1.7 KiB
E/F wrapped-label experiment evidence
Neither profile is promoted. Start with validation and limitations, then the locked E and F readings. Source gates and artifact conservation passed; dependable first-reading ownership did not.
Run from the repository root with Python 3, without launching the application:
python harness/experiments/wrapped-label-20261001/evidence/verify.py
python harness/experiments/wrapped-label-20261001/evidence/verify.py --index
python harness/experiments/wrapped-label-20261001/evidence/verify.py --source
Default verification uses only included packet bytes. --index additionally
requires this packet staged in Git; --source reads the pinned d089 source
objects from Git. The helper checks the inventory, pristine E/F SVG conservation,
full exported Project delta, reopened-F logical coordinates and process receipts.
It does not render SVG or replay native interactions. Original checkers and
launchers are preserved byte-for-byte under evidence/scripts;
some retain historical local paths, so use the wrapper for portable offline checks.
The earlier independent review predates fresh reopen tests. The later saved-F review covers their model/export files; parent observations describe the actual native events and limitations. Profile selection is external and not persisted. The ordinary Baseline reopen showed one line; it is not a claim of ordinary wrapping parity or a source-save round trip.