Has anyone deployed a nextjs project that uses the supabase javascript sdk v2 on google cloud run , i wanted to know how did you set your environment variables without committing a env file or hardcoding keys in the dockerfile ?

Once in a while I encounter a situation where it would be fantastic to have the ability to determine dynamically not only whether given rows can be accessed, but also which columns of those rows may be accessed. Coming from a "vanilla" Postgres background, a

grant

is often the tool I would turn to, but this is complicated by not quite using the built-in Postgres role system in the usual way. In addition, I'd be interested in being able to determine the columns dynamically based on other properties of the user/context besides the user role. What I'm wondering is if there is something similar to the

using

clause in `create policy`—where I can use whatever functions etc. that I wish—that can be used to determine which columns should be accessible. Even if this was table-wide, like

grant select (...)

, it would be a start, but of course it would also be nice to be able to determine it dynamically based on the content of the row, if necessary.

So your project gets paused after 1 week of inactivity, whats stopping someone from setting up a cron job to avoid this?

Two questions on this? 1. The foreign key must be a uuid? The following will not work unless I add uuid: post_id uuid references posts (id), 2. The posts table id must also be a uuid? An error is also returned: "foreign key constraint "comments_post_id_fkey" cannot be implemented" A post on stack overflow mentions this is because the table you are pointing too must also have an id of uuid. Why is this? It's my understanding that uuid's are difficult to search, so often it is better to use SERIAL/INTEGER. Thanks much

Is it possible to change the db shema from the cli ? i uderstand this could be risky for production workload. As such i wanted to see if there is a way i can modify my schema for devlopment purposes without the risk of affecting the production schema. And also be able to migrate the production schema to the new schema ?

I'm trying to redirect the user to stripe connect to create their account, but I keep having this issue and I don't know how to resolve it. This is the response that I am sending from the edge function:

return new Response(null, {
      status: 302,
      headers: {...corsHeaders, location: accountLink.url, "Content-Type": "application/json"},
    });

I took

corsHeaders

from supabase edge function docs. And this is the error that I have on my client:

Access to fetch at 'https://connect.stripe.com/XXXX' (redirected from 'https://XXX.functions.supabase.co/stripe_onboarding') from origin 'http://localhost:3000' has been blocked by CORS policy: Response to preflight request doesn't pass access control check: No 'Access-Control-Allow-Origin' header is present on the requested resource. If an opaque response serves your needs, set the request's mode to 'no-cors' to fetch the resource with CORS disabled.

So I'm trying to make a storage bucket so that when new users sign up, it will create folder with their user ID and only allow them access to that bucket. Luckily, there is a template that will do just that. But I cant figure out how I can actually generate the folder when a user signs up. They wont have any access to the bucket with RLS, so I cant just generate one when they sign up. Is there a way I can do it on supabase itself thatll trigger a folder to be made when it detects a new user signed up? Thanks in advance.

Hi there, I was wondering if Supabase would be a suitable tech for my application. I will need to make a request to the database every second (sometimes even more) as I will be storing data from the stock exchange there and be calling it too. Do you think Supabase would be a viable option for this

I think Im doing something wrong already😀 How can I just that field automatically, a user will add goal by clicking button and table should add uuid of user, who clicked button

Hi, is there a setting in Supabase where I can force the user for prompt every time they login with Google? This is so they can choose to login with different Google account. The issue am having is that even when I sign out of Google, everytime I log back in I automatically get signed in without prompt. The only way I can login with a different Google account is if I clear my history and cookies. It might be a Google Login specific question, but I thought I'd ask if anyone here has any advice

We are unable to connect to our project and when we click 'still connecting'. It is asking us to check the health of our database. When we click database health, it doesn't show any info about the health. Only details about the DB for the API. We created this project way back when supabase was still in Alpha/beta I think. I'm not sure if we have reached the limit or anything but we can't determine it with the error that's being provided. The data stored are not that big and we only have few users, little traffic, so this shouldn't be an issue with the DB health. Hope anyone from the Supabase team can help us. Thanks 🙂

I have a messages table with RLS and RealTime enabled. I have RLS policies like SELECT, INSERT for all

authenticated

users. Inserts work fine and I can select records from the table, too. I have a realtime channel set as follows, but it doesn't receive any updates. Looking to the browser network logs, the websocket seems to be connected as I see some heartbeat messages going back and forth. I have

@supabase/auth-helpers-nextjs@0.5.4

and

@supabase/supabase-js@2.8.0"

.

If I want to join the author record on a posts record, I can just add it like so:

ts
supabase.from('posts')
.select('*, author!inner(*)')

But how do I do this in pure SQL so that the results are inner json like Supabase does? I'm thinking

sql
select * from posts
join profiles on profiles.id = posts.author

But this won't make the results nested json like I want or give my desired results: There seem to be lots of confusing examples with

json_agg()

,

row_to_json()

,

array_agg()

and

json_build_object()

, none of which are simple or clear. What's the simple translation for this? Thanks, J

here is the log

Hello, I am new to Supabase. I would like to know if there is any way to check registered email is verified or not through API. I googled for awhile but I cannot find related posts. I might be missing something. Do I just need to simply make database for checking the status of email verification? ### When I call supabase.auth.signInWithPassword, I will get a following response which telling email is not verified. I would like to have better way to check status of email verifiation.

const res = await supabase.auth.signInWithPassword({
      email: data.email,
      password: data.password,
    });

    console.log(res);

{
  "data": {
    "user": null,
    "session": null
  },
  "error": {
    "name": "AuthApiError",
    "message": "Email not confirmed",
    "status": 400
  }
}

After running

supabase db remote commit

I get the following error. Is that an issue with my linked project?

