nix flake lock
does not update transitive inputs to the upstream transitive inputs, even if the input itself changed #465
Labels
No labels
Area/build-packaging
Area/cli
Area/evaluator
Area/fetching
Area/flakes
Area/language
Area/profiles
Area/protocol
Area/releng
Area/remote-builds
Area/repl
Area/store
bug
crash 💥
Cross Compilation
devx
docs
Downstream Dependents
E/easy
E/hard
E/help wanted
E/reproducible
E/requires rearchitecture
imported
Needs Langver
OS/Linux
OS/macOS
performance
regression
release-blocker
RFD
stability
Status
blocked
Status
invalid
Status
postponed
Status
wontfix
testing
testing/flakey
ux
No milestone
No project
No assignees
2 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: lix-project/lix#465
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Related to #8.
Changing a flake input URL to a newer version and then running
nix flake lock
incorrectly does not update any of the inputs of that input, even if there is nothing holding them at their current version.Initial state: URL is set to the one for Lix 2.90.0 as follows:
lock1.json (repro gist commit
a8b4606
), initial stateflake.nix
:flake.lock
Technically even here the lock file is out of date because of this bug previously affecting the Lix installer. But don't mind that, it would happen even if it were not broken that way.
lock2.json (repro gist commit
54af1e0
), clearly broken stateThen we update the
2.90.0
in the URL above to2.91.0
and runnix flake lock
.This commit describes the broken state.
lock3.json (repro gist commit
35a7298
)From this broken state, we can update the inputs of
lix
by runningnix flake update lix
:Observe that it updates the inputs of Lix in the lock file even though it does not update lix itself.
repro gist
You can
git clone https://gist.github.com/lf-/dc9386df57c20df7561ea475664791b7
to get the full reproduction above as git history.Expected behaviour
I would expect that updating a flake URL would cause the input to be re-fetched and its inputs updated to the latest ones. After all, to me, this state semantically means that the input is using the upstream inputs.
Version
nix (Lix, like Nix) 2.91.0-dev-pre20240807-529eed7
This issue was mentioned on Gerrit on the following CLs: