A Radically New Internet - A Study on P2P Protocols and Mesh Networks

zelf · April 25, 2019, 4:36pm

Illustration of the Scuttlebut protocol

Mesh networks are local network of routers that are interconnected. Usually they are a way for communities to share WIFI connectivity with each other and in many cases these mesh networks are run by communities as a means of free Internet connection within an area, or at least cheaper Internet connection within an area. Mesh networks can also be found within companies for more local infrastructure, these are seldomly, by those active in the mesh network scene, considered mesh networks though.

What’s specifically interesting is when you take mesh networks in combination with peer-to-peer protocols, such as Scuttlebutt or the DAT-protocol as new features are enabled, seldomly else even imagined. They happen to fit very well together as the two technologies merge and creating something quite unique:

Some interesting use qualities are,

offline communication
Communication which can happen in the same geographical region without requiring http/“The Internet”. Be that file sharing, such as in the case of dat or actual person to person communication such as in the case of SSB.
Free Data - The only cost is the hardware, hardware similar to the ones already found in almost every single household in western society.
More private data, less vulnerability - As the data is stored locally there’s a much smaller risk in terms of third party infringement, be that in the form of a data leak such as the facebook scandal

I’m currently looking at use cases for these technologies and have sketched out three different areas for these kinds of implementations.

Underdeveloped regions where infrastructure are not always available. Having mesh networks in combination with peer-to-peer enables for communication through WIFI and that communication can still happen.
Local communities, it’s very efficient means for communication, for example for eco villages, off-grid living - mesh networks are ideal.
Privacy / National Infrastructure Security - We are seeing a rise of cyber security attacks, specifically cyber security attacks which are targeting national and corporate infrastrucutre - taking communities and entire nations offline. Mesh networks in combination with peer-to-peer protocols could be a solution for a lot of privacy issues that we are currently seeing in contemporary society.

If mesh networks and peer-to-peer protocols are implemented, especially in combination with each other, it would be a much more durable communciation infrastructure for society. This also goes for company infrastructures and the flaws that we see - as a means for both international, and business2business espionage is happening via cyber platforms.

If companies use and communicate via meshnets and peer-to-peer it is safer and not as easy to hijack.

One of the biggest saefty issues in terms of privacy is also human errors, some of those issues are eliminated with peer-to-peer solutions. Passwords are eliminated. User IDs and device are tied to each other. A hacker would actually have to have physical access to someones device or gain their private keys rather than password accounts which is more easily cracked.

Currently there are a lot of different mesh networks in the world. I do not have a complete overview at the moment. Guifi covers most of southern Europe. Freifunk is a big one in Germany. Then we have some areas in Greece which communicates via satellite connections, but local ones.

Those are the most interesting ones right now. There are actually no examples as of now having mesh networks plus peer-to-peer protocols implemented at the same time.

In regards to specifically Sweden and the Nordic region I am in looking into this right now. I am keen to explore this as with the three target areas mentioned above. On one hand our communication infrastructure is not as durable as we would like to think. Both seen through examples as in Northern Sweden as infrastructure can be down and out for longer time periods when there is snow for example. Also with cases when there is highly fragile infrastructure which from a cyber security perspective can be taken offline entirely by a small team of 5 qualified hackers.

This research is funded by the EU fund Designscapes for exploring the implementation of mesh+P2P in Northern Europe. Any leads, potential contacts or groups to discuss implementation with is highly welcomed!

felix.wolfsteller · April 26, 2019, 11:23am

Love it.

I had a list of technologies like IPFS, scuttlebut, dat, SOLID (Bernes-Lee) and so on, but don’t find it at the moment. If you could document your findings in a wiki topic (like the resource planning software list) or somewhere else and share the link, I’d be very gladful. Noawadys, a lot of projects take the blockchain hype (and I assumed 95% of it will be gone/given up within the next few years), where it is very difficult to access the real value and basis of these projects - at least for me. Doing stuff “blockchain” seems to often comes with “having a supershiny overpromising webpage” (have to rant a bit here).

And at the last Makers4Humanity lab there was something going a bit into the crossover you mentioned: a local file-sharing hot spot with chat capabilities. I do not remember the projects name, but might stumble upon it again. I think I took a picture of the printed-out usage instructions somwhere.

johncoate · April 26, 2019, 3:02pm

You are part of the Designscapes group?

johncoate · April 27, 2019, 2:35pm

I meant zelf. It looks like an interesting group and project.

johncoate · April 27, 2019, 2:36pm

There are actually no examples as of now having mesh networks plus peer-to-peer protocols implemented at the same time.

So in other words, there are these networks set up for communication within the network, and they may or may not plug into the main 'Net in some way, but there is not an example of these mesh networks communicating with each other directly?

zelf · April 28, 2019, 11:28am

I had a list of technologies like IPFS, scuttlebut, dat, SOLID (Bernes-Lee) and so on, but don’t find it at the moment.

