I need to use the http extension but I have no way to hide my API keys

Hey please i keep getting this error trying to save to the database "new row violates row-level security policy for table "cartItem", this is my code, what do you suggest i do

Hi! My supabase account has my github email linked to it, but I will not use that email anymore. I changed my github email but I can't find a way to change my email in supabase. Thanks!

Not sure how to investigate. I am just on the free tier atm so perhaps it is this but I feel like most of tables only have a couple of rows each in them as I am still in development atm and just testing. Any pointers?

drop function if exists get_member_log;
create function get_member_log(membertag text)
  returns table (
    playername text, 
    team text[], 
    opponent text[], 
    t int,
    map text, 
    ticket int, 
    trophychange int
    )
language plpgsql
as $$
begin
   return query
   select (
      club_members.playername, 
      club_league.team, 
      club_league.opponent, 
      club_league.time,
      club_league.map, 
      club_league.ticket, 
      club_league.trophychange
    )
    from club_league
    inner join club_members on club_league.playertag = club_members.playertag
    where club_league.playertag = membertag;
end;
$$

I have an rpc like the above but when I try to call it using supabase py wrapper it says 'Returned type record does not match expected type text in column 1.' Why does it return a record type? and how can I fix this?

Please read the documentation as this is documented.

How do I format the URL for self-hosted realtime? I tried something like 'ws://realtime-dev.192.168.1.157:4000/socket'; but the client cant use this address

I'm a newbie working on a Sveltekit + supabase project. I make heavy use of edge functions and I found myself rewriting a lot of the same utility functions for both. I'll appreciate any guides/examples of how to set up a mono repo like this: Packages: - webapp (pnpm, sveltekit) - supabase (/functions - deno) - utility (can be imported both by sveltekit and deno)

Hello, I just recently created some constraints for some of my columns, However I couldn't find anything displayed in the dashboard that shows it. How can I find all active constraints on my columns?

I've set my database up carefully with cascading tables so that when a profile is deleted all traces of their activity are gone - however I am left with an empty auth.user column, since I do not seem to be able to set up a cascading relationship there. Is there any way for me to delete the auth.user column when the profile is deleted?

Hello, I have the user UID and I want to fetch the user to get the email. Is that possible? Thank you.

The GUI shows that I've set 1 option for a table column, which is the

unique

option, but the option is missing from the dropdown. Is this a known/recent issue?

I am using Angular. The login was already working and I wrote CSS and now Google Auth doesn't work from one minute to the other. I get error messages from the Supbase API

error=server_error&error_description=Unable%20to%20exchange%20external%20code:.....

I have already checked the Google Creds, they have not changed... Can it be that there is a problem in the backend, because I am briefly logged in and redirected, then the session is immediately destroyed and I am redirected back to the login page.

Is there a way to make sure that supabase rejects every request to manage the Auth/DB, unless it comes from a browser and from a specific domain? Some sort of whitelist?

I am importing a former postgres csv data set csv into supabase. It has a simple autoincrement key I have tried multiple options but ultimately I consistently get this problem:

Failed to run sql query: relation "community2_id_seq" does not exist

I set id as the primarykey and identity. I am sure it is properly sequenced but it won't let me import it directly. So i just remove the identity key setting on creation, and then I just set it to identity after the table is created from the gui.. However, once I do that, the autoincrement key is set at 1, I want to reset it to the length of the import, but when I go to do that in the SQL editork, the SQL is failing to allow me to alter the sequence and says the relationship does not exist * select pg_get_serial_sequence('Community', 'id') * I try runnning this command in sql editor but then i get

Failed to run sql query: relation "community" does not exist

Could anyone please help!

I'm trying to link my local database to my remote one, and I'm getting this error Remote database Postgres version 15 is incompatible with db.major_version 14. If you are setting up a fresh Supabase CLI project, try changing db.major_version in supabase/config.toml to 15. when I change the major version in the config, I get the error: Error: Failed reading config: Invalid db.major_version: 15. any thoughts?

I want to use the twitter api and supabase sign in says provider token is null?

how to copy the one record from one table to another table using Supabase service. As I want to copy the data from one to another.

I no longer want to ask Twitter users for their email address, so I turned it off in the App settings. But now Supabase tells me "500: Error getting user email from external provider", unable to find email with Twitter provider. Why is it still trying to get their email address? Is an email address necessary if Twitter Auth is the only type of Auth I want for my web app?

