A Festschrift for Danny Cohen

VINT CERF: We are about to
embark on a professionally recorded event. This is Danny Cohen’s
Festschrift. [APPLAUSE] VINT CERF: And we’re so pleased
that so many of you were able to join us. Some others we hope will
show up a bit later. Here’s the plan. Neil Gershenfeld is going to
manage the sequence of events. And he also has a lot of your
slides sitting on his laptop. I’m just here to welcome you to
Google, to tell you that if you need to find restrooms,
please go out this door over here that you came in. And look at the signs. Oh, you can go out
that way, too? There’s another door
over there. That works too. If there’s a fire, please
exit anywhere you can. But walk, don’t run. My name is Vint Cerf. But I think everybody
knows that. And as an employee of Google,
I’m very pleased to be able to provide facilities here for
this special occasion. We are recording
professionally. We have four cameras
going right now. So you may not know when
you might be on camera. So you should act accordingly. We’ve asked you to sign these
releases so that the videos, which should be ready post-edit
by the 25th month, will be available. And we’ll be putting
them up on YouTube. So with that, welcome
all of you. Welcome Danny, especially. I’ll turn this over to Neil
to begin the show. NEIL GERSHENFELD: That’ll be
for the speaker, I think. VINT CERF: Those of you who
are presenting, we would appreciate it if you would stay
within the white lines here– this is for
video purposes– and also to avoid being blinded
by the projectors. And use the hand-held
microphone. That would be very helpful. OK, Neil, you’re on. NEIL GERSHENFELD: My modest role
is to keep time, which I will do firmly. And just to start, I think this
is the easiest meeting I’ve ever organized
in my life. VINT CERF: That’s because
you, Carla. But usually when you organize a meeting, you have to shepherd. I said I didn’t even
get Danny out. I was part way through Danny,
and just everybody said yes. And I think just that alone
speaks volumes for why we’re here today. So in 10 minute blocks, we’re
going to go through this arc of people who’ve lived and
worked with Danny, roughly in order as sorted by Danny. And implicitly, for me, what’s
part of the magic of today is this wonderful microcosm of the
arc of the whole internet as lived by Danny. So with that, I’m delighted
to start with Larry. LARRY ROBERTS: Well, I’m going
to be particularly short, because I don’t remember
everything from that era. But what I particularly remember
was that I met Danny, evidently the first
of most other people, in the late ’60s. But we were talking about
graphics at that point, which I was involved in. But what I particularly remember
was his work with voice, because this has truly
changed the world in terms of how the voice is used
on the network. And I think that his early work
on that, which was in the early ’70s, was particularly
important because it told us all that we had to have a mode
in the network where we didn’t delay things for
error checking. And that was fairly critical. And we really hadn’t
copped up to that. NCP, the original protocol,
didn’t support that, and Vint’s original TCP didn’t. And eventually we
got it right. And voice works. But what it caused me to do is
make a presentation to the National Academy of Sweden,
when Len and I won the alternate Nobel Prize. It’s funded by Ericsson, since
Nobel wouldn’t allow mathematicians. But in any case, we talked
about it there. And my speech was saying
that in 20 years– and this was in ’81 or ’82– in 20 years, we will have
voice on the network. And it will grow to
be all voice. And that is the time scale
that was correct. And 20 years I made up because
I knew the telcos. And transitions are slow. But anyway, Danny, I think,
was very important to the world in that particular
respect. And that’s about all I’m going
to comment on today. NEIL GERSHENFELD: Thank
you very much. LARRY ROBERTS: I hope everybody
else got their check shirts only message. No? VINT CERF: It’s this
crawling video. LARRY ROBERTS: Well, I didn’t
know it was going to be videotaped until I
put my shirt on. Anyway, I wrote some stuff. So I would rather have a desk. But I’ll honor your request to
stay in the middle here. VINT CERF: We want you to
look good on television. LARRY ROBERTS: Thank you. The subject of my talk
is gifts from Danny. I’ve known Danny since 1966. Danny walked into my office at
Digital Equipment Corporation. He was working on a project. And he had some suggestions
that he wanted to have implemented in the equipment
that he wanted. And of course, we had already
completed the entire design and were in manufacturing. And so Danny and I started our
relationship in a somewhat hostile, adversarial way. But then again, Danny
is Israeli. So it shouldn’t be any
surprise to anyone. When he came into my
office, I thought he was another customer. He was representing a software
company that happened to be building equipment for none
other than Larry Roberts for his trip down to ARPA. And so gift number one from
Danny to me was the introduction to ARPA. And as a result of that and the
discussions we had and the relationship we had, I wound up
being a program manager at ARPA from 1967 to 1970. And during that time, I think,
up until about 10 years ago, Danny and I had a very close
relationship, working relationship and personal
relationship. But I’ll get to some
of that later on. The interesting thing about
that time frame– and this is sort of
the next subject– was that Danny, after he
lambasted me about the design of the equipment, we had a
chance encounter walking through the halls of MIT. And he said, well, what
are you doing? What are you taking? Because I had been an
undergraduate at MIT. And Danny had just come. And I was telling him that I
was taking this fabulous course at Harvard given by Ivan Sutherland on computer graphics. And Danny rushed over to
Harvard and tried to get into the class. And I guess you didn’t
get into that class. If you can get that
on video, please. Gift number one was the
introduction to ARPA. Gift number two from Danny
was an introduction to non-commercial flights. So some of you are experienced
in this. And I’ll just go through
two incidents of flying with Danny. The first one was, in fact,
during the ARPA time frame. I happen to be in Boston. We were probably looking
at Ivan’s project at Harvard at the time. And Danny said let’s
go to dinner. And I said, fine, let’s
go to dinner. And Danny said, well, we’re
going to go to my favorite restaurant. And we’re driving out
to the airport. And we get to Logan. And we get in a chartered plane,
or the fleet that he was involved in. And he flies to Cape Cod,
to Hyannis Airport. Just to show you how long
ago this was, parks just off on the side. There’s a hole in the fence. And we go through the hole in
the fence and wind up eating at this little shop. And then at the end of dinner,
we get back on the plane and fly back. So that was number one. Not a bad trip. It was kind of entertaining. My second trip with Danny,
I was visiting much later in Los Angeles. It was at ISI. I guess you’ll hear all about
these places, if you don’t know already. And I was on my way up
to San Francisco. He says, I’ll fly you. I said, fine, I guess. Because the first
trip was great. And so I hop in the plane. But they give them
the wrong plane. They give them a 172 instead
of what he wanted. 172 has a maximum speed
of 120, 130 knots. And there’s 40 knot headwinds. And so I won’t tell you
the whole story. But as we’re flying up, it got
more and more painful to the point where we had to land the
plane, not to refuel, but to empty instead. Not to be too graphic
