Hans Moritz Günther

Astropy is currently running a user survey to find out what operating system and version use. We don't have answers yet, but I've seen an uptick in users asking about...

While reviewing #2252 I've taken ad look at the scripts that we use to generate (parts of) `sherpa/astro/xspec/src/_xspec.cc` and `sherpa/astro/xspec/__init__.py` or the first time. So, please forgive me if what...

Note: This is not the code in main, but already adjusted for a PR I'm working on right now, but the change compared to main is minimal and the conclusion...

It seems that #2309 doesn't change anything - but again, it's possible that I have some odd setting in `.sherpa.rc` or something like that - I'll need to investigate more.

The mystery deepens: It fails in one of my conda environments, but passes in another with the same Python and Numpy version. They also use the same environment, but clearly...

In its current state, this PR also changes the behavior of three-column tables that are read with `ncols=3, dstype=Data1DInt1`. Before this PR, that would give ```Python import numpy as np...

Since this is for **guessing** not for fitting, I suggest to consider a simple approach. If someone needs a more complex model setup, then they likely have prior knowledge on...

Interesting, because I run into that error when I tried to build XSPEC a few days ago. **Edited to add (because my fingers press enter too fast**: while using the...

I think we should actually switch both io and plotting backends to use entrypoints once we drop Python 3.9. If we do that, we'll have to do a slight rework...

And while we are at it, can we have a `TemporaryIOBackend` context manager al la #1382?