From fbecd0c858cd1e73aabbea17db630df5e90c170c Mon Sep 17 00:00:00 2001 From: tannevaled Date: Fri, 30 Oct 2026 22:05:47 +0100 Subject: [PATCH] =?UTF-8?q?ci:=20pin=20the=20ratchet's=20runner=20image=20?= =?UTF-8?q?=E2=80=94=20a=20gate's=20environment=20is=20not=20a=20moving=20?= =?UTF-8?q?label?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Closes #838. main has failed its own conformance gate since 2026-10-27, and nothing in the repository changed. Four pull requests are blocked behind it -- #840 (a peer's #833 fix), #836, #839 and #792 -- none of them the cause. THE CONTROL. Run 40355659600, commit e54df08, re-run 28 hours later: attempt 1, 2026-10-26, Image: ubuntu-24.04 -> success attempt 2, 2026-10-27, Image: ubuntu-26.04 -> FAILURE core/process/setrlimit_spec.rb: 28 passing, baseline 29 (-1) Same commit. Same go-version: 1.27.3. Same SPEC_SHA. `runs-on: ubuntu-latest` rolled between the two attempts, and that is the only variable. Read from each attempt's own `Runner Image / Image:` line, so the cause is measured rather than hypothesised. It reproduces across three runs, two of them on an unrelated branch. That the file is `setrlimit` is not a coincidence: it measures resource limits, among the first things to differ between two Ubuntu majors. `kernel/require` moved +1 in the same runs -- one of the five the BASELINE header already records as environment-sensitive. WHY THIS LANE AND NOT THE OTHER THIRTEEN. BASELINE pins everything else about the measurement: the corpus SHA, a real path rather than a symlink, TIMEOUT at 60s, an isolated snapshot, three sweeps agreeing on all 2,206 records. Every variable but the operating system, which was named `latest`. A gate whose environment is a moving label is a reading, not a gate -- and the failure is the quiet kind, because nobody re-runs a green lane: history says fine, today says otherwise, and the next PR to touch anything is told it caused a regression it did not cause. The other lanes keep `ubuntu-latest` deliberately. They build and test, where an image roll is news worth having rather than a baseline to protect. MOVING TO 26.04 STAYS POSSIBLE, as a deliberate change with its own measurement and its own BASELINE delta, like a SPEC_SHA bump. What must not happen is re-recording blind: the same run also showed kernel/require +1, equally environmental, and UPDATE_BASELINE=1 would bank it as progress. actionlint clean. One line of configuration; the rest is why. Co-Authored-By: Claude Opus 5 --- .github/workflows/rubyspec-ratchet.yml | 26 +++++++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/.github/workflows/rubyspec-ratchet.yml b/.github/workflows/rubyspec-ratchet.yml index e2c76648..45e4283f 100644 --- a/.github/workflows/rubyspec-ratchet.yml +++ b/.github/workflows/rubyspec-ratchet.yml @@ -18,7 +18,31 @@ permissions: jobs: ratchet: name: ruby/spec ratchet - runs-on: ubuntu-latest + # ⛔ PINNED, and this is the one lane in this repository that must be. + # BASELINE pins everything else about the measurement -- the corpus SHA, a + # real path rather than a symlink, TIMEOUT raised to 60s, an isolated + # snapshot, three agreeing sweeps -- and then `ubuntu-latest` re-pointed + # without notice and the gate changed its answer. + # + # Measured (#838): run 40355659600 at commit e54df08, re-run 28 hours later. + # attempt 1, 2026-10-26, Image: ubuntu-24.04 -> success + # attempt 2, 2026-10-27, Image: ubuntu-26.04 -> FAILURE + # core/process/setrlimit_spec.rb: 28 passing, baseline 29 + # Same commit, same go-version, same SPEC_SHA. `setrlimit` measures + # resource limits, which is among the first things to differ between two + # Ubuntu majors. Four pull requests sat blocked behind it, none of them the + # cause. + # + # A gate whose environment is a moving label is a reading, not a gate; and + # the failure is the quiet kind, because nobody re-runs a green lane. + # Moving to 26.04 stays possible -- as a deliberate change with its own + # measurement and its own BASELINE delta, like a SPEC_SHA bump. Re-recording + # blind is what must not happen: the same run also showed kernel/require +1, + # equally environmental, and `UPDATE_BASELINE=1` would bank it as progress. + # + # The other lanes keep `ubuntu-latest` deliberately: they build and test, + # where an image roll is news worth having, not a baseline to protect. + runs-on: ubuntu-24.04 steps: - uses: actions/checkout@v7 - uses: actions/setup-go@v7