> > There is some firewall trickery that we could probably use > > to get most protocols to work, but there are still going to > be things > > that won't. > I view this as not necessarily being a bad thing. it > provides a natural mechanism for controlling bandwith use. I strongly oppose blocking any traffic at all. If you pay for an internet connection, you should be able to pass any traffic you want through it unmolested. Traffic shaping is easy to implement though, so I don't forsee any bandwidth problems that can't be solved with a few rules. > Horwath pointed out; my original statement mentioning BGP may > not be the best way. It's just the first thing that came to > mind. OSPF or something else may be better, but I don't know > squat about dynamic routing protocols BGP is typically used as an external routing protocol. The only reason you would want to run BGP internally is if you have a large network with internet connections all over in different places, then you could propagate your BGP tables internally and traffic would always leave the network at the best internet connection for the network you are trying to reach. OSPF is not necessarily optimal for mobile IP though either as it will flood the network everytime someone connects or disconnects if you are propagating /32's. I don't really see this as being a mobile ip type network though, it would require vast amounts of equipment and money. The reason Ricochet was able to do it is because their product does not need direct line of sight and has a range of up to a mile. 802.11 (don't know about Canopy), almost always needs line of sight, and your range won't be more than a couple hundred feet without a directional antenna, and those are hard to keep pointed while you're driving. :) (I half-slept through > those chapters in my networking classes, and even what I > didn't sleep through didn't sink in very deep). > > Carl Soderstrom. > -- > Network Engineer Scary. :P