No dummy atom appears when adding centre of mass

[matterhorn103]

This is with the current nightly AppImage, haven't tested with other versions yet.

Avogadro version: (please complete the following information from the About box):

• Avogadrolibs: 1.99.0
• Qt: 5.15.2

Desktop version: (please complete the following information):

• OS: openSUSE Tumbleweed with KDE Plasma 6.0.3
• Version 20240416

Describe the bug
Selecting atoms and then choosing Build > Add Centre of Mass doesn't cause a dummy atom to appear.

However, the dummy atom does appear in the Atomic Coordinate Editor as an atom with element Xx, as expected.

Can anyone reproduce this on a non-Linux system?

[ghutchis]

Is this still an issue?

[matterhorn103]

Testing adding the centre-of-mass with benzene with the coordinates:

C     -6.131895    2.769078   -0.000105
C     -6.094009    1.385029   -0.000311
H     -2.746382    0.936655   -0.001085
H     -4.982015    4.574150    0.001859
C     -4.952279    3.494148    0.000656
H     -7.014433    0.819106   -0.001192
H     -2.814100    3.400660    0.001216
H     -4.846684   -0.354298   -0.002091
C     -3.734624    2.834893    0.000433
C     -4.876447    0.725715   -0.000551
H     -7.082097    3.283518    0.000824
C     -3.696736    1.450806    0.000007

With my local Qt5 build, the stable (i.e. v1.99) Flatpak (Qt5), the beta Flatpak (Qt6), and the nightly AppImage (Qt5), a dummy atom is added and also appears, but has the incorrect coordinates:

...
H      1.281812    0.659986    0.000154
Xx    -0.409527    0.175829    0.000001

For reference that looks like:

Meanwhile with my local Qt6 build, when I add the centre-of-mass all atoms disappear and the coordinates of the dummy atom have a NaN in them:

...
H      1.281812    0.659986    0.000154
Xx         -nan    0.162200    0.000006

It seems odd that one build should be different to the others so let me checkout all the current master branches and rebuild.

[matterhorn103]

Nope, this definitely still happens on an up-to-date Qt6 build. Even though it doesn't with the Qt6 Flatpak.

[matterhorn103]

Oh, and similar to what I just wrote in #1657, if I add the CoM in the problematic build without any atoms selected, it shows the same behaviour as all the other builds (the same incorrect position i.e. Xx -0.409527 0.175829 0.000001)