You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
pcapkit/vendor/ipx/socket.py:255-256 credits a ruling with wording the maintainer never used:
The owner's original ruling on #775/#841 held Registered by Xerox out of this set as a real ownership fact rather than a placeholder.
The phrase is not his. A fully paginated search of 2118 issue and pull-request comments, 334 inline review comments and ~989 bodies finds a real ownership fact in exactly two bodies — PR #861's, which is where this prose came from, and #994's own — and in no comment at all. The nearest comment variant, "(an ownership fact)", is an agent's own analysis on #775 and describes this same Xerox row, not a ruling.
The ruling's actual reason is on #847 at 2026-09-27T13:15:36Z: a proprietary protocol may not expose a name of its own, so the company name becomes the name. His criterion, #847 at 13:06:33Z, is whether a name is the final concrete assigned name or merely a notation for readers.
Same defect class as #994, which fixed five sites under tests/. This one is separate because:
it is production source, not a test, so it is outside a docs(tests): tranche;
tests/vendor/test_ipx_socket_unit.py:121-122 carries the phrase too, but hedged ("on the theory that"), so it needs a lighter touch;
docs/source/changelog/1.5.0.rst:361 also carries it and is exempt — the changelog is a record of what was said at the time.
A second, pre-existing tension in the same comment block, worth settling while it is open: the prose says the ruling kept the row minted and then that #775's final round converts it. #775 carries no later ruling reversing the mint call — the conversion is PR #878's, whose body says it closes #775's remaining scope. So the prose credits an issue with what a pull request did.
pcapkit/vendor/ipx/socket.py:255-256credits a ruling with wording the maintainer never used:The phrase is not his. A fully paginated search of 2118 issue and pull-request comments, 334 inline review comments and ~989 bodies finds
a real ownership factin exactly two bodies — PR #861's, which is where this prose came from, and #994's own — and in no comment at all. The nearest comment variant, "(an ownership fact)", is an agent's own analysis on #775 and describes this same Xerox row, not a ruling.The ruling's actual reason is on #847 at 2026-09-27T13:15:36Z: a proprietary protocol may not expose a name of its own, so the company name becomes the name. His criterion, #847 at 13:06:33Z, is whether a name is the final concrete assigned name or merely a notation for readers.
Same defect class as #994, which fixed five sites under
tests/. This one is separate because:docs(tests):tranche;pcapkit/vendor/ipx/socket.pyis a generator template, so the paired const file has to be brought into line rather than hand-edited — the same constraint that holds for docs(tests,pcapkit): paraphrase the remaining quoted rulings instead of quoting them (#719) #987's remainingpcapkit/constandpcapkit/vendorspans;tests/vendor/test_ipx_socket_unit.py:121-122carries the phrase too, but hedged ("on the theory that"), so it needs a lighter touch;docs/source/changelog/1.5.0.rst:361also carries it and is exempt — the changelog is a record of what was said at the time.A second, pre-existing tension in the same comment block, worth settling while it is open: the prose says the ruling kept the row minted and then that #775's final round converts it. #775 carries no later ruling reversing the mint call — the conversion is PR #878's, whose body says it closes #775's remaining scope. So the prose credits an issue with what a pull request did.