| 23 Mar 2025 |
 @alex:acrawford.com | What are people using instead? | 16:28:07 |
 emily | deploy-rs and colmena are the big ones I think | 16:28:18 |
| emily | but they only do a subset of the stuff. for the rest there's Terraform, Pulumi, etc. | 16:28:29 |
 flare | I made a semi organized mess of make files | 16:28:34 |
| flare | a framework to do all secrets and config deployment | 16:29:14 |
 @dmjio:matrix.org | We need nixops back man, it's all we have and its the best | 16:34:13 |
 K900 | It's far from all we have | 16:34:21 |
| K900 | And it's far from the best | 16:34:24 |
| K900 | (the reason it's dead is because no one wants to be within blast radius of that code base) | 16:34:53 |
| @dmjio:matrix.org | as emily mentioned in #haskell:nixos.org none of the others I've seen use both nix and have provider support | 16:35:11 |
| K900 | Just use Terraform | 16:35:29 |
| @dmjio:matrix.org | bro | 16:35:34 |
| emily | you could try to help with nixops4 | 16:35:35 |
| K900 | Terraform is significantly better at that part than Nixops ever was | 16:35:45 |
| K900 | And Terraform isn't even that good at that part | 16:35:50 |
| K900 | You can use terranix if you don't want to write HCL | 16:36:06 |
| emily | you can use agenix for secrets | 16:36:43 |
| @dmjio:matrix.org | we could just rewrite nixops in haskell | 16:37:17 |
| K900 | You can go talk to @Robert Hensing (roberth) for that | 16:37:26 |
| K900 | He's already rewriting it | 16:37:29 |
| emily | https://github.com/nixops4/nixops4 | 16:37:41 |
| @dmjio:matrix.org | if its ready I'll use it, but is it at feature parity | 16:37:56 |
| emily | how can you say "we could just rewrite nixops" but also not use/contribute to something if it's not at feature parity? | 16:38:12 |
| emily | it's more at feature parity than an empty repository… | 16:38:19 |
| @dmjio:matrix.org | I don't know the rust ecosystem like I know the Haskell one, I've already used hnix and other tools like amazonka to make API calls, I know how I'd do it in that ecosystem. With rust I'd have to relearn all that. | 16:40:28 |
| @dmjio:matrix.org | my understanding is that current nixops is like a wrapper around boto or python fabric | 16:41:07 |
| @dmjio:matrix.org | Robert Hensing (roberth): hello ! đź‘‹ | 16:42:09 |
| @dmjio:matrix.org | how's the nixops4 development effort going, I'd love to use it | 16:45:15 |
| flare | I prefer to use sopsnix | 16:51:15 |
| flare | with age | 16:51:25 |
