| 5 Aug 2022 |
 profpatsch | Effectively it’s against protecting the value of nixpkgs from upstream. | 09:57:43 |
| profpatsch | Maybe impure isn’t part of nixpkgs, but that it stopped working is sign of deeper problems with how nix handles backwards compat. Maybe tomorrow it will be some random thing a few packages in nixpkgs depend on. Maybe the day after, haskellPackages stops working with nix-build | 09:59:30 |
| profpatsch | At that point, having a policy that says “no you have to fix it and release a patch or revert the thing that broke it” is a life saver | 10:00:35 |
| profpatsch | Also I fear the old-style commands will just unceremoniously be removed in a few months time, and every company on this planet will be stuck on outdated versions of nix and nixpkgs. | 10:01:32 |
 FRidh | In reply to @profpatsch:augsburg.one
Maybe impure isn’t part of nixpkgs, but that it stopped working is sign of deeper problems with how nix handles backwards compat. Maybe tomorrow it will be some random thing a few packages in nixpkgs depend on. Maybe the day after, haskellPackages stops working with nix-build Did __impure ever work with nix-build? | 10:03:17 |
| profpatsch | FRidh: that’s not the question here | 10:03:50 |
| profpatsch | I’m not gonna start arguing technicalities, sorry. Strawman. | 10:04:17 |
| FRidh | All Nix development nowadays seems to focus on features which are only supported with nix build and not nix-build, meaning newer features cannot be used without switching to nix build as well. Either these newer features are going to be supported in nix-build as well, or it means we just can't use them in Nixpkgs yet for a while. That's really no different from how we already work with minver. If I am correct your worry is about that bumping this minver and thereby forcing nix build (and maybe even dropping nix-build) can happen real quickly. In that case I think the best way forward is to get ahead and get an RFC out stating what is acceptable and what not. Supporting new cli is more effort so it is very fair IMO to expect more transitioning time there than when just using newer features in Nixpkgs. | 10:12:14 |
| profpatsch | There is absolutely no reason a nix-the-language feature like impure could not be implemented for nix-build as well, unless it requires CLI support | 10:15:59 |
| profpatsch | new builtins as well | 10:16:12 |
| profpatsch | It’s just a very basic thing about not burning bridges | 10:16:45 |
| profpatsch | It’s okay to require people to update nix versions at a reasonable rate, but it’s not okay to change the CLI interface of nix without giving people time to adapt. 2 years is already on the low end, Python2 migration was well over a decade before EOL (and even that is questionable given that there is still lots of python2 out there) | 10:18:30 |
| profpatsch | I mean, I guess I sound like it already went south, but I’m pretty sure it might go there if nobody starts thinking about this now. | 10:20:16 |
 Wanja Hentze | yes, and unsupporting nix-build now, when the new thing is not even stable yet, means there is essentially a negative amount of grace period, a time window in which nothing at all is formally supported | 10:28:41 |
| Wanja Hentze | that would be pretty painful | 10:28:53 |
 mkaito | This tribalism is troubling. We are Nix. There is no “them”. I do not support this adversarial attitude in the team. | 11:59:49 |
 Alyssa Ross | As a Nixpkgs developer, I do not feel represented or listened to by the Nix developers. I tihnk profpatsch feels the same way. | 12:36:45 |
| Alyssa Ross | We are different groups, and there's a sizeable subset of Nixpkgs developers that the Nix developers have not succeeded in getting on board with their plans | 12:38:05 |
| Alyssa Ross | That's not our fault | 12:38:20 |
 K900 | As kind of an outside person, or at least a rather new contributor, it definitely feels like the communication has broken down | 12:39:22 |
| K900 | Between the Nixpkgs contributors and the Nix contributors that is | 12:39:35 |
|  Bryan Honof joined the room. | 13:13:57 |
| mkaito | Well if that's how we feel, then that's the first problem we should be tackling. If we go into this with an "us versus them" mentality and resentment from the outset, we might as well not bother with anything we said we want to try. | 13:51:42 |
| mkaito | We are Nix. All of us. If that's not true, then we start there. | 13:52:13 |
 infinisil | While I think it's important to fix that, I don't think this team is the right one to tackle that | 13:53:16 |
| mkaito | Your call | 13:54:15 |
| mkaito | But if this is how we plan to conduct things here, I'm afraid I don't have the time or patience to put up with petty grievances and resentment. | 13:54:58 |
| mkaito | Please keep that in mind when reviewing my application | 13:55:14 |
| infinisil | I guess if we discuss the idea of splitting up nixpkgs this could become relevant for this team | 13:58:37 |
| infinisil | Since flakes will be an option | 13:58:49 |
