Hey ,can we create a policy for a table in which users in that table can only access it.

Hi there, I'm facing this issue here: https://github.com/supabase/supabase/issues/4883 And ran the SQL suggested as a fix in the comment thread - but now hitting another error of: "URL using bad/illegal format or missing URL" In the postgres logs - could someone please help investigate? 🙂 I had just unpaused my project today after a few weeks

Is there any issues with ordering by foreign table column via JS SDK?

js
const { data, error } = await client
  .from('otherdeeds')
  .select('id,prices!inner(price)')
  .not('prices.price', 'is', null)
  .order('price', { nullsFirst: false, foreignTable: 'prices' })
  .limit(10)

Always returns unordered list.

Is there any chance to describe models and authorization rules (RLS) as code instead of using the supabase UI?

Hi Folks..I am facing issue with supabase realtime.I am working on my project after 6 months but when i left the project everything was working perfectly fine with current configurations but now realtime has stop working unexpectedly.I have all the current configuration in my supabase dashboard as suggested by latest supabase docs and tried different suggested ways but still not able to make it work

Hello is there way to call the https://supabase.com/docs/reference/javascript/auth-update to update a user's email from serverside? without having the jwt of the user?

I just recently starting getting the error

canceling statement due to statement timeout

when trying to truncate or drop a certain table. Weirdly it is showing that there are 0 rows but still 11mb of data in it. What could be causing this to happen? (I tried creating a new table and deleting it and it works fine so it seems to be a problem with this specific table)

i have a couple of tables, all of which have something along the likes of

id: uuid

and

created_at: timestamp

, and the following additions: - the table

urls

has a

host_and_path: string

column containing a normalized url - the table

website_data

has (among other columns) a

url_id: uuid

column pointing to a

urls

entry both were created through the ui, no special sql wizardry going on here when creating a

website_data

entry from my input data (which includes a url that i normalize) i'd like to create a

urls

entry if one doesn't exist, then take its id and use it as the newly created

website_data

entry's

url_id

, all in one go, and possibly for many entries in one call. is that possible today or do i need to resort to an ugly and inefficient multi-step solution (upsert url returning id and normalized address, iterate through input data to match url ids to data objects, create

website_data

entry for each input datum using the info i need and the correct

url_id

)?

Is it possible to install/use ffmpeg in an supabase function?

Hi I am using graphql to interface with my Supabase database. I've added RLS to my table and attached individual policies for SELECT, INSERT, UPDATE, and DELETE (attached image). I am able to SELECT without any issues, however my INSERT fails with:

new row violates row-level security policy for table \"projects\"

I've been stuck for a while now and not sure what to do. This is my mutation query:

mutation MyMutation {
  insertIntoprojectsCollection(objects: {assets: ["test01"], name: "test", scene: "testscene", sensors: [{name: "camera1"}], user_id: "ca3ca3b8-9146-48e3-9106-1d8e42f8f5b5"}) {
    affectedCount
  }
}

Hey guys! I'm trying to post some form data to Supabase in an async function.

const makeRequest = async (formData, userId) => {
  const supabase = getSupabase(userId)
  const { data: users, error } = await supabase
    .from("users")
    .update({ 
      user_type: formData.userType,
      first_name: formData.firstName,
      last_name: formData.lastName,
      is_onboarding: false 
    })
    .eq('user_id', userId)

    error ? console.log("Supabase error:", error) : console.log(data);
  }

But I'm getting this error: secretOrPrivateKey must have a value utils/supabase.js (18:20) @ supabase.auth.session

16 | 
  17 |     supabase.auth.session = () => ({
> 18 |       access_token: jwt.sign(payload,  process.env.SUPABASE_SIGNING_SECRET),
     |                    ^
  19 |     })
  20 |   }

Here's where I'm signing the JWT:

// utils/supabase.js
import { createClient } from '@supabase/supabase-js'
import jwt from 'jsonwebtoken'

const getSupabase = (userId) => {
  const supabase = createClient(
    process.env.NEXT_PUBLIC_SUPABASE_URL,
    process.env.NEXT_PUBLIC_SUPABASE_KEY
  )

  if (userId) {
    const payload = {
      userId,
      exp: Math.floor(Date.now() / 1000) + 60 * 60,
    }

    supabase.auth.session = () => ({
      access_token: jwt.sign(payload,  process.env.SUPABASE_SIGNING_SECRET),
    })
  }

  return supabase
}

I have my SUPABASE_SIGNING_SECRET in my .env.local file in the root directory. I'm signed in and retrieving a user id from the session, but when I submit the form it throws this error. Anyone able to point me in the right direction?

