Breakwater

Backtest / #1320

perf: add __slots__ to Requirement

pypa/packaging#1320 · merged 2026-07-07 · base 2b6726cec7 → head d5fedc90bb · ran on Docker

PASS

No behaviour divergence found by any probe.

† Merged after Nemotron 3 Ultra was released (2026-06-04): the model cannot have seen this PR or its fixes.

Known regressions in this PR

Performance · Requirement

No realistic input family measured slower.

Measured in backtest packaging-v2 (same commits, run bt-1320-81a5cc, image bw-pair:9728e5e1fd432d33). The semantic results above are from a later run; perf doesn't use the semantic planner, so it was not re-measured.

head slowerhead faster1×100input size (characters)F2F2: 1.00× at ~21 charsF2: 1.00× at ~24 charsF2: 0.99× at ~46 charsF2: 1.00× at ~138 charsF2: 1.01× at ~266 charsF2: 1.01× at ~266 charsF2: 1.01× at ~266 chars
  • F2
Head/base cachegrind instructions per call, by input size (log-log). Each line is one input family Nemotron proposed; points above 1× are where the "faster" code does more work.
FamilyInputsRatio range Slope base → headVerdict
F2Repeated attribute reads on many Requirement instances (slots make attribute lookup faster and reduce per‑object memory).claimed speedup · plan0.99× – 1.01×0.68 → 0.68neutral

Receipt

TargetBackendImageRun AuditAdjudication
packaging.requirements:Requirementlocal Dockerbw-pair:f8d0aa2a4ebfdaf6bt-1320-9c396e$0.11$0.000

Nemotron cost on Nebius Token Factory. Backend: where the untrusted code ran. local Docker (an image cited by its content-addressed tag). Every image is content-addressed (repo at each commit + harness), so every card re-runs from the image and seed 0. Raw data: audit.