Re: Issue 381 in lilypond: input/regression/accidental-tie.ly

[Joe Neeman]

On Wednesday 20 June 2007 13:15, address@hidden wrote:
> Issue 381: input/regression/accidental-tie.ly
> http://code.google.com/p/lilypond/issues/detail?id=381
>
> New issue report by gpermus:
> \version "2.11.27"
>
> input/regression/accidental-tie.ly
>
> % gis after the line break should have an accidental

This seems to have been caused by commit b80bb48c. Before line breaking is 
done, tie->original () will always return 0, so the accidental will always 
suicide. I've reverted this in the jneeman branch but I don't want to merge 
it until I work out why the patch was originally made. The only things I can 
think of were that the accidental was interfering with spacing or with slur 
positioning.

The spacing code deals with this situation explicitly -- it looks out for 
accidentals and leaves them out of skylines that they shouldn't be in.

I don't really understand the slur code, but I think that it doesn't get 
called before line breaking anyway.

Have I missed something?

Joe