/netplus/ - Networks and Plus

I'm still kinda curious about the real purpose of mesh networks.

I mean, you want a flexible and self-repairing network? Well then fuck security.

You want something secure? Either embrace a centralized, non-flexible network with no anonymity or go fuck yourself.

You want anonymity? Well you can I guess, network wise, shame about that unsecure endpoint of yours though, it's unsecure as fuck and can easily leak data without you knowing it, not to mention, say goodbye to security.

You just wanted an alternate, more reliable network? Too bad its antennas can easily be taken down, and its frequencies jammed.

The real internet is simply better on all points, except the last.

The weakest element remains endpoints though, no one can be sure that any device at least tries to be secure and not implement a dozen easily exploitable backdoors for the NSA. My trust in hardware is absurdly low, and frequent findings do nothing to help.

I'd love to establish and use mesh networks, but you'll have to point me towards a safe, decent and widely available endpoint with communication capabilities first. The closest we could get would be a gameboy, audio player, scientific calculator, or other uncommon device, communicating using a very simple protocol and encapsulation implemented 100% software wise and transmitting over the most basic antenna, wire or IR led. And I guess we all see the problem with that.

Why can't I trust anything these days? Paper seems to be the last frontier, and I even came to use fully transparent pens only , just so I can be sure there's nothing inside.

>>181

really there's no replacing the current networking scheme. It's entrenched and it is very good at what it does.

>>178

interesting.

opennicproject supports namecoin's .bit domains btw.

>>181

great points, though i'd argue that these problems exist on the regular internet too. they're just significantly amplified on a mesh.

for example, you could onion route on a mesh network but the bandwidth cost goes up dramatically.

the regular internet doesn't really offer anything significantly more secure at the wire level – you always run encrypted services on top. same goes for a mesh, the cost is just much higher.

part of what makes meshnets difficult is that the available bandwidth goes something like 1/log(N) where N is the number of hops. definitely a problem in need of solving

>>226

First time visting this board. Just wanted to give a shout out to this server, been using it for awhile with dnscrypt-proxy

https://dnscrypt.pl/

>>181

Isn't this where hardware can fill the void? Design a box with a simple network connection, two separate boards with a single one-way connection between the two that only carries a request. Let the connected side run over a low frequency open network with no security, and then once the data moves over to the twin system you can encrypt it there before transferring to a more complex machine for uer interaction.

Of course people could eventually determine your ownership by inserting some form of proprietary data packet and track it to your everyday machine, then match it to the transmitting board- but if we can make the 'connection box' cheap it could be replaced periodically to fix that issue.

Just put this to words a second ago so please correct me if I missed something.

Byzantium needs seeders. http://project-byzantium.org/

>OLSR/byzantium meshnet in each city and town

>forban, tox, URC, i2p-bote/bitmessage allow them all to talk to one another

>most or some nodes on the mesh probably have some connection to the greater internet, anything that needs to go outside the town/city can do so though those nodes, through Tor.

>edgenet

http://theedg.es/#/