A decentralized net!

### /g/ternet - name subject to change ### Mark your location on the map if interested: https://www.zeemaps.com/map?group=2784880 irc (SSL): irc.jollo.org:9999 #gternet wiki: http://mesh.gentoo.today pastebin: https://paste(.)ee/p/0AQU4 last thread: >>63542914 → # What is /g/ternet? /g/ternet would be our own decentralized p2p network running on our own infrastructure, powered by something like ipfs. Similar things have already been done: http://www.businessinsider.com/cuban-youth-built-a-secret-internet-network-2015-1 This project is less than 3 days old, so we need all the expertise and help we can get. Every contribution is valuable! Gather a group of friends and join the irc and wiki. To do: - [s]decide on standards for software[/s] We are most likely using libremesh w/ b.a.t.m.a.n. - decide on standards for hardware - create resources for newbies on the wiki - gather more people Hey Lainons, we need your help! We’re trying to create a decentralized net. If you’re interested, there’s current,y a thread running on /g/ and you can join the IRC. Thanks!

hey this is exactly what i was hoping for! A cubano style streetnet! You can read more about it in the pirate book. Another thing I like are those dead drops. Perhaps we should create some more dead drops in tech cities. They're basically public fridges for files. If I were richer, I'd buy a bunch of 128 gb flash drives and make a dozen dead drops. Maybe you could buy a raspberry pi, which, when viewed externally, would be an ethernet port, and you can connect to it as a node to /g/net.

>>6150 Following his idea, it would be interesting to release instructions for joining or even image files to run on RPis to either use them as clients or servers to this network. Like they do in folding@home for example.

what about GNUnet or IPFS?

>>6154 those are higher level in the osi layers, they can run over mesh. Although batman might have performance issues with large dht swarms beyond a certain point. It saturates my uplink and lan... but i use it as a batching service for HTC type stuff

I dont see how this will work out

>>6158 Because it won't. It's /g/ Remember a few years ago where everyone got scared by some law and they decided to back up a bunch of anime on i2p, it's exactly that. It will be forgotten in three or four weeks.

>>6158 >>6159 Throw soykaf at a wall and see what sticks. All you will lose in this is some time, and you may gain some knowledge in exchange.

It appears as though they have finally kind of decided in a name. The main contender is ‘AlterNet’. We also already have decided on using LibreMesh + Batman-adv. People are now discussing about what hardware to use. Anyway, it appears as though there is some progress, I guess.

contribute resources to an already existing meshnet. Or, maybe use the meshnet technology already invented and build the infrastructure. Some of it works very well.

>>6160 This. That is why I want to help.

I'm really interested in how this goes. I hope the hardware is cheap and ubiquitous like raspberry pi or something like that. Anyone from NYC there is already a meshnet thats being deployed https://nycmesh.net/

>>6185 that's what they seem to be doing (what i gather from talking to people in their IRC anyway). i don't care if there's ever a cohesive /g/net so long as it gets more people interested in this stuff. which mesh you join is likely to come down to which one happens to have presence in your city anyway.

>>6224 more like microtikous ahahaha :::---DDD Cheap hardware like raspberries is unreliable and slow. There is nothing wrong in spending thousand dollars on something that will replace Internet backbone and work for more than 5 years without reboot.

>>6150 I don't know how much/little money you have lainon, but there are places online where you can get 128gb usb 3 drives for ~$33 US. They're getting cheaper all the time. Good hunting.

I hope this workouts but all this sounds so post apocalyptic stuff to be honest but its a good community project nontheless

>>6613 always a good idea to get in on the ground floor if things do "go down" in our lifetimes.

I'm really hopeful this turns into something great, and I have a couple thoughts you might find useful... >DAT instead of IPFS see thread https://www.lainchan.org/%CE%A9/res/142.html for a case against the use of IPFS. tl;dr IPFS has some underlying scaling and usability problems which DAT solves. See https://beakerbrowser.com/ for a functional implementation of DAT. >GNUnet I've always been interested in this since I first heard about it, but I haven't seen anyone implement it in a meshnet. I'd love to see how it turns out. >LibreMesh https://mesh.gentoo.today/wiki/Protocol Here you disqualified LibreMesh because of two reasons. I would like to hear a more fleshed out answer as to why you aren't using it and why those two cons are so important. Especially because seems to be the most popular and well supported stack.

>>6677 IPFS is not done yet but as is, works and is great. DAT is a scam.

lol a guy on the zeemaps lives like 5 minutes away from me

>>6677 I'm interested in hearing more about why LibreMesh was disqualified as well, seems like the best choice in my eyes as well.

