Re: Reasons for cross-voice limitations?

[Urs Liska]

Am 30.03.2015 um 00:20 schrieb Simon Albrecht:
> Am 29.03.2015 um 23:46 schrieb Kevin Barry:
> > For sure the voice context limitations are a pain, and if I knew how, I
> > would write a function for starting and finishing slurs without the need
> > for creating a hidden voice, but I don't even know if it is possible. 
> > In my
> > own head, I imagine that LilyPond `thinks' in voices and there isn't 
> > much
> > that can be done about that.
> I think the idea is something similar in some way to \change Staff, 
> which shows that there is a possibility to switch contexts. The 
> difference is that \change Staff moves an entire (voice) context to 
> the other staff, whereas a slur is only one object in the bottom 
> context. Only brainstorming: what about creating an ephemeral 
> below-bottom ‘spanner’ context, which only lives as long as the slur 
> (or other spanner? probably only for tie, slur, phrasing slur) is 
> there, and supports writing \change Voice = "foo"? Or one might be 
> reminded of a Lyrics context being associated with a Voice context – 
> in a similar way might a ‘Spanner’ context be associated to a Voice 
> context, and thus the slur get linked to the notes?
> Unfortunately, I can’t say anything about implementation, though I 
> like the idea … ;-)
>
> Yours, Simon

As far as I can understand this goes very much along the line I would 
think it should be approached. And I can't imagine that there are 
fundamental obstacles against it. Although of course I can't think about 
an implementation either.

Urs

> >   Practically 100% of my work is either piano
> > music or music reduced to two staves, so I bump up against this issue 
> > all
> > the time. I never use the part combiner.
> >
> > I obviously don't know how necessary this limitation is, but iirc the 
> > old
> > DOS program SCORE required that the whole score be reentered for the
> > purposes of adding slurs.
> >
> > How hard would it be to write a cross-voice slur function?
> >
> > On Sun, Mar 29, 2015 at 6:49 PM, Urs Liska <address@hidden> wrote:
> >
> > > Hi,
> > >
> > > this has been discussed numerous times, but I think I'll have to 
> > > bring it
> > > up once more: the limitations that slurs, dynamics and other 
> > > spanners can't
> > > cross voice borders. This limitation is a major inconvenience for 
> > > users:
> > > New users are regularly confused, using hidden voices to work around 
> > > the
> > > limitation is actually an ugly workaround, and it's not acceptable 
> > > to be
> > > forced to use such workarounds for reasonably common things. And now 
> > > I've
> > > come to suffer from an issue where this limitation is a real PITA: the
> > > partcombiner.
> > >
> > > I find it quite obscure to tame the partcombiner to do its work anyway
> > > (for example, when you're facing an issue you don't have a real way 
> > > to tell
> > > what "mode" it currently is in because the last switch could have 
> > > happened
> > > about anywhere, and in any of the voices taking part in the 
> > > combination).
> > > But I've come to the conclusion that very often the partcombiner has to
> > > make ugly choices because it's bound to keep voice contexts alive 
> > > during
> > > the whole spanners. Often it remains in partcombineApart mode for
> > > ridiculously long ranges, just because a slur has to be closed in 
> > > the same
> > > voice as it has begun. And this kind of issue is usually sooo 
> > > awkward to
> > > solve because you might end up inserting a hidden voice to achieve the
> > > slurring but have to restrict that to the partcombined staff while not
> > > using it for individual part printing etc.
> > >
> > > So I want to bring that up once more: Why do we still have this
> > > limitation? Is it an inherent problem that can't be fixed, is it just
> > > because noone cared (or had the chance) to fix it, or is it "only" 
> > > because
> > > we didn't explicitly think about the right way to deal with it 
> > > semantically
> > > and with regard to syntax?
> > > I mean, I can't imagine it makes a serious difference when it comes to
> > > *engrave* a slur. It may make a difference for managing and maintaining
> > > contexts, but just as it is possible to add a hidden voice to 
> > > accomplish a
> > > cross-voice item it should be possible to create a built-in solution 
> > > for
> > > that.
> > >
> > > This wouldn't fix the partcombiner limitations, but it would at 
> > > least make
> > > it possible to think about improvements.
> > >
> > > Any ideas?
> > > Urs, frustrated ...
> > >
> > > _______________________________________________
> > > lilypond-devel mailing list
> > > address@hidden
> > > https://lists.gnu.org/mailman/listinfo/lilypond-devel
> > _______________________________________________
> > lilypond-devel mailing list
> > address@hidden
> > https://lists.gnu.org/mailman/listinfo/lilypond-devel