[rajnisht7]

What this changes

test_the_javascript_side_agrees (tests/test_integer_by_value.py) ran node --version and node --test via subprocess.run(..., text=True) with no encoding= set. Without it, Python decodes the child's stdout/stderr using the ambient locale's preferred encoding. The CI "Portability" job runs the suite under LC_ALL=C PYTHONUTF8=0, which resolves that to plain ASCII, so any non-ASCII byte in Node's output (UTF-8) crashes the decode instead of letting the test assert pass/fail normally.

Type of change

• Editorial (typo, link fix, clarification: no normative effect)
• Non-breaking spec change (new optional field, new platform profile, informative addition)
• Breaking spec change (requires 14-day comment period and Project Lead sign-off)
• Schema change
• Example addition

Spec section

None

Tests for a fix

Checklist

• DCO sign-off on all commits (git commit -s)
• CHANGELOG.md updated (for any normative change)
• Breaking changes marked with <!-- CHANGED: #NNN: description --> in spec text
• Backward compatibility statement included (for breaking changes)

[lywinged]

Approved. Reviewed at 2dfb0fe on da69b31. The diagnosis is right: node --test defaults to the spec reporter from Node 23, ubuntu-latest is moving to the Ubuntu 26.04 image, which carries Node 24, and the spec reporter's ✔ and ℹ are what the ASCII codec of LC_ALL=C PYTHONUTF8=0 refuses. With Node 22 given --test-reporter=spec through NODE_OPTIONS here, this test is the one failure in the root suite on main, and it passes on this head, on 3.11 and on 3.12. The measurements are on #624.

#624, which dinakarjs opened first for the same failure, made the same change to the node --test call and is merged now. This PR merges cleanly on top, and what it adds is the node --version call, which prints an ASCII version string such as v24.21.0: with that encoding removed and the other kept, the test passes under the same conditions, so it changes no outcome, and it keeps the two calls alike.