A large number of rule-based program that self-organise could be used to move information more efficiently around the internet. These agents, modelled on bees and ants, treat user requests like food to move video, etc, around the networks. The Bio-Networking Architecture site has research papers.
This makes a lot of sense. The simple rules that ants use to find the shortest route between, say, food and the nest are extremely powerful when applied in parallel. This kind of transport level syndication would could make caching and higher level syndication methods unnecessary, and would be ideal. I wish I could remember where the comment in slashdot was, but it mentioned a potential network that was transport mechanism independent. As such it could have no ping or traceroute, but it was completely decentralised. Whether these kind of ideas will come into a network of the future I don't know, but it's really not going to be any time soon - look at how long it's taking to bring in IPv6.
A large number of rule-based program that self-organise could be used to move information more efficiently around the internet. These agents, modelled on bees and ants, treat user requests like food to move video, etc, around the networks. The Bio-Networking Architecture site has research papers.
This makes a lot of sense. The simple rules that ants use to find the shortest route between, say, food and the nest are extremely powerful when applied in parallel. This kind of transport level syndication would could make caching and higher level syndication methods unnecessary, and would be ideal. I wish I could remember where the comment in slashdot was, but it mentioned a potential network that was transport mechanism independent. As such it could have no ping or traceroute, but it was completely decentralised. Whether these kind of ideas will come into a network of the future I don't know, but it's really not going to be any time soon - look at how long it's taking to bring in IPv6.