Nice! Great to hear!😊 But does Supabase have api methods for e.g sum, avg and group by? If not, is it correct that you can write custom SQL queries to your db instead, to achieve Sum, Group By etc.

It doesn't have separate methods for that from what I know u can just them inside select function and it wud work

supabase.from("cities").select(" max (population)")

U can pass the count type estimated, exact, etc. And head in the options after the string in select fn

anyone deployed supabase react project in aws. can i have your suggestion youtube tutorial

Hey all. I made a little NPM module to bridge Firebase Functions with Supabase HTTP triggers. Would love some feedback https://www.npmjs.com/package/firebase-supabase-function-bridge

(it basically makes sure there is a special API key HTTP header when calling your functions and it automatically inserts each HTTP trigger for each function)

https://github.com/burggraf/SupabaseLogQuery I rewrote this today - hope you find it useful.

hey there, is there a way to send data from the HTTP request of Function hooks? As far as I can see from the UI only url params are supported. What about sending JSON?

Is the storage policy changed? I remember I have to add select policy manually to make files public even the bucket is public. But now the files are public as default if you create a public bucket.

Hi everyone, I got an error when using upsert. This error means I have to give all fields right? But what if I just want to update one field? It doesn't support or something wrong with my code? My code is like:

js
await supabase
    .from('design_files')
    .upsert({ id: 'c9a68fc0-757a-4556-b339-659c19b551fe', name: 'Heron' })

upsert = update if already present, insert if not present. There are 3 columns: `id, name, owner`; all 3 mandatory. Case 1: Entry is already present. It updates the row with the new

name

, leaving

owner

unchanged. Case 2: Entry is not present. It tries to insert a new row with that

name

and `ID`; but

owner

is a required field, so insert will fail.

So you need to specify

owner

for

upsert

to work, or else

insert

(case2) will fail. If you're okay with it failing, just use

update

insert of

upsert

.

like i asked before how do i implement rate limiting ?

do i have to create a new table for that and integrate it into rls maybe?

Postgres rate limiting

Thanks @User , I'll use update and insert.

Greetings everyone. I am having a self hosting issue. I was able to get the project running based on the docker configuration with the default settings. I then changed the JWT secret and the anon and service keys and now I am getting an 'Invalid authentication credentials' error. I changed the keys in dockerfiles/kong/kong.yml and in the environment section of docker-compose.yml. I noticed the self hosting guide didn't mention changing that but I figured the keys needed to be the same. Is there something that I am missing?

Make sure you've also updated the following env vars for the containers: -

supabase-rest

container:

PGRST_JWT_SECRET

-

supabase-realtime

container:

JWT_SECRET

-

supabase-auth

container:

GOTRUE_JWT_SECRET

-

supabase-storage

container:

PGRST_JWT_SECRET

Editing them, and then restarting/redeploying the individual containers should fix the issue

@User The JWT_SECRET is read from the .env file for those containers so they all should be using the updated value.

Hi Juuun your issue is simple... You have null fields and in your table you check the NOT NULL CHECKBOX... Edit your table settings 😉

That's breaking the data model!

Yes... he need check if the fields can be null as well 🙂

Could this be a caching issue? I shutdown down the containers, deleted the folder and did a fresh clone of the repository without changing anything and I am getting the same error.

Is anyone using Supabase with Next.js? Do you prefer to implement auth on the server side or the client side (and why)? I've read tutorials doing it both ways but I don't really see any advantage to the server-side method, just increased latency + more limited deployment options. I'd love to hear pros/cons of doing it this way.

We use it with Next.js and use a combination of both server and client side. Somethings we use the service role key for inserting data and that stays strictly on the server side where users can’t grab the key. We use the supabase client on the client side with the anon key + appropriate row level security policies in place

Thanks, that's helpful 😊

Also, I've been searching about whether it's possible to impose rate limiting with Supabase and noticed a few people wondering the same but I couldn't find any conclusive answers. Would it be possible to impose a constraint on login attempts without having to involve a backend?

This is awesome >>> I wanted to use micro-functions/FaaS, but supabase functions are alpha, and I need stable. So I figured, Ah ha! I'll use Cloudflare workers. And so I get an account and look for a javascript example. And I see "Build data-driven applications with Workers and PostgreSQL", so I get excited. That looks good. I can probably adapt that to Supabase. Well, no..... Because... . . The example uses Supabase https://developers.cloudflare.com/workers/tutorials/postgres

I am stuck though on one thing. I cannot determine the URL I am supposed to use to connect to postgrest

I am starting to think the example is outdated.