Glissandi disappear in extracted parts

[julenComposer]

Issue type

Engraving bug (incorrect score rendering)

Bug description

Dear all, I'm extracting parts from a large orchestral score. Some parts I had to create from scratch. When creating a new part and adding instruments to it with the eye button of the instruments panel (F7) glissandi that were in the score disappear.
This happened to me only when porting an instrument into a new part that I created myself.

Steps to reproduce

1. Create an ensemble score with instruments playing glissandi.
2. Go to "parts".
3. Create a new part.
4. Open that new part.
5. From the instrument panel import an instrument performing glissandi into that new part.
6. Check if glissandi are missing from the part.

Screenshots/Screen recordings

No response

MuseScore Version

MuseScore version (64-bit): 4.2.0-233521124, revision: github-musescore-musescore-eb8d33c

Regression

No

Operating system

Windows 11

Additional context

No response

[oktophonie]

Sounds like #16096.
If you can attach a score (.zip it) where this problem can be reliably reproduced, it would be very helpful.

[julenComposer]

Oh sure, I apologise for being this late.
Orkeststuk I - in vier scénes 03_Jan_2023.zip

Thank you!

[MarcSabatella]

I can reproduce this from scratch as well.

1. new score for one instrument
2. add two notes with a gliss between them
3. go to File / Parts
4. Open all

Result: gliss is there, as expected

But then:

5. go back to File / Parts
6. create new Part
7. Open selected
8. in Instruments panel, show the instrument

Result: no gliss. You don't actually need to follow steps 3-4; that's just to show gliss in parts does work normally.

So it seems the default part containing the instrument gets the gliss, but subsequent ones don't. This issue appears to only affect gliss, not tremolo, slurs, or hairpins.

Screen.recording.2024-01-15.4.24.36.PM.webm

[MarcSabatella]

Same behavior in 4.1.1 and 4.0, BTW. MU3 didn't have this same workflow so it's tough to compare directly, but the analogous steps there don't show a problem.