WireGuard: Interface - Unofficial WireGuard Documentation

[Sob]

C. Use the IP addresses as a mechanism to create Dynamic Routes from the Local Node for:
(this is most common as it MIMICS existing linux/android apps that AUTOMATICALLY CREATE ROUTES from allowed IPs.)

i. Local originating traffic entering the tunnel heading outbound
ii Remote returning traffic (coming from local servers or the local WANIP for example) entering the tunnel outbound.

As you may have surmized, the traffic identified in i and ii, contain BOTH the destination IPs of the local users and the source IPs of the remote users.
Hence the IP addresses are used to cover the ALLOWED IPs on the local node. and this is exactly what many wireguard apps do!!

I told you it’s completely wrong. Several times. I explained why. Several times. I asked you nicely to stop with this nonsense. I even used “please”. So help me here. What should I threaten you with? Or what else will do the trick?

You’re mimicking nothing. You don’t know what you’re doing. At least try it. If you have e.g. this config:

[Interface]
PrivateKey = <key>
Address = 192.168.44.30/24

[Peer]
PublicKey = <key>
AllowedIPs = 192.168.44.0/24,192.168.45.0/24,192.168.46.0/24
Endpoint = my.server.tls:12345

Then Linux client does this:

# wg-quick up wgtest
[#] ip link add wgtest type wireguard
[#] wg setconf wgtest /dev/fd/63
[#] ip -4 address add 192.168.44.30/24 dev wgtest
[#] ip link set mtu 1420 up dev wgtest
[#] ip -4 route add 192.168.46.0/24 dev wgtest
[#] ip -4 route add 192.168.45.0/24 dev wgtest

Do you see it assigning random 192.168.45.x/24 and 192.168.46.x/24 addresses anywhere? You don’t. Why might that be, you ask? Because it would be nonsense. So once more, get a grip and just do what wise Linux (and other) people do.