Re: skyline-centered dynamics

[Keith OHara]

On Mon, 02 Sep 2013 10:29:45 -0700, Kieren MacMillan <address@hidden> wrote:

> re: <http://code.google.com/p/lilypond/issues/detail?id=1127>
> Any movement on this issue?

Nothing other than what you see on the tracker.

The simple rule used for comment 8 line 3 seemed to make people happy:
"Center each independently-placed dynamics group between the lower skyline of the 
staff above and the upper skyline of the staff below, then space Dynamics and Staves as 
normally."

The rule is simple because the shape of the Dynamics line does not depend on 
the distance between staves after they are spread to fill the page, only on the 
shapes of the protrusions from neighboring staves.

If the Dynamics can find the data structures for the Staff above and below (not 
obviously easy within a system that allows for ossia staves, and frenched 
scores, and align-above-context, and several Dynamics between staves, etc.)
then that rule would be set the Y-offset of each DynamicLineSpanner after 
line-breaking to the average if its Skyline::distance() to the staves above and 
below.