The Secret History of the Internet
01:25:19 Share

Support for Waveform comes from AT&T. What's it like to get the new iPhone 16 Pro with AT&T NextUp anytime? It's like when you first light up the grill and think of all the mouth-watering possibilities. Learn how to get the new iPhone 16 Pro with Apple Intelligence on them and the latest iPhone every year with AT&T NextUp anytime. AT&T, connecting changes everything.

Apple intelligence coming fall 2024 with Siri and device language set to US English. Some features and languages will be coming over the next year. Zero dollar offer may not be available on future iPhones. Next up, anytime features may be discontinued at any time. Subject to change, additional terms, fees, and restrictions apply. See att.com slash iPhone for details.

Bored with your boring cardio? Stop pedaling that snooze cycle to Nowheresville and try some cardio that's actually fun. Supernatural Fitness, available on MetaQuest. Isn't that right, Jane Fonda? Cardio will never be boring again. Sweat to the beat of thousands of chart-topping songs inside stunning virtual landscapes. Bet your stationary bike can't do that. Visit GetSupernatural.com and join the next fitness revolution. Supernatural VR Fitness, only on MetaQuest. Wait a team for team.

All right. Welcome back to Waveform. We got another special Longform episode for you today. And this one is a long time coming. We actually did that ICANN and the Seven Keys of the Internet episode way back in November. We published it in December. Turns out we did so much research for that episode, we ended up with two stories.

We tried to slam both of those stories into one story and it just became too tail-end-y, tangential. So we just decided we were going to cut it up into two. And now, almost six months later, we have the second part. Welcome to part two. Welcome to part two. If you haven't seen I Can and the Seven Keys of the Internet, I do recommend going to watch that. There's a lot of relevant stuff. They're not directly related, but they are tangentially related. So you should go watch that either before or after this episode.

But yeah, you guys ready to get started? I am ready. I'm trying to recall as much from that as possible just so I have that context. But yes, fully. I almost forgot about this, to be honest. The most context you need is David never ordered that pizza at the end. Yeah, that was a cliffhanger. Spoiler for those of you who didn't yet see that episode. The pizza was never ordered. Yeah.

But we do own .pizza now. This is a lot more like a prequel rather than a sequel. Yeah, it's like a prequel. It's the lore. Yeah. It's Star Wars, you know, like backwards. This is about to be like the Phantom Menace of Waveform podcast episodes. I don't know if that's a good thing or not. It's about to be the attack of the clones of Waveform podcast episodes. I like that one. I like that. I like that. All right. So if you guys could take a guess, where do you think that our story begins today? Oh, thanks.

David brainstorming something and four hours later making up a different story. I should say, when do you think our story begins? The second you ordered that pizza. As has been the case for the at least two other long-form podcasts that we've done so far, our story begins on October 4th, 1957. CBS television presents a special report on Sputnik 1, the Soviet space satellite. Now,

Now, on October 4th, 1957, Russia launched the Sputnik satellite into orbit, right? And I think we were both surprised and also completely unsurprised that this is where the story begins. Because as it turns out, an insane amount of the technology that we use today comes from the United States' reaction to that one day.

So when the United States realized that Russia had put that satellite into space, it like went on full on panic mode, right? Because it meant that there was another world superpower that was ahead of the United States technologically. And if you know the United States, you know that that's not going to fly. Can't just sit around and let that happen. No, no, they don't. They don't like that. I mean, the U.S. was already freaking out about needing to have the strongest military around. And this was also 10 years into the Cold War, which lasted 45 years.

So when they threw that satellite in this space, they were like, oh, God, oh, God, we are very behind here.

So if you remember from any of the other space episodes that we did, like the James Webb Space Telescope episode or the new Space Race episode, what did the U.S. do immediately after Sputnik got launched into orbit? I think we started working on our own launches. Didn't we put man on the moon like 15 years later? Yeah. Like the 60s? Yeah. So what organization did they launch to do that? Oh, NASA. Well, that's where you're wrong, suckers. Oh, okay. Baited. Baited. Got him. Baited.

So, like, technically, yes, that is what happened. But what actually happened is that four months after Sputnik launched, the U.S. spun up something called ARPA, which is the Advanced Research Projects Agency. So basically, if Russia was going to be pumping millions of dollars into missile, satellite and space technology, the U.S. was going to do that, too, right? DARPA, shaping the future.

Creating opportunities for new capabilities strategically, tactically.

DARPA takes on the most difficult technical challenges for the Department of Defense. So you said ARPA, but that said DARPA. Right. So you might have noticed that. It's because it switched back and forth between DARPA and ARPA like a million times. It was ARPA and then DARPA and then ARPA like a few times. And that's because it kept getting pulled in and out of the Department of Defense. Okay. Yeah. So the Defense Advanced Research Projects Agency was DARPA and then Advanced Research Projects.

projects agency was arpa just saying i like darpa better defense championship uh yeah okay anyway so the actual space program that the u.s wanted to launch was arpa uh instead of nasa right just a couple months after that satellite went into space but pretty quickly the u.s realized that it didn't have a lot of public support for war right like

Putting a ton of public funding into a military program was extremely unpopular. This was right off the heels of World War II. And add that to the fact that the U.S. was kind of at this time seen as this like aggressive colonial nation. And they kind of realized that it was going to be a lot easier to fund a missile program if they could make it public facing and sciencey and fun. Right. So a couple months later, like literally a couple months after they founded DARPA, they founded U.S.

NASA. We have one of the most challenging assignments that has ever been given to modern man. Expansion of human knowledge about space. We've been assigned the mission of launching a scientific earth satellite.

It's crazy how everyone from the 50s had the same voice. I know. And I wish I had that voice. Recognizable voice. I want the radio voice from the 50s. I'm David Amell and brought to you by... And this is Waveform. Coming up. Now, Sam. Okay. Anyway. So...

So, yeah. So NASA is doing serious research, right? They're doing serious rocket science. It's still obviously related to what the U.S. wanted to achieve with DARPA. But it was also more focused on the science-y aspect of it because, you know, that's what the public was into at the time. But the nice thing is that when you're trying to build missile defense and anti-air systems, building things for NASA is basically the same thing. Hmm.

You're basically doing what you wanted to do with DARPA, but you're just doing it in a public facing way. Right. But it's still not exactly what they wanted to do. Right. Because the specific things that they wanted to be building were like literally anti aircraft, anti missile, like detection systems, like nuclear detection systems, all this stuff. They really wanted ARPA. But within a few years of that, all of the civilian space programs of this new agency were being transferred to NASA. Right. So DARPA had existed, but everything was being funneled into NASA. And all of a sudden, um,

They basically didn't have a lot of people working at DARPA anymore, and they didn't really have any funding. So you end up with this extremely lean, no-money group, and they became a high-risk, high-reward, and far-out research program. They were kind of happening behind the scenes. And the point of DARPA became to surprise other nations before they were surprised. So basically, be the next Sputnik. NASA was spun out, the military side was spun out, and a few years later in the early 60s,

one part of that agency focused on application of computers and how to make big leaps forward. So that was Steve Crocker. He's super pivotal in the early standardization of a lot of the stuff we're going to talk about later, but he was the one that we contacted for this story originally. That office that was created funded

advanced work in computer science, time-sharing systems, artificial intelligence, advanced graphics, new architectures, multiprocessors, and so forth. Big stuff at the time. So since all those funds were from DARPA were getting funneled to NASA, and since it was this big, fun, public-facing agency, it was getting a lot of support.

And DARPA had no money. So but everyone knew that DARPA was really important, right? Because like especially all the people in the government were like, we still need to be putting resources into this.

And the first director of DARPA apparently left a $160,000 job at GE for an $18,000 job at DARPA. Do you know how much money that is today? Like a $160,000 job? This might be double. That's $1.7 million today. Yeah. Insane. Anyway, so when you can't employ that many people because you don't have that much money, like that director was getting paid $18,000 a year. Yeah.

Maybe a smarter idea is to just sponsor all the smartest people in the world. You know, there's a famous quote that like all the smartest people don't work for you. And so what they did was they decided to sponsor universities instead. But here's a problem, right? These really big computers that researchers wanted to use, they were like really, really expensive.

So, there weren't a lot of them. And it was really inefficient to have all these amazing state-of-the-art computers at some universities that were being sponsored by DARPA, but not be able to use those resources at other universities for the same research purposes.

