Tom Nicholas

I would be - this is something that @abarciauskas-bgse and I want to get funding to work on... Don't let that stop anyone else having a go in the meantime...

VirtualiZarr isn't the right layer for that feature - Zarr itself (or Icechunk) would have to support it on disk.

> I opened https://github.com/pydata/xarray/pull/10000 as a side PR. Issue number 10,000! > More functionality can be implemented in follow-up PRs I support merging this as soon as possible, because it...

> namespace packages I like this idea, with the one exception that `xarray` itself is our namespace for xarray manipulations. > The fact that we have two ways to read...

Regardless it seems like you're a fan of the idea of making xarray an optional dependency @maxrjones ? (With #389 as a fairly separable issue)

Thanks @scottyhq ! > lightweight metadata mappings that can encode many things (file and byte locations, arbitrary metadata, "on read" computations like scale and offset, subset, reprojection). I see the...

> VRT is already an extremely lightweight virtualization I don't think the idea here would be to replace or add a new layer, but instead to create tools that can...

Coming back to this conversation from #432 and in a post-icechunk world. The recommended serialisation format for VirtualiZarr is now effectively Icechunk. But the points @mdsumner and @scottyhq raise above...

> Can GDAL work with virtualized Zarr V3? I'm going to focus on 2 :) (It would be nice to discuss this on the icechunk issue tracker instead, because it's...

Cool! Might it be possible to write a virtualizarr reader for this just using the python xml standard library module? That's basically what the [DMR++ reader](https://github.com/zarr-developers/VirtualiZarr/blob/develop/virtualizarr/readers/dmrpp.py) we have does.