| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
cause problems on restarts and uPnP rule bloat.
|
|
|
|
interface a full sockaddr_storage instead of an int identifier, which turns out to be better for multi-homing and other uses.
|
|
routers.
|
|
|
|
|
|
remote addresses.
|
|
|
|
|
|
|
|
including automatic driver update from NDIS5 to NDIS6. Also a bit more robust on creating new ports, just in case.
|
|
setupapi and using these functions directly.
|
|
|
|
|
|
slight performance improvement. We do our own cryptographically secure authentication so UDP checksum is worthless.
|
|
|
|
make best path choice aware of path rank.
|
|
pushing changes. Move path viability check to one place, and stop trying to use link-local addresses since they are not reliable.
|
|
|
|
|
|
enumeration.
|
|
|
|
leak?)
|
|
about 8MB already.
|
|
|
|
|
|
|
|
renamed accordingly. NetBSD and OpenBSD do not support tap, only tun, so they will have to wait.
|
|
|
|
|
|
|
|
significantly simpler way to move all of this logic entirely into the containing service, liberating the core from any concern over the nature of its pipe to the outside world.
|
|
fallback tunnel hostname.
|
|
|
|
|
|
|
|
and clean up service start/stop in installer.
|
|
|
|
|
|
start on a NodeJS class for talking to the network controller.
|
|
|
|
that later. Should be able to re-use FreeBSD port for OpenBSD, but we will see.
|
|
it will be easier than retrofitting the old Qt client for the new API.
|
|
|
|
stuff and test.
|
|
|
|
control bus at all yet.
|
