We're using django-stubs and we've been pretty happy with it. We've started the project with it, which was fairly simple. There's one or two model field types that are not supported but those can be skipped. We're using pre-commit to run mypy, this makes sure that we never commit a state that isn't type-checked.

Idk if its just me but I find the testing frameworks for expressJS to be _awful_ -- chai and mocha work great but throw in sinon and its like pulling teeth.

The lack of good documentation is painful and often treats me like I don't already know the concepts involve.

Sequelize isn't as intuitive as Django's ORM either imo.

I recently set up a good old Django backend with REST framework. Frontend in Next.js

Turned out to be painfully slow to iterate. The ORM needs constant attention to join or prefetch queries.

Switched to Prisma and it's an utter joy. If the typescript has no errors then there won't be runtime errors.

Sequelize is a multi generational mess. It's a huge time waster.

> Sequelize is a multi generational mess. It's a huge time waster.

Could you elaborate on that ?

I personally like the functional aspect of sequelize.

Just that there are several generations of API. This makes TypeScript pretty much useless / impossible.

Working with Prisma: if it compiles (and you don't have a business logic error), then you usually won't get any runtime sql errors.

You don't have type safety with sequelize.

Ah yes, that's definitely true, automatically generated getters and setters are a pain with sequelize. It's not impassive to have type safety though, it's just complicated.

I had to write my own types around the existing ones to get getters and setters to work.

prefetch_related and select_related are pretty straightforward, but for more complex stuff the Django ORM can be a hindrance.

I haven’t checked out Prisma. Thanks for the tip.

Sure, I always have to spend time checking logs and tuning the queries like that. By default, Django selects everything. I used to discover that joins and prefetches were hauling in large text fields when I only wanted an id or a slug.

Prisma has a great API for this, and it's easy to tune. By default is prefetches and selects related even several levels of joins deep.