Dave Farber: I first became famil­iar with some of the tech­nol­o­gy when I was at Bell Laboratories in the late 50s and ear­ly 60s and had read some of the doc­u­ments that Paul Baran had writ­ten talk­ing about pack­et switch­ing. And then when I left Bell Laboratories to go to the RAND Corporation, I actu­al­ly worked with Paul and con­tin­ued to be inter­est­ed in it.

I then went down to the University of California at Irvine as an aca­d­e­m­ic and actu­al­ly got involved with it in two ways. One, stu­dents were inter­est­ed, and I was inter­est­ed, and a good friend of mine from many years ago, Bob Kahn, was work­ing at DARPA and was lead­ing that project. So by the time it was done I end­ed up with a small research project look­ing at how to secure cer­tain aspects of the ARPANET. And one of the relay machines, the IMPs there, my stu­dents there includ­ed a stu­dent from University of California at Los Angeles, was Jon Postel. So he was my grad­u­ate stu­dent. We were work­ing on a the­sis on net­work reli­a­bil­i­ty and net­work per­for­mance. But that got me intro­duc­tions to oth­er stu­dents, Steve Crocker, Vint Cerf, etc.

So I start­ed get­ting more and more involved. Also at Irvine, we were cre­at­ing prob­a­bly the world’s first real oper­a­tional dis­trib­uted sys­tem. And we were using a fair­ly nov­el local net­work called a token ring and devel­op­ing many of the tech­niques that are now being used in the cloud ser­vices. So that’s actu­al­ly what I’m get­ting my fun­ny lit­tle award for, among oth­er things.

A num­ber of my stu­dents down there like Paul Mockapetris went on into these things. So in some sense my aca­d­e­m­ic chil­dren became some of the fathers of the Internet. It’s why some peo­ple mum­ble that I’m the grand­fa­ther of the Internet. 


Oh, that’s fine. Being a father’s a pain in the neck. Being a grand­fa­ther’s great. Believe me, I’ve done it. 

After stay­ing in LA for awhile, I went to the University of Delaware (go back east where I was from orig­i­nal­ly), and again got involved in some of the ear­ly net­work stuff with stu­dents. Some of them, Dave Sincoskie and oth­ers went on to do sig­nif­i­cant work in that area.

But what I did along with some of my col­leagues around the coun­try is we noticed that— This was dur­ing the post-Sputnik era, and sud­den­ly tech­nol­o­gy was every­thing. And there were lit­tle com­put­er sci­ence depart­ments in almost every uni­ver­si­ty around the world. And espe­cial­ly around the United States. And all of them were small. And we real­ized that that was­n’t going to work for long because they had nobody to talk to. So we proposed—actually three of us—proposed to the National Science Foundation that they fund an exper­i­ment in cre­at­ing a net­work using basi­cal­ly Internet tech­nol­o­gy but also using telephones—whatever we could get our hands on—to allow uni­ver­si­ties’ com­put­er sci­ence depart­ments to talk to each oth­er through email.

And we got mon­ey to do that, cre­at­ed the serv­er envi­ron­ments we need­ed. We were using first ARPANET tech­nol­o­gy, but we also estab­lished inter­con­nects between the CSNET and the ARPANET, which had some of the big schools on it. And we also extend­ed it to indus­tri­al research lab­o­ra­to­ries. So pret­ty soon we had a web of large­ly com­put­er sci­ence peo­ple who were count­ing on that.

We then wrote a cou­ple of papers for Science and made some pro­pos­als and extend­ed it to uni­ver­si­ties as a whole—everybody in the university—and went on from there. It grew almost out of con­trol at that point. So I was PI in almost all those activ­i­ties. I chaired the National Science Foundation advi­so­ry board on net­work­ing. Once you get in that mess, you stay in that mess. So that’s how I got into it, largely.

Intertitle: Describe one of the break­through moments or move­ments of the Internet in which you have been a key participant.

Farber: Some of the break­through moments were the estab­lish­ment ear­ly on of an inter­tie between the ARPANET (the mil­i­tary exper­i­ments; they were on mil­i­tary net­works) and the CSNET. That involved basi­cal­ly mak­ing an agree­ment with the gov­ern­ment to allow us to pass traf­fic back and forth and estab­lish the notions of peer­ing rela­tion­ships that are now stan­dard. So that was done. 

