Benjamin Rodenberg

I tested the partitioned heat equation (https://github.com/precice/tutorials/commit/15338dea52f3f512f081eed69eb36fed833a2b78) with the quasi-Newton implementation provided here and it did not converge. The following changes are applied to `partitioned-heat-conduction/precice-config.xml` compared to https://github.com/precice/tutorials/commit/15338dea52f3f512f081eed69eb36fed833a2b78: ``` -...

> A more memory efficient approach could be to internally use the existing bspline to calculate the derivative using finite differences or similar. One thing that worries me with finite...

We should also keep in mind that we only enforce $C^0$ continuity at the window boundaries. If we offer an API that allows users to access derivative information this can...

This is the use case I had in mind: https://github.com/precice/precice/blob/14839fa6353ad468417a52af9309513e818fe303/tests/serial/time/explicit/parallel-coupling/helpers.cpp#L63-L69

To me formalizing our architecture by introducing such a data centric approach sounds like a good idea. The whole issue reminds me a bit of how you create visualization pipelines...

I just stumbled over this issue, when trying to develop a safe system for the configuration of coupling schemes and interpolation/extrapolation. Currently there are many `int`s and it is very...

Strong types could have saved us here: https://github.com/precice/precice/pull/1590 (I mixed up a mesh ID with a size in the test)

Since we removed extrapolation in #1503: Does this issue need any updates? I saw some second order extrapolation above.

What's the status here? I guess it would be nice to use two-level initialization as the default option and delete the alternative implementation with the breaking release, because it should...

Shifting it to `3.x.x`, because it is not critical in my opinion. Feel free to shift it back to `3.0.0`, if you are willing to work on it.