Re: markup: give the padding functions effect in skyline spacing; issue 3255 (issue 9295044)

[Keith OHara]

On Sun, 09 Jun 2013 12:37:27 -0700, <address@hidden> wrote:

> https://codereview.appspot.com/9295044/diff/11001/input/regression/text-script-vertical-skylines.ly#newcode6
>
> This is hard to understand IMHO.  What about
>
>    @code{pad-around} computes the (smallest) bounding box of a stencil
> and pads this box by the given amount.  @code{with-dimensions}
> completely...

The point of this test, though, is not how the box is computed (which is done 
differently for the different \pad-* functions).   The point is to check that 
these boxes show up in the skylines, so that they have effect if the stencil 
ends up in a TextScript spaced around music.

I'll reword that doc-string with the forthcoming patch-set that cleans up the 
PostScript output.

One possibly-interesting choice is whether the stencil-padding functions should always replace the 
skyline with a box, or make the union of the skyline with the box.  Clearly \with-dimensions needs 
to replace the skyline with a box for your kludge \with-dimensions #'(0 . 0) #'(0 . 0) .  However, 
what would you expect from  {c'1^\markup{ \pad-to-box #'(-1 . 5) #'(0 . 1) "High" }} ?  I 
think I should pass both the box and the "High" to the skyline-creator so that we use 
their union in this case.