two-scale-heat-conduction DuMuX dependencies setup

[MakisH]

In the two-scale-heat-conduction tutorial, we currently have:

• Run scripts that depend on the custom DuMuX scripts macro_dumux and the micro-manager.
• A script setup-dumux.sh, which:
  • Removes any previous builds
  • Gets and builds DuMuX 3.7 in the current directory
  • Gets and builds the DuMuX adapter, the module dumux-phasefield, as well as the external dependency spgrid
• A script compile-dumux-cases.sh which:
  • calls make on the (already built?) macro-dumux and micro-dumux modules
  • moves the built binaries to the respective case directories

In the rest of the tutorials, we don't provide such scripts. For example:

• For deal.II cases (see perpendicular-flap and run-dealii.sh), we ask the user to install deal.II and the deal.II adapter. If these are not found, we inform the user.
• In the Quickstart, we ask the user to install all the dependencies, hinting them how to do that.

There are multiple reasons to do this like this:

• Such scripts tend to be very system-specific. Already the line mv micro-dumux/build-cmake/appl/micro_sim.cpython-310-x86_64-linux-gnu.so micro-dumux/ demonstrates that.
• If such a script fails, then the user is clueless and can only blame preCICE (us).
• Such scripts might conflict with our documentation on how to install the respective adapters. For example, I already built the DuMuX adapter in the home directory of the VM, but now I am asked to repeat such an installation in a local folder.
• We have to maintain such scripts.

I think we can improve the situation here:

1. Inform the user what the dependencies are
2. Inform the user that the solvers need to be built
3. Describe how to build these using a given DuMuX adapter installation

We can still provide some helper scripts, but I would rather not. Instead, I would focus on demonstrating in some CI workflow file / Dockerfile what steps to take to run the tutorial. This could be demonstrated in the system tests, or in the CI of the micro-manager.

cc @IshaanDesai @Fujikawas

[MakisH]

Maybe interesting for @mathiskelm as well.