Pol Febrer Calabozo

I liked the custom class thing, because there the user explicitly gets what they want and there is no confusion, while you can still easily get the neighbors in supercell...

Cool, I think it's the best approach seeing that there might be different users interested in different things. And also it makes sure that the user knows what they are...

What do we do with the `find_neighbors` method? That one returns a list where each item contains the neighbors of a given atom. Should each item in the list be...

This goes back to the never ending discussion :smile: My view is that as long as you have periodicity, it makes sense to translate to the unit cell regardless of...

Even more so because `nsc` is many times [1,1,1] because it is simply unset, but we know that there is periodicity. For example when a geometry is read from a...

For the moment they should, right? SIESTA assumes that everything is periodic when it calculates things. But yeah when the PSolver makes it possible to not treat things periodically I...

I would see this being useful if some function is not implemented in ASE, or it has an API that is more complicated to acheive a certain goal. If someone...

I will leave the issue here, I don't want to work on it for now, it was just an idea.

Hmm ok, but could we have some kind of wrapper where we get the type annotations and help messages from the function definition so that we don't have to redefine...

There are some failing tests, but I don't know if they are related to this PR. Apart from that, could we set `nsc=[1,1,1]` in the tests that check the modified...