Re: cadenzaToMusic function, odd behavior causing barcheck failures

[James Harkins]

On Jul 29, 2014 7:29 AM, "James" <address@hidden> wrote:
> On 29/07/14 04:46, James Harkins wrote:
>> Some time ago, a user here gave me a "cadenzaToMusic" function, which could stretch rests in non-cadenza parts to match the duration of a cadenza. (That's because I was having problems with compressing the cadenza music to fit in a normal-length bar.)
>
> More for information to the list really@
>
> Part of the original thread you had it seems was from 2012  - it's only fair when you say 'sometime ago' as this will make a difference on the version of LP you were using at the time.
>
> http://lists.gnu.org/archive/html/lilypond-user/2012-04/msg00125.html I think is that thread.
>
> Which seems to imply this was originally written for \version 2.14.x and you're now using 2.18.x - again just to confirm so that the dev who helped (Thomas Morely by the looks of it) has a chance

Good point. I just double-checked the source for the 2012 piece. I used 2.15.36 for it. (And thanks for tracking down the old thread -- now I can give proper credit in the new piece's acknowledgments!)

I also used a modified version of the function, which skips setting the measureLength, for a 2013 piece that I compiled with 2.16.2.

> In those 'so far' pieces, were they all for LP 2.14.x or was it working in a case for LP 2.18.2 (i.e. is this the first time you are using 2.18.2 and find it no longer works as expected)?

I had never used it with 2.18.x before -- first try with 2.18, first time it failed.

It's got to be something with the set Timing.measureLength -- the time signature engraver puts 6/8 in the right place rhythmically, but the barlines appear as if the 2/4 bar were actually 3 beats. Will play with it some more in a little while.

Thanks,
hjh