[freenet-dev] Wierd clustering behaviour again

[Ian Clarke]

On 12/19/06, Oskar Sandberg <oskar at freenetproject.org> wrote:
> Matthew Toseland wrote:
> > - Does anyone have any idea how to proceed from here? Any theories? Any
> >   idea for intelligence gathering?
>
> I have previously recommended that all nodes should drop their position
> in favor of a random selection on a regular (actually random) basis.
> That could help against this as well.

But isn't that really just sweeping the problem under the carpet - or
is there a good reason why we would see this type of clustering using
the algorithm without some kind of re-randomization?

Ian.