Re: Measure number collision

[Joe Neeman]

On Monday 23 April 2007 05:49, Maximilian Albert wrote:
> Graham Percival schrieb:
> > This has been fixed in 2.11; please wait until the next stable release
> > (2.12) comes out.
> > http://code.google.com/p/lilypond/issues/detail?id=98&can=1&q=collision%2
> >0numbers
>
> On that page it says that the fix already happend in 2.11.5. Looking at
> the git diff output, Joe at that time introduced the line
>
>    (outside-staff-priority . 100)
>
> in the settings for BarNumber in scm/define-grobs.scm. Later, however,
> it was removed again with the following explanation in the log:
>
> "BarNumbers collide with a treble clef's bounding box but not with the
> treble clef itself so setting outside-staff-priority pushes them up too
> much."
>
> Reinserting the above line manually in scm/define-grobs.scm and
> comparing it with Mats' solution yields exactly the same result for
> staves with brackets, but for ungrouped staves it does indeed look ugly.
>
> Is there an easy way to make this setting only work in the former case?
> (Hm, probably not, otherwise it would probably already be implemented;))
> Of course, it's not a big deal to add the padding property manually to
> each score if needed, but I think eventually lily should be able to
> detect these cases itself.

The solution for this is to use more accurate glyph outlines than bounding 
boxes. I hope to do this at some point, but it is non-trivial.