And so they were all kind of working on these different advanced computer science projects, but they were in their own little isolated versions of the future. They weren't talking to each other. And I guess a critical thing to understand is that, of course, at each of our places, we had already been selected and been working on advanced computer science activities. So we were, in some sense, living in the future in each of our little worlds. So pretty much they just realized that their resources were not available.

being as efficient as possible because it's yeah essentially you're just doing the same work right multiple different areas right okay you have all these super computers at these universities and it's like they're not talking to each other starting to think we're finding the connection to the episode what if we could like connect all of them together maybe in some sort of like interconnected network of

MARK MANDEL: That'll never work. MARK MANDEL: Would that be a good place to start? MARK MIRCHANDANI: If you had that idea 60 years ago, you'd be considered a genius. MARK MANDEL: And you would have only gotten paid $18,000 this year. MARK MIRCHANDANI: Yeah. MARK MANDEL: Interesting. MARK MIRCHANDANI: Yeah. So they obviously wanted to connect all these computers together, because it became clear very quickly that there was a lot that you could do if you connected all of these supercomputers together. You could do time sharing, like we talked about in the last episode. You could be sharing information so you weren't doing redundant research. It made a lot of sense.

So, in 1966, there was a proposal for the ARPANET, right? The Advanced Research Projects Agency Network.

Which was a network of computers that would connect all these different universities and research institutions and would allow the researchers to access the computing power of each other's supercomputers, share information, all this kind of stuff. You're basically doing like game streaming. Like right now, game streaming is like you're using a computer somewhere else. Someone else's processing power. Right. To do your own research. Right. I mean, game. Right.

I mean research. Mom, I'm doing research. I'm doing research. Do you guys remember folding at home? Have you ever done folding at home? No. I think it's called origami. Yeah.

Holding at home is like a genetics project that was very popular where you could, like while you were sleeping, use your computer's computing power to do research for the World Genomics Association. Sounds like the scientific version of mining Bitcoin when you're sleeping. Effectively, yeah. We're going to go to BitTorrent. It's pretty much the same thing as mining Bitcoin, actually. I see. Anyway, same idea. Just real quick, before you go any further, for my own sake.

What approximately what year we are right now because like a lot has happened is started around like the 50s and yeah So the ARPA net was proposed in 1966 Okay, so there was a decent amount of time between our being spun up and this network being proposed Yeah, right That's kind of a theme throughout this episode is there are large chunks of time between these new technologies being invented and actually getting rolled out Like you're gonna see the timeline gets kind of wild, but this is actually pretty good timing. Um

because this guy that was working on the development of the ARPANET at a US government think tank, as well as a separate guy in London, came up with an idea called packet switching. Do you guys know what packet switching is?

I've heard the term, but if you asked me to explain it, I would not be able to. It sounds like one of those things where someone's like, you're a nerd. What's packet switching? I was like, you wouldn't understand that, but I actually have no idea. Okay. Well, so before packet switching, everything was circuit switched, which means that you had to have a direct line connection from like one node to another. So think about an operator that would connect your phones together. What he was actually doing was literally switching the circuits. Yeah, operator, where's the nearest telephone?

You were literally connected directly to the other phone. Which is insane, right? Yeah, that's really insane. But a packet switch network is pretty different. So basically, the data that we have on our computers that we're sending to each other can be broken into packets. And a packet usually is like, you know, eight bits or something, a byte of data, some sort of data.

And it gets broken into these packets and sent through various routes in the network, depending on which line has the least load. So usually on a circuit switch network, you have to go directly to each other. But now you kind of learn about this in like queuing theory and computer science 101. If you have all of these different lines, there's going to be different amounts of load on each line, right?

So if you break your data up into a bunch of packets, each of those packets can go on a different line that has the least amount of traffic. So you can see it like eight lanes on a freeway, right? I was just going to say, I wish I could do this driving to work in the morning. Yeah. Waze is just packet switching for cars. Right. Funny enough, there's actually a thing in traffic theory where this is like, it's crazy. Traffic theory is very similar to...

to like liquid flow theory. - Fluid dynamics. - Fluid dynamics, where adding more lanes doesn't actually make traffic go faster, so it's not the perfect analogy, but imagine you're at the grocery store, right, and you wanna make hamburgers for dinner, and you have five family members and there's five ingredients, and there's five checkout lines. It would be faster for each person to take one thing, check out at the same time, go home and build the hamburger, right?

Instead of waiting in one line and then checking things out individually. You're basically doing it five times faster. So that's basically the idea of packet switching. You're breaking data up into little packets and then you're sending it over this distributed network. So in a network, this actually helps create something that is decentralized or more accurately distributed. Because there's three types of networks. There's centralized, decentralized, and distributed networks. And this is actually a distributed network.

So imagine that DARPA connects four different universities, right? They're each connected to each other. They're not just connected in a line.

And they can send data in packets. And that allows them to send a ton of data through the queue with the least stuff in it. So the data knows what order to rearrange itself in because it's got a little tag on it that says, this is the first part of the data and this is the second part of the data. Because data is just like bits, right? Streams of ones and zeros. Streams of ones and zeros. But it knows like this is the first chunk, this is the second chunk, this is the third chunk. And it can rearrange itself. That allows for a lot of data to pass through in a much faster distributed way because data is traveling at the speed of light, right? Yeah.

So even if you drop packets, that's not a problem because you just resend it.

If you're traveling at that speed, it's like, oh, no, I dropped. You've probably seen this before, like dropped packet loss, dropped these packets. There's like interference or something in the network. But it's like it doesn't matter because you're just sending data through. The best thing to use when you like miss a kill in Valorant. Just be like, oh, packet loss. Damn it. My ping. My ping. Yeah. Anyway. So, yeah. So they get the first node set up in 1969 for this new ARPANET. Right.

And this was running on the 1822 network protocol, which was implemented in 1970. So it's just like time is going by slowly. Like they get the node set up. The next year they get the protocol implemented. And the protocol is basically saying, like, how is this data going to transfer across the network? How are the computers going to handle the data flowing between them? So the network gets turned on in 1971 and it actually works really well.

And there's this really funny story about when they were at the ARPANET and they were first starting to use it. And they connect all the computers together and they're getting all excited. And they want to log into another computer remotely because that's what you're doing if you're doing packet switching, right? You're like, let's do remote desktop. So they type L and the other person on the other side of the line says, yep, got the L. They type O and then the other person's like, yep, we got the O.

they're trying to type log in. They type G and the network crashes. But, you know, it works. And they had to do some modulation to fix it and make it a little better and like actually make things work. But that was the foundation of the ARPANET. So they got the first computers connected and it gets rocking. Andrew just brought up a really good point that

the very first electronic message ever sent might have just been L. - It was the original taking an L on the internet. - Yeah. - Imagine you took the first L on the internet. - The first L on the internet. - And you didn't even know it. - Yeah. - So over the next few years, the packet switch network idea starts to really take off. All of a sudden you get a ton of universities and organizations that are using their own packet switch networks with things like ethernet that had just gotten invented. And there was something called Token Ring from IBM

which had a really small community. And Xerox invented a new protocol called PUP, and there was a French Cyclades network, right? Like this was a protocol, the way of doing packet switching was just how you send data over a network. So all of a sudden we saw all these networks getting set up. And these are all individual networks only connected within themselves. Correct. Not quite what we have today. Correct. So it was just, yeah, like the French had their own network. And then everyone was creating their own little like

protocols right because the arpanet was running on the 18 22 protocol but token ring was running on something else just at ibm some universities make their own protocols that only ran at those universities so it sounds like at this point we have a bunch of little internets a bunch of little networks

Yeah, I have to keep reminding myself that they're just logging into these other computers. They're remoting into these other computers. This isn't like creating this crazy internet that we have today. Right. This is super basic. They're using each other's processing powers. They're sharing data. They're sharing files. But it's within a small network. A bunch of local networks. Correct. A bunch of intranets. Intranets, if you will. Networks, yeah.

Also, there isn't standard networking infrastructure yet, which is why a lot of these networks have to run on separate protocols. Because the 1822 protocol was DARPA, right? They weren't going to just tell people how to run these networks. And also, in the United States, we had this full network of phone lines set up. So when these first universities...

were setting up, wanted to connect via the ARPANET, they literally just reached out to the phone companies and said, hey, can we lease some time on this line? So the University of Hawaii actually had this thing called the AlohaNet, which was...

basically doing packet switching over radio, which was really awesome. That is really cool because before we were doing straight phone lines. So radio, this is like kind of like the first wireless. Yeah. If you want to think about it. They showed that you could do packet switching over any type of network, which is really cool. Yeah. We're going to take a quick ad break, but there's a lot more to the story. It gets very exciting. So make sure you stick around and we'll catch you on the other side.

Support for Waveform comes from AT&T. What does it feel like to get the new iPhone 16 Pro with AT&T next up anytime? It's like when you first pick up those tongs and you know you're the one running the grill. It's indescribable. It's like something you've never felt before. All the mouthwatering anticipation of new possibilities, whether that's making the perfect cheeseburger or treating your family to a grilled baked potato, which you know will forever change the way they look at potatoes.

With AT&T NextUp Anytime, you can feel this way again and again. Learn how to get the new iPhone 16 Pro with Apple Intelligence on them and the latest iPhone every year with AT&T NextUp Anytime. AT&T, connecting changes everything. Apple Intelligence coming fall 2024 with Siri and device language set to US English. Some features and languages will be coming over the next year. Zero dollar offer may not be available on future iPhones. NextUp Anytime features may be discontinued at any time. Subject to change, additional fees, terms and restrictions apply. See att.com slash iPhone for details.

Is it time to reimagine your future? The right business skills may make a difference in your career. At Capella University, we offer a relevant education that's designed to focus on what you need to know in the business world. We'll teach professional skills to help you pursue your goals, like business management, strategic planning, and effective communication. And you can apply these skills right away. A different future is closer than you think with Capella University. Learn more at capella.edu.

All right, we're back. Welcome back to the Packet Switch podcast. Like we talked about before, there was all these networks that were being set up, right? But when you set up a network, you're buying this box from whatever the vendor you were using gave you, right? So you ended up with a software stack of applications that you could use on top of it, and that was basically it. It was like they gave you this black box to set up your network, and they ran all the protocols, they ran all this stuff, and you couldn't really do a lot to it.

So now all of these vendors have their own protocols, their own data rates that they're transferring data over. None of these companies cared if their networks interconnected and they probably actually liked that. It was like using Android and iOS. Apple likes that iMessage doesn't work with Android because it keeps you in their ecosystem and all these companies wanted to keep you in their little network ecosystem. It's the same thing.

And there's some level of control maybe over reliability or materials used, things like that. All these things. Yeah. Yeah. It's just a thing we have in tech. Yeah. Well, and also it was all like separate, like in order to for everything to work together, there would have had to been a conscious effort from all these different groups to be like, let's build one thing. Right. Which they don't want to do. I mean, you can see this playing out today with iMessage, WhatsApp, Telegram. They all basically do the same thing, but none of them actually talk to each other.

right there hasn't really been a universal messaging protocol the internet mess for internet messaging that rcs yeah well the best thing we have is rcs but you know how that's played out so far uh hopefully that comes into play eventually but yeah rcs

RCS. But pretty damn quickly, the Department of Defense started taking notice of all these networks, right? And as more of these distributed package switch networks got set up, the government was like, wait, we should use this in war. Yeah.

Right. Pretty classic. Evergreen. Yeah. Classic government move. Yeah. They could take a distributed communications network and they could hook it up to, you know, like planes and tanks and automobiles and all of these different things. Because at this point, computers are getting more popularized and computers that were running on tanks or planes were going to have a strategic advantage over a plane or a tank that didn't have a computer in it. That was just the way it went. Yep.

Some people think older people don't know how to use the internet.

I have news for you. We invented it. It's possible to invent something and be totally behind on it. Yeah, for real. I'm just saying. Well, you wouldn't say that to Vint Cerf. He was working on the early ARPANET design as well, and he helped get a lot of the stuff online. Also, he's one of the most important people to ever contribute to the internet. He is in the Internet Hall of Fame. He is considered, like, Internet Zaddy. Oh, yeah. Internet Hall of Fame, yeah.

The Internet Hall of Fame. The Internet Hall of Fame, which is like the most important people to ever contribute to Internet networking. How much Reddit karma does he have? Probably a lot. Yeah. If you talk to any like data systems engineer or any like computer science graduate, they he's like they're like Wayne Gretzky.

Yes. I don't know that reference, but probably. A lot of goals. Probably yes. A lot of goals. He's a very important person. Long career. The blank stare was everything I've ever wanted in life. His title at Google is Internet Evangelist. It's Chief Internet Evangelist. And he works at Google and his title is literally Chief Internet Evangelist.

Man, as if the internet needs evangelizing. Well, you think it doesn't, but surprisingly, we'll get to that later. Okay. And in the spring of 73, a guy named Bob Kahn, who had worked very much on the ARPANET design, left Bolt, Baranek, and Newman and went to DARPA.

And he showed up in my office at Stanford in the spring of '73 and he says, "We have a problem." And I'm thinking, what do you mean we? And he says, "Well,

The Defense Department has reached the conclusion that computers could be useful in command and control. But the implication of using computers in command and control is that you'd be putting computers in mobile vehicles, ships at sea, and in airplanes, in addition to dedicated computing facilities. Well, the ARPANET was built out of computers that were in dedicated computer rooms with air conditioning. Everything's all interconnected by dedicated telephone lines.

And you can't connect an airplane...

to a telephone line. That doesn't work. Cars don't work, or the tanks, they run over the wires. That doesn't work. Ships get all tangled up. Funny guy. Funny guy. But it's true, right? So the government noticed how well packet switch networks were working, so they started spinning up satellite-based packet switch networks because that doesn't use a line, right? You could actually connect that to pretty much any type of military vehicle. And they called that SatNet, right? Yeah.

So they made a radio-based packet switch network because the military realized it's crazy useful to command and control. And suddenly they want computers and airplanes and they want them in tanks and boats. And they spun up like all of these different packet switch networks with different data rates and different packet sizes and different latencies because they're all package switched, right? So you need different data rates and latencies and stuff with like, you know, tanks and airplanes.

airplanes and stuff because they're far away. You have the inverse square law for data transfer. And they all do different things and probably all have different requirements for bandwidth and what types of things you're actually sending over that network. Right.

Yeah. And the thing is, like the government and the military have different, you know, there's the Air Force and then there's the there's the the Navy, the Navy, the Army. Right. So they're all spinning up their own different packet switch networks, but they're not really communicating that much. And that becomes a problem. So in 1973, they go to vent and they say.

we've got a problem. We need something that can allow all of these networks to talk to each other because we need all of our military applications to be able to communicate. We're looking at four different packet switch technologies, all of which are different. They're different data rates, different packet sizes, different addressing, different latency. It's just all different, but they're all packet switched.

So the question is, how do you take all those different kinds of packet networks and connect them together in a way that makes everything look uniform? That was the internet problem.

And we spent from the spring of 73 to the fall of 73 trying to figure out how to do that. Okay. Can I just take a second and like the internet problem, right? You have all these networks. They're all packet switch networks. Nets. Networks, network, network, network, inter networking problem. Yeah.

This is the connection of all the other networks. We're finally bringing everything together. Yeah, because inter means everything under one umbrella.

And we just take the word internet for granted right now, but it's like, it means all the networks are now connected. You're making a giant network of smaller networks. And this was, this was just the government's networks. This wasn't every network from all the schools and everything else, but it was combining several networks that were being used for several different things into one internet network. Okay. I don't know why this reminds me of it, but, but this is like when, when

When Xbox first came out and you played like split screen and then you started going to your friends and you played like interconnected like LAN parties. Yeah. And Xbox Live comes out and now it's like

I don't have to go to my friend's house with eight other friends to play eight player. All of us can say that's exactly correct. You have your own little network when you're doing like a LAN party and then all of a sudden you're on the internet. Yeah. A bunch of little networks. It's funny to think that at the time there were computer scientists working on networking and then there were computer scientists working on internetworking. Yeah. Yeah. Really awesome. Um, so, um,

How do you get all these computers to talk to each other? That's a big problem because they're all using different protocols, different data rates, different latency, all of these things. So like Vint just said,

between just between like six months they came up with a way to solve the internet working problem we eventually came up with a protocol we call the transmission control protocol or TCP have you guys heard of TCP before oh we're starting to get into modern terms yeah TCP IP all you guys are starting to sound familiar TCP IP okay insane clown posse

So TCP, yeah. Yeah. So transmission control protocol and internet protocol. And these are the protocols that allow all these different networks to talk to each other on the internet. Right? Very cool. Vint Cerf invented TCPIP.

probably the most important thing ever. - Okay, I can see how he's in the Internet Hall of Fame. That makes sense. That guy's pretty good. - In like six months. - Okay, that's pretty good. - Yeah, he's very important. But yeah, TCP is basically a list of agreed upon actions that allow these networks to communicate and pass data between each other and let that data stay intact so that if you drop data, you can just throw it back in.

Back then, it meant that if your computers and your network can use TCP, you can connect to any other network that's also using TCP. The local network at the University of Southern California could now connect to the local network at the University of London. They could share their research. It was like the USB of internet protocols. Basically. Like a standard. It was a standard. A standard. Right. So scientists can now communicate with very early forms of email across the entire world. Right.

