[WIP] fix: possible fix for CCI breakage (#1238)

* [WIP] fix: possible fix for CCI breakage

This is intended to resolve #1237.

For some unknown reason, when I execute the notebook at
`examples/workflows/plot_simulate_somato.py`, after installing
`mne-1.11.0`. I can reproduce the issue. The issue reproduces only if
Python is set to 3.13 or 3.12, but not 3.11.

I think the key line in the script that is the problem is this
https://github.com/jonescompneurolab/hnn-core/blob/master/examples/workflows/plot_simulate_somato.py#L45
I honestly don't know why this is suddenly an issue *now*, since it's
been months since we changed the CircleCI Python version we use, and
this notebook hasn't been touched in 3 years, and as far as I can tell
the version of MNE being used in our CircleCI runs hasn't changed
between this week and last week (both 1.11.0).

I will note that the error only occurs if there is no pre-existing
folder named `MNE-somato-data` in your `~/mne_data` folder. You must
delete that folder every time in order to redownload it, and trigger the
bug. The bug doesn't appear if you already have that folder.

This PR may be good enough to fix our issue, but it's possible that
we've identified some sort of weird upstream bug with `from mne.datasets
import somato` that only occurs in rare situations. We should
investigate and then tell MNE about it if the problem is inherent to
them instead of to our notebook here.

* debug: force full CircleCI rebuild py3.11

Author: asoplata

.circleci/build_cache

@@ -1,2 +1,2 @@
 # Change the following version number to trigger a complete rebuild of the doc in CircleCI.
-v4
+v5

.circleci/config.yml

@@ -28,7 +28,7 @@ jobs:
               command: |
                   export PATH=~/miniconda/bin:$PATH
                   conda update --yes --quiet conda
-                  conda create -n testenv --yes pip python=3.13
+                  conda create -n testenv --yes pip python=3.11
                   source activate testenv
                   # necessary due to https://github.com/mpi4py/mpi4py/issues/335#issuecomment-1486366039
                   rm $CONDA_PREFIX/compiler_compat/ld