When I use createServerSupabaseClient in

auth-helpers-nextjs

I get the following error

[TypeError: context.res.getHeader is not a function]

It's coming from the setCookie method and causes my app to throw an unhandled error. I'm currently using Next 13 and auth-helpers-nextjs 0.5.2

I have a 2 tables related with each other. Table A has 2 columns ( start_date , end_date ) that i want to filter when querying table B. so when i do

supabase.from<B>(Table B).select("*, b(*) ") i get all the information.

but when i get do

const date = new Date().toISOString();
supabase.from<B>(Table B).select("*, b(*) ").gte("A.start_date", date);

i only get table B information even though i have the b_id present on the table. I also tried filter but i did not manage to get any results. I will go on the view approach. question is if i'm doing anything wrong or it is a limitation.

This may be a more generic question than supabase specific but I'd like to make my sign in emails prettier. Any pointers on good email template makers / guides. I'll do my own digging in meantime

When running commands like

supabase db push

or

supabase db dump

, it takes around 4 to 5 minutes to ask for my database password. Once I enter it, the command finishes executing quickly. Essentially, anything that connects to my hosted Supabase instance via CLI is extremely slow. I tried restarting my hosted instance, but that did not fix the issue.

The following code hangs in the supabase console:

select * from net.http_collect_response(
  (SELECT net.http_get('http://httpbin.org/headers', '{}'::jsonb, '{"Content-Type": "application/json"}'::jsonb)),
  async := false
);

I'm struggling to figure out how to test the

pg_net

if I can't even run this.

Pretty self-explanatory. I'm interested in the

pg_vector

and

pg_repack

extensions which were just merged into the build in https://github.com/supabase/postgres. Right now it looks like the AMI build is broken due to some internal problems, but once it's fixed, how will I actually get my existing instance running that AMI?

Ok I just updated from V1 to V2 of the SDK . The auth.API.setToken function is now gone and the best thing I found to replace it was the createServerClient helper for remix. Before the update I had a single instance of supabase and would retrieve a token from my session cookie set the token into the supabase client. I'm essentially doing the same thing but now I am also instantiating a new serverClient on each request. It feels like it smells bad. Is there a better way? Is this dialing a psql connection on each request? Is there an alternative to the setToken that I missed?

Hey crew! I have this bit of SQL that works fine in the SQL Editor, but I am not sure how to write it properly with the JS tool.

sql
select id,username from profiles where id = (select profile_two from friends where profile_one = 'user-uuid-string')

Tried this, but does not work.

js
await supabaseClient
  .from("profiles")
  .select("username, id")
  .eq("friends.profile_one", "user-uuid-string");

> Cannot apply filter because 'friends' is not an embedded resource in this request Have tried this with no success

js
await supabaseClient
  .from("profiles")
  .select(`username, id, friends(profile_two)`)
  .eq("friends.profile_one", user.id);

> Could not embed because more than one relationship was found for 'profiles' and 'friends' the

friends

table is just

uuid (pk), profile_one (fk), profile_two (fk)

the

profiles

table is

uuid (pk), username

Trying to capture where the current user is in

profile_one

column and return profile username and the

id

and

profile_two

would be the same, so that part doesn't matter. Just trying to gather the other profile's username. Thanks for any help

Hi, I'm trying to get the nextjs boilerplate working with Supabase following the guide here https://supabase.com/docs/guides/getting-started/quickstarts/nextjs however I'm getting the following error (screenshot of error & code below) .. seems like something pretty basic

In my app there are organizations that can create tasks. Instead of having a serial task ID, I would like to have an incrementing task ID for each organization ID. Just like in GitHub, where repo + issue number leads to the issue. So I define tasks like:

sql
create table "Tasks" (
  "organizationId" int references "Organizations"(id),
  id int check (id > 0),
  description text not null,
  
  primary key ("organizationId", id)
);

But Users should not "know" the ID before creating it, so I add a trigger:

sql
create function increment_task_id() returns trigger language
plpgsql security definer as $$
declare
  next_id integer := 1;
begin
  select coalesce(max("id") + 1, 1)
    from "Tasks"
  where "organizationId" = new."organizationId"
  into next_id;
  new.id = next_id;
return new;
end;$$;

create trigger task_insert
before insert on "Tasks"
for each row execute function increment_task_id();

This works fine until I run multiple insertions in parallel. The triggers seem to be executed in parallel, leading to several tasks having the same ID. Inserts might fail when two or more people happen to create a task at the same time. How can I avoid this? Is there a better way?

Hello, implementing RLS to ensure users (org members) can only affect table rows that belong to their company (their inventory, planning,..) can be done in at least 2 ways: 1. every table that needs RLS gets an extra column companyId and a policy attached that checks if the user jwt claim companyId matches the correcponding row's companyId 2. using extra join queries to retreive the corresponding companyId for every row + policy that checks other 3. other What would you recommend/ pros and cons assuming that the company id of a table row never changes (so normalization is less of a concern, performance gain of 1 wins?)