Dan Allan

I'm curious whether @garryod has given thought to this or encountered this interpretation of RBAC before.

Thanks, that's useful feedback. I think where @nmaytan are landing is this: 1. Tiled should grow a new column in the `nodes` table, i.e. _next to_ the node `metadata`. From...

It's interesting that that was a compromise based on: - NSLS-II wanting to store one key, `data_session`, relying on an external system to resolve groups/users - ALS wanting to store...

Discussion continued in #951, where we're focusing in on the plan.

Hello @toqduj! I agree. How's this branch work for you? https://github.com/bluesky/tiled/pull/889

This looks like a very recent backward-incompatible change in dask. Recently there have been a wave of changes to dask APIs that I would have considered "public" based on their...

Actually, I take that back. This code _does_ address a backward-incompatible change in dask, but it happened in 2024.9.0. https://github.com/bluesky/tiled/blob/819d194253a888b2a87208524675036baac8a8fa/tiled/server/etag.py#L1-L9 In the current version of dask, the function is still...

Then, the question why you are ending up the `dask.base` section of the code quoted above. You should be in the `dask.tokenize` section, I would think.

As @skarakuzu noted, we can hard-code that tabular datasets stored in SQL always have `1` partition. Partitioning has values in file-backed storage but not for database-backed storage.

Notes from conversation: 1. Delete `generate_data_sources` method, as that is used in a file registration use case (`tiled register ...` or `tiled serve directory ...`) which is not applicable here....