anki: 24.06.3 -> 24.11

[cything]

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 25.05 Release Notes (or backporting 24.11 and 25.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

---

Add a 👍 reaction to pull requests you find important.

[Defelo]

tested on x86_64-linux

[martinetd]

passed on aarch64-linux but one of the rust tests failed on x86_64:

---- deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step stdout ----
thread 'deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step' panicked at rslib/src/scheduler/answering/mod.rs:557:61:
called `Result::unwrap()` on an `Err` value: InvalidInput { source: InvalidInputError { message: "bug: card modified without updating queue: id:1733439758805 card:1733439759 entry:1733439758", source: None, backtrace: None } }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
failures:
    deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step

There have been other flaky tests that had been disabled in the past, would need to investigate these..
(iirc upstream fixed them so we could re-enable them)

[cything]

That's weird cause it passes on my computer which is x86_64-linux too.

test deckconfig::update::test::should_increase_remaining_learning_steps_if_unpassed_learning_step_added ... ok
test dbcheck::test::tags ... ok
test deckconfig::update::test::should_increase_remaining_relearning_steps_if_unpassed_relearning_step_added ... ok
test deckconfig::update::test::should_keep_at_least_one_remaining_learning_step ... ok
test dbcheck::test::note_fields ... ok
test dbcheck::test::note_card_link ... ok

I also enabled the tests that were disabled for being flaky like you said and those pass too.

[martinetd]

I could get deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step to fail on my laptop by just running it in a loop (after ~1-2k tries; about 2 mins)

$ cargo test -vv should_keep_at_least_one_remaining_relearning_step
...
grab the very long line that actually ran something
$ i=0; while BUILDHASH='' CARGO=/home/asmadeus/.rustup/toolchains/1.82.0-x86_64-unknown-linux-gnu/bin/cargo CARGO_MANIFEST_DIR=/home/shared/anki/rslib CARGO_PKG_AUTHORS='Ankitects Pty Ltd and contributors <https://help.ankiweb.net>' CARGO_PKG_DESCRIPTION='Anki'\''s Rust library code' CARGO_PKG_HOMEPAGE='' CARGO_PKG_LICENSE=AGPL-3.0-or-later CARGO_PKG_LICENSE_FILE='' CARGO_PKG_NAME=anki CARGO_PKG_README=README.md CARGO_PKG_REPOSITORY='' CARGO_PKG_RUST_VERSION=1.80 CARGO_PKG_VERSION=0.0.0 CARGO_PKG_VERSION_MAJOR=0 CARGO_PKG_VERSION_MINOR=0 CARGO_PKG_VERSION_PATCH=0 CARGO_PKG_VERSION_PRE='' DESCRIPTORS_BIN=/home/shared/anki/out/rslib/proto/descriptors.bin LD_LIBRARY_PATH='/home/shared/anki/target/debug/build/blake3-9cf1804407af6b48/out:/home/shared/anki/target/debug/build/libsqlite3-sys-521a7856f28c3ca9/out:/home/shared/anki/target/debug/build/ring-1f5a96420d9a73dd/out:/home/shared/anki/target/debug/build/ring-de37fd4f38f3f6b1/out:/home/shared/anki/target/debug/build/zstd-sys-ceda26dcbb83d3fc/out:/home/shared/anki/target/debug/deps:/home/shared/anki/target/debug:/home/asmadeus/.rustup/toolchains/1.82.0-x86_64-unknown-linux-gnu/lib/rustlib/x86_64-unknown-linux-gnu/lib:/home/asmadeus/.rustup/toolchains/1.82.0-x86_64-unknown-linux-gnu/lib:/opt/wayland/lib64' MACOSX_DEPLOYMENT_TARGET=10.13.4 OUT_DIR=/home/shared/anki/target/debug/build/anki-913348601b716ed8/out PROTOC=/home/shared/anki/out/extracted/protoc/bin/protoc PYO3_NO_PYTHON=1 STRINGS_PY=/home/shared/anki/out/pylib/anki/_fluent.py STRINGS_TS=/home/shared/anki/out/ts/lib/generated/ftl.ts /home/shared/anki/target/debug/deps/anki-cb2bb8435ea2d543 should_keep_at_least_one_remaining_relearning_step; do i=$((i+1)); done; echo $i
....

---- deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step stdout ----
thread 'deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step' panicked at rslib/src/scheduler/answering/mod.rs:557:61:
called `Result::unwrap()` on an `Err` value: InvalidInput { source: InvalidInputError { message: "bug: card modified without updating queue: id:1733554619996 card:1733554620 entry:1733554619", source: None, backtrace: None } }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace


failures:
    deckconfig::update::test::should_keep_at_least_one_remaining_relearning_step

1759

If you have time reporting it like I did in ankitects/anki#3353 and just skipping it here is probably the way to go.

[cything]

Thanks! I was able to reproduce it exactly like you did. First, it failed after 1.6k tries and during some high CPU usage, it failed after 400 tires. I also created an issue upstream: ankitects/anki#3619

[martinetd]

Thank you!

[donovanglover]

LGTM. Grats on your first contribution to nixpkgs!

[cything]

Thank you so much!!

[dotlambda]

Please use fetchCargoVendor so we don't have to vendor the lock file.

[donovanglover]

Forgot about that but agree. For anyone that has time, feel free to make a PR updating anki to use the new fetchCargoVendor