Re: RehearsalMark breaks system's ref-point? (was: The ultimate slur shaping function)

[Ralph Palmer]

On Mon, Nov 25, 2013 at 5:22 PM, Janek Warchoł <address@hidden>wrote:

> Hi all,
>
> it seems that when there is a RehearsalMark, the reference point of a
> system changes and the relative extents are calculated in an
> unreasonable way.
>
> Here's the background - i got a bug report about \shapeII:
>
> 2013/11/25 Ed Gordijn <address@hidden>:
> > Hi Janek,
> >
> > This [\shapeII] is a very nice function you have made. But I get
> unexpected results
> > when I use a rehearsal mark.
> >
> > Please see the included snippet.
>
> I did some tests and here's what i got.  I have a function (it's very
> lame, i know) that does some bogus override of a slur's "positions"
> property, and in the meantime it prints the relative Y-extent of
> slur's left NoteColumn:
>
> \version "2.17.96"
>
> foo =
> #(define-music-function (parser location item)
>    (symbol-list-or-music?)
>
>    (define (quux grob)
>      (let* ((orig (ly:grob-original grob))
>             (bound (ly:spanner-bound grob LEFT))
>             (ref (ly:grob-system bound)))
>        ;; print the extent of the left NoteColumn relative to the system:
>        (display (ly:grob-extent bound ref Y))
>        ;; bogus value:
>        (cons 3 3)))
>
>    #{ \tweak positions #quux #item #})
>
> \relative c'' {
>   \foo Slur
>   d1( c')
> }
>
> In this case it outputs (0.449994 . 1.550006) - as expected: the
> extent of the d note, measured relative to middle staffline, is
> roughly (0.45 . 1.55) staffspaces.
>
> But when i add a rehearsal mark:
>
> \relative c'' {
>   \foo Slur
>   d1( c')
>   \mark \default
> }
>
> the output is (-4.100012 . -3.0), as if the point against which the
> extent is measured was on top of the higher note (instead of middle
> staff line).  Indeed, if i change the higher note to e
>
> \relative c'' {
>   \foo Slur
>   d1( e')
>   \mark \default
> }
>
> the extent is further "lowered" by 1 ss: (-5.100012 . -4.0)
>
> This behaviour seems extremely weird to me.  Do you know anything about
> this?
>
> best,
> Janek


Hi, Janek - I went ahead and submitted this as Issue 3677 :
https://code.google.com/p/lilypond/issues/detail?id=3677

Hope I wasn't out of line!

Ralph