It's Been a Good Run, ISPs.

This node might be able to reach anywhere in the world, which definitely shouldn't be possible. Not because the antenna is magical, but because each node like it [music] forms a mesh network automatically. Now, in my last video, I discovered something called Reticulum, an open-source internet protocol that might let you build networks without ISPs. And that video ended up reaching over 2 million people, which I'm sure caught the attention of even my most loyal viewers. There was just one problem. All those tests were between two nodes about this far apart. So, yeah, of course it worked. They were basically touching. So, naturally, people started wondering, "Cool, but how far can this actually go? Across a house? Across a neighborhood? Across a city? What about across the entire United States?" Because if the tech delivers, we might be able to build an open-source internet, one that doesn't depend on ISPs at all. There's just one rule for this video. The data must flow solely through the Reticulum network. So, the challenge is simple. Can I be here in Florida and talk to someone in, say, Venezuela all using Reticula? Let's find out. Now, I've spent the last few years building what I like to think of as a parallel economy of tech. Basically, open-source, decentralized, and affordable tech that almost anyone can run. And the whole point is to get rid of subscriptions, gatekeepers, [music] and middlemen, and put the power back into the hands of the people. And if you're watching this, you probably want the same thing I do, which is a network you don't need permission to use. So, let's see if we can even get the first brick to hold. We're going to do two nodes, one text message, and no ISP. This is the house test. Quick heads-up, today's video is sponsored by Claw Stage, an open-source personal assistant that runs on a Raspberry Pi 5. But more on that later. Okay, so for our first test, we're just going to see if we can send messages across my house. Very, very modest test, but it'll lay the found works for some future, more exotic tests. So, um let me give you the mental model here. So, for the home-based node, we have a green Haven node, which is essentially a Raspberry Pi. It's running a fork of uh open-source routing system called OpenWRT. The specific variant is called OpenMANET. And it has traditional Wi-Fi, it has Ethernet, but what makes it special is that it has a sub-gigahertz uh 900 MHz um Halo chip. Halo chip and Halo antenna, or sub-gigahertz antenna. And so, what that lets us do is it bridges all those different Wi-Fi interfaces, but it's able to um make a connection at a lot longer of a distance than uh traditional 2.4 or 5 GHz um Wi-Fi. So, when we send these two Reticula messages, they're going to be going across the Halo radio. That's why we're going to hopefully be able to get more distance. So, Reticulum is actually not running on this router. It's just going to be making use of this router's network, and it's going to be a mesh network. There's technically backhaul um that goes out to uh public internet, but that's actually not fundamental to what we're doing here. So, um it would be able to function either way. But um we're going to be doing a little bit of internet testing here and there, so that's why I want to have that. But um so, where does Reticulum run then? Well, Reticulum's going to run on the client. So, my computer right here is going to be running a program called MeshChat, which is it's a messenger, but it's also um it it runs the RNS service, which is actually like Reticulum itself. Um this is where we'll send messages, but it makes use of this network. So, my computer is going to connect to this guy's regular 2.4 GHz Wi-Fi. Um the network name is green, and you can see I'm connected to it. My computer has no other connections, no Ethernet, nothing like that. So, the only way it's able to get a network connection is from this guy's Wi-Fi, and we're connected. We should be good to go. And what we're going to be doing is using the auto interfaces. With Reticulum, you can define interfaces uh on the actual hardware, like on a USB serial, on a radio itself. It's quite flexible and quite agnostic, but you can also just do this auto interface, where if your device has Ethernet or Wi-Fi, it'll just automatically discover that and make use of that, which is nice. Um and I think this kind of uh keeps the overhead here pretty minimal. So, mental model base node, base network node, Reticulum client. Now, there's nothing stopping us from we could add other devices to the setup that also connect to this node here, um but you need at least one Reticulum device, and then you need a network for it to make use of. And um and so, that's the setup we have here. Now, for node number two, because it's uh something that we're going to be moving with, like on the go, uh I'm going to use a much smaller device. What we have here is this is technically um the Heltec HDT01, which is a it looks like an MCU, it's not. It's actually a full Linux board. So, this also runs OpenWRT. It doesn't run OpenMANET, but it runs OpenWRT, same software as what's running on our Raspberry Pi here. It's not going to have the speed that the Raspberry Pi has because it's a much smaller footprint, but it it does have the Halo radio. It has an Ethernet jack. It has an onboard 2.4, just like this. And it has our SMA connector. So, what we can do is we can take another um sub-gigahertz antenna, same sort of antenna we'd use for Meshtastic, Mesh Core, etc., because it's that sub-gigahertz. Pop that on here, and this is going to um bridge to the Halo on the Haven. Now, the question is how do we get Reticulum working on this? Well, we don't need to get Reticulum working on this, the same way we didn't put Reticulum on this. This is just going to provide the connection to our network, and then Reticulum is going to run on another client app. Um the best sort of on-the-move app is an app called Sideband. It's a public APK, meaning you can just download it off the GitHub releases page. You can put it on your Android device. This device is going to connect to this. We're going to use the same sort of auto interface on Sideband. We're not going to do anything fancy there. And this Retic- um Sideband runs RNS, which is Reticulum, but it also gives you the chat um interface. So, we'll be able to send messages to our base node here. So, I should be able to text uh type on my desktop, receive messages on my um Android, type on my Android, receive messages on my desktop. Now, a couple of things about this setup. This setup is not something you've probably seen a whole lot, but there there's something that's kind of cool that I've been messing around with, which is this. So, put this aside for a sec, but let's just take my iPhone. That that Android device is something I use for testing, but um my iPhone is something that I use day in and day out. And I actually forgot I had this around my neck. Something that's really interesting that you can do is if I go into my iPhone and let's turn this down a little bit. So, I go into my iPhone and I turn off the network connections. I'm turning off Wi-Fi, turning off cellular, and I'll turn off Bluetooth, just and AirDrop. So, this guy has no network connection. Okay? But what's interesting is if I go over to the settings page, what I can do is I can actually plug this guy directly into my router using a USB-C to Ethernet cable. And when I do that, we see Ethernet pop up here as soon as it's plugged in. And I get an IP address and all that good stuff. And what's cool about that is you get internet speeds like this. I actually didn't know that. I didn't know that the um smartphone supported Ethernet like that. So, are we going to be putting our phone in our home router over Ethernet? Probably not. But what we can do is we can take that same concept, and