tinc 1.1pre10 slower than tinc 1.0, experimentalProtocol even more

[shikkc]

On 2014-04-17 04:23, Henning Verbeek wrote:
...
> 
> Sorry, my bad here. I misread the config. The two tinc daemons run on
> separate ports. All good *phew* :)
> 
...

Just out of curiosity, are the daemons in separate physical hardware? I ask 
because I have tried a wide variety of configurations with multiple daemons 
on the same machine, and have noticed that without fail the CPU will fully 
saturate if two daemons on the same machine are on the same VPN. This has 
been true both inside and outside of various visualization technologies, as 
well as in cases where the two instances should not be able to talk to each 
other (e.g. even when they do not have each other's host file, and 
TunnelServer = yes). I have not really done any further digging aside from 
taking this as a good sign I should not do it.