Topic for discussion ... Found Here (http://sourceforge.net/tracker/index.php?func=detail&aid=554707&group_id=34939&atid=412930)
You have 5 machines, each connected to the internet.
Each of these machines is hooked to the main server.
Each machine can handle let's say a maximum of 200
Users. That means 1000 users connecting to 5 machines
throughout the network. These 5 machines transmit
respective data in real-time to the main hub server.
Each machine transmits it's own set of chat messages
and file catalogues. The central server then transmits
recieved data from each machine to the others, making
a complete cluster, so each machine has a complete
network snapshot from chat, users and network data and
addresses.
When a user requests files or chat, the machine he is
connected to has all the required information needed
for communication, thus the user can bypass the
network as long as there is no lag in the address
information.
I understand this may not be as easy to code as it is
to plan, but maybe clustering is in the future. If
developed, it would certianly accelerate the growth of
the DC Protocol.
------
Reading this got me thinking of ways that this would be possible ..
HubSoft Cores in general dont (nowdays) consume any huge amount of cpu ...
so what other benefit would that be for hub servers?