Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
k0nserv pushed 3 commits to feature/bwe-loss-based-controller algesten/str0m
algesten closed a pull request on algesten/str0m
Correctly handle per m-line TWCC
So I've been noticing that some receivers report `PeerStats.egress_loss_fraction` sitting at 20-40% with no other indications of packet loss, and `MediaEgressStats.loss` reporting ~0%. Further inv...
alexlapa created a comment on a pull request on algesten/str0m
> Can you split out the change to AbsSendTime in a separate commit and mention it in the changelog? Sure, removed it.
k0nserv created a comment on a pull request on algesten/str0m
I'm happy with this, @algesten do you wanna weigh in too?
algesten pushed 1 commit to fix/cleanup-packet algesten/str0m
- Clean up unused code in packet c6c0769
algesten created a branch on algesten/str0m
fix/cleanup-packet - A Sans I/O WebRTC implementation in Rust.
k0nserv created a comment on a pull request on algesten/str0m
Can you split out the change to AbsSendTime in a separate commit and mention it in the changelog? > but I'm really not sure that its worth it: iterating over 3-6 elements comparing enums should...
alexlapa created a comment on a pull request on algesten/str0m
Also I've been considering caching if TWCC and AbsSendTime are enabled in `Media` to avoid iterating over `ExtensionMap` on each packet, but I'm really not sure that its worth it: iterating over 3-...
k0nserv created a comment on a pull request on algesten/str0m
Oh nice, I've been spotting these high loss reports periodically and have been meaning to investigate it. This sounds like what I've been seeing too!
alexlapa opened a draft pull request on algesten/str0m
Correctly handle per m-line TWCC
So I've been noticing that some receivers report `PeerStats.egress_loss_fraction` sitting at 20-40% with no other indications of packet loss, and `MediaEgressStats.loss` reporting ~0%. Further inv...
k0nserv pushed 782 commits to feature/bwe-loss-based-controller algesten/str0m
- ice: client to ice_lite should not expect binding req 9da9599
- Fix bad unsafe that breaks in nightly 59f4f3c
- Remove frivulous unsafe f4cd770
- No unsafe DTLS version 31499e9
- Forbid unsafe code 409c964
- Fire ChannelOpen once the SCTP channel is ready for prenegotiated channels 4c80e0b
- Move Config->DcepOpen to in-band code path 2d7172c
- Test channel label in out-of-band open event 30e5376
- make clippy happy 0dbfb46
- Allow stats to be turned off 70c923b
- chat stats doc 38deed5
- Fix for flappy ice-lite state 60b7c55
- Lower STUN_MAX_RTO_MILLIS to 3000 0134f15
- Test for flappy ice state 2779363
- .expect, not panic 874241f
- Emit new BWE with tolerances 3f9e480
- Remove unused PADDING_FACTOR 105cc47
- Document constants dd369cc
- poll_estimate dispatch responsibility moved to session c9384f9
- sdp: allow m-sections to have no c-line. c41fd64
- and 762 more ...
algesten closed a pull request on algesten/str0m
Configure RTX ratio cap
https://github.com/algesten/str0m/pull/566#issuecomment-2377171154 > Another thing is it would be great to be able to configure that `0.15` RTX cache drop ratio, mind if i do it in another PR? P...