about it. Oh, sorry, I’m drifting. Is that it? VINT CERF: I noticed
that, too. But I wasn’t going to say
anything about that. Go ahead. BARRY WESSLER: Is somebody
going to tell me? Because I have a lot. The third and sort of the worst
thing was Danny was in Washington actually visiting. And my parents were there. And my parents were going
back to New York. And Danny offered to fly them
back from Washington to New York, because he
had his plane. He had a plane with him. And they were thrilled. You know, here’s an Israeli
taking them back to New York. And I got a call from my mom
when she got off the plane. She said it was wonderful
trip. But don’t ever, ever
do that again. [LAUGHTER] BARRY WESSLER: So that’s
gift number two, non-commercial flying. Gift number three was actually
a business thing. Danny called me up. I was responsible for
international business it at Telenet. And we were building networks
for companies and for administrations for international telephone companies. And Danny said, I know people
at the Ministry of Communications in Israel. So we hop on a plane. And we go to Israel. And we had a fabulous time. We actually got the contract. It was one of these made in
heaven kinds of things. And for those of you who have
not been in Israel with Danny, you should do it, because he
has friends there that are beyond belief. Unfortunately, I guess, none
of them showed up. Or you didn’t put any of them–
or who knows where they are these days. So that was gift number three
was the Israeli MOD. Gift number four was painful. Gift number four occurred
in England. Danny was giving a presentation
on this thing called IP– I wasn’t sure what he
was getting at. I didn’t really understand
all this stuff. I didn’t understand
the implications. We were doing something else. And the unfortunate thing is,
instead of listening to Danny, we did something else. Gift number three. I should have listened. Gift number five, and my last
gift from him to me is the personal friendship that
Danny and I have had over 40-odd years. And that’s exemplified. Probably the most interesting
gift that Danny’s given me is I had an accident. And I was unable to drive. And I was staying at home
or taking taxis. Or my wife or friends
were driving me. And Federal Express shows up. And they show up
with a package. It’s from Danny. And it was a tricycle, because
I had fallen off my bike, which is how I had
injured myself. So that’s physical
gift number one. Physical gift number two
I wear around my waist. It has to do with a tradition
that Danny and I have, which is when I’m in Palo Alto,
virtually every night that I’m in Palo Alto– and for the last eight, nine
years my daughter has been here, so I’m here a reasonable
amount– Danny and I walk over to the
classic gelato shop and have gelato together, sit and talk
outside the gelato shop. And as my gift to Danny– I actually ate the gelato. But if you’ll take the cup. Thank you very much. [APPLAUSE] NEIL GERSHENFELD: Ivan next. IVAN SUTHERLAND: I’m
Ivan Sutherland. I am Danny Cohen’s
faculty advisor. He got a Ph.D. At Harvard. And I think I was his advisor
at that time. And 15 years or so later, he
called me up one time. And he said, I need some advice from my faculty advisor. No, no, Danny. That was 15 years ago. Forget the whole thing. He says, no. He says, it’s a tenured
position. And Dave Patterson used that
statement in a paper that he wrote, I think, in the ACM about
the relationship between graduate students and faculty. And faculty advisor is
a tenured position. And Danny makes that
perfectly clear. When he was a young student,
he came to my house for Thanksgiving dinner
on one occasion. And every Thanksgiving ever
since, I’ve received a phone call from Danny. It’s now 46 years, every
year faithfully. Somehow he’s tracked me down. I haven’t made it very easy. But he’s not hard to shake. I have two stories I’d like
to tell about Danny. The first one has
an introduction and a couple of chapters. The introduction has to do with
the fact that I lived in Salt Lake City. And Danny used to come and
visit me in Salt Lake. He said he liked to come to Salt
Lake City, because it was the only city in the world where
he could be a Gentile. And we used to drive from Salt
Lake Airport up towards the university. And he would say,
stop the car. Stop the car. And we would stop the car at the
bridge that went over the Jordan River. And he would get out and admire
the Jordan River. I don’t think he reminded
him of home. But I think he rather
enjoyed being there. And I think that was part
of the thrill of visiting Salt Lake. Now the Evans & Sutherland
Company at that time– so this is chapter one– Evans & Sutherland at that time
had just finished a major product, which was
called LDS-1. Any resemblance between the name
of this products and the local religious organization
was purely intentional on my part. But I did check it out
with Dave Evans before installing it. And he said, well, it’s
not offensive. And it might be memorable. So it’s OK. And we wanted to make an
advertisement for the LDS-1. You have to remember that in
those days memory was so expensive that you couldn’t have
a raster display and have enough memory to remember
the pixels. So you had a telegraphic
display that would display lines. And I thought it would be nice
to have some text in this advertisement. And so I said to Danny, could
you program up some text for this advertisement? And Danny said, sure,
no problem. And a week or so later, this
text appeared on the screen. Now I had failed to mention
that it might be in Roman characters. It was in Hebrew. [LAUGHTER] IVAN SUTHERLAND: And I thought, this is not offensive. And it’s interesting. It might be memorable. Let’s use it. So this is now chapter two,
which has really nothing to do with Danny, but I think rounds
the story out appropriately. The Evans & Sutherland Company
used as its advertising agency the Evans Advertising Agency. The Evans Advertising Agency at
that time was the largest advertising agency west of
the Mississippi River. They are the guys who invented
the pop out thermometer for your turkey, for example. Now the account representative
we had there was a man named John Dwan. Yes, that really was his
name, John Dwan. And I sent this photograph of
the screen down to John to get the advertisement made. And a little while later, a
beautifully printed, ready copy to go to press arrived. And I looked at it. I don’t read Hebrew. But it was left justified. AUDIENCE: Oh, no. IVAN SUTHERLAND: And I
suspected, because Hebrew’s written right to left, it
should have been right justified and ragged
on the left. But I didn’t know. And there was no one in the
Evan’s Advertising Agency who could check this out. That’s got to be the only
advertising agency in the world that had no Hebrew
capability. So what are we going to do? Well, there was a wonderful
delicatessen in town run by a man named Lou Dornbush. Now Luke Dornbush carried with
him at all times a souvenir of his activities during
World War II. It was a numeric tattoo. And I figured that Lou Dornbush
would be able to answer this question. So John Dwan and I went into Lou
Dornbush’s delicatessen. And we put this advertising
copy on the counter. And I said to Lou,
is this correct? And he looked at me. And he looked at John Dwan. And he looked at the ad. And he looked at me. And he looked at John Dwan. And he looked at the ad. And he said, it’s wrong. That’ll be $150. [LAUGHTER] IVAN SUTHERLAND: Did
you know that chapter two of this story? Now when Danny was a
graduate student, I invited him to the house. And my children were
that size, very well-behaved children. So they called him Mr. Cohen. And he didn’t correct them. He simply started calling Juliet
Miss Sutherland Dean Master Sutherland. And pretty soon, he
became Danny. And now that’s a good thing. Now Juliet worked for
Danny at ISI. He had his first summer job. I arrived at LAX with no way
to get to our home in Santa Monica except taxi. And I thought Juliet is probably
at ISA, which is halfway between. And I could take a taxi there
and ride with Juliet. So I dialed Juliet’s
phone number. And the phone rang. And Danny answered. And I hemmed and hawed a bit. And he said, is this my
friend Ivan calling? Or is this Juliet’s daddy? And because I wanted Juliet, I
admitted I was Juliet’s daddy. And he said, parents of
employees are not permitted to call, and hung up. [LAUGHTER] VINT CERF: Thank you, Ivan. Wow. BOB SPROULL: Hey, Danny. Thank you. VINT CERF: Actually, you can
go back, if you want. Just don’t go any farther
than that. That’s all. BOB SPROULL: You can put a fence
up, you know, like a prisoner in the dock. So I was one of those students
in Ivan’s class at Harvard. I was an undergraduate
at the time. And Ivan had a scheme
for the class. He’d lecture. And then the students, actually,
in the class would go upstairs to a smaller seminar
room where there would be more discussion and work
assignments and so on. And the lecture was actually
open to everyone. And it was quite
well-attended. Ivan was a new arrival on
campus, but known because of Sketchpad, of course. In fact, there were many people
auditing the class who weren’t even from Harvard. So the students, we were a
little puzzled about all this. And it became very clear in the
first session or two of the class, there were a lot of
questions coming from the back of the room, one guy in
particular with very aggressive questions,
and not all that intelligible, but questions. And Ivan was, at first, I
think, actually a little flummoxed by this. But after a while,
things proceeded. But this was not one
of the students admitted to the class. Now at that time, the way you
got into classes of this sort, Harvard had a very quaint little
phrase called admission by consent of the instructor. And sure enough, we had
all traipsed up. And Ivan had consented. So about a few weeks later in
the sessions upstairs, this questioner from the back
of the room showed up. Of course, it was Danny. And we all figured that
Ivan had consented. According to Ivan, it’s
more like relented. But nevertheless. So actually, Danny had scored
a much bigger prize, as you all know, which is he managed
to talk Ivan into being his thesis adviser– for life. VINT CERF: Ivan didn’t
understand what Danny was saying at the time. BOB SPROULL: Well,
maybe not, yeah. So Danny and I actually had
parallel careers in many ways, in which we would connect
in various ways. Obviously at Harvard, and
most recently when Danny was an Oracle. But I remember vividly after
the VLSI multi-chip project had been moved ISI. And Danny was helping stand up
MOSIS, the ARPA sponsored version of that. And Danny was the technical guy
who would go around and explain things. And there was a user’s
group, of course. And there was a user’s
group meeting at MIT. I have no idea what
year it was. And Danny was up talking about
the new things to come in MOSIS and so on. The users, of course, we’re
pretty aggressive and asking lots of questions. And Danny said, now I know
all of you would like second layer metal. You can tell how old this was. I want a show of hands. How many of you would like
second level metal? All hands go up. Then he says, how many of you
would like to wait until second level metal
was available? And of course, the technical
side of this is the MOSIS balance at that time was very
much trying to provide a reliable service for getting
things fabbed, without being so aggressive that things
failed, not because of students making mistakes but
because the fabrication was not all that reliable. So I thought that was the
typical Danny way of making the point. He could have done it
much more crudely. During the VLSI phase, I’m
very proud to say that I helped Professor Finnegan
actually publish an important paper. And Danny may actually
not remember the punchline of that paper. But it solved one of the most
important and vexing problems, or computational complexity
problems at the time. You know, of course, about
Moore’s Law that said, basically, life is good. And Professor Finnegan’s version
of that was, the more VSLI processors in your system
design, to better the paper about it. Remember that, Danny? So around that time, Danny and
I actually collaborated on a paper together. This was for a special issue the
IEEE proceedings that Bob Kahn edited on communications. And it was titled “High Level
Protocols.” And it basically drew on his experience in
doing speech to speech protocol, and some of mine
in doing graphics things. And we were kind of railing
against the complexity that people felt was involved
in designing protocols at the time. And you may remember all those
quaint little diagrams of packets and octets and bytes
that are all made out of every ASCII character known to man. And you can and can’t very
well figure out. And then of course, Danny
pointed out that you had to use either red or green glasses
to look at them, depending upon whether your
computer was Big-Endian or Little-Endian. So we wrote a little paper that
said this is really all much simpler. You really should think about it
more using data structuring kinds of ideas. It wasn’t a deep paper. It was pretty obvious, and
fell like a stone. Got very little notice. So I often exchange emails with
Danny over the years. There have been some
celebrated ones. Some of them are unprintable– at least in this audience,
even though you’re a good audience. But one thing that has come up
over the years, more times than you might imagine, is
frustration with our management at the time. And without fail, though,
there’s something in any of Danny’s messages that
makes me smile. And it was a trace,
if you will, of Finnegan in every response. And I think that’s part
of what we’re honoring here today. Thank you. [APPLAUSE] NEIL GERSHENFELD: Chuck. CHUCK SEITZ: My experiences with
Danny are a lot like the others you’ve heard about. It happened that Danny I went
to Courchevel in the French Alps to teach a class to
professors at European universities on how to
teach VLSI design. This was sponsored by IBM. In fact, they even gave us
some gratuity for that. And so we went. And of course, when there was an
opportunity to go climb an Alp, we did. So that’s the story behind
that particular picture. We’ve had a lot of shared
experiences. I think of them as being in sort
of four epics, plus all those having to do
with children. The initial one– I’ve mentioned in other talks
about how grateful I am to Ivan for opportunities
he created for me. I’m also grateful to him for
being the reason I met Danny, which was at Harvard while we
were all working on computer graphics on steroids. We’ll call it that
for right now– being able to do things in real
time that hadn’t been done before. And that all migrated then to
Evans & Sutherland, which was in Salt Lake City. So actually, when Danny came
to visit you, he was also visiting me. And of course, that’s where he
started with motorcycles. I’m sorry. And he was at my house
in Salt Lake. And I showed him how to work the
motorcycle, how to start it and all that. And then it was gone. He decided he liked that. So he was going to take
that for the day. And he did. So there’s a little extra thing
for that, a little epic. Then there was the VLSI
revolution, which Ivan, again, dragged the both of us in. That was a very, very exciting
period at Caltech. In fact, Danny was briefly on
