Skip to content

Conversation

@huitema
Copy link
Contributor

@huitema huitema commented Dec 3, 2025

This PR is a companion to a picoquic PR updating the C4 code. It update the spec, design and test documents to reflect the changes, and loosens 4 tests:

  • the "max of 100" completion times of the ECN test rose from 5s to 5.6s
  • the "max of 100" completion times of the "bad wifi" test rose from 4.3s to 4.5s
  • the "max of 100" completion times of the "compete with cubic under bad wifi" test rose from 12.1s to 12.5s
  • the max latency observed in 100 "media short then long RTT" test rose from 125 to 126ms.

Most of test changes are due to the stricter evaluation of the Max RTT. The previous (buggy) code set the early "Max RTT" to match the final RTT observed in the Initial phase, which was too high and made C4 too aggressive. This was mostly visible in the "bad wifi" tests. Other performance changes are explained by the random consequences of changing initial conditions.

@huitema huitema merged commit cc635d7 into main Dec 3, 2025
1 check passed
@huitema huitema deleted the fix-c4-rtt-bug branch December 3, 2025 01:50
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