| 18 Jul 2021 |
Narice | Alright, someone is going to file a PR to resolve this issue! That's awesome! | 17:00:26 |
Narice | Thanks David Arnold for the confirmation that this was most probably an upstream thing! | 17:01:34 |
David Arnold (blaggacao) | fufexan: Here is my (very) WIP Docs PR for digga https://github.com/divnix/digga/pull/90/files | 20:33:27 |
David Arnold (blaggacao) | * fufexan: Here is my (very) WIP Docs PR for digga https://github.com/divnix/digga/pull/90/files (it's basically just copying over files from devos) | 20:34:10 |
@teutat3s:pub.solar | Did anybody else notice issues with caches when trying out the develop branch of devos?
I can see multiple substituters entries in /etc/nix/nix.conf the last one probably overriding the first one. The last entry only contains the following (notably https://cache.nixos.org is missing here):
substituters = https://nrdxp.cachix.org https://nix-community.cachix.org
trusted-public-keys = nrdxp.cachix.org-1:Fc5PSqY2Jm1TrWfm88l6cvGWwz3s93c6IOifQWnhNW4= nix-community.cachix.org-1:mB9FSh9qf2dCimDSUo8Zy7bkq5CX+/rkCWyvRCYg3Fs=
| 20:36:37 |
@teutat3s:pub.solar | Noticed something strange must've happened when the devshell started building gcc from source :D | 20:36:57 |
@teutat3s:pub.solar | * Noticed something strange must've happened when the devshell started building gcc from source 😀 | 20:37:18 |
@teutat3s:pub.solar | * Did anybody else notice issues with caches when trying out the develop branch of devos?
I can see multiple substituters entries in /etc/nix/nix.conf after running nixos-rebuild switch, the last one probably overriding the first one. The last entry only contains the following (notably https://cache.nixos.org is missing here):
substituters = https://nrdxp.cachix.org https://nix-community.cachix.org
trusted-public-keys = nrdxp.cachix.org-1:Fc5PSqY2Jm1TrWfm88l6cvGWwz3s93c6IOifQWnhNW4= nix-community.cachix.org-1:mB9FSh9qf2dCimDSUo8Zy7bkq5CX+/rkCWyvRCYg3Fs=
| 20:38:26 |
@teutat3s:pub.solar | Maybe related to https://github.com/divnix/devos/commit/6c25e77f6fc3e4c0fa0eb009aae48a817b6bac51 ? | 20:38:57 |
David Arnold (blaggacao) | https://github.com/divnix/devos/commit/6c25e77f6fc3e4c0fa0eb009aae48a817b6bac51 is supposed to be persisted on a per-repo basis. | 20:45:12 |
David Arnold (blaggacao) | Using extra-* is supposed to behave well with existing configuration. | 20:45:33 |
David Arnold (blaggacao) | However, this line adds optional configuration to your nixos system. | 20:46:07 |
David Arnold (blaggacao) | It was supposed to not conflict, but maybe we need to revise it's implementation? | 20:47:03 |
David Arnold (blaggacao) | (btw. the cache that is advertised there is just a tiny bit short of its existence, yet) 🤣 | 20:49:54 |
David Arnold (blaggacao) |
the last one probably overriding the first one.
If this is what seems like being happening, we probably need a better implementation of nix.extraOptions, in general *and in nixpkgs. Sure we need to provide an "anti-corruption-layer" in the meantime within digga.
| 20:51:44 |
David Arnold (blaggacao) | Let me know if we found something here... | 20:51:51 |
David Arnold (blaggacao) | Is nix.conf supposed to be a ini style file? | 20:59:57 |
David Arnold (blaggacao) | (we could make use of https://github.com/NixOS/nixpkgs/blob/master/pkgs/pkgs-lib/formats.nix#L60) | 21:00:34 |
| 19 Jul 2021 |
David Arnold (blaggacao) | I've just run across a case, where I want to manage my collection of devshells as part of a system environment, but don't want all of those devshells to be combined as my devos devshell. Sounds familiar? | 00:04:19 |
David Arnold (blaggacao) | Thinking about this, I guess we have still room left for improving digga's devshell api container before v1.0... It would be nice if bud would allow me to enter arbitrary devshells across those defined in my personal system environment at any time. I'm not even sure any more if we should provide a way to compose the devos devshell at all through digga (we could easily implement a ./shell.nix file for that job in devos). | 00:07:10 |
David Arnold (blaggacao) | Help! https://github.com/divnix/digga/pull/92/files produces :
error: infinite recursion encountered
at /nix/store/6kr1yqygz4p3ph3pdyd9ps9z37sppfgm-source/examples/infiniteRecursion/flake.nix:6:14:
5| digga.lib.mkFlake {
6| inherit self;
| ^
7| nixos = ./os;
| 01:50:21 |
David Arnold (blaggacao) | ... and I have no idea what's going on. 🤷 | 01:50:42 |
fufexan | In reply to @blaggacao:matrix.org fufexan: Here is my (very) WIP Docs PR for digga https://github.com/divnix/digga/pull/90/files (it's basically just copying over files from devos) nice, I'll take inspiration | 12:20:08 |
fufexan | David Arnold: can you give me a few pointers on this when you have time? https://github.com/gytis-ivaskevicius/flake-utils-plus/pull/84#discussion_r672162523 | 12:32:45 |
fufexan | thanks!
| 12:46:33 |
| 20 Jul 2021 |
@gytis-ivaskevicius:matrix.org | David Arnold: can you take another look at that PR? | 15:11:41 |
David Arnold (blaggacao) | Gytis Ivaskevicius: fufexan what are your thoughts on merge any? this feels like a forward-compatibility layer towards nickel or alternatively a substitute of the module system merge semantics. | 18:14:45 |
David Arnold (blaggacao) | * Gytis Ivaskevicius: fufexan what are your thoughts on lib.mergeAny? This feels like a forward-compatibility layer towards nickel or alternatively a substitute of the module system merge semantics. | 18:15:01 |
David Arnold (blaggacao) | * Gytis Ivaskevicius: fufexan what are your thoughts on lib.mergeAny? This feels like a forward-compatibility layer towards nickel or alternatively a substitute of the module system merge semantics. So I wonder if we can give it more concise / niche semantics in the big theater of configuration merging. | 18:15:30 |
David Arnold (blaggacao) | * Gytis Ivaskevicius: fufexan what are your thoughts on lib.mergeAny? This feels like a forward-compatibility layer towards nickel or alternatively a substitute of the module system merge semantics. So I wonder if we can give it more concise / niche semantics in the big theater of configuration merging. | 18:15:47 |