Simple self-cooked time integration faking the interface of OrdinaryDiffEq

[ranocha]

The self-cooked RK method doesn't support resize! yet, so you can't use AMR with it. However, I don't think that's necessary to demonstrate how to create our own time integrator, which is the main purpose of this PR.

[codecov]

Codecov Report

Merging #272 into dev will increase coverage by 6.96%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##              dev     #272      +/-   ##
==========================================
+ Coverage   79.74%   86.71%   +6.96%     
==========================================
  Files          66       68       +2     
  Lines       11885    12108     +223     
==========================================
+ Hits         9478    10499    +1021     
+ Misses       2407     1609     -798     

Impacted Files	Coverage Δ
src/Trixi.jl	100.00% <ø> (ø)
src/callbacks/analysis.jl	83.41% <0.00%> (+23.61%)	⬆️
src/timedisc/timedisc.jl	43.44% <0.00%> (-44.90%)	⬇️
src/callbacks/save_restart.jl	51.78% <0.00%> (ø)
src/callbacks/save_restart_dg.jl	48.71% <0.00%> (ø)
src/callbacks/save_solution.jl	87.93% <0.00%> (+1.88%)	⬆️
src/mesh/mesh.jl	78.72% <0.00%> (+2.12%)	⬆️
src/callbacks/save_solution_dg.jl	93.54% <0.00%> (+3.22%)	⬆️
src/solvers/dg/interpolation.jl	75.26% <0.00%> (+3.22%)	⬆️
src/equations/2d/compressible_euler.jl	91.85% <0.00%> (+3.51%)	⬆️
... and 13 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0452e33...a0ca542. Read the comment docs.

[sloede]

This is really great to have! Especially since now it will be possible to also port the 3S* scheme such that we can fix the test setups that use that scheme. I just have a few questions and some polite requests for more commentary