Sup Forumsternet - We are back and better than ever edition

Mark yourself on the map with your contact info (can be a cock.li throwaway if you don't want to dox yourself)
zeemaps.com/map?group=2787145

IRC (SSL): irc.jollo.org:9999 #gternet (uses self signed cert)
Git: git.gternet.me

What is Sup Forumsternet?
Sup Forumsternet would be our own decentralized p2p network running on our own infrastructure, totally disconnected from the clearnet.
It will be a combination of DIY meshnets on a city wide basis, and a platform for interconnecting these city wide meshnets over p2p openvpn bridges and high powered wireless dishes for City to City relays.
The Inter-Meshnet part of the project will allow any existing DIY meshnets to join our network by adding a bridge node or a city to city relay.
The short term goal is to have a few anons set up their own local meshnet and link them together using the bridges as a proof of concept.
A long term goal is gain a large enough user density that we can disable the brige nodes and have our own decentralized network idependent of the internet.

Other urls found in this thread:

open-mesh.org/projects/open-mesh/wiki
ubnt.com/airfiber/airfiber5/
security.stackexchange.com/questions/113532
twitter.com/NSFWRedditGif

Where have you been? I thought this project was dead?
This project's primary contributors all have real lives and real jobs that get in the way of development. We figured that it would be best to just work without threads for a few months to try
and hammer out some plans and details. We now need help to set up the meshnets and intermesh nodes for testing and further development. This project is not dead.

Well what do you have to show us now?
We have one 3 block square meshnet implemented on raspberry Pi 3s from one contributor. The mesh devices are battery operated and get about 16 hours of life before they need to be charged.
The mesh network allows non mesh devices to join as if it is a normal wifi network. The mesh is then connected to a p2p node waiting for a peer to connect to.

What can I do to help?
Currently we need people to set up their own local meshnets (can be small just to prove the concept) and then set up p2p nodes. You can use gternet-cli from git to help with meshnet and
p2p setup. Its still a work in progress and for some reason you have to run the deploy mesh command twice for it to work on some devices. Help us modify it to be better by submitting change requests
to git and knots will review them and merge if the code is improved

To Do:
- decide on hardware standards for both City to City relays and Meshnet end user equipment
- start building internet bridges for the intermesh connections (Linux Sysadmin tier)
- start designing diy kits for meshnet end users (Should be newbie-midlevel friendly tier)
- look into croundfunding options for purchasing city to city relay backbone equipment (Richfag tier)

Things NOT To Do:
- Talk about logos
- Talk about names
- Shill crypto bullshit for funding

Here is a screenshot of gternet CLIs help command.

How will operators protect themselves once this gets popular and the police are wardriving around triangulating the locations of mesh nodes? Every darknet ever is mainly used to share content that would get you trouble if transmitted over the clearnet, so I don't see how this would be any different.

We have lightly discussed this in IRC and we have a few ideas revolving mostly around enforcing encryption network wide giving node operators plausible deniability for the content that goes through their node. But this isn't the best solution. We are open to any legal advice someone could offer us for this issue.

fuck off, nobody on this board knows enough about networking or they're too busy with irl or being shitposting faggots in the threads and irc

I think your best approach would be to focus on good strong opsec and accept that some of you will be going to jail. The Pirate Bay wasn't breaking any laws when it launched, so the law was changed to go after them. I see absolutely no way for your project to remain lawful.

Essentially we would be operating as a free WIPS so the laws inbplace have some protection for these kind of services already. Also the design allows for a nebulous nature so nodes can keep popping in and out of existence to outpace any shutdowns.
Thanks for the input!

Aarnon, saw your Aarhus pin. No contact info was there, but it said you watch the threads. Are you on IRC?

>I think your best approach would be to focus on good strong opsec and accept that some of you will be going to jail.

If they change the laws, then every person involved will have to evaluate how far they're willing to go. We're not there yet. There are also the options of adapting the system to the laws. How that will work will depend on the laws. Meanwhile I've been to jail. I'm down.

This is what a winblows machine sees when connecting to the mesh.

bump

...

Never seen this before. What benefit does this offer over the standard internet I get from my ISP?

You won't have to pay for an ISP if this gets big enough. We are decentralizing the internet and making a way for eveyone to get connected without getting fucked by a provider every month.
Also allows for net neutrality to be built in by default, and ensures large companies don't keep their monopoly on internet access.
But these are LONG term goals way down the line.

If it reaches a useful number of nodes then decentralization. We want to the people to have control over how traffic is handled as opposed to the government or for profit entities.

That would be nice. But how would infrastructure improvements happen? Or upkeep on existing infrastructure?

Let's say in 20 years this becomes so massive that it's standard procedure, but there is so much data being transmitted that the current cabling/info transmission system can't keep up with it. How will that be upgraded? Who will pay for it?

