The use of "8va" works perfectly in the rosegarden GUI, but leads to a lilypond error:
Parsing...
rosegarden_tmp_F23993.ly:41:14: error: GUILE signaled an error for the expression beginning here
#
(set-octavation 1) d''' 4 #(set-octavation 1) a''' f''' f''' |
rosegarden_tmp_F23993.ly:41:41: error: GUILE signaled an error for the expression beginning here
#(set-octavation 1) d''' 4 #
(set-octavation 1) a''' f''' f''' |
Interpreting music...
Preprocessing graphical objects...
Finding the ideal number of pages...
Fitting music on 1 page...
Drawing systems...
Layout output to rosegarden_tmp_F23993.ps'...
Converting to./rosegarden_tmp_F23993.pdf'...
Unbound variable: set-octavation
Unbound variable: set-octavation
fatal error: failed files: "rosegarden_tmp_F23993.ly"
Testcase: http://www.davidfaure.fr/2015/8va.rg
lilypond 2.18.2
http://www.lilypond.org/doc/v2.19/Documentation/notation/displaying-pitches says we should be outputting \ottava #1 ... \ottava #0 instead?
lilypond generation (and then checking that lilypond doesn't give errors) is something that would benefit from being unit-tested.
I would contribute a first unit-test, but the buildsystem doesn't make it easy. What happened to the switch to cmake mentionned in https://cmake.org/pipermail/cmake-promote/2006-December/000446.html ? If the current developers would consider it, I could contribute a cmake buildsystem.
We lost our build system in the port. I am not a build system guy at all, and apparently neither was anyone else on the team at the time.
Our build system is a train wreck, and I would be beside myself with joy if you would be willing to contribute a cmake build system. I can't imagine anyone else disagreeing, but I will post a note on the devel list for good measure.
OK, keep me posted.
I'm curious, which port was that? Qt3-Qt4?
It was a port from KDE3 to Qt4. It was brutal and ugly.
Which reminds me that if we were to end up with a new build system, it would be good to have the option to compile with either Qt4 or Qt5. In principle, either one should work now.
Should be fixed in rev. 14277.
It seems the exporter was using a very old and deprecated LilyPond syntax.
It was probably broken since years.
Thanks! This indeed fixes my first testcase. But it still gives an error with this other testcase:
http://www.davidfaure.fr/2015/8va-broken.rg
rosegarden_tmp_z29882.ly:40:44: error: syntax error, unexpected EVENT_IDENTIFIER
\ottava #1 gf''' 16 \ottava #1
[ ef''' bf'' af'' ] \stemDown gf'' 8 \ottava #0 r r2 \stemNeutral |
rosegarden_tmp_z29882.ly:21:5: error: errors found, ignoring music expression
Should be fixed in rev. 14278.
I really can't find any point to output the second "\ottava #1".
I remove it hoping it doesn't break anything else.
This fix needs more testing.
Thanks Yves, now the 8va works on my file. Moving on to the next issue... :-)
You can close this bug as far as I'm concerned.