Ryan Abernathey

Looks like RTD has finally updated. Not sure why it was so delayed. I added the release at https://github.com/zarr-developers/numcodecs/releases, I I guess we can say 0.13.0 is officially out! @mwlon...

@d-v-b - rather than duplicating the grid_mapping attribute on each variable, it could be a group-level attribute. However, the broader problem with what you're proposing is that it will break...

Creating an [empty] array is not that expensive. Having multiple different standards for the same thing also adds complexity to implementations. Literally copying NetCDF layout, attributes, etc. is what we...

Look, I totally understand where @d-v-b is coming from! Obviously if you are starting from scratch and only considering what is the best way to map the data model to...

> [@rabernat](https://github.com/rabernat), what would make you more comfortable with this change? It's not a question of making me personally comfortable. The question is: who is going to be responsible for...

> This is fundamentally a _Zarr problem_. This is not a Zarr problem at all. Zarr says nothing about coordinates period. It's a low-level container for storing ND-arrays and metadata,...

> [xproj](https://github.com/xarray-contrib/xproj) is trying to address this issue, at least in the Xarray (broader) ecosystem. This looks very promising. I wonder what it would take for odc-geo, rioxarray, xee, etc....

> We should probably figure it out in conjunction with (hopefully) the adoption of `rasterix` in those packages as well, since both `xproj` and `rasterix` share the same goal: provide...

Thanks for the reply @Psykopear! I tried to run your example, but it didn't work. I got this error ``` TypeError: (src/run.rs:133:49) worker error Caused by => TypeError: (src/worker.rs:149:14) error...

Thanks so much for your reply. > having windowing operators able to work with any kind of orderable/comparable data type makes sense Yes, exactly, this is exactly what I was...