Tobias Megies

Sounds good from the PR description above, I'll just need a tiny bit of time to look through the patch, hope to be able to do that as soon as...

A couple of minor questions. Otherwise good to go, I think, if CI is OK with it.. hard to say with such big edits, but it seems old tests are...

> If I understand correctly, option 3 would probably be my choice such that we essentially just juggle both and map as needed, truncating if need be. May need to...

My guess is this is a duplicate of #3320 and fixed on master branch by #3306

> What I propose, it to extend the **Inventory** hierarchy to also include site characterization metadata. This would be implemented with a set of new classes to describe the siteXML...

After looking more at the details, here is the main thoughts that come to mind that need to be figured out: ### 1. Objects not treated by StationXML That means...

> The site characterization metadata will be associated with a station by adding a new attribute at the Station class for the new top level **SiteCharacterization** class. > > A...

Just a brief summary of the most important points we discussed in zoom.. - Probably best keep it all separate from `core.inventory` and just keep everything in `io.sitexml`. - Many...

My first guess is that there might be small gaps of subsample size that accumulate across the whole day and then lead to a gap of the order of sample...

> But I think obspy reads one day at a time and calculates the end time from the first record header, the number of samples and the sample rate. If...