Tobias Megies

rebased and force pushed

> About the redirection, it was the case, but a lot of clients does not follow HTTP 300 response, so we removed it. I think we had that discussion somewhere.....

I can't follow, what's the problem exactly?

OK, I see. Well this is embarrassing. I added this functionality on request by someone and asked them to please check it out. So much for the urgency of that...

Should be fixed. But this functionality is still untested. Would be good if somebody took the time to contribute a test.

Hmm.. actually I added test data already in `sac/tests/data/IU_ANMO_00_BHZ.sacpz` and `sac/tests/data/IU_ANMO_BH.sacpz` that can be tested against `station/tests/data/IU_ANMO_00_BHZ.xml` and `station/tests/data/IU_ANMO_BH.xml`. Just needs a test written down..

If you have everything in place to compile ObsPy from source you can.. ``` bash $ conda remove obspy $ pip install https://github.com/obspy/obspy/archive/releases.zip ```

Can you add `obspy.signal.invsim.pazToFreqResp` to the comparison (see http://docs.obspy.org/tutorial/code_snippets/frequency_response.html)? Just to avoid anything coming in there from your low-level use of scipy routines..?

Actually, I also had some issue recently with a memory leak, didn't have time to track it down so far, though.. maybe it's the same reason.. (was also processing across...

Yeah, let's drop the mem leak issue here, please open a new ticket for it with a small self-contained example if possible..