Re: A tricky example -- polyMETER against polyRHYTHM

[David Wright]

On Wed 02 Nov 2016 at 22:52:12 (-0700), mclaren wrote:
> Took some skull sweat to figure this one out. But someone may get a kick out
> of it. 
> What I wanted to do was to set up two polymeters (different accent patterns)
> with a shifting polymeter against 'em. 
> Here's a png image of the score: http://i.imgur.com/RtYUgRR.png
> <http://i.imgur.com/RtYUgRR.png>  
> 
> Here's the Lilypond code:
> 
> \version "2.18.2"
> \header { 
>   tagline = ""  % removed 
> } 
> % Lilypond bug #1 -- placing the \paper command anywhere else fails to
> %  produce the desired output, but placing it here generates an EXITED WITH
> %  ERROR CODE 1 message even though the score output looks correct and is
> %  as desired.
> 
> \layout {
> 
>        \paper  {#(set-default-paper-size "a4" 'landscape)}
> indent=0
> 
>    \context {
>     \Score
>     \accepts "TimeLine"
>     \remove "Timing_translator"
>     \remove "Default_bar_line_engraver"
>    
>   }
>   \context {
>     \Staff
>     \consists "Timing_translator"
>     \consists "Default_bar_line_engraver"
>     \consists "Time_signature_engraver"
>   }
>   
>   % End of code.
> }
> 
> {
>   
>   
> <<
> 
> 
> \new Staff { \clef "treble"
>              \relative c'''
>              
> {
> 
> 
> \time 9/8  
>  { c8-> d f e-> b c d16-> f e8  b} 
> \time 8/8
>  {c8[-> d16 f e8 b] c16[-> d f8 e b]}
> \time 9/8
> { c8[-> d f] e[-> b16 c d8] f[-> e b16 c]}
> \time 8/8
> {e8[-> b16 c d8 f] e[-> b16 c d8 f]}
>  
>  
> \time 9/8  
>  { d8-> b c a-> b g e16-> f g8  d} 
> \time 8/8
>  {c8[-> d16 f e8 f] e16[-> g a8 b a]}
> \time 9/8
> { g8[-> f d'] e[-> g16 c, d8] g,[-> a b16 c]}
> \time 8/8
> {a8[-> b16 c d8 e] g[-> f16 e f8 d]} 
>  
>  
> }  
> }
> 
> 
> 
> \new Staff { \clef "treble"
>              \relative c'
>               
> {
> \time 8/8
> {e8[-> b c d16 f] e8[-> b c d16 f]}
> \time 9/8
> {e8[-> b c16 d] f8[-> e b] c[-> d f]}
> \time 8/8
> {e8[-> b16 c d8 f] e[-> b16 c d8 f]}
> \time 9/8
> { e16[-> b c8 d] f[-> e b] c[-> d f16 e]}
> 
> 
> 
> }
> \relative c'
> {
> 
> \time 8/8
> {e8[-> b c d16 f] e8[-> b c d16 f]}
> \time 9/8
> {g8[-> a c16 d] e8[-> a, b] g[-> b c]}
> \time 8/8
> {b8[-> c16 a d8 g] f[-> c16 b a8 f]}
> \time 9/8
> { e16[-> b c8 d] f[-> e b] c[-> d f16 e]}
> 
> 
> 
> }
> 
> }
> 
> 
> \new Staff { \clef "bass"
>              \relative c
> {
> 
>  
>  \time 11/8
>  \once \omit TupletNumber
>   \set Timing.measureLength = #(ly:make-moment 8/8)
> \tuplet 11/8 {  g8[->  e' b c d f e b16 c d8 f e]}
> %\scaleDurations 8/11 {  f8[->  e b c d f e b16 c d8 f e]}
> |
> \override TupletNumber #'text = #tuplet-number::calc-fraction-text
> \set Timing.measureLength = #(ly:make-moment 9/8)
> \tuplet 11/9 {  b16[-> c d8 f e16 b c8 d f e d f g]}
>  
> 
> |
> \once \omit TupletNumber
> \set Timing.measureLength = #(ly:make-moment 8/8)
> \tuplet 11/8 {  f8[->  e b c d f e b16 c d8 f e]}
> \override TupletNumber #'text = #tuplet-number::calc-fraction-text
> 
> \set Timing.measureLength = #(ly:make-moment 9/8)
> \tuplet 11/9 {  e8[-> b c d16 f e8 b c d16 f e8 b c16 d]}
> \bar "|"
> \break 
> 
> \time 11/8
>  \once \omit TupletNumber
>  \set Timing.measureLength = #(ly:make-moment 8/8)
> \tuplet 11/8 {  c8[->  b d b a g f c16 b e8 g f]}
> %\scaleDurations 8/11 {  f8[->  e b c d f e b16 c d8 f e]}
> |
> \override TupletNumber #'text = #tuplet-number::calc-fraction-text
> \set Timing.measureLength = #(ly:make-moment 9/8)
> \tuplet 11/9 {  e16[-> c e8 d a'16 b c8 e g f a g b]}
>  
> %\scaleDurations 11/8 {  b16[-> c d8 f e16 b c8 d f e d f g]}
> |
> \once \omit TupletNumber
> \set Timing.measureLength = #(ly:make-moment 8/8)
> \tuplet 11/8 {  c8[->  e, a f g f e c16 d b8 c g]}
> \override TupletNumber #'text = #tuplet-number::calc-fraction-text
> \set Timing.measureLength = #(ly:make-moment 9/8)
> \tuplet 11/9 {  e8[-> b c d16 f e8 b c d16 f e8 b c16 d]}
> \bar "|"
> \break 
> }
> 
> }
> >>
>      
> }
> 
> ===========
> 
>   Incidentally, I seem to be getting a bug, and it's an odd one. If you try
> to compile this Lilypond score from scratch, you get an exit with ERROR=1.
> Then if you comment out the lines
> 
> \layout {
> 
>        \paper  {#(set-default-paper-size "a4" 'landscape)}
> indent=0
> 
>   and recompile, the score runs fine. Here's where it gets weird... If you
> then immediately uncomment out those 3 lines, you still get an exit with
> ERROR=1, but this time the score gets engraved properly in landscape mode.
> 
>    Can't figure this one out. What am I doing wrong with setting landscape
> mode?

As posted:

GNU LilyPond 2.18.2
Processing `poly.ly'
Parsing...
poly.ly:12:8: error: syntax error, unexpected \paper
       
       \paper  {#(set-default-paper-size "a4" 'landscape)}
Interpreting music...
Preprocessing graphical objects...
Finding the ideal number of pages...
Fitting music on 1 page...
Drawing systems...
Layout output to `poly.ps'...
Converting to `./poly.pdf'...
fatal error: failed files: "poly.ly"

The output is landscape.
With your \paper line moved to between the first comment and the \layout line:

GNU LilyPond 2.18.2
Processing `pol1.ly'
Parsing...
Interpreting music...
Preprocessing graphical objects...
Finding the ideal number of pages...
Fitting music on 1 page...
Drawing systems...
Layout output to `pol1.ps'...
Converting to `./pol1.pdf'...
Success: compilation successfully completed

but portrait.
Removing the "-default" from the \paper line:

GNU LilyPond 2.18.2
Processing `pol2.ly'
Parsing...
Interpreting music...
Preprocessing graphical objects...
Finding the ideal number of pages...
Fitting music on 1 page...
Drawing systems...
Layout output to `pol2.ps'...
Converting to `./pol2.pdf'...
Success: compilation successfully completed

and the landscape output is identical to the first (failed) run.

A second run producing different results from the first suggests
either a platform problem (unlikely but possible) or that you're
running a different source version, eg people using frescobaldi
have reported confusion about whether the workspace version or a
saved version is actually being processed.

So really you need to report which LP version you're running, and
the commands being used to run it, preferably with the logs as I've
posted above.

Cheers,
David.