Martin Frausing

@lwouis I think that could be a good solution. You had good points for my original suggestion and I have to admit I didn't consider them before posting this issue....

Example of an query where the performance between subselect and join matters (https://pastebin.com/raw/gw7DmXuD): `db=[db]# EXPLAIN ANALYZE SELECT f.id, d.id FROM feature f JOIN dna d ON f.dna_id = d.id; ┌─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐...

+1

It seems like it is mostly __perlbrew_set_env that is causing the slow down, at least in zsh