Re: Did VerticalAxisGroup default-staff-staff-spacing stop respecting padding in 2.17.10?

[address@hidden]

On 8 févr. 2013, at 16:57, Xavier Scheuer <address@hidden> wrote:

> On 8 February 2013 09:27, address@hidden <address@hidden> wrote:
>> 
>> (snip)
>> 
>> It's gonna take me some time to have a look at it - I'll do my best to check 
>> it out on Sunday.
> 
> Is this related to the not-yet-added-to-the-tracker "chord name collide
> with staff" issue?
> http://lists.gnu.org/archive/html/bug-lilypond/2013-02/msg00019.html
> 
> Should Trevor's report be added to the tracker?
> 
> Cheers,
> Xavier

The two problems are related but not identical.  The problem here is that 
skyline.cc uses an EPS value to weed out too-tiny buildings.  This means that a 
skyline that has tiny extents (including 0) for its elements could turn out 
empty.  That causes the new skyline code to puke because it does not know how 
to measure distance between an empty skyline and a non-empty skyline (or an 
empty vs empty for that matter).

Running regtests now on a solution, but Joe's input would be good.  I only got 
rid of EPS in skyline.cc where it causes problems in this particular instance, 
but perhaps it'd be a good idea to get rid of it all together.

Cheers,
MS