Each city meshnet could act as its own governing body and could implement group funding efforts to upgrade their side of the network. This means that the infastructure itself is owned and operated by the users.
A city wide mesh where users cannot afford the upgrades could be supplemented by other crowd funding options across the wider user base.
Havent thought about this much though. Good points.

Bamp. How can a newb in Alberta help with this?

Invest in two or 3 raspberry pi3s (have been tested) and try learning how to use gternet cli to get them meshed together jump On IRC for help

In a true mesh ultimately the node owners have to pay for upgrades. The cool thing is the more distributed the mesh gets the less sections of the network start to look like back bones.

There are a lot of what ifs that we haven't answered. It will fall on the participants to find a way. The cool thing about these what ifs in particular is that they are good problems. I much prefer to live in a world where new solutions have to devised in order to allow mesh networks to deal with congestion due to their massive popularity. It beats the one where you get to choose between 2 government supported companies and pretend it's a free market.

>Let's say in 20 years this becomes so massive that it's standard procedure, but there is so much data being transmitted that the current cabling/info transmission system can't keep up with it.

Ultimately I'm saying let's cause this problem. It's awesome. Then we can solve it. That's even more awesome. Then we can cause it again. That's how innovation happens.

I've got one raspberry pi3 I'm trying to set up as my own nextcloud server, would I need one more, or two more?
Bamp for inspiration. Fuck the ISPs and the CIA

At least one other pi is needed to test a mesh network. 3 is better because then you can have a real test of connectivity.

Hm.. I'm a studentfag poorfag though. Is this project still going to be around in two years?

Yes. plan on it. I am one of the primary contributors and im not going anywhere.

Also, is there any way to know whether the Raspberry pi3s have backdoors built into them? Would be a shame to rebuild the internet only for the CIA niggers to have root level access to everything

The way I understand the plan is anything that can run openvpn can be a node. So if you ran gternet on a the nextcloud server it would be usable in this. My plan is once we get the core vps nodes going then I'll throw up a wifi access point that allows people to access the content. I play on setting up nodes for people I know to allow them to do the same.

This.
Right now gternet cli works with debian systems and one member already ported it to fedora. But any toaster tier computer that can run openvpn and has batman in its OS can work in theory it just has to be tested

Ultimately it's really hard to prove a negative. Iirc there are some blobs involved in raspberry pi firmware. Raspberry Pis are being suggested because this has been tested on them. Porting to different systems. Anything that can run OpenVPN is a potential node. If you want to run this on a FPGA configured to be an open source CPU, then you have my ax.

>Porting to different systems
I meant
Porting to different systems is encouraged.

How do you do intercontinental internet?

Fair point about the massive "what if".

As I understand it, what allows you to connect to the internet is a line that goes from your house directly to the ISP which consists or tons of switches that all direct traffic. How would/do the bridges work/bypass this?

Btw second semester network specialist. I'm at that point where I kind of know my shit but also kind of not

Internet p2p bridges for now. Licensed low frequency transmissi9nbhas beem discussed but its highly regulsted and pretty expensive. We are going to stick to using the openvpn bridges as the intercontinental back end for now

At the physical layer for now we are still using the internet as a back end for now with the openvpn nodes operating at layer 2.
The long term would see this project acting as a kind of wireless ISP with community owed/rented intercity hardware. Right now we are looking at ubiquity fiber over air equipment to accomplish this

Low frequency transmissions
Sorry on phone right now and still suck at touchscreens 10 years on.

I'm not against afraid of tunneling using existing infrastructure. Particular if it's multiple tunnels owned by different people. Non-profits putting satellites in orbit is something that has been done before as well.

>Btw second semester network specialist.

You're probably better equipped to understand this than me then:
open-mesh.org/projects/open-mesh/wiki

My knowledge is more in line with the EE side. I'm learning the ins and outs of the networking as we go. Right now what I understand is that dynamically decided routing is something with a few solutions and room for improvement.

aren't you guys afraid that there will be gatekeepers, sabotaging your project

That is a problem that any large scale effort will run into. The current way we have things implemented, each p2p or intercity node operator needs a personal level of trust to other operators. To connect meshes these operators need to exchange certificate pairs manually to connect. For now this is the best way we can establish that someone isnt trying to fuck the network but we will need to revisit this at scale. Another big "what if" that will need to be addressed at some point

I'm not sure what a gatekeeper means in this context. The idea of abusive nodes is one that we've discussed. Other meshes have probably run in to it. I'm open to hearing suggestions.

I asked about interconti internets. It seems like a bottle neck since solutions are rather expensive. So these nodes could be used to limit access to a world wide internet.

With p2p openvpn bridges running on VMs hosted in datacenters we can spin up as many intercontinental nodes as we want on the fly. I don't see the bottleneck unless they shut down undersea cable access to the project which is really unlikely because who the fuck are we?

my assumption was w/o using undersea cables.

