| Age | Commit message (Collapse) | Author |
|
|
|
COM is pushed with multicast, which makes tremendous sense in retrospect.
|
|
|
|
|
|
more cleanup. Also fix it so certificates wont be accepted unless they are newer than existing ones.
|
|
|
|
and on receipt of MULTICAST_LIKE.
|
|
TRACE, flesh out multicast tracing a bit.
|
|
even if there are no frames passing between them.
|
|
well. Time for some more simulation before 0.5.0!
|
|
independently and does so frequently enough to prevent expires. Also add a multicast debug facility for use on the testnet.
|
|
|
|
Also fix mac-tap build on new version of Xcode CL tools. Must use old llvm-g++ instead of clang for i686 -mkernel.
|
|
|
|
|
|
|
|
algo a bit more before finalizing.
|
|
|
|
|
|
variety of system types to test the result of this.
|
|
|
|
still some issues with tap.
|
|
do random propagation for more efficient coverage with less bias in sparse graph cases.
|
|
a start.
|
|
|
|
|
|
development.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
testnet before merge to master.
|
|
|
|
|
|
of a diet. Work in progress, wont build yet.
|
|
|
|
cleaned up and clarified it a bit.
|
|
|
|
duplicate multicasts back to where they came from
|
|
the code.
|
|
to help debug multicast propagation
|
|
prevents a race that can cause the first NAT-t to fail where the NOP arrives before the WHOIS reply from the supernode. Now NAT-t initiators will push their own public keys anyway so that doesnt matter.
|
|
|
|
|