supabase-mailer VS plv8

Supabase developer here. Just to clarify, Supabase has server-side functions (edge functions), emails can be sent with something like my repo here (https://github.com/burggraf/supabase-mailer) or through edge functions -- but you're right, while we do have the ability to send any auth-related emails built-in (sign up, sign in with magic link, MFA, reset password, etc.) we don't have an out-of-the-box solution for sending emails, that's why I wrote that repo. And lastly, it's quite easy to schedule tasks using pg_cron, which is built into every Supabase project by default.

Hello, I've tried these instructions https://github.com/burggraf/supabase-mailer/blob/main/Mail_Providers/Sendgrid.md when I tried to execute the first line INSERT INTO private.keys (key, value) values ('SENDGRID_API_KEY', 'aaaaaaaaaa'); in SQL editor (with my API key) in Supabase I've got this error: "Failed to run sql query: relation "private.keys" does not exist"

Here's a repo where I use the http extension extensively: https://github.com/burggraf/supabase-mailer

I was excited to find the supabase-mailer project, but unfortunately it seems that only the Mailgun provider is fully supported and working. Mailgun only has a 30 day trial, which didn't work for my needs.

You can't call the http function like that. You need to use a SELECT statement. See https://github.com/burggraf/supabase-mailer/blob/main/02A_send_email_mailgun.sql for an example: SELECT content INTO retval FROM http (('POST', ....etc

I'm not 100% sure you could do this with PostgreSQL functions, but you probably can. Since we have http access from these functions, I've written interfaces to some of the transactional email apis here: https://github.com/burggraf/supabase-mailer . (It's probably even easier if you use straight SMTP mailers.)

Not sure what you mean. I have a couple open-source repos on github that show you how you can send emails or text messages from a Postgres function using third-party providers. I also have a repo that shows how you could roll your own auth using functions. Assigning permissions and roles is definitely something you can do in functions -- that's standard postgres stuff....

a bit more than an experiment at this point. pljs, even in its early state, has some very good results: https://github.com/plv8/plv8/issues/531#issuecomment-1627883...

passing through v8's javascript/c++ membrane has always been painful, and appears to be getting worse.

From here you'll write a send_email function in Postgres that calls the API. I initially wrote the entire function in PLpgSQL and spent an entire day to get it debugged and working. And even then I wasn't happy with it. In my opinion that language is unintuitive and difficult to learn and debug. I switched over to PLV8, an extension for Postgres that supports writing functions in Javascript. It takes one click in the Supabase UI to enable this extension, and it will save hours and hours of time.

You jest, but the evil geniuses at plv8 have already done it for Postgresql.

To be honest I haven't used it, but I've extensively used pl/pgsql and a little pl/ruby, and know that https://plv8.github.io/ exists - it might be what you're looking for and it's on my list of things to play with

but, let's take your straw man a little further. let's suppose that all of the actual parsing is done for you already, and all you're doing is iterating through the data structure, creating objects through the c++ api, and calling it good. that should be faster than calling the c++ JSON.parse(), shouldn't it? since we don't have to actually parse anything, right? no, it's actually much slower. you can see this in action at https://github.com/plv8/plv8/blob/r3.1/plv8_type.cc#L173-L60...

again, we're not talking about whether javascript in an interpreter is faster than c++, we're talking about whether v8's api causes enough slowdown that some workloads that require a lot of data between c++ and javascript are slower than the same workload that requires very little data between c++ and javascript ... because passing through v8's c++/javascript membrane is slow.