Tobias Megies

Hmm, the inventory says the instrument is a LE3D-5S. To me it looks a bit strange for that.. e.g. the poles don't match what Lennartz specifies for LE3D-5S in the...

No idea why that `.type` is there in the first place.. comparing `.dtype == np....` would be what I'd do

Also, numpy is advertising `is` for comparison: https://numpy.org/doc/stable/reference/arrays.dtypes.html  But yeah, somehow it is also advertising `dtype.type` for comparison..

Can you reproduce that problem without the seedlink client, after reading the problematic miniseed you are writing to disk? Having a reliable regression test would be the most important thing...

> we could hack a test that changes `dtype.type` to `np.intc`? sounds good to me :+1: essentially what mock does in testing

Never seen that `np.intc` btw.. really weird. Kind of curious how we end up with it, might be even a numpy version dependent issue? Or even something worth reporting to...

@s-schneider exception handling should be as specific as possible to avoid catch-all situations, so I made some changes. I'd still like the `except` clause to be more specific.. what kind...

Oh, I see. OK.

Hmm.. not sure about this one, and I'd have to dig deeper, not sure when I get to that. Anybody else wanna look at this one?

I guess we could try to not drop that info and see if it breaks anything in our tests :rofl: