nix flake show: don't evaluate derivations for foreign systems by default #6988

max-privatevoid · 2022-09-01T20:09:13Z

Simple implementation of the idea, based on how legacyPackages is handled already. Doesn't cover hydraJobs, unfortunately.

Fixes #6985

…ault

thufschmitt

I like this idea, although it conflicts a bit with the idea that the system hierarchy layer should be special-cased as least as can be. But probably that's worth it.

thufschmitt · 2022-09-05T14:59:01Z

src/nix/flake.cc

@@ -957,6 +957,7 @@ struct CmdFlakeArchive : FlakeCommand, MixJSON, MixDryRun
 struct CmdFlakeShow : FlakeCommand, MixJSON
 {
    bool showLegacy = false;
+    bool showForeign = false;


I'm not keen on is having showForeign false by default, because it would be bad for backwards-compatibility – esp. for the json output which is designed to be used in scripts and is more dangerous to break.

I totally understand your sentiments, though having it defaulting to true is a major annoyance for the human.

Evaluation of nix flake show will take much longer than needed

Evaluation of nix flake show will show a lot of redundant information

Evaluation of nix flake show will fail for a majority of flakes

We can avoid all of these 3 problems by defaulting it to false.

At the same time, breaking the interface is fine, as long as we do it now. After 3.0 has been released we can not do it anymore.

Back when 2.4 got released, no one cared that it broke a lot of scripts. Back then it has been pointed out that its at your own risk to use an experimental API. So… If someone complaints, why did they use an experimental API if they do not want things to change?

Also, it is much easier to add --foreign to a handful of scripts, but annoying to have to remember to add it to each and every invocation to shave off minutes of eval time and annoying IFD errors…

I'd even suggest to make this a nix.conf option, such that every user can decide on their very own default in the systems or users nix.conf.

This should definitely not be a nix.conf option, as tempting as that is, since it's a bad idea for scripting to have options that fundamentally alter what the output of a command is. You can end up with a script that unexpectedly fails on another machine or user account because foreign is set to a different value...

NobbZ · 2022-09-05T16:45:46Z

I like this idea, although it conflicts a bit with the idea that the system hierarchy layer should be special-cased as least as can be. But probably that's worth it.

I think we can easily agree on the fact that the way the system is currently handled has a lot of flaws, and doing special handling it will be the only way around it (besides rewriting "flakespec" from scratch).

NobbZ · 2022-09-05T20:44:32Z

And I also want to add:

Specialcasing the system within nix' source code wasn't necessary that mouch if we were doing checks and shows in nix expression language, as briefly suggested in #6453 and #6454.

Of course I see, that we would still need to tell the nix expressions what the current system is.

edolstra · 2022-09-06T11:44:39Z

--foreign is not very descriptive as an option name. Maybe --all-systems or something like that?

nixos-discourse · 2022-09-06T12:00:24Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/deduplicating-a-package-definition-between-nixpkgs-and-a-flake/21460/4

yajo · 2022-09-22T06:31:12Z

Or maybe --systems=all (default value for flag: --systems=local). That would let us also give a specific list of systems to eval (--systems=x86_64-linux,aarch64-linux).

L-as · 2022-10-26T11:38:27Z

Is there any reason why we need this instead of just catching errors that happen due to not being able to build a drv and omitting those parts of the output?

yajo · 2022-10-26T11:42:54Z

For nix flake show, probably it would be the same. However, for nix flake check we would need a way to check the flake only for the current arch. Otherwise an IFS would always become an error, even if it's not.

nixos-discourse · 2022-11-30T10:38:29Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/announcing-stacklock2nix-easily-build-a-haskell-project-that-contains-a-stack-yaml-lock-file/23563/5

domenkozar · 2022-12-19T11:37:00Z

Lost ~hour today due to this unexpected behavor while working on cachix/cachix#476

adrian-gierakowski · 2023-01-26T23:23:16Z

What is stopping this fix from being merged?

thufschmitt · 2023-01-27T09:17:18Z

What is stopping this fix from being merged?

Nothing really.

@max-privatevoid I've merged master back into this and added a few tests. Let me know whether that looks good to you

thufschmitt · 2023-01-27T09:18:21Z

Oh also: I've noticed that this doesn't apply to legacyPackages. Is it intentional?

(I don't think it should be a blocker, and I think it makes sense either way, just wants to make sure it's not an overlook)

max-privatevoid · 2023-01-27T23:36:03Z

Thanks for writing the tests. I've added the functionality to legacyPackages as well now. I definitely meant to have the flag work there too, just forgot about it.

Don't hardcode “x86_64-linux” as this won't work too nicely on other platforms

thufschmitt · 2023-01-30T09:46:51Z

Awesome, thanks. I've just pushed a fix for the tests because they were broken on non-x86_64-linux. If the CI is OK, it'll just merge

Things leading to another...

