Pianoroll and Drumroll Editor information should disappear if there's no score and update if looking at another
Priority
P0 - Critical
Type
Functional
Frequency
Few
Severity
S1 - Blocker
Reproducibility
Always
Status
closed
Regression
No
Workaround
Yes
Project
Information in the 'Pianoroll Editor' should disappear if there's no score loaded and update if looking at another.
Using MuseScore 2.0 Nightly Build (771004d) - Mac 10.7.5.
Comments
I marked as critical, because of crashes it can lead to:
1. Open score.
2. Right-click stave.
3. 'Pianoroll Editor'.
4. Close score.
5. Open score.
6. 'Play'.
4. Make change.
5. Close score.
6. 'Don't Save'.
7. Open score.
Result = Crash.
Using MuseScore 2.0 Nightly Build (771004d) - Mac 10.7.5.
Also the 'Drumroll Editor' - I think that's all of them?
Using MuseScore 2.0 Nightly Build (deb8c96) - Mac 10.7.5.
I believe the attached log was produced by using the same, or similar steps.
Using MuseScore 2.0 Nightly Build (28f6ddb) - Mac 10.7.5.
The crash with pianoroll editor also happens with the following steps:
Windows 8 & Linux Mint 13, commit 1f41a8570 (self-compiled).
The Pianoroll editor window is trying to access the score which has been deleted: this causes the crash.
Attached the address sanitizer log under Linux.
No crash for me currently, but pianoroll is definitely not happy.
Crashes right after closing current score for me.
Concerning the crash PR is created: https://github.com/musescore/MuseScore/pull/4174
Concerning updating view perhaps we should make a separate issue as I believe it is not really clear how exactly such auto-update should behave.
Fixed in branch master, commit 313231b403
fix #18400: handle absent score and staff correctly in piano roll editor
This also adds some overrides marks and removes some empty member
functions from PianorollEditor class.
Fixed in branch master, commit 8be8835205
Merge pull request #4174 from dmitrio95/18400-pianoroll-close-score
fix #18400: handle absent score and staff correctly in piano roll editor
Automatically closed -- issue fixed for 2 weeks with no activity.