Happy Fathers of the Internet Day

were gonna say Punky Brewster, not that one. Yes, I do remember my two dads when they asked us if we wanted to do a Father's Day episode. We were trying to figure out how we could do something that related to tech stuff, and so Jonathan and I decided we wanted to talk about the fathers of the Internet. And as it turns out, there were two to some

massive exponential power. We're going to be talking about a whole lot of fathers of the Internet because well, there were no two or one, or or seven people that were responsible for the Internet as we know it today. There are tons and tons of people who worked on different aspects of the Internet, right, Yes, this is uh, the Internet is a thing that grew out of the

contributions of dozens and dozens of people. But really, in order to kind of narrow our our scope a bit, we thought we would really concentrate on our Pannet, which in itself is sometimes considered a father of the Internet because it's a predecessor of the Internet. It itself was

not the Internet. It was a It was, however, a network of computers, of of heterogeneous computers, so in other words, they were not all the same type of computer, which was a big deal because, uh, prior to the Arpanet, there really there was really no way of of networking together various different models of computers because they all had their own proprietary languages that they worked on, and one computer's systems was not necessarily compatible with any other computer

system to the Internet. From arponnet, I am your father. End of line. Um, yeah, that's true, it's not possible.

Well anyway, that and whole that end. Uh, the quote I know is not exactly accurate, but because someone will write in um, yeah, that that was the big thing, And there is a myth, if you will, that the Arpanet was designed by the United States government to create a defense network or a network to share information between different parts of the defense network real life meat space network UM sharing for nation in the event of a catastrophic attack of some sort two ferry information from one

group to another. And that's not exactly true, although hey, it is a nice reason. UM the government. Different parts of the government were at that time looking for ways to share information between computers, and so were UM different parts of UH university networks. And those were really that the two UH types of organizations that were really interested in networking computers back in those days. But they all

had UM. If you are a regular tech stuff listener, you know that all these different kinds of machines back then were really very proprietary. You didn't have UM. You know, you didn't have a system UH where you made computers, and they would all run the same operating system like you do now with Windows or for example, or Linux. UM. Even IBM systems. I'm sure you know the system three sixty would run in a different operating system than a different kind of IBM computer because it was designed to

run on on that specific machine. And of course you had dozens and dozens of different computer manufacturers on making these machines. So you know, you had the problem of trying to communicate between UH one and another, so you needed a protocol that would bridge the gap. Yea. So let's take some steps back before we get into the development of that actual protocol, because I think the journey

there is pretty interesting. Oh yeah, I just wanted to say that if you were going to network these machines, which was the goal, that would that would be one of the things you have to do. So in fact, the one of the first people to sort of envision something that that kind of encapsulates what the Internet is, but in a in a smaller way. Is h Venever Bush? I s. Yes. He was a very important fellow in the United States history. He worked on the Manhattan probably ject, yes,

which we've talked about before. That was the project where for a few beads we were able to build an entire city. Wait, no, I'm thinking of something else. Yes, yes you are that was that was actually Manhattan. Yeah. Uh. Bush was very involved with the Defense Department in during during World War Two. UM a very very smart guy and uh he helped with a Manhattan project as well as as other projects. UM and one of the things he really wanted to do, and you can read about

this in an article in the Atlantic. It's still up on the website called as We May Think. Yep, he um. He came up with well, he was a founder of Raytheon, and he came up with this idea he called microwaves,

he called memes and memeics. In his mind, it was going to be a micro film based machine, because that was sort of the epitome of the technology at the time, and the idea being that he would have a database of all the information that he would ever need, an entire library of information, and also that he could create information on this system, and that he would be able to access this information in a very natural way because

the system itself would mimic the way people think. So this was kind of a or not maybe not even mimic compliment, perhaps will say, complement the way humans think. So that way, when you search for something, it would be able to bring back the relevant information to you and not just any information that just happened to have that term in it. We you know, we talked about this in our our episodes about semantic web and artificial intelligence.