we can actually plug our um Halo board into our phone. Um and so, that's exactly what I'm going to do here. So, let me just situate this, and turn on my phone here. Okay, so I'm going to just go over to settings, and I'm going to do the same thing. I'm going to turn off Wi-Fi. So, Wi-Fi's off. So, what we're going to do is we're going to have to turn this guy on. And what we did um just to uh show you up close um here is that this guy has Ethernet like that. And we just plug it in like that. And all we have to do now is power this guy up, and I'm going to be using to do that um an 18650. And actually, it's the it's similar to what we have inside the Haven. The Haven's actually 21700s, but this is an 18650 and what's interesting about it, actually has a side USB-C port. Seems weird, but actually very um practical and I actually get a lot of use out of these. So, I'm going to plug a USB-C in there and then I'm also going to, if I can, um I'm plugging this dongle in here just so that we can monitor the rate of the power, uh but it's definitely not needed. So, uh don't get hung up on that. But, like for instance, it will allow us to um see the energy draw and things like that. And we see 5 volts, um less than a watt of power draw, which is cool. Certainly not going to get those kind of numbers from the Raspberry Pi. Um so, good to get those numbers. So, that router board is starting up and what we're waiting for is for our connection to show up here. And I'm just going to give it a second. It usually takes a second for it to show up. And then on my computer over here, I have MeshChat going. I've already set up some messages with the device, so hopefully I can just keep using that conversation in MeshChat, but let's get um let's get this guy connected to this guy's network and see if we can't send some messages and then I'll go across the house with this and um and we should be able to get those messages to go. So, one way to know how you're looking in terms of the network is to go to something like WiFiman. WiFiman will tell you if you're getting any reads yet. Okay, so we just saw it show up on the mesh. So, that's looking good. No IP address because funny enough, the mesh is actually Batman ADV, which is layer two, so it's actually working off MAC addresses, not IP addresses. But, that's beyond, I think, what's uh relevant to this video. So, now we should get a connection up here. I'm going to go over to settings. Okay, there it goes. Do you see the ethernet? Just showed up there. It doesn't give you a whole lot of details because I don't I don't know how many people are using ethernet on their phone. And uh we'll talk about You actually don't need to use ethernet. I just thought it was neat. Um this guy could actually use this guy's 2.4 just fine and that's probably a better way to do it. And then this would just need to be like in your backpack or something and you could just work off your phone like you normally would. But, I just thought the ethernet part was cool and uh we're exploring new new ways to do things today. So, anyways, um so, do we know if we have a network connection? So, we go back over to WiFiman. When it says gateway there, that's the Heltec and it's it's got an IP address, 10.410.3. You can see we're on a slightly different subnet than like traditional home network, right? Um in most cases. Um but, you also see uh we have an upstream router. So, the Heltec is a mesh node on the mesh gate, which is the Haven, and because the Haven has internet, we can do a speed test and we should get a connection. Hm. Um let's try just going to a website. Let's go to Amazon and we can see that it does work, but we can see it's crawling. And we can also see that WiFiman doesn't even register as a connection. So, what's going on there? What's happening is we're using the 1 MHz channel width for Halo. That's configurable, but the thing about 1 MHz is it has the best range, but it has the lowest bandwidth. It actually caps out at I think it's around 3 megabits per second, but in reality it's more like one or two. So, I think WiFiman's kind of saying you don't have an internet connection for the most part. Um but, if we change that to like 8 MHz, you can get as high as like 43 megabits per second, which is more akin to high-speed internet. See if we can't get something. It should be able to. Why is it giving us a hard time with that? Well, you know what we could do? Let's do an internet speed test just on like Google. That should work. Come on, you can do it. Look at that internet speed. But, the thing about 1 MHz is like it's still pretty capable for like IoT stuff. And funny enough, I was actually doing like I would I FaceTimed Amelia yesterday using the 1 MHz channel and she said the quality was like perfect, which is kind of bizarre. Um so, you can do a lot with a couple megabits per second. I mean, it's orders of magnitude more than like LoRa and LoRa does text messages, GPS, and stuff like that. So, um yeah, so you it's and it's real internet, right? Where LoRa is just kind of its own protocol. This Wi-Fi Halo is like true internet. So, yeah, those those upload speeds are pretty modest, but um you can do a lot with that and we can always tweak that later. So, I wouldn't get too hung up on that. So, let's do the Reticulum stuff, right? So, we have a mesh. We have two different devices. So, on my computer, I'm going to go over to MeshChat cuz that's what we're going to need. And on my phone, I'm going to go over to Sideband. And we're going to see if we can't send a message. So, Sideband um you know, has uh runs Reticulum, has all the configurations around that, but also has the actual messenger itself. So, if I go to um conversations, I see MeshChat. So, I select that and I'm going to say um are you with me? And it's going to try to discover a path. Well, it already says it was delivered and I see a dot over here. So, actually that was pretty that was pretty quick. Uh let's see if we can do um something a little bit more here. I'm going to do a voice note. So, I'm going to say high-quality voice. We're going to do allow while using app. Testing 1 2 3 4. This message is it's being sent across Reticulum. And then I think you send a note with that. Just going to put voice. Okay, the note came across, but where is voice message? Oh, I didn't click save to message. Okay, let's do it again. Testing, this is a voice message that will be sent across Reticulum. Stop recording. Save to message. Test. Okay, I think that's going to work here. Sometimes you got to refresh. Yep, there it is. Testing, this is a voice message that will be sent across Reticulum. Cool. So, we just sent a voice message across Reticulum and then maybe we can do an image real quick. So, we're going to do high-res image, um DCIM, camera. All right, and there are our images. Says it was attached. I'm going to do IMG. Just showing you a little bit kind of what you can do with Reticulum, right? So, I just sent a message. It looks like on MeshChat, you kind of have to like navigate away and then navigate back and there it is. There's our image. So, all right, we're successfully sending messages across Reticulum from client to client, but what's cool is the clients are powered by uh our Wi-Fi Halo mesh network and everything here is basically open source. So, um pretty impressive what we're doing already. But, I recognize antenna number one is very close to antenna number two. So, let's go across my house and see if we can't continue to get these messages to send. Mr. Bruno, >> [snorts] >> code name Silver Fox, the world needs you, Mr. Bruno. We have a very important mission for you, Mr. Bruno. I know you've been in retirement for several years, Mr. Bruno, but you are the only one who can help us. We need your expertise for one last mission to save the world, Mr. Bruno. Will you help us? Will you help us, Mr. Bruno? Okay. So, can