On Sun, 9 Sep 2001, Don Simons wrote:
> In a private nessage, Neil Killeen wrote
>
> > On Sun, 9 Sep 2001, Don Simons wrote:
> >
> > > Several bugs crept in when I added automatic shifting of accidentals and
> > > noteheads in chords. I think I have them fixed. Version 2.332 is at
> > ...
> > does that include the dot shifting buglet ? its
> > the only one relevant to me presently, so
> > i wont upgrade unless its in there
> >
> A sample of the problem is with "g2d zc zd" where the dot from the c
> interferes with d . I don't think this is a new problem (but correct me if
> I'm wrong). I'd call it a bug if I hadn't moved the dot with its own
> notehead, but it looks like I did. I've had some private correspondence
> with Bernhard about this too. I think it would be really hard to come up
> with a general set of rules for dot shifting to avoid this sort of problem.
> If anyone can come up with one--an algorithm that specifies how much to
> shift all dots, horizontally and vertically, and covers all possible
> cases--I'll try to implement it. Meanwhile, there's a facility in PMX to
> manually shift the dots. BTW, I think I read somewhere that the dots should
> always form a single column, but I'm not sure I like that approach.
>
> --Don Simons
>
i was thinking it should be a similar algorithm to getting
the accidentals to find some free space except the dots
trail rather than lead. I would see the accidental as
a shape of size [nx,ny] The dot is a shape of size
[nx2, ny2] where nx2 << nx and ny2 << ny of course but
otherwise no different. Presumably you already have a constraint
on how much space the accidental can lead the note by, similar for
how much the dot can trail the note by.
I realize there is the extra dimension of verticality as well. One could
start with auto-horizontal shifting and leave verticality as manual.
I would expect that would deal with the majority of cases and
probably makes the algorithm tractable rather than a longer walk
through 2-d phase space.
I never attempt to make all my dots line up in a vertical column,
i have a suspiciion that would look odd. However, it would
make for a very easy algorithm !
regards
Neil
_______________________________________________
TeX-music mailing list
[EMAIL PROTECTED]
http://sunsite.dk/mailman/listinfo/tex-music
