Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
k0nserv created a review comment on a pull request on algesten/str0m
It's percent expressed as a value in the range `[0, 1.0]`
xnorpx created a review comment on a pull request on algesten/str0m
inherent_loss, is it fraction or percent? either in name or newtype?
xnorpx created a review comment on a pull request on algesten/str0m
can't you use the DataRate type on all bps?
k0nserv opened a pull request on algesten/str0m
Add support for loss in BWE
This PR, which resurrects # 147, adds support for BWE based on loss. Prior to this the BWE system relied solely on congestion manifesting as increasing delay, but it can also manifest as loss, with...
k0nserv pushed 10 commits to feature/bwe-loss-based-controller algesten/str0m
- Correctly handle per m-line TWCC e69b496
- Clean up unused code in packet f487aa9
- Correctly handle per m-line Absolute Send Time 36ead9e
- Dedupe remote ICE candidates The changes in `5c4c9d5` caused problems because we'd end up creating new identical pa... 41ce6cf
- Make use of sha1 crate an optional feature 7c28a1f
- Fix remote candidate de-duplication When considering whether to de-duplicate a remote candidate it's sufficient to v... bc98163
- Tweak log and copy order f4e8558
- Implement goog_cc inspired LossController Ports relevant parts for the V2 loss controller from libWebRTC. This is ba... 17bfff8
- Refactor unnecessary ownership transfer 5a4d257
- Integrate loss controller in BWE dd40752
k0nserv pushed 2 commits to main algesten/str0m
k0nserv created a branch on algesten/str0m
fix/fix-remote-candidate-deduplication - A Sans I/O WebRTC implementation in Rust.
algesten created a comment on a pull request on algesten/str0m
The asm feature is still not working.
thomaseizinger created a comment on a pull request on algesten/str0m
> > FWIW, changing code paths with features can be problematic because features get aggregated by cargo. As a rule of thumb, features should only add new APIs, not modify existing code paths. > ...
algesten created a comment on a pull request on algesten/str0m
Just testing to see if the asm feature works on windows now.
algesten created a comment on a pull request on algesten/str0m
> That is a semver-breaking change then. Is the plan to make the corresponding version bump too? Yeah, let's bump to 0.7.0. > FWIW, changing code paths with features can be problematic becaus...
thomaseizinger created a comment on a pull request on algesten/str0m
> cfg feels a little bit much in this case, but whatever @algesten see fit. I tend to agree. The risk is more on your end because there is no way to disable the `sha1` feature if it is enabled s...
xnorpx created a comment on a pull request on algesten/str0m
cfg feels a little bit much in this case, but whatever @algesten see fit.
thomaseizinger created a comment on a pull request on algesten/str0m
FWIW, changing code paths with features can be problematic because features get aggregated by cargo. As a rule of thumb, features should only add new APIs, not modify existing code paths. To sel...