@felix.wolfsteller, that’s awesome! If you do find it, I’d be very keen to see.
I keep my lists of fun things in this spread.

Yeah, reg. blockchain I feel similarly and tend to stay away from the hype. That’s also why I’m so excited about the new realm of distributed webs, it has real practical implementations and solves a lot of the issues we see with current web protocols. IPFS too even though they are more silicon-valley oriented than SSB or dat. I’m very keen to learn more about Solid! So far I’ve only heard briefly about it… I can’t help but wonder if there’s a reason I for why I haven’t heard anyone talk about it much.

And at the last Makers4Humanity lab there was something going a bit into the crossover you mentioned: a local file-sharing hot spot with chat capabilities.

This sounds interesting as well! If you remember what it was, link me up pls!

zelf · April 28, 2019, 1:29pm

You are part of the Designscapes group?

Not directly part of it, the research I’m doing currently on mesh-networks and P2P was funded by them though

So in other words, there are these networks set up for communication within the network, and they may or may not plug into the main 'Net in some way, but there is not an example of these mesh networks communicating with each other directly?

No, and yes. haha, to give a correct answer.
No as in, that’s not what I was referring to at in that section. There simply aren’t any mesh networks which run with P2P protocols as of yet, only prototypes. This is even though technologically they are a perfect fit due to the local data hosting.

Yes, as in there are no direct connections between the distributed protocols although the ssb and dat community are starting to look into how they can merge better, especially SSB and DAT are a very good combo. If you are on Scuttlebutt, there’s a bit more about this in this link:
%jRIyWsJZwKgxfA3K/pJfZv4sMv4DQMKl8Eq+JHPo3p4=.sha256

zelf · April 28, 2019, 1:32pm

And at the last Makers4Humanity lab there was something going a bit into the crossover you mentioned: a local file-sharing hot spot with chat capabilities.

Was it perhaps Subnodes you were thinking of?

Yannick from iMal mentioned it over on Scuttlebutt and it sounded very similar to what you described.

felix.wolfsteller · April 28, 2019, 7:13pm

Right now pretty sure it was PirateBox . On the way to finding it, I also stumbled over David Darts - Offline File Sharing . Too sad we often forget that the world is full of crazy and cool.

hugi · April 29, 2019, 9:21am

Thank you for this excellent post @zelf!

We actually just got our hands on a bunch of RaspberryPis at Blivande, the cultural hub, workshop and co-working space we run in the harbor of Stockholm. Maybe we should extend a mesh network around the house and the harbor? I think @jonas might be interested in some meshing.

notme · April 29, 2019, 8:20pm

Check out freifunk.net - Freifunk steht für freie Kommunikation in digitalen Datennetzen
They seem to be the largest community that is active at the moment. They have been active for quite a few years.
For something more ephemeral but in the same idea of other networks, also have a look at https://hyperboria.net/
Hyperboria had a very interesting scene going a while ago, I lost my key so I have not been there for a while. There were some people from the Nordic region in Hyperboria when I was there.

There are quite a few P2P protocols or Delay Tolerant Networking systems available, but not very many that would be easy to adopt. At least I have not found a favourite yet. Freifunk seems to have a relatively robust solution going.

Maybe you know all this, but I did not see it mentioned.

johncoate · April 29, 2019, 11:50pm

Meshin’ around.

alberto · April 30, 2019, 1:05pm

Hello @zelf, welcome and thanks for the thoughtful post. I am myself a user of Scuttlebutt (SSB), and would like to qualify your proposition that

Scuttlebutt’s presentation video states:

When Scuttlebutt is running, it looks to see if there are any other Scuttlebutt accounts on the local network, that is to say, the coffee shop wi-fi.

In the video, the two cute characters Jackie and Martin are indeed using SSB from the same coffee shop. She follows him, he is well connected, and all is clear skies and smooth sailing from there. Indeed, they do not need “the Internet” to share stuff with each other, because they sit on the same local network. Neither do they need servers, because the content lives in folders on their laptops. So, technologically, the features you identify in your post ensue.

But what happens socially? Not much, it turns out. Maybe New Zealand is different that way, but If I go to a Brussels Café the chances of someone else running SSB from the same café are close to zero. So, nothing is enabled for me. To use SSB, i need to use the hack of pubs:

pubs are servers that have a static IP address, are always online, and always follow you back when you follow them (thus growing their network). In this way, they can act as a waypoint for you to find the location of your friends and catch up with them. When you first get started on Scuttlebutt it can appear incredibly empty until you “join” a pub.

In other words, I could not use SSB without “the Internet” (or moving to a hipsterish area of New Zealand, perhaps).

None of this invalidates what you have to say. All I’m saying is this: let’s all be very wary of technological determinism. Just because the tech enables something, it does not mean this something will happen. I am curious: do you have any experience of SSB working locally, as in your areas 1 and 2? Where? Do you have any observation to report as to what people use it for?

