• Joined on 2026-05-22
lars created repository lars/gebos-landing 2026-06-23 19:15:14 +00:00
lars pushed to klaus/skeleton-infra at lars/Gebos 2026-06-01 19:26:10 +00:00
a84b279747 fix nodejs version
lars opened issue lars/Gebos#24 2026-06-01 16:04:29 +00:00
Test Issue
lars commented on pull request lars/Gebos#23 2026-05-30 00:08:21 +00:00
Infrastructure skeleton — flake, three hosts, Supabase compose, ingester, deploy-rs

@Klaus, that sounds great. Please add sops-nix to the current PR. Can we also add default secrets to run locally in development without having to create any? The same goes for other env vars like…

lars commented on pull request lars/Gebos#23 2026-05-29 23:49:46 +00:00
Infrastructure skeleton — flake, three hosts, Supabase compose, ingester, deploy-rs

@Klaus What would be a good way to store those secrets in code (encrypted of course)? I'd like to have everything in one place and preferably only manage one secret (the private key to decrypt the…

lars commented on pull request lars/Gebos#23 2026-05-29 23:45:28 +00:00
Infrastructure skeleton — flake, three hosts, Supabase compose, ingester, deploy-rs

@Klaus Quick question. What secrets will be required for deploying the full app?

lars created repository lars/klaus-bot 2026-05-28 19:46:05 +00:00
lars commented on issue lars/Gebos#21 2026-05-27 17:17:31 +00:00
Project Setup

@Klaus For the studio: "Don't expose it publicly — SSH tunnel when you need it.". Yes please create a PR

lars commented on issue lars/Gebos#21 2026-05-27 17:15:16 +00:00
Project Setup

@Klaus I think the choice of using RPC and ditching Huma is good. PostgREST + /rpc/ is the public surface. Yes, then let's go with the setup of Caddy and Kong chained. In order to get the best of…

lars commented on issue lars/Gebos#21 2026-05-27 17:08:09 +00:00
Project Setup

Where does the Supabase stack actually run? My choice is (b) Run the compose stack but point its services at external db-host Postgres.

**PostgREST vs. Huma — pick one public REST…

lars commented on issue lars/Gebos#21 2026-05-27 17:01:55 +00:00
Project Setup

@Klaus These are good points:

  1. I agree on the single-node HiveMQ: "Accept single-node with fast restart + persistent session storage on disk; devices reconnect and replay. Probably fine for…
lars commented on issue lars/Gebos#21 2026-05-27 16:47:35 +00:00
Project Setup

@Klaus Please continue our discussion about the system architecture above

lars commented on issue lars/Gebos#21 2026-05-27 16:46:44 +00:00
Project Setup

@Klaus Can you hear me?

lars commented on issue lars/Gebos#21 2026-05-27 15:36:57 +00:00
Project Setup

Hello?

lars commented on issue lars/Gebos#21 2026-05-27 15:23:05 +00:00
Project Setup

Ok thank you. I've made a more detailed plan about this. We need multiple different machines (microservces) in order to separate failure modes.

Architecture

lars commented on issue lars/Gebos#21 2026-05-25 18:24:14 +00:00
Project Setup

@Klaus for now let's ideate on this together in this issue. Once I approve, you can start with the implementation.

lars created branch main in lars/Gebos 2026-05-22 23:19:11 +00:00
lars pushed to main at lars/Gebos 2026-05-22 23:19:11 +00:00
71d74d3b83 first commit
lars created repository lars/Gebos 2026-05-22 23:06:08 +00:00