Few Observations About the Freenet Project

No Comments

While studying different, PRACTICAL, non-academic, real-life, routing algorithms for the raudrohi project, I obviously came across the famous Freenet project.

As this is a free-style blog post, which means, I do not necessarily have to be thorough, all-precise and all serious, I say/write that the Irish, like the Feenet's founder, Ian Clarke, and a kind of romanticized, a bit over the top, fight for freedom sure fit together.

A link to a Wider Version of the Photo.

I did not study the source of the Freenet project, but I did try to build it. According to the presentation that Ian Clarke gave at 2004 DEF CON

the Freenet architecture is based on a swarm of agents, where some of them are out of date, "previous generation" Freenet nodes, some are created by the attackers of Freenet (NSA, KGB, China Wall, Sharia fanatics), etc.

The key points that I gather from that presentation are:

swarm of agents;

to see, how a property of an agent effects the system as a whole, the swarm has to be simulated;

some agents in the swarm are rogue, created by attackers;

agent version updates are practically impossible, unless newer generation agents ignore older generation agents;

system load might be measured as a percentage of that resource that the agent that uses the different types of resources, exhausts most easily;

I think that for all that great work, the Freenet founder, Ian Clarke, definitely deserves to be in the uncensored list of influential software developers, alongside with the developers of Samba, Vim, Ruby, etc. However, what regards to the implementation of the Freenet, despite an appealing (subjectively speaking) theoretical architecture, the implementation of the Freenet is in violation with my subjective taste to such an extent that I do not trust to use it. If I really needed something like that for privacy, I would spend some time on writing some almost-speed-optimized C++ version of my own.

None-the-less, I do emphasize that from theory's point of view the Freenet guys and gals have done a great job. It's just that the "industrialization step" is missing, which is a common issue with almost all European academic projects.

Comments are closed for this post