@monkeyBanana2048

Decided to create an implementation of RLS. It’s kind of sloppy, but it should work for it’s purpose. It works by having hidden columns on every table that dictate what keys have access to the rows, and what permissions those keys have. Service key is able to set RLS status (enabled/disabled) and individual keys are generated per user. I think the project is ready to be shipped, because honestly SheetDB is pretty slow because it uses apps script and google sheets. So if you want to scale up and have multiple users, its better to use Supabase or something of that sort.

Finished all the required stuff to have a sorta working database, still need better auth. I was trying to abstain from AI, but sometimes CORS errors really get on your nerves. People can start using this today, but the authentication is pretty much non-existent. Future updates will probably focus around getting the authentication to be better, and maybe implementing indexing emulation and RLS. Website is deployed here. You can find more details on gh bc I already spent a lot of time writing docs lol.

Made auth better

I didn’t think putting the password as a argument in the URL itself was a good implementation, so I took the liberty to beef things up a bit more.
Now:

• Separate master key
• Separate read-only key
• Keys submitted in the body of post requests

Additionally, I found a bug with removing rows. It didn’t cascade down leading to holes in indexing so I fixed the indexing part of the script.

TODO:
Tbh, this is almost complete imo, but if I did have to do some more work, then:

• Implement RLS
• Implement indexing/cache (optional ultra pro max)

Made auth better

I didn’t think putting the password as a argument in the URL itself was a good implementation, so I took the liberty to beef things up a bit more.
Now:

• Separate master key
• Separate read-only key
• Keys submitted in the body of post requests

Additionally, I found a bug with removing rows. It didn’t cascade down leading to holes in indexing so I fixed the indexing part of the script.

TODO:
Tbh, this is almost complete imo, but if I did have to do some more work, then:

• Implement RLS
• Implement indexing/cache (optional ultra pro max)

Finished all the required stuff to have a sorta working database, still need better auth. I was trying to abstain from AI, but sometimes CORS errors really get on your nerves. People can start using this today, but the authentication is pretty much non-existent. Future updates will probably focus around getting the authentication to be better, and maybe implementing indexing emulation and RLS. Website is deployed here. You can find more details on gh bc I already spent a lot of time writing docs lol.