We have reviewed version 2.0, and we will not be adding support for it for the foreseeable future. We understand that many users here were expecting that we will add support for it, so before you feel disappointed, let us explain why.
Supporting AmneziaWG 2.0 means that we need to drop support for native WireGuard which is powered by a kernel module which is a lot faster than the Go implementation of AmneziaWG 2.0 to support it on all existing VPN servers.
Alternatively we could deploy brand new and separate infrastructure only for AmneziaWG 2.0 and separate it from WireGuard, but that is costly (not only in terms of money, but mainly developing and keeping it safe and up to date) and we already have Stealth Proxies for this.
AmneziaWG 2.0 is very close to Hysteria 2.0 that we already support.
Supporting only AmneziaWG 1.0 is not that bad and DPI vendors still need to keep the block rules for it. More block rules = more computing power. At some point some DPI vendors might even drop block rules for it, because everyone will focus on 2.0 version. Of course more modern DPI system might keep all the possible protocols in the block list, but it does not apply to every country/city.
Adding support for AmneziaWG 2.0 won't solve the issue with blocked VPN infrastructure that we are using for WireGuard/OpenVPN and I have explained this in my previous message here.
We will keep checking the status and development of AmneziaWG, but for now we will not support AmneziaWG 2.0 or any further versions of it. We think it's fair game based on how many protocols we already provide today.
You don't have permission to do this.
You're going a bit too fast! Take a break and try again in a moment.
Something went wrong! Please reload the page and try again.
Your session has expired. Please reload the page and try again.
We have reviewed version 2.0, and we will not be adding support for it for the foreseeable future. We understand that many users here were expecting that we will add support for it, so before you feel disappointed, let us explain why.
Supporting AmneziaWG 2.0 means that we need to drop support for native WireGuard which is powered by a kernel module which is a lot faster than the Go implementation of AmneziaWG 2.0 to support it on all existing VPN servers.
Alternatively we could deploy brand new and separate infrastructure only for AmneziaWG 2.0 and separate it from WireGuard, but that is costly (not only in terms of money, but mainly developing and keeping it safe and up to date) and we already have Stealth Proxies for this.
AmneziaWG 2.0 is very close to Hysteria 2.0 that we already support.
Supporting only AmneziaWG 1.0 is not that bad and DPI vendors still need to keep the block rules for it. More block rules = more computing power. At some point some DPI vendors might even drop block rules for it, because everyone will focus on 2.0 version. Of course more modern DPI system might keep all the possible protocols in the block list, but it does not apply to every country/city.
Adding support for AmneziaWG 2.0 won't solve the issue with blocked VPN infrastructure that we are using for WireGuard/OpenVPN and I have explained this in my previous message here.
We will keep checking the status and development of AmneziaWG, but for now we will not support AmneziaWG 2.0 or any further versions of it. We think it's fair game based on how many protocols we already provide today.