Found during an unattended triage pass. origin/modernize_type_annotations (single commit 40cdbac, authored by thorwhalen, 2025-11-17, no PR ever opened) modernizes typing.Dict/List/Optional/Union to dict/list/X | None in qh/base.py, qh/core.py, qh/main.py, qh/scrap/*, qh/stores_qh.py, qh/tests/test_base.py, qh/trans.py.
Why I did not land it: its merge-base (1ada85a, "edit setup.cfg") is 57 commits behind current master — before the "Phase 1/2/3" architecture rewrite (convention-based routing, type registry, OpenAPI/client generation, etc.). Rebasing it onto master fails:
qh/main.py and qh/trans.py no longer exist on master (deleted during the rewrite) — conflict as modify/delete.qh/base.py and qh/stores_qh.py have real content conflicts (both files were substantially rewritten since).
Its diff is also no longer representative of the codebase: master today has ~19 modules (types.py, rules.py, async_endpoints.py, jsclient.py, openapi.py, …), most added after this branch's base, and those still use the old typing.Dict/Optional/Union style throughout — this branch only ever touched a slice of the pre-rewrite tree.
Recommendation: delete origin/modernize_type_annotations as superseded (nothing in it is salvageable without a full rewrite against current master). If the type-annotation modernization is still wanted, it should be redone as a fresh branch against current master, ideally as a repo-wide ruff/pyupgrade pass (pyupgrade --py310-plus + a ruff rule such as UP006/UP007/UP045) rather than a hand-written diff, so it doesn't go stale the same way again.
Why this is a manual-task and not something I did myself: deleting a remote branch is outward-facing/irreversible per this fleet's branch-hygiene policy (confirm before deleting), so I'm leaving the branch in place and flagging the decision here rather than deleting it unattended. No code changes made; master is untouched.
Found during an unattended triage pass.
origin/modernize_type_annotations(single commit40cdbac, authored by thorwhalen, 2025-11-17, no PR ever opened) modernizestyping.Dict/List/Optional/Uniontodict/list/X | Noneinqh/base.py,qh/core.py,qh/main.py,qh/scrap/*,qh/stores_qh.py,qh/tests/test_base.py,qh/trans.py.Why I did not land it: its merge-base (
1ada85a, "edit setup.cfg") is 57 commits behind currentmaster— before the "Phase 1/2/3" architecture rewrite (convention-based routing, type registry, OpenAPI/client generation, etc.). Rebasing it ontomasterfails:qh/main.pyandqh/trans.pyno longer exist onmaster(deleted during the rewrite) — conflict as modify/delete.qh/base.pyandqh/stores_qh.pyhave real content conflicts (both files were substantially rewritten since).Its diff is also no longer representative of the codebase:
mastertoday has ~19 modules (types.py,rules.py,async_endpoints.py,jsclient.py,openapi.py, …), most added after this branch's base, and those still use the oldtyping.Dict/Optional/Unionstyle throughout — this branch only ever touched a slice of the pre-rewrite tree.Recommendation: delete
origin/modernize_type_annotationsas superseded (nothing in it is salvageable without a full rewrite against currentmaster). If the type-annotation modernization is still wanted, it should be redone as a fresh branch against currentmaster, ideally as a repo-wideruff/pyupgradepass (pyupgrade --py310-plus+ a ruff rule such asUP006/UP007/UP045) rather than a hand-written diff, so it doesn't go stale the same way again.Why this is a manual-task and not something I did myself: deleting a remote branch is outward-facing/irreversible per this fleet's branch-hygiene policy (confirm before deleting), so I'm leaving the branch in place and flagging the decision here rather than deleting it unattended. No code changes made;
masteris untouched.