the faculty at Caltech. But he mostly was at USC/ISI. And there were all of these
people at other universities. The number of VLSI courses in
the United States went from one university to 2 to
10 to 140, and so on, in successive years. So it was a great
success story. Danny was given the job of
setting up MOSIS, DARPA silicon broker. And Neil was, I think, wanting
to have a little bit of semi-technical rather than
merely comedic content. So he asked me to say a couple
things about MOSIS, which we’ll do later. By the time the rubber meets the
road with designing chips, we had done some at Caltech that
were sufficiently useful that Danny and Bob Feldemann
used them to make a network at ISI. And this atomic land was a kind
of research prototype of the networks built by Myricom
in their early years. I did want to also say a word,
one slide about our SDI adventures. SDI was the strategic
defense initiative. Is [? Sevaro ?] here? I saw his name on the list. VINT CERF: He was supposed
to be here, but we haven’t seen him. CHUCK SEITZ: Well, because he
is typically opposed to all such things. But I wanted to show you
a little bit about what that was all about. And Danny chaired that
committee, which did, I think, some rather exceptional work. And then, of course, the early
years of Myricom, that I won’t say anything about because
there are others here– Bob, Ruth, and so on–
who were there. And in fact, Bob Parker was
the program manager for Myricom’s DARPA contract. So here’s the token
story about MOSIS. That was Danny’s name,
of course. Is it a Mel Brooks movie where
they get to this one thing? How about we kill the first
born male child? And he says, no, no–
the name stuck. And it’s another acronym built
on an acronym, because MOS stands for metal oxide
semiconductor. So to try to get as much mileage
as you could out of each fabrication run, something
Carver Mead did in the early days was they put
multiple projects on one die. I’m sorry, I moved
out of my place. VINT CERF: Well, we’ll fix
that with Photoshop. Don’t worry. CHUCK SEITZ: And so that’s
an interesting one. Can you see the one on
the upper right? That was a self-time project
done by a student named Eric Barton, who I believe was
your thesis student. And you can see the
clock on there. The paper that Danny and George
Lewicki wrote about MOSIS as of 1981, they mentioned
that when the MOSIS people first looked at the wafer
under the microscope, it was at the right time. But by the time they took the
photograph, it was about three minutes late. And of course, you had to apply
power to the chip in order to get the
hands to move. And there was a second
hand on it as well. But you can’t see it at the
resolution of this photograph. Was all that correct, sir? DANNY COHEN: About. CHUCK SEITZ: Then in order to
multiply things yet further, the early MOSIS project
started composing many different diatypes on one wafer,
something that Carver Mead et al had not done but Lynn
Conway did in the MPC79. So there’s some history
to this. Some of you may be interested
that the recent IEEE computer magazine has a whole issue about
the VLSI revolution. What’s your batting average
in ARPA contracts, Danny? DANNY COHEN: About 49. CHUCK SEITZ: Yeah,
pretty high. Above one, right? He has got more ARPA than he
has written proposals. So the way APRA works, they
asked IS I to do this because ISI obviously had the necessary
people in computer resources, infrastructure,
and so on. And as long as they were going
to do it, they certainly did a terrific job of it– faster turn, added new
technologies all the time, but within keeping it reliable. Now I’ve asked myself many
times, why did Danny agree to lead this project? I mean, this is not a project at
the same intellectual scope as the other things that you’ve
done in your career, if I may say so, my judgment. And I think it was because ISI
allowed reimbursement for use of general aviation aircraft. And so Danny had a MOSIS
express service between Southern California and
Silicon Valley. And I loved it, because
I was frequently allowed to come along. And sometimes he would be flying
really cool airplanes. That looks like a Grumman
Tiger there. But more typically, he was
flying a Cessa T201 or something like that. DANNY COHEN: Or Mooneys. CHUCK SEITZ: Or Mooneys, yeah. I remember one time, at night
you get tail winds coming from the Bay Area back to Southern
California. And so the plane has
what’s called DME– distance measuring equipment– which will tell you what your
ground speed is relative to a particular VORTAC. How many of you have
pilot’s licenses? Of course, I had to get
a pilot’s license in self-defense later. AUDIENCE: [INAUDIBLE]. CHUCK SEITZ: What? VINT CERF: Self-defence part. CHUCK SEITZ: My theory, Danny’s
a very good pilot because people taught
him how to fly. Now ask me what I think
of his driving. I mean, Danny is the source of,
well, your Honor, why did I run the red light? It was the only color available
at the time. And relative to me getting him
addicted to motorcycles, he had a particularly frustrating
day one day, and went out, took a ride, wrote me
an email about it. He said, oh, Chuck,
it was wonderful. After all this crummy
frustration, I went out. And there it was, the sun
and the wind and the bike and the cop. And it was, of course,
a motorcycle cop. And Danny thought that there
was professional courtesy. So I think, my sneaking
suspicion is, Danny agreed to take this job so that he could
do some more flying. Something for the record– this
was a lot of work that Danny did and I did, and a
committee of 10, 12 people. And the next following year
after the summer meetings in ’85, there was another
group who met. And I served on an office of
technology assessment panel. You remember when the Congress
had an office of technology assessment so the Congress
critters could become a little bit informed about scientific
matters? It was a wonderful
organization. Of course, they got canned. Who canned it? Anyway. They cancelled it. They didn’t want to
be so advised. And actually, the National
Research Council does a good job of producing reports
on things that they’re interested in. This was a more interesting
government panel than most, because of one initial member
who was a believer in software engineering and a promoter
of software engineering. And he went to the first meeting
and then decided that was infeasible. And so he sent a letter of
resignation, not to the people who appointed him but to “The
New York Times,” which, of course, they were eager to
get such a lot to learned description of why all of
this was impossible. Danny and I remain interested
in this subject to this day. IVAN SUTHERLAND: Of course, the
point of the system was not to shoot down missiles,
but to keep them from ever being launched. If you ever had to see whether
it worked, it had failed. And so it didn’t have to work. It just had to be plausible. In that sense, it worked
magnificently. The detractors were wrong. CHUCK SEITZ: And we did our job,
because this report which went out explained how plausible
it is to do this. And furthermore, the problem
was the original Fletcher report about assumed a globally
consistent database of objects and assets
and so on. And we said no, that’s
not necessary. It isn’t necessary anymore than
it’s necessary for the operation of the internet. But of course, the detractors
were constantly, well, yeah, that couldn’t possibly work. But the question is, did our
efforts make any difference? I believe they did. Because what we saw in the
strategic defense initiative office was a complete acceptance
that the way to make this all feasible is to
select an architecture where the programming became easy. But it was still fun to watch
