Comment by MarsIronPI
2 days ago
So do Gentoo and Nix, yet they have packaging separate from the source code. The source is fetched, but the build is sandboxed from the network during the configure, build and install phases. So it's technically possible.
Nix definitely does not allow most things to be built offline (at least in the way Debian means it).
With Nix, any fetcher will download the source. It does so in a way that guarantees the shasum of what is fetched is identical, and if you already have something in the nix store with that shasum, it won't have to fetch it.
However, with just a mirror of the debian source tree, you can build everything without hitting the internet. This is assuredly not true with just a mirror of nixpkgs.
> With Nix, any fetcher will download the source.
OK, I see how the Debian idea differs from the Portage/Nix/etc. idea. For Portage and Nix it is enough that the build proper be offline, but the source code is fetched at the beginning of the package build. Not only do I find this sufficient, I prefer it because IMO it makes the package easier to work with (since you're only wrangling the packaging code, not upstream's).
There are probably still ways to maintain a source archive with a ports system. Just analyze the sources used by builds, create a mirror, and redirect fetches to use the mirror. It's not that crazy. The packaging would still be a separate affair.
This is exactly what Nix does, if I understand correctly: the "mirror" is cache.nixos.org as mentioned in XorNot's reply to the same parent post.
3 replies →
Nix and specifically nixpkgs is IMO very bad at this. It's not a distro: it's a collection of random links that in many cases now only exists in cache.nixos.org. The tarball server frequently doesn't have content, can't represent some content at all (recursive hash types), links have rotted away completely (broadcom driver zips referencing a domain which is now advertising online gambling).
Nix isn't functional: it's a functional core that moved every bit of the imperative part to an even less parseable stage, labelled it "evaluation" and then ignored any sense of hygiene about it.
No: your dependency tree for packaging should absolutely not include an opaque binary from a cache server, or a link to a years old patch posted on someone else's bugzilla instance (frequently link rotted as well).
Nothing has made me appreciate the decisions of mainstream distributions more then dealing with an alternative like this.
Not saying you're wrong, but in the ~year I've been using NixOS I've never noticed anything like that.
The heavy asterisk here is that none of this actually makes using NixOS impossible because it obviously still works. But when you get into the problem I am - and one of the major purported benefits of it which is reproducibility and traceability - this is a pretty serious issue.
So long as the NAR files in cache.nixos.org exist, everything will work - that's not a problem. But if you actually choose to exercise that traceability - which is what I've been working on - suddenly you start finding all this stuff. The problem is nixpkgs doesn't expose or archive the code: it archives a reference to code that existed somewhere at some time, and worse it obfuscates what the code was - I can obviously still go get it from the NAR files, but I can't get any of the context surrounding it.
By contrast, things like the Fedora and Debian patching systems have - crucially - actual archives of what they're building, the patches they're building them with, and the commit messages or other notes on why those patches are being applied and the change record of them. With NixOS you get a bunch of hashes that terminates on "wefu123r23hjcowiejcwe.nar" and you don't know what that is until nixpkgs happens to evaluate it and calculate it, which means it's impossible to even know up-front what's going to be pulled in.
Then of course you get to practical matters: just because you can exactly specify dependencies doesn't mean you should - we all realized with containers that having a couple dozen versions of libraries kicking around is a bad idea (and lo and behold that's what traditional distro packaging tries to minimize) - and that's where all those calculated paths burn you anyway. Nix is a fairly freeform programming language, so it's nigh impossible to stop some snowflake package from pulling in a different version of a compiler or library even if I can see it happening (example I currently have: 5 different version of Rust, 5 different versions of Golang - and the invariant I want on that is "no, it's this version and you deal with it" - but there's a lot of ways nix will let you make this which are very resistant to static analysis or automated correction).
that's a pretty damning detail about nix which everyone else seems to speak so highly about. are there any articles that explain this in more detail?