Chris Allan

Thanks, @copiesofcopies. The BEIPA is the best public representation of the intents we want to convey. As constructed it may just not work well outside the US. We're global so...

Firstly, I'd just like to echo @jbms in saying that this is sorely-needed functionality. It's exciting to see thought and real effort being put into chunk localization. I would however...

If we've settled our collective intent to be solely addressing the read-optimized, chunk-localized use case I will definitely table the TileDB discussion in favour of not creating additional noise here....

Thanks, @jbms. With respect to Google Cloud Storage and Azure Blob Storage "conditional reads and conditional writes" are you referring to the optimistic locking options (such as `ifGenerationMatch`) available to...

> Re: conditional reads and writes: Yes, I am referring to the ifGenerationMatch and similar options. I was not proposing that zarr-python necessarily make use of optimistic concurrency, merely that...

See also #857.

I think I've managed to convince myself that @jcafhe is 100% correct. That is: > Thus, the Array.__setitem__ method is not atomic. There is probably room for the [documentation](https://zarr.readthedocs.io/en/stable/tutorial.html#parallel-computing-and-synchronization) to...

> An Amazon Elastic File System can be shared between Lambdas within a Virtual Private Cloud. Yes it can, however that just feels like using a sledgehammer to kill a...

Thanks, @jcafhe. Certainly the general consensus of how to deal with deadlock for multiple lock acquisition scenarios like the one you mention is to formalize the lock acquisition order. However,...

For our projects the complexity of dealing with the blosc native library dependency has certainly come up numerous times (glencoesoftware/bioformats2raw#34 for example) and the default we currently have to say...