all of the detractors. The most important
thing, children. My children have known Danny
almost as long as I’ve known them, or as long as
they’ve known me. And here’s, for example, a
picture from probably 30 years ago of Danny with
my son Russell. He’s trying to show him some
of the elements of flying. VINT CERF: Like gravity,
right? CHUCK SEITZ: Danny is equally
popular with my daughter. You can go head. And all the flying I’ve done
with Danny, he has taught me a few things. [LAUGHTER] CHUCK SEITZ: Do you like
the animation? Have you ever seen airport
lighting like that? Actually, don’t pay any
attention to the not here, because they the other rest
of it is implied. So Danny, I salute you. My kids salute you. My mother salutes you. Everybody I know salutes you. Not because you’re a general,
it’s because you’re a wonderful person. [APPLAUSE] NEIL GERSHENFELD: Thank
you, Chuck. LEONARD KLEINROCK: Everybody,
this is great. Danny, big day for
you, wonderful. DANNY COHEN: Thank you. LEONARD KLEINROCK: So I’ve known
Danny for [INAUDIBLE] to 40 years. And we have certainly a
professional relationship, but very strong personal
relationship. And he also has a very strong
personal relationship with my wife, Stella. In fact, we share her
in many ways. We both get free medical
advice in her field of expertise, which is
always welcome. But Danny always calls her for
his dose, even before I’ve gotten mine. It’s been a wonderful thing. So there’s the gentleman that
we recognize today. First time I met Danny– well,
among the first times– we were at a den. We were at a den in someone’s
apartment in Venice near the Pacific Ocean Beach. And everybody’s playing
this game called Mem. Tolly Holt had invented
this game. And I walked in this room. And some of the people in this
gathering here were there. There was smoke rising, smelly
smoke of some kind, rising to the various corners. They were playing this
game of Mem. And Danny was there diving in. He was so much involved with
puzzles, gadgets, games, challenges, anything else he
could get his hands on. So let me tell you one of the
early adventures with Danny. It was 1973. We had a meeting on computing
communications in Brighton, England at the University
of Sussex. Many of the people in this
room were there. And we were put up
into dormitories. I had to go home a day early. I got home, start unpacking. And I realize, son of a gun, I
left my electric razor in the dormitory in Brighton. So I said, what night owl might
be up now and on this thing called the internet so I
can connect to him and maybe get my razor back? So I figured the only person I
know who would be crazy enough was not this guy but this
guy, Larry Roberts. So I fired up a piece of
software called [? Viso ?] Sharing Executive, a piece of
software that knew it lived in the network. And I typed in “where
Roberts.” What this thing did was to log
on to every machine on that small ARPANET at the time,
check the who list. And it comes back a few minutes
later on my little machine at home and type out
“Roberts logged in as terminal number so and so at BBN.”
So he was up at 3:00 AM. It was now 3:00 AM in
London, in Brighton. And he was logged on to BBN
doing some damn thing. So I set up a chat session. I explained to Larry what
my problem was. And the next day, basically the
criminal courier arrived with my razor from across the
ocean and gave it back to me. Well, it’s criminal because it
was a personal use of that government sponsored internet. And I think we may have
collectively committed the first crime on the
internet today. Take that off the video. So this is just the kinds
of things you’ve seen Danny talk about. He’s been referenced in all
of these technologies. Let’s go on. This work, of course, on 3D
image manipulation preceded some of the virtual reality. The MOSIS you’ve heard
about as well. And the idea that Larry talked
about when Danny was among the first to say we’ve got to do
something about a technology use that we hadn’t anticipated
in the ARPANET, namely streaming traffic. So let’s go on. Danny came up and began to use
what are called barycentric coordinates. OK, that rotated. [INTERPOSING VOICES] LEONARD KLEINROCK: We were
talking about that. At any rate, the idea here is
when you’re close to an axis, you have the attribute by
which it’s labeled. So if you’re near the horizontal
axis, you have high reliability in your
communications. If you’re in the left-hand axis,
you have short response time, on the right-hand
axis, high throughput. So this blue point is obviously
interactive traffic. You don’t care how much
throughput you get. But you do care about high
reliability and a short response time. The green button points to
high reliability, lots of throughput. You don’t care about
the response time. The thing the ARPANET was not
designed to do is what Danny pointed out. You don’t care about
reliability with streaming traffic. If it gets there late,
forget about it. If it’s wrong, forget
about it. If it’s out of order,
throw it away. But you do want short response
time and high throughput. And you said the ARPANET
can’t do that. And that led to some
revolutionary ideas that helped adapt the internet now
to do all the streaming, voice, media, video,
et cetera. Those are very major
contributions. You heard about Endians
argument here. You can read all
about it here. But you put Endian into Google,
you’re going to get over four million hits today. He was very much concerned
about which bit gets transmitted first. So Danny, as an internet
pioneer, was in there in earliest days when he had these
very beautiful display machines, clinking along
at [? 110 ?] characters per second. So move it along. I know everybody was afraid that
I might deliver a poem. I’m not going to
deliver a poem. I’m going to deliver one
of Danny’s poems. And Danny had this wonderful
comment about the internet, the ARPANET. “In the beginning, ARPA
created the ARPANET. And the ARPANET was without
form and void. And darkness was
upon the deep. And the spirit of ARPA moved
upon the face of the network. And ARPA said, let there
be a protocol. And there was a protocol. And ARPA saw that it was good. And ARPA said, let there
be more protocols. And it was so. And ARPA saw that it was good. And ARPA said, let there
be more networks. And it was so.” And here we are today
with a lot of networks connected together. Danny, you said it right
way back then, and well-articulated. So here’s the accomplice
in crime. We can buzz through
that very quickly. AUDIENCE: How come it didn’t
burn up in the 220 AC? Oh, well. LEONARD KLEINROCK: That’s
why I wanted it back. It was an adaptable one. So here’s Danny at ISI. And you notice his
dropped desk. As a pilot you, he wanted
things down low where they should be. But– click– where’s the phone? It turns out Danny needed
his phone repaired. He left the office at ISI and
told the secretary, when the repairman comes, just send
him into my office. The repairman comes in. He spends a few minutes there. He walks back out to the
secretary and says, I can’t find the phone. So she said, he’s a pilot. Guy went back in. And he found it down on the side
of his desk, just where it should be, almost
in a holster. So Danny was very much
interested in flying, as you understand, both good and bad. And one wonderful story
he told me– and you may have seen this
in his plane as well. He went into someone else’s
plane one day. And he found that all
the dials were twisted in strange ways. None of them read the
English correctly. Rotated 45 degrees,
90 degrees. And Danny told me the reason the
guy did that is when the plane is flying with all of its
proper measurements going, they all point vertically. So if you just scan it with an
eye, you can see what’s awry. Well, I don’t know, Danny, did
you do that to your plane and drive these folks crazy when
you’re flying with them? but that was a brilliant idea
that Danny pointed out. So here is Danny the sportsman,
skiing, biking. Here he is doing voiceover
IP with sunglasses. Here he is without sunglasses. Here he is in a turtleneck. And there he is more recently in
a bow tie, which is not an easy thing to do. So finally, we see Danny
in all of his glory. Danny, congratulations. It’s a pleasure to be here. [APPLAUSE] ROBERT KAHN: Well, Danny and
I go back a long time. And this is a time to celebrate
Danny, both as an individual, as a friend,
and for all of his accomplishments. And I’m not going to go deep
into the accomplishments. I think most of you in this
room, perhaps all of you, know what they are. I’ve been involved in many of
them, mainly from the DARPA side on much of the occasion,
but sometimes just personally between the two of us. But I just wanted to highlight
some of the things that we’ve done together or encountered
together over the years. You may remember some
of them, Danny– like the trips in the plane. We went once through the Grand
Canyon in a very lovely trip from Los Angeles out there and
then back, below the rim of the canyon, right through
the middle of it. And not only was that trip
interesting for me because of just the venue and the visage
and all of that sort of stuff. But this was a trip in which
Danny taught me about hypoxia. [LAUGHTER] ROBERT KAHN: And so
I could add a few numbers in a few seconds. By the end of this trip, it was
taking me a minute or more to add the same thing. And I was, let’s
see, 4 and 3– I think that’s more than 6. Maybe it’s 8. Let me try 9. So I learned all
about hypoxia. We had a great time at the
Paris Air Show– if you recall, Danny– where I got to see we of the
various planes that were being performed, and the names that
I didn’t understand. Like [? Mr. Falcon ?] I thought was M-R
period Falcon. But it was [FRENCH],
the French word. And I think we consumed four
dozen oysters that night at the [INAUDIBLE],
if you recall. I don’t think it was each. I think that was collectively. I don’t know how many of you
experience the unexpected calls from Danny. He’d show up at Dulles Airport
at 10 minutes of 6:00. Fast asleep, I get a
call from Danny. I need to see you right away. We can meet at DARPA sometime. You always had interesting
things to talk about. But it was urgent to
do it right then. And we always did. The story about Danny bringing
the razor back, I have a similar one. I have an old Porsche. And the engine was sinking. And it needed transmission
mounts to hold it. These were kind of oil-impregnated cardboard mounts. So I went to my local dealer. And he said, I’m sorry. They’re no longer available. You’ve got to go to the office
in Stuttgart to get them. Well, these are the days when
the ARPANET was still out there early. And just like Len’s story, one
of the people that worked for us at DARPA was [INAUDIBLE] Jack Hammett. He was an Army colonel
at the time, I think. Jack was en route to Germany,
in London at the time. So I sent a note that night. I had been at the Porsche dealer
at 5:00 in the evening. So about 5:30 I sent a note. Jack was still up. And he said, fine,
I’m going over. I’ll stop. I’ll be in Stuttgart at
the command center. I’ll stop and see if
we can get you the transmission mounts. Next morning, he tries. And of course, they
don’t have any. There’s a corner on the market
by an automobile dealer in Hermosa Beach, California
that I never heard of. It’s called Elijah Polak,
or something like that. So I said, Danny would probably
know where that is. So I sent a note. Danny is up and on the net. Do you know where this is? He said, yeah, it’s
right by LAX. I said, on one of your trips
to Washington, would you be willing to stop by there and
see if they have two transition mounts, because
I need them? He says, well, I’m flying to
Washington in the morning. So why don’t I stop there and
get them and bring them on? So here he is 10 o’clock the
next morning, or whenever. He was on the overnight
flight. He delivers two transmission
mounts. So my Porsche dealer,
I went back and told them I went to Stuttgart. They didn’t have them there. So I went to California here
that I thought the next morning, courtesy of Danny. You remember that? I think Danny had to be among
the most motivated of people I’ve ever seen. Not only is he visionary. But he’s very action oriented. He likes to take ideas
and actually see them come into being. And I’ve always appreciated
that. And let me just give you one
sense of his transportation system, the global express that
you used to run up to Silicon Valley. I actually have a different take
on it, Chuck, than you do, because I used to interact
a lot with Keith Uncapher. And Keith used to run ISI. And Keith and Danny didn’t
always see eye to eye on everything. Danny was sort of like
Jean Valjean. And Keith was like Javert. He had the rules. And so Danny, who had a project
to run, get some wafers done, Friday night he’d
get in this plane, take them up to Silicon Valley,
[INAUDIBLE] them through. And by Monday morning, they
were ready to go. Keith would show up at CNR
sometime, the organization I was running, or sometimes
even before. And he complained about Danny,
because the rules of USC were you were not supposed to
use private planes. And so here I could clearly see,
Danny was trying to be very helpful to the community,
get the job done. And Keith was trying
to hold the rules. And there was no give
here at all. I think Danny finally
won that one. Danny and I first met probably
in the late ’60s when he was doing his computer graphics
stuff at Harvard. And he was trying to ship stuff
between Harvard and MIT. And we had a lot of interactions
and discussions. And when I was at DARPA, I was
trying to get the program and packet speech going. And Danny and I sat down,
had a long discussion. He was talking about what his
future should look like. And of course, ISI was a natural
place to do some of the work on packetized voice. And as a result of that,
Danny could probably explain it more directly. He ended up going to ISI. In the packet speech work in
those days, they were the people working on speech
compression who basically were trying to figure out how to
take speech and put it out with good understandability at
the lowest possible rate, and those people who are trying to
make it work as a system. And those weren’t always the
same types of people. Danny was clearly on the
system side of things. But he had an intense interest
on the other, as I could see it. So when he went on to
ISI, he masterminded the operation there. And the network voice protocol
which he developed became essentially the thing that
made voice happen. We have to work with
BB&N to get them to introduce type 3 packets. BB&N at the time said they
couldn’t do that. But they finally relented and
decided they could when I volunteered to do it myself. They knew that would
be a disaster. But in recognition of that, I
think people now understand that those early experiments
that were done at DARPA really where the basis of what today
is called VIO IP. And it was viewed as an IEEE
milestone recently. And I think, Danny, you deserve
a very large part of the credit for making that
happen, and not only on the packetized voice. But they also demonstrated it. Stephen Casner, I guess, was
involved in quite a bit of that as well on the
packet video side. I think those of you who have
known Danny know that he has a profound sense of ironic humor,
as I would put it, coupled with a kind of
self-effacing modesty. And those two together produce
Professor Finnegan– probably the perfect iconic
example of those two traits put together. Ed Feigenbaum recently
commented about telling stories. It was one of the Computer
History Museum events that some of you may have attended. And he said, in order to have a
story, you can’t just have a lot of successes one
after another. You have to have some highs. And you have to have
some lows– and maybe higher highs
and lower lows. And I think Danny definitely
has a story. He may not tell you the story. But it’s a compelling
one, in my view. Not only has he had many
technical accomplishments, which I said I won’t
go through. But he’s also had some family
losses and other unfortunate episodes in his career that he’s
rebounded from in a very fundamental and powerful way. So it’s to his credit
that he’s been able to navigate that. And of all the things that
he’s done, the one that I appreciated the most by far was
a paper he wrote, which many of you may never have read,
called “Ethics in the Pursuit of Ethics,” where he
was commenting upon people trying to do things in the right
may often do wrong in the process of trying
to do right. It was a wonderful paper. And I always felt very happy to
have known Danny for having written that paper. So I think we would like to
just celebrate Danny for everything he’s done, and
for his friendship. You’ve made our lives
a lot better. Thank you. [APPLAUSE] VINT CERF: You have
a microphone. AUDIENCE: Oh, you wanted
to shake Bob’s hand. He just wants to shake
Bob’s hand. VINT CERF: I thought you
wanted to shake the microphone. Who’s up? Do want to do the movie first? [VIDEO PLAYBACK] -All of the voices in this film
will be heard as they would to a conference
participant. They have been digitally
processed to reduce the amount of data which is transmitted by
a factor of more than 20. It will sound like this. [TYPING] -You are now a conference
participant. -Drew, thanks for calling. You’re the first participant. -Oh, great. Only Tim and Rick are here. But they are familiar
with the project. [BEEPING] -Oh, it looks like someone’s
trying to call in. -It looks like it’s Parker
from Salt Lake City. Let’s see what he has to say. -Hello, Randy? I’m anxious to hear what
Drew’s group has found. -Well, they’re on and
can hear you. All we need now is Danny. We haven’t heard from him. I don’t know where he is. -Oh, you know, it’s Thursday. I’ll bet he’s sailing. -Good afternoon. Beverly Hills Rolls Royce. You’re now connected with
the ISI, [INAUDIBLE] digital voice conferencing
system. -Here’s Danny now. -Randy, Sorry I’m late. Are the others still on? -Yes, they’re on and
can hear you. In fact, Parker would like
to verify your numbers. Do you have your
data with you? -No, I don’t have the exact
numbers with me. I could talk now in round
terms, if you’d like. -Danny, Drew seems to
be wanting to speak. I’d like to check back
with Parker first. And then I will give
Drew the floor. -Drew, Randy wants to
hear from Parker about what I just said. Then he’ll give you the floor. Sorry. Parker, I hope it’s finished. -Yes. We’re just typing it in. You can expect a message
from us this afternoon. Drew, you can have the floor. -Randy, I think it’s impossible
for us to build a program around general
numbers. We really need Danny’s exact
figures before we can continue this. I’d rather talk when
he has them. What does anyone else think? -Well, [INAUDIBLE]. Can we reconvene this call
tomorrow at the same time? OK, Parker, we really don’t
need you anyway. -Do you have all of your team
members on hand at that time? -Danny, just be on
time tomorrow. Fine. We can all make it. We should be able to get
this finished tomorrow. Now remember, Danny, you have
to be on an ARPANET phone so we secure the password. -This was a hypothetical
conference using features of two real systems developed
at the USC Information Sciences Institute. [END VIDEO PLAYBACK] BOB PARKER: So you may have
noticed in that film that Danny’s accent had changed
significantly. There is a short story
behind that. It turns out that Danny’s voice
drove the vocoder so crazy that the speech
was unintelligible. So we had to do a voiceover
for that. It’s also kind of said that’s
the only movie I’ve ever played a supporting role in. And when I looked on YouTube, it
had totaled about 500 hits. So I probably made a good
career choice, actually. So I’m Bob Parker. And I’m going to share this
talk slot with Randy Cole. Randy Cole is a colleague
from the ISI days. So we’re now switching into
the ISI period in the ’70s and ’80s. And he and I worked on several
projects together, and also worked for Danny. I think I’m going to
trust me with this. So Randy was not able to
make the meeting today. So he wanted to me to add a
few comments in for him. So we collaborated on the second
half of this talk, which is really the
lessons learned. It was sort of a theme that
seems to be going on here as we got some lessons learned. So the last half of this is
lessons that Randy and I remember from those
early days. So I’m an electrical engineer
by training. And so my interest is in the
hardware of the ARPANET. And so that’s all I’ll
be focusing on talking to you today. So we have to go back
to the early ’70s. I was also a founding member
of ISI, started in March to July of ’72. We came from the Rand
Corporation to start this whole activity at ISI. And ISI has really stood up to
push that vision that DARPA had for the ARPANET. And we had one customer. And we had one mission. It was pretty simple. And we had one contract. So that happened [INAUDIBLE] for quite a few years
that way. So I want to take you back now
to the early ’70s, and representative of a couple
things that I was working on. That terminal up there in the
upper right-hand corner– it’s interesting what
this has done to the background of the slides. But anyway, portable terminals,
I think they call them laptops or something
these days. I don’t really know. It’s been a while. So in 1973, we were billing
these things. And there’s some DARPA
alum here in the room, some follow alum. And some of you were there
at DARPA at that time. Probably carried one of these
things around with you in ’73. The other thing– I’m going to move out
of my zone here. This is a network attached
printer, is what they call it these days. That thing, Xerox Graphics
Printer, was about the size of a washing machine and the dryer
stood on top of each other, about that big. It had that dedicated PDP
11 in the background. It had some of my interfaces in
it connected as a host to the ARPANET. And so of course, the whole idea
was that you could begin to manage documents on the
ARPANET, send them around, remotely print them off. And then that was ’74. By ’81, it was a Rapacom 450,
which was a fax machine. Fax machines were not on
ships in those days. In fact, that thing was
about the size of a washing machine as well. And we had a project called
Faxie that Steve and I worked on at that time. We connected that thing
to the ARPANET. So if you think the ARPANET of
things was only invented in the last couple years, we were
putting a few things on the ARPANET a long time ago. So the whole point of that one
was that we were beginning to manage faxes not as point to
point scans from one machine to another, but as documents
that you could put on the ARPANET, edit and forward
and so forth. So then the natural thing next
to do was to work on voice. And so I have to stop for a
minute and back up a little bit and talk about when
I first met Danny. So I was a pretty young, impressionable engineer at ISI. I’d been there a couple years
when Danny showed up. I remember distinctly– somebody’s actually mentioned
