Re: Stem Direction Bug

[Erik Sandberg]

On Tuesday 23 November 2004 05.07, David Raleigh Arnold wrote:
> On Monday 22 November 2004 01:46 pm, Erik Sandberg wrote:
> > On Monday 22 November 2004 16.52, David Raleigh Arnold wrote:
> > > Using both lilypond-snapshot2.4.2 and lilypond2.2.6, the default stem
> > > directions are wrong at the place indicated in both input files below.
> >
> > That is not obvious to me.
>
> Sorry.  The original has been in print since 1838, and I don't think
> you can find a single instance different, from any publisher.  I
> noticed, that's all.

I'm really sorry if I was impolite, all I meant was that I wasn't convinced 
that it was a bug.

Have you seen more than one example that does stemDown?

> > Upward stem works fine IMHO, it could even be
> > considered better than down stem (this might minimise the maximal stem
> > length, which may be considered good). My standpoint is not very solid
> > though, if you have good arguments I might change my mind.
>
> Not my problem.

Your previous argument was however a good one. Lily's goal is to typeset 
professionally; and if professionals consequently typeset that piece of music 
as stemDown, then there is in some sense a difference between lily's 
behaviour and professional behaviour.

I suppose the most useful thing is to add a bunch of examples where stem 
direction is incorrectly chosen, as some kind of regression test. So I added 
your example to beam-direction.ly (& renamed it to the more correct name 
stem-direction.ly).

Erik