| 5 Nov 2024 |
tjni | What is happening when this is printed?
❯ nix flake update
[44.9 MiB DL] unpacking 'github:NixOS/nixpkgs/a04d33c0c3f1a59a2c1cb0c6e34cd24500e5a1dc' into the Git cache
On my company Mac laptop, we run MS Defender (wdavdaemon process), and each time this happens it takes minutes (sometimes 9-10) before it finishes. So I'm trying at first to get some help understanding what is happening during this step. | 18:06:33 |
K900 | Unpacking a tarball with a lot of files | 18:07:40 |
tjni | Is it unpacking into the Nix store? | 18:09:58 |
K900 | Yes | 18:10:09 |
emily | actually, "the Git cache" isn't in the store | 18:17:31 |
emily | it's copying it into a Git repository(!) in your ~/.cache | 18:17:40 |
emily | the unpacking into the store comes after | 18:17:45 |
tjni | Oh hmm. I see, thank you | 18:19:08 |
emily | FWIW, this is a 2.19+ thing, 2.18 may be faster (but is going EOL soon) | 18:19:26 |
tjni | Yeah, I'm using nixVersions.git. But maybe I can try to repro this outside of Nix if it's just untarring into this cache directory. | 18:20:06 |
emily | nah it's using libgit2 | 18:20:47 |
emily | https://github.com/NixOS/nix/pull/10038 | 18:20:48 |
emily | see also https://github.com/NixOS/nix/pull/11330 which helped the macOS perf regression somewhat | 18:21:08 |
| 6 Nov 2024 |
| Viorel-Cătălin Răpițeanu changed their display name from Viorel-Cătălin Răpițeanu to @catalin:one.ems.host. | 07:02:25 |
| seapat joined the room. | 13:45:44 |
| 7 Nov 2024 |
Mic92 | tjni: try this: git+https://github.com/NixOS/nixpkgs?shallow=1 | 06:32:33 |
Mic92 | * tjni: try this: nixpkgs.url = "git+https://github.com/NixOS/nixpkgs?shallow=1"; | 06:32:56 |
Mic92 | I found shallow clones both to be faster on the first but also on subsequent runs. | 06:33:26 |
Mic92 | Robert Hensing (roberth): I am trying to drop nix 2.19, but your hercules-ci-cnix-store still depends on it. | 06:34:24 |
Robert Hensing (roberth) | right, I'll update it this week | 08:27:50 |
Martin Schwaighofer | Hi 👋 my name is Martin, I'm new in this channel!
I have two things that I would like to discuss with someone from the Nix team.
* My comment on https://github.com/NixOS/nix/pull/11749#issuecomment-2462223730 and
* [my paper on Extending Cloud Build Systems to Eliminate Transitive Trust](https://www.digidow.eu/publications/2024-schwaighofer-scored/Schwaighofer_2024_SCORED24_CloudBuildSystemsTrust.pdf), also covered by [my talk about that work at NixCon 2024](https://www.youtube.com/watch?v=UlJUpUQc9Lc).
Would one of the regular meetings be suitable to discuss this, or would some other way be better? 😊 | 15:29:08 |
Mic92 | Martin Schwaighofer: both would be fine. If it requires a lot of back and forth, meeting might work better. Checkout this: https://github.com/NixOS/nix/tree/master/maintainers#meeting-protocol | 17:40:48 |
Martin Schwaighofer | Thanks, I will try to join the next work meeting on Monday and see how that goes. 😊 | 22:39:55 |
| 8 Nov 2024 |
p14 | Observation: If I do nix build .#{foo,bar,baz,qux} in a dev checkout of nixpkgs, I appear to see all of nixpkgs get copied to the store repeatedly. At least, it takes a while and the 'copying' message is shown. Is there a cheaper way to eval a large number of packages during development? | 11:27:13 |
Mic92 | p14: nix build -f . .#{foo,bar,baz,qux} | 13:24:38 |
Mic92 | for nixpkgs | 13:24:45 |
Mic92 | maybe also: nix build -f . .#{foo,bar,baz,qux} -L | 13:24:59 |
Mic92 | p14: actually you need to do this: nix build -f . foo bar baz qux -L | 13:26:23 |
p14 | In reply to @joerg:thalheim.io p14: actually you need to do this: nix build -f . foo bar baz qux -L That is fine for nixpkgs but what about flake-only projects? | 13:34:16 |
| moved to @amadaluzia:tchncs.de changed their display name from (artur 'manuel) to @amadaluzia:tchncs.de. | 14:28:52 |