pnoel

Must be some dirty ETL pipelines :')

Yes it does :)

That's a nice diagram! Yeah that's roughly it. We'll be adding support for more sources of truth in the future to expand coverage, like the ones you mention but also NoSQL like MongoDB

Good question -- the primary difference is the method of integration with Postgres. ZomboDB is a Postgres extension, which limits their compatibility with Postgres serivces like AWS RDS, while Retake is compatible with any service where you can enable logical replication

* If the “real-time server” goes offline temporarily, will it catch up on any newly added rows in the interim?

Yes, we're built on top of Kafka

* Do you intend to emit any OpenTelemetry metrics? I’d like to monitor lag in production.

We don't have that yet, but open to it

* Will I be able to deploy this as a single container on ECS/Kubernetes?

Yes

In December 2022, we shut down Whist after ~3 years. For those who didn't know us, we were a Mighty competitor, building a cloud-streamed browser.

I spent a long time reflecting on what happened, our successes and mistakes, and compiled them in a post-mortem. A few people close to me read it and some found it insightful, so I figured I'd share it with the wider community.

A story on hard-learned lessons that feel obvious in hindsight. Here's to hoping my mistakes can save you some!

Whether you have 1 cloud tab or 100 cloud tabs, the resource utilization on your machine will be the same. If you open more local tabs, though, those will take up RAM/CPU on your machine.

Charging a subscription fee is the primary way for which we would pay for cloud compute time -- but we're thinking there will be a version of Whist which will have very limited amount of compute time and be free (a sort of free trial if you will)

Noted, we'll keep this in mind as we go forward then

You absolutely can!