Jeremy Cohen

@ssimeonov It's a really fair question! 1. Existence (+ relation type) is the only information dbt needs for all relations _before_ running, to populate its runtime cache. As each model...

Yes, I think that's right. Let's aim to include this change in a v1.1 beta, and ask folks to test it out in their own projects to ensure it yields...

dbt uses `show table extended like '*'` for two purposes today: 1. To cache relations at the start of each run, whether for one model or many 2. To access...

@ryanClift-sd Thanks for the amazing comment + experimentation! I agree, I'm leaning pretty strongly in the direction of approach 2 (`show tables` + `show views`). It performs _much_ better at...

Thanks for the context there @ssimeonov. I agree, dbt should be able to pick it out as either `database` or `namespace`, whichever is available. I sense that this may be...

Thanks @TalkWIthKeyboard!! Yes, I'm happy with the simpler if/elif/else solution. I don't imagine they're planning to change this again anytime soon.

@oleksandrkovalenko This is really interesting. It is possible to reimplement/override the basic materializations, by copying-pasting-editing from `dbt-presto` into your own project. In particular, this is the offending bit of logic:...

@nssalian Is there any value in using a hosted instance of Apache Spark (e.g. AWS EMR), instead of maintaining these Docker images ourselves to run in CI? Or possibly in...

I don't think we should do this :) Newer MSFT databases should support nested CTEs (https://github.com/dbt-msft/dbt-sqlserver/issues/457). We shouldn't expose the full `Compiler` interface to adapters (https://github.com/dbt-labs/dbt-core/pull/9134).

After discussion with @callum-mcdata, let's pause on removing the deprecation warning & backwards compatibility for now, given greater changes that we're envisioning for the metrics spec in the next few...