[Nix-dev] Publish All of Hackage

[Oliver Charles]

On Thu, Nov 19, 2015 at 11:14 AM Peter Simons <simons at cryp.to> wrote:


> One issue is that checking that ~50MB file into Nixpkgs might be a bad
> idea, because it sets a dangerous precedent. Arguably, if we check all
> of Hackage into Nixpkgs, then we cannot reasonably say *no* to someone
> who wants to check all of CPAN into Nixpkgs too, and before we know it
> our Git repository triples in size. So it might be wise to put the full
> Hackage variant of hackage-packages.nix into a separate repository that
> we fetch via "builtins.fetchurl" if the corresponding option is enabled.
>

It seems a bit of a shame to me that the current state of the world is that
we are forced to essentially check in a build result into the nixpkgs
repository. The code that you really want to commit is the scripts that
*generate* the 50mb file as some sort of Nix expression. Then, when I as a
user choose to evaluate the set of Haskell packages, I will be forced to
generate all the Nix expressions - or, this being Nix, ask a binary
substitution server for that.

I believe this requires some sort of "hierarchical Nix" or "Nix-in-Nix"
that we don't have, but realistically the approach of checking in huge Nix
expressions does have scaling problems. Not to mention I'm always nervous
of checking in files that should not be edited by humans.

Ollie
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.science.uu.nl/pipermail/nix-dev/attachments/20151120/b419f857/attachment.html