we get a message to send through? Let's take a look here. So, just go ahead and [clears throat] try something out here. So, I'm going to do test 1 2 3 and it was immediately delivered. So, that's really what we'd be looking for. And then I guess for fun, let's go ahead and take a picture of Bruno. Bruno's going crazy over there, but he's helping us with our mission. I just snapped a little picture of Bruno. So, I should be able to come back over to Sideband and then inside band should be able to do one of these. We're going to do high-res image, DCIM, camera. And then it's that, I believe. And we're going to say Silver Fox because that is his code name. There it is. Oh my god, look at that shot. Okay, so we are able to send messages uh across just fine. I can even open up Sideband here. We see our picture of Bruno that we sent to MeshChat. And then when I go back up to my office, I should see these messages on the MeshChat chat. Oh my god, Bruno. Um so, across the house, pretty trivial, actually. It seems to be working fine. Um so, what we're going to do is we're going to push this a little bit further and we're going to try to do messages across my neighborhood, so like several kilometers, and see if this continues to hold. Okay, so now we're back up here, and I'm just taking a quick look at mesh chat. I'm going to refresh it. Okay, and we see that the messages came through. So, there Bruno is fighting some imaginary I don't even know how he contorts like that, but anyways, I don't think people understand just how disruptive uh this space is going to be over the next couple years with things like Leo from Amazon. Starlink is all already a multi-billion dollar company, and just the advent of AI and LLMs when it comes to coding. I mean, the barrier to entry to this stuff just keeps going down and down and down. And we're here to take advantage of that. Okay, so across the house we have been able to achieve. The next mission is going to be to send messages across my neighborhood. So, in order to make that work, um I want to actually uh upgrade our antenna setup here, because my house is a cement brick. So, yes, sub gigahertz can penetrate dense material, but we're not exactly giving it kind of the best advantage by permeating the um signal from within uh an enclosed structure like that. So, what I want to do is I want to mount an antenna outside my house, just so that it is able to go even further. So, let's do that. Okay, so like I mentioned before. Before we do the neighborhood test, I want to put an antenna outside my house, because again, house is made out of cement brick. So, if I put the antenna inside, it immediately kind of hits the brick and starts reflecting, and it can penetrate through the brick, but it's not going to be optimal. If we could just get it on the other side of that concrete, I think we're pretty much going to extend the range considerably. So, the way I'm going to go about that is um this is a fiberglass outdoor antenna with 9 dBi gain, and this is um I think the center frequency is 915 for this one. I got it on Amazon. I think in retrospect, I probably would have gone with uh just a reputable antenna supplier, because it's hard to verify if the antenna is actually tuned properly, and what the center frequency actually is. And they probably expect most customers not to do that. So, I've been getting good results out of this, but I would probably just go with a vendor that has a really good reputation as opposed to just sort of like a generic brand on uh like Amazon or something like that. But, this has been working well, so we can stick with that. Now, I don't immediately want to um start drilling holes in my house or the facade or anything like that, because I don't know where I'm going to put the antenna ultimately, and this is really just for an initial test. So, what I'm actually going to do is this uh antenna uses a BNC connection. Um and so, it mounts a particular way. But, what I'm thinking is this. I have a kind of uh window sill here, and what I think we could do is put a tension mount um between the two sides of that. And so, this will run, because there's like two pretty solid concrete sides that we can uh set this up against. Uh and it's a tension mount, so I don't have to drill any holes. And then, what I'll do is once this is up, I can use something like this. Um this is just an antenna bracket mount. Um I can use So, I can use something like this to actually get the antenna on, and it's going to look something like Let me take this this off. And we will set that up something like this. And then, now that we have this like that, we'll be able to, if I unscrew this this nut real quick, we'll be able to get this fiberglass antenna fixed onto this. That comes down like that, and then uh all we have to do is put this nut on. That will secure it in there. Okay, right? So, you have something like that. Okay? And then, uh and then the final piece is this needs to run into uh the Haven node. So, now because this is a BNC connection, uh and our Haven node is SMA female, uh we're going to need to uh convert those two. And I just have a uh BNC male to SMA male, which will work with uh with the Haven. So, apparently um the longer the cable, the more loss you have. So, try to keep that cable uh as sort of um short as you can. But, I'm not sure what that loss looks like. I'm sure it's not it's not that bad. Um okay, so what we'll do is we'll screw in the BNC connection here. Right? So, that looks like that. And then, this guy just goes right into uh into the Haven node on top, and we'll put that right down here. So, anyways, that's our mission, and this should allow us to get sort of a temporary antenna uh on the outside of my house. Okay, so first thing, I do want to grab chair, so that I can actually reach outside the house. So, I'm going to put this right here. And then, now that I'm going to get up here. Okay. And then, we want this to be like that. So, just [snorts] make sure that that is like that. Come back in here. Come back over here. And then, now I need to tension this up. All right, I think we're good >> [snorts] >> for a temporary mount. So, the next step is let's get this Let's get this antenna put in here. And then, the final step is going to be running this guy. Running the BNC connector into here. So, ideally we'd have a flat cable. We don't have a flat cable. So, this guy is just going to get squished a little while we have this going, but that's okay. So, Okay, so now we would just have to hook the Haven node up to this. So, I'm going to have to probably put it over here. Let's see if it can reach there. Yeah, should be able to reach. So, before I turn her on, that comes off. That is SMA SMA female. And hopefully, get this guy. And it's barely going to reach. Right on the edge. Okay. Cool. And then, plug this guy in. Okay, and because we turned on the Haven node, I should be able to log in over here. I can. Looks like everything is up and running. We don't have any other mesh devices right now, because they're powered off. But, I should see some information here. Yep, 915 MHz. Um and so, now the Haven green node is making use of the fiberglass antenna with the 9 dBi that is outside the house. It's still not perfect, cuz it's pretty close to the wall, but should give it a fighting chance once we start moving around the neighborhood. So, anyways, we have our antenna set up outside. It is a high gain uh high altitude, probably I don't know, 20 or 30 ft off the ground, which is pretty decent. But, to run the neighborhood test, we're going to have to have a node that is mobile, a node that we can use on the go. And for that, I have something up my sleeve that I want to show you. Which brings us to today's sponsor. I'm not going to lie, when this thing arrived at my house, I didn't really know what to expect. But, I was pretty sure I was about to experience a real-life episode of Black Mirror. But, once I started playing around with it, I was pretty impressed. This little guy is a Kickstarter that brings Open Claw and Home Assistant to a Raspberry Pi, and pairs it with natural language understanding to give you a powerful local agent without buying expensive Mac minis, or sending your chat data to Bezos or Tim Cook. Claw stage runs on Open Claw, which if you haven't heard, recently got acquired by OpenAI for reportedly around a billion dollars. So, there's definitely interest in this space. But, beyond queries, weather, and calendar integrations, Open Claw has a hub of skills. So, I decided to download a Tesla skill to see if I could control my car. Open Claw guided me through the entire process, including creating a virtual key and provisioning access. Miko, what's the battery on my Tesla right now? Smaug's battery is currently sitting at 53%. Which gives you about 190 miles of range. Plenty for now. Unless you were planning a spontaneous road trip. Miko, can you honk the horn on my car? Consider it done. So, if you are anywhere near Smaug, you should have just heard him growl. So, that's kind of wild. And yes, I named my car Smaug from Lord of the Rings. Miko gets the joke. So, now I'm controlling my fantasy dragon car from open source software running on a $65 computer. This timeline is so dumb, and I love it. Anyways, guys, uh this project is really a great entry point into Open Claw and their Kickstarter is kind of going bananas right now. So, check it out in the link in the description below. All right, back to the mission. Okay, so I may have done the unthinkable. Yes, I have put a node on my car with a magnetized antenna like this. And it was actually kind of funny because this week I was washing my car and I kind of forgot that I had this guy on here and the guy had to had to take it off. Thankfully, he reminded me. Otherwise, this thing would have end up probably with a circuit fried on it. So, anyways, the way this works is we have a magnetized antenna board here. This is currently running a kind of experimental antenna. It's a right-hand circularly polarized antenna built for the sub-gigahertz frequency. Now, we'll mess around with that later. It has certain properties. It doesn't immediately give you like way better range, but it does have some certain interesting properties particularly if you pair it with an LHCP and do MIMO. Then you can get a lot of new capabilities. But just for this test, we're going to use the standard linearly polarized antennas because if you're going to use these circular ones and you want them on both sides and the one we have on the side of the house is linear. So, we're going to stick with linear linear. So, I'm actually going to just unscrew this and this is just a standard um standard SMA female connector. And so, what I can do here is I reach into my pocket. Um we're going to go ahead and just use one of these stub antennas. This one actually works pretty well. Um it's not exotic, but it actually like when I do the RSSI reads and things like that, it's pretty good. So, like you end up with I don't know, pretty clean look. You're still running a cable into the trunk, which is probably not ideal, but um it gets the job done. And I'll show you what I got going on on the inside. So, if I just pop this. Okay, [snorts] so throw the antenna in here. I don't need that right now. So, the Tesla has a 12-V um one of these, right? So, I can just pop that in there. But what I then have is um just a USB-C. So, what I'm thinking is this will power the the Heltec that we got going on. Um that that is going to be our mobile node. And then right here, we can just see we have the SMA male connector, which will just go right into the Heltec. So, I guess let me just show you that setup real quick. So, really all we do is we don't need the battery on this guy now because um we have stationary power here. So, I'll put that to the side. But again, right? We have our This is the same device that we were using upstairs. Um it is this guy. We'll keep the power read on there cuz I like to just keep optics around um the energy. So, okay, so now we just need to hook up the SMA. Again, this is the one that's magnetized on the hood of the car. So, we'll just um get this in there like so if I can if I can get that to work. I like the right angle one cuz it just doesn't put as much pressure on um or as much strain on it. Um and then the USB-C will just go into here like so. Yeah, so it's already kicking out power. You can see like the LED going and all that stuff. Okay, so um it's already going. Now, okay, in terms of getting this connected to the Reticulum client. Like I still have my Android phone here and we could ethernet it in like [clears throat] like we did last time. Um and what that would look like is just getting Well, we would just grab this cable here. We would plug that right into the Heltec. And then we would get the USB-C side and we would just plug that into phone like so. See if this works. Okay, we just saw ethernet come through. So, um we should be good there. Let's do a real quick test. Okay, Wi-Fi man doesn't like that for some reason. Um let's just go to Amazon. See if we can get any internet. Okay, so internet is working. It's slow, but that's what we'd expect. So, so let's do internet speed test. I think Wi-Fi man just says oh, it's so slow, it's probably not working, but it actually is working. Okay, so we do have internet. It's just crawling, which is fine. So, let me show you guys another thing real quick. The Heltec actually has a 2.4 GHz Wi-Fi network. So, if I actually take this out, right? So, I just took out internet ethernet. But I go back over to network. So, we're not at the mercy of that ethernet. I just thought that was cool. But we can also, if we go to Wi-Fi, so we turn on Wi-Fi, this guy should have a local see Heltec naked. I call it naked cuz it doesn't have it doesn't have any uh any enclosure. So, look, we're already connected. And now, if we do the same thing, we have internet. So, like you don't have to use the ethernet. You can This thing can just be in the car and that's actually fine. Let's actually just do that. Let's get rid of the the ethernet cable. Um so, this guy is just going to be in the back. I'll be up front connecting over its 2.4 and then the the Wi-Fi sub-gigahertz antenna is what's going to end up doing sort of like the backhaul to our our network that actually has the internet. Hopefully, that makes sense. Okay, so before going out, I just wanted to validate that we're able to send messages right here. And we are. So, let me show you the setup real quick. So, again, this guy is connected to the green node upstairs. Connected to the green node upstairs. It's running Reticulum right here in an application called MeshChat and I'm going to say MeshChat and we're just going to send that across, right? Says it was delivered right away. Cool. Um this guy this guy is connected to the Heltec over its 2.4 Wi-Fi and you can see we just got the message MeshChat. So, they are talking over Reticulum and that's what we're looking for at this stage. Um and the antenna that it's doing that talking over Reticulum is our uh the antenna that's on top of the car. Which is what we're looking for. So, I can say you know, hello. Delivered pretty quickly. Little quirk of MeshChat is you have to um you have to like go away and then back. But there it is, right? So, okay, so they're talking to each other. I'm going to leave this computer up in my office. And then this one I will just take with me into the passenger seat and we will see how far we can send messages. And again, these messages are going over Halo at a 1-MHz channel width and yeah, we'll see how it does. Okay, so I'm little literally in my driveway. Let's just make sure that we can get a message through. Test one. Delivered. Okay, so we're still in business. Okay, so we're about 1,000 ft out. Let's see if we can't get another message to go through. Okay, delivered. Delivered at about 1,000 ft. So, that's pretty sweet. Okay, just got another message through. You can see test three. Probably at about 1,400 1,500 ft. And we're going to keep pushing it here. And yes, I have full self-driving on. So, all I have to do is pay attention. All right, looks like we might have hit our limit here. Yeah, just failed. Um yeah, we hit a limit at about like 2,000 ft. There's quite a few things you could do to like improve the range and stuff like that, but you're not going to get like orders of magnitude uh further range. Um so, I think like you get messages around your neighborhood. That seems doable even with the same setup we had before like the Wi-Fi Halo. But how do we actually get messages across a city? That's like a lot more challenging. So, anyways, that's our next objective. We got messages reasonably across the neighborhood. So, now we're going to up the ante here and try to get some messages uh to the neighboring city. And just to reiterate, guys, like the reason we're doing this here is that we have a really cool open-source protocol on our hands that's very dynamic, very capable, very flexible. It's called Reticulum. Um and frankly, there's a lot of interest there. But we're trying to figure out like practically speaking, like what could this actually look like in practice? Could it scale up to something like the internet? I mean, we know that the architecture of the uh meshing algorithm should be able to support that because it doesn't do repeat flooding and it does it it stores routes and it scales linearly and not exponentially. These are all very key things that frankly aren't true of other mesh networking technologies that are out there. Like a lot of those mesh networking technologies kind of uh they basically stop working once you get past several nodes because the overhead just gets more and more complicated and complex. So, technically Reticulum kind of excels in that regard, but you know, can we spin up this sort of open-source internet across the country that could actually go toe-to-toe with sort of the big ISP big Telco sort of legacy internet? That's the question. Theoretically, it should be possible, um but you know, we want to kind of push it to the limits here on that channel and answer that final question. And then just like why does this matter? It matters because Reticulum's kind of like the blockchain of communication. No one owns it. It's decentralized. It's community-run. It's unstoppable. It's uncensorable. Um you know, countries can make it illegal. I mean, just like I think in the US like the smart contract Tornado Cash. I think to to interact with that smart contract I think is technically illegal. So, you know, they can make legislation, but they can't physically stop you, which is interesting. Um and they certainly can't impose their agenda on you and kind of limit your action potential in the digital world, which is something that Mark the the founder of Reticulum, the creator of it, kind of emphasized and it rightfully so. So, you know, for me it's it's good to see Reticulum kind of getting the attention that I think it deserves. And I also think that you know, this is the this is the next model, the V2 of communications. It's not going to be centralized. Everything's going decentralized and and there's a reason for that. But the powers that be, the incumbents, are going to push back just like we see with some of the legislation coming across in California around like 3D printers and stuff like that. Solar, 3D printers, cryptocurrency, decentralized mesh networks, these things all threaten the incumbents. But they benefit the people. So, you know, what side of that equation do you want to be on? I think it's interesting and helpful and useful and ultimately good for humanity to um to put the power back in the hands of the people. I mean, the whole part of one of the founding tenants of America is that government should serve the people and people shouldn't serve the state, right? And you have checks and balances and you have decentralized even within the government of having different states, right? Like we don't have to have different states, but we do so that they can kind of govern themselves and then it kind of bubbles up into you know, what makes up the entire government of the United States. That has worked pretty well for for 100 years. So anyways, this kind of just takes that ethos into the technological realm and brings it into the communication space specifically. So, the main question I get from people is really just how to get started with all this stuff. So, I finally put together a course called the parallel primer that walks you through the entire tech stack, the software, the hardware, the radios, the physics and gets you capable with the technology. We build a node from beginning to end and we even use some of the most cutting-edge AI techniques to move as rapidly as possible. We remove all the fluff and just bake in all the hard-won insights that we've learned over the years. So, if you want that, check out the link in the description below. So, we're able to get a couple thousand feet, maybe one or two thousand feet around my neighborhood. I will say that all the houses in this development are cement brick, so it is pretty dense with uh obstacles for the uh signal to have to penetrate, so probably not ideal, but technically we were able to get a signal around the neighborhood. And this is Florida, so there's always something going on with the weather, which honestly just makes this whole project feel a lot more necessary. So, for the next test is the city test and we're basically trying to see if we can get a signal to go from one city to the next. Okay, so usually when I go aerial, I try to use the kite. [clears throat] I like the kite cuz it's low-tech. You can keep it up there for as long as you want and frankly it can go really high up. Um but there's literally no wind today, so we're actually going to use my drone, which should also work. So, I decided actually I'm going to bring my computer with me. So, this is Amelia's computer and this will be our base Reticulum client. It's also running mesh chat. It's connected to the Green Haven node upstairs. And then the Haven node has the Wi-Fi Halo antenna that's hanging outside the house. So, that's the network that this is on and it's we already have a conversation going with the Tesla node, so we're able to get messages through here, but now we're going to go several miles out kind of across the city and we're going to see if we can't get a message back to our base node right here. So, now we're going to run the city test and I know some of you might be thinking, well, to get you know, data across several miles like just use ham radio. And I can definitely appreciate that. In fact, there's even a project out there called AREDN, which is amateur emergency radio disaster network and that's basically the idea behind it. But the technology we're using is different than ham radio. So, ham radio first off, you need a license because you're going above the 4 W EIRP that our technology has to adhere to, but second off, you can't encrypt ham radio. So, anyone who intercepts those signals can can basically can get in on your conversations. So, your conversations are not private. Also, it's analog and it's not full internet. So, what we're talking about here is the ISM band, which is unlicensed. It can be encrypted and it's going to be full internet. We're leveraging the Wi-Fi Halo protocol, but it's the same band as things like LoRa, Meshtastic, that sort of thing. So, I think ham radio has a place and I'm definitely going to get my license and I think it'd be really cool to make a IP mesh radio that was just like 50 W or something. But we're at the mercy of keeping this unlicensed, but we also want to keep it secure and that's what Reticulum really excels with. Everything's like cryptographically complete. It's it's encrypted by default. So, it has all that sort of privacy baked into it. So, I have a script here that basically tells you the health and status of your mesh. So, right now I'm near my house. So, you see all three devices by MAC address and that beeping says that there's a good signal. You see an SNR of 73. Um but this is what we're going to be checking once we kind of go several miles out to see if we can get Okay, so now we can only see two nodes. That's because the gate fell off. We are now where we want to do our range test. We're out kind of in the middle of nowhere. We want to do our range test out here. We're going to pop the drone up. We only see two nodes by MAC address. That's the Tesla node. That's the drone. And we hope that when the drone goes in the air that we get a third node here. Um that is kind of the hope here, so we shall see. Okay, so this guy is going in the air. Should be pretty um pretty intuitive as to what's going on here. All right, with the 8650, we have another Heltec mesh node. We have a the Mouser Works 915 antenna and this is going to be able to reach my base node at home, but when this is in the air, I need a way to get sideband onto this network, so that's where the Tesla node comes in. So, we'll have a three-way mesh. Sideband will run Reticulum. So, none of these devices run Reticulum. Reticulum's just going to make use of this network and again because it's it's sort of protocol agnostic, it doesn't care that we're running everything over Wi-Fi Halo or whatever. And you can even do more exotic things with like mixing radios with LoRa and and Halo and traditional Wi-Fi and stuff like that. So, so yeah, it's a pretty cool setup and if it all works, sideband on my phone is going to connect to the Tesla node over 2.4. The Tesla node is going to connect to this over sub-GHz. This is going to connect to my base node over sub-GHz. My base node is going to connect to my mesh chat desktop over 2.4 and we should be able to get a message back to base across the city here. Okay, so we're getting node number one set up here, which is the the Tesla node. This guy is just going to sit in the car and then um we'll have another node. Let me just get this guy plugged in first. We'll have another node that goes into Okay, we see this guy booting up. So, okay. So, that's the Tesla node. Um and the second node that's actually going to give us the range to be able to reach back to our house, which is about 4 miles away, is going to be actually this guy. >> Battery level is low. Aircraft will return to the home point in 10 seconds. Okay, so yesterday we tried to get the signal at about 4 miles using the drone. The drone wasn't going past 100 ft because I needed to update the firmware and I needed to log in and by the time I did all that and then got it back into the sky, it was dying. It was like low battery. So anyways, we're back. We're going out to our spot today. But as fate would have it, we're having a bit of weather today. So hopefully hopefully there's not so much rain that we can't get up there. But what's interesting is there's wind. So I might pivot and try to use the kite instead of the drone to get our node up into the sky. So let's see how it goes. All right, and then So I want to show you real quick. We got going on here. All right, so we only got two nodes on the mesh right now. Kind of as expected. We're hoping when we get the kite into the sky though that we that we end up getting three. Okay, so you can see our kite is way up there. And what do we have here? We have our third node. So now the question is can we send a message over Reticulum? Okay, so I spent the last couple days really trying to make the the Halo test work. We put a node in the sky with the kite. I I tried all different things. I was able to get a three node mesh working. The problem is the kite would see the the home base node. It would spot it for like a second and then it would take like another 10 seconds until it would see it again, which just isn't adequate to actually send data. There's a bunch of things going on, right? Like when I was in Miami, when I was in Brickell, I would turn on these radios and I would kind of took for granted that they would just kind of work. But the difference was I was hundreds of feet in the air in a high-rise apartment. Now I'm here in a kind of normal home and I don't have the same sort of altitude. And yes, we have our antenna right there, but it's still on the side of the house. The house is made out of concrete. It's not above the house. There are other houses like in the vicinity that are kind of blocking the signal. So it's just not um as advantageous like a sort of vantage point as a a lot of the stuff that I was doing when I was in Miami. So anyways, um I think what we're going to do by the way, I think if you did a better setup with Wi-Fi Halo, you could still get the cross city test to work, but I just I don't really have the the time right now to get that working. We'll come back to it at some point. But what I'm thinking is because these antennas are all sub gigahertz like same band as Laura. I want to pivot to trying Laura. Laura works fine with Reticulum. Actually, Reticulum doesn't care what radio you use and Laura has um a lot of special properties that really make it the king of range. It can operate below the noise floor, which Halo just can't do. So it can like tease out the data as long as it can identify a pattern even when there's just tons of noise around it, which is something that yeah, Wi-Fi Halo just can't do. So we're going to try Laura and also with Laura, you have a bunch of like kind of levers that you can tweak like things like coding rate, spreading factor, bandwidth, stuff like that that you can kind of optimize the signal to go even further. So we can use the same antenna. I just need to set up a bunch of Laura nodes, which is going to take a second. And then hopefully we can get a message to go across town. Now the trade-off is we don't get quite as much bandwidth. It's not real internet, but that's fine. We're just trying to send messages over Reticulum. So anyways, we're going to get working on setting up those nodes and seeing if this doesn't solve our problem and seeing if we can't finally send a message across town. So I'm still trying to get these devices to work. So I'm trying to because I don't have the highest powered devices or even frankly like the latest chips. I'm going to try to get Reticulum provisioned on some of these devices here so that we can try to get this cross city test working. I've been having a lot of trouble, but I think it all has to do with like my setup because technically it should all work. The physics of Laura and all that. So anyways, still chugging along here, but putting Reticulum on a couple new devices here to see if we can't improve our range. Okay, so just wanted to give a little update here. So we've switched over to Laura. I have the R node running here and it's actually hooked up to our external antenna that's outside the house. So it's a high gain omnidirectional antenna. But right now I'm just doing a little like local troubleshooting. We have our other device here. This is another essentially it's an R node, a different board. I'm just messing around with different antennas, but I don't think antennas are the issue here. And then two mesh chat clients. And I'm just not getting the range that I should. I'm able to get these two to talk like it when I go out to the neighborhood, but I should be able to go much much further. So I'm not 100% sure what's wrong here. I'm starting to look into like the RSSI, which is the signal. And I might recruit some other boards. I do have some newer boards that have higher transmit power. But something and like we can also configure different things like the coding rate, the spreading factor. What's the other one? Bandwidth, stuff like that. So I'm going to keep tweaking here, but Laura should be able to go much further and like with Meshtastic, I've been able to get way more distance. So I'm not 100% sure what's going on here, but I'm going to continue to iterate here until we can get the city test to work. Okay, so I didn't have much luck in my neighborhood. I I don't know what it is. I think there you know, there's a lot of dense cement brick houses and uh yeah, it's just it's a lot of cement to try to penetrate. So what I'm going to do instead is there's a big long straightaway section in a certain part of my town. I'm going to try to set up one node on one side and then I'm going to drive the car down this straightaway and see how far we can try to get a signal. And there's no major obstructions here. Like there is a bend. So it's not perfect line of sight, but it should be better than the neighborhood I'm hoping. So we'll see. And then this time so we have two Heltec V4s, which are running 27 DBM. We've changed the parameters so that the bandwidth is 125 kilobytes. The spreading factor is 12, which is the highest. The coding rate is eight. Um and yeah, I think those are the major things for the parameters. So so we're really trying to optimize range here and we're just sending text messages. So I'm going to go place the first node and then we can drive with the second one attached to the car and see see if the messages keep coming through. We have a script on the Haven node that's driving one of the Reticulum instances. There's a script that sends a ping every 60 seconds. So as long as we get those pings, we are connected. Okay, so I have it elevated. It's probably about 7 or 8 ft. Um and I'm on a bit of a sort of berm here. So I'm hoping that I'll be able to get a signal going all the way down there because there's really no major obstructions. So um that should help with Laura not having to penetrate all this cement brick. And then on the car, we just have the Moosie whip. Um and it's going to be right on the back of the car like that. Okay, so I'm in the car. Um and I don't want to disrupt this too much, but we do have ping 23. So we should continue to get pings. So I'm going to start driving out. Okay, we are now driving and we just got another ping. RSSI was -44, which I don't think is too bad and SNR was 5 dB and that was ping 25. Okay, so we're at about a half a mile and I just got a ping. Um ping 26 at about a half a mile. Okay, we just got ping 27 at 1 mile. -85 RSSI and -3.2 SNR. So we'll see how far we can keep pushing it. So I'm going to try with the Alpha on there and then I'm also going to go in a different direction and see if we can't get more distance. Okay, so putting it on a little berm, we were able to get over a mile. The problem is if I go in one direction, then there's a school between me and between the two nodes. If I go in the other direction, then there's a factory between the two nodes. Metal always acts as a bit of a Faraday cage. So I'm going to try one more thing. So I'm not going to elevate the node, but there is a straightaway that's basically unobstructed for a long period. I'm going to try just putting one node on one end and then driving out the other end and the only potential thing that might be an obstacle is other cars on the road, but anyways, this is the last thing I'm going to try, so we'll see. All right, finally having some success. So, we just got a message at 1 and 1/2 miles. Our SSI is -92, which is not great. SNR is 3.2, which should be okay for now. And that was ping 67. Okay, we just got 2 miles. Ping 68, SNR is 4 dB, which is pretty good. RSSI is -91, so that's 2 miles with line of sight. Just got another ping at 3.4 miles. So, finally having some success. We're at -6.2 SNR, so I'm expecting it to fall out pretty much anytime now. It's just interesting, this node isn't even elevated. I mean, it's elevated on a tripod, but it's it's in a flat It seems to be a little bit more sensitive to obstruction than needing the elevation. So, I don't know, you never know with these things. There's always some element of randomness, but good to finally be getting some success here. Okay, so it seems like it really likes line of sight. We knew this, but I had had success in the past with like reflection and urban setting, so I was a little confused. And also, elevation didn't seem to help us as much as I thought it would. So, just interesting. I'm going to continue to play around with like these antennas and radios and positioning and stuff like that. Ideally, I would have liked to get one on a kite or a drone, because then you can really just like crank it as far as you want. But anyway, so we got 3 and 1/2 miles. That dropped out on my way out, then when I was coming back, reacquired signal at about 3 miles. So, I'm pretty happy with that. Is it a true citywide test? I don't know. Depends depends what city we're talking about. But, I think you get the gist there. And if you can get more line of sight, you would get more range as well. So, I feel pretty confident about that. So, I think the next thing is, how would we get it from one state to another state? Or like a large How would we get it across the country? I'm here in Florida, and if I wanted to talk to someone in California, could we use Reticulum to do that? That's the question. Okay, so with the LoRa radios, we were able to get 3 and 1/2 miles. It wasn't a perfect test, but it was just basically on an open road, so decent test. No elevation, no altitude, and really no exotic setups there. But, if we did the math on that, right? If we assumed a 3 and 1/2 mile hop using LoRa nodes, distance from here to say even California is 2,500 miles, then we would need 800 nodes to achieve that connection. And although theoretically, you could stick a LoRa repeater in the sky, like a CubeSat, to get around the curvature of the Earth, but you'd still probably need multiple satellites to cover a full coast-to-coast path. Or you can do what projects like Mesh Sat are exploring, which is to use the LoRa mesh, a gateway over USB serial, and then end up connecting to something like an Iridium modem. So, that one gateway can carry a message for the whole mesh. But again, you need to broker an agreement with Iridium, and you do have to pay for the right to use their infrastructure. Still a cool setup, I'll probably explore in the future, but not exactly something I can get up and running with the gear I currently have. At this point, I'm staring at the math and thinking, cool. All I need is 800 nodes, a space program, and a small national budget to get this to work. But then a thought came to me. Because here's the interesting thing about Reticulum. It doesn't actually care what kind of network the data travels across. So, you know, Wi-Fi, Ethernet, tunnels, LoRa, Reticulum doesn't care what the link is. And even something ridiculous like Morse code theoretically could work. Basically, if it can move ones and zeros, Reticulum can route it. To Reticulum, they're all just links. So, instead of building hundreds of radio nodes across the country, what if we just connect two distant mesh networks together with a tunnel? One here in Florida, and one in Venezuela with my editor Adolfo. But, how would we actually do that? So, to bridge that gap, we're using something called Tailscale. Now, we all know the internet is just a series of tubes. And building on that example, Tailscale is basically your own private set of tubes. Tailscale creates what's called a mesh VPN. And that just means that multiple machines can join the same private encrypted network and communicate directly with each other. Which, if we set up