Radia Perlman: Primarily my role has been in inventing technology that has been getting used in the Internet. So, my philosophy on technology is that people shouldn’t have to understand it in order to be able to use it. And I’m not a real fan of gadgets. I’m not a first adopter. I’m a last adopter kicking and screaming of all these things.
But I understand things conceptually. So I’ve designed things that make the network self-configuring. It just…you plug it together and it works. Robust. Meaning that without me, some of the design, some of the alternate designs, if you just blew on the Internet it would fall over and die. Which is very…bad, because on your a personal computer, if it gets into a bad state, which it does all the time, you know that you turn off the power and turn it back on again. But in a network there’s no on/off button. So the network can’t get into these weird states.
So the things that I designed made it so that the network would…you know, be self-fixing and self-organizing. There’s a lot of things that other people designed that are fancy features that work if you’ve configured them exactly right but if you make mistakes it just will be horrible. And so I’ve tried to talk to people like that sometimes, and they claim their customers love to configure things and they never make mistakes. And I don’t believe that. So my contributions have been a sort of doing the opposite, of making things very simple designs, very scalable, and very robust.
Intertitle: Describe one of the breakthrough moments of the Internet in which you have been a key participant?
Perlman: So, the one that I’m kind of most known for is the spanning tree algorithm. Now, with that is is that Ethernet was this technology that allowed a bunch of machines to all sit on the same wire. Just like in a conference room, if there’s no moderator calling on people who raise their hands and you just have to talk when you want to, it’s good if you follow the Ethernet rules. It CSMA/CD. CS means “Carrier Sense,” means listen before you start talking. If someone else is talking, don’t talk. MA is “Multiple Access,” meaning be aware there’s other people that are sharing the same bandwidth. And CD is “Collision Detect,” meaning if you’re talking and somebody else talks, stop because you can’t be understood if both of you are talking at the same time. And I’m kind of amazed at how many people don’t follow that. But at any rate, that only works for maybe a few hundred nodes, or else there’s just so many collisions when people try to talk at the same time.
So, I understood Ethernet as being a cheap way of hooking a few hundred nodes together, but it still was only one kind of link in the network. And the stuff I was designing was how you move data around from one link to another. But, kind of to my horror, people thought that Ethernet was the new way of doing networking. So actually I believe that if they called it Etherlink instead of Ethernet the world would not have been quite as confused. But because of that people were not doing the stuff that I had done, and instead only building applications to talk over Ethernet. Which would mean only in a single building, only within a few hundred nodes.
So I tried to argue and people said “Oh, go away Radia. You’re just upset because nobody needs your layer anymore.” And I said, “Oh, but you may want to talk from one Ethernet to another.”
So as it turned out, a few years later people realized “Hey, we do want our applications to work beyond a single building.” So that was where my manager, because I do distributed algorithms, he said “Oh, invent a magic box that will sit between two Ethernets and move packets around without the end nodes implementing the same language as the switch.” You know, like IP as a Layer 3 thing. So without that, with only Ethernet.
And so with the constraints that we had at that point in time, which was that there was no spare fields in Ethernet, the packet had a maximum size so that you couldn’t add any fields to it, I had to figure out a magic way of moving them around. So it was a fairly simple idea, but the problem was that there had to be only one way to get from any place to any other place, or else packets would just kind of move around and multiply and it would be a disaster.So that’s why the spanning tree algorithm, where spanning means “reaches everybody,” tree means “no loops.”
So the story of inventing it was kind of funny. Which was my manager asked me to design this on a Friday, and he was going to be gonna vacation the whole next week. So that night I realized oh my goodness, it’s really simple. You know, with no configuration whatsoever it just all works, and I could prove that it worked. So Monday and Tuesday of next week when my manager was gone, I wrote the spec. And I did it in enough detail that the implementers got it working in just a couple of months without asking me a single question. But then I had three more days to the week when my manager was still gone, and this was before people would send email or cell phones or anything. So I couldn’t concentrate on anything else because I was so excited.
So I spent the remainder of the week working on the poem that goes along with the algorithm. And it’s the abstract of the paper in which I published it. So the poem is called the “Algorhyme,” because every algorithm should have an algorhym. And the poem is
I think that I shall never see
A graph more lovely than a tree.
A tree whose crucial property
Is loop-free connectivity.
A tree which must be sure to span
So packets can reach every LAN.
First the Root must be selected
By ID it is elected.
Least cost paths from Root are traced.
In the tree these paths are placed.
A mesh is made by folks like me
Then bridges find a spanning tree.
An algorithm for distributed computation of a spanningtree in an extended LAN
So another contribution that I think that I’ve made is the books that I’ve written. Where most books are just telling you the exact details of what’s currently deployed, as if students should just memorize that but it doesn’t teach you to think critically at all, and it’s full of acronyms and marketing FUD and so it’s hard to read but not technically deep, it’s sort of very superficial, mine are quite the opposite. They’re easy to read but thought-provoking. And I don’t just say “Here’s all of the details of the one thing that’s deployed,” I say “Here’s a conceptual problem. Here’s seven different ways that it could’ve been done. Here are the pros and cons.” And then I’ll say, “And by the way IP did this, AppleTalk did this, IPX did this.”
And some professors, according to the publisher, say “Why is this book telling my students stuff that they don’t need to know?” Because no recruiter is going to ask you whether you know AppleTalk. As if students’ brains are very very tiny and if you fill them with any knowledge that you don’t need to know, then other stuff won’t fit. But it actually empowers them to be able to see other ways of doing things.
But at any rate. So I wrote this— Oh. Before that, even though I did really important things, it was kind of the stereotypical thing that at a meeting people would not hear what I said until someone told tall and pompous said it. So unfortunately you know, relentlessly self-promoting people do really well, and other people…you know, don’t get recognized. And that would’ve been the case for me had I not written that book.
So I wrote the book and it became kind of the book at exactly the right time. And that changed my professional life, because I didn’t have to act all pompous and sneering and condescending in order for people to take me seriously. It would still be the same thing that they’d be introduced to me and just take one look at me and just sort of yawn. It was like “Yeah yeah.” But then they’d hear my name, and I was like “Oh! I learned the field from your book.” So that was really cool.
But the funny story about it was that a couple years after it was published, I had a four-hour layover in a Chicago airport. And I was just so tired, and I didn’t know what I was going to do for four hours. So I was just wandering around in a daze. And for some reason the airport was almost deserted. I passed by this group of like five men sitting together and I thought I heard some of the right words, like “bridges” and “routers.” But I wasn’t sure, because I wasn’t quite close enough. And one of them had a book on his lap that all I could see was the edges because there were papers on top of it.
So, without thinking I just decided in this empty airport to maneuver myself to sit right next to them. At which point they stopped talking and they stared at me. And I realized oh, I’m acting bizarre. This is embarrassing. But then I said to them, “Oh I’m just curious, what is that book?” So he pulled out the book and it was my book. It was Interconnections. So I said, “Oh! I’m Radia Perlman.” And they all turned completely white and said “No!” And then when one of them finally kind of recovered well enough to talk, he said “We are in a panic trying to prepare a customer presentation. And we don’t know what we’re doing. And right before you showed up, one of us said ‘If only the author of that book were here.’ ”
And I said, “Oh, it’s not a problem. It comes with the book. You wish for me, I appear. How can I help?” So I helped them with the presentation, I signed their book, and I never had a better time at an airport.
Intertitle: Describe the state of the Internet today with a weather analogy and explain why.
Perlman: Yeah, that’s a very good question. It’s…excitingly wonderful, and excitingly terrifying at the same time. So, I’m not quite sure what kind of weather that signifies. But it’s transforming society. The fact that you can buy things from all over the world. The fact that someone with an innovative product doesn’t have to do millions of dollars of marketing, or have a storefront where lots of people can reach them is amazing. The opportunities for bringing education and knowledge of the outside world to remote places is unbelievably exciting as well.
But there are really terrifying things, which is all of the scams that happened, just…hundreds of them every day that you see. And the fact that there’s still people getting taken in by this is really kind of astonishing. There’s just the annoying in-your-face pushing advertising at you. Like making you watch videos before you can watch the video that you want. Not ever doing things in written form, so if you’re just trying to quietly read some news, your machine suddenly blares some ads for things.
There’s misinformation as well that gets propagated. So, how do you know whether anything’s true or not. There’s no overhead whatsoever to posting things. So a lot of people think that society is getting so much better-informed because there’s so much data on the Internet, but it actually can be just the opposite. Because you can choose what you want to look at. And so whatever your bizarre views are, you can find fifty people across the world that believe the same views. And if you only read the things that this little community posts, it winds up with people getting less-informed and society getting more polarized. And that’s very scary.
Intertitle: What are your greatest hopes and fears for the future of the Internet?
Perlman: One of the concerns is that privacy is just absolutely gone. So we can blame the younger generation for posting stuff on Facebook without any notion of privacy. But that’s not really the problem. Because even if they didn’t, all of this information is harvestable by stuff on the Internet. Everything’s on the Internet.
There used to be a separate phone network, and telephone calls would work. And these days since even if it’s not a cell phone a lot of the links are going to be over the Internet, and we just are accepting the fact that phone calls are going to be of unacceptable quality a lot of the time. And this is really unfortunate.
But yeah, in a society where you can’t hide, no matter…every transaction that you do, somebody can figure out where you are, is extremely scary. So with a fairly benevolent government, it’s not so bad. But if the government decided to do some sort of genocide and hunt down everybody of a—it would be extremely difficult to do any sort of resistance or hiding with basically everything that you do being mirrored and accessible somehow over the Internet.
Intertitle: What action should be taken to ensure the best possible future?
Perlman: I’m not sure I know. Because it’s so wonderfully convenient to have everything connected, and everything accessible. To be able to pay your taxes online. To be able to vote online. To be able to shop online. All of these things are wonderful. But it also leads to the complete lack of any ability to have privacy. And the ability for anyone to post any information is also incredibly wonderful. But how do you filter out correct information from incorrect information. And I’m not sure that I have any answers that can help either of these things.