`nix flake show` now skips derivations for foreign systems: NixOS#6988 This commit borrows from that to implement the same behavior for `nix flake check`. See "nix flake check breaks on IFD in multi-platform flake" NixOS#4265

peterbecich · 2023-02-06T00:05:27Z

Thanks for this PR, I am trying to use it to implement the same behavior for nix flake check: #7759

`nix flake show` now skips derivations for foreign systems: NixOS#6988 This commit borrows from that to implement the same behavior for `nix flake check`. See "nix flake check breaks on IFD in multi-platform flake" NixOS#4265

nixos-discourse · 2023-02-28T17:52:05Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/nix-2-14-0-released/25900/2

`nix flake show` now skips derivations for foreign systems: NixOS#6988 This commit borrows from that to implement the same behavior for `nix flake check`. See "nix flake check breaks on IFD in multi-platform flake" NixOS#4265

nixos-discourse · 2023-04-21T11:57:44Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/nix-team-report-2022-10-2023-03/27486/1

`nix flake show` now skips derivations for foreign systems: NixOS#6988 This commit borrows from that to implement the same behavior for `nix flake check`. See "nix flake check breaks on IFD in multi-platform flake" NixOS#4265

`nix flake show` now skips derivations for foreign systems: #6988 This commit borrows from that to implement the same behavior for `nix flake check`. See "nix flake check breaks on IFD in multi-platform flake" #4265

nix flake show: don't evaluate derivations for foreign systems by def…

02bff90

…ault

thufschmitt reviewed Sep 5, 2022

View reviewed changes

showForeign -> showAllSystems

6da4cc9

NobbZ mentioned this pull request Sep 21, 2022

nix flake check breaks on IFD in multi-platform flake #4265

Open

Théophane Hufschmitt added 2 commits January 27, 2023 09:46

Merge remote-tracking branch 'nixos/master' into pr-flake-show-foreign

ab424a3

Add a test for nix flake show

79c084c

apply showAllSystems to legacyPackages as well

02e81cd

peterbecich mentioned this pull request Jan 29, 2023

Mac & Linux support of Flake template input-output-hk/haskell.nix#1825

Open

Fix the flakes/show test

bc6e65e

Don't hardcode “x86_64-linux” as this won't work too nicely on other platforms

thufschmitt approved these changes Jan 30, 2023

View reviewed changes

thufschmitt enabled auto-merge January 30, 2023 09:45

Fix the flakes init test

3ac9f16

Things leading to another...

thufschmitt merged commit 575d0ae into NixOS:master Jan 30, 2023

peterbecich mentioned this pull request Feb 5, 2023

nix flake check: skip derivations for foreign systems #7759

Merged

7 tasks

peterbecich mentioned this pull request Feb 6, 2023

nix flake check fails input-output-hk/haskell.nix#1097

Closed

Kha mentioned this pull request Mar 19, 2023

Not commiting the flake.lock file makes the nix code virtually unusable leanprover/lake#165

Closed

peterbecich mentioned this pull request Apr 15, 2023

nix flake show should be filterable by the current system #6985

Closed

jamesdbrock mentioned this pull request Jun 6, 2023

flake nixConfig justinwoo/easy-purescript-nix#225

Closed

peterbecich mentioned this pull request Aug 22, 2023

behavior of show and check changed in Nix recently nix-systems/nix-systems#9

Open

bew mentioned this pull request Sep 20, 2023

nix flake show is cluttered with information I'm not interested in #9011

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

nix flake show: don't evaluate derivations for foreign systems by default #6988

nix flake show: don't evaluate derivations for foreign systems by default #6988

max-privatevoid commented Sep 1, 2022

thufschmitt left a comment

thufschmitt Sep 5, 2022

NobbZ Sep 5, 2022

edolstra Sep 6, 2022

NobbZ commented Sep 5, 2022 •

edited

Loading

NobbZ commented Sep 5, 2022

edolstra commented Sep 6, 2022

nixos-discourse commented Sep 6, 2022

yajo commented Sep 22, 2022

L-as commented Oct 26, 2022

yajo commented Oct 26, 2022

nixos-discourse commented Nov 30, 2022

domenkozar commented Dec 19, 2022

adrian-gierakowski commented Jan 26, 2023

thufschmitt commented Jan 27, 2023

thufschmitt commented Jan 27, 2023

max-privatevoid commented Jan 27, 2023

thufschmitt commented Jan 30, 2023

peterbecich commented Feb 6, 2023

nixos-discourse commented Feb 28, 2023

nixos-discourse commented Apr 21, 2023

nix flake show: don't evaluate derivations for foreign systems by default #6988

nix flake show: don't evaluate derivations for foreign systems by default #6988

Conversation

max-privatevoid commented Sep 1, 2022

thufschmitt left a comment

Choose a reason for hiding this comment

thufschmitt Sep 5, 2022

Choose a reason for hiding this comment

NobbZ Sep 5, 2022

Choose a reason for hiding this comment

edolstra Sep 6, 2022

Choose a reason for hiding this comment

NobbZ commented Sep 5, 2022 • edited Loading

NobbZ commented Sep 5, 2022

edolstra commented Sep 6, 2022

nixos-discourse commented Sep 6, 2022

yajo commented Sep 22, 2022

L-as commented Oct 26, 2022

yajo commented Oct 26, 2022

nixos-discourse commented Nov 30, 2022

domenkozar commented Dec 19, 2022

adrian-gierakowski commented Jan 26, 2023

thufschmitt commented Jan 27, 2023

thufschmitt commented Jan 27, 2023

max-privatevoid commented Jan 27, 2023

thufschmitt commented Jan 30, 2023

peterbecich commented Feb 6, 2023

nixos-discourse commented Feb 28, 2023

nixos-discourse commented Apr 21, 2023

NobbZ commented Sep 5, 2022 •

edited

Loading