Re: \once causes segfault

[Mats Bengtsson]

When you say \once, the setting basically only applies to the next note,
that's why I asked about what you really had in mind.

However, your example points to another strange bug/feature, namely
that when you set the properties subdivideBeams or beatLength,
the setting applies also to the beam of the last note before the
setting. Example:

\relative c' {\time 2/4
  \set subdivideBeams = ##t
  f16 e d c f e d c |
  % The setting affects also the first beam!!!
  f16 e d c  \set Score.beatLength = #(ly:make-moment 1 8)
    f^"setting before this tone" e d c
  c2 \set Score.beatLength = #(ly:make-moment 1 4)

  % Moving the setting 1-4 16th notes ahead gives
  f16 e d e f \set Score.beatLength = #(ly:make-moment 1 8)
    e^"setting before this tone" d c
  c2 \set Score.beatLength = #(ly:make-moment 1 4)
}


I think the following is an answer to your example:
\relative c' {\time 2/4
  \set subdivideBeams = ##t
  \set Score.beatLength = #(ly:make-moment 1 8)
  | f8 f f16( e) \times 2/3 { e16( d c) } |
  % ...
  f16 e d e  \set Score.beatLength = #(ly:make-moment 1 4) f16 e d c
}

   /Mats


Christoph Ludwig wrote:
> On Wed, Sep 29, 2004 at 10:06:02AM +0200, Mats Bengtsson wrote:
>
> > I can confirm that the SIGSEGV remains in 2.3.19.
> >
> > Just being curious, what did you expect the \once to
> > mean in this situation? Should it only apply to the first
> > 16th note?
>
>
> In the score I am transposing I have 
>
>   \time 2/4
>   % ...
>   | f8 f f16( e) \times 2/3 { e16( d c) } |
>   % ...
>   r8 f f16 e d c
>
> The group with the triplet should be subdivided for better readability
> but the group with the four 16th should not be subdivided.
>
> I am not very familiar with the LilyPond concepts and terminology yet,
> I am not a Scheme or Lisp programmer, and I don't completely
> understand yet the difference between \override and \set as well as
> the scope of these settings. Therefore, I am mostly relying on trial
> and error... 
>
> I tried first to set Score.beatLength before the group with the
> triplets and reset it to 1/4 in the following measure, but for some
> reason it didn't work. Either all beam groups were subdivided or none. 
>
> I then decided to try it with \once because it resets the value after
> a time step (or so say the docs). I am not sure about the definition of
> "time step" - if it is actually related to beatLength then I can
> understand that my attempt confuses lilypond. But I considered it
> worth a try - and the result (a core dump) is certainly not what the
> developers intended whence I reported it here.
>
> Of course, I'd appreciate any hint how to subdivide a single beam
> group only. (We can take it to lilypond-user, I am subscribed there,
> too.)
>
> Regards
>
> Christoph
>
>
> > Christoph Ludwig wrote:
> >
> > > Hi,
> > >
> > > the attached file causes lilypond to dump core:
> > >
> > > address@hidden:~/lilypond/music/test> uname -r -m -o
> > > 2.4.21-243-default i686 GNU/Linux
> > > address@hidden:~/lilypond/music/test> lilypond -v | head -n 1
> > > GNU LilyPond 2.3.16
> > > address@hidden:~/lilypond/music/test> lilypond beat-length.ly
> > >
> > > Now processing `beat-length.ly'
> > > Parsing...
> > > Interpreting music... Segmentation fault
> > >
> > > If I comment out the `\once' then everything works fine. However, then all
> > > 16th groups are subdivided every 8th but I want this only in the first
> > > measure. 
> > >
> > > Is there a known workaround? Or is this already fixed in 2.3.19?
> > >
> > > Thanks
> > >
> > > Christoph
> > >
> > >
> > >
> > > \version "2.3.16"
> > >
> > > \score {
> > >   \context Staff {
> > >       \time 2/4
> > >       \relative c'' {
> > >           a8 a
> > >           \set subdivideBeams = ##t
> > >           \once
> > >           \set Score.beatLength = #(ly:make-moment 1 8)
> > >           a16 a a a
> > >           a a a a a a a a
> > >       }   
> > >   }
> > > }
> > >
> > >
> > >
> > > _______________________________________________
> > > lilypond-devel mailing list
> > > address@hidden
> > > http://lists.gnu.org/mailman/listinfo/lilypond-devel

--
=============================================
        Mats Bengtsson
        Signal Processing
        Signals, Sensors and Systems
        Royal Institute of Technology
        SE-100 44  STOCKHOLM
        Sweden
        Phone: (+46) 8 790 8463                         
        Fax:   (+46) 8 790 7260
        Email: address@hidden
        WWW: http://www.s3.kth.se/~mabe
=============================================