One sophisticated

[Rune Zedeler]

Never claim that you have fixed the last bug wrt. something... - A new bug 
always shows up the minute after.
Or, well, this is actually a very old bug. Fixing it will be rather easy but 
will result in a tremendous increase in the spaghetti-level of the accidental-
engraver - which already has a far too high spaghetti-level, and hence I 
suggest that we just leave this bug alone.
The problem is the idea that was introduced with the tie-accidental-system - 
about using SCM_BOOL_T as a "multi-accidental" signalling that the next note 
needs an accidental nomatter what note it is. This thingy removes information 
about the old accidental and hence it destroys the EXTRA NATURAL algorithm:

cisis2 cis2 | cisis1 ~ | cisis2 cis2

The first cis correctly gets an extra natural, but the 2nd cis does not.
I have used the idea a bit more in the new accidental algorithm, and hence 
there are also other situations (wrt having simultainous i.e. c's and cisses) 
where extra naturals don't show up as expected.

Fixing the bug would require something like replacing the SCM_BOOL_T-flag with 
a list of used accidentals. Easy bug spaghetti-like.

BTW, please cc this adress (address@hidden) as my normal - subscribed - adress 
(address@hidden) is temporarily out of order.

Rune Zedeler
Otte Ruds Gade 44A 4tv
8200 Århus N
Tel: 86786563