active-repositories: Don't override fully-deprecated packages by erikd · Pull Request #11760 · haskell/cabal

erikd · 2026-04-21T05:17:25Z

This PR takes over #8997 from the original author Alexander Esgen who unfortunately passed away late last year.

The PR includes tests and additions to the documentation.

When active-repositories includes a repo with :override, and that repo's preferred-versions marks all its versions of a package as deprecated, the index-combining step previously still applied full override semantics, hiding all versions of that package from earlier repos.

Fix by consulting preferred-versions when combining indexes: if lookupDependency finds no preferred version for a package in the override repo, fall back to merge semantics so earlier-repo versions remain visible.

Two changes:

PackageIndex: add overrideOrMerge, a per-package Override/Merge strategy
IndexUtils: add deprecationAwareStrategy, wired into getSourcePackagesAtIndexState

Fixes #8502

This code was originally authored by Alexander Esgen and sumbitted in a PR over 2 years ago. Erik manually rebased that code onto master and used Claude to add tests and do some very minor refactoring to Alexander's code to make it more testable.

Please read Github PR Conventions and then fill in one of these two templates.

Include the following checklist in your PR:

Patches conform to the coding conventions.
Any changes that could be relevant to users have been recorded in the changelog.
Is the change significant? If so, remember to add significance: significant in the changelog file.
The documentation has been updated, if necessary.
Manual QA notes have been included.
Tests have been added. (Ask for help if you don’t know how to write them! Ask for an exemption if tests are too complex for too little coverage!)

erikd · 2026-04-21T06:30:50Z

Tagging @andreabedini because he had some comments on Alexanders original PR.

@Mikolaj In July last year you asked if Alexander's PR could be moved forward. As I stated at the top, I have taken this over after Alexander passed.

Mikolaj · 2026-04-21T09:22:20Z

@Mikolaj In July last year you asked if Alexander's PR could be moved forward. As I stated at the top, I have taken this over after Alexander passed.

Oh, I didn't now about Alexander. I'm honoured to help carry onward his work. Let me also mark @geekosaur, who reviewed the original PR. @erikd, please don't forget to set the needs-review label as soon as you'd like to widen the pool of reviewers.

Mikolaj · 2026-04-21T09:26:33Z

An initial technical question: does the test have the property that it fails without the PR and succeeds with it? I assume the tests from #11684 don't have this property and so they just check this PR doesn't break how active-repositories should work, right?

erikd · 2026-04-22T01:20:08Z

@Mikolaj If I back out some of the changes in cabal-install/src/Distribution/Client/IndexUtils.hs then 3 out of the 610 tests (running cabal test cabal-install:unit-tests) fail. They all passed before I backed out the changes.

Otherwise, yes, the tests in #11684 were just to test the "active-repositories" features so that these changes did not break that functionality.

Mikolaj · 2026-04-22T08:08:06Z

Good to know. Are the failing tests the newly added overrideOrMergeTests and/or deprecationAwareStrategyTests?

erikd · 2026-04-23T03:05:26Z

Inspecting the failures more closely, they are a) intermittent and b) not related to the changes in this PR. For some reason a couple of tests in the UnitTests.Distribution.Client.UserConfig section fail intermittently with things like:

  UnitTests.Distribution.Client.UserConfig
    nullDiffOnCreate:                                                                        OK (0.02s)
    canDetectDifference:                                                                      FAIL
      Exception: /home/erikd/Git/Haskell/cabal/cabal-install/tests/fixtures/project-root/
         test-user-config.tmp: withFile: resource busy (file is locked)
      
      HasCallStack backtrace:
        ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/FD.hs:331:17 in
          ghc-internal:GHC.Internal.IO.FD
      
      Use -p '/canDetectDifference/' to rerun this test only.

I did some further refactoring to expose addIndex so I could add unit tests specifically testing the Merge/Orverride/Skip logic.

There are still separate tests for addIndex and addStategy. Having separate (and cheap execution wise) tests is still worthwhile even though addIndex with an empty [Dependency] list behaves identically to applyStrategy for all three strategies.

Rebased against master as well.

philderbeast · 2026-04-23T10:00:19Z

Inspecting the failures more closely, they are a) intermittent and b) not related to the changes in this PR.

Yes, they're not related. These failures are reported in #11686 with a fix in flight with #11759.

erikd · 2026-05-07T07:54:34Z

The CI Bootstrap failure has nothing to do with my PR. The failure is in Python code and my PR does not touch any Python code.

Mikolaj · 2026-05-07T07:59:49Z

The CI Bootstrap failure has nothing to do with my PR. The failure is in Python code and my PR does not touch any Python code.

Right, just a transient network failure. Feel free to restart the failed jobs. Let me know if you can't.

erikd · 2026-05-08T03:18:52Z

Right, just a transient network failure. Feel free to restart the failed jobs. Let me know if you can't.

Pretty sure I already tried restarting it and it failed second time just like the first.

Restarted again 🤞 .

Mikolaj · 2026-05-08T08:40:16Z

You are right, the network failure wasn't transient at all. But I've restarted once more and it's gone, finally. :)

angerman

🫰

Mikolaj · 2026-05-20T07:50:39Z

@sebright: does it look to you like this impacts the solver, which would necessitate analyzing the consequences, or is it unrelated, in which case we could probably land it right now?

sebright · 2026-05-21T07:44:32Z

@Mikolaj I didn't get a chance to review this PR, but it sounds like it only affects the contents of the source package index that is passed to the solver. The solver should be able to handle any set of available packages and versions, so I don't think that this would have much effect on the rest of dependency solving. I'm also guessing that the removal of preferred versions from Hackage (haskell/hackage-server#1498) simplifies this change.

erikd · 2026-05-22T06:20:05Z

@sebright Since you were requested as a reviewer, in order for this to be merged, you either need to review this or remove yourself as a reviewer. Thanks!

sebright · 2026-05-22T07:51:37Z

@erikd Is the review request blocking Mergify? I had thought that PRs only needed two approvals, without requested changes.

