[fippo]

With EnablePictureID, PictureIDs below 128 were sent in the 1-byte (7-bit) form and larger ones in the 2-byte (15-bit) form. Both are valid per RFC 7741, but the changing descriptor size means an SFU that rewrites PictureIDs, e.g. when switching simulcast layers, may have to grow or shrink the packet instead of overwriting two bytes in place.

libwebrtc changed to always sending the 2-byte form in
https://webrtc.googlesource.com/src/+/eb94436b386afb7165d91724893b6085365ca1ad

[fippo]

(and yup, libWebRTC is in violation of RFC 7741)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.48%. Comparing base (cbf78a2) to head (d564428).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #377   +/-   ##
=======================================
  Coverage   87.48%   87.48%           
=======================================
  Files          28       28           
  Lines        3172     3165    -7     
=======================================
- Hits         2775     2769    -6     
+ Misses        397      396    -1     

Flag	Coverage Δ
go	87.48% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[JoTurk]

I think it's expected for SFUs to handle this correctly, Since the spec mentions it explicitly? example livekit https://github.com/livekit/livekit/blob/master/pkg/sfu/codecmunger/vp8.go#L192

The receiver MUST NOT assume that the number of bits in PictureID stays the same through the session

I think it's generally bad to adjust our code to bad / lazy implementations that violates specs, but as you said "(and yup, libWebRTC is in violation of RFC 7741)", and it's small change that doesn't break anything so LGTM :)

thank you.

[fippo]

well, this simply avoids the problem, pragmatic @ggarber style ;-)