Until we can license low band radio transmissions or something else for traversing oceans then the p2p notes or the best we can do.
Open to any other ideas though.

I think we should start off by giving this project a name and a nice logo to start pushing it to other people. "Sup Forumsternet" doesn't exactly roll off the tongue. And if you think about it, great companies like microsoft and apple would be nowhere without their identities (logo and name)

Well if we assume the undersea cables are inaccessible we're left with line of site satellites with relatively high latency or HF communications with low bandwidth and reliability. Really unless they're willing to heavily limit international vpn connections, this isn't a problem.

Oh right these threads again, I still think lasers are the best idea

Gigawatt EIRP moonbounce baby.

Are you the laserbro who bought stuff to build a laser based relay?

If you get a laser link up we're cool with that. Physical layer is something both ends of a link have to agree with. Nobody else really has to worry about it.

bump

Bumpu~

...

...

>shoot signal lazors at moon
>reflectors by nasa return signal to earth
>cia niggers see signal as coming from the moon
>????
>profit

>hurr lets build the internet guizeeee

I like it, honestly. If nothing else, everyone walks away a little smarter on how shit works and how to fix other problems.

I'm seriously considering joining. Can we go back to web 2.0 though? js was a mistake.

...

Not sure how you enforce this while staying with the project goals, but I totally agree js was a mistake.

No, it really wouldn't work without creating severe restrictions (which would go against the project to do.) I just wanted to shit on js. :(

It's not totally disconnected from the internet if it uses the internet.

Also you have a nice idea but lacking in implementation.
No you cannot use shell scripts to implement a p2p protocol. You cannot and should not do this under any circumstance.

>No you cannot use shell scripts to implement a p2p protocol.

Have you read the code? The shell scripts are just automated ways of configuring the servers. OpenVPN and B.A.T.M.A.N. are where the heavy lifting gets done.

the end goal is to not use the internet. Yes you can be isolated from the internet while on the internet thats how VPNs operate. By creating isolated bridges. Thats how TOR works, thats how I2P works, thats how ZeroNET works.

the shell scripts are used to set up existing solutions and tieing them all together. Most people here cant set up an openVPN server. So we automated it. Most people dont know how to use BATMAN-adv, so we automated it. We arent out here writing custom p2p protocols in bash. That would be retarded. How about you ask questions first and jump to assumptions later?

I have one node in my city, what's next?

get on IRC and learn how to set up a local meshnet. Go to map and find user who put pin down. get in contact with them and work together to get your two meshes connected

It's not isolated from the internet, but any data traversing it can be protected from observation by someone on the internet.
You can use encryption to "isolate" yourself as mentioned before but the bridges will still be dependent on internet infrastructure.

No sorry I haven't read the code, but the repo could use some documentation.

not once we hit critical mass and can turn off the internet bridges. As you can see in the diagram and the OP we are planning on turning off the internet bridges once we have enough city to city relays set up with wireless equipment.
We are looking into ubiquity fiber over air stuff right now.

sorry about the sparse documentation. I'm working on it.

That's pretty dope actually.
I can help out with whatever if you guys need it. I'm a bit rusty on shell scripting and other HLLs but I have experience writing production C/C++ code. I can also help out with the docs if I take some time to get acquainted with the architecture.

Will be following the project in my free time

cool! glad to have the help. Jump on IRC and join the discussion if you want. People who are willing to help right documentation and how tos are really needed right now

I had an idea a while ago, where I would build a mesh network for the 4G protocol, and distribute SIM cards to my users. Should I just build a standard WiFi mesh network instead?

>20 and 30 miles to nearest nodes
Ain't gonna work. Some questions though:

1. How do nodes even connect to one another? Wifi? You won't get the required density for that in an "open beta" kind of scenario like this. Anything else I can think of that works over longer distances requires extra equipment and too much fucking about for an average Sup Forums user and is probably cucked by the FCC anyway.
2. What sort of security is provided, how do you deal with rogue nodes, MITM and the like?
3. Is it targeted at stationary desktop users or hipsters going around with their laptops?
4. Can it work and scale without even remotely central management, ie. nodes just coming and going as they please without affecting the reliability of the network? Because central management is always the breeding ground for pushing agendas, power trips, divide and conquer destruction from the inside by glow in the dark cia niggers.
5. Once it's up and running, do you plan to gate to the Internet somehow? If not, what content do you even plan on having hosted there? I mean there's hardly any content on TOR that you can't find on surface web, and 99% of what you can't will get you vanned. Are you sure you want to create a "network" as such (which I understand to be a backbone for hosting other services) rather than a decentralised distributed communication service, something like less-than-realtime IRC or Usenet maybe?

Not sure if it helps or even makes sense, that's my 2 cents for now though.

What about dns servers? Because using ip adresses to connect to a service might too much trouble especially with ipv6.

Just make a dedicated DNS server array with public-net passthrough, and have your nodes use it.

I will addreess this post later tonight when im back at my desk. Thanks for the input. These are good questions

>check map
>somebody else from leipzig (germany)
nice.

Cheers.

Cyber bump

If you can tie your 4g based network to our intermesh connections then go for It! The idea is that any DIY mesh network should be able to interface with the rest of gternet regardless of the local mesh architechture

> (uses self signed cert)
What with Let's Encrypt, AWS Certificate Manager, etc., there is no excuse anymore.

There is nothing cryptographically wrong with self signed certs.

>How do nodes even connect to one another? Wifi?

Physical link is completely up to the two people making the links. Wifi seems to be preferred. You'd be surprised what is doable with a line of sight and high gain antennas. 30 miles links will take some hefty investment though.

>What sort of security is provided, how do you deal with rogue nodes, MITM and the like?
I haven't seen much discussion of possible attacks. I'd be interested in hearing more about threats and mitigations. If you're concerned about MITM attacks though you should definitely be using some kind of protocol that allows you to protect against this.

>Right now we're mostly thinking about getting stationary working. I believe the batman protocol is okay with roaming. I could be wrong.
>Can it work and scale without even remotely central management
Putting the infrastructure in the people's hands changes the power dynamics but yeah those people are vulnerable to mind viruses. A lot of this is becomes social questions I don't think are settled. To me the two social dynamics that are most important is how do you decide who to link with and what rules do you set with them about what happens over that link.

>Once it's up and running, do you plan to gate to the Internet somehow?
If someone is willing to act as the gateway, then I'm okay with them doing this. I also don't see how we can prevent it even if we weren't. I'm wary of being an exit node and concern number one when trying to port this to my OS has been am I sure I understand what's going on well enough that I know this isn't going to send traffic through anything but an encrypted tunnel.

For anyone wondering this is what I'm looking into buying a pair of for the fist long range city to city link

ubnt.com/airfiber/airfiber5/

...

...

...

...

that girl is 9 at most

One little doubt it came to me when I was scouting the thread. Since I guess untill better move, the mesh will be being tunneled through existing ISPs, that means that mesh participants will be able to cofound the best and fastest speed they can bought from whatever ISPs, which will improve the speed a lot, and not only that, if I'm right, that means that each mesh will be able to spend whatever they consider necessary if they don't need a lot.

How much am I right or wrong?

I think you've got the idea. Each person has a lot of autonomy on deciding how they connect.

was meant for
pic was meant for

Seems like a very feasible project then. I lack knowledge and probably some money, although I would have speed. I'm here just trying to learn more about nftables and how to do a really specific thing, so I think I won't be able to help.

But fuck me, user, I will be lurking and I hope you get more and more people around.

Feel free to lurk IRC and learn along with us. I have almost no idea what I'm doing but every once in a while I figure something out and get closer to what I'm trying to accomplish.

I never went on IRC because I know I can expose myself too much if I don't set myself right, and I know I can be too brainlet soometimes so I stayed more safe than sorry.

theres one node in my country.

hi OP, do you have any more doccu for the project? I wrote a research paper on mesh networks in uni and I would love to play around with the project and get involved if I could.

>I know I can expose myself too much
What do you mean? Just connecting to IRC is pretty safe.

We're really low on documentation right now. I'll be picking through the gternet-cli code soon trying to port it. Maybe I'll take some notes. We had a wiki but it died because of drama.

>What do you mean? Just connecting to IRC is pretty safe.
I always heard things like ip address or the proper messages could be exposed. Maybe I'm uneducated but I found things like this when I try to know more about it:

>security.stackexchange.com/questions/113532

Your IP address is exposed to other users on this server I think. Something to do with the way they set up hostmasks. You could set up a proxy if this really bothers you. It doesn't for me. I'm hanging out with my wang out. As for the messages this server only allows encrypted connections. So nobody can read your packets in route. However this is a public chat room. Assume everything you say in here is recorded because it is. I keep logs. I keep them on an encrypted disk but anyone could be publishing them. This is certainly not the place to share secret information or start a criminal conspiracy. You could use asymmetrical encryption to communicate securely with a single user using pgp or something on here. That's a whole different story. There's only 4 billion or so ipv4 addresses. You're not losing a lot of security by exposing that. If you're not currently wanted by a law enforcement agency or organized criminals, then I'd say it's safe to come say hello.

I have to apologize since probably I overstated what I wanted to say. I meant, if that kind of data (which, as you say, there's other issues more serious than those) is exposed, I don't know what more I can shit up. Although I'm seeing IRC clients right now. Some recommendation for Windows? I will research some ones on Linux later.