René Widera

We need to make the domain object a general object, maybe in pmacc and the helper too. We should not access the classes from another plugin.

Often the problem is coming from broken chunking in HDF5. This could be a solution: https://github.com/ComputationalRadiationPhysics/picongpu/issues/4845#issuecomment-2009408453

note: `consteval` is making very often problem e.g. if you build something line `getAccName(acc)`. `constexpr` works in many cases better.

please have a look into this PR https://github.com/ComputationalRadiationPhysics/picongpu/pull/5158 If the data get moved to `validation` folder within the setup they will be copied. To be copied on `tbg` call the...

I like the second idea to let the destructor optionally wait for an event or queue. We could add a method like `buffer.destructorWaitFor(any)` and keep within the buffer a list...

As documentation that I not forget about it during my vacations, this is how it works in alpaka3. In alpaka3 everything allocated with `alloc()` is a ManagedView. It tracks the...

The difference is that a Buffer allocation wull happen on the Constructor Call andd not via a free allocation function. The free allocation function is very similar to a C...

@fwyzard Can you point me in our meeting about alpaka2 to this topic again, please. For math functions, I have a solution.

Yes, the calorimeter plugin is only collecting particles those leafing the global domain on absorbing particle boundaries together with all particles within the domain. Monitoring only leaving particles is not...

@zwjlpi @ikbuibui opened a PR to create histograms out of particles leaving the outer simulation box https://github.com/ComputationalRadiationPhysics/picongpu/pull/5039 It is not collecting particles on periodic boundaries or reflective boundaries.