I know more about IPFS than I do DAT so maybe I jumped to conclusions. Could you please elaborate on why DAT isn't a better choice.

Is there any chance this could take off or is the project already dead? (4 day without any activity here)

What about ZeroNet? It seems to be exactly what those guys are trying to do. Its IPFS based with bitcoin addresses for well... addressing, has IIRC a Git-like structure for updating files and already thousands of nodes. The really unforgivable thing is that they are using Python 2.7 even though it started in 2015.

>>6834 >The really unforgivable thing is that they are using Python even though it started in 2015. FTFY https://github.com/HelloZeroNet/ZeroNet >Decentralized websites using Bitcoin crypto and the BitTorrent network I don't understand what the connection is to bitcoin. Sounds like marketing hype, not a serious technology. I have some blockchain iced tea you might be interested in purchasing. >Each visited site is also served by you. Essentially broadcast to the whole world every website I visit? Even if I use Tor you can still link together all my browsing sessions right? >Hide your IP address with Tor but use the same easily correlated public identity for every connection woot i iz anonymus Did the developers actually think this through or is Tor just a bolt on checkbox feature?

>>6835 Could you explain why using Python in 2015 is a bad choice?

>>6835 >I don't understand what the connection is to bitcoin. It uses bitcoin addresses as a substitute for IP addresses. That way, the addressing is backed up by key pairs.

>>6859 So it's just public key crypto but they use the bitcoin name for hype?

>>6835 >Essentially broadcast to the whole world every website I visit? Even if I use Tor you can still link together all my browsing sessions right? No, not when you use tor. Please read up what tor is, how it works, what it can and cannot do and then come back here. >Did the developers actually think this through or is Tor just a bolt on checkbox feature? The devs thought this through quite thoroughly, unlike you before you pressed "post". See above. >FTFY >I am 12 and what is this FTFY >I don't understand what the connection is to bitcoin. Sounds like marketing hype, not a serious technology. Again. Please do some research before spewing uninformed nonsense. BC adresses are not just PKI with hype as >>6865 claims (i suspect it also you). They follow a set patterns and have useful properties which make them ideal for adressing. On top of that this also means that you do not need to tack on another payment technology on top of websites that need to send or receive money. Its integrated into the technology. Honestly I would not mind if the average grandma on Facebook could not see how awesome that is, but a user here on this board, thats a bit disappointing.

>>6873 BC relevant books

>>6874 Mastering Blockchain is superb book

Hi

Oh boy, yet another /g/ project "made" by clueless LARPers that don't even understand how related technology works in the first place. I can't wait until it's dropped like Netrunner was. Why not contribute to I2P, GNUnet (which by itself is finished, but a more user-friendly client is desperately needed) or any of the other decentralized nets that already exist? Because /g/ doesn't understand how they work and some anon posted "botnet" and the scary parentheses, so they believe every other project is compromised or something.

Have you guys looked at Byzantium Linux? People have done most of the work for you guys already. There was a whole distro that booted up, put your wifi card in ad-hoc mode, connected to other nodes, and routed data using OLSR. On top of the OLSR routed wifi mesh it ran applications like Groundstation for colaborating and Forban for file sharing. IRC even ran over it someone. If one node in the network had a connection to the greater internet, it was shared to everyone else as well. It worked. It was great. And no one used it. Now you guys and a bunch of other other people are all trying to accomplish what was already done years ago. Download Byzantium. It works. Build upon it. >>6956 Those things are not meshnets. Those are decentralized VPNs inside the internet.

>>7018 http://project-byzantium.org/

>>7018 >Those are decentralized VPNs inside the internet. This is not correct at all. i2p and Gnunet are peer to peer routing protocols. They are not bound to the "internet" in any way. You can run either of them ontop of ad hoc wifi to your friend across the street instead of ethernet to your isp. Or run them over cjdns instead of ipv4. If there is no bridge for you to contact existing nodes then you are effectively creating your own isolated network though. >Those things are not meshnets. Depends what you mean by "meshnet". You mentioned OLSR which is also just a routing protocol. >Byzantium Linux https://github.com/Byzantium/Byzantium >Latest commit Aug 11, 2014 NOPE. sage because I don't like this thread or /g/ or anything that comes out of /g/

>>7022 The project was good. Linux is good and it's like 20 years old or something.

>>7359 If no one has touched it in 4 years then it is full of security holes and it will be difficult to add new software because of outdated runtime and libraries. If you want to shll a distro at least make sure it is actively maintained.

>>7365 I'm not saying use it now, I am taking take from it the work that has already been done for us. There is no need to reinvent the wheel again.