Activity for Yves Guillemot
-
Yves Guillemot posted a comment on ticket #1705
Sorry, I forgot to remove the #include: misc/InstrumentsTr.h (and .cpp) was part of a first attempt to fix the issue and are no longer needed. Please, thanks to merge again.
-
Apart from the initial translation issue there was four other problems when running lupdate via the make-ts script: 1 - errors with ::StaffHeaders /home/yves/src/Rosegarden/rosegarden.git/src/gui/editors/notation/StaffHeader.cpp:148: Qualifying with unknown namespace/class ::StaffHeader /home/yves/src/Rosegarden/rosegarden.git/src/gui/editors/notation/StaffHeader.cpp:437: Qualifying with unknown namespace/class ::StaffHeader /home/yves/src/Rosegarden/rosegarden.git/src/gui/editors/notation/StaffHeader.cpp:562:...
-
Fix bug #1705: Bank and Program names are translated
-
Disambiguation string wouldn't work in that case. We have to use a specific context for the instrument names strings. AutoloadStrings.cpp should use QT_TRANSLATE_NOOP in place of QObject::tr and QCoreApplication::translate should replace QObject::tr in MIDIInstrumentParameterPanel.cpp and other places where the translation is only allowed from a given context. I already did most of the work. I still have to finalize a script modifying the .ts files. Without it, the translators will need to change...
-
French translation
-
Thank you, Philip. You got it. I confirm the stretcher works fine when the NDEBUG condition is removed.
-
The problem occurs in release mode only. When RG has been compiled in debug mode, the audio time stretcher works fine.
-
Thanks Philip, its fine that you fix it. As far as I understand, crotales are used in pairs or larger groups, but each element of a pair or group is a crotale. So probably there is some spelling ambiguity here.
-
Some categories of segment presets are no more obtainable.
-
Thank Philip, but this is not the cause of the problem: libsndfile 1.0.31 is installed here and its devel package as well. Moreover audio files can be read and write without any problem and I presume it's done with libsndfile. Nevertheless, as it works on some system, I suspect the problem is related to a library. Concerned libraries may be libfftw3 and libsamplerate. I'll try to trace data inside the AudioTimeStretcher as soon as I'll have some time.
-
Message was not coming because I didn't run RG in debug mode. [AudioSegmentRescaleCommand] AudioSegmentRescaleCommand: segment file id 2 , given ratio 0.945976 [AudioFileTimeStretcher] getStretchedAudioFile(): got source file id 2 , name "/home/yves/rosegarden/sounds/stretch-conv-Sunny Afte-20240227-232331-3.wav" [AudioFileTimeStretcher] getStretchedAudioFile(): got derived file id 3 , name "/home/yves/rosegarden/sounds/stretch-stretch-conv-Su-20240228-222741-4.wav" AudioTimeStretcher::initialise...
-
I tried with 23.12, the last git and another about two or three years old version (I currently don't remember). I tried again with 23.12 and all preferences set to default, but it changes nothing. The audio files are ordinary wave files. Here are the descriptions of two short files before and after the stretch. The files have been recorded with RG itself. Test 1 - Record audio files set in preferences as: 16 bits PCM WAV format Before stretch: --------------- $ sndfile-info 'rg-[testaudio.rg]-[Audio_#1]-2024-02-27_22.03.57-16.wav'...
-
It works for me on three different systems: Mageia 8/Qt 5.15.2, Mageia 9/Qt 5.15.7 and Arch Linux/Qt5.15.11. Without any hint on how to reproduce, I can't do anything.
-
For some reason the title position of the TimeWidget (which is a QGroupBox) is misplaced. I don't find the real cause of the problem, but enforcing the current style of this widget fixes it. Should be fixed in branch Fix1678. Thanks to merge.
-
-
Eventually I reproduce it. It only occurs with the "Native (Light)" theme. There is no problem when the "Classic (Medium)" or the "dark" theme is used. I didn't imagine that a change of the background colour may have such an effect on the text location.
-
Audio time stretcher erases all audio data
-
It's weird. I can't reproduce it, either with Plasma 5.20.4/Qt 5.15.2 or Plasma 5.27.5/Qt 5.15.7.
-
LilyPond doesn't draw the repeat double bar when there are alternative endings
-
LilyPond doesn't draw the repeat double bar when there is alternative endings
-
LilyPond issues deprecation warnings while processing RG output
-
Update French translation
-
Should be fixed in branch "segLabelsTranslation". Thanks to merge.
-
Fix for ticket #1667
-
The segment parameter box displays a translation of the segment label
-
The segment parameter box displays a translation of the segment label
-
In NotationGroup::calculateBeam(), the maximum duration of initial and final notes of a beamed group was set to quaver and not to dotted quaver. This should now be fixed in branch "beams_and_dots". Thanks to merge.
-
Fix for ticket #1666
-
A group beginning of ending with a dotted eighth note can't be beamed
-
A group beginning of ending with a dotted eighth note can't be beamed
-
Thanks Philip, you are totally right. I was selecting the slur with the first two notes and I didn't notice it. If I select the notes carefully, without touching the slur, all works as expected. Sorry for the noise.
-
Thanks Philip, I tried your fix and it works fine. While testing it, I just notice a strange behaviour which preexists the fix: Entered at least four consecutive notes with a slur. If the two last notes are selected then deleted, the slur is kept on the first notes. But if the same action is performed on the two first notes, the slur is totally removed. This is a bit weird, but not harmful and not directly related to the current issue. I'm not sure it deserves a ticket.
-
Removing the first note of a slurred group creates an unreachable rest.
-
Update French translation
-
Update French translation
-
Track is now a ClefKeyList
-
ClefKeyList object code now replacing original code in Segment
-
Add code for ClefKeyList object, but don't use it still.
-
testWeak : Fix insertion of track in composition
-
Some tests to look how it's currently working.
-
Clean creation of three segments with some clefs and keys
-
Start writing a new test for weak clefs and keys
-
These gaps are intended and are unrelated with any LilyPond version. LilyPond versions prior to 2.12 are no more supported (2.12 was released in 2008) and the exporter now produces the same LilyPond code whatever the compatibility level is. The attached example shows what happens: Staves A and B : the syllable oo and aa are followed with a dash ("oo-" and "aa_") and the exporter write LilyPond code giving a hyphen between oo and aa and an extender after aa. In staff A, there is no syllables under...
-
Thanks to have tested it, Ted. Lilypond may remove or reduce the size of dashes to optimize the simultaneous rendering of lyrics and music. This can be shown by adding rests between to syllables as in the attached example in bar 5. Dash between "se" and "gar" is invisible in the first line. On the contrary, dash between "gar" and "den" is invisible in the third line. The two kinds of underscore in version 3 are the result of a bug: the protection of a double underscore syllable didn't work as intended....
-
Fix the double underscore bug (in Fix for bug #1652)
-
LilyPond exporter: Fix the rendering of the syllable "__"
-
The separators of the lyrics syllables are not correctly exported to lilypond.
-
It should now be fixed. Merge requested. Thanks to merge. The table below shows how LilyPond export is now behaving: Notation editor LilyPond input LilyPond output xxx- yyy xxx -- yyy xxx - yyy xxx- - - yyy xxx -- _ _ yyy xxx - yyy xxx_ yyy xxx __ yyy xxx__ yyy xxx_ _ _ yyy xxx __ _ _ yyy xxx______ yyy
-
Fix for bug #1652
-
Merge branch 'master' into lilylyrics
-
Fix for lyrics containing some special characters
-
Restore LilyPond export test baseline
-
Fix a small bug related to numbers in the lyrics
-
Cleanup
-
Rework the lyrics processing
-
Rename
-
Remove old code
-
Lilypond hyphens and underscores: provisional version
-
The separators of the lyrics syllables are not correctly exported to lilypond.
-
Fix memory leaks in test/reference_segment
-
I tried ASan using your instructions in the wiki. It's far easier to use than valgrind. I fixed the memory leaks in reference_segment . Please, see merge request. Not sure it's really useful as this memory leaks are not harmful in a test being not run inside a loop. BTW, this probably should be in another ticket...
-
Fix memory leaks (shown with ASan) in test/reference_segment
-
Fix for bug 1640
-
Thanks Ted. Indeed your suggested modification fixes the bug. I should have seen it. The worse is that it's me who wrote this bit of code and therefore introduced this bug. I'm going to send a merge request.
-
Fix for bug #1640
-
The problem occurs when a "hairpin" (crescendo or descrescendo) is limited to a single note. In this case, Lilypond text uses an expression involving two invisible rests. With ArchLinux, the code writing this expression fails and gives a wrong text. Among the RG examples, this is only found in "glazunov.rg" and "glazunov-for-solo-and-piano-with-cue.rg". In the code below, "str" (last line) is an ofstream used to write the LilyPond file. The strings "obtained" and "expected" are found when comparing...
-
I was able to download Arch-Linux-x86_64-basic-20230429.146445.qcow2 yesterday (or maybe the day before) with the initial link. It take me some time to use it as I never used qemu before. Eventually, I was able to run extra-x86_64-build. It compiled for hours and find no error because I kept PKGBUILD unchanged. After modifying PKGBUILD: before: xvfb-run ctest --output-on-failure --test-dir build/test -E '^(lilypond_export_test|test_notationview_selection|convert)$' after: xvfb-run ctest --output-on-failure...
-
Should have been fixed along with bug 1649: see merge request for this bug.
-
Should be fixed: see merge request for branch lilylyrics. Also fix the bug 1650.
-
Fix for the bugs 1649 and 1650
-
Fix the LilyPond exporter autotest and update the baseline
-
Thanks Philip, I tried your fix and everything works fine with it.
-
I didn't see the problem in the event editor because I forgot selecting the rests. The whole rest in the last modified bar is not replaced with a half rest. This only happens in "restricted", "simple" and "matrix-overlay" paste types. Normalizing rests (Ctrl-N) just after the paste fix the problem. Before normalizing rests it may be necessary to reselect the last bar. In "restricted" mode, the pasted notes are no more selected when pasted with Ctrl-V (or "Edit->Paste") but are kept selected when...
-
Copy paste in notation editor behaves strangely
-
Cleanup
-
Add comments
-
Cleanup and comment
-
Rename
-
Cleanup
-
Fix bar number of verses in exported LilyPond files
-
Cleanup
-
Fix various problems.
-
Remove old code, now unused.
-
Rename various elements related to the word "volta"
-
Comment
-
File forgotten in the last commit
-
Fix the removing of an underscore at the end of a bar of lyrics
-
Cleanup
-
Exporting multiple lyrics verses with repetition is now fully working
-
Fix the computing of the "Multiple verses factor".
-
Add comments
-
Fix the printing of supplementary verses
-
LilyPondSegmentsContext::getNumberOfRepeats() modified
-
Add new LilyPond versions. Remove debug messages.
-
Remove unused methods
-
Cleanup and remove compilation warnings
-
Cleanup. LilyPond versions < 2.12 are no more supported.
-
The vertical alignment of verses is now working.
-
Exporting to Lilypond with volta/alternatives and lyrics is now working.
-
LilyPond exporter simplification. New methods to help exporting lyrics.
1 >