Re: warning: no feasible beam position

[Stephen]

----- Original Message ----- 
From: "Han-Wen Nienhuys" <address@hidden>
To: "Stephen" <address@hidden>
Cc: "lilypond devel" <address@hidden>
Sent: Tuesday, August 02, 2005 9:28 AM
Subject: Re: warning: no feasible beam position


> Stephen wrote:
> > Hi,
> >
> > Is this a bug? The last note of the first system is flipped up probably
> > because of a negative beam length with the warning "weird stem size"
> > after
> > the warning in my subject heading. I don't see what is causing this. It
> > does
> > not happen in 2.4, only 2.6. So perhaps this is a regression?
>
> I'm not sure, it depends whether 2.4 also places the linebreak just at the
> beam. Can you isolate that beam in a snippet? This looks like an

It is layed out in 2.4 and 2.6 exactly as it it layed out in the original
edition, down to it being on page 5 and starting with the 10th measure. This
is clearly a bug that cropped up in 2.6, 2.4 does not do this at the break.

UPDATE:
I never followed up with this properly, but now I can report that the bug 
dropped out with version 2.6.4. Does anyone happed to know how the bug was 
fixed?

> interesting bugreport. You'll need to pad said beam with some skips to
> force the linebreak at the same place.
>
> I am rather astonished that Lily handles this piece of notation so well.

My goal is to test Lilyponds limitations by trying to notate something 
really challenging. I am gratified that you are astonish since that is the 
point (not to astonish you but to demonstrate what Lilypond is capable of.). 
What is fascinating to me is that Lilypond is capable of more than its 
creators and developers are aware of. In this way it corresponds to a 
computer language since it also is extensible and is capable of more than 
its creators can anticipate. And like a computer language Lilypond has an 
internal logic that sometimes allows the user to give it compands that it 
responds to in a predictable way that its creators do not anticipate.

Unfortunately, Lilypond may be moving away from this design philosophy now.

> How long is the entire piece? I'm wondering whether it's worth the effort
> to ask Universal whether we can use this snippet (I guess that 2 systems
> is enough to get the point across) as a demo.

I could email the ly files if you want. I started the second movement before 
I stopped.

Stephen

> BTW, I've been thinking about a
>
>   \override Slur #'details #'edge-attraction-factor = 21
>
> syntax. Perhaps you'd be interested in sponsoring it?
>
> --
>  Han-Wen Nienhuys - address@hidden - http://www.xs4all.nl/~hanwen

beam_break_bug.ly
Description: Text Data

beam_break_bug.png
Description: PNG image