Skip to content

📝 docs(msrv): 1.95 is a promise above the graph's 1.90 floor (#722) - #728

Merged
kbrdn1 merged 4 commits into
devfrom
chore/#722-msrv-floor
Nov 12, 2026
Merged

kbrdn1 merged 4 commits into
devfrom
chore/#722-msrv-floor

Conversation

@kbrdn1

@kbrdn1 kbrdn1 commented Nov 12, 2026 •

Copy link
Copy Markdown
Owner

Description

rust-version = "1.95" was justified by libsqlite3-sys 0.38.1, which declared no floor and used the std cfg_select! (stable since 1.95.0). That reason was already stale from #585 on: libsqlite3-sys 0.38.2 defines its own cfg_select! with macro_rules! in build.rs (checked on the .crate files: 0.38.1 calls the std macro at build.rs:110/643, 0.38.2 and 0.39.1 define it at build.rs:11). #720 (#723) then brought rusqlite 0.41.1 / libsqlite3-sys 0.39.1, which also declares rust-version = "1.88.0". Nothing failed on the drift because tests/msrv_tests.rs pins the MSRV claims, not the rationale.

This PR re-measures the floor by compiling, keeps the declared floor at 1.95, and corrects every explanation of it.

Closes #722

Measurement

Against the lockfile of 8e2eed25 (dev at branch time), on aarch64-apple-darwin, RUSTFLAGS="" like the msrv job, with rust-version temporarily set to the toolchain under test (otherwise cargo's gate fires on gwm-cli's own 1.95 before reaching the graph), each toolchain installed with --profile minimal and a throwaway CARGO_TARGET_DIR:

toolchain command exit
1.90.0 cargo +1.90 check --all-targets --locked 0
1.90.0 cargo +1.90 check --all-targets --locked --no-default-features 0
1.89.0 cargo +1.89 check --all-targets --locked 101
1.89.0 cargo +1.89 check --all-targets --locked --no-default-features 101
1.89.0 cargo +1.89 check --locked (lib + bins, what cargo install builds) 0
1.89.0 cargo +1.89 check --locked --no-default-features 0
1.89.0 cargo +1.89 check --all-targets --locked --ignore-rust-version (default features only) 0

The 1.89 failure, verbatim, identical for both feature sets:

error: rustc 1.89.0 is not supported by the following package:
  ordered-float@5.5.0 requires rustc 1.90

So the graph's floor under the msrv job's definition is 1.90, set only by ordered-float 5.5.0's declared rust-version. It is reached through saphyr 0.3.3, a [dev-dependencies] entry (cargo tree -i ordered-float@5.5.0 -e all), and the code compiles at 1.89 once the gate is lifted: no language or std feature needs 1.90. Highest declared floors in the resolve (cargo metadata --filter-platform) are the same on darwin, linux-gnu and windows-msvc: ordered-float 5.5.0 at 1.90, then rusqlite 0.41.1 and serde-saphyr 2.1.1 at 1.89. Linux and Windows were not compiled at 1.90 locally; they are bounded below by that platform-independent declaration and above by the green 1.95 job.

Decision: keep 1.95

docs/6.development/3.stability.md §MSRV policy has no rule for lowering: it defines rust-version as the floor the crate is expected to compile against (1.95 still is, the msrv job holds it) and only ever describes bumps. The issue's own bar decides: lowering is a support promise, not a cleanup. A 1.90 promise would rest on a test-only dependency cargo install never builds, and the next saphyr / ordered-float release raising its floor would become an MSRV change (minor release + changelog) under that policy. So the floor stays, and the prose records the gap between the declared floor and what the graph needs as a dated decision, without turning it into a policy rule.

Type of change

  • ✨ Feature (new functionality)
  • 🐛 Fix (bug fix)
  • 🚑️ Hotfix (critical production fix)
  • ♻️ Refactor (no behaviour change)
  • 📝 Docs
  • ✅ Tests
  • ⚡ Perf
  • 🔧 Chore / CI / build

Changes

Tests

  • cargo test passes locally (GWM_HISTORY_FILE pointed at a temp file): full suite on 9249cca8, 3589 passed, 0 failed, 4 ignored (all pre-existing); after the review fixes, the 17 test binaries that read the edited files re-ran on fff1845a: 527 passed, 0 failed. CI on fff1845a: 15/15, the three msrv (declared floor, …) jobs included
  • cargo fmt --check passes
  • cargo clippy -- -D warnings: not run, no Rust source touched
  • New tests added under tests/: none. Exception argued: comments and docs only, no behaviour change. tests/msrv_tests.rs keeps pinning every live MSRV claim to Cargo.toml (all LIVE_CLAIMS substrings are byte-identical) and the msrv job still compiles the declared 1.95 on all three runners.

