Docker Workflow Enhancements #1081

Problem it solves

A follow-up to #1080,
Better Tracking of MFC Containerization, More Containers, and Docker Hub README Auto-Updating.

Proposed solution

• Simply put, every release is containerize with specific build instructions, but there is no way yet to test for corrupt binaries. GitHub-hosted runners' limits can be exhausted easily and can't test all variations. Therefore, I propose a weekly/bi-weekly CI test with the (latest-gpu) images on the self-hosted runners & (latest-cpu linux/arm64) on macos-latest runner.

• Pushing previous releases for MFC GPU OMP, and updating docker.yml to do so.

• Updating Docker Hub repository description with this GitHub Action (https://github.com/peter-evans/dockerhub-description) which should reference a .github/README.md, instead of manual updates.

Alternatives

N/A

Additional context

Those items and more can be implemented to be more robust with docker use.
v5.1.3 failed unnoticed for about a week now.

------
 > [linux/amd64 4/6] RUN echo "TARGET=cpu CC=gcc FC=gfortran" &&     cd /opt/MFC &&     if [ "cpu" = "gpu" ]; then       ./mfc.sh build --gpu -j $(nproc);     else       ./mfc.sh build -j $(nproc);     fi:
0.071 TARGET=cpu CC=gcc FC=gfortran
0.121 mfc: ERROR > Python 3.10.12 (python3) is out of date. Required >= 3.11.
------