| 25 Jun 2021 |
 David Arnold | * But yeah, it laks a point about nix 3.0 entanglement (or disentaglement) | 03:05:32 |
| David Arnold | * But yeah, it lacks a point about nix 3.0 entanglement (or disentaglement) | 03:05:37 |
| David Arnold | * But yeah, it lacks a point about nix 3.0 entanglement (or disentaglement) | 03:05:46 |
| David Arnold | I also think we should revert the outputs schema and let it grow via a separate RFC-like process or provide outputs.name.__contract semantics, where the tooling can consume the contract and validate against it | 03:06:36 |
| David Arnold | And we should refine the inputs semantics | 03:07:04 |
 ash | i don't care what happens to flakes, i just want it gone so it stops being an impediment to the healthy ecosystem | 03:07:22 |
| David Arnold | In reply to @kity:kity.wtf
i don't care what happens to flakes, i just want it gone so it stops being an impediment to the healthy ecosystem Nah... that stance needs still a little work, I'd guess. | 03:07:57 |
| David Arnold | There is already some work consolidating around the input stuff here: https://github.com/NixOS/nix/pull/4641 | 03:08:13 |
| ash | no it doesn't. flakes can come back when it's ready. | 03:08:20 |
| David Arnold | In reply to @kity:kity.wtf
no it doesn't. flakes can come back when it's ready. Make a distionction | 03:08:32 |
| David Arnold | In reply to @kity:kity.wtf
no it doesn't. flakes can come back when it's ready. * Make a distinction | 03:08:39 |
| David Arnold | ... where it's approriate. | 03:08:45 |
| David Arnold | flake vs nix 3.0 | 03:08:50 |
| David Arnold | or nix flake | 03:08:56 |
| David Arnold | vs flake.nix | 03:09:00 |
| ash | flakes should never have been introduced to nix unstable in the first place | 03:09:22 |
| David Arnold | I hear you point as such: nix flake go away so we can release nix | 03:09:29 |
| ash | they should've been developed separately | 03:09:30 |
| David Arnold | What are flakes? | 03:10:36 |
| David Arnold | * What are flakes? | 03:11:33 |
| ash | the thing i'm seeing with this document is a lot of discussion that would be welcome in an rfc | 03:13:02 |
| David Arnold | In reply to @kity:kity.wtf
they should've been developed separately A plugin system for nix would be nice. | 03:13:04 |
| ash | but this isn't an rfc | 03:13:09 |
| ash | it's a discussion about what to do right now | 03:13:17 |
| ash | which doesn't mean "rethinking flakes" or changing the implementation | 03:13:29 |
| David Arnold | In reply to @kity:kity.wtf
it's a discussion about what to do right now I would be suffering if they where removed omorrow) | 03:14:10 |
| ash | In reply to @blaggacao:matrix.org
A plugin system for nix would be nice. nobody needs a plugin system. flakes can be implemented in pure nix at least as a proof of concept, and it would've been easier to hack on the implementation then | 03:14:14 |
| David Arnold | In reply to @kity:kity.wtf
it's a discussion about what to do right now * I would be suffering if they where removed tomorrow | 03:14:16 |
| David Arnold | In reply to @kity:kity.wtf
nobody needs a plugin system. flakes can be implemented in pure nix at least as a proof of concept, and it would've been easier to hack on the implementation then Sure, we need one. I'm no Cpp expert. | 03:14:31 |
| David Arnold | (I don't know even not a littl bit) | 03:14:47 |
