Denis Hirn
@kryonix.de
Postdoc at Database Systems Research Group @ University of Tübingen
Reposted by Denis Hirn
This is a briljant feature!
I have actually used it to solve a challenge I faced for a customer, and it was a lot quicker than a more traditional SQL-based solution. To be fair, it was slower than an equivalent algorithmic solution.
Anyway thanks @teggy.org and @duckdb.org for contributing this!
I have actually used it to solve a challenge I faced for a customer, and it was a lot quicker than a more traditional SQL-based solution. To be fair, it was slower than an equivalent algorithmic solution.
Anyway thanks @teggy.org and @duckdb.org for contributing this!
July 10, 2025 at 7:10 PM
This is a briljant feature!
I have actually used it to solve a challenge I faced for a customer, and it was a lot quicker than a more traditional SQL-based solution. To be fair, it was slower than an equivalent algorithmic solution.
Anyway thanks @teggy.org and @duckdb.org for contributing this!
I have actually used it to solve a challenge I faced for a customer, and it was a lot quicker than a more traditional SQL-based solution. To be fair, it was slower than an equivalent algorithmic solution.
Anyway thanks @teggy.org and @duckdb.org for contributing this!
Well… sorry, I guess? 😅 I could not find a smaller test case to trigger a bug in DuckDB‘s old bottom up decorrelation, so I’ve added this monstrosity as a test case. Do you mind sharing why do you need to debug it?
June 6, 2025 at 10:41 PM
Well… sorry, I guess? 😅 I could not find a smaller test case to trigger a bug in DuckDB‘s old bottom up decorrelation, so I’ve added this monstrosity as a test case. Do you mind sharing why do you need to debug it?