https://wesmckinney.com/blog/looking-back-15-years/?s=09 A Very thoughtful read

Malloy like prql is really cool

I see, so Malloy is from Looker, make perfect sense.

so in this space specifically: • we have prql and malloy as much more expressive sql alternatives that compile to regular old SQL. • Toby who makes

sqlglot

also has a funded alternative to

dbt

caled sqlmesh which appears to be an interesting countera-rgument to jinja hell that we know all to well My biggest takeaway from my 10ish years in data is that all roads lead back to SQL, revert to the SQL mean if you will. People have been trying to kill it since the 70s and it just wont die. You then ask yourself the question why? The best answer I can come up with is that the barriers to entry is so low that nothing else can compete in the long term. It does nearly everything OK, you’ve got people like duckdb breaking the spec in the spirit of expressiveness. The minimum bar python programmer (kedro users for that matter) is much higher than a SQL engineer. Having someone who is able to write clean, well documented and tested code is much taller order in Python world. I’m not sure exactly where the great decoupling lands but I’m excited to see what happens.

Interesting indeed. Thx for sharing 👍🏼

Also look at this - https://github.dev/lloydtabb/imdb, install the

malloy

extension when ready and refersh until the notebooks render nicely. The coolest part is that the SQL compilation part works locally without codespaces: