Re: Issue 1472 in lilypond: Collision between MMR and key signature

[lilypond]

Comment #15 on issue 1472 by address@hidden: Collision between MMR and  
key signature
http://code.google.com/p/lilypond/issues/detail?id=1472

Carl (re: 13 &11) I do like defaults to both handle the common case, and to  
be unsurprising.  Zero is usually the most unsurprising.  Here, though, the  
0.1 acted as a tie-breaker, like the +EPS we sometimes resort to in  
floating-point work to test for differences outside of rounding error.  0.1  
might be the least-surprising flavor of zero in this situation.
Maybe that's just a rationalization (sour grapes); I tried putting  
extra-spacing-height only on the items who need it, and got frustrated  
trying to keep things from breaking, like { g''( c''-> geses'' a'') }.

Mike (14) I did not see these changes in my tests.  The images look like  
what I would expect if the patched define-grobs.scm was used with an  
executable from current master (specifically current master  
separation-item.cc).