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
{{ message }}
Repository navigation
fix(const): LLACode.Wilcard misspells RFC 5568's code 0, and LLACode.MH names code 2 differently from the RFC #1848
pcapkit/vendor/mh/lla_code.py:25-30 hand-writes the table.
Code 0 is 'Wilcard', a typo: its own comment and RFC 5568 §6.4.3 say Wildcard.
Code 2 is 'MH', where the comment and the RFC say Link-Layer Address of the MN. This one is a naming judgement rather than a typo: the library's only use of it is the MH link-layer address option.
The generated LLACode.Wilcard = 0 and LLACode.MH = 2 have shipped since 1.0.1.post1 (55cde1b33) and are in 1.4.1, so a rename is breaking. In-tree users of LLACode.MH are protocols/internet/mh.py:3257,8105 and tests/protocols/internet/test_mh_unit.py:367,801,1301. #1845's probe also names LLACode.Wilcard.
Recorded: option 1, for the typo. LLACode.Wilcard becomes LLACode.Wildcard, with a breaking-change versionchanged:: 1.5.0 and no deprecated alias, as in #1090. A worker is on that part.
To confirm code 2: RFC 5568 §6.4.3 (as RFC 4068 did before it) names it Link-Layer Address of the MN, and IANA keeps no separate registry for these codes. Following the RFC therefore means renaming LLACode.MH to LLACode.MN as well, which touches protocols/internet/mh.py:3257,8105. Shall it become MN, or stay MH as shipped?
Recorded: LLACode.MH becomes LLACode.MN, following RFC 5568, alongside the Wildcard fix. Both go in the same breaking-change PR, with versionchanged:: 1.5.0 and no deprecated alias. The in-tree users (protocols/internet/mh.py:3257,8105 and tests/protocols/internet/test_mh_unit.py) are updated with them.
Activity
JarryShaw commentedon Oct 16, 2026
take option 1 - rename the typo and for Code 2, keep what the IETF/IANA/RFC says, as it may actually be a historical choice by the writers.
JarryShaw commentedon Oct 16, 2026
Recorded: option 1, for the typo.
LLACode.WilcardbecomesLLACode.Wildcard, with a breaking-changeversionchanged:: 1.5.0and no deprecated alias, as in #1090. A worker is on that part.To confirm code 2: RFC 5568 §6.4.3 (as RFC 4068 did before it) names it Link-Layer Address of the MN, and IANA keeps no separate registry for these codes. Following the RFC therefore means renaming
LLACode.MHtoLLACode.MNas well, which touchesprotocols/internet/mh.py:3257,8105. Shall it becomeMN, or stayMHas shipped?JarryShaw commentedon Oct 16, 2026
MNthen.JarryShaw commentedon Oct 16, 2026
Recorded:
LLACode.MHbecomesLLACode.MN, following RFC 5568, alongside theWildcardfix. Both go in the same breaking-change PR, withversionchanged:: 1.5.0and no deprecated alias. The in-tree users (protocols/internet/mh.py:3257,8105andtests/protocols/internet/test_mh_unit.py) are updated with them.fix(const)!: name LLACode codes 0 and 2 as RFC 5568 does (#1848)