On Tue, 23 Feb 2021 at 06:04, Joe McCool via LilyPond user discussion <
lilypond-user@gnu.org> wrote:
>
> David, I really appreciate your help. But I still have a wee problem:
>
> When I look at newmend.pdf, it looks spot on to me. But if I pull down your
newmend.ly and compile it here, the result is completely different. See attached.
>
> No doubt this is another cockup on my part :-(
Hello,
You compiled the file with version 2.18.2 whereas David compiled it with 2.22.0 (looking at the tagline in his PDF).
Apparently 2.22 handles incomplete measures and \alternative better than 2.18.
The doc of 2.18.2 also explains when it becomes necessary to set the Timing.measureLength in the case of alternate endings with incomplete measure.
This part is not present anymore in the doc of 2.22.0.
If you use 2.18.2 then modify your \alternative by setting
Timing.measureLength:
\alternative {
{ g8 [ fis8 g8 ] e8 [ fis8 g8 ] |
\set Timing.measureLength = #(ly:make-moment 5/8)
a8 [ fis8 d8 ] c8 b8 } % b4 } % ←
{
\set Timing.measureLength = #(ly:make-moment 6/8)
d8 [ cis8 d8 ] fis8 [ e8 d8 ] |
c8 [ a8 fis8 ] g4 } % g4. } % ←
}
Cheers,
Xavier
--