Is this whole idea of contextual information that a machine might be able to recognize. Now, this was something that he was just sort of theorizing about, but it was a very important element of what would go into our bonnet and the Internet. Now, the next person I would talk about besides Bush is Joseph Carl Robnett Lick Lighter or j c. R or Lick to his friends. Um, it's it's funny because I've read a paper that Lick wrote. Um,

and uh that why. I don't want to get too far ahead of where Jonathan's going with this, but um he if you think, well that this this machine, this memics um would be able to call up any basically you would you would take a document and take a photo of it and create a microfilm from that and store it. Because you know, computers back then we're huge. They weren't um reliable like they are now. They were giant. Um, so that was kind of impractical, and so are a

large cabinets full of paper. So that was kind of the idea was, Hey, I'll shrink this down into his size where I could just keep it in a desk and find stuff. Um. Well, and you say, wow, that that's remarkably like the Internet. Well, in a way, yes it is, but it's also not that remarkable because Lick has said or um, of course I read this a long time ago, but I remember him saying in this, uh this article that it's not an accident that the

Internet sort of mimics the memics. It's the idea was so compelling to the early founders of the Internet, and they said, we like this, Let's try to make this happen with the technology. We have to change it from theory to reality. Yeah. Yeah, So it's it's not completely an accident that what the memics. Um, the concept of the memetics is very much like a computer attached to the Internet and being able to pull up information. And so lit he was working in back in nineteen sixty two.

He was working at Bolt Baroneck Newman Newman, which is also better known as BBN. That's going to come up

again in a little bit, yes it will. He worked for bb IN and then came up with an idea called he was calling the Intergalactic Computer Network because why think small, And this was the idea of being able to network computers together so that you are able to share information between machines in a very efficient way because you know, the the challenge was that you had all these really smart people working on at the time, very

sophisticated machinery, but all of their work was siloed. All of that work was contained within the physical building that that machine was in. Yeah, if you want to share that, that means you have to box that up and send it or or you had to go and use it. You had to you had to yourself plan a trip out to wherever that machine was and share information that way. There was no way of sending it electronically at all, and so you could call somebody and read it to them,

but that's still that's electronic still. But the Intergalactic Computer Network was this idea that you would be able to actually have computers communicate with one an they're in share information. Well, he became the head of Computer Research Program at the Advanced Research Projects Agency which is ARPA, and he became the head of that in nineteen sixty three and he began he renamed the Computer Research Program the Information Processing Techniques Office or i p t o IP two and

um UH. The Advanced Research Projects Agency is now known as DARPA. We added a d on that UM and it's part of the Department of Defense, which is where that whole concept of they wanted to create a network that could withstand a catastrophic failure UM sort of came from.

And it came from because there were there were movements within the United States government to try and create something like that, but that was not the purpose of ARPA net Um, which would grow out of link Lighters projects now link later to Liqui liquid It or rather taught to several other people and kind of talked about his his concept of this interglad to computer network and got them really excited and they sort of jumped on board and also helped champion the cause and get UH funding

money and partnerships with various research institutions to agree to try and make this become a reality. A couple of those people include Ivan Sutherland, who was another is another really remarkable guy in computer science history because not only was he part of trying to get this network off the ground, he also invented sketch Pad, which was an early graphical user interface sort of the predecessor to things like Windows and the Mac operating system and really any

operating system on any touch device kind of has. You know, you have to go all the way back to Sutherland to thank him for his work. He also was a pioneer in computer graphics. He was a pioneer in augment a reality, in virtual reality, so he was It's interesting because his his focuses were on its folk. I we're on a different area of computer science than network technology,

but he was still very much interested in this. He became the director of I p t O in nineteen sixty four after lick Lighter left the program, so lick

Lighter leaves before Arpanette begins. Another person that lick Lighter taught to and got interested in this idea was Bob Taylor, and Bob Taylor became the director of I P t O from nineteen sixty five to nineteen sixty nine, and he was the one who actually sent out the request for quotation the r f Q for building out our Pinett And this was the project where they wanted to connect multiple computers together through a network and create the

infrastructure that would make that possible. Now, this infrastructure would have to be able to work of various different kinds of computers and send information between them so that the different computers could understand what they were. Each one was saying because, like we had said before, these computers were working on proprietary operating systems that were not compatible. So they send out this request for quotation to one hundred

and forty bidders. Now, most of the companies and research institutions that received these h r f q s thought that the whole idea was just too difficult, outlandish, impossible, just not practical at all. There were quite a few of them probably said why, Yeah, so they want to

do exactly there were some that did say why. They received responses from twelve of the one and forty bids they sent out requests for bids uh and out of the twelve, four of them were considered viable by our PA, And it ultimately went down to a couple of different candidates, and the one that stepped down in the end was bb N, which was awarded the bid on April seventh,

nineteen sixty nine. Yeah, they're they're a very interesting company to don't want to get into too much depth, but they're sort of I think there's sort of raytheon like in that they have done all kinds of different engineering projects over the years, and I always associated them with computing in the Internet, but they their history goes back quite a ways. UM, and they did they they have done much much more than just participate in this pro project.

