What — Review ElastNetMT's adoption of MolSysSuite Python support libraries and developer tools under uibcdf/molsyssuite#56. The central inventory still marks both reviews pending.
How — Record the four runtime integration boundaries, fix the import-time PyUnitWizard policy override, and retain the LinDelInt auto-engine failure as a provider dependency under uibcdf/lindelint#8. Pin a published Pytest Receptor release in every hosted test environment and select its ci profile in the maintained pytest command. Keep adoption states separate.
Why — Source 6705363187762a69048d147853f1018097c76a5e declares ArgDigest, DepDigest, SMonitor, and PyUnitWizard and uses them in model paths. _pyunitwizard.py unconditionally sets process-wide defaults. The latest CI run 36311638612 failed four of six cells in the trajectory test because LinDelInt raised ImportError: CuPy is required for the 'gpu' engine instead of reaching its vectorized fallback. The Python 3.11/3.12 test environment lacks Pytest Receptor, and the CI command uses plain pytest. Published GH Run Receptor 1.0.0 inspected the run; native failed logs identified the upstream traceback.
Record — devguide/pending_proposals/review_python_ecosystem_policy_adoption.md
Acceptance — Keep both reviews partial until user-selected unit policy survives import, the provider fallback is verified on the affected CI cells, and all maintained pytest CI routes use an exact published receptor pin with --receptor=ci. Record the exact run and boundary tests before claiming adoption.
What — Review ElastNetMT's adoption of MolSysSuite Python support libraries and developer tools under uibcdf/molsyssuite#56. The central inventory still marks both reviews pending.
How — Record the four runtime integration boundaries, fix the import-time PyUnitWizard policy override, and retain the LinDelInt auto-engine failure as a provider dependency under uibcdf/lindelint#8. Pin a published Pytest Receptor release in every hosted test environment and select its
ciprofile in the maintained pytest command. Keep adoption states separate.Why — Source
6705363187762a69048d147853f1018097c76a5edeclares ArgDigest, DepDigest, SMonitor, and PyUnitWizard and uses them in model paths._pyunitwizard.pyunconditionally sets process-wide defaults. The latest CI run 36311638612 failed four of six cells in the trajectory test because LinDelInt raisedImportError: CuPy is required for the 'gpu' engineinstead of reaching its vectorized fallback. The Python 3.11/3.12 test environment lacks Pytest Receptor, and the CI command uses plain pytest. Published GH Run Receptor 1.0.0 inspected the run; native failed logs identified the upstream traceback.Record — devguide/pending_proposals/review_python_ecosystem_policy_adoption.md
Acceptance — Keep both reviews partial until user-selected unit policy survives import, the provider fallback is verified on the affected CI cells, and all maintained pytest CI routes use an exact published receptor pin with
--receptor=ci. Record the exact run and boundary tests before claiming adoption.