Re: two bugs in ties

[Heikki Johannes Junes]

On Tue, 22 Jan 2002, Jan Nieuwenhuizen wrote:

> I used:
> 
>     \score {
>       \context Staff \notes\relative c'' { <c4 g' e'>~<cis4 gis' e'> }
>     }
> 
> which creates no ties, but also doesn't warn.  The code says:

I found the same behavior for this example. Nevertheless, I would expect a
tie between e'-notes (only).

> I get sharps on cis and gis.  Because you say (flats, etc.), maybe
> there's something broken with your font setup?  Have you tried running
> buildscripts/out/clean-fonts?

I usually compile make first 'make uninstall' for the earlier version
(1.5.26), extract the whole source (1.5.28) in new directory and compile
it with 'autoconf && ./configure && make all && make install'. I tried
'buildscripts/out/clean-fonts && make all && make install', but it didn't
change the result. All my accidental work fine.

> >  2. accidentals (sharp, flats, etc.) should not be omitted
> 

If I modify just a bit your example and add a 'Voice'-context for chords
and change one e' to es':

    \score {
      \context Staff \context Voice = i { 
         \notes\relative c'' { <c4 g' es'>~<cis4 gis' e'> } 
      }
    }

Now I get all my errors and benefits. Something works already and
something needs to be fixed.

If tied chords are in the same 'Voice'-context:
 - midi fails with first tied chords, while succeeds with second,
   third, etc...
 - there are slurs between notes with are aligned vertically in the same
   level at the score, but differ chromatically: 
     c4 and cis4, g' and gis', es' and e
 - for the notes which differ only chromatically there are no
   accidentals in the latter tied chord:
     cis4 and gis' need  sharp (#)
     e'            needs natural (cancels the effect of the previous flat)
Except for these bugs, I found the behavior of tying with '~' nice and
effective.

Greetings,

      Heikki Junes