Hi all, with a supabase function hook, how would you send record data with the httprequest? eg:

{
  "record": {
    "email": "asd@gmail.com",
    "id": "16c34376-54ff-49dc-a374-f32b063ef5da"
  }
}

Hey guys i setup custom smtp with supabase and sendgrid folowing the instructions exactly as they said but i still get error sending magic link on frontend when i send the login email

Hi all, i'm trying to join tables using the REST api (postgrest resource embedding). Goal: get the projects.name and something.externalId in a single request. Tried: /rest/v1/projects?id='0805c7c6-ebdf-4d51-84fe-feaa50a17b05'&select=\*,something(externalId). Response: PGRST200 - could not find a relationship between projects and something in the schema cache. Question: how to join the table projects & something when the table workspaces is __not__ a join table. Thanks! any input is very much appreciated 🙂

Mjml doesnt work in supabase template

Hi All, I'm unable to drop a column though I removed the foreign key constrain on the column using the UI and made the cells NULL. What is going wrong? thanks!!!!

Is it possible to remove the extra level of depth on select request? For instance, I have this select request:

js
            .select(`
                organization(id, name)
            `)

Which gives me the following result: [   { organization: { id: dqzdqz, name: "dqdqdq" } }, { organization: { id: dqdqdqdqzd, name: "dqdqdq" } } ] And I would like to have: [ { id: dqdqdq, name: "dqdqd" }, ... ]

Row level security best practice for join tables -- curious if anyone has advice / experience setting RLS on a join table? I have a many-to-many relationship between artists and tags. Users can assign many tags to different artists. Is it necessary to have a user_id column on the join table?

This is a question for the team, particularly the functions crew. I see that edge functions are using the deno deploy platform; based on that I tried to deploy the fresh framework as an example and hit a "Only POST or OPTIONS requests allowed". Is this an artificial requirement or a technical limitation?

I am trying to work the Edge Functions locally but I can't seem to connect directly to the database using https://deno.land/x/postgresjs@v3.2.4. I keep getting

Connection refused

. Do I need to allow Deno to access the locally running DB? If so how can I do that? I tried appending the

--allow-net=localhost:54322

flag via

supabase functions serve

it didn't like that.

any ideas why

let {data, error} = await supabase.from('food').select('*');

might just "fail" with a

Require cycles are allowed, but can result in uninitialized values. Consider refactoring to remove the need for a cycle.

warning? Inspecting the network shows it's not even making a call to the api. I have calls to superbase.auth.signUp working fime...

js
  let {user, error} = await supabase.auth.signUp({
    email: email,
    password: password,
  });

i'm not quite sure what i'm doing wrong here...

Are there any examples anywhere showing how to connect directly to the Postgres DB in an Edge Function? I am trying to connect but I keep getting

502 Bad Gateway

. I am using the provided

SUPABASE_DB_URL

environment variable when connecting to the DB.

Is there a way to receive the user metadata in the auth response?

Hello gang, new to supabase and loving it. I wondered if there was any way to get a swagger.json or openapi.json spec of the postgrest api.

noob question for you all: just started using supabase and i have created some tables via Prisma and i can see them in the webUI. If i want to write a sql query against those tables, how do i do that? I go to the sql editor but it doesn't let me select either of the tables i created.

I get

Error: unknown flag: --no-verify-jwt

when I call

supabase functions deploy my-function --no-verify-jwt

. Am I misunderstanding something?

Realtime suddently doesn't work

I'm not quite sure of exactly what I want, but is there a callback to run something when authentication happens or a token is refreshed? I'd like to keep the user object in react-query, which works by wrapping up

supabase.auth.signIn()

into a react-query custom hook, but of course that doesn't stay in sync when supabase refreshes the token and etc..

Bug? When editing a row, the timestamp defaults to 12:30PM of today, which throws an error. Moving the date to sometime in the future makes it work, but obviously the created_at time shouldn't change on an edit. Not sure how it's possible to not touch it at all? This is just using the standard timestamp column as suggested by default.

I'm using Supabase DB with Prisma in an Express REST API project. My dataset is in CSV. My dataset is extremely large. So I have a smaller dataset that I want to seed into the database. How can I seed a csv into the database with Prisma? Second question is my dataset is large. Two csv. One with 500 rows and another with 3 million 100k rows. Is it too big for free tier? The csv size is 300mb and 36kb.