Achilleas Koutsou

I think the mistake I make in NIXPy isn't replicated in NIX: The creation of DataSets with size `(0, ...)` even when we know the size. I think fixing that...

With #432 the DataSets backing metadata properties are now initialised with the shape of the initial data (and defaults to 8 if there are none). We should keep this issue...

On a similar note, the same assertion error is triggered when trying to store an empty `nix.Value`. ``` python >>> nix.Value() Value{[Nothing] } >>> section.create_property("test", nix.Value()) ... nix/backend/hdf5/PropertyHDF5.cpp:305: static nix::hdf5::h5x::DataType...

Right, no null-type in the backend. Understandable. By catching the error earlier, do you mean removing `Value{[Nothing] }` so that the error is caught on creation of the property value?

Empty properties can be created if a dtype is specified. E.g., ```python sec.create_property("empty", values_or_dtype=nix.DataType.String) ``` Should we close this?

Hey Ajay. I think I know what's going on here. This line ```python replace(micro, "u").replace(mugr, "u") ``` is meant to replace the character `μ` in a unit string with `u`....

Cool. Thanks for the extra info. I'll try to poke this bug to death later today... or tomorrow. Or soon. Most likely _soonish_.

Hello @s0nskar. That's an issue that arises when using the C++ bindings (`backend="hdf5"`) instead of the pure Python backend (`backend="h5py"`). The problem here is that some functions simply call the...

https://github.com/G-Node/nixpy/blob/632dae8ab46a770111b529c2c8ee5c4acb87aaab/nixio/file.py#L148-L163 Initially, add a new UUID in the header with key `"id"` when creating files and check for it in the `_check_header()` method.

I vote it be a separate module completely. I'd like to avoid having nixpy (nixio) depend on matplotlib and scipy. A second module called nixtools or something similar with submodules...