Apparently the original implementation had issues. Apparently NetGate shipped some bad code after hiring a FreeBSD contributor for their implementation. Kyle Evans has even decided to step away from maintaining wireguard-freebsd. Seems like a mess.
> Kyle Evans has even decided to step away from maintaining wireguard-freebsd
That is not true. Everyone acknowledges the code is not up to the required standard so it is being removed but Kyle and others will continue to work on it with the aim of adding it back.
Edit: as @asmr below linked to, he has has since announced that he is quitting.
To be double-clear - the problems being discussed are in the freebsd kernel-mode implementation. The freebsd userspace implementation is fine, and implementations on other OSes are fine :)
Apparently the original implementation had issues. Apparently NetGate shipped some bad code after hiring a FreeBSD contributor for their implementation. Kyle Evans has even decided to step away from maintaining wireguard-freebsd. Seems like a mess.
> Kyle Evans has even decided to step away from maintaining wireguard-freebsd
That is not true. Everyone acknowledges the code is not up to the required standard so it is being removed but Kyle and others will continue to work on it with the aim of adding it back.
Edit: as @asmr below linked to, he has has since announced that he is quitting.
https://lists.zx2c4.com/pipermail/wireguard/2021-March/00652...
2 replies →
I think it's just lack of formal peer review [1]
[1] https://lists.zx2c4.com/pipermail/wireguard/2021-March/00650...
To state this clearly: there are not concerns about WireGuard itself.
To be double-clear - the problems being discussed are in the freebsd kernel-mode implementation. The freebsd userspace implementation is fine, and implementations on other OSes are fine :)
There were questions around the implementation. See https://news.ycombinator.com/item?id=26475519