Re: vertical movement without anchors

[David Kastrup]

Thomas Morley <address@hidden> writes:

> Just back from vacations, reading up the mails...
>
> I tested the code below with 2.16.2 and 2.19.44 and colored the ref-points.
> Images attached. Up to now they are simple observations, have to think
> about further implications...
>
>     \relative d' {
>       \stemUp
>       \textLengthOn
>       \set Staff.instrumentName = #(lilypond-version)
>       d2^"foo"
>       d'^"foo"
>       d,^\markup \raise #4 "foo"
>       d'^\markup \raise #4 "foo"
>       d,^\markup { \null \raise #4 "foo" }
>       d'^\markup { \null \raise #4 "foo" }
>       d,^\markup \box \raise #4 "foo"
>       d'^\markup \box \raise #4 "foo"
>       d,^\markup \box { \null \raise #4 "foo" }
>       d'^\markup \box { \null \raise #4 "foo" }
>       d,^\markup \box \line { \null \raise #4 "foo" }
>       d'^\markup \box \line { \null \raise #4 "foo" }
>     }

That looks like the anchor point _is_ making a difference and that it is
part of the skyline ultimately used for positioning in 2.18 (a skyline
which fits beside the skyline of the opposing stem, resulting in a
tighter adjustment).  However, I don't see this observation explain the
difference in behavior in the manual example.  Does \textLengthOn have
anything to do with it?

-- 
David Kastrup