After the recent somewhat discouraging stuff with clickbench that made the news, I must say this blog post is very cool.
Sounds like you found an excellent way of structuring this problem to solve it with Claude (for those who don't click into the article - writing hand/claude-optimized Rust code for each query, checking its assembly, and trying to make their own optimizer generate something close to that same assembly), and the writeup is very interesting and readable too!
Kudos, looking forward to more blog posts like this one.
agumonkey•15m ago
Are the pgrust guys lisper historically ?
code_martial•11m ago
Not answering your question directly, but FWIW, that’s the default IR notation Claude reaches for.
agumonkey•7m ago
oh really ? I never poked at that kind of problems using claude. still interesting that claude's training placed lisp at the top for IR stages.. even though historically it's obviously one of the culture that worked on interpretation and compilation the most, it's not mainstream. But I guess for sql engine teams, it decided that it was perfectly on point.
cube2222•18m ago
Sounds like you found an excellent way of structuring this problem to solve it with Claude (for those who don't click into the article - writing hand/claude-optimized Rust code for each query, checking its assembly, and trying to make their own optimizer generate something close to that same assembly), and the writeup is very interesting and readable too!
Kudos, looking forward to more blog posts like this one.