Full transcript
0:00Your phone is great at keeping you connected, as long as you keep paying the bill.
0:04But soon, you might be able to stay connected across continents and do it all over a network
0:10you actually own.
0:11Something the legacy telecom model still can't give you.
0:14This radio doesn't need a cell tower, and you don't need to pay a monthly subscription
0:18for the right to use it.
0:20You can modify it, upcycle the hardware, and extend what it can do.
0:24It transmits on a frequency about 100 times lower than traditional Wi-Fi, the same kind
0:29of radio people use to carry their voices across oceans.
0:32But instead of talking into an analog microphone, I'm bridging it into Reticulum, a fully
0:37digital, permissionless, open source network.
0:39I want to find out if we can connect communities hundreds or even thousands of miles apart
0:44over the air for free, and help build an internet that belongs to the people using it.
0:49The problem is, today, our communication passes through gatekeepers.
0:53They decide what we're allowed to do, keep a record of us doing it, and then charge us
0:57for the privilege.
0:58I want to build a network where nobody has that kind of power over the people using it.
1:03Which might sound like wishful thinking, but incredibly cheap, low-power technology, open
1:08source code, and community-run networks are already functioning today.
1:13Projects like Meshtastic and MeshCore have proven out that model and continue to gain
1:17adoption.
1:18In fact, a few weeks ago, I was driving north to see some family from Central Florida all
1:23the way to New Hampshire.
1:25And for fun, I stuck a magnetic 915 MHz whip antenna on the back of my car with a small
1:30Neo One listening on the public Long Fast channel.
1:34And over the course of the trip, the radio picked up messages and connected us with people
1:38along the route.
1:39Some cities like Concord, New Hampshire, were absolutely teeming with activity.
1:44And by the end of the trip, we had seen more than 450 nodes on just that one protocol.
1:49That's what this looks like when enough people get curious, put up a radio, and invite somebody
1:53else to try it.
1:54You start with something you want to understand, share it with a few friends, and little by
1:58little a community builds around it.
2:00But what happens when the people you want to reach are hundreds of miles away?
2:04That's the gap I've been trying to bridge for over a year now.
2:07Because right now, if you want long-range communication, you basically have three options.
2:12You can use cell towers, but you need a large number of them, and they're really only economically
2:17viable for big telcos and ISPs.
2:20You can use a satellite service like Starlink, Garmin, or Iridium.
2:24These are good for off-grid situations.
2:26And in my last video, I showed how you could hook an Iridium modem into Reticulum and send
2:31a text message from anywhere with a clear view of the sky.
2:35But the satellites were proprietary, you need to broker a monthly plan, and each message
2:39still incurred a toll.
2:41Or you can build a community network with something like Meshtastic.
2:44And that model is real, local, and actually belongs to the people using it.
2:49But covering serious distance requires a lot of nodes, and enough people between here and
2:54wherever you're trying to reach.
2:56But is there another option?
2:58I think so.
2:59And if this works, privately owned networks could connect across hundreds or even thousands
3:04of miles without traditional internet backhaul.
3:06But for that to have even the slightest chance of being possible, I need two miracles.
3:12Miracle number one is I need to somehow cross this, on a path we own.
3:17And miracle number two, I need to do it legally.
3:20Now the physics behind this isn't new, and neither is sending data over radio.
3:25What is new is Reticulum, and who can build with it.
3:28Radios that once cost thousands are suddenly becoming affordable.
3:32Open source networking software can stitch completely different links into one single
3:36network.
3:37And AI lets one reasonably stubborn person work through problems that once required an
3:41entire engineering team.
3:43For the first time, something like this doesn't feel completely insane.
3:47But that doesn't mean it's a great idea.
3:49After all, maybe there's a reason nobody has done this before.
3:51But really, there's only one way to find out.
3:54And fortunately for you, circumventing corporate and government overreach is one of my favorite
3:58hobbies.
3:59So I'm going to try it.
4:00One real Reticulum message, hundreds, maybe thousands of miles.
4:04No cell tower, private satellite, or corporate network in the middle.
4:08And the technology that makes that possible is one a lot of people have already written
4:12off.
4:13This is Ham Radio.
4:14And depending on who you ask, Ham Radio is either one of the last long-range communication
4:19systems ordinary people can actually own, or an elaborate way for retired men to discuss
4:23the weather.
4:25But these radios can do something your Wi-Fi router can't.
4:28They can potentially cross this.
4:30But they can carry digital data too.
4:32Small packets sent directly over the air.
4:35Could one of those packets travel thousands of miles?
4:38Potentially.
4:39But could we turn that connection into a bridge for a modern open source network?
4:42That's what I'm here to find out.
4:44But I wasn't here to chat with random operators.
4:46I wanted to take the thing making those conversations possible, the radio link itself, and make
4:51software talk through it instead.
4:53Because Ham Radio isn't limited to voice.
4:58That noise is data.
5:00It's called APRS.
5:01Small packets carrying locations, weather reports, status updates, and short messages.
5:07Nearby stations can repeat those packets, while internet gateways collect them and plot
5:11their reported positions on maps like this.
5:14So sending data over amateur radio isn't new.
5:17But APRS was built for a different job.
5:19And this particular APRS channel is on VHF.
5:23Great for local coverage, but not the long-distance skywave link that I'm attempting.
5:27I could build another one-off HF messaging system, but then it would connect only radio
5:31operators running the same hardware and software.
5:34That's why I want to use Reticulum.
5:36Reticulum is transport agnostic.
5:38It doesn't care whether data travels over LoRa, Wi-Fi, Ethernet, the internet, or a
5:44custom HF link.
5:45To Reticulum, each one is simply another path through the network.
5:49So this wouldn't create a tiny new network for a handful of HF operators.
5:53It could bridge existing Reticulum communities together, even if the people inside those
5:57communities never touch an HF radio themselves.
6:02Someone using LoRa in one town could potentially reach someone on an entirely different kind
6:06of link hundreds of miles away, with HF quietly carrying the data across the gap.
6:12That means I need to design a custom HF packet that carries what our bridge needs.
6:17If it works, this radio doesn't become another isolated system.
6:21It becomes a long-distance bridge between networks that already exist.
6:25Now, I have a bad habit of refusing to ship these projects until they're absolutely
6:29perfect, which is why I use sometimes wait months between videos.
6:33So this time, I gave myself three weeks.
6:36When Sunday comes, testing stops, whether it works or not.
6:39I have to ship what I have.
6:40So I have three weeks to build the bridge.
6:43To attempt this, I need one more radio.
6:45Which is saying something, because my office is already full of them.
6:48I have boards for Bluetooth, Wi-Fi, LoRa, Halow, even an Iridium modem.
6:54But those radios either stay relatively local, or depend on somebody else's infrastructure
6:59to go farther.
7:01This is the AirMay Lite 2, and it's the key to this entire puzzle.
7:05It's a relatively inexpensive, open-source, software-defined radio, rated to cover frequencies
7:10from 0 to 38.4 MHz, including the part of the spectrum known as HF, or high frequency.
7:17A traditional HF radio is built primarily around a person with a microphone, but the
7:23AirMay is built around a computer.
7:25It can still carry voice, but I can also have my software generate the exact data frame
7:30I need, put it directly on the air, and decode it on the other side.
7:35And HF has a trick that none of the other radios scattered around my office can perform
7:40on their own.
7:41Under the right conditions, an HF signal can travel into the ionosphere, reflect back toward
7:47the earth, and land hundreds or even thousands of miles beyond the horizon.
7:52It's called skywave propagation.
7:54There is no chain of community nodes or private satellite carrying the messages.
7:59The atmosphere itself becomes part of the path.
8:02So this might be the right radio that lets us cross this map.
8:06There was just one minor problem.
8:08Mine arrived in three pieces.
8:09Okay, so let's try to put this guy together.
8:12So we have the enclosure, we have the main board, and then we have the filter board,
8:19which we're also going to need.
8:20So three parts, and that should be all we need to put this guy together.
8:25So let's go ahead and open this up.
8:27We slide the board into these rails here.
8:30I kind of like that.
8:32So basically this goes on like this.
8:36There's nothing holding it together.
8:37What's going to hold it together is the face, the front face plate, which has the four screws
8:42in the corners.
8:43So this part, this is the filter.
8:45This is kind of like the sideboard.
8:47The tricky part that I've seen is on the main board.
8:52And this is our board right here.
8:54And this thing is interesting because its connection is an ethernet.
8:58So it shows up as a network device line.
9:01But what I'll do is I'll push the screw out just, I should be able to do it like this.
9:07There it goes.
9:08So that's our first half right there.
9:10I believe all we do is slide it in the rails.
9:15So now and then these connect through the head pin here, doing one of these.
9:21I don't think there's a direction.
9:23Apply pressure evenly here.
9:25The other side of the enclosure actually doesn't slide.
9:28It just, you can go right on top.
9:30And I think we're going to be good to go.
9:32No, I think we're good though.
9:35This thing is, is pretty solid.
9:38We've got it set up here.
9:39Good to go was perhaps slightly optimistic because while the Airme was now capable of
9:45transmitting, I still wasn't legally allowed to press that button.
9:48Remember miracle number two, I have to do this all legally.
9:52Transmitting data on this part of the amateur radio spectrum requires a license and I didn't
9:57have one yet, but receiving is different.
10:00Anyone can listen.
10:01So while I studied for the ham radio technician class license exam, I started learning about
10:06the receive side of the radio.
10:08And a lot of that studying came from two YouTube channels, ham radio crash course and ham radio
10:132.0.
10:14I basically had their videos playing in the background while I tinkered with the rig.
10:18So thank you to both of them.
10:20The first thing I learned was that the radio itself was only half the problem to hear anything.
10:25I needed the right antenna.
10:27The 900 megahertz radios I normally use have a wavelength of about 13 inches, but at 28
10:33megahertz, the wavelength is roughly 35 feet.
10:36The antenna doesn't have to be a full wavelength long, but it does scale with it, which is
10:41why this thing has a much longer antenna.
10:45This is a Falcon dipole, basically two lengths of wire size for the band.
10:50Ideally it would be outside high in the air with plenty of open space.
10:54You'll notice my installation took a slightly more interpretive approach.
10:58Okay, so this thing is a nano VNA and it helps you understand how effective your Intuna setup
11:04is.
11:05This thing intimidated me for a long time.
11:07You can see there's a lot going on here and it's not clear what everything does.
11:12So I'm going to simplify it for you.
11:13So what we can do is we can remove some of these traces, display trace.
11:20We can get rid of this, we can get rid of this, and we can get rid of this.
11:25What we really care about is the SWR, which tells us where the center frequency is for
11:30the antenna and how well tuned it is at that frequency.
11:35So, and we do need to specify a range, but basically what we can do is we can take an
11:39antenna like this.
11:40So this is a Muziworks 915 megahertz whip antenna.
11:46So what we can do is we can just screw this on like so, and it's going to start to give
11:51us a readout here.
11:53We have the end of the range here at 900.
11:56So what we're going to want to do is move this so that we can actually see what we want.
12:00So we'll do start at say 850 megahertz and then end stop at 960 megahertz.
12:11Let's see here.
12:13Okay.
12:14So we can start to see the center frequency.
12:16So now if we want to see like exactly where that dip is, we just drag this over, bring
12:22it down to the dip.
12:24So you can see, you can see it moves a little bit like based on like how, you know, how
12:28we have this thing set up, but we can see it's around 900 megahertz, right?
12:32One other thing I'll show you is you probably have some tri-band antennas or dual band antennas.
12:39This right here is one of those.
12:40I just had to get a special connector for this, but when you put in an antenna that
12:45is dual band or tri-band or what have you, you'll get multiple sort of like resonant
12:50zones.
12:51If I'm transmitting on 28 megahertz, is my antenna actually tuned around 28 megahertz
12:57and does it have an efficient reading?
12:59Yeah, these things aren't that intimidating.
13:01You can get better results by calibrating it each time with these three sort of dummy
13:04loads right here.
13:05You pop it on, go over to calibrate, and then you hit calibrate and you have open short
13:11load and you just do all three of those and it'll calibrate it.
13:14So pretty easy, definitely a useful tool in my toolkit to see like how effective an actual
13:19antenna is.
13:20And it's been interesting.
13:21Now, one of the cool things is you can get that to move based on like moving the arms
13:26of the antenna.
13:27So like see how the center right now says 26.840.
13:32Well, if we see one is clipped over there and one is clipped over there.
13:39If I move, like let's unclip it real quick.
13:43And I just tossed it down there.
13:46And now you can see it's much deeper just by moving it, right?
13:50But the problem is we want the center frequency to move to 28.
13:54That's the problem.
13:55Yeah, you can see how like moving the arms affects the shape there.
14:03So like now it's more towards 28.160, which is where we're going to transmit.
14:08If you zoom out a little bit, like you can see it's relatively flat, flat, flat, flat.
14:13But then when we get to our 28 megahertz, we're dipping down.
14:16And that's basically what we want.
14:18So that's really just the process that I use to get these antennas tuned properly.
14:22But it was connected and somehow it started hearing things.
14:26So this is my first day with the Airme and I have Spark SDR going here and I didn't get
14:34any voice.
14:35Well, I did get a little bit of voice, but not a whole lot when I started playing around
14:39with it.
14:40But I'm listening right now on this band here, 21 megahertz.
14:46The channel that I'm ultimately going to use or the frequency I'm ultimately going to use
14:49is 28 megahertz, which is what my antenna is tuned for.
14:53But nonetheless, when I listen on this frequency and decoding FT8 data packets, I see basically
15:02a whole conversation going on here.
15:05FT8 is a digital mode built for what's called weak signal communication.
15:09It makes the same basic trade-off as LoRa, a very low data rate in exchange for working
15:14at extremely low signal to noise ratios.
15:18Sometimes recovering messages from signals you can barely distinguish from the noise.
15:22It's basically lightweight text messaging over HF.
15:26And the first message that I got was actually from Bogota, Colombia over HF.
15:33And this right here is what a message looks like.
15:36It's like a thin line goes out for several seconds and then it gets decoded and then
15:42it gets turned into the actual message here.
15:44And this is like the call sign, the station, and it's basically saying like received, and
15:50then it's giving a DB, like received at negative 14 DB.
15:55And you have the timestamp and all that stuff here.
15:58So like my first message I received from Bogota, Colombia, I'm here in central Florida.
16:06That's 1600 miles.
16:09The air may rate here, the antenna setup I have is pretty janky.
16:14Let me show it to you.
16:17So this is the antenna.
16:20This is called a center fed dipole antenna.
16:23It's basically two wires that go left and right and you can do horizontal, vertical,
16:28L-shaped, et cetera.
16:29But you can see one of the arms is literally just in my office right here.
16:35And then the other arm is dangling off out the window.
16:39So definitely not an optimized antenna rig by any means.
16:44It should be outside.
16:45This is a cement brick house.
16:46Say hi Bruno.
16:47This is a cement brick house, so not ideal.
16:50But like we see, like we already see more messages coming through here.
16:54Like we see this data being sent and that's the whole powerful aspect of that.
17:01Like the only reason the message can hit us from Bogota, it's blocked by the curvature
17:06of the earth.
17:07The only reason it can hit us is because it bounced off the ionosphere.
17:12And that's kind of like the interesting characteristic of this band.
17:16This band also like makes use or is very sort of affected by the solar environment and like
17:23flares and nighttime versus daytime and other factors like that.
17:27But you can see it's pretty active right now.
17:29Well, that's miracle number one.
17:31Once FT8 proved the radio could hear a message from 1600 miles away, I started turning the
17:37dial.
17:38And at 4.84 megahertz, I found an AM voice broadcast, most likely WWCR out of Nashville,
17:47about 727 miles away.
17:57But they were transmitting at 100,000 watts.
18:00So they were definitely doing most of the work.
18:03Either way, receiving clearly worked.
18:06Now I needed to earn the right to transmit.
18:08So now we're going to start messing around with like transmitting and what data packet
18:13we want to use.
18:14We might not use FT8.
18:15You can see these messages are very short.
18:18That's probably too short for our needs.
18:20So we're going to need something with a little bit more bandwidth.
18:22So I'll start sending, coming up with that sort of frame shape and then start sending
18:28some data and we'll listen with this guy right here.
18:31This is just an RTL-SDR and then we'll work on establishing range and stuff after that.
18:38I had heard a message travel 1600 miles, but listening was still the only thing I was legally
18:44allowed to do.
18:45If I wanted to send our own packet, it was time to take the exam.
18:49And by this point, I had been studying for a few weeks.
18:52Now you can take these tests online, but I guess I wanted the full experience.
18:56So instead I drove down to Miami to Unified Radio Group, a place they call The Farm, where
19:02they were having a meetup and doing in-person exams.
19:05Okay, so I am at a place here called The Farm.
19:09Basically I'm here to get my ham radio license or at least take the test.
19:13Not that nervous because you can literally take these online, but it would be nice to
19:17knock this out in one go so I can actually use this.
19:20And this will just allow me to make use of other bands and it'll allow me to use higher
19:25transmit powers on the existing bands I've been using, which is like Laura and Halow.
19:31So I can go beyond the 4-watt total EIRP.
19:34There will be some things to consider around on encryption, stuff like that.
19:37But for the most part, this will allow me to do a whole lot of other neat and interesting
19:42experiments as well as potentially create community nodes that people can make use of
19:47that make use of longer range signal.
19:52So hopefully I pass my test.
19:54We'll see.
19:55So I'm excited for the exam to start.
19:56One of the guys offered to walk me through a rig he had set up that wasn't too dissimilar
20:01from what I ultimately wanted to do.
20:03It was called a VARA gateway.
20:04So the one with the coax coming down is for HF.
20:07Yeah, the one with the box right there.
20:10Okay.
20:11Yeah, yeah, yeah.
20:12And then that antenna way up there, that's for the VARA thing.
20:17Okay, so star link.
20:19Does VARA need internet?
20:20Yes.
20:21Oh, because it needs a gateway.
20:22Yeah, it is the gateway.
20:23What we run here is the gateway.
20:27What we connect to connect from doesn't need internet.
20:31And that's for FM.
20:32VARA also works like globally.
20:33So it's kind of dirty in here, but that's a little VARA station.
20:39Okay, so the data goes over VARA and then it bridges to the internet through star link,
20:44right?
20:45Yep.
20:46What kind of like, like images, voice?
20:48It's very subtle.
20:49Like you don't want to send anything over a meg.
20:50And that's like really pushy.
20:51I see.
20:52It'll take a couple minutes.
20:53I see.
20:54Okay.
20:55And it depends on like, there's audio levels that you adjust on this mode to make sure
21:00that it's not overdriven, underdriven.
21:02And then of course the signal strength.
21:04And so can you control that remotely or do you have to be on site?
21:07I have like remote access software to the computer.
21:11Just so I don't have to come here.
21:12Yeah, exactly.
21:13So you can transmit from wherever technically, right?
21:17Yeah, but this isn't going to be transmitting.
21:19This is just like listening.
21:20It'll transmit.
21:21Oh, okay.
21:22But it's just monitoring one frequency.
21:24Oh, okay.
21:25And then when I connect to it from say my house, it picks up.
21:28It's like, oh, they're calling me because it's attached to a call sign.
21:32The demodulator side over here demodulates that data says, oh, they're looking for me.
21:36Let me respond.
21:37I acknowledge your request.
21:38Okay.
21:39Send me the data back and forth, back and forth.
21:41Okay.
21:42Okay.
21:43You know?
21:44So, and what kind of range can you get on that?
21:45With this, it's line of sight.
21:47So it really is super dependent upon like the station that I'm trying to connect to.
21:52I've connected to it.
21:54I mean, if I got a bad-ass antenna that's up like on my tower, I could probably get
21:57like 50 miles, you know?
21:59So pretty far.
22:00Yeah.
22:01Are there enough other Vara like nodes to like...
22:04Not around here.
22:05That's why I set this one up, but there are some as you go north.
22:07Okay.
22:08So I think Port St. Lucie, a little bit north, maybe a little south, there's some there.
22:11What you can do is I could digi-peat from this one to another one to another one.
22:14Digi-peat, okay.
22:15I think you're limited at two hops.
22:16Okay.
22:17So we can hop from here to there and then we can have communication back and forth.
22:21Cool.
22:22That's awesome.
22:23I'll show you the website where it shows all the different nodes.
22:25Okay.
22:26So heading back, just got out of the meeting.
22:31I took the test.
22:32I passed.
22:33Very happy about that.
22:35But I got to say, the group was awesome.
22:39You know, there was UHF, VHF.
22:41They were running a bunch of tests.
22:43They had active gear going, but they had some mesh core nodes.
22:47So some people had some LoRa nodes.
22:48One of the guys showed me the setup he did where he connected a Starlink to a VHF radio
22:53and uses this thing called Vara to be able to send data over like analog radio and stuff
22:59like that.
23:00So really cool stuff.
23:01Two of the guys had seen my videos.
23:04So that was really interesting.
23:05So yeah, I'm going to go back for sure in the future.
23:08I grabbed some contact information and I'll be able to loop these guys into my projects.
23:12I'm located a little bit outside Miami now, so it'll actually prove as a cool kind of
23:18range test to be able to see what we can get between each other.
23:21But I'm going to loop them into my projects and just be able to kind of bounce some ideas
23:24off them.
23:25But great group, really nice turnout, and just I think a lot of subject matter expertise
23:30among all the participants.
23:32So pleasantly surprised with the turnout there.
23:34Happy to get my license.
23:36That is a bit of a relief.
23:37Hopefully the FCC will send me my call sign in a day or two.
23:41And I will be able to start transmitting and get on the air, as they say.
23:47And yeah, I'm really going to be focusing on the digital component of this stuff.
23:50I'm not going to be doing UHF/VHF so much.
23:52I'm going to be doing things like APRS, overdriven LoRa, overdriven Halow, decrypted, and maybe
24:01some HF bridging with reticulum and just stuff that's really never been done before.
24:06But I would need to do it the right way.
24:08Hence me trying to get my license today.
24:11Anyways, overall, really good experience.
24:14Definitely would encourage you to get your license if that's an option for you.
24:18Oh, and I forgot to mention, something that was really interesting is, so they have this
24:22kind of gazebo and music playing and stuff.
24:24But at one point, the power went out.
24:26But the VHF radio they were using was hooked up to a Life PO4 battery.
24:33And they were still transmitting over to Key Biscayne, which is pretty far away.
24:37The power went out, so it was literally a grid down scenario.
24:40And they were just talking over the air.
24:44So really kind of demonstrates that use case for hurricanes and all that sort of thing.
24:50So it was just really cool to see that in action.
24:52So thank you to Unified Radio Group of Miami.
24:54And a few days later, I got a very welcome message from the FCC.
24:58I now have an official call sign, which feels really good.
25:02But it also means that anything I attempt over the air will now be associated with me.
25:06So I have to make sure I'm doing things the right way, or I could be kicked off the air.
25:11We've shown that the physics are there.
25:13It's not easy, but text messages can travel thousands of miles.
25:18The second half of this challenge is going to be keeping everything legal.
25:21And it's not obvious how we're going to do that.
25:24As I mentioned before, we're not interested in analog voice.
25:27We want to send data.
25:29Audio is a waveform.
25:31Data is a package of bits.
25:33With the license I have, if we want to use HF for skyway propagation and send data, it
25:38has to happen between 28 and 28.3 megahertz.
25:42This is interesting because a lot of my work on ISM band projects like LoRa and Wi-Fi Halow
25:47have been around 900 megahertz.
25:50But 28 megahertz is far lower.
25:52And the longer wavelength should improve penetration and refraction in addition to providing atmospheric
25:59skipping.
26:00So we can use that frequency for our package transmissions.
26:03But we also need to make sure our frames are designed around amateur radio rules.
26:08They need to include my call sign.
26:10We can do that.
26:11The codec has to be public.
26:13We can also do that.
26:14But we're forbidden from obscuring the meaning of these communications by any means, which
26:19obviously includes encryption.
26:21That felt a bit like we were dead in the water because Reticulum doesn't merely use encryption.
26:27It's built out of it.
26:28Reticulum's creator Mark put it this way, "In Reticulum, encryption is gravity.
26:33Cryptographic identities tell the network who destinations are.
26:37Signed proofs validate paths.
26:39And encrypted packets are part of what allows information to move across multiple hops without
26:44trusting the nodes in between."
26:46So removing encryption wouldn't give us a less private version of Reticulum.
26:51It would break the machinery that makes Reticulum work.
26:54And this is where I think most sane people would throw in the towel.
26:57But while I was studying, I came across one interesting exception.
27:02Amateur radio allows encrypted commands when they're being used to control a satellite.
27:07And that makes sense.
27:08You don't necessarily want anybody who hears your command link to be able to take control
27:12of your spacecraft.
27:13So either every message we send also slightly adjusts the position of a satellite, or we
27:18find another way.
27:19We're going to find another way.
27:20And that's when I realized I don't actually need to send the encrypted Reticulum packet
27:24over HF at all.
27:26At least theoretically, you could terminate encryption at the RMA gateway.
27:31Messages could be addressed to that gateway, which would convert them into an amateur radio
27:35frame and send them over HF with a destination LXMF address.
27:40Anyone who heard the HF message could then bring it back into Reticulum and forward it
27:45to its final destination.
27:46This has obvious drawbacks.
27:48We're not going to solve spoofing in this video, although cryptographic signatures could
27:52still prove mathematically that a message originated from a specific sender.
27:56But more importantly, the long haul HF leg would be transparent and open for anyone to
28:01read, which kind of felt like taking two steps backwards.
28:04But honestly, maybe that's an acceptable trade off for now.
28:07I mean, Meshtastic has public channels where the private keys are widely known, and those
28:11channels still have utility.
28:13APRS does something similar when it sends short audio bursts that get decoded into messages.
28:19In fact, amateur radio is fundamentally open communication, and people still use it every
28:24day.
28:25Because if you're stranded on a boat or lost in the mountains, do you care if someone
28:28can hear your SOS?
28:30No.
28:31I think it's still useful.
28:33So I fired up Cursor, launched an agent, and started drafting the architecture using Crosstalk
28:38as the end-user application.
28:40Now that sentence compresses about five days of work.
28:43The codec went through multiple versions as I changed the modulation and symbol rate,
28:48started error correction, built benchmark tools, and ran test after test over the air.
28:53But now that we're transmitting, we need a way to listen so that we can test.
28:57And to do that, we're going to use this device right here.
29:00This is an RTL-SDR.
29:02Mine costs $37.
29:04It can't transmit, but it can listen.
29:07And for this side of the bridge, that's all we need.
29:10Now it doesn't have a proper antenna yet, but that's okay here in the office because
29:14it's so close to the station that it should still hear our transmissions.
29:17We're going to use it to listen for the transmissions coming from the air may.
29:21And what's awesome about this is that you don't need a license to listen.
29:25Theoretically, a user could have a reticulum gateway that listens for our HF traffic and
29:30acts as an ingress point, bridging any messages it hears back into the reticulum network.
29:35This is similar to how websites like ADS-B Exchange work.
29:39So we already know listen-only nodes have utility.
29:42For testing, I'm going to connect this to one of my prototype Haven builds, which is
29:46really just a Raspberry Pi 5 built for field work like this.
29:50We have our initial data package set up, and it's going to look like this.
29:54We'll be able to choose the transmit power that we use to send it.
29:58But now comes the tricky part.
29:59We need to get this working with reticulum.
30:01Now I'm going to use crosstalk, which is a mesh chat fork I created so I could build
30:06an experiment with it.
30:08Reticulum lets you create an interface over anything that can move ones and zeros.
30:12But if we create a native interface with the RMA, it will move encrypted reticulum frames
30:17directly over HF, which isn't legal here in the United States.
30:22So instead, I built a section called Bridge Extensions, and it lets you create a connector
30:27for either the RMA or the RTL-SDR that runs in parallel to reticulum.
30:32If the RTL receives a message, the connector reads it, brings it back into reticulum using
30:38the native frames, and forwards it to its destination if a route is available.
30:43Transmissions work in the other direction.
30:45I can send a message in crosstalk.
30:47If I have an HF gateway available, no direct path to my destination, and the message isn't
30:52sensitive, I can hand it to the gateway node.
30:55That gateway reads the reticulum packet, writes an amateur radio frame, and throws it over
31:00HF.
31:01If a receiver hears it, that receiver turns the open message back into an LXMF message
31:07and forwards it through reticulum.
31:09Now there's definitely some magic sauce under the hood, but everyday users shouldn't
31:13need to deal with that.
31:15It should mostly just work.
31:16Because most people are never going to own an HF radio.
31:18They shouldn't have to.
31:20If one person on your mesh has the radio and they're licensed to transmit, then the whole
31:25mesh has a long hop.
31:27It's the same idea as a repeater or a digipeter.
31:30You don't all carry the tower.
31:32You just have to have reach to the person who does.
31:35So we're going to attempt to send a message.
31:37Before we can transmit, we do need to publish our codec.
31:41So I'm going to add that to the crosstalk github repo.
31:44There's going to be a markdown file here.
31:47And you can see it just gives basically the recipe to decode one of these messages.
31:52So that if someone wanted to, if they record these and later want to read them, they can.
31:59And I have crosstalk.
32:01And this looks good.
32:02So I'm going to go ahead and generate a comment.
32:05OK, we're going to go ahead and commit.
32:07I'm going to push to origin.
32:10And I'm going to pull it up on github.
32:12Crosstalk updated just now.
32:15And then it should be under docs.
32:19We're going to go to HF codec.
32:20OK, so it's literally public now, the decode process here.
32:25So that is part of keeping this legal.
32:28OK, so we are going to run our first end-to-end reticulum test here.
32:35So basically, we have four different nodes.
32:38And I'll show you a network diagram.
32:40It's a little bit of a complex setup.
32:42Node number one is just this Android phone.
32:46And it's just running a reticulum client called Columba.
32:49And this only has one interface.
32:52And that interface is our public TCP internet interface, rns1.buildwithparallel.com.
33:01So this is a very vanilla reticulum client right here.
33:06And this is where we're going to send the message to.
33:08So this is going to be the final destination message.
33:10Now this guy has a path to this guy, which is also running a reticulum client, except
33:16it's something we created called crosstalk.
33:19So this guy is our Raspberry Pi.
33:23And you can see its crosstalk instance over here.
33:26And what it has for interfaces is also the public rns1 internet interface.
33:32So that's how the Pi can talk to Columba.
33:36But what's interesting about the Pi is it also has down here the bridge extension for
33:42the RTL-SDR.
33:44And it's set up to ingress any traffic into reticulum.
33:49This runs alongside reticulum.
33:53It doesn't run a native interface because again, we want it to run as an extension.
33:58If we put native reticulum frames over HF, that's going to be illegal.
34:03So what this guy can do is it can talk to anyone who has a connection to rns1, but it
34:08can also listen for our HF traffic.
34:12And if it hears anything, it can ingress it into reticulum.
34:16So these are one mesh, these two devices here.
34:19Now the second setup is the Airme, which is going to be right over here.
34:26Again another crosstalk instance.
34:28And what this has for an interface is only one interface, which is a local TCP interface.
34:37So that local TCP interface is going to be what it uses to talk to another instance of
34:43crosstalk, which is going to be where I send the message from.
34:47So what I want to do is have a regular reticulum instance talk to another regular reticulum
34:54instance where neither of them have any of this HF stuff going on.
34:59And also they can't talk directly to each other.
35:01Their only bridge would be a shout over HF.
35:05So this is going to be where we send the message from.
35:07And again, it just has like a local TCP that allows it to connect to the Airme.
35:12Now the Airme has a bridge extension down here where it can egress HF traffic.
35:18And this is where all those configurations are held right here.
35:22And so what we're going to see is because the origin node here, because it has access
35:27to an HF gateway, when I open up the conversation here, it's going to give you this little toggle
35:35where you can elect to do a public broadcast.
35:38When you toggle this, you say, "Hey, if there's any HF gateways and I try to address an entity
35:43that is not on our mesh, try to do an HF transmission."
35:48This is an optional thing.
35:49You don't have to turn it on if you don't want to, but this is how we signify to crosstalk,
35:53"Hey, send this to the gateway and have the gateway throw a Hail Mary over HF."
35:59I should be able to send a message from here and hopefully get it over here.
36:05So I'm going to give this a go.
36:07We're going to say, "Crossing Tim's office."
36:15And when I send this, what we should see is, or what we should hear rather, is the Air
36:20May keying.
36:21Because, again, this reticulum instance is going to hand it to the node that has an HF
36:27gateway.
36:28And the HF gateway is going to toss it over HF.
36:31The RTL should hear it and decode it and then ingress it back into reticulum.
36:35Here it goes.
36:38Okay, we can hear it keying.
36:42And again, it takes like seven seconds to get the message across.
36:49Okay, so it just stopped keying.
36:52That means it just transmitted the whole frame.
36:56And if we come over here, we can see there's our message.
37:06So we just sent a reticulum frame over HF legally between two nodes, basically a reticulum
37:15instance on my computer and a reticulum instance just on this Android phone that do not have
37:23a path between them.
37:24And we just made that work.
37:26There's a lot that goes into these bridge extensions and things like that, but it's
37:29possible.
37:30And we kept it legal because the frame is sent over plain text.
37:35When we check this, we're basically saying, "Hey, if there's no path, then actually address
37:41it to the gateway node so that the gateway node can decrypt it, repackage it, and toss
37:47it over HF."
37:49And when it tosses it over HF on the dipole antenna that we have over there, the RTL listens
37:55for it.
37:56And when it hears it, it says, "I just heard an HF frame, it's a reticulum frame, and it
38:00has the final LXMF address."
38:02And then it wraps it into a new proper reticulum packet addressed to right here.
38:09Yeah, so that's how we're able to relay a message over HF.
38:13Well, that's miracle number two.
38:16We created an HF frame designed around amateur radio rules that we can use with reticulum.
38:22And yes, learning enough about this space to even attempt a project like this took a
38:26lot of work.
38:27So I put those hard-won insights, the hardware, software, radio systems, and how they all
38:33fit together into Parallel Primer, a course designed to help you start building your own
38:38infrastructure.
38:39And I'll leave a link to that in the description below.
38:42But this message crossed only my office.
38:44Now we have to take it into the field and find out how far it can actually go.
38:48I'm trying to cancel the middleman.
38:50Apparently, there's some assembly required.
38:53Across my office, the mismatched antenna didn't matter much.
38:56The radios were practically sitting beside each other.
38:59But if we wanted to test any meaningful distance, we needed an antenna actually designed for
39:0428 MHz.
39:06And since I wanted to drive around while the RMA stayed in my office, I grabbed a 10 meter
39:11mobile whip antenna for the RTL.
39:13It uses a few magnetic mounts to attach to the hood of my car.
39:17I spent a few days messing with this thing and using this nano VNA to understand how
39:22it was tuned.
39:23Go to display trace.
39:25Kill off this.
39:26Kill off this.
39:28Kill off this.
39:29And then let's put this in the right range.
39:33So we're gonna do stimulus.
39:34Start 20 MHz.
39:37Stop 40 MHz.
39:41Okay.
39:42So now we can see that dip, right?
39:46So we want to see where that dip actually is.
39:54So it's around 29 MHz.
39:57It's not ideal.
39:58We want 28.
39:59Sometimes changing the coax.
40:02So I just adjusted the coax.
40:04Now it's deeper, but it's still around that 29 MHz area.
40:12I think that'll be okay because look, if you come over to 28 MHz, well, you can see it's
40:17not ideal, honestly.
40:19Like we would want that dip to be closer to 28 MHz and that's something I can mess around
40:23with by adjusting like the length of it.
40:25But I think for now for our testing, I'm just going to go with this.
40:29Just not exactly where we're going to be transmitting, but I think that's okay.
40:32I'm going to be honest.
40:33I don't know whether I have this ground plane set up correctly.
40:36I just know it's closer to 28 MHz than the LoRa whip antenna.
40:39And on the feed line, the shield and the core have to stay apart.
40:42At first they weren't, but now they are.
40:45So this right here, that is actual core.
40:49So this is actually going to carry the signal.
40:52And this bolt right here is also core.
40:55So this part is also core.
40:57So it can't be touching the ground plane, which would be this or the shielding underneath.
41:02And that's why we have this plastic washer right here.
41:05But let me show you the core even further right here.
41:09Yeah, I don't know if you can see it, but there's a, there's an insulating ring of plastic
41:14around the core that separates the nut underneath.
41:18And then you use this to keep them separate.
41:22So this comes down like this and it keeps that core off of the ground plane essentially.
41:29And then this comes and secures that down.
41:32And I didn't have that right at first.
41:34And so the signal was all over the place.
41:35So now you have a proper kind of ground plane, which is the shell, the coax, this, potentially
41:40even the car itself.
41:42But if you touch this, you mess up the signal because this is, this is core right here.
41:46Yeah.
41:47So to mount this, all we do is grab this and then screw it on like this.
41:56Okay.
41:58And then we have our whip.
42:00The Haven will be in my car, listening for the shouts from the Air May back in my office.
42:04The Air May is on the second floor of my house.
42:06So it's somewhat elevated, but it's still far from ideal.
42:10The homes in this neighborhood are concrete blocks.
42:12So our Air May is caged inside one of them.
42:15And we'll be driving through a neighborhood with dozens more between us and the transmitter.
42:20But it's what we have for now.
42:21We also want to be meticulous about logging distance.
42:24So I'm adding this GPS receiver, which we can put on the hood.
42:28After each range test, we'll know exactly how far the message traveled.
42:32So in the driveway, the message was working.
42:34Then came the first drive.
42:36Failure.
42:37As soon as I got down the block, this thing died, but it was hard to tell whether the
42:41problem was the signal, the packet fidelity, or the decoder itself.
42:45This system isn't just about getting the message out.
42:48The message needs to arrive clearly enough for the Pi to decode it.
42:51So I went back numerous times, getting a little more distance with each test.
42:56Now we're outside my neighborhood.
42:58But it's still failing when we enter the next neighborhood.
43:01At 1 milliwatt, it worked.
43:0310 milliwatts worked too.
43:04But eventually we hit another wall.
43:06The Air May can transmit up to 5 watts, so I still had plenty of power available.
43:11Or at least I thought I did.
43:13Instead of more range, I got this.
43:19Above 300 milliwatts, the radio became unstable and packets stopped getting through.
43:24And that led me back to the cheap power brick I had grabbed from my office.
43:27OK, so for the Air May, we had been using this power volt, because this is what I have
43:33around the office.
43:35It is a 12 volt, 3 amp, 36 watt, which should be adequate for this.
43:44This guy says it wants 2 amps, continuous.
43:48But it's not, because basically when I set the TX, like the transmit power, above 300
43:54milliwatts, this guy starts to fail and the messages don't get through.
43:58And this guy can do a total of 5 watts, so I'm really losing a ton of power there by
44:04not being able to go above 300 milliwatts.
44:06So I bought this guy.
44:09This is more of a true ham radio power bank.
44:13I think this guy is rated to, let's see here, yeah, amperage.
44:195 amperage constant, 7 amperage surge.
44:23So this guy should be a little bit better equipped to provide the power that the Air
44:27May needs to really transmit at those higher transmit powers.
44:31So that's what I'm hoping for here.
44:34So the only thing about this guy is we have to get a barrel jack attached to these guys.
44:40I'm not 100% sure how to do that.
44:43So I tried the new power supply and it turns out we're still not able to use more than
44:47just 2 watts without current issues.
44:50Which is funny because AI specifically told me this was the power supply to buy.
44:55Then when I started troubleshooting with it, it told me I needed another power supply.
44:59So AI isn't always right.
45:01I'm out 60 bucks and I don't really want to buy another one right now.
45:05So we're going to see what we can do with what we have.
45:07Okay, so we're out here doing another test.
45:09When I transmit, it takes several seconds to finish sending the frame.
45:13Then the Pi attempts to decode it.
45:15The decoding is computationally heavy.
45:18It was taking a long time and failing pretty often.
45:21But I had an idea.
45:22If you're in our subreddit, Modern Radio, you may have seen Ken, also known as Frosty,
45:28building PRNS, a high performance implementation of Reticulum written in Rust.
45:33It now even has a browser flasher at prns.dev.
45:37So that gave me the idea to try the same thing with the most computationally expensive parts
45:41of the decoder.
45:42I had Cursor port them from Python into Rust, which should run faster.
45:47And that actually worked.
45:49Decode times dropped to just a few seconds.
45:51The office test was easy to control when I could stand next to both radios.
45:55But for the range test, the Airme had to stay home while I drove the receiver around.
46:01The first few times I would queue a message with a delay, run out to the car, drive the
46:05receiver to a test point, and wait for the Airme to transmit.
46:09Then I had to drive all the way back to my office just to find out whether the Raspberry
46:12Pi had decoded it.
46:14Every failed test meant another complete round trip, which is not a particularly efficient
46:19way to debug a radio.
46:21So I built another app.
46:22So I actually came up with an app here.
46:25And so this app runs on the Pi.
46:28And it uses some MeshVPN dark magic to basically make it so that we can read RTL-SDR as transmissions
46:38come through.
46:40But we can also control the Airme and adjust parameters and send messages like ad hoc.
46:46So like I could do like a 10 milliwatt message.
46:48I could do a 5 watt message.
46:50I'll do 1 milliwatt because both devices are in my office right now so it doesn't need
46:53much power.
46:54I can adjust the message and say like, I don't know, test.
47:00And then also, down in advanced here, I can adjust the gain on the RTL-SDR.
47:05So because the RTL is in the same room, we're going to do a gain of zero and it should still
47:09be able to hear the message.
47:11So like now when I drive out, I can send a message and it will tell me if the message
47:17got through or not.
47:19And if it was decoded properly by the Pi, if the fidelity was high enough that it could
47:23decode.
47:24And it also gives me some readouts here like SNR, the receiver gain, et cetera.
47:28So let's just try sending this message test.
47:31And again, everything is in my office right now.
47:34But once we go on the road and navigate about, we can do the same process here and we'll
47:39know right away if the message got through.
47:41So I'm just going to go ahead and click send.
47:45Okay, that click you heard was the AirMaze keying and it's still keying.
47:51It will tell us when...
47:53Okay, so it just stopped keying.
47:55Yeah, so that was very quick.
47:57Basically it said that we sent the message, but then the Pi had to decode it.
48:01And so you can see the entire frame here.
48:03You can see the message test and you can see that it was successfully decoded by the Pi.
48:10So now because we have this app, I can bring this into the car and the Pi will be connected
48:16to this guy's hotspot.
48:17That's what like that little ring is right there.
48:20And this phone is on tail scale.
48:22My home network has tail scale set up.
48:24And so that's how we're able to control the AirMaze remotely and then also get the results
48:30in real time.
48:31So, yeah, I know that's pretty complicated, but it's going to make this testing a hell
48:35of a lot easier.
48:36Okay, so we've got our rig back there.
48:40GPS antenna is actually right there.
48:45It's on the roof.
48:49And then what we're going to do is drive out a little bit and we're going to try to send
48:53some messages here.
48:54So let's go out on the street a little bit.
48:58Okay, so we're just a little bit up the street and we're going to try to get a message through
49:03from here and see if it works.
49:06Okay, so I am a little bit up the street now.
49:10So what we can try to do is send a message here.
49:12So I'm just going to type something in here.
49:15I'm just going to do range 07.
49:20So that's our message.
49:21Our transmit power is going to be one milliwatt.
49:23We still have zero gain on the RTL-SDR because we're so close that I don't think we're going
49:27to need any.
49:28And yeah, one milliwatt is like nothing.
49:32But interestingly, this has been working for me.
49:34So let's go ahead and send this.
49:36Okay, it says transmitting and you can see SNR hop up to 32, which is very healthy.
49:44Okay, it's done.
49:46And so if we come down here and we can see it was immediately decoded by the Pi.
49:50So range 07, that's our frame.
49:53So in my neighborhood, I'm pretty easily catching these data packets.
49:57So we're going to keep going out a little bit further.
50:00Okay, so we are over two miles away from our base station.
50:05And you can see that I have the quarter wave whip antenna.
50:10I have the GPS right here.
50:14And then I have the RTL-SDR right here.
50:20What we can do is try to send a message together.
50:23So I'm going to go over to the app here.
50:27So I can see what's going on here.
50:30Okay, so you can see the app here.
50:32We're getting live reading.
50:33So let's try to send a message right here.
50:36So I'm going to send hello.
50:46And let's see if we can't get this to go through.
50:48Okay, it just transmitted.
50:53It's waiting for the Pi.
50:56I'm not sure that one's going to get decoded here.
50:58So let's see.
50:59Okay, so it doesn't actually look like that went through.
51:02And sometimes we have to tweak the gain on the RTL or the transmit power.
51:06Although it was working a second ago, so I'm going to give it another shot.
51:09All right, we're going to bump up the transmit power to 940 milliwatts and we will do H3
51:15this time.
51:16And then we'll go ahead and click send.
51:18I'm going to do a short message R1.
51:23Let's see here.
51:27Okay, we just got one through.
51:31Took a little bit of finessing, but you can see R1 decoded.
51:35So basically you have two parameters you're working with here, which is the gain on the
51:39receiver antenna, the RTL gain, and also the transmit power.
51:43And actually also the size of the frame, so which is like how much data you're putting
51:46in the message.
51:48So there's a couple of different parameters there, but we're able to get messages.
51:54We're over two miles right now and this is not line of sight.
51:57We have neighborhoods between us, so I feel pretty good about that.
52:01Boom, right there.
