David Schneider

We still face occasionally failing acceleration tests (see e.g. [this job run](https://github.com/precice/precice/actions/runs/3996108195/jobs/6855773734)). mac users also reported, that the acceleration tests takes sometimes suspiciously long.

Not sure about the hack: using a conservative mapping will mostly be wrong or must be well handled in the adapter: a conservative mapping to the SU2 mesh will lead...

I stumbled across the same issue and dug a bit deeper. I think the problem here is more with our building script. We use the following lines to 'detect' success...

> But the intention with asking ldd was to check that everything will work even building, when loading the adapter. > > Do we need an additional check, a different...

I set up a test cases on SuperMUC NG in three different cores sizes: - input mesh size 192 cores, output mesh 480 cores - input mesh size 192 cores,...

We still need to find a solution for storing the event files, i.e., creating a directory.

In my opinion we should replace the raw `bounds` vector by two vertices (representing the corners of the bounding box) and clean-up the remaining functions accordingly. The comparison operators etc...

Thanks a lot @playcatchup and @jeremiahbintang, The reason why this issue is a 'good-first' issue is primarily that you don't need to look into the whole preCICE library, but almost...

> I like, but why not also making the backend an attribute? Then, we could define a default in the long run (probably partition-of-unity). I'm not sure whether we want...

> max-iterations There is no such a tag in the mapping, so that's a good question. However, solver tolerance and similar attributes will only be required for one type. POU...