Bug: Key signature lost when previous system with section break is deleted

• Nov 27, 2019 - 06:37
Reported version
S4 - Minor

To reproduce:

1) New score
2) Create a system with random key-signature, and finish it with a section break.
3) The next system is of the same key-signature as the first as to be expected.
4) Work in that second system (add some notes), relying on that key signature.
5) Now select all of the first system and delete not just the content but the actual measures.

Result: The second system becomes key-less and so all notes that had accidents related to the key signature are now all explicit.

This might take the user by surprise because it looks as if the second section "has" a key signature. When the user deletes the first system, the second system will no longer have that said key signature. It might be argued that this is "by design", but even so, this might want to be reconsidered.

Either the key signature should be reset (no accidentals) after placing a system-break, which I advise against but only very slightly, or the score needs to "re-iterate" internally that there's a key-signature on the next system so that if the previous system is deleted, there's no loss of information and the key-signature remains.

If anyone wanted to test this and is lazy enough to rather download a ready made mscz file with the above steps, included is such a file :)

Attachment Size
keysig_after_section.mscz 8.6 KB


I would not call this a bug. Perhaps you expect a section break to signify more than it currently does. The fact is, you removed the key signature that determines the key for the following systems, so those systems have been updated to reflect this. It is easy enough to add the key signature back after deleting the system. However, if you reiterate the key signature yourself before deleting the previous system, the key signature will remain after the system is deleted.

In reply to by mattmcclinch

The current behavior under consideration isn't necessarily about section breaks and their expected behavior, but rather one of the possible surprise a user may receive in losing a key-signature if previous systems are deleted.

Of course the mentioned key-signature was removed, and of course it can be added back explicitly, but If a user puts a page break, starts a new title frame, etc. and sees that the old signature "shows," it's easy to think that that is the key signature in which one is continuing as a basis of a 'new section'.

To continue in such a manner, and then for instance delete the previous page because of a mishap will result in that next page being keyless, and it's easy to overlook this. The software 'imo' ought to accommodate the potential mishap, and the mentioning of a system break was a means to express a stronger sense of the issue because of the 'organizational properties' of such an element of 'starting anew' withing the same score. Maybe this should be considered a suggestion regarding the user experience. I can see why some would think this shouldn't be reconsidered, as it's 'logical' enough as it is, and the title might rather be changed to a suggestion instead of a bug in that sense.