Track guessed as N from the FAILING STEP — line 1 of 4, ./compiler/pascal26 test/test_nilpy_float_repr_roundtrip.npy /tmp/test_nilpy_float_repr26, which names test/test_nilpy_float_repr_roundtrip.npy. Not from the job's name or its src: those describe what the job is ABOUT, and this job's recipe spans 2 source file(s). The ranker reads frontmatter, so this line — not the body — decides who works it; correct it if the guess is wrong.

origin/master has advanced 4 commit(s) since this sha. Re-verify at current HEAD before acting — the callback is tagged to the sha that was tested, which may no longer be the state of the tree.

regression: test-nilpy#src:test/test_nilpy_float_repr_roundtrip.npy at 3daf4bc16cc1 in step 1/4, ./compiler/pascal26 test/test_nilpy_float_repr_roundtrip.npy /tmp/test_nilpy_float_repr26 (auto-filed by twatch)

• Type: regression (auto-filed by Track T watcher, host borg, twatch 1f06bcf02d3f). Untriaged.
• Found: 2026-09-22T10:58:52Z
• Test source: test/test_nilpy_float_repr_roundtrip.npy tools/expect_same.sh
• Failing step: line 1 of 4 of the job's recipe; it names test/test_nilpy_float_repr_roundtrip.npy.

  ./compiler/pascal26 test/test_nilpy_float_repr_roundtrip.npy /tmp/test_nilpy_float_repr26
  
  

Repro

tools/testmgr.py --tier full --job 'test-nilpy#src:test/test_nilpy_float_repr_roundtrip.npy' at 3daf4bc16cc15bc3a670f4115ef40ad6cbb27c25

Range

The named sha 3daf4bc16cc1 CANNOT be the cause — it touches no buildable file (docs / tickets / tstate only). It is the sha that was TESTED, i.e. the upper bound of an untested range; the cause is somewhere below it.

bad 3daf4bc16cc1, last good 5fccc890af91, 4 commit(s) in range — the watcher narrows this by idle bisect; check tstate/TSTATE.md for the current range.

Log tail

pascal26: error: a write to the output file stored fewer bytes than asked: /tmp/testmgr-scratch-1141223/test_nilpy_float_repr26
(tail)
pascal26: error: a write to the output file stored fewer bytes than asked: /tmp/testmgr-scratch-1141223/test_nilpy_float_repr26
  check all four -- the first is the commonest and the last is the one that fools people:
    df -h <dir>   free BYTES
    df -i <dir>   free INODES -- can hit 100% while df -h reads 9%
    ulimit -f     a file-size limit truncates at a plausible size
    another pascal26 writing THIS SAME PATH -- two writers interleave,
      and the file can then end up the RIGHT size, so its size proves nothing.

Stub ticket: signal only. Track T agent (face 2) enriches or a dev track takes it from the repro line.

Log

• 2026-09-22 — auto-closed by the borg watcher: test-nilpy#src:test/test_nilpy_float_repr_roundtrip.npy passes at c194231297b1 (tier full); it was red at 3daf4bc16cc1. Reopening is by a fresh NEW-RED stub, since a second red is a second finding with its own range.