Re: Simplify the NullVoice context (issue 117050043 by address@hidden)

[dak]

On 2014/09/08 23:28:31, dan_faithful.be wrote:
> On Sep 8, 2014, at 03:54 , mailto:address@hidden wrote:

> > Again: I would strongly suggest that you backpedal and consider the
> > question "what was my first approach which did not work, for
> > reasons?".
> >
> > And then we see what it would take to address the reasons.

> How about placing the voices inside a VoiceGroup context, moving the
> Accidental_engraver there, and leaving the NullVoice outside, like in
my recent
> partcombine experiments?

> VoiceGroup would by default consist of nothing, accept any kind of
Voice, and be
> accepted by any kind of Staff.

A solution that reorganizes the default voice hierarchy will cause
interference with all user code fiddling with hierarchy.  At the current
point of time, such code has some place for cross-voice placements of
items like stems, ties, or slurs.  That kind of thing has its place in
piano music.  So touching the default hierarchy should be a last resort.
 One can use music functions for juggling with \accepts/\denies in local
context modifications and thus change the hierarchy ad-hoc just at the
place where an effect is desired.

I am not sure I understand your proposal correctly, namely whether you
are thinking of changing stuff like the \defaultchild of Staff.  If you
don't, all VoiceGroup staves would have to be instantiated manually.  If
you do, many things will change.

The former could be used for providing music functions solving a
particular problem.  But it sounds like a solution that has quite a bit
of potential of interfering with other solutions focused around similar
problems.  It would, of course, at least spare us the hackery in the
internals of the grob implementation.  But I think we can do better, see
<URL:https://code.google.com/p/lilypond/issues/detail?id=4093#c6>

https://codereview.appspot.com/117050043/