I'm just throwing this out there with no knowledge of Prisma or migration, but I had issues using datagrip over the Supabase connection (using Postgres user) and doing table change operations (not things in the table). https://discord.com/channels/839993398554656828/902772250791542785 this was the thread someone else found the solution for that. Sorry if this is not even close to what you are encountering, but with weekend coming figured I throw it if no one else comes around. https://github.com/supabase/supabase/issues/3688#issuecomment-953370874 the issue does mention I/O error....

@User Thanks!

I tried running those two statements, but get "ALTER SYSTEM cannot run inside a transaction block"

Am I in the right place trying to run them?

That is where I would have run it, and it worked, but I can't even swear exactly what I used in the train of tracking it down.... it was from the discord thread though. That was also a ways back and the environment might have changed. Edit: Wild ass guess run one at a time.

Ha! whaddayaknow. It worked 🙂

Now let's try my ALTER statement again

It worked! Thanks a lot @User!

Hi there, I just read https://supabase.com/blog/2021/07/30/supabase-functions-updates about Supabase Hooks. It seems only newly created projects have access to it. Is there a way to convert my existing project to a new one? Or do I have to set it up again? (Schema, and data, …)

You can use

pg_dump

to migrate your current project to a new one. https://supabase.com/docs/guides/database#migrating-between-projects

That way you don't have to manually create everything again

You might have to restart the server after running the migrations so that all the API docs and stuff update properly. Also, I noticed that the database extensions won't be automatically enabled when migrating, so be sure to toggle those on before doing the migration.

is there a way to clone a project

Thanks for the explanation. But what does it mean by database extensions? Can you give me an example?

This view right here. So for example if you have turned on the pg_cron extension, you should turn it on in your new project before migrating the data in.

See my link above about migrating between projects. That is the only way to do it currently

Hey, thanks! I found the issue. Basically I disabled the RLS and the realtime started to work 🙂

I've never used it so didn't know about that. Thanks!

hey. I try to use API only auth. No JS. and when I try to call: supabase_url/callback?code=&state= i get error, that its unable to exchange external code. But, if call directly github - i get response with access token. i checked values i entered into supabase dashboard. and i copy pasted them into my code. Check this image ->

repo to replicate:

and to test go to: http://localhost:3000/api/auth/github

code: https://github.com/leimantas/remix-caravaggio/tree/main/app/routes/api/auth

is there a way to change the default filename format in storage buckets?

When listening to Realtime data, is it possible to only retrieve some of the columns of the e.g. updated row? 🙂 (I have a profile table for example where I only retrieve the name.. But when it updates the listener will send the whole row down)

Does anyone know if upsert works with a different key? I'm currently migrating from firestore to supabase and in the first phase, I'd like to mirror the data and then once everything works do the switch.

supabase.from("companies").upsert([ { ...data, legacyId: "some-id" } ], { onConflict: "legacyId" })

May it work like this? Insert as usual, but update if legacyId reports conflict?

You’re gonna need a unique constrain on the columns you use

If you do, then yeah, it’s gonna be fine

Though...

> Primary keys should be included in the data payload in order for an update to work correctly.