But they were certainly well suited because they were were UM electrical engineering experts. Certainly and UH Taylor left I p t O because of the Vietnam War actually, and Lawrence Larry G. Roberts stepped in and he was program manager and office director at Areva before he became the director of I p t O. And UH he was someone who had who was leading a team within this, UH, this department within I p t O that was working

on packet switching. Now, packet switching is something that was being worked on by various individuals throughout the entire world independently. So this was not necessarily When when I say that Larry Roberts and his team worked on packet switching, this is not where packet switching necessarily just comes from. There was a a pioneer in packet switching called Paul Baron who was very much working on this, as well as

another person in the UK named Donald Davies Yes. UM also Leonard Kleinrock Yes who ended up working with our pannette through u c l A because u c l A was the University of California at Los Angeles was one of the research facilities one of the universities that became part of the initial Arpanet project, and so klein Rock began to work through our pa net, not directly from ARPA itself, but through u c l A. Yeah, the three those three guys are are known really for

creating the packet switching part of the Internet, the concept behind packet switching on the Internet, the way it's done. Um and uh actually have a quick overview of packet switching again. UM, well, I was going to get to that in just a second. Um. Actually is a very uh, very diverse team. Baron actually came from from Poland. Um and Uh it's actually now in Belarus, but at the time it was in Poland. And Um, he he had an idea of a network. Now there there's several different

kinds of networks. Um. Obviously, you can have one computer in the center to which other machines connect, which is known as a centralized network. And then you can have a decentralized network. Um. If you think about I think Napster kind of used that concept. We have multiple servers, but the machines connect to those main computers, which are in turn connected to a main main computer. UM. Or you know, there are many important computers and lesser computers

connect to those, but those that's a decentralized network. UM. But uh, but Paul decided that what he should do is create a or what what you should use for this particular project is a distributed model. UM. And that's the way the the internet works. Uh. It's it's a network of networks and each computer has redundancies. UM. So

what does that mean? So if we were using a decentralized network, um, and you sent an email to the net to the main computer that your computer was hooked up to, but that computer was down, your email wouldn't get where it's going. Um. In a distributed network, there are multiple routes that that email can go. So the computer that you were supposed to send it to is down, but there's another way to go. So it goes through that computer and then from there to another computer and

it basically follows that path. Now. UM, what they realized, these guys were that this redundancy is good, but they they're actually using redundancy of redundancies on the Internet, the network of networks and said, okay, well, uh, it would be kind of cumbersome to send this all as one file. And if we send it all is one file and it gets bogged down in the direction it was supposed to go, it's not going to get there. So they broke it into pieces called which they decided to call packets.

And um, each packet is represented more than once. So let's say, just for the sake of convenience, that you send an email and that's broken down into ten packets. Um. Those ten packets are sent to multiple locations or through multiple locations to get to the end point. And one way or another, those ten packets are supposed to reach there and be reassembled on the other end in the

correct order to complete the transmission. So the packets, uh, you know, if if one of those computers is down, let's say, and packets three and seven are lost, they have also been sent to other computers in the path, and they will be reassembled in the correct order. Um. And using using that method, Uh, packet switching has become the way we send information over the Internet, and that

goes for tiny little files to great big chunks. So Larry Roberts had decided that this packet switching technology, which had was in its infancy at this point, was probably the best way to go when you're talking about having a networked series of computers and you want to get this information from one to another. UM and so that's

kind of the direction that the team went in. And also the folks at the various institutions where the computers that were going to be connected together, uh, they also were working on this along with the folks over at our path. So let's, uh, let's let's talk a little bit more about where these computers were. There. There were four initial host computers for our pannett. All right, these four hosts were the four computers that were going to

