| 15 May 2024 |
 vringar | * I mean, I have been lurking and also read a bunch of blogposts.
Started of with https://xeiaso.net/ and https://jade.fyi/blog/flakes-arent-real/ and this series by Amos
And then I was forced to set up a dev shell, because arch doesn't ship an LLVM that is matching the current Rust llvm, which is required by LibAFL QEMU.
So I first tried with docker, but found the mounting issues too annoying and set up my first flake, which was literally just creating a dev shell and installing a bunch of packages | 20:35:09 |
 @jade_:matrix.org | ok noted that my blog post is very good :3 we should probably go Put It In The Docs when we get to that point | 20:35:50 |
 Qyriad | Jade your blog post is literally the actual best single piece of writing on flakes we have seen so far | 20:36:28 |
| vringar | In reply to @emilytrau:matrix.org
actually you've come to the right place! 😅
desperately needs some dusting off but should do the trick https://github.com/emilytrau/nixpkgs/tree/ghidra-extensions Then emily nerdsniped me into updating her POC and getting it into nixpkgs and for the last week I've had the nixpkgs/pkgs docs and the stdenv page I linked open, along with constantly searching nix doc <term I care about> | 20:36:58 |
| @jade_:matrix.org | actually, asking for some nix support: when is lib.mkIf required in nixos? i am reviewing a PR and I think it is not required in a certain position. i know that it is deferring to deal with evaluation inflooping | 20:37:03 |
| @jade_:matrix.org | but i think that's not a problem if you're just a leaf value rather than some attr set? | 20:37:22 |
| vringar | Also having this channel as a judgement free zone to ask questions was really helpful 😊 | 20:37:45 |
| Qyriad | It's not just to deal with infinite looping, but it is to deal with eval errors in general, is our understanding | 20:37:45 |
| @jade_:matrix.org | mkMerge [(mkIf ...)] is the specific pattern i see here, and it's just making a list | 20:37:45 |
 Grimmauld (moving to @grimmauld:grapevine.grimmauld.de) | In reply to @qyriad:katesiria.org
Jade your blog post is literally the actual best single piece of writing on flakes we have seen so far it also was (and persists to be) something some people take every chance to quote despite him still using flakes :P | 20:37:51 |
| Grimmauld (moving to @grimmauld:grapevine.grimmauld.de) | In reply to @qyriad:katesiria.org
Jade your blog post is literally the actual best single piece of writing on flakes we have seen so far * it also was (and persists to be) something some people take every chance to quote despite them still using flakes :P | 20:37:56 |
| Qyriad | In reply to@jade_:matrix.org
mkMerge [(mkIf ...)] is the specific pattern i see here, and it's just making a list If it's just making a list, and if the values in that list don't depend on config. values, then that's probably equivalent to lib.optionalAttrs? | 20:38:28 |
| vringar | In reply to @grimmauld:grimmauld.de
it also was (and persists to be) something some people take every chance to quote despite them still using flakes :P :D I mean yeah, flakes are my way of ensuring that if I get something to work, I can commit it and keep iterating without fear | 20:38:39 |
| @jade_:matrix.org | In reply to @grimmauld:grimmauld.de
it also was (and persists to be) something some people take every chance to quote despite them still using flakes :P it's SO funny that people think i hate flakes, when in fact, i do dislike flakes but use them every day lol | 20:38:48 |
| Qyriad | Which is basically the story of most of the Lix team's opinions on flakes lol | 20:39:16 |
 mjm | i quote it to people to help them understand where flakes start and end, when they seem fundamentally confused about that | 20:39:17 |
| Qyriad | Very much | 20:39:28 |
| Grimmauld (moving to @grimmauld:grapevine.grimmauld.de) | In reply to @jade_:matrix.org
it's SO funny that people think i hate flakes, when in fact, i do dislike flakes but use them every day lol oh its a fair point to have, but it gets really funny and ironic at times | 20:39:34 |
| @jade_:matrix.org | yes, it's my fave way of making people stop thinking flakes are magic bullshit that they absolutely need all the time | 20:39:36 |
| Grimmauld (moving to @grimmauld:grapevine.grimmauld.de) | and then they say germans don't have humor | 20:39:40 |
| Qyriad | In reply to@qyriad:katesiria.org
Very much It does not help that outside of flakes Nix has like 4 different common entrypoint patterns | 20:39:46 |
| mjm | In reply to @jade_:matrix.org
yes, it's my fave way of making people stop thinking flakes are magic bullshit that they absolutely need all the time it's very good at communicating that, i think | 20:40:00 |
| Qyriad | It's wild to see how many questions there are in the #flakes channel of the NixOS Matrix space that have absolutely nothing to do with flakes at all, the user in question just happens to be doing the thing in a flake and doesn't understand that there's no difference | 20:41:04 |
| Grimmauld (moving to @grimmauld:grapevine.grimmauld.de) | wait thewre is a flakes channel? oh dear flying monster of sphagetti.... | 20:41:43 |
| Qyriad | Though to be fair this problem is rather widespread in the ecosystem. There are lots of questions in the #nix-language channel that are not language questions at all, and likewise we see tons of people complaining that "the Nix language" is incredibly confusing or hard to read, and then go on to describe the NixOS Module system or the mkDerivation API, etc | 20:42:42 |
 samrose | Flakes hurt me, but I got over it | 20:44:55 |
| Qyriad | Flakes suck but no lockfile and no schema sucks more (most of the time) | 20:47:47 |
| mjm | no lockfile is plenty easy to solve, no schema hurts tho | 20:48:14 |
| @jade_:matrix.org | no schema really hurts yes | 20:48:31 |
| @jade_:matrix.org | that's the big thing i want to fix in a flakes replacement | 20:48:47 |
