Re: hinterfleisch of natural

[Werner LEMBERG]

> > IMHO there is still a problem with naturals and sharps occurring
> > in a chord at the same time: They don't appear vertically aligned;
> > [...]
> 
> That's because they are positioned independently.  I don't consider
> this a high priority bug.
> 
> (I remember that you contributed a better algorithm for accidental
> positioning... If you feel tempted, go ahead :-)

Yes, I would like to do that if time permits -- and if I manage to
understand heavy C++...

Anyway, before starting to fiddle with the algorithm, I wonder whether
we can improve horizontal stacking of grobs.  Recently, I had to work
with Intellifont stuff (the fonts built into older LaserJet printers).
While Intellifonts are obsolete today, they have a very nice feature
called `sector kerning' which might be quite useful for lilypond,
especially for accidentals.  I've attached some images which show the
concept.

Implementing this in the metafont sources should be straightforward.
The natural vertical distance from sector to sector would be a second
interval.  While outputting the data for the mf-to-table script, we
could also emit the vertical position of the lowest sector, the number
of sectors, followed by the left and right sector values, to be stored
in some new files.


    Werner