correctly, could potentially give us all the long haul we would ever need in our network. In fact, I already use a setup like this. So, for this YouTube channel, I've struggled for a long time to build a good workflow with my editor Adolfo, who lives all the way in Venezuela. We've used things like Adobe's frame.io, we've tried remote storage, cloud tools, etc. Moving hundreds of gigabytes back and forth was constantly breaking. So, I ended up building a NAS in my home lab, running an open-source OS called TrueNAS, based on Linux. And then I installed Tailscale on it. Now, Adolfo just connects to my private tailnet, and through some encrypted VPN tunnels and Samba black magic, the drives show up on his computer as if they were local to him. So, when I drop a file onto the NAS here in Florida, it virtually instantly appears on his machine in Venezuela, like it's sitting on his own hard drive. And so far, that's been getting the job done without issue. So, one of the really cool things about the NAS that we've set up is that I can generate a piece of content, like a video, say for instance, with my phone. So, I just hit record. I'm snapping a video right here. And what I can do is once I have that content ready to go, I can go ahead and AirDrop it over to my computer. And that uses Bluetooth, and that should show up in my downloads. So, that puts it onto my computer, but my NAS, which is over here, has a bunch of different drives. And my editor Adolfo over in Venezuela has Tailscale. So, what's cool about that is if I take the video that I just recorded, and I pull it over onto the NAS, he should be able to see that show up on his NAS in real time. And go [clears throat] ahead and play that, Adolfo. I think it's cuz it's a .mov. Oh, never mind. There we go. Okay, so it took a second there, but um So, I don't have to do anything. All I have to do is take my files, put them on the NAS, which is right there, and he's able to access and watch and edit all that content from wherever in the world, because he just connects to my tailnet. I have Tailscale running on my TrueNAS server, and it really makes our whole editing process a lot faster and a lot more streamlined. So, that's just like one example of how you could use mesh VPNs to do some really creative things. So, I thought if that works for sending videos, there's no reason it shouldn't work for connecting two disparate mesh networks. Now, to clarify, this is not the public internet. Now, technically, the packets are still traveling over existing internet infrastructure, but they're not traveling over the public internet in the way most services do. They're moving inside a private encrypted network. Think of it like a really long encrypted Ethernet cable stretching across the country, or in this case, across the world. All the messages are encrypted and addressed inside Reticulum. So, the underlying infrastructure can carry the packets, but it can't read them. And if you don't want to rely on a private service like Tailscale specifically, there are completely open-source ways to do the same things, using tools like WireGuard or Headscale, which lets you run the entire coordination layer yourself. But, the concept's the same. And what's cool is over time, those links don't actually have to be internet tunnels. We could end up using long-range radio, satellite, or thousands of community nodes. And then it's at that point that the internet becomes optional. [music] So, if everything works, a packet should leave my LoRa node here in Florida, hit my computer, travel across the Reticulum network over Tailscale, and appear on a terminal in Venezuela on Adolfo's screen. [music] So, let's try it. Okay, so this is where we are stringing everything together, and we're going to send a message locally from here in Florida, actually all the way to Venezuela. And the current setup that we have is the origin node is a LoRa Heltec node. So, we're sending the message over LoRa. We're using the RNode software. The client is Android, which is running Sideband. And that's going to send the message to my computer, which has another RNode going. But then my computer has another interface that is connected to TCP/IP. And that TCP/IP is my Tailscale IP address. Adolfo, who is in Venezuela, is also connected to my Tailscale instance. And on his computer, he has a TCP interface listening on my Tailscale IP address. Now, what you're seeing on his screen right now is our chat. But, we haven't sent any messages yet. And my goal is to send a message right from here, to hop to my RNode on my computer, to hop across an encrypted mesh VPN Tailscale, and land in Venezuela, all over Reticulum. So, you're going to find out in real time right now if we can make that happen. So, I'm going to go into Sideband here, and I'm I'm to go into my messages and I'm going to see if I can't send a message. So actually Adolfo, can you click the button announce now? Ooh, anonymous peer. Anonymous peer. Anonymous peer. See? Yeah. So that was the message here. All right. So we just sent the message sent from Florida from this Android phone that has no external network via the R node over this circularly polarized antenna to our R node over here which is connected to my computer which is in connected to our telnet instance that connected to a mesh VPN and then tunneled its way all the way to Venezuela to Adolfo's computer on his mesh chat instance all over Reticulum. So I think I'm confident to say that Reticulum is flexible, it's powerful, and it really can go toe-to-toe with a lot of the latest cutting-edge technology. And it's just really cool that everything's encrypted end-to-end. Sure it went over the internet but it went over a private instance of the internet. So it didn't go over the public internet. And the idea here would be over time we could sort of phase out our reliance on traditional ISPs. But until we're there in terms of infrastructure roll out and having satellites and cell towers and all that stuff, we can still use it to kind of complement the networks that we set up until we're able to be more independent. But yeah, it's just really kind of amazing to be able to get this to work like that. And the possibilities kind of definitely get the wheels turning in my mind about where this technology goes next. But so anyways, we got our message across my house, across the neighborhood, across the city, and now across the world. Part of the path crossed the internet but not the public internet. It crossed a private encrypted mesh VPN between two nodes which means we're borrowing the infrastructure but the network itself is ours. But replace those links with radios, satellites, or community nodes and suddenly the internet becomes optional. And if we end up going the way a lot of other countries have, having another internet might not be optional. Because a free decentralized mesh immune from state coercion is going to be pretty valuable in the coming years. At the end of the day, this channel's always been about two main things which is communication freedom and running it on hardware that you actually own. And Reticulum might be the most serious answer to both of those things that I've ever tested. And with Reticulum, we don't need permission, we just need a path. So let's make a path cuz there's a reason cyberpunk is having a major comeback right now. And it's that the barrier to entry in this space is just collapsing. Radios are getting cheaper. Open source projects are deploying AI to build applications the legacy incumbents would never be able to approve of. Devices get more capable with each passing day. And all of this has me stuck on one idea that I've been turning over for the last several months. A bold idea that will threaten some and inspire others. But nonetheless, an idea I believe whose time has come. Anyways guys, and as always, thanks for watching. For more, click here.