Keep up the good work.

Pasting the link to the SSB presentation video for the many folks who do not use it.

pbihr · May 10, 2019, 12:17pm

Hi @zelf, I love this!

I’m sure you’ve considered it already, but just in case: I think another really interesting use case for this is just local communities, even within well-connected areas. A fully localized mesh could be built up as a hub for all kinds of creativity that has a distinctly local flavor in a way that often the internet isn’t optimized for (even though of course localized communities exist there, too).

For a fictional, but really cool exploration of this I’d highly recommend the (still brand new) debut novel by Tim Maughan, Infinite Detail (Goodreads). Set in a kind of post-crash world without internet, there’s a community there that uses a local mesh as the foundation for all kinds of local creative sharing from music to augmented reality installations in public space, out on the street.

Not exactly what you’re looking for, but I figured it might be close enough to your research interest that it might be worth taking a quick look there.

That said, the notion of more resilient networking is a tempting one, even outside of purely military use - say in conflict/post-conflict or disaster settings (war zones, earthquakes, etc.), or even just in very remote locations where local file sharing can be really useful among (concurrent or time-shifted) visitors.

zelf · May 14, 2019, 7:31pm

This sounds awesome! I’d be happy to include you in the potential collaborators for a the prototype stage at the end of the report for the research phase, does this sound good to you @jonas?

zelf · May 14, 2019, 7:38pm

Yes, they are awesome! I’ve been considering getting in touch with them. Plan for now is to try it out in the nordics and then potentially expand. I’ve researched their Mesh-structure a bit and apparently their network structure makes it so that the entire network would act as a singular node for the SSB protocol. Will look further into it next week though

I think Guifi might be the largest network in the world though

Hyperboria are cool too! Just saw that they’re hosting a “Battle of the mesh” event on July 8-14th too!

There are quite a few P2P protocols or Delay Tolerant Networking systems available, but not very many that would be easy to adopt. At least I have not found a favourite yet. Freifunk seems to have a relatively robust solution going.

Hmm! This I haven’t heard about, what is it?

zelf · May 14, 2019, 7:43pm

It’s true that you can’t, at the moment, use SSB without wifi, thus the proposed merge with Mesh-Networks. This would make it so that you could be anywhere in a mesh-network region and be “at the same cafe” so to speak. Decentralized protocols would also solve the issue of mesh-networks running over https usually having a very big issue with stable connectivity since SSB is “offline-first” locally hosting the data.

I am curious: do you have any experience of SSB working locally, as in your areas 1 and 2? Where? Do you have any observation to report as to what people use it for?

Nope! And that’s why we are conducting this project of ZMesh, to explore the combination of P2P protocols and mesh-networks in a real life scenario, as it has never been done before, as of yet.

zelf · May 14, 2019, 7:46pm

Coooooooooooool!
Awesome recommendation! Will definitely try to get my hands on Infinite Detail!

That said, the notion of more resilient networking is a tempting one, even outside of purely military use - say in conflict/post-conflict or disaster settings (war zones, earthquakes, etc.), or even just in very remote locations where local file sharing can be really useful among (concurrent or time-shifted) visitors.

Yes!! Definitely. I’m very keen to explore this aspect. Also from a resilience perspective, preparatory for the inevitable crash of the centralized communication structure which is the internet in its current form.

jonas · May 15, 2019, 2:20pm

I would be happy to be on the list even though at the moment there are too many things going on and I’m having a hard time focusing on this subject, might organize myself better soon and explore it more…

I bet @Andrusha would also be interested in collaborating <3

btw, I also heard recently about http://pjodd.se/ from from this %P9i5KKgtlzPR3ybqrvyCAdGoAGtwa+PHckimUYNyjaE=.sha256 @zelf, it might be also relevant to you

alberto · May 16, 2019, 7:02am

On my list. Thanks @pbihr!

I do not see that as a severe problem. Rather, I was pointing out to a social network problem: a local social network will only work if enough local people use it for it not to be an empty space. Until we went together to the matchmaking event in Stockholm, I had never had the pleasure to fire up Patchwork and see SSB users on my local network. People were taking screenshots, which means it does not happen often to others, either.

In Belgium we have a hyperlocal social network called Hoplr. When you sign up, you have to add your address, and they check that you really live where you say you live. You then are added to the same database as everyone else, but you can only see, and interact with, people that live in your same neighborhood. And neighborhoods are small, at least in Brussels: they use a statistical unit called quartier, of which there are 145 in the city (1.3 million inhabitants). I am a member for my quartier, but the activity is not enough to sustain a real conversation, even through “the Internet”. My corner of Hoplr has defaulted to a kind of hard-to-search Craigslist, with people selling used furniture and the like.

Of course, local Hoplr and local SSB could be socially great, if everyone used it. But not everyone will adopt them until they are socially great, so we are stuck. The pubs solution makes SSB viable for me, but it would not be possible without “the Internet”.