walex VS pg_bitemporal

The WAL CDC approach: https://github.com/cpursley/walex?tab=readme-ov-file#publica...

CREATE PUBLICATION news_item FOR TABLE news WHERE (topic IS "AAPL");

What do you mean testing with processes?

I won't suggest these are the best written tests, but I test various processes, supervisors, etc like this:

https://github.com/cpursley/walex/blob/master/test/walex/sup...

You don’t send the entire WAL, just what you subscribe to - and you can even filter via SQL: https://github.com/cpursley/walex?tab=readme-ov-file#publica...

What we do is much if the business logic in Postgres. But then there’s all the other stuff like external integrations, etc.

We handled that by having an Even system built on the Postgres WAL that we use like a callback system.

I put together a little library in Elixir (that originally started out as forked Supabase realtime) for this:

https://github.com/cpursley/walex

Recently added the ability to configure WalEx to forward events to webhooks or EventRelay (so you don’t need to know Elixir).

If you're using Elixir, check out https://github.com/cpursley/walex

I've actually been thinking about turning this idea into a product where you can just point it at your postgres database and select the tables you want to listen to (with filters, like you describe). And have that forwarded to a webhook (with possibility of other protocols).

I'd love to hear folks thoughts on that (and if it would be something people would pay for).

Neat, this is the pattern I've been thinking about for a while now. Also glad to see this is Elixir based.

I've been using https://github.com/cpursley/walex (basically a fork of cainophile via a fork for subabase) to listen to Postgres changes in Elixir.

https://github.com/cpursley/walex

Which is based on WAL logs and doesnt have the same limitations.

Great writeup, just added this to the WalEx readme:

https://github.com/cpursley/walex

(WalEx is an Elixir lib for listening to the WAL and perform callback-like actions with the data)

Very cool.

I'm using Elixir to listen to change events via https://github.com/cpursley/walex (which I basically ripped off from Supabase).

I feel like i keep yelling the following, but bitemporal tables.

- https://aiven.io/blog/two-dimensional-time-with-bitemporal-d...

- https://github.com/scalegenius/pg_bitemporal

4 timestamps and some ugly queries.

People here may also be interested to see this analysis of the state of SQL:2011 "temporal table" feature adoption: https://illuminatedcomputing.com/posts/2019/08/sql2011-surve...

I don't think much has really changed since, and I'm not sure Postgres is any closer to addressing this natively (although there have been extensions, e.g. https://github.com/scalegenius/pg_bitemporal).

Something more sophisticated would be https://github.com/scalegenius/pg_bitemporal

Sure, I can appreciate that native support for a feature like this is nice.

As I understand it, most implementations (including another one for bitemporality[1]) involve either audit tables, as you mention, and/or additional support columns. It's as if the "now" representation is simply a narrow lens onto the underlying data.

That said, PostgreSQL encodes and has battle-tested decades of database functionality including an ecosystem around those, so I'd be a little wary of switching technology even if it does solve one individual problem thoroughly. Everything has to start somewhere, though.

[1] - https://github.com/scalegenius/pg_bitemporal