be connected together. One was at u C l A and it was the University of California at Los Angeles. It was an s D S Sigma seven computer running the Sigma Experimental operating System. Uh. Then you had the University of California's color Fried Interactive Mathematics Center, which had an IBM three sixty slash seventy five running on the O S slash m VT operating system. You had the University of Utah, which had the d e C p DP ten computer with the with the ten X operating system,

the old Deck machines. And then you had Stanford Research Institute and they had an SDS nine D computer which ran on the Genie operating system. So all four of these machines are different machines running different o s s. I'm sorry, did you mention that two of those machines were made by the same company yet they use completely different operating systems? No? I did not, and so then yes, but that is true. There are two two of them were made, Um, we're sts computers, but both of them

were running different operating systems. So yeah, even the same company's machines wouldn't necessarily talk to each other in the same way. That's true. And so those were the those are the four institutes as well. So you've got u c l A, Stanford, University of California, and um, you've got the University of Utah. Uh. Now, the team over

at our PA that was in charge of putting this together. Um, actually what happened was not really the team at Arbo, the team at BBN, because that was the company that won out the bid. The team at BBN consisted of Frank Hart who was the leader of the team, Dave Walden, who was a programmer and real time systems expert. You had Will Crowther who was another programmer. You had Bernie Cosell, who was a d bugger, so he was someone who would find all the stuff that's wrong with the programs

that everyone else is making. You had, Yeah, you had Bob Cohn, was a computer theory expert and an error control specialist and also was the guy they went to when it came to how do we send data across telephone lines? That was kind of his master network protocols. You had several Orne Stein who was a hardware specialist UM, and you had another hardware engineer by the name of Ben Barker who joined the team a little bit after the initial UH team got together, and Uh, I think

that's everybody. I think I got everyone who was on the that initial team. Now, the approach they made was kind of interesting. You know, how do you create this network where these four computers that do not speak a common language, how can you get them to communicate with each other? And they came up with a hardware and software solution. So really it's a hardware and protocol solution. The hardware approach they took was to create interface message

processors imps. YEP. A guy named Mike wing Wingfield actually came up with the the interface that would connect a computer to the IMP. Yeah. So these are these themselves are computers. These are but these were unlike now the computers that we're talking about at these these colleges were those enormous computers that you think of in the old you know, pictures from the sixties room full of computer. Now, the were comparatively speaking, much more simple than those. Yeah.

I actually have a photo um from u c l A that shows uh Klein rock and a picture of an IMP, which I'm gonna show Jonathan. Um. It's it kind of looks like a stereo amplifier actually built into a wall. It's it's not it's not very exciting to look at, but very necessary. At these lights are blinking out of sequence. Will give them the blink in sequence. Yeah, it's it's definitely a Star Trek computer. Yeah, or stuff blink stuff what blinks? And Yeah. So the the imps

were designed to be the go between. They're actually kind of like routers. There were gateways is what they were. Yeah, and you connected these physically to the host machines with a serial connector, a proprietary Cereal connector, and these then the imps themselves would then connect to phone lines that had been LEAs for the project. Uh. And they could run at about fifty bits per second. That's the data transfer speed that they were capable of reaching. And the

original imps were Honeywell d DP five sixteen computers. And actually I read an interesting where I thought I thought it was a kind of a funny story Ben Barker when he joined the team, he was one of the hardware engineers. Uh they you know, he was. They received the delivery of the Honeywell five sixteen computer to hook up to a host computer, and he wanted to test it. He wanted to run some code on and it didn't work.

And then he realized nothing was working, and so he physically took the computer apart and physically unwrapped the wires that were wrapped around pins and then reattached them to different pins because he figured out that it had been wired wrong, so he had to rewire the machine from the ground up in order to get to work. Worked sixteen hours a day until he got it to work properly. That wasn't one of the computers, that was just one of the imps. So these MPs, it's surprising that he

didn't just pitch the thing out the window. Of course, I imagine that that would be kind of heavy. So these MPs provide the interface, UH from a physical standpoint, so that you know, you would send a command to the imps. The MP would then send that command along to the MP connected to the host computer you wanted to send messages to, and then that would UH MP would send its message to that host computer that would

receive whatever the command was UM. And there's a lot more to it than that, but that was the basic idea and UH. The protocol that first was used on Urbanet was called Network Control Program in CP. That was actually used for quite a while. It was it was not It was not the most elegant protocol. It took a while, and it wasn't the most uh flexible protocol I there, but it was serviceable. It worked well enough so that it could create these connections between these different computers.

This was the set of rules that the computers had to follow in order for the communications to happen. And uh, eventually they started to add more stuff on top of the basic commands that were built into our ponett at the beginning. All Right, So those those other things included things like email and file transfer that that was built on top of n c P. But n c P was kind of it was not the strongest foundation for that. So we'll we'll talk a little bit about what took

over n c P a little bit later. Um. Anyway, so that was your basic design, and uh, did you do you have the story about the first message that was sent on Arpanett? So so well, first I'll say Arpanett gets connected, uh in nineteen nine. Um, it's kind of funny it actually it actually moved pretty quickly. Um. The the contract was awarded in April, right, but the

the first connection was made later that year. Uh. And in fact, all four of the hosts were connected by December five, nineteen So April the contracts awarded by December five, they've actually managed to create, uh, the the connections that that put these four hosts in communication with each other. Now, the first permanent link what went up on November twenty one nine, and that link was between the imp at u c l A and the MP at Stanford Research Institute.

The other two would join online before the end of the year. And the first message was sent by a student, Charlie Klein, who was a student at U c l A. And the first message was supposed to be asking was it asking for an extension for a term paper? No? No, actually it was. It was being supervised by Klein Rock at the time, so I'm pretty sure that that when the he could have just turned around and asked, dude,

my boss is reading over my shoulder. L O L here. Yeah, the first message was supposed to be log in alright, so he typed the L and you typed the O, and then the system crashed. So the very first message sent across arpanet was low. It should have been Hello world. Yeah, I was just a little short of that, which is you know, of course, the traditional greeting when your first starting up a new program. Yeah, so anyway, that was

the first message was low sent across uh that. Now, after about an hour of fiddling with the system, they got it back online image to actually do a remote log in. So it was a successful test, which is sometimes shorter than the amount of time it takes me to log in once my systems crashed, that's true. And yeah, and so Arpanet was once once again. Towards the end of nineteen sixty nine, the Arpanett was was working. Now,

that was a great achievement, but that was just the beginning. Uh. That's when we started seeing the various people working at these research institutes as well as within our PA and at BBN build on things like email file transfer protocol uh and that sort of stuff, where um it became a way to expand the features of arpanet. Uh. In nineteen seventy two, it was Ray Tomlinson who developed the

electronic mail system for Arbonett. Yes, he's he's the one responsible for finding new life for the AT sign on your keyboard. Yeah. He he chose that as the symbol to join the names of the recipient and the host computer so that the the system would know what computer to send it to and then what user on that computer would get the message. And that's why our email

addresses today have that AT symbol on there. It's a carry over from that, yes, yes, which I actually kind of like that he chose that particular symbol because since it has the A in the center. It kind of helps, you know, newcomers recognize, you know, it gives us a mnemonic device to go at at. Of course, people have been trying to come up with the actual name for that thing for many years. Um, and it's been hotly

debated in some circles. Anyway, some circles. There was also there was a student at U c l A who worked in Klein Rocks department. Who have I heard of this guy? Yes, you probably have. He met Bob Cohn at U c l A. While Cohn was working on the Arpanet deployment. This is a fellow named venton Surf. Oh yeah, vent Surf. Yeah, you may know him from such things as Google. Um, well, hopefully you know him

from such things as the Internet. Right. So he started over at U c l A. And then he moved to Stanford and UH and then in nineteen he actually went to work for Darba itself, but he and Cohn

began to work together to start designing something that would replace. Ultimately, it would replace the n c P protocols because, like I said, they the n c P protocol worked well enough for initially connecting these computers together, but it lacked a lot of the the the flexibility that they needed once they started coming up with new things to add on top of just just communication protocols. So he and Bob con began to work on what we now call t c P I P and uh the Transmission Control

Protocol on the Internet protocol. Uh. This was much more sophisticated compared to n c P, and it took time to design it, uh, and once it was designed, it took some time to transition to it. In fact, it wasn't until January one that n CP was officially retired and our bonnet changed over to use t c P

i P instead. And that I think some people say that day, which they also call Flag Day, that that day marks the birth of the Internet, because that was when we started using the protocols that the Internet relies on in order to get you know, it's the set of rules that we all follow, or that all the machines follow in order to get information from one place

to another. And that's funny because uh, it really it's sort of the birth of the Internet proper is It's sort of hard to uh define in in that way because uh, you know that the drafts for TCP were predated that by ten years um nearly seventies seventy three. Um. Yeah, and and Bob con was working on a way of getting satellite data networks to connect to land based networks. And that's kind of like the The actual growth of the Internet was sort of a natural evolution of various

communication networks connecting together. I mean that is the Internet. It is a network of networks and Arpanet became part of that in a way. So Arpanett itself was not the Internet. Arponnet was a predecessor to the Internet, and in some ways was connected to other elements of the Internet. UM. But a lot of the stuff that was developed within our bonnet would become the backbone of what, at least from a protocol standpoint, of what we think of as

the Internet. UM. Yeah. If you're interested, there is a a really uh cool article called How the Internet Came to Be that Surf wrote and he gets into the history of the arpanet and the Internet. UM. But it's really kind of kind of cool because it's written from an insider's point of view. UM. One of the things that I found kind of interesting was that one when

they were working on it, UM, they graduate students. Surf included, UH, this was such a high level project that you know, a quote unquote lowly graduate student wouldn't be allowed to speak with authority. So UM, they had requests for comment

basically saying, so what do you guys think? UM, So, without treading on toes UM, people like UH vent Surf and the other people who we consider instrumental and you know Internet experts were able to uh to speak their mind and I think, um, you know, now we go Wow, they had to ask him to please comment on this. I mean he's he's like, you know, one of the people. But you know, it's just kind of funny how that works.

And UH also mentioned that as their work progressed, the military became more interested in using the network in the late nineteen seventies. Again yeah, because again you're talking about the ability to to communicate between centers very very quickly

share information that could be pivotal to national security. So it was obviously something that the military would be interested in and in taking advantage of these systems that were created for our pannette and either creating a separate military network, which was done UM, or adding on to existing networks, which was also done. And in fact you could there were windows from the Internet into the military networks. Um yeah, this was this was sort of what gave rise to

what we think of today as the Internet. Now. Our pinett itself did not last forever. It was decommissioned on February nine and was Vince Surf had something to say about that. He wrote a poem. Okay, here's here's the poem. Vince Surf wrote to honor the Arponnet when it was taken offline in nineteen. He said, it was the first, and being first was best. But now we lay it down to ever rest. Now pause with me a moment, shed some tears for all langsyne for love, for years

and years of faithful service, duty done. I weep, laid down thy packet now, O friend and sleep. Um yeah. The funny thing was that the Arpanet was the the forerunner of the Internet. And um, but what we learned from that, the all the people who were what we now consider the fathers of the Internet, basically what they learned from the Arpanet was, um, the best technologies to

use in networking, many many computers over long and short distances. Um, and you know, the the the arpon nets spawned many different kinds of networks, one of which I used extensively in college bittenet UM. But you know there were used net and and and many others. UM. And these technologies, you know, are are really the result of the arpon nets UH influence as you know this this primitive and

later you know, more sophisticated computer network. UM. So we definitely derive what is now the Internet and the ability to uh to switch packets and communicate with reliable protocols. Um to to the efforts of these many, many, many many people UH and and UH many of those many many people have been awarded countless awards by governments around

the world. They've received received um honorary degrees by universities and other uh you know, other kinds of accolades from all kinds of people for the work they've done in bringing the world together in communication. So I think that it's only fitting that they also receive from us Happy Father's Day, fathers of the Internet. We got you a goodie bag and a tie you'll you all have to share. It's a networked tie. It's got you all have to wear it at the same time together. It's all got

it's just got zeros and ones all over. It's a binary tie. It's either it's either tied or not tied. That's kind of binary. So anyway, Yeah, that's that's our discussion about Urbanett, the father of the Internet. If you will, and we hope you've enjoyed it. If you guys have any suggestions for our topics you would like us to cover in future episodes, let us know you. Let's go

on Facebook or Twitter or handled. There is text stuff h s W or send us an email or addresses tech stuff at discovery dot com and Chris and I will taught you again really soon. For more on this and thousands of other topics, visit how Stuff Works dot com, brought to you by the reinvented two thousand twelve camera. It's ready, are you