Half duration shortcut crashes musescore when time signature is 5/4

(you're all but one, and crashes are critical)
I can't reproduce in 3.6.2 in Windows 11

I can in a Debug build though. But that is nothing a mere mortal user would/should ever use. The released version os MuseScore, as available from this site's download page, should not have that issue (being build in Release mode, asserts disabled), so I guess you'd need to bring this up with whoever you got your version from.

Easy fix might be to just remove that Q_ASSERT()?

Code:

// Longest TDuration that fits into Fraction. Must fit exactly if truncate = false.
TDuration::TDuration(const Fraction& l, bool truncate, int maxDots, DurationType maxType)
      {
#ifdef NDEBUG
      Q_UNUSED(truncate);
#endif
      setType(maxType); // use maxType to avoid testing all types if you know that l is smaller than a certain DurationType
      setDots(maxDots);
      truncateToFraction(l, maxDots);
      Q_ASSERT(truncate || (fraction() - l).numerator() == 0); // check for exact fit
      }

Possible fix:

$ git diff
diff --git a/libmscore/cmd.cpp b/libmscore/cmd.cpp
index f06d59eeed..e43140643d 100644
--- a/libmscore/cmd.cpp
+++ b/libmscore/cmd.cpp
@@ -2693,7 +2693,7 @@ void Score::cmdIncDecDuration(int nSteps, bool stepDotted)
 
       // if measure rest is selected as input, then the correct initialDuration will be the
       // duration of the measure's time signature, else is just the input state's duration
-      TDuration initialDuration = (cr->durationType() == TDuration::DurationType::V_MEASURE) ? TDuration(cr->measure()->timesig()) : _is.duration();
+      TDuration initialDuration = (cr->durationType() == TDuration::DurationType::V_MEASURE) ? TDuration(cr->measure()->timesig(), true) : _is.duration();
       TDuration d = initialDuration.shiftRetainDots(nSteps, stepDotted);
       if (!d.isValid())
             return;

I have no idea about any possible negative consequences though

Same crash with Shift+Q, W and Shift+W. So I think my proposed fix above is correct, it fixes them all and in the routine that handles those change duration shortcuts.

I'll add it to https://github.com/musescore/MuseScore/pull/9000 and check whether the same bug and fix applies to master (which will become MuseScore4) too.

Edit: apparently it does

See https://github.com/musescore/MuseScore/pull/12084

Please tell the Arch maintainer to build releases in Release mode rather than in Debug mode

But meanwhile, best to skip those unofficial third-party builds and instead download the official version of MuseScore directly from this site (see Download link above). For Linux, the AppImage is the supported build. The versions individual distribution maintainers build themselves are generally not tested and often contain errors like this.

Which is what I wrote above ;-)

Unless the maintaine did some further changes, a "Release" build does #define NDEBUG which in turn disables the abort() inside the assert() marco (which is used by Q_ASSERT()), so it won't crash (not there at least). But "release" is not neccessarily the same as "RELEASE" or "Release".
Whatever: just use the AppImage, this is provided by the MuseScore developers and the only supported version for Linux. And it won't crash in this situation

Seems fixed, I can't reproduce in 4.0.1 nor in a Debug build from the master branch

Should not be fixed though, guess my testing was flawed

Fixed in branch master, commit d6188228fb

_Fix #332925: Half duration shortcut crashes musescore when time signature is 5/4

Crash happens only in Debug builds._

Fixed in branch master, commit f3ce3c94c2

_Merge pull request #12084 from Jojo-Schmitz/cmdIncDecDuration

[MU4] Fix #332925: Half duration shortcut crashes musescore when time signature is 5/4_

Automatically closed -- issue fixed for 2 weeks with no activity.