Hi! I am working with PsychoPy on a presentation computer (Win10) for MRI, this computer is not connected to internet.
We currently have a multisite study ongoing at our centre that very specifically needs a PsychoPy version from 2023 because of its reliance on “Advanced Vision Science” plugin. I have since been building experiments using PsychoPy 2024.2.4 that will not run on that version
I have found that updating PsychoPy and then adding the following code to the old study fixes the problem
from psychopy import plugins
plugins.activatePlugins()
However, given that it is a multisite study, there is resistance to changing/adding to the code at all.
I am trying to find the best path forward, prioritizing this multisite study and its strict guidelines, and I cannot in good conscience change anything on the presentation computer without considering the impact on all of the ongoing studies. So I am looking for help.
I have seen discussion of using “use version” to switch between versions within one stand-alone install Best practices for using multiple PsychoPy versions on one computer But would this require me to specify which version in the multisite study and any other older psychopy studies as well?
I have also seen it suggested that I could rename the older standalone version, and install a new one, thus having two stand alone versions. But is there a way to keep the older version as the default (the one named PsychoPy) and have the option of a newer one for newly built studies?
Thank you so much for your help on this. I did try to run the experiment i programmed in 2024.2.4 on the 2023 version and it gave me a very helpful error, something like “it looks like you made this on a newer version that includes functions that do not exist in this version!'“
There are other studies coming in that want to use PsychoPy too, so I would like to have the new version but make it backward compatible with older studies. Maybe the best option is to get the new one, then troubleshoot any old ones that will not run.
as long as your experiment runs, you can ignore this warning. It can become an error when you use a parameter or function that does not exist in an older PsychoPy-Version.
I think so too. Sticking to old versions for too long usually ends up in a lot a work. Rather make gradual changes.
I think you are misunderstanding what Jens is saying
Your base psychopy version would be the most recent stable release (e.g. 2024.2.4). Then for older experiments, you open it with that “base" version, and would select an older version of psychopy (e.g. 2023.2.3). Within the settings window (see below):
Yes, I understand this, thank you. At this point, I updated to the most recent stable release (2025.1.1) and have checked/fixed as many of our psychopy-based experiments as I could find.
I was just looking for the least disruptive solution for our centre.