Checklist

  • Branch follows <type>/#<issue>-<description>
  • Commits follow Gitmoji + Conventional Commits (see CONTRIBUTING.md)
  • CHANGELOG.md updated under ## [Unreleased]
  • README updated if CLI surface changed: no surface change
  • examples/gwm.toml.example updated if config schema changed: no schema change
  • No unwrap() on user-facing paths (return GwmError instead)
  • No println! in TUI render code (status bar only)

Linked issues / docs

Notes for reviewers

  • rust-version itself does not move, so the msrv jobs still install 1.95: green there is the same check as on dev, not a test of the new prose.
  • Not done: a guard pinning the rationale to the lockfile. It would have to re-implement cargo's gate or fire on every patch bump of the named crate; a follow-up issue if wanted.

The rationale for `rust-version = "1.95"` named `libsqlite3-sys` 0.38 and
its use of the std `cfg_select!`. Since #720, `rusqlite` 0.41.1 brings
`libsqlite3-sys` 0.39.1, which declares 1.88.0 and defines its own
`cfg_select!`, so that reason no longer holds.

Re-measured against the lockfile of 8e2eed2 on aarch64-apple-darwin,
with `rust-version` set to the toolchain under test:
`cargo +1.89 check --all-targets --locked` stops at cargo's gate on
`ordered-float@5.5.0 requires rustc 1.90`, and `cargo +1.90` passes with
and without default features. `ordered-float` comes in only through the
`saphyr` dev-dependency, and 1.89 compiles every target once the gate is
lifted, so the 1.90 is a declaration, not a language or std need.

The declared floor stays at 1.95: lowering it would promise 1.90 to users
on the strength of a test-only dependency. Cargo.toml, the stability
policy (EN, FR), the msrv job comment, ROADMAP and the roadmap pages now
say so; CHANGELOG gains a Docs entry.
@kbrdn1 kbrdn1 added the chore Repo maintenance label Nov 12, 2026
@coderabbitai

coderabbitai Bot commented Nov 12, 2026 •

Copy link
Copy Markdown
Pre-merge checks: Passed 1 · Inconclusive 1

Failed Failed checks: 1 inconclusive

Check name Status Explanation Resolution
Title check Pending This assessment has not completed. The check will be retried on the next pre-merge check run.
Passed Passed checks: 1 passed

Check name Status Explanation
Description Check Passed Check skipped - CodeRabbit’s high-level summary is enabled.
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

…the msrv job (#722)

Review of #728. `libsqlite3-sys` 0.38.2, which #585 brought in, already
defined its own `cfg_select!` (0.38.1, the version #491 measured, used
the std one), so the 1.95 rationale was stale from #585 on and #720 only
added the declared 1.88.0. The msrv job comment, one of the three places
#722 lists, now says what sets the graph's floor and points to Cargo.toml
for the measurement. The measurement is qualified as macOS-only and the
gate-lifted 1.89 run as default features only; CONTRIBUTING gets the
same "kept since #722" as ROADMAP. The stability policy no longer gains a
rule for lowering the floor: the PR corrects its explanation, it does not
change the policy.
)

Review of #728. The stability policy's MSRV section gained a rule ("may
sit above what the graph needs, never below it") and Cargo.toml a
procedure for lowering the floor; neither was asked for, and the PR body
said the opposite. The section now keeps its original wording and only
records the measurement and the decision, dated to #722. The changelog
says the rules are unchanged. The msrv job comment qualifies the graph's
floor as measured on macOS only; CONTRIBUTING and ROADMAP name 0.38.1,
the version that actually raised the floor.
…asurement (#722)

Review nits on #728: ROADMAP said the graph needs less than 1.95 without
the macOS qualifier every other surface carries, and "since #722" in the
msrv job and Cargo.toml dated to the measuring PR a gap that #585 and
#720 opened.
@kbrdn1
kbrdn1 merged commit be8550b into dev Nov 12, 2026
15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

chore Repo maintenance

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant