Tom Nicholas

Sounds good. I'm going to have to remind myself of how all this works 😅 My email is Tom at Earthmover dot io

> I think there should just be one way to specify a backend. Could we consolidate these parameters? I agree that this has become messy. Could consolidating everything just look...

> I would prefer a keep_variables argument over a drop_variables argument (similar to this https://github.com/pydata/xarray/issues/1754). I don't really have a strong opinion about this, but I do think there will...

> Why do you want to work with ReadableFile? It would be better to work with the raw obstore calls. I'm struggling to imagine what that would look like in...

> I would be most excited to open an obstore-backed alternative to https://github.com/zarr-developers/VirtualiZarr/pull/458 That did not occur to me but it's an intriguing idea!

This is still using the wrong installed version 😞 We would appreciate you help with this @maxrjones

Now that https://github.com/zarr-developers/zarr-python/pull/2874 was merged upstream our normal upstream tests should be able to run against the updated zarr-python. However I messed this PR up by merging from `main` (instead...

I love the forward-thinking-ness here, and I've also been mulling over what the world of findable virtual zarr stores could look like. However, whilst I agree there is a lot...

I'm pretty agnostic as to which GH org the parsers live in. I think one argument against putting them into `zarr-developers` is how I keep having to ask @joshmoore to...

The [format specification](https://huggingface.co/docs/safetensors/index#format) seems very straightforward. If it really is that simple then I think this reader could potentially be implemented without even using the safetensors library, instead just using...