Avogardro Version: 2.0.0
system: MACOS Tahoe 26.2, Apple M3
Only “orbitals” function works under Semi-Empirical QM(xTB), the other functions do not work (no response).
RDkit does not work.
Python settings: can not make selections, only “other“appears.
I’m not sure why orbitals would work, but not other functions. Can you give a bit more detail on what you’ve tried?
Can you be a bit more specific?
On MacOS and Windows releases, we use pixi to install environments for plugins now. This ensures everything is installed cleanly.
@ghutchis I assume everything works correctly on your mac?
Yep. Same with the students in the undergrad lab that served as beta testers. Optimization + Orbitals + Vibrations + Energy.
I tested with a simple molecule, and the orbitals function worked (attached below).
But the other functions just don’t respond. To test the function, I drag one of the carbon atoms from its original position so that, if optimisation occurs, I can see it clearly. Only the geometry optimisation under Open Babel and the one under Extensions worked.
The “python settings” does not allow any changes.
I’m not doubting you - it’s just hard to understand how some features of the plugin would work and others do not. (For example, if the calculation converges for orbitals, it should be able to give an energy.)
@matterhorn103 - correct me if I’m wrong, but it should be possible to see the last calculation on disk, right?
I reboot my Mac, and it starts working for semi-empirical QM. However, I found that the optimised and smart opt under semi-empirical methods do not always work. It seems these functions are sensitive to structural deformation. An example is here:
When draw a structure, one can end in these two scenarios:
Both optimise and smart opt accept the first one and perform optimisation, but the second one (no response at all). I tested about 10 times; Avogadro crashed twice. Open Babel does not have such an issue.
RDkit still not working.
I click the pop-up window when using the old version to upgrade, so I am not sure which one I have downloaded.
Sure. If you have a highly distorted structure, quantum mechanical methods like GFN2 will not converge - the bond lengths are too large.
I would appreciate details on the crashes.
The crashes happened when clicking the smart opt for the second structure.
Correct, that’s what the Go to Calculation Files menu command is for, it takes you to where the calculations are stored (and the most recent calculation is the one in the last directory).
@Renyun_Zhang What would be really useful would be if you could upload the xtb output file itself (last/output.out) and the easyxtb log file ( ~/Library/Application Support/easyxtb/log.log) after one of the calculations that crashes.
