I've been learning GraphQL (and various other parts of the stack) over the past week or so and I feel I'm finally over the "oh god how does this all work" part.

I think the abstractions you can get in the Node/Javascript ecosystem now-a-days is pretty bonkers. I basically just go "I define my database schema in a file, then I create my graphql objects/queries/mutations in code" and it just... generates the endpoints, queries etc for you.

It's really odd where I've often had a "DIY mindset"

However it's also quite nice to break out of that, especially for hobby projects. I often found myself getting bogged down in the weeds, writing the best framework to suit my needs, doing tons of research on new tech and just burning out on the project.

Whilst I did also learn a bunch of new tech during this project, it was a more "how do I use it" instead of a "how do I implement this".

I actually really like GraphQL as an endpoint, but I can see how painful it would be without frameworks

For those curious, the rough current stack:
- NextJS (honestly just using it host the GraphQL server atm, I still need to convince myself to do the frontend...)
- GraphQL Yoga
- Prisma (for model management and database access)
- Pothos (for generating GraphQL types/interfacing with Prisma)
- PostgreSQL

And that's actually about it. It's Typescript (wouldn't dare to try this in just JS) and a few other util libraries for parsing but thats it