Mikolaj · 2026-05-22T08:46:30Z

I'm satisfied by the non-review. :) Thank you @sebright!

When active-repositories includes a repo with :override, and that repo's preferred-versions marks all its versions of a package as deprecated, the index-combining step previously still applied full override semantics, hiding all versions of that package from earlier repos. Fix by consulting preferred-versions when combining indexes: if lookupDependency finds no preferred version for a package in the override repo, fall back to merge semantics so earlier-repo versions remain visible. Two changes: - PackageIndex: add overrideOrMerge, a per-package Override/Merge strategy - IndexUtils: add deprecationAwareStrategy, wired into getSourcePackagesAtIndexState Fixes haskell#8502 This code was originally authored by Alexander Esgen and sumbitted in a PR over 2 years ago. Erik manually rebased that code onto master and used Claude to add tests and do some very minor refactoring to Alexander's code to make it more testable. Co-Authored-By: Erik de Castro Lopo <erikd@mega-nerd.com> Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

erikd · 2026-05-23T01:21:54Z

@Mergifyio queue

mergify · 2026-05-23T01:22:38Z

Mikolaj · 2026-05-23T07:35:56Z

@erikd: BTW, we normally merge by setting a label to give people some more time to react (but in this case there was ample time, so no problem). Please see CONTRIBUTING.md.

coot · 2026-05-27T07:51:21Z

Have we bumped the cabal-version? This will allow one to require in a *.cabal file that only cabal that supports this change can be used to build a package - a handy feature to make an ecosystem safe.

ulysses4ever · 2026-05-27T15:58:08Z

This is not a part of the cabal format, and I don't think we want a precedent of bumping the format for unrelated changes in the tool.

erikd force-pushed the erikd/dont-override-if-all-deprecated branch 3 times, most recently from 38c15e5 to c257e5a Compare April 21, 2026 05:28

erikd mentioned this pull request Apr 21, 2026

Don't override a package if all versions are deprecated #8997

Closed

8 tasks

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from c257e5a to b5a6079 Compare April 21, 2026 06:04

erikd added the attention: needs-review label Apr 21, 2026

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from b5a6079 to 02f4672 Compare April 23, 2026 03:05

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from 02f4672 to 71dde16 Compare April 23, 2026 09:31

sebright added the cabal-install: solver label Apr 25, 2026

erikd force-pushed the erikd/dont-override-if-all-deprecated branch 3 times, most recently from 689e85c to 2924f25 Compare April 30, 2026 05:23

erikd force-pushed the erikd/dont-override-if-all-deprecated branch 2 times, most recently from 14f76fa to 2379e8c Compare May 7, 2026 06:16

erikd force-pushed the erikd/dont-override-if-all-deprecated branch 3 times, most recently from 5268e09 to a41c69f Compare May 12, 2026 03:12

Mikolaj requested a review from sebright May 14, 2026 17:29

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from a41c69f to 8117942 Compare May 15, 2026 05:49

erikd force-pushed the erikd/dont-override-if-all-deprecated branch 4 times, most recently from 7f044d4 to a621875 Compare May 17, 2026 23:19

andreabedini approved these changes May 20, 2026

View reviewed changes

angerman approved these changes May 20, 2026

View reviewed changes

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from a621875 to 5f04334 Compare May 22, 2026 06:17

Mikolaj removed the request for review from sebright May 22, 2026 08:46

erikd force-pushed the erikd/dont-override-if-all-deprecated branch from 5f04334 to 3a31161 Compare May 22, 2026 23:02

mergify Bot added the queued label May 23, 2026

mergify Bot merged commit 6989d21 into haskell:master May 23, 2026
100 of 102 checks passed

mergify Bot removed the queued label May 23, 2026

Mikolaj removed the attention: needs-review label May 23, 2026

erikd deleted the erikd/dont-override-if-all-deprecated branch May 24, 2026 23:31

Conversation

erikd commented Apr 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

erikd commented Apr 21, 2026

Uh oh!

Mikolaj commented Apr 21, 2026

Uh oh!

Mikolaj commented Apr 21, 2026

Uh oh!

erikd commented Apr 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Mikolaj commented Apr 22, 2026

Uh oh!

erikd commented Apr 23, 2026

Uh oh!

philderbeast commented Apr 23, 2026

Uh oh!

erikd commented May 7, 2026

Uh oh!

Mikolaj commented May 7, 2026

Uh oh!

erikd commented May 8, 2026

Uh oh!

Mikolaj commented May 8, 2026

Uh oh!

angerman left a comment

Choose a reason for hiding this comment

Uh oh!

Mikolaj commented May 20, 2026

Uh oh!

sebright commented May 21, 2026

Uh oh!

erikd commented May 22, 2026

Uh oh!

sebright commented May 22, 2026

Uh oh!

Mikolaj commented May 22, 2026

Uh oh!

erikd commented May 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mergify Bot commented May 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merge Queue Status

Uh oh!

Uh oh!

Mikolaj commented May 23, 2026

Uh oh!

coot commented May 27, 2026

Uh oh!

ulysses4ever commented May 27, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

erikd commented Apr 21, 2026 •

edited

Loading

erikd commented Apr 22, 2026 •

edited

Loading

erikd commented May 23, 2026 •

edited

Loading

mergify Bot commented May 23, 2026 •

edited

Loading