Conversation

Notices

1. Embed this notice
   Tito Swineflu (tito_swineflu@sfba.social)'s status on Wednesday, 24-Apr-2024 01:40:12 JST Tito Swineflu

   Wow, sqlalchemy's serialization is SLOW.

   • Embed this notice
     Philip Weiss (Phil in SF) (kingrat@sfba.social)'s status on Wednesday, 24-Apr-2024 01:40:11 JST Philip Weiss (Phil in SF)

     in reply to

     @tito_swineflu

     So's Django's. Seems to be common among Python ORMs.

   • Embed this notice
     Tito Swineflu (tito_swineflu@sfba.social)'s status on Wednesday, 24-Apr-2024 01:43:30 JST Tito Swineflu

     in reply to

     • Philip Weiss (Phil in SF)

     @kingrat Beyond serialization, I usually give up on ORMs after a point because they get too hard to optimize. I wrote a tiny ORM for php that I use. It can get one table's data with where clauses and look up items by id, but nothing more.
     I find that, once a bunch of joins are involved, figuring out what's slowing down an ORM is more time consuming than just manually writing the query.