| 22 Jun 2021 |
 tomberek | Ideally, yes. It was on HN yesterday for a while. I’d suggest one aspect of the marketing is to describe why this is a good thing and what this allows. The Nix/NixOS marketing team meets tomorrow, we can bring it up to have a coordinated thing. | 20:05:47 |
|  ollijh joined the room. | 20:12:40 |
| 23 Jun 2021 |
 siraben | tomberek: is there an invite link for the marketing meeting? | 03:21:32 |
|  @anubhavkini:matrix.org joined the room. | 06:52:49 |
 raboof | we should definitely make sure it makes it to the reproduce-builds monthly newsletter, I'll write something up unless someone beats me to it | 07:40:03 |
| raboof | I'd really like to see https://github.com/NixOS/nixpkgs/issues/125380 fixed before making more noise, but I guess the cat is out of the bag :D | 07:40:27 |
| tomberek | @siraben: https://nixos.org/community/teams/marketing.html | 13:33:38 |
| 24 Jun 2021 |
|  tadfisher joined the room. | 03:33:06 |
|  @timdeh:matrix.org joined the room. | 17:04:11 |
| 26 Jun 2021 |
|  @grahamc:nixos.orgchanged room power levels. | 01:31:35 |
| 27 Jun 2021 |
 davidak | CW: RANT
the post was on hackernews and got 23.5k views. most probably hear about nixos and reproducible builds for the first time. the post didn't explain what either is about. so what people will remember might be "obscure project does obscure things and i don't see why i should care". this is a missed chance for us to make nixos more popular. i would say it's a marketing disaster that damages nixos reputation. why are we unable to coordinate and do something correctly? this makes me angry, because nixos has so much potential :((( | 01:14:44 |
 @matthewcroughan:defenestrate.it | In reply to @davidak:matrix.org
CW: RANT
the post was on hackernews and got 23.5k views. most probably hear about nixos and reproducible builds for the first time. the post didn't explain what either is about. so what people will remember might be "obscure project does obscure things and i don't see why i should care". this is a missed chance for us to make nixos more popular. i would say it's a marketing disaster that damages nixos reputation. why are we unable to coordinate and do something correctly? this makes me angry, because nixos has so much potential :((( Functional package management isn't going anywhere. It is grounded as a concept, in my view. | 01:44:58 |
| @matthewcroughan:defenestrate.it | Doesn't matter how long it takes to get there, we can only speed it up. | 01:45:21 |
| @matthewcroughan:defenestrate.it | * Doesn't matter how long it takes to get there, we can only speed it up or slow it down. | 01:45:26 |
| @matthewcroughan:defenestrate.it | What people need are probably simpler explanations in the form of videos. I have a few videos I want to make:
- Declarative Samba
- Declarative Windows VMs
- Declarative Android system image
Things like this are going to shock and wow.
| 01:46:45 |
| @matthewcroughan:defenestrate.it | Offtopic though. Unsure if we have a marketing channel to discuss. | 01:52:56 |
 Dandellion | #marketing:nixos.org | 01:53:44 |
| tomberek | I agree it was a missed opportunity. This was brought up in the marketing meeting. We also don’t have a clean story for the “so what” for reproducibility. In fact, this is true for Nix overall. There’s been an acknowledgment that documentation needs to improve, but I’m also thinking a clear/short enumeration of “why” would be helpful. Additionally, an even shorter slogan (I’m thinking of Rust’s) may help. | 03:52:52 |
 Jonas Chevalier | It would be nice to come up with a way to prevent regressions on the minimal ISO. Like mark the derivations with reproducible = true and extend Hydra to support that. That way it will become possible to extend the reproducible build surface without making it an uphill battle. | 11:54:44 |
 andi- | Would it be sufficient to have the installer a different jobset in hydra and use that reproduce option that is already has? I vaguely remember that hydra was able to build things twice (or more often). | 11:56:05 |
 Gytis Ivaskevicius | In reply to @zimbatm:numtide.com
It would be nice to come up with a way to prevent regressions on the minimal ISO. Like mark the derivations with reproducible = true and extend Hydra to support that. That way it will become possible to extend the reproducible build surface without making it an uphill battle. Maybe it would be cool to have anonymous statistics when running nixpkgs-review and show em in PR. Possibly with history as well 👀 | 19:35:41 |
| Gytis Ivaskevicius | Probably scope is a little too large. Kinda bothersome | 19:36:01 |
| 28 Jun 2021 |
| @matthewcroughan:defenestrate.it changed their display name from matthewcroughan to matthewcroughan - nix.zone. | 02:20:50 |
|  teto joined the room. | 09:30:14 |
|  anodae joined the room. | 09:34:16 |
|  fzakaria (Old) joined the room. | 15:24:38 |
| fzakaria (Old) | Hello! | 15:24:43 |
| fzakaria (Old) | Jonas Chevalier: that's not a bad idea; kind of in between having a fixedOutput but not as deterimental to contributions or changes (having to change). | 15:31:20 |
| fzakaria (Old) | Collecting statistics on the # of nixpkgs that are fixedOutput or reproducible via eval would be great | 15:31:48 |
| fzakaria (Old) | (an attribute would allow that) | 15:31:54 |