52:04So you could see that we just turned up the transmit power and it helped us get a message
52:08through.
52:09And to be clear, TailScale was carrying only the remote control and telemetry.
52:14The packet, whose range I was measuring, still had to travel entirely over HF.
52:19After driving around town, I eventually got a clean decode two miles away.
52:24It was still a local test, but the bridge had carried a real reticulum message from
52:28the air may in my office to the receiver in my car.
52:32Now it might occur to you that this is an awful lot of work just to own my own communication.
52:37It is, I agree.
52:38But if this were only about sending a text, it would be insane, but it isn't.
52:43The problem with middlemen isn't merely privacy or monthly fees.
52:47It's actually their ability to control our action potential in the digital world.
52:52This is another concept that Mark, Reticulum's founder, harped on when he spoke about it.
52:57Remember back in 2021 when GameStop stock got caught up in that massive short squeeze?
53:01There was unprecedented volatility around the stock.
53:05And to day traders, volatility is how money is made.
53:09But Robinhood was a centralized middleman, and it restricted trades.
53:14So imagine spending a lifetime developing an investment thesis, and when your window
53:18of opportunity finally arrives, the rug gets pulled out from under you.
53:22Whatever you think of about that week, people discovered that someone was standing between
53:27them and the market.
53:29Someone who could say no.
53:31But Robinhood's hands were somewhat forced.
53:33To remain legal, it couldn't have infinite exposure.
53:37And that's what ultimately led to the restrictions.
53:40Now I'm not critiquing a corporation.
53:42I'm critiquing the architecture that made that situation possible.
53:46Decentralized exchanges like Uniswap flip that architecture.
53:50The website can still be pressured or restricted, but the underlying protocol has no broker
53:55in the middle who can simply turn off the buy button.
53:58But middlemen don't only gatekeep.
54:00They also know things about you that they can later be compelled to hand over.
54:04In 2004, that's exactly what happened when a Chinese journalist used his Yahoo account
54:10to send an email summarizing government instruction given to the media.
54:14His offense was literally sending an email.
54:16Yahoo provided account information that helped Chinese authorities identify him.
54:21He was ultimately sentenced to 10 years in prison.
54:24Yahoo's defense was that they had to obey local law.
54:27And again, that's precisely the point.
54:30The architecture put his identity in the hands of a middleman who could be forced to surrender
54:35it.
54:36Silicon Valley tried to deal with these unprecedented levels of power by creating principles.
54:40Google famously had one called "Don't be evil."
54:43But greater still is "Can't be evil."
54:46And that's the ambition behind decentralized blockchains and permissionless networks like
54:51Reticulum.
54:52Use math to reduce the amount of trust we have to place in anyone standing in the middle.
54:57And it's a better model.
54:58Okay, stepping off my soapbox now.
55:00Every time I extended the link, another fundamental part of the system needed work.
55:05Sunday was two days away.
55:07I could move messages around the neighborhood, but I still hadn't built the tests represented
55:11by this map.
55:12At that point, this wasn't about local range anymore.
55:15I wasn't going to prove Skyway by taking another drive around the neighborhood.
55:19A real test needed someone hundreds of miles away, listening on the same frequency with
55:24my decoder while the band was open.
55:26They wouldn't need another AirMay or even a license to transmit.
55:30Potentially they would only need a $37 receiver, an antenna, and a computer.
55:35But they would need to be in the right place at the right time, running software I had
55:39only just finished building.
55:41I had built everything I could test alone.
55:43The final obstacle wasn't purely technical anymore, it was logistical.
55:47And the deadline was now only a few hours away.
55:50So I kept working.
55:51I kept testing the decoder, fixing the power problems, and trying to turn all this radio
55:56gear into something another person could actually operate.
56:00Because I had made myself one promise at the beginning of this project.
56:03When the clock ran out, I was going to ship whatever I had.
56:06And Sunday came, and Sunday went.
56:08And I had never made it to the SkyWave test.
56:11And this is where my perfectionism would usually get the better of me.
56:14I would keep working, and you would keep waiting, and I wouldn't show you anything until the
56:18result looked like the one I had imagined.
56:21But this time I had committed to publishing the good, the bad, and the ugly.
56:24So whether I liked it or not, it was time to ship this part of the project.
56:28Which was a bit of a bummer.
56:30But honestly, I had come a long way from where I had originally started.
56:33A few weeks ago, I didn't know how to use an SDR like this, or interpret a waterfall.
56:39I had never listened for APRS traffic, or picked up a real FT8 message.
56:43I didn't know what it meant for the band to close, and then reopen.
56:46I had no idea how to verify that an antenna was tuned for the frequency I was using, why
56:51touching the coax could move the center frequency, or how ground planes worked.
56:56Well, I still don't understand ground planes, but the other things I do now.
57:00I built the antennas.
57:01I learned how to listen.
57:03I heard other people doing the thing I came here to do.
57:06Across distances, I honestly wouldn't have believed this little radio could cover.
57:10And somewhere in all that static, I started to understand what this was actually going
57:13to take.
57:14Maybe the status quo exists for a reason.
57:17Because after a year of trying to escape what I half-jokingly call the "communication
57:21cartels," I had somehow subscribed to nearly everything they sold.
57:25Fiber, Verizon, Starlink, and now Iridium.
57:28In trying to cancel the middleman, I was somehow also becoming their best customer.
57:32Still, we proved a few key things.
57:35First, this receiving setup decoded a digital transmission from Bogota, roughly 1600 miles
57:41away.
57:42That proves we can push beyond the horizon.
57:45Second, we proved our bridge worked.
57:47We used an ordinary reticulum node to send a message to a licensed gateway.
57:52The gateway converted it into a public, call sign-identified HF frame, and transmitted
57:57it over the air.
57:59The receiver decoded it, rebuilt the message, and delivered it to an ordinary reticulum
58:04node on the other side.
58:05But in spite of all that progress, we still hadn't done what we originally set out to
58:09accomplish.
58:10We wanted to prove that two people separated by hundreds or even thousands of miles could
58:16still reach each other without a cell tower, a private satellite, or anyone in between
58:21giving them permission.
58:23I haven't proved that yet, but I'm a hell of a lot closer than when I started.
58:27But I still haven't fully succeeded, and that bothers me.
58:30The goal of this channel is to build a free, open-source internet that we actually own,
58:35using tools built by the people, for the people.
58:39Because if we can get a message to originate from a reticulum network in, say, Germany,
58:43bounce off the sky, touch down in North America, and bridge back into a reticulum network here
58:49in the States, well that might change everything.
58:52Somewhere on the other side of this map, somebody already has the other half of this test.
58:57Now I just have to find them.
58:59For more, click here.