Ian Rose

The `IDrive` interface is purely client-side. The `jupyterlab-github` extension does have a small server-side component, but it is currently just used to attache authentication information to the GitHub requests. The...

I did some experiments on this PR yesterday to try to draw out a bit more when using a shuffle-based approach is a good idea (warning: highly-faceted charts incoming to...

Everything looks green except for gpuCI, which is unrelated: #9391

Thanks for promoting this to an issue @jsignell. @pavithraes and I have seen issues with `sorted=True` before (cf. https://github.com/dask/dask/issues/8999#issuecomment-1150087728). I'm not 100% sure that is related, but linking so I...

I agree that the issues around custom `__init__` functions and `init=False` fields make it difficult to fully support traversing, unpacking, and rebuilding dataclasses. I am a bit agnostic as to...

xref: https://github.com/dask/dask/issues/8377

> but I have a feeling we may have better luck encouraging the use of map_partitions when the user/dev knows munual fusion makes the most sense . Yeah, that's fair....

I have no objections to this content (even if I have some personal disagreements about squash merges :wink: ), thanks for writing it up @jrbourbeau! I wonder if some additional...

Reported upstream [here](https://issues.apache.org/jira/browse/ARROW-16838). I think changing to `schema=None` (or supplying a manual schema to `ddf.to_parquet`) is the correct workaround until it is fixed upstream.

Okay, as I feared, this does indeed make cupy unhappy -- I'll work on pushing some of this into the dispatch system