As P2P applications are proliferating, are there also dangers to this?

Francois Rey wrote us the following cautionary words:

“Ripple, Peeple, freenet, Skype, file sharing, media broadasting, content distribution, etc. It seems that no computing area that is left untouched by the P2P phenomenon. With so many P2P programs becoming available, will it become possible for them to run together on the same machine? The question may not seem so important nowadays, but with several P2P programs running there are a few problems that may arise.

First the number of connection can exceed your machine capacity and without an overall management for each p2p connection you run the risk of slowing your machine to a halt and not being able to watch properly the P2P TV channel. QoS (Quality of Service) for P2P will certainly become a concern. Secondly each P2P program requires certain doors to be opened on the host machine, thus making it more vulnerable to internet attacks. Each P2P program exposes its host to its vulnerabilities.

The time where such questions will become poignant may not be so far after all: many people run Skype already, many also run a file sharing program at the same time. Add a couple other P2P programs and you’ll probably see the limits of unmanaged P2P applications. In order to resolve elegantly such problems some form of communization will be needed, at least at the level of network connection. One could imagine a single component for managing the number of peer connections and the quality of service (e.g. to ensure you can still watch P2P TV while downloading some other content). However beyond such simple common service it would not be very difficult to imagine other services that could be shared between several P2P applications: peer discovery, peer capacity discovery, service advertising, etc. Eventually this may lead to the concept of P2P application servers managing P2P application so that each node can safely run several P2P applications, and from where P2P application can easily be installed.

JXTA is certainly the first technology that comes to my mind which has the potential to lead us in that direction. However I’m not aware of all P2P research and work, so feel free to share your insight if you have any.”