Looking at anoth­er one, we estab­lished an accept­able use pol­i­cy, which I lived to regret but I wrote the first one. And that said what could run on that net, what could it be used for. And rather than mak­ing a very con­strict­ed one, we tried to make it as lib­er­al as pos­si­ble. And that meant a lot of peo­ple could use it. They did­n’t have to only do com­put­er sci­ence. That was one.

Also the estab­lish­ment of inter­con­nect agree­ments with for­eign… We were Johnny Appleseeds. We’d walk around with a tape to Japan, we find some bright grad­u­ate stu­dent there and he’d bring up that soft­ware. And then we had an agree­ment with them that we could inter­tie with them. And so those mech­a­nisms which we prop­a­gat­ed to almost every coun­try in the world I think were the begin­ning of the idea that the Internet (well, it was­n’t the Internet at that point), that the net was a glob­al phe­nom­e­non. Not sub­ject to cen­sor­ship, not sub­ject to pay­ing for it. You pro­vide your own facil­i­ties, we were there to accept your mail and deliv­er mail, and oth­er things.

And that was one big thing. And then work­ing with­in the NSF to pri­va­tize some of the net­works. Because orig­i­nal­ly they were net­works designed for research, aca­d­e­m­ic things, and it was much more than that. And putting into place the mech­a­nisms which allowed any­body to use it I think was a very key point. And then all the con­trol­ling the growth and every­thing. I think once you got to that point it was unstoppable.

Intertitle: Describe the state of the Internet today with a weath­er anal­o­gy and explain why.

Farber: I think it’s stormy for a num­ber of rea­sons. I think pri­va­cy, secu­ri­ty, are major prob­lems. And it’s going to cre­ate many a storm. You’ve seen some of them over the last month or two with the Snowden thing. But that’s a small plot. 

Governments have caught on to the fact that the Internet is pow­er­ful. It’s uncon­trol­lable. Governments don’t tend to like to have things that are uncon­trol­lable. So just the attempts of gov­ern­ment to reg­u­late, to con­trol the Internet, in an envi­ron­ment where it’s almost impos­si­ble to do that. A friend of mine put it nice­ly, he said the Internet routes around cen­sor­ship. You know, it’s very dif­fi­cult to con­trol it. And that’s going to cause a lot of stress. Countries, and you see that with Europe, are try­ing to say, Ah! We now have an excuse to say we want to for­bid US com­pa­nies from oper­at­ing here because we don’t trust them.” That type of balka­niza­tion I think could destroy the util­i­ty of the net.

The oth­er things I think that are very dan­ger­ous is the fact that governments—and they’ve always been doing it—want to know every­thing about what you’re com­mu­ni­cat­ing with. And who you’re com­mu­ni­cat­ing with. And that I think is going to cause a lot of prob­lems. Some of them will take a while to devel­op. But peo­ple are begin­ning not to trust, you know, and the Internet is all about trust.

Identity prob­lems on the net are a big prob­lem. There was that famous New Yorker car­toon that says, with the dog on a com­put­er say­ing, On the Internet nobody knows you’re a dog.” Well, nobody knows who you are. And that cre­ates addi­tion­al problems. 

So I’d call it stormy, in the sense that a lot of poli­cies are being made now large­ly by peo­ple who don’t under­stand the net cul­ture. Or the net tech­nol­o­gy. And they’re mak­ing some bad mis­takes, and are going to run into real trou­bles with it.

The oth­er thing is the net is not a reli­able beast. It was nev­er built to be stressed the way it’s been stressed.


We were try­ing to see if it could be done. It was a rel­a­tive­ly small group of peo­ple. We trust­ed each oth­er because we all knew each oth­er. And so issues like secu­ri­ty and robust­ness… Robustness was engi­neered in, but secu­ri­ty you did­n’t wor­ry about. So you built a set of of rules, pro­to­cols, which large­ly ignored secu­ri­ty because who’s going to—

Well, now it’s so huge that it resem­bles soci­ety as a whole, and you you have these end­less pen­e­tra­tions. And it’s not an easy job to fix them. We’re sit­ting here at the IETF and try­ing for years to fix them. We could’ve maybe done it years ago when they were a thou­sand peo­ple on the net, ten thou­sand, but now they’re mil­lions and mil­lions and try­ing to do any­thing is dif­fi­cult. Because it’s not just the net­work it’s the com­put­ers you use on it.

So…I’d call it stormy. And it’s going to get a lit­tle stormi­er for a while.

The oth­er thing that’s hap­pen­ing is that tech­nol­o­gy is chang­ing very fast. And there are lim­i­ta­tions in the net­work that’re going to mean that it’s going to have to be redesigned at some point in the future. And the redesign will be fun. But how to ever get it out into the field is not at all obvious.

Intertitle: What are your great­est hopes and fears for the future of the Internet?

Farber: One of my hopes is that it becomes a very ubiq­ui­tous thing that peo­ple can trust, peo­ple can use with some assur­ance that it will not bite them back. (Best way to say.) When you use the tele­phone— We were always able to wire­tap a tele­phone. But there were a set of laws that made it dif­fi­cult. And in fact wire­tap­ping a tele­phone with tech­nol­o­gy we had, we could do it but there was no way to process it.

Now, cour­tesy of Google and many oth­ers, you can absorb all that, you can ana­lyze it. And so you can find inter­est­ing things like you’re talk­ing to him, and he talked to him, and he talked to a bad guy and there­fore by def­i­n­i­tion you sud­den­ly are a bad guy. No, that’s not a log­i­cal thing but it will hap­pen. People will be pun­ished for things that they do. And that’s dangerous.

My hope is that we can do some­thing about that in the short term and it’s not just tech­nol­o­gy. It’s law. It’s con­vinc­ing law­mak­ers who nat­u­ral­ly don’t like that, that their eco­nom­ic health depends on a healthy, open Internet. And there are a lot of peo­ple around here work­ing on that—ISOC works on it, and oth­ers. But when you go to Washington it’s not clear that they under­stand it. Or any oth­er capital.

Intertitle: Is there action that should be tak­en to ensure the best pos­si­ble future?

Farber: Depends on the coun­try you’re in. Education. Education not only of the cur­rent lead­er­ship, but future lead­er­ship. That’s hap­pen­ing in most uni­ver­si­ties. Something hap­pens, though, as peo­ple get old­er they maybe get con­ser­v­a­tive. But they for­get what they did as kids, as col­lege stu­dents. And sud­den­ly they pass laws that they would not like if they were— But that’s the way to world works. 

I think you have to edu­cate. You have to edu­cate how impor­tant the Internet has become in the world econ­o­my. If you sud­den­ly could stop it, it would­n’t just be incon­ve­nient, it would be a cat­a­stro­phe. A lot of things would­n’t work. Lights would go off, every­thing would go plooey.” And so it’s a beast the health of which I think is going to dri­ve the world for­ward. Without it I think we will prob­a­bly spi­ral now into a very bad eco­nom­ic depres­sion, which is not healthy.

I think edu­ca­tion of peo­ple is very very impor­tant. And edu­ca­tion of peo­ple who will help deter­mine that. For instance the uni­ver­si­ty should have intro­duc­to­ry cours­es in some of the social pol­i­cy issues. The tech­nol­o­gy, they all know how some­how it works, maybe a lit­tle bit of that. But more the soci­etal issues. How it’s used, what its con­straints are. And that will help. But it’s going to be a rough time because it’s like the auto­mo­bile. A lot of old busi­ness­es go out of busi­ness. And we have to make sure we replace them.

And we have to make sure that we don’t destroy things. Like we’re well on our way of mak­ing it almost impos­si­ble for whistle­blow­ers to con­tact a reporter. We’ve removed the clas­sic ways of doing it and the Internet has tak­en its place but peo­ple don’t like it so they try to some­how stop it by con­trol­ling the net.

Intertitle: Is there any­thing else you would like to add?

Farber: I think it’s very impor­tant to edu­cate peo­ple of all lev­els, all dis­ci­plines, of the…let me call it the ethics of the net. What’s an unac­cept­able role of gov­ern­ment and of pri­vate enter­prise and what’s an accept­able role for cit­i­zens. And every coun­try has a dif­fer­ent— I don’t think we will ever get to a point where there’s a uni­form thing. But with­in coun­tries, we should be edu­cat­ing them what their rights are and what their oblig­a­tions are. And we don’t. 

Further Reference

Dave Farber pro­file, Internet Hall of Fame 2013