Skip to content

chore(release): v0.36.0 — every member is a signal that lied - #894

Open
avrabe wants to merge 1 commit into
mainfrom
release/v0.36.0
Open

chore(release): v0.36.0 — every member is a signal that lied#894
avrabe wants to merge 1 commit into
mainfrom
release/v0.36.0

Conversation

@avrabe

@avrabe avrabe commented Sep 5, 2026

Copy link
Copy Markdown
Contributor

Version bump + CHANGELOG for v0.36.0. Scope taken from the
release: v0.36.0 query, not from prose.

v0.36.0:  10 artifacts, 10 verified, 0 proposed

The theme

Every member is a check that ran, reported success or a confident number,
and measured less than its name claimed:

a gate satisfied by the empty stub it exists to catch REQ-306 / REQ-295
a validator that cries wolf and cannot fail REQ-322
a parser that drops documents in silence REQ-321
markers invisible in shell, so CI evidence cannot count REQ-319
tests that resolved a binary by luck, invisibly to CI REQ-314
a trailer that links but never transitions REQ-315
an unmodelled row indistinguishable from a forgotten one REQ-320
CI alerts that name the wrong cause REQ-316 / REQ-317

Six were reported by people using rivet (#807, #862, #870, #871, #876), three
the tooling found on itself, and one came from a fuzz run on a six-byte
input
.

Notes reconciled mechanically

Not by re-reading — release notes drifted from scope twice in this project's
history, so it's diffed:

IN SCOPE, NOT IN NOTES:  none
IN NOTES, NOT IN SCOPE:  REQ-308, REQ-323  (deliberate cross-refs, explained in context)

Known residuals — stated, not omitted

  • alternate-backlinks link types are still unvalidated — a check that does not
    fire, rather than one that fires wrongly.
  • verification-evidence matching is tighter but still not full
    command-reachability.
  • The fleet-restart root cause was found by the infra maintainer (needrestart
    under unattended-upgrades) and the fix is deployed, but suppression of a real
    restart is not yet demonstrated — REQ-323.

Also files REQ-324

The mutation gate reports mutants in #[cfg(kani)] code that no test it runs
can ever kill
— found on #893 while shipping REQ-320. That's this release's own
theme arriving in the tooling, so it's recorded rather than quietly ignored.

Gates

cargo fmt --check · clippy --all-targets -D warnings on 1.97.0 ·
cargo test --workspace (exit 0, 66 ok) · rivet validate · rivet docs check
· yamllint — all exit 0.

docs check caught a forward reference to the next version in my own draft; it's
rephrased. Version is 0.36.0 in Cargo.toml, Cargo.lock and
vscode-rivet/package.json.

Not tagging until this is merged and main's CI is verified green.

Ten artifacts, all verified, scope taken from the `release: v0.36.0` query
rather than from prose. Six were reported by people using rivet (#807, #862,
#870, #871, #876), three the tooling found on itself, and one came from a fuzz
run on a six-byte input.

The theme is not "bugs". Every member is a check that RAN, reported success or a
confident number, and measured less than its name claimed:

  a gate satisfied by the empty stub it exists to catch      REQ-306 / REQ-295
  a validator that cries wolf and cannot fail the process    REQ-322
  a parser that drops documents in silence                   REQ-321
  markers invisible in shell, so CI evidence cannot count    REQ-319
  tests that resolved a binary by luck, invisibly to CI      REQ-314
  a trailer that links but never transitions, so readiness   REQ-315
    reports shipped scope as unshipped
  an unmodelled row indistinguishable from a forgotten one   REQ-320
  CI alerts that name the wrong cause                        REQ-316 / REQ-317

Notes reconciled against the query mechanically, not by re-reading: all ten
scoped ids appear, and the only two extras (REQ-308, REQ-323) are deliberate
cross-references explained in context.

Also files REQ-324 for the next release: the mutation gate reports mutants in
`#[cfg(kani)]` code that no test it runs can ever kill, found while shipping
REQ-320. That is this release's own theme arriving in the tooling, so it is
recorded rather than quietly ignored.

Known residuals are stated in the entry rather than omitted: alternate-backlinks
are still unvalidated, verification-evidence matching is tighter but still not
command-reachability, and the fleet-restart fix is deployed but its suppression
is not yet demonstrated.

Confirmed with cargo fmt --check, clippy --all-targets -D warnings on 1.97.0,
cargo test --workspace (exit 0, 66 ok, 0 failed), rivet validate, rivet docs
check, yamllint on the CI paths — all exit 0. Version is 0.36.0 in Cargo.toml,
Cargo.lock and vscode-rivet/package.json.

Refs: REQ-324
Trace: skip
@github-actions

github-actions Bot commented Sep 5, 2026

Copy link
Copy Markdown

📐 Rivet artifact delta

Change Count
Added 1
Removed 0
Modified 0
Downstream impacted (depth ≤ 5) 0

Graph

graph LR
  REQ_324["REQ-324"]:::added
  classDef added fill:#d4edda,stroke:#28a745,color:#155724
  classDef removed fill:#f8d7da,stroke:#dc3545,color:#721c24
  classDef modified fill:#fff3cd,stroke:#ffc107,color:#856404
  classDef overflow fill:#e2e3e5,stroke:#6c757d,color:#495057,stroke-dasharray: 3 3
Loading
Added
  • REQ-324

📎 Full HTML dashboard attached as workflow artifact rivet-delta-pr-894download from the workflow run.

Posted by rivet-delta workflow. The graph shows only changed artifacts; open the HTML dashboard (above) for full context.

@codecov

codecov Bot commented Sep 5, 2026

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant