Alpt wrote: > yep, no incoming connection for mobile nodes. Is it sacrificable right? > (generally you don't need to host a server in a mobile node) well, maybe not a server like sshd or httpd, but some voip softphones work really better without nat. I think we should not think of mobile-nodes just as mobile-phones or embeded devices, but also as someone wardriving with his/her laptop, or things like that.... I agree it is sacrificable, but for me it's better to keep it if it's possible.
> if the node goes down: some hacks must be used. In the previous mail, I > imagined a situation where you can send packets of a same connection by > different IPs. This isn't impossible, but it requires some good hacks on > netfilter (f.e. see http://lab.dyne.org/Ntk_IGS) ok, but this (as reported in the link) is very o.s.-dependent. It might be a problem for embedded device, since every vendor should then modify ntkd code if they want their device to act like a static netsukuku node... and the vendors should make sure that it doesn't mess up with the device's firewall... But maybe it's the most simple solution, you're right. > the other advantage to use a NAT/swarm solution is to completely separate the > mobile > part from the static one, this is relevant because you can't think in terms of > maps for mobile nodes (maybe) yes, it's the easiest way, but it's a shame not to use all those devices for routing :p Sorry, but I don't understand why mobile nodes should not have maps... Is it becouse when they move they would generate too much traffic to update its and other nodes' maps? I'll try to test something when i have time, i'll let you know if i have better ideas :p Luca. _______________________________________________ Netsukuku mailing list [email protected] http://lists.dyne.org/mailman/listinfo/netsukuku
