Another minimal example from
http://code.google.com/p/lilypond/issues/detail?id=2795#c16
\paper { indent = 0\mm
line-width = 41\mm
ragged-right = ##f }
\score { << \new Staff { g1_"retaliate" }
\new Staff << { e''4 e''' r2 } \\ e'1 >> >> }
Workaround
\override DynamicTextSpanner #'skyline-horizontal-padding = #0.5
Oops, the correct override is
\override TextScript #'skyline-horizontal-padding = #0.2
where 0.2 is the value Janek is considering.
Added as <http://code.google.com/p/lilypond/issues/detail?id=3325>
Would it make sense if the "skyline horizontal padding" was only applied
in inward direction and/or limited/clipped by the outward padding? In
that case, one could use comparatively larger values for stopping
interleaving while not causing outward problems.
Not in this case, because the near-collision is just as disturbing if the word ends at the 'i'
"retali" and a stem comes within the usual inter-letter space. See examples, 2, 3
and 5 in Reinhold's original report of the problem
<http://lists.gnu.org/archive/html/bug-lilypond/2012-08/msg00190.html>
It would also make it possible to "close" something like the \downbow
symbol with liberal internal padding without causing outward problems.
In that case, maybe ... if there are outward problems. Should the mordent go
around one side of the downbow, while the downbow stays clear of the staccato,
and is the reason because the mordent only wraps around one side? If so, we
could in give the mordent less horizontal-padding in scm/scripts.scm
{ \override Script #'skyline-horizontal-padding = #0.2
f''4 -. \downbow -\tweak #'script-priority #200 \upmordent
Skylines that are the convex-hull of the stencil, or whatever Janek's magnetic
skylines are, might be nice, but while waiting for them we can at least set
usable defaults for the skylines that Mike has implemented.
_______________________________________________
bug-lilypond mailing list
bug-lilypond@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-lilypond
