databend VS polars

Databend vs. Snowflake: https://github.com/datafuselabs/databend/issues/13059

Indeed, under a pay-as-you-go model, if there's a lack of precise control over the warehouse, such as a 10-minute suspension, it could lead to significant waste. This is because most queries might only take a few seconds, and the rest of the time is wasted. If you find Snowflake expensive, consider Databend. It's an open-source, cost-efficient alternative to Snowflake, and it maintains a consistent product experience with Snowflake.

Open-source: https://github.com/datafuselabs/databend

They're advertising as an open source direct competitor with Snowflake, with the ability to store data in parquet files. Github repo (5.6k stars) here.

Databend has recently introduced an SQL function that generates SQL statements from natural language. This feature can significantly reduce the time required for writing and debugging SQL statements.

Hi, we used to use sqlparser in [Databend](https://github.com/datafuselabs/databend). But at last we decide to write our own sqlparser using nom-rule.

Link to Github: https://github.com/datafuselabs/databend

This is because 0.1 is in actuality the floating point value value 0.1000000000000000055511151231257827021181583404541015625, and thus 1 divided by it is ever so slightly smaller than 10. Nevertheless, fpround(1 / fpround(1 / 10)) = 10 exactly.

I found out about this recently because in Polars I defined a // b for floats to be (a / b).floor(), which does return 10 for this computation. Since Python's correctly-rounded division is rather expensive, I chose to stick to this (more context: https://github.com/pola-rs/polars/issues/14596#issuecomment-...).

https://github.com/pola-rs/polars/releases/tag/py-0.19.0

That turned out to be related to pola-rs/polars#11912, and this linked comment provided a deceptively simple solution - use PARSE_DECLTYPES when creating the connection: