Maximilian Linhoff

@mhvk Any idea how to deal with these endianness problems?

Ok, I think I solved it (bruteforce, by switching endianess and switching back, however, this is still faster than the previous numpy code, just a bit slower than the native...

I also added a cythonized version of `Latitude._validate_angle`. The benchmark in #13479 now looks like this: Main: ``` Longitude 50 µs ± 816 ns per loop (mean ± std. dev....

> More on the implementation, I wondered if you had considered making this a ufunc? It has two advantages, that numpy deals with the shape, and that astropy can deal...

> We use some ufuncs, but all are quite a bit more complex than what is needed here. Best example probably is https://github.com/astropy/astropy/blob/main/astropy/stats/src/fast_sigma_clip.c You just made me realize I wasted...

> As for moving the shape-stuff into cython, that would indeed just be to keep the python implementation simple; it would also mean that one could more easily replace with...

I am confused about the test failures... the code contains explicit checks for the byteorder and swaps if not native, yet I get these errors: > ValueError: Big-endian buffer not...

> `>>> # Find all neutral beauty hadrons` > `>>> Particle.findall(lambda p: p.pdgid.has_bottom and p.charge==0)` I find it strange that something like `has_bottom` or `charge` lives on `Particle.pdgid` and not...

Does that return int or Foo for the second case?

> Hence we are currently blocked by that. We hope to get this merged up and a new version released as soon as the required packages become available. conda-forge had...