Re: forcing hshift

[David Raleigh Arnold]

On Sat, 22 Jun 2002 22:28:18 Han-Wen wrote:
> address@hidden writes:
> > David Raleigh Arnold wrote:
> > 
> > > {a2 a4} + {a4 a2} = {a4 a4 a4}
> > > 
> > > This spacing has been bad for years.  I hope it's fixed now.
> > 
> > I don't agree with you.
> > You seem to think that spacing should be calculated globally on
> score 
> > level without thinking about individual voices. I.e. this would mean
> 
> > that in your opinion two shifted voices with quarter notes
> > 
> > < { c4 c c c }  {c8 c4 c c c8} >
> > 
> > should get same spacing as one voice with eight notes.
> > Furter more your approach would mean that the first measure of this 
> > example should get same spacing as 2nd measure, which I concider
> very 
> > bad spacing.
> 
> I tend to side with Rune on this issue.  Dave's request would mean
> that Lily would have to have the intelligence to detect what kind of
> rhythmic patterns are in the piece, and that seems like a hairy issue.

Another way of looking at it would be spacing according to
where notes start in relation to beats.  You are spacing according
to where notes end.

if I put this

|        |   |
o        *   |
*     o      |
|     |      |

in tab, it would be like this:

|     |     |
*-----------*----|
*-----*----------|

Because tab is always one voice.  That's what's wrong with it.
Tab has the beginnings of notes, but takes no account of
where they end.  But the column spacing is unhairified.
If only you spaced the notes according to the tab.  That's
why it's too bad you didn't do tab long ago.       

> 
>  A semi manual solution is possible. if the user tells what kind of
> notes should be regularly spaced.

You have been spacing notes instead of beats.
Spacing with times instead of events.  The time
value of a note for spacing purposes is the time to the next
note anywhere in the score, or to the end in the case of the
last note.  Notes are played according to their *positions*
in time, not their durations.

And the solution in the past for me has been to add a
third part.    

>  I once implemented this, but I
> threw it away together with the old spacing engine. The original
> motivation was standchen.ly, a piece where there are running 8th notes
> in the accompaniment.  Currently, their spacing gets botched every
> time an 8th triplet is played against it.

The important thing is to have the 2nd eighth visually midway
between the 2nd two notes of the 8th triplet.  That seems to
be fine.
> 
> With the feature you could specify regularSpacing = #(make-moment 1
> 8), and get regular 8th notes.  it was introduced in 1.5.2, and i'd be
> very glad if someone would front-port it again (hint hint :-). It also
> makes a ragtime piece that I tried to do with lily a lot easier to
> read.

Would that help with < { a4 a2 } { a'2 a'4 } > in 3/4?

------------------------------------------------------------
Information is not knowledge.           Belief is not truth.
Indoctrination is not teaching.   Tradition is not evidence.
         David Raleigh Arnold   address@hidden