've found a method in lily/multi-measure-engraver.cc called
'clone()', we should probably look there"
- "did you try looking at scm/define-grobs.scm, which is where
MultiMeasureRest is defined?"
- "'clone' is a basic term of scheme programming, you idiot, why on
earth are you trying to do any lilypond programming if you don't know
even know that?"

Any of those answers would be an improvement on my current knowledge
of this bug, and would be appreciated.  If nobody has any ideas I'll
try grepping for MultiMeasureRest and clone, and start poking around
blindly.

Cheers,
- Graham