Skip to content

chore: bump prop for NSEC compression + per-record multicast floor - #375

Merged
jonasbark merged 1 commit into
6.3from
chore/bump-prop-nsec-and-record-floor
Aug 4, 2026
Merged

chore: bump prop for NSEC compression + per-record multicast floor#375
jonasbark merged 1 commit into
6.3from
chore/bump-prop-nsec-and-record-floor

Conversation

@jonasbark

Copy link
Copy Markdown
Collaborator

Submodule bump only — the fixes are in OpenBikeControl/prop#4.

Context

Tester on 6.3.0+34: still requires the tnc trick. The capture does confirm the previous round landed — all three interface sockets now receive multicast answers where only one did — but it exposes two more defects:

  • NSEC records spelled their 'Next Domain Name' out in full. RFC 6762 §6.1's restricted form (the only one implementations are required to parse) puts a two-byte compression pointer there; a restricted parser reads the Type Bit Map at a fixed offset and got garbage from our 24-byte inline name.
  • The §6 multicast floor was keyed per service, not per record. An A answer stamped the whole service, so the NSEC answering AAAA for the same hostname was suppressed and only ever went out unicast — QU AAAA … → unicast ×81, never once multicast. On a shared 5353 that is the indeterminate delivery path, which matches the long-standing "disable IPv6" workaround in INSTRUCTIONS_WINDOWS_IPV6.md.

Expectations

Both are real, spec-cited defects and the second is a regression I introduced two rounds ago. Neither is proven to fix the tester — the client is resolving the current hostname and still failing. If this build still needs tnc, a Wireshark capture on udp port 5353 is the honest next step.

🤖 Generated with Claude Code

https://claude.ai/code/session_01PanTG9kRBi4QTQ2psvpTxP

…t floor

Picks up two mDNS fixes from the 6.3.0+34 capture: NSEC records now compress
their 'Next Domain Name' to the two-byte form RFC 6762 §6.1 says
implementations are only required to parse, and the §6 multicast floor is
keyed per record instead of per service, so an A answer no longer suppresses
the NSEC that answers AAAA for the same hostname.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PanTG9kRBi4QTQ2psvpTxP
Copilot AI lite review requested due to automatic review settings August 4, 2026 06:39

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR updates the prop Git submodule pointer to pick up upstream mDNS fixes (NSEC “Next Domain Name” compression and per-record multicast floor behavior) referenced in the PR description.

Changes:

  • Bump prop submodule from 23e03e23af77cb9529a04bbdbaf56c3f3197bb4c to 3c470f3bb272cca2cc29bb205b675c23a1042927.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@jonasbark
jonasbark merged commit c3c137e into 6.3 Aug 4, 2026
1 check passed
@lw1321

lw1321 commented Aug 4, 2026

Copy link
Copy Markdown

LGTM

@jonasbark

Copy link
Copy Markdown
Collaborator Author

I appreciate your opinion on this matter

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants