That’s not true, they have been developing it as red knot for a good while :)

Like I said, it only got its name a few weeks ago.

You basically pay databricks a “fee” to choose the more appropriate and modern stack for you to build on, and keep it up to date. Never used it, but it handles with lots of the administrative bs (compliance, SLAs, idk) for you so you can just ship.

This looks so dope!

sqlalchemy doesn’t really accepts strings - if you do, you need to pass them into a “conn.execute(text(…))”, so end users should not face a breaking change.

Type checkers to the rescue ahaha I think db.get could also raise if the type does not match?

OpenAI coping so hard

he's coping so hard.

Oh you maintain PyTorch right?

I think GP is https://www.ekzhang.com/ (hi Eric!), not http://ezyang.com/

Absolutely. No second thoughts.

also a whole bunch of new features (--script mode), speed of development, a stacked team.