Re: cross-voice slurs

[Janek Warchoł]

2013/11/9 David Kastrup <address@hidden>:
> Janek Warchoł <address@hidden> writes:
>
>> Hi,
>>
>> 2013/11/9 David Kastrup <address@hidden>:
>>> Colin Campbell <address@hidden> writes:
>>>
>>>> I have an exercise which requires starting a note on one string, then
>>>> adding a second note while the first is still sounding.  Graphically,
>>>> it looks like a slur. My problem is that, in order to get timing
>>>> right, I have to use two voices, and of course, I can't slur from one
>>>> to the other.
>>>
>>> Well, if we are talking about a cello, you'll never have more than one
>>> slur at a time, right?  So you might just move the Slur_engraver from
>>> Voice level to Staff level.
>>
>> I have a quick thought.  What about creating a new grob named
>> CrossVoiceSlur, which would be identical to Slur in all ways except
>> that its engraver, Cross_voice_slur_engraver, would be placed in Staff
>> context by default?  This way, one could use either normal or
>> cross-voice slurs without messing things up.
>
> I don't see how.  The information still needs to be given to LilyPond.
> My particular proposal of moving the Slur_engraver up actually suffers
> from a logical problem: if the Slur_engraver is placed at Staff level,
> of course it will also get to see all notes and note columns at Staff
> level.  So it will not pick sane attachment points.

Hmm.  Good question.  Maybe it could be attached to NoteHeads, not NoteColumns?
(As i understand it, the problem with attachments results from the
fact that slurs are attached to notecolumns (the bound of the slur
spanner is the NoteColumn), and at staff-level the NoteColumns from
different voices are joined which makes them not specific enough.
However, if the bound was the notehead, the information should remain
specific enough).

Janek