Oliver Ruebel

Attaching things in HDF5 is useful, in particular for MatNWB and folks that look at the raw HDF5. However, I think for PyNWB it may be useful not to just...

Defining some include_type statements in the tutorial sounds find. For warnings, I think we'd need to create an error when including a namespace that contains typenames that are already defined....

I don't think you should need to have to search across namespaces, but all types should appear in the namespace as they are included there and there should never be...

Could also briefly mention ``/intervals/invalid_times`` and other time intervals tables as appropriate (even if it just for further reading)

In addition to updating tutorials, I think this goes in the direction of having documentation for "Best practices" It may very well be that the best practices would just point...

@ajtritt can you have a look. I think this looks Ok. The details on specific needs/preferences for this sort of metadata are a bit out of my expertise but it...

@bendichter thanks for working on this. I assume you are referring to https://github.com/NeurodataWithoutBorders/pynwb/issues/522 on the PyNWB repo. I don't think we should use autogenerated classes in the main PyNWB API....

@bendichter are you planing to work on this in the next month or should I change the the milestone to 2.x?

Thanks @bendichter. Just trying to finish triaging tickets.

@bendichter can this issue be closed?