Re: Issue 1290 in lilypond: Skyline compaction is going overboard sometimes

[David Kastrup]

"Keith E OHara" <address@hidden> writes:

> On Mon, 25 Oct 2010 05:38:24 -0700, <address@hidden> wrote:
>
>> 2) write better docs for this particular [skyline-horizontal-padding] tweak
>
> I talked around my point but didn't make a conclusion.  Docs for the correct 
> way to solve this, system-system-spacing, are already good.
>
> One might ask :  Why not increase, or document, skyline-horizontal-padding ?
>
> When skylines are padded, they make a nice cushion shape (image attached) 
> like David was asking for.  The cushion does not extend vertically, but that 
> is what the various 'padding properties are for.
>
> However, skyline-horizontal-padding consistently pads skylines
> *wherever* they are used, and I found more places where I would not
> want even half a note-width of padding (attached piano music images
> are with and without).

I was never suggesting to pad individual objects.  The horizontal
padding would need to be applied to the skyline of the whole _system_,
not the skyline of individual objects while the system is being
assembled.

> In the example that opened the issue, I think it is the fact that the
> notes are logically unrelated that makes their interleaving so
> objectionable.  Therefore the response should be tied to the fact that
> they are from different systems, so bump system-system-spacing back
> up.

But that affects the vertical spacing of _any_ system, not just systems
with interleaving parts.

I am afraid that you are trying to evaluate my proposal by tweaking
existing knobs.  The existing knobs are no use for a reasonable
automatic resolution of the problem.

-- 
David Kastrup