ORMs Bug Me 🐜

Today, I find myself thinking about #ORMs (Object-Relational Mappers). It's a complicated relationship, to say the least – love them or hate them? - something akin to a roller coaster of emotions. 🎢

On one hand, ORMs can be incredibly efficient tools, streamlining the process of database interaction and making our lives as developers that much easier. The ability to write database queries without the need to switch between SQL and our programming language? Count me in!

But, as with all good things, there comes a catch. 🤔

Every now and then, I find myself tangled in a web of performance issues, all pointing back to the ORM. Large projects - with lots of engineers - tend to amplify this issue, transforming the once helpful tool into a source of headaches and long debugging sessions.

🔍 My running theory? ORMs create a veil of abstraction that, while helpful, breaks the "locality" between the developer and the database. This distance, this gap in direct interaction, leaves room for mistakes and inefficiencies to creep in.

Sure, ditching the ORM and going back to raw SQL might slow things down initially, but let's not forget: the bulk of a software's cost comes post-development, during maintenance. Are the upfront savings really worth it if we pay the price later?

I’m not quite ready to part ways with my trusty Django ORM, but I’d be lying if I said the thought hasn’t crossed my mind.