Jeremy Maitin-Shepard

Seems reasonable. Note that there are actually 2 lz4 formats, the lower-level "block" format and the higher-level "frame" format. The lz4 command-line tool only supports the frame format. blosc lz4...

Agreed that this is not intended to specify the actual API in any implementation.

No, because the in-memory representation is up to the implementation and the stored representation in chunks is up to the array to bytes codec. The bytes codec specifies 1 byte...

> In-memory representations are always up to implementations > > The point of the spec is to ensure that when we write out data others understand what we have written...

I approve this license change.

I agree that a separate `extensions` object doesn't necessarily help --- I argued against that previously because I don't see a strong benefit in distinguishing between what was in the...

@maxrjones Here are some responses to your comments: What I originally had in mind with the registered attributes proposal is that at least most registered names would include a clear...

This is nice in that it cleanly separates encode and decode parameters, but on the other hand it makes it more verbose in the common case where all codecs apply...

Variant 1 is not allowed by the spec. Tensorstore allows that as a shorthand in a tensorstore json spec but not when reading stored metadata.

> > Variant 1 is not allowed by the spec. > > per https://zarr-specs.readthedocs.io/en/latest/v3/core/index.html#array-metadata-codecs: > > > Each codec is an [extension point](https://zarr-specs.readthedocs.io/en/latest/v3/core/index.html#extensions-section) and MUST conform to the [Extension definition](https://zarr-specs.readthedocs.io/en/latest/v3/core/index.html#extension-definition)....