Damien Ayers

I've found an "interesting" undocumented feature lurking in the ODC configuration code. If you run `python -m datacube` (or call the function `datacube.config.auto_config()`, a new configuration file will be written...

The documentation for the configuration file needs improving too. The canonical source in the docs seems to be [Database Setup -> Create Configuration File](https://datacube-core.readthedocs.io/en/latest/installation/database/setup.html#create-configuration-file). It is mostly made up of...

Oh my, there's more complexity in choosing the "Datacube Environment" to use. The priority for choosing the environment is (to the best of my understanding): 1. The `env=` value passed...

Okay, after thinking I understood this now, and tearing my hair out.... there's more. The above logic is only called when creating a `datacube.Datacube()` in code. If you're using the...

This seems like a nice usability addition. Eventually people learn to work around it by specifying a really high or really low date as one end of the range, which...

Thanks Simon. Sounds pretty serious for any _pixel drill_ applications. @vnewey has discovered the same issue, I'm not sure if she has managed to dig into or start fixing it.

The plan is to deprecate `GridSpec` in v1.9 and remove it in v2, so this will no longer be relevant. Any errors will be via [opendatacube/odc-geo: GeoBox and geometry utilities...

There is code inside [opendatacube/datacube-ows: Open Data Cube Open Web Services](https://github.com/opendatacube/datacube-ows/) which handles this and should be ported into Core.

We're going to leave the PostgreSQL schema as `agdc` in the existing drive, but the new PostGIS driver already uses ODC for the schema. Closing

This is really hard to do with the current implementation, but should be possible with the database changes coming for ODC 2.0.