Error: ERROR: relation "supabase_migrations.schema_migrations" does not exist (SQLSTATE 42P01)

I'm looking into migrating from firebase and I have a few questions about realtime database change listeners. With firestore I can query for and then listen for updates on any arbitrary query. Is that possible with Supabase? The docs seem pretty sparse about listening to row level changes. They only mention the following about filtering: > You can listen to individual rows using the format {table}:{col}=eq.{val} - where {col} is the column name, and {val} is the value which you want to match. Does that mean that the only possible filter you can apply to an update listener is

eq

? If so am I correct in thinking that Supabase's realtime change listener capability is significantly more limited than Firestore's? It seems to me from reading the docs that Supabase has the following limitations: * You must use a different API for querying data and for connecting to and filtering realtime updates * You cannot fetch and begin listening in a single operation, you have to fetch data then set up a listener as a separate step * The only operation available for filtering change listeners is

eq

* There are significant performance limitations with listening to document updates (the docs only say

By default, listening to database changes is disabled for new projects due to database performance and security concerns

) Are those accurate? If not where can I find more info? I'd like to figure out if Supabase can fill the requirements to migrate my project from Firebase. Thanks!

If you have a table with user_id -> auth.users.id that semantically represents the owner of a record, an insert policy like the following opens the risk of any user creating any record owned by some other user. That is, if I have the following policy:

sql
CREATE POLICY "policy_name" ON public.messages
   FOR INSERT TO authenticated WITH CHECK (true);

Any authenticated user can do the following:

javascript
await supabase.from("messages").insert({ user_id: some_other_users_id, ...other_fields });

Or am I missing something about Supabase REST API? And is the following a way to shield against that?

sql
CREATE POLICY "policy_name" ON public.messages
   FOR INSERT TO authenticated
  WITH CHECK ((auth.uid() = user_id));

Is there a way to get the status shown here: https://supabase.com/docs/reference/javascript/insert I basically want to know if my insert is successfully and ideally through the response message.

const { error } = await supabase
  .from('countries')
  .insert({ id: 1, name: 'Denmark' })

From the docs, it shows the response message below, but is it accessible anywhere?

{
  "status": 201,
  "statusText": "Created"
}

I am only using supabase for a project and would like to know when to use RLS and API, cause rls interacts directly with postgres and api go through gotrue so they are slower but what if I want to only use api should I even enable rls? I couldn't find anything related to this so my concepts might be wrong please correct me.

Am trying to implement a role based permissions in my app (Nextjs 13) and it seems difficult any suggestions??

Any way to voluntarily authenticate using a JWT from within a native Postgres connection? I guess yes, but haven't seen it in the docs.

I am trying to add Supabase as a one click app to Caprover. Here's the pull request: https://github.com/caprover/one-click-apps/pull/833 Everything deploys correctly but the issue I am facing is when I open dashboard it is unable to connect with the postgrest API and results in 404. Though all individual apps are working fine and can be accessed from their individual urls without any 404.

Calling Supabase's

signOut()

method does not clear the session or current user on Flutter. Both appear to retain their state. An auth changed event for log out is fired, but those variables remain not cleared.

Hello, I was just curious if it was possible to disable some containers during local development to reduce resource usage? For example, I never use

imgproxy

and

storage

, so turning them off saves some resources. I didn't see any post on Discord or Github.

I have a really strange situation. I have a table that I cannot even do a simple insert on. It accepts data when using SQL, but when using the supabase client it fails. The error message (as referred to in the subject) really seems off. And like I said, the following simple SQL INSERT query does work:

sql
INSERT INTO section 
(poll_id, slug_id, "label", description) 
VALUES(4, 'welcome', 'Welcome', 'Test');

For the record, the table is (was) empty. RLS is enabled, with one relevant policy:

sql
(EXISTS ( SELECT 1
   FROM poll
  WHERE ((poll.id = section.poll_id) AND (poll.project_id = ( SELECT project.id
           FROM project project
          WHERE (project.owner_id = auth.uid()))))))

A

project

has an

owner_id

. A

project

can have many

poll

records via

poll.project_id

. A

poll

can have many

section

rows. The relevant DDL is attached to the post. This is the client call I am making:

ts
  async create({
    poll_id,
    slug_id,
    label,
    description,
  }: {
    poll_id: number
    slug_id: string
    label: string
    description?: string
  }) {
    const { data: sectionIds, error } = await supabase
      .from('section')
      .insert([{ poll_id, slug_id, label, description }])
      .select('id')
    return { id: sectionIds?.[0]?.id, error } as { id: Number; error: PostgrestError }
  },

Digging deeper, I find this error-entry in the Supabase "postgres" logs: ERROR: more than one row returned by a subquery used as an expression The content json is really interesting, but I am not able to figure it out. Please see the attached file. Any assistance or tips on working around it would be appreciated.

Ive been using react supabase for a long time now and its worked well until now. This line has always worked and returned correct data and it still does.

js
  const { data, error } = await supabase
    .from("stories")
    .select()
    .order("index");

This I'm trying to do the same thing for the

user_settings

table and it's returning an empty array in data

[]

js
  const { data, error } = await supabase.from("user_settings").select();

i don't know what to use, i have to store user data for my company's internal application. the user wont be able to update or delete it themselfes. including name, language preference, title, department, telephone and fax extension. whats the correct policy?

Hello! Anyone using supabase only for auth + storage with a different database? I am curious how you guys create the user in your own db when he logs in ( when using supabase also for db, you add a trigger, But in this case i cannot do it)

Hello. I updated the supabase-js and auth-helpers to the latest versions and my typescript linter started triggering. It does not give anymore the inferred types. example here my type that i intend to return from the function and the second image shows host: [x:string].