Skip to content

spec(t27os): the t27b leg runs after a red t27c gate, so an @extern spec gets its compiler (Closes #12332) - #12333

Merged
gHashTag merged 10 commits into
masterfrom
night/170-t27b-leg-always
Oct 26, 2026
Merged

gHashTag merged 10 commits into
masterfrom
night/170-t27b-leg-always

Conversation

@gHashTag

Copy link
Copy Markdown
Collaborator

Closes #12332. Refs Trinity-S-AI-DNA/t27os#3.

Stacked on #12285 (base commit c3ebc76). The diff shows #12285's commit until it merges; this PR's own change is d3c39fc.

Lab run 40170146414 (PR #12284): glass_window.t27 is blocked on t27c, the verdict gate exits 1, and the t27b job (needs: run with no if:) is skipped. So line_verdict's @extern excuse can never be reached.

  • T27B_ALWAYS: if: ${{ always() && needs.run.outputs.resolved != '' }}, placed in the T27B_JOB_YAML header between needs and runs-on.
  • runs_after_red_gate(y) and the test t27b_job_runs_after_a_red_t27c_gate, whose negative control is the deployed header.

Not verified yet. The Mac load hook refuses both compilers here. The lab pair (d3c39fc, specs/t27os/lab_legs.t27) is queued behind lab run 40170786895. This PR stays a draft until the receipt shows both legs green. After the merge, TRI CLAW copies the header line into t27os lab.yml line 560.

🤖 Generated with Claude Code

gHashTag and others added 3 commits October 25, 2026 12:18
… t27b's last 20 lines in the log (Refs #12219; Refs Trinity-S-AI-DNA/t27os#3)

Lab run 40163543896: specs/t27os/hdmi_grab_main.t27 passed 2/2 on t27c and got
tests=NA on t27b, while the leg had sent t27b's output to /tmp/b.txt only.
T27B_NA_TAIL is the law line; the receipt line keeps its key=value shape.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…pec gets its compiler (Closes #12332)

Lab run 40170146414 (PR #12284): glass_window.t27 blocked on t27c, the
verdict gate exited 1 and the t27b job (needs: run alone) was skipped, so
line_verdict's @extern excuse was unreachable. T27B_JOB_YAML now carries
if: always() && resolved != ''; new test with the deployed header as the
negative control. Stacked on #12285. Refs Trinity-S-AI-DNA/t27os#3

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-25 07:44:40 UTC

Summary

Status Count
Total Open PRs 50
PRs with Failing Checks 25
PRs with All Checks Green 25
READY 16
FAILING 25
PENDING 0
NO CHECKS YET 0

These columns do not partition: 16 + 25 + 0 + 0 = 41, and there are 50 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=55c0f1b1b32d != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

@gHashTag

Copy link
Copy Markdown
Collaborator Author

Night tick 173: #12333 was DIRTY against master (#12285 rewrote the t27b job text in specs/t27os/lab_legs.t27). Merged master in 143b062, keeping this branch's hunks: master's T27B_JOB_YAML is byte-identical to this branch's apart from the if: ${{ always() && ... }} line. Both hunks are additive: T27B_ALWAYS, runs_after_red_gate and the test t27b_job_runs_after_a_red_t27c_gate.

Not yet verified: the Mac load hook refuses a local compiler run, so the lab has to check it. Lab run 40173880388 (#12294) is still queued, and only one run may be in flight. This head is queued in pending_dispatch in place of d3c39fc, and the PR stays a draft until its receipt shows both legs green.

@gHashTag

Copy link
Copy Markdown
Collaborator Author

Night tick 175: the emit-bitexact red here (run 40174626397, specs/git/diff.t27 refused by typecheck) is inherited from master eada185 (#4634, a Rust match). Repair: #12350. This PR's own spec is not involved.

@github-actions

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-25 08:21:03 UTC

Summary

Status Count
Total Open PRs 50
PRs with Failing Checks 23
PRs with All Checks Green 27
READY 19
FAILING 23
PENDING 0
NO CHECKS YET 0

These columns do not partition: 19 + 23 + 0 + 0 = 42, and there are 50 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=55c0f1b1b32d != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

This was referenced Oct 25, 2026
@github-actions

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-26 01:54:27 UTC

Summary

Status Count
Total Open PRs 39
PRs with Failing Checks 31
PRs with All Checks Green 8
READY 2
FAILING 31
PENDING 0
NO CHECKS YET 0

These columns do not partition: 2 + 31 + 0 + 0 = 33, and there are 39 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=75249fe1efad != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

…12332)

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
@gHashTag

Copy link
Copy Markdown
Collaborator Author

Auto-merge enabled on explicit owner instruction (no-auto-merge overridden by the owner in the OpenCode agent chat, 2026-10-26). No --admin used.

@github-actions

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-26 02:05:06 UTC

Summary

Status Count
Total Open PRs 35
PRs with Failing Checks 19
PRs with All Checks Green 16
READY 2
FAILING 19
PENDING 0
NO CHECKS YET 0

These columns do not partition: 2 + 19 + 0 + 0 = 21, and there are 35 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=8ddf64a0178e != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

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

Labels

no-auto-merge Opt out of the issue gate's default squash auto-merge (issue-gate.yml)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

lab: the t27b leg is skipped whenever the t27c gate is red, so an @extern spec never gets its compiler

1 participant