Firstly, the initial configuration of the network has little effect on the network's long-term behavior (networks all gravitate towards a stable state) which is why Theo was happy to use an unrealistic ring architecture.
If you really want to make things more realistic, your approach depends on whether you want to simulate announcements or not. The 0.3 network relied on a central seed list, when a new node came online it registered itself with this seed list and is given a random selection of 10 of the previous 100 nodes which registered themselves. The 0.4 network is more complex, and the method of seeding is evolving fast. Some initial seeds are obtained (currently just a hardwired list of volunteer nodes but this will change) and an announcement message is sent to these. This announcement message is routed randomly until its HTL runs out, and every node it hits gets a reference to the announcing node, and the announcing node gets the references to every other node that was hit by the announcement. All in all, I would probably start-off with the ring topology, and then improve it after everything else is in place - it shouldn't be a high priority. Ian. On Mon, Nov 26, 2001 at 02:22:39PM +0000, Thomas Gill wrote: > > I'm continuing work on my Freenet search simulation. One problem in > particular is bugging me: How do I start off the network? Am I correct in > saying that previous simulations use a simple ring? That's obviously not > realistic, but i have no idea whether it's a valid approximation. > > It's difficult to imagine how the network in the real freenet might start > (I've been reading all the seed node threads, but It's not obvious how > this would behave in practice) > > Any bright ideas? > > ned. > > > _______________________________________________ > Devl mailing list > Devl at freenetproject.org > http://lists.freenetproject.org/mailman/listinfo/devl -- Ian Clarke ian at freenetproject.org Founder & Coordinator, The Freenet Project http://freenetproject.org/ Chief Technology Officer, Uprizer Inc. http://www.uprizer.com/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20011126/e51ecf73/attachment.pgp>
