Fix gcc version to fix c++ standard mismatch

[VincentRouvreau]

With the proposed environment.yml, I had gcc 9.5.0 installed and an error:

Warning in cling::IncrementalParser::CheckABICompatibility():
  Possible C++ standard library mismatch, compiled with __GLIBCXX__ '20191114'
  Extraction of runtime standard library version was: '20220527'

cf. jupyter-xeus/xeus-cling#415 (comment)

I fixed it by setting gcc=9.4.0 in environment.yml

[nthiery]

Ah yes, I had fixed this in my classroom environment, but forgot to backport it
to this repository. Thanks!

Glad to see someone else is looking at this repo :-) Maybe even using it?

[srpgilles]

@nthiery At Inria Saclay we are using (vanilla) Laby for scientific outreach, using a cheatsheet which is really akin to your documentation.

Something unsatisfactory is that we do not provide a way for students who would like to keep playing with it at home an easy way to install it, especially (but not only) for user not using Ubuntu.

By looking into the issues I found the issue you submitted and the discussion you had there with Stéphane Gimenez, and we (@VincentRouvreau and I) investigated a bit. I was stuck when trying to use it but @VincentRouvreau found the reason and submitted this MR.

So to put in a nutshell, no not using it at the moment but we will have a look at it.

[nthiery]

Something unsatisfactory is that we do not provide a way for students who would like to keep playing with it at home an easy way to install it, especially (but not only) for user not using Ubuntu.

Yeah; that's exactly what prompted me to port laby to Jupyter.

So to put in a nutshell, no not using it at the moment but we will have a look at it.

Cool; feel free to ping for help: it's working for us here, and we did not invest much yet into making sure it
work flawlessly for others, but if there is interest for it, that's a good incentive :-)

You may be interested by the Python version which is less demanding in bleeding-edge
technology (xeus-cling). Also Python presumably fits better scientific outreach. It also allows -- experimentally --
block programming.

https://gitlab.dsi.universite-paris-saclay.fr/nicolas.thiery/Laby