Because they had email when they first set up these networks. It was very primitive, but it was like there was an email protocol that

Which is why, this is the reason that you can use Gmail or Spark or whatever. You don't have to, like, if you think about like Telegram versus WhatsApp versus like iMessage, those all use different protocols, but there is an email protocol. So that Gmail and Spark and Outlook and all. Yeah, it doesn't matter what client you're using, you're sending the same data and it's getting across, which is dope. Like I wish you could send a Telegram message and it came through on WhatsApp, but you can't. Anyway.

So they were working on TCP for a long time. They had four different iterations of it. They made a lot of mistakes and it took them 10 years, right? Like they came up with the first iteration of it in like six months, but it took them 10 years. Within that span of time, the first cell phone came out, the first digital camera came out, the VCR came out, the first Apple computers came out, the GPS, the Walkman, everything.

Basically, an insane amount of technology came out in the 1970s. The first season of retro tech right there. Yeah, exactly. Pretty much all from the 1970s. And Vince says they pretty much finished the protocol in 1978, but it took them five more years for it to get implemented. Huh.

Yeah. So there are large spans of time between any of this stuff happening. So we shouldn't expect Mastodon to have any sort of reckoning in the next year. It's probably going to take a while for a new standard to hook up. If we ever have a universal standard for that, it's probably going to take 10 to 15 years. It's going to take a long time. Yeah. And it's probably because it's not being implemented by the government. This was the government. Yeah. Right. Right.

That feels like the big difference between back then and today is this all started through the government. Right. This is kind of a tangent, but the only reason that the internet or that email was able to do that was because it was kind of run by the government. We just didn't have...

internet messaging that the government created, right? Internet messaging was created in an era where private or public companies, private companies, companies created all these messaging clients. So it wasn't the government that said, you guys all have to use this, right? Anyway, to get TCPA IP implemented, you still had to force everybody to use it.

That was a problem, right? For a universal protocol to work, everyone has to be on board, which is why I was so excited for this Matter episode because everyone started using it and Matter is basically TCP IP, which is awesome. It already exists. I mean, Matter used to be called Chip, which is connected home over IP. It literally is IP. It's internet protocol, which has been invented. Crazy, right? And since everyone's signing on to use the protocol, everyone becomes happier.

But Vince, like, we need people to use TCP IP. So he makes this absolutely baller move. He knows that everyone's going to be happier if they're using it. And there are benefits to the military for the military for that happening, too, because DARPA is a military organization at the end of the day and they need command and control to be able to communicate. And researchers are going to love it. And Vince just like, no, I have to force people to use TCP IP, even if they don't know it yet. Right. That they're going to love it.

So Vint goes to everyone on the ARPANET, which again, there's all these research universities and stuff, and it's being funded by the government, but they're still research universities doing their own thing and they create their own little networks and they don't really want to use TCPIP because there's no little networks. They don't really care. So he goes everyone to the ARPANET, which at that point is about 400 computers, and he tells them, you all need to switch to TCPIP by January 1st, 1983, or you're off the network.

Power move? Power move. He's basically like, we are cutting off your funding. We are taking your computers. You have to use TCP IP. And he has that type of control? Yeah. Because it's being funded by ARPA, which is the government. They were funding these research projects. And they're like, we literally will come in and take your computers. Turn off your access. Yeah.

And they didn't want that. By this time-- this is like 15 years into this project-- they were hooked into using computers. Imagine if somebody came and were like, we're going to take your computers away if you don't just switch this protocol. I guess we're switching. So by the time 1978 rolls around, we are basically ready to freeze the protocol spec after having gone through four iterations and start implementing like crazy. We get to the point in 1982

late 81, early 82, where we are now ready to force every computer on every one of the networks supported by ARPA to switch over from the earlier ARPANET protocols to the internet protocols. And so we set a deadline of January 1, 1983. There were only 400 computers involved. And said, you all have to be running TCP IP by January 1, 1983, or you're off the net. Right.

And, you know, so everybody sort of, there was some grumbling about that. But since I was running the program at the time, I could say things like, you know what, if you're not on the net by January 1 of 83, I'm not funding your research anymore. And what was that program that you were running technically? Was there a name for that?

Oh, well, it was the internetting program. The internetting program. Wow. Yeah. Like we take that for granted now, but. Yeah. Serious. Someone had to flip that switch to get us all together. Yeah. To get them all together. Yeah. Yeah. So he gets that switch flipped. Now all the 400 computers, they were all using whatever older protocols are all now using TCP IP. They can all now talk to each other and all future computers that use TCP IP can all talk to each other.

It's a pretty good starting point. So in between, or right around now, TCP and IP actually got split. Right. Which is why we just call them IP addresses instead of TCP IP addresses. And the reason for that is pretty nifty.

The important part of the IP part is the addressing part, which if you listen to part one of this whole thing, you'll know is like this big complicated thing that needed to get figured out. So they said that IP addresses should be universal for all devices that are now connecting to this inter-networking project.

But TCP isn't the best way to get data across in every scenario. For example, TCP sends confirmation back like, hey, I received each packet, which is useful if I'm trying to send you a Word document and I need to know that you missed a few words. But for example, the lights behind you in the studio...

They work on an internet protocol, but if one of the packets gets lost when I'm sending stuff there, I don't need to know the packet gets lost. I can visually see that the packet gets lost. So we use a different protocol called UDP, the user datagram protocol, which is like a slimmer, less reliable, but faster sort of thing. And all of these things are now working under IP. Right.

Yeah. So everything has an IP address, which is just a address for your computer. So it knows where the data should go. But they split TCP IP. So it's officially called TCP slash IP. That's what it's called now. So we turned the internet on on January 1st, 1983, basically. And it's been running ever since. Hmm.

No big deal. We turned the internet on. This guy's Wayne Gretzky. Yeah. For sure. Yeah. Yeah. Most shots on goal. Most goals scored. Yes. Amazing. That's pretty sick. So the internet's on, baby. It's going. A lot of people have access to it. Suddenly you've got this distributed network that can send packets between computers, super long distances. You can share information. You can remote log into super powerful computers. This internet thing is pretty rad. Yeah.

And before we break into the next act, I just want to tell this really funny story. First, you got to know that Vint Cerf wears a three-piece suit to literally everything. I looked him up on Google Images. Yeah. You are correct. Yes. Like he is like this white beard, white hair, three-piece suit. Like he is a serious man. He's so dope. Yeah, he looks really good in it.

And I've been calling him Internet Zaddy for quite a while now because that's what I... Anyway, yeah. Title of the podcast. Internet Zaddy. I have that written down for some reason. Anyway, he has this funny story where he's at an Internet Engineering Task Force meeting, which is like these meetings that they were having to talk about internet protocols and stuff in 1992. And he does this strip tease in his three-piece suit.

And under the suit, he has a shirt that says IP on everything. Because at the time, he was trying to push the adoption of internet protocol being on all networking equipment. And, like, nobody was really having it. So he does a freaking strip tease at this internet meeting. I'm glad the inventor of the internet was an OG memer. Yeah. Yes. It was like...

Off the net memeing right here. Anyway, it works and it becomes the foundation of the internet and home networking and the internet of things and basically all the technology that we use right now that does networking is based on TCP IP. And this guy made it in like six months. Incredible. Well played. Yeah. So there's more to that story. Quite a bit more to the story and it gets very interesting. So we're going to explore that after the break.

If you've heard that sound from Babbel before, I bet you do. Babbel is the science-backed language learning app that actually works. With quick 10-minute lessons handcrafted by over 200 language experts, Babbel gets you on your way to speaking a new language in just a few weeks. With over 16 million subscriptions sold and a 20-day money-back guarantee, just start speaking another language with Babbel. Right now, up to 55% off your Babbel subscription at

Support for the show today comes from NetSuite. Anxious about where the economy is headed? You're not alone. If you ask nine experts, you're likely to get ten different answers. So unless you're a fortune teller and it's perfectly okay that you're not, nobody can say for certain. So that makes it tricky to future-proof your business in times like these. That's why over 38,000 businesses are already setting their future plans with NetSuite by Oracle.

This top-rated cloud ERP brings accounting, financial management, inventory, HR, and more onto one unified platform, letting you streamline operations and cut down on costs. With NetSuite's real-time insights and forecasting tools, you're not just managing your business, you're anticipating its next move. You can close the books in days, not weeks, and keep your focus forward on what's coming next.

Plus, NetSuite has compiled insights about how AI and machine learning may affect your business and how to best seize this new opportunity. So you can download the CFO's Guide to AI and Machine Learning at netsuite.com slash waveform. The guide is free to you at netsuite.com slash waveform. netsuite.com slash waveform.

All right, we're back. So the internet's on and it's working and it's scaling really, really fast. Originally on the-- yeah, it scaled very quickly. Originally on the ARPANET, there was no central address book for these computers. Each network node had its own address book that was sort of maintained by the network administrator. But there was no way to prevent duplicates in the address book or update the address book network wide.

So when all these networks came together to form the Internet, that was a problem. And it was scaling really quickly. You know, this is similar to this is the act two is called the DNS. And I have a it's very abridged because we did an entire episode on the DNS. I was going to say this sounds like it could lead straight into the last episode. It basically could. What you could do is you could like listen to last episode and then come back if you really wanted to.

But think about this, when TCP/IP was introduced, it introduced the idea that every computer needed an address. So if you think about packet switching again, you realize it's important that each packet of data knows where it's going. So of course it's bouncing off all these network nodes trying to find the path of least resistance to get there as fast as possible, but it needs to be sent to a specific address. So back then, if you were a computer at UCLA and you wanted to get on a computer at UC Berkeley, you had to type in the IP address to connect to that computer.

This was mostly fine at the beginning when there was only a certain number of computers on the network. But, you know, there was only four universities connected at the very beginning. So you had this very small list of IP addresses. But the Internet scaled very quickly. There were already 400 connected when the Internet got turned on. And can you imagine having a spreadsheet of like 400 computers, each with different IP addresses? And every single time you wanted to connect to one, you had to like...

Look through the spreadsheet with your granny glasses and find you know Which IP address was correct like not great IP address is always in the format that they are today which is like

There have been different, there's like IPv4, IPv6, and that basically just like adds extra digits, which exponentially scales the amount of computers. That's what I'm thinking. Like at the time, oh, there's only 400 computers. So if we just have like a four digit IP address, we're good. Yeah. And then suddenly that's not enough. Well, there are many stories of people being like, oh yeah, IPv4, like that'll last for the next 500 years. No, it did not. Not even close. Yeah.

Now I believe we're on IPv6 right now, and that should last a while, but who knows? We're making a lot of devices. Because every individual device that connects to the internet has a new IP address. It's crazy. Yeah, we talked about this in the ICANN and the 7 Keys of the Internet episode, but the original solution to this IP address indexing problem was one guy named John Postel.

He created a document called host.txt that would contain every single computer's IP address. So, you know, yeah, you'd call up John, you'd be like, I want to connect my computer to the internet. And he'd be like, okay, so you're this IP address and I'm going to put you. And then he would distribute this host.txt spreadsheet to

to the entire internet and you download the host.tech spreadsheet every day. The yellow pages of the entire internet. Yeah, basically. Because he was handling the assigning of IP address, he became known as the IANA or the Internet Assigned Numbers Authority. Very similarly, yeah, there's a lot of redundant information, but I'm just trying to skim over it. Go watch our I Can and the Seven Keys of the Internet episode. Do it. Yeah.

So this was not a scalable solution. So a call of proposals on how to improve that system was made. And a man named Paul Macapetris came up with the DNS or the domain name system, which was a way to automatically map domain names or word versions of IP addresses to the server's actual IP address. Became a lot easier. It's very lightweight. It's very elegant. It's great.

Yeah. So while all this DNS stuff was happening, Vid decided maybe he should go and join the private sector. Right. He'd been working on government stuff for like a very long time and he was interested in trying something new because, you know, he just created the Internet basically. He's like, yeah.

He could go do something in the private sector and make a lot more money. Side hustle. I was going to say, is he still getting paid like garbage at this point? Pretty much. Yeah. No, he was literally, he literally told us like, I want to send my daughter to college. So I think he was, yeah, he was like, I could make a lot more money in the private sector. Um, though his two sons to college. Sorry.

So right before ARPA turns the internet on, like right before they turned the internet on, but he had it ready, everything was ready, he left and went to go work for this company called MCI, where he created something called MCI Mail. It starts with the office equipment you've already got. It ends with a message delivered to anyone, anywhere, faster, cheaper, or both. It's MCI Mail.

And it's what business uses to communicate when communicating means business.

This is very much business. Yeah. 1983. I was going to say, that's the 80s encapsulated. Oh, yeah. Yeah. So effectively, MCI mail was basically an early version of commercial email because before this, digital messaging like email was only available to government employees and the university researchers. There was no like public version of email at that time. Yeah. Yeah. So nobody had a lot of access to this stuff.

So MCI started putting computers in places like hotels, business centers, airports, and

Or you could get them at home if you were rich enough to have an actual computer. You could boot up your MCI software. So it was basically email for rich people or business people. And it wasn't connected to this new internet. Like MCI was leasing their own telephone lines from the phone companies and building their own. That's why it had to be a separate thing. Because like David said, you had to be doing government research if you wanted internet access at this point.

The internet was still only for government people and military. And this is the private sector. Yeah, because the technology was there. So some MCI spun up and they were like, we could do this on our own network, right? Yeah, so MCI mail was actually really cool because it did some other really wacky things. Like you could enter a physical address into the address bar and it would send a letter.

Oh, I was going to make that joke. No, literally. Okay. Literally. Yeah. You could, you would pay different rates based on like what kind of tele of communicate, like communications platform you wanted to use. That's hilarious. So if you thought like, Oh, I want to send this MCI mail to this other person, but I'm not sure they're going to get it in time. Send it in letter format to this guy's hotel. It was like business, you know, big business. A pigeon would take off with it. Yeah. Yeah.

So Vin actually launched MCI Mail. He went to MCI to build MCI Mail for MCI, and he launched it very quickly in September of 1983, which was the same year they turned the internet on, right? Like he left ARPA right before they turned the internet on, built this email client for MCI, and they launched it that same year, which was pretty wild. And I think it went kind of okay, like the launch of MCI Mail. This was like the first commercial email service.

Think they were expecting it to be like insanely profitable because mint saw like the value of the internet and being able to communicate like that But it didn't really turn out that way They tried a lot of different marketing campaigns to try to get people to use MCI mail But it just didn't gain a ton of hype. It took a while. They were just too early basically and that'll happen, you know, I

So Vince stayed at MCI for a while like eight years, but he got pretty bored again because you know He created the internet. He's probably wanting to do it. You're just itching for that next. Yeah. Yeah. Yeah, and you're right. Hi Yeah, and so Bob Kahn who was actually the guy who went to vent to try to get him to build TCP IP for the military and

had just created this new nonprofit called CNRI or the Corporation for National Research Initiatives. Basically a think tank, if you know what a think tank is. Just like we do all this research and then we sell it to the government or something like that. So Vint went to do that instead because it was just way more fun. Like they got to like just build a lot of random different stuff.

So they gallivant together on this eight year exploration of what the Internet could be used for because I just got turned on. They were the only ones that knew how valuable it was. And so, you know, they started making all this crazy stuff like digital libraries, mobile software, knowledge robots, which were basically early forms of AI. Yeah. For eight years, they were just like, let's go freaking wild and just made all this crazy stuff.

But in the background, this is over the course of eight years, there's a lot of stuff developing in the internet community. The National Science Foundation decided that they wanted to connect 3,000 universities to the internet.

And remember when the internet got turned on, it was only 400 computers. Yeah. Computers, not even just in universities, like 400 computers. Oh yeah. And they want to connect 3000 universities. So what they were going to do was build a multi-network system or something called an internet backbone. So basically like there are clusters of computers of servers that run the internet. Right. Right now we have like Google and Facebook and like,

These are all these servers that are just giant clusters that route your data around. That's called an internet backbone. And then all of a sudden, the Department of Energy and NASA decided they also wanted to create these big clusters, these internet backbones. And by the mid 80s, there were suddenly four different massive government internet backbones that were growing. And that required a lot of networking equipment.

So during this period, there was suddenly a huge influx of networking equipment manufacturers that were popping up to create the backbones, right?

Because suddenly you're like, oh yeah, we need 3,000 universities connected, but we have no equipment. To fill that need. Yeah. Yeah. So people are going to come in and make money. But something happened around this time that made Vint realize that he needed to get the internet into the hands of the public. And that the internet was going to be a very big deal. Because again, still only for research universities. Yeah.

but it's growing so quickly. But Vint has this story where in 1988, he went to an internet trade show, which just became a thing at the time, which is basically just networking equipment manufacturers that were showing off their stuff. Yeah. But he saw all these companies that had these huge booths and huge displays and huge TVs that were like,

And he's like, oh, my God, like people are putting real money into networking equipment. Like this is like millions of dollars of stuff. Yeah, it made sense. I mean, over this time, it became pretty clear that personal computing was blowing up. The Apple Macintosh came out in 1984. Personal computers were getting really popular. They had their own personal software that ran locally. People were getting computers in their homes. And Vince like, oh, these are computers and we have the Internet. I mean, yeah.

they were made to go together, right? - Hey, you got your computer and my internet. No, you got your internet and my computer. - But the governments and the universities were the only ones that had access to the internet, which for the people who knew how useful the internet was, was really freaking useful. So Vince at this trade show and he's seeing how much money is being thrown at it and he realizes that if the general public could get access, that the internet could be a self-sustaining entity.

Because, again, like these millions of dollars are being put into the industry that are only being sold to the government, military and universities right now. But what if the public suddenly became a customer? You know, you have this really limited amount of people and it's still blowing up. Then all of a sudden it's just like everybody, the biggest industry of all time is about to explode.

Yeah. And I mean, and government funding always is like, it's a question mark. Like this in the start of this all happened because the government basically stopped funding ARPA, DARPA, whatever it was back then. Exactly. They got to the point they were so small that they had to work on this project. Yeah. I'm sure they've got that in the back of their minds. Like we could lose this again. Yeah. Yeah. Yeah. It could be like a huge waste of money. Right. But the problem was one big thing needed to change before this was going to be put into the hands of the public was

That was the law. Because even with stuff like MCI mail, there's only one application that people had access to. If you wanted to be on the actual Internet and you wanted to grow, something needed to be done to get people on that thing. And Vint went full gangster mode. He's like, how can I break the law? Nice. Yeah. And so I remember thinking in 1988, how are we going to get the general public access to the Internet?

And I thought, well, there's a problem because there was something called an appropriate use policy of the government backbone networks. You could only use them for government-sponsored research. That was the basic rule. And so I had to figure out how could I break that rule? How do I break this rule that only allows the government access to the Internet? So he has this idea. What if he's able to put MCI mail, which he just made for a private company,

the commercial email service, not just on MCI's network, but on the internet, like the internet internet. He's like, well, first of all, this wasn't even necessarily made to be on the internet. Like I don't really know the protocols were made for a local network, not necessarily the internet. And I don't even know if this is going to work, but I also just want to break the rules because I want to see if I can get commercial traffic on the internet. Right. Yeah.

So in '88, he goes to the Federal Networking Council and he asks them, "Can I hook up MCI mail to the internet?" And they said yes. Nice. There are no rules. Yeah. It's easy to break the rules when they just let you do it right off the bat. Exactly. It was like he was very, like, I'm not even sure if they knew what they were agreeing to. They probably didn't because he's an internet person and they're just like, "Oh, technology. I don't know. You do your thing." Right?

But they did. And so now, all of a sudden, anyone that was using an MCI mail was also interacting with people on the Internet. So suddenly you had commercial traffic flowing over a government service hosted by public universities. None of these things are ever supposed to touch. These are three separate organizations that are not supposed to be intertwined. Can't be commercialized unless we get the government to change its policy.

So as a relatively new employee of CNRI, I left a couple of years into working there, I went to the Federal Networking Council and asked them for permission to hook up MCI mail, which is a commercial email package, to the internet. My purpose was twofold. I wanted to see what would happen if we did that because the two systems were not necessarily designed to do their work. And second,

I wanted to break the policy logjam that said you couldn't put anything commercial on the Internet. I figured until we did that, we'd never be able to turn it into a self-supporting system, and it would never spread, at least not much.

Well, much to my astonishment, they said yes. Which is just crazy. Yeah. So naturally, other email providers would start saying like, wait, that's not fair. Because at this point in time, MCI mail was not the only. Oh. Yeah. There were other ones that were popping up. Yeah. And everyone's like, why does MCI get to be on the internet and we have to have our own little local network? That's a private organization that has like a weird deal with the government. Very strange. Yeah.

So the government becomes in a jam or as it puts it in a log jam. I mean, they let this happen. They open the floodgates. Right. So they had to let other businesses connect to the Internet, too. They couldn't just only let MCI connect to the Internet. So you can imagine the surprise when people suddenly start. Suddenly they're not in these little silos of services anymore. Like imagine that you're using WhatsApp.

and you get a whatsapp message in telegram or i message all of a sudden they've got these separate apps and you're used to using whatsapp and all of a sudden someone who messages you on through telegram just comes into your whatsapp you're like that's that's really weird that's basically what was happening like all these independent protocols just opened up and pretty rapidly infrastructure started to pop up to route people to the internet these were these became internet service providers right

Because all of a sudden the public has access to the internet through these commercial services. And these things like MCI mail or their competitors became the internet service providers that we know today. So you needed the person to be the at-home private citizen having their stuff connected into this now network of government and private citizens and universities and everything because they couldn't get to there before. So someone had to

MCI was connected now to the internet, but they had to connect you into the internet because you had just been connected to MCI before. So this is a funny thing. People are technically, you know, like on the internet now, but in a very limited way, like they're just trading data. They're not browsing the web. And notice that I used the used web for the first time. Things don't actually start heating up for another 10 years.

There's just all these periods of time between all of this stuff. And now we get to the creation of the World Wide Web. So what year are we in now? 1999. Hey, Ellis and Adam here from the Waveform podcast. Despite the fact that we've been researching this story for almost a year, both of us somehow missed David saying that the World Wide Web was invented in 1999.

The World Wide Web was actually invented by Tim Berners-Lee in 1989. So if you will, for the next three or so minutes, just imagine David saying 1989 every time you hear 1999. Thanks. So now we get to the creation of the World Wide Web. A man named Tim Berners-Lee at CERN came up with an idea. You may have heard of this man.

So the internet's already running. It's already connecting the world with very basic applications. But isn't the point of the internet to share information and not just share data and applications? Yeah. Yeah. Are you going to say something? I agree. Okay.

Yeah. So that's happening, of course, but like primarily in the sense that you were going to be using individual web pages, you're going to specific web addresses and websites and you're consuming the information that was available on those websites. That was all that was happening at the time. So like, how do you research now? Like what's the best way to jump around to different web pages on the internet and find what you need? Search search engine. What happens when you search on Google?

It crawls all of the internet. And what does it show you? And shows you a list of websites related to your search. And what is that when you click on one of those things? What is that called? When you click a URL? Yeah. Forwarding? Hyperlink. Hyperlink. Oh, you click the hyperlink. Hyperlink, right? Hypertext. Yes. Yes. Hypertext transfer protocol. Yeah. We're getting there. We're getting there. Boom, boom. Hypertext. Hypertext.

Yeah. So it's information that tosses you to other information. It's like the foundation of information discovery. Like there's a, there's those Wikipedia games where they're like, how do you get from this Wikipedia page to this Wikipedia page as fast as possible? Fun fact. This is total tangent. My friend made back in high school, he made an app.

He called it wiki golf and literally the entire point was like it would give you a par four and it would give you the two subjects and it was your job to get from one to the other. That's awesome. And we couldn't believe it didn't take off. I'm so good at that game. Yeah. I don't want to brag. We need to bring that back. We do. I'll take anyone on. It could be a studio video. Yeah. Yeah. We used to play that in middle school all the time. It was pretty fun. Yeah. Oh, I'll lose. Yeah.

So hypertext was actually invented way before 1999. The term was actually coined in 1963, but originally it was useful for jumping around a single document, right? Like quickly getting to the references page or that kind of stuff. But like what if hypertext, the thing that was actually very useful for finding stuff on one document, could link to other stuff on the entire internet?

because at the time it was just like click here to go down to this part of the page but what if you clicked here and it went to a website you didn't even know about yeah like a table of contents for the entire internet fire very cool yeah so at cern all these physicists are showing up with their own little networks and their own little protocols and the scientists would need to access information from all these universities around the world so if you connected hypertext to the internet

Suddenly you could just jump around the internet and find all this information so much faster. Like imagine you're just trying to like do research and you know it's somewhere on the internet, but you have nowhere to look and you can't do search.

Like there's no hypertext on the internet. That's annoying as heck. Yeah. There's no Dewey decimal system. Right. For like knowing everything. Yeah. It's like how we used to remember people's phone numbers. Like you have a limit to that. Right. You can't remember everybody's phone number. Yeah. Well, maybe. I don't know. The brain's pretty cool. Anyway, Tim Berners-Lee creates the World Wide Web, which most people don't know is actually different from the internet. Like a lot of people say like the internet and the World Wide Web.

But just like the Internet is a super network that covers all networks, the World Wide Web is all the hyperlinks that connect to all the other hyperlinks. Yeah, it's a web of hyperlinks because they're it's like this interconnected thing. Right. The Internet was already there, which was a network to cover all networks, but it only became a web when they started linking, connecting to one another.

So Tim's like, okay, well, sick. Let's connect all of these sites together with hypertext. So the next couple of months he works and he builds a very early internet browser, which he literally calls World Wide Web, which is the first internet browser. And then later he renamed it to Nexus because he wanted to avoid confusion with the World Wide Web, which was like the idea of all the hyperlinks connecting to each other and the product World Wide Web. Okay. Yeah.

It wasn't the kind of browser you're probably familiar with today. That really started with Netscape and Mosaic, which came out a couple years later. Classics. Yeah. Yeah, it wasn't like a GUI-based interface. But it was basically a document reader that just allowed you to jump around to other documents. Can I interject real quick? You guys keep saying GUI. Oh, yeah.

Graphical user interface, which is just having things to click on and having photos. I dig it. Yeah. I'm team GUI. Team GUI. The first, what is pre-GUI called? Text UI? DOS? TUI? I feel like it's referred to as DOS a lot of times. DOS was a text UI. Yeah. Okay. And you had to know all the commands and it was like in terminal or command. Oh, okay. Was it TUI?

I don't know if it had an umbrella term, but I'm going to call it 2E. A command line interface, CLI. Okay. Yeah. Something like that. Yeah. So this picks up really, really freaking fast. And like that, the browser wars are suddenly on in full force. You've got all these browsers that are popping up that are trying to make the most market share possible. And they all offer different services and all this different kind of stuff. Toolbars. Toolbars?

And that's generally where the history of the internet podcasts, books, and other stories begin is 1999 with Tim Berners-Lee. Because pretty soon after that, the browser war started. You got Mosaic, which was trying to make the web look like a magazine, and Netscape and AOL. There's so many podcasts that cover this era in really intense detail, so I didn't want to go over that again.

But that is the secret history of the Internet, how the Internet got started. But what's next, right? That was the question I had. And it turns out Vint is working on a version of the Web and the Internet that he thinks is going to be the future of the Internet right now.

Right now. He's working on that now. He's working on that now. Wayne Gretzky's working on a new thing now. We should probably hear about this. Do you have a guess for what you think it is? Web 3. You say Web 3. It's not Web 3. It's not Web 3. Okay, that's good. This is Internet Zaddy we're talking about. Okay, okay. Web 4, baby. Web 4. Four-piece suits. No, I don't know. I imagine it's got to be a way to connect an even bigger set of things, but I can't really think of a bigger... The Mind. The Mind.

Oh, wow. I don't know. I don't know. You're correct. What is it? An interplanetary internet. Oh, true. The biggest set of things. The biggest set of things. That is the biggest set of things. Yes. That we know. So, yeah. The reason I wanted to cut off the history of the internet and not do the browser wars and stuff, there's so many podcasts, so many books that cover that in such intense detail. I wanted to cover the secret history of the internet, which was what came before it. And then...

the interplanetary internet. Yeah. Right? Okay. So now we're gonna take a left turn pretty hardcore here. Let's go to Mars. But it's very cool. It will not take that long, but it's very cool. Okay. All right. So, the first successful Mars mission was in 1976 with two Viking landers, right? Yeah.

But at that point, for about 20 years, they couldn't do it again. They couldn't land more stuff on Mars again for like 20 years. Things were just crashing. They weren't having any success. And then miraculously, about 20 years later in 1997, there's this mission called Pathfinder and it finally works.

And it's driving around, it's taking photos, and it's transmitting data from Mars. Pretty cool. I think that we probably think of the more modern Mars rovers right now, but we actually had them back in 1997. It was actually taking data and sending it to us. Very cool. So Vint, in 1997, flies out to JPL, the Jet Propulsion Laboratory.

After that mission and they start seriously talking about something they might need 25 years down the line Right because they just set up the internet they were like well We should get started on like everything that we do related to the internet takes like 15 years We should start work on the next thing now. Mm-hmm. So they're like well, what if we had an interplanetary backbone network and

There's these backbone networks that were getting set up by the National Science Foundation and stuff on the internet. What if we had an interplanetary backbone network? They're originally just going to use TCPIP for the interplanetary backbone network. But because of the delays, because the distance between planets, it would be...

the delay would be way too high. Right? Yeah, so what's the stat now? So if we want to send a command to the Curiosity River on Mars, we're a certain number of light minutes away from Mars. Right. And so that latency is gigantic. Yeah, so from Earth to Mars, when they're closest together, when Earth and Mars are the closest possible together, they're 35 million miles together.

Which means that it'll take three and a half minutes to get a signal at the speed of light from Earth to Mars. And when Earth and Mars are the furthest apart, they're 235 million miles apart. So that's like a 40 minute round trip. TCP, not exactly built to handle round trips that take 40 minutes. Because remember, if you're dropping packets right now, speed of light on Earth, you just resend the packet, you notice no change. Right.

But imagine it takes 40 minutes to get some data over there and you drop some packets. The ping is crazy on Mars. It's crazy on Mars. And then you have to take relativity into account and like all this stuff and time and all that stuff's weird. Okay. We talked about that in the James Webb Space Telescope episode. Little plug. Go listen to that if you haven't listened to it already.

Yeah, so it's not built for 40 minute round trips. That's definitely not going to work. And there's also the fact that the planets are rotating. So you've only got a small period of time when you can transmit data from the planet to a node in space because the nodes in space, the like satellites that are rotating around the earth and stuff go behind the earth and then they can't send the data anymore. Right. Big problems.

So they decided they basically needed to make a mesh network that stores information that can be sent. They were going to need a whole new suite of protocols that took relativity and the speed of light into account. Yeah. So they start getting heavy on this math stuff. Yeah. Yeah. So they started working on this new protocol suite.

And they're working on this. And then suddenly in 2004, there's another successful Mars mission. And the original idea was to just send data directly from those rovers to these huge satellite dishes that were scattered at very strategic points around Earth. There was one in Canberra, Australia, one in Madrid, Spain, and one in Goldstone, California.

So these are like as separately apart as you can make them. And they're like 70 meters wide. They're massive satellite dishes because you're sending data through Mars. You've got like, you know, the information is going like this. You got to have the biggest bucket as possible to collect as much data as you can. Just on the same page, 70 meters. That's about 220 feet.

Is that true? You're welcome. Did you do that just now? Yeah, it's like 3.1 feet per meter. Oh, that's easy. He's European. 229 feet, Marcus. Sorry. Just for those United States weirdos. Anyway. Imagine a 230 feet satellite dish. Yeah, that's huge. That's huge. Yeah. Yeah. Anyway, they try to do this and the data rate is about 28 kilobits per second.

Not fast. That's very slow. It feels like the internet in this building. Anyway, yeah, it's really slow. And the researchers did not love that. So while the data is transmitting, well, they start trying to do this, right? They're sending data from the Mars rovers. They're transmitting the data. And the radios on the rovers start overheating. That's really bad because you don't want to burn out the radios on Mars rovers that are like insanely expensive and took forever to get there.

Like if you imagine you just send these rovers over and they just start transmitting data and they just burn up like they just that's no fun. Yeah, they overheat. So at this point, they came up with a new thing called store and forward, which is different from packet switching.

So we talked about packet switching before where you're just sending all this data through the path of least resistance. You drop a packet. You send it again. It's fine. Storen forward is a little bit different where instead of just sending like a steady stream of data and like taking a bunch of different routes, you send all of the data directly to one node and you make sure that all of that data gets to one node before it goes somewhere else. But it becomes a mesh networking system.

So instead of the rover sending all of the data through space and hitting these satellite dishes to try to get all the information there, you're actually sending it to like these different satellites in strategic points in space. Okay. So when you say mesh network, we are literally talking about...

several satellites strategically positioned between Earth and Mars that will relay data to each other on the way to Mars to minimize packet loss. And I don't know what that does to speed, but I imagine that's faster. Because if you create a network and all of these different satellites are closer together, there's this thing called the inverse square law where the power of data being sent through space is

like goes down over time. Okay. So if you, you can send a lot more data if the things are closer together effectively. I feel like I have a baseball analogy that Ellis might appreciate. It's like a cutoff throw from the outfield, right? Rather than throwing from center field to home base and letting it just like

dribble in really slow you hit someone else on the way in and then they throw it again right so you had multiple points on the way imagine if you had like seven cutoff throws you could throw a much bigger object than a baseball no that's kind of it's exactly it it's like picture instead of an outfielder trying to throw someone out at home you give them nine baseballs and they have to throw over and over again each base one baseball to each base

It's like the percentage of those that are actually going to hit is very low. But if you write little numbers on the baseball, you know, you can reorder them into the correct thing. Whereas this space internet thing is much more like waiting for the perfect moment and then throwing one baseball. Make a giant baseball. Make it have everything. Throw it.

Catch it. Yeah. Yeah because so much so much of normal packet switching TCP IP stuff is just Blasting out information and assuming you can send it again. Well, no and then putting it back together again at the other end Right, whereas this is much more about like wait for it. Wait for it Yeah, okay, yeah, that's interesting that sounds like a lot of expensive hardware - it's called store and for store and forward networking, right? So at this point they've been working on this new protocol software and they're like wait, I

So they stopped sending the data from the Mars rovers because they're burning up. And there's like, hmm, there are X-band radios on the rovers, but there's also X-band radios on the orbiters that are orbiting Mars that they were using to create a map of the surface of Mars to know where they wanted the rovers to land. Mm-hmm.

So you've got satellites that are orbiting Mars, and you've also got the actual orbiter on Mars. And they both have this specific type of radio. So what if you could upload these new protocols that Vint had been working on for the last number of years, because again, he got started on this a while ago, and see if you can use this store and forward method instead of the packet switch TCP IP direct transmission, right? Yeah.

So they push these new protocols to the orbiters and to the rovers. They basically update the software to be a completely different protocol. And now when the orbiters circle overhead of the rovers on Mars...

The rovers now push all of their image data, their sound data, whatever data they collected to these orbiters at 128 kilobits per second instead of 28 kilobits per second. They're not burning up because of their proximity. They're way faster.

And this kind of kicks off this whole push for this new interplanetary internet system. Because imagine we have different strategic orbiters or satellites placed at different parts of the solar system. If you have all these nodes that are closer to each other, you can just start ping-ponging information between these different orbiters. Wow, I have so many questions. Yeah.

Just like the possibility. I mean, we talk about like having a whole bunch of things in orbit around Earth and how that's kind of a lot of debris and possibly interference for observation of space because now there's just stuff in the way. I'm sure those are all things being considered. Definitely. Yeah. But I mean, the cool thing is like in the James Webb Space Telescope episode, we talked about Lagrange Point 2, which is a very special metastable point where the James Webb Space Telescope is being pulled out.

at the same velocity in all directions, which makes it basically super stable. And every planet has five different Lagrange points, right?

And I asked Vint, can we put one of these nodes at a Lagrange point? Because then you could just place different satellites. Instead of orbiting a planet, you can basically have them stationary in space. Relative to the Earth. Yeah. Or whatever, you know, Lagrange point. Because the Earth is moving around the sun, and then the sun is moving around the center of the galaxy, and the galaxy is moving, so it's all relative. It's all relative, but stationary. Yes, but stable relative to Earth. Right. And...

Because Mars is going to keep spinning separately from Earth. And so we probably just have a ton of stuff. But relative to Earth, at least you have this satellite that is super far away from Earth but is way closer than Mars is to Earth at any given time. And if you can place enough of these network nodes throughout the galaxy –

Is that so cool? This is pretty wild. Solar system. Solar system. Eventually, maybe the galaxy. Yeah, we'll start with our solar system and we'll work on it. I just like the idea that you have little nest Wi-Fis just floating in space, transmitting data. This reminds me of a MrBeast video I want to pitch to where to see if you can get a Wi-Fi access from the East Coast to the West Coast by just planting...

mesh network points all the way across the country. I'm pretty sure it would lose so much power. It would just probably end up with all we need is 128 kilobit per second. If you can even get one bit, that would be pretty awesome. Yeah. Yeah. So one thing that Vince really excited about is the amount of commercial interest that's been pouring into this stuff, even in just the last three years, since like commercial interests have finally caught up and kind of over like surpassed government interests when it comes to

Space exploration with things like Starlink SpaceX, you know all these companies now want to throw these orbiters into space So it's gonna be cheaper and easier to maybe get SpaceX or somebody to be like throw up an orbiter at Lagrange point 2 so you can create this interplanetary Black backbone network. Yeah. Yeah, I guess they're all slightly more unique missions now because

Planting things on the way to Mars is seems like a bunch of spots We haven't been to in space before yeah, but yeah, that is a really fascinating idea I think I remember having this conversation maybe Maybe with Neil deGrasse Tyson. I'm not sure but about how the internet on Mars right now could be Completely different from the internet on earth because we're accessing all these servers on earth. We talked to the satellites We're all using earth internet. Yeah, and then on Mars

It could just be a brand new internet. So obviously, Vint has been thinking about this much longer than I have because he's thinking about how to connect the two to the point where maybe a live stream from somebody...

Twitch streaming on Earth could be live streamed to Mars and actually kind of work, maybe. Yeah, exactly. I like the idea that Vint was like, I'm in the Hall of Fame for the Internet, but that's not good enough. I want to be in the intergalactic. We were doing this interview with him and we're trying to be all respectful and polite and proper. And we end the interview, we stop recording. And we were just like, wait a second. How did you also know space stuff?

You just asked him, like, you're like this internet god. You're a computer. You're like on the cutting edge of computer science. How did you also know how rocket ships work? And he just casually, he was like, oh, yeah, like right after high school, I did some engineering work for NASA. JBL. Yeah. Just casually. Yeah. Also, should I share this? I've got a fun fact. So...

We didn't really know where to put this in the episode, but it's too fun a fact not to mention. And that's that John Postel, the IANA, the guy we focused on last episode. In the DNS episode. Exactly. Vince Cerf, the main guy of this episode, and Steve Crocker, the first internet guy that we... All three huge foundational people in the internet.

all went to high school together. - Yeah. - What high school? - Van Nuys High School about 10 minutes away from where I grew up. Vince Cerf and I have the same cell phone area code. - That's pretty intense. - Yeah, they literally all went to high school together. They all separated and didn't really know each other. And then they did graduate work together and built the ARPANET together.

Wow. Isn't that crazy? That school's Hall of Fame goes so hard. It goes hard. That's sick. Yeah, it does go hard. Yeah, so there you have it. That's the secret history of the internet and the upcoming potential interplanetary future of the internet, which is quite cool. I'm excited for Mars internet. I hope Mars internet is... This guy is still scoring goals, man. He's still putting up satellites and thinking about things.

I hope Mars Internet is pretty fire. Yeah. Yeah. So before we depart, well, first I want to have a little conversation. Did you guys learn anything new that you didn't really know before? I knew all of that. Yeah.

No, I learned a lot. Yeah. Yeah. That's anything. I learned that if I ever meet Vint Cerf at a bar, I will buy him a beer. I thank him for the Internet. Whenever I tell people like, oh, yeah, I've got an episode coming up. We interviewed this guy named Vint Cerf. They're always like, you talked to Vint Cerf. Like he is like an Internet god to a lot of people. Twice.

Yeah, twice. Twice. Yeah, twice. And the second time we called him and he was like, good to see you again, David. And I was like. You start bawling your eyes out. Yeah. So that's the secret history of the internet. But I have one more thing that I want to do with Adam and Ellis right now to sign off. I just have one message that I want to tell you guys and also all the viewers that are watching and listening right now. I have buttons. Oh, I didn't get them. Unbutton it. Unbutton it. Unbutton it. What? What?

Wow. Is it in focus? Nice. For audio listeners, they've all coordinated shirts that say IP on everything. It's spelled I-P-E-E. No. For the audio listeners. This is the shirt that Vint wore to the internet task force meeting. Very nice. I assume there's a striptease coming. Thanks for watching.

That was it. Yeah, sure. That was good enough. That was the strip tease. And yeah, that was a good learning experience. The secret history of how we accidentally made an internet and we'll soon have intergalactic internet, which will be awesome. Yeah. It's kind of all folded into a lot of the other episodes that we've done. Go watch the I Can and the Seven Keys of the Internet, all about the DNS, which we skimmed over here. We talked about the...

New Space Race, which is about sort of the satellites that are throwing into orbit. The James Webb Space Telescope episode, which is called How the James Webb Space Telescope Sees Into the Past or something like that. We got a lot of cool space and internet and light and data episodes. So go check those out. Leave your requests for other crazy long form stuff you want us to tackle in the comments below because I will be reading those and we'll get some fun ideas from that. And we'll make David go crazy over them. And we'll have some shirts made, of course. And maybe I'll order pizza. Thanks for watching.

Catch you guys in the next one. Peace.

Apple intelligence coming fall 2024 with Siri and device language set to US English. Some features and languages will be coming over the next year. Zero dollar offer may not be available on future iPhones. Next up, anytime features may be discontinued at any time. Subject to change, additional terms, fees, and restrictions apply. See att.com slash iPhone for details.