nix-env -i haskell-haskell-platform-ghc7.6.1-2012.2.0.0 # assertion failed

[skull-squadron]

[root@freebsd ~]# nix-env -i haskell-haskell-platform-ghc7.6.1-2012.2.0.0
installing `haskell-haskell-platform-ghc7.6.1-2012.2.0.0'
error: assertion failed at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/development/compilers/ghc/7.0.4-binary.nix:7:1'
(use `--show-trace' to show detailed location information)
[root@freebsd ~]# nix-env -i haskell-haskell-platform-ghc7.6.1-2012.2.0.0 --show-trace
installing `haskell-haskell-platform-ghc7.6.1-2012.2.0.0'
error: while evaluating the builtin function `derivationStrict':
while instantiating the derivation named `haskell-haskell-platform-ghc7.6.1-2012.2.0.0' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/build-support/cabal/default.nix:36:13':
while evaluating the derivation attribute `buildNativeInputs' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/stdenv/generic/default.nix:67:15':
while evaluating the builtin function `getAttr':
while evaluating the builtin function `derivationStrict':
while instantiating the derivation named `ghc-7.6.1-wrapper' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/development/compilers/ghc/wrapper.nix:8:3':
while evaluating the derivation attribute `propagatedBuildNativeInputs' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/stdenv/generic/default.nix:69:15':
while evaluating the builtin function `getAttr':
while evaluating the builtin function `derivationStrict':
while instantiating the derivation named `ghc-7.6.1' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/development/compilers/ghc/7.6.1.nix:6:3':
while evaluating the derivation attribute `buildNativeInputs' at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/stdenv/generic/default.nix:67:15':
while evaluating the function at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/lib/meta.nix:41:13':
while evaluating the function at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/lib/meta.nix:13:28':
while evaluating the function at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/lib/customisation.nix:98:35':
while evaluating the function at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/lib/customisation.nix:59:24':
while evaluating the builtin function `isAttrs':
while evaluating the function at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/development/compilers/ghc/7.0.4-binary.nix:1:1':
assertion failed at `/nix/store/rcwjawql0ck2nbrnpkqgrgxjm2k42z3q-nixpkgs-1.0pre21403_def0189/nixpkgs/pkgs/development/compilers/ghc/7.0.4-binary.nix:7:1'

[skull-squadron]

let
  supportedPlatforms = ["x86_64-linux" "i686-linux" "i686-darwin" "x86_64-darwin"];
in

assert stdenv.lib.elem stdenv.system supportedPlatforms;

[skull-squadron]

What is obvious to you or I would not be obvious to an end-user.

[peti]

@steakknife, what do you suggest we do?

[skull-squadron]

Is there syntax semantics to decorate an assertion with a failure message?

[peti]

There is 'throw', which does supply an error message. It's my understanding, though, that assert and throw differ in their laziness, i.e. it is in general not possible to replace assert with throw and to expect the same outcome. I'm not sure whether that would cause problems in our particular use case. @kosmikus should know?

[kosmikus]

I have no idea. We'd have to try.

[edolstra]

"throw" behaves in the same way as "assert" except for the error message. You may be thinking of "abort", which cannot be caught by tryEval.

[peti]

Unfortunately, throw cannot easily replace assert in ghc-7.4.2-binary.nix and friends. At least I don't see how without changing the structure of that expression quite a bit. Am I missing something?

What is the point of that assertion anyway? Isn't the throw statement in the srcattribute good enough?

[kosmikus]

I think the throw is good enough.

[peti]

Commit b1169e5 removes the call to assert. I guess that change fixes this issue?

[edolstra]

Closing, appears to be fixed (and in any case not a Nix issue).