this– the first few probably months that I remember him
being there, I couldn’t understand a word
that he said. I just had no idea at all. And we used to speak
phonetically. And so there was this alpha,
bravo, delta stuff going on. And well, interestingly enough,
that continued for the decade that I worked for him. So never got any better. But after I began to understand
some words, then I discovered that I really didn’t
understand what he was talking about. So it didn’t get much better. The lead into this
is an example of the day in the life. I was really Danny’s
toy builder. I was responsible for building
toys for Danny for more than a decade in the ’70s and ’80s. And I build lots of
different toys. And I remember one day, Danny
used to call us to his office. And we would have no idea what
the heck was going to happen. He was going to rant
about something. We didn’t really realize at that
point that a lot of those things were truly visionary. But when we were living it,
it wasn’t quite the same. And so we called to
Danny’s office. OK, sure, I’m an engineer. Let me do my engineer stuff. And leave me alone. So we go to his office,
sit down. And he starts ranting. So he’s got a computer
over here. He’s got a telephone
over there. And he says, I do all my
work on my computer. I can send email. I can communicate. I can compute. I can do all this stuff. I want to do everything here. Why is it that I have to get
up and go over here when I want to use my telephone,
and they’re not connected in any way? And so that led into several
years of hardware devices and protocol development and trying
to figure out how to integrate Danny’s phone
onto his computer. And it started off with these
little control boxes up in the upper left of the
picture here. And those things were control
boxes for a variety of different early voice
conferencing systems. So that was in ’75. So you see this spans about
six years or so of development activity. And the movie that you saw– sort of some snaps from it
down here at the bottom– that came in the middle of the
time between the ’75 and ’81. And so a lot of the things were
simulated in the movie. It wasn’t until the Wideband
voice came along, capability in the ARPANET, that they were
actually able to connect. That board in the upper right
is a board that I built that connected the Ma Bell telephone system with the internet. And so at that time, that was
the piece of hardware that enabled that connection. And so when Danny was stepping
into this odd shaped box– for those of you that are young,
that’s not the TARDIS from “Doctor Who.” But it
is a phone box. And you put coins in it
and make phone calls. So really what happened is, as
soon as we got that STNI card working, then we could
connect the world out to the last mile. We used Ma Bell for the last
mile of these conferences. And then we did everything else
over the ARPANET, which is, of course, the way
it all ended up. And that other person in the
movie on the right there is Randy Cole. So he gets in the
movie as well. So now I’m going to turn
to the lessons learned. We’re getting on in years. And so we have a lot of valuable
experience to offer to the next generation. So it’s very important to be
able to articulate these lessons so taht people will
benefit from them. And that’s kind of odd. These first three relate
to aircraft. It seems to be a theme of this
discussion, is that these things that relate to Danny, and
experiences that we had. So the first one, “when you’re
flying a Cessna below a certain speed with the landing
gear retracted, its best to tell your passengers that that
kaxon sound that they’re hearing doesn’t mean that
they’re about to die.” The second one– “the FAA recommends
that you verify that the oil filler cap is on
really tight before takeoff.” And it turns out that if you
don’t do that, then that oil’s really hard to get off
the wind screen. It’s much harder to see
out of the airplane. The third one– “when you’re
flying a Cessna at 40,000 feet, the control actuators,”
flaps and that kind of stuff, “may freeze.” And I don’t really
remember whether it was 40,000 feet. But I do remember that we were
well above the commercial jet aircraft traffic that was going
up and down between San Francisco and LA. And we were using oxygen. So now we come back to earth. It’s important to “check to see
if gale force Santa Ana winds are in the forecast before
you sail your small boat out of the Marina.” Do you
remember that ride, Danny? This is one. So these are sort of to trigger
your imagination. You kind of have to fill in
the details on your own. The second one is one
of my favorites. It’s “being Danny’s secretary
can be fun. And it’s even more fun if you’re
stoned.” And there is a story there. The third one– “never hang your
dry cleaning from your handlebars of your bicycle. Flying is best done
in an airplane. And a three point landing
does not mean both hands and your face. You can compress speech as much
as you want, if you’re not too picky about
the quality. Don’t buy an SPS-41.” It’s
a signal processor in the early days. And some of you probably had the
pleasure of working with that thing. And the third one, this is
typical Danny. “The speed limit on Lincoln Boulevard
is not 80 miles an hour.” Number three– “sometimes it’s better to beg
forgiveness than to ask permission.” That seemed to
serve Danny pretty well throughout his career. This is sort of an ISI one.
“When you really want something, always ask Keith. Tom is only authorized
to say no.” So the Keith here is the Keith
Uncapher that was mentioned earlier, the original
founder of ISI. Tom Ellis was the
deputy director. And it’s absolutely right. His job function is clear. Third one– “it’s a good idea to
turn the shower off before you leave for Israel.” And
lastly, “when the Union 76 station catches fire with your
car inside, it’s good to know someone like Danny, who will
lend you his old car, even if it is a Pinto wagon.” That was
an experience that Randy had. Most importantly, “always look
for someone like Danny to work with.” If we had to live it all
over again, Randy and I would not change a thing. And thanks for the wild ride. And Danny’s always been very
involved with the families of several of us here. And so my wife and I– certainly best wishes
in your retirement. And [? Carolyn ?] Cole also had a nickname that
Danny had given her. It’s Randy and [? Carolyn ?]
“Crew Cut” Cole. She got her hair cut
at one time. And so she got a lifelong. Thank you. DANNY COHEN: I’d like
to shake his hand. NEIL GERSHENFELD: Larry
Miller next. VINT CERF: No, he wants
to shake hands. DANNY COHEN: Thank you. LARRY MILLER: So I can’t top
that in terms of jokes. And Danny gave me very specific instructions today to be funny. But I’m not going to be. And this is a little bit
of an interlude, and a very personal one. And like, I think, almost all
of you here, I was a grad student at ISI when
I first met Danny. And what we had in common from
day one was certainly not my intellectual ability. But it was the fact that
I was a pilot. And no matter what stories any
of you have about Danny’s flying, I can top them. [LAUGHTER] LARRY MILLER: The time
we landed at Van Isle with a flat tire. The time Danny buzzed El
Mirage in a glider. So I say we could
go on and on. But Danny, for me,
was not a mentor. Danny was a champion. Danny promoted me in ways that,
frankly, I don’t really understand why. Because I don’t think I
was deserving of it. But from day one, Danny really
worked with me to make my life a lot better– my time at UCLA and back
at ISI, time off to write some books. And we’ve had a long, long
period of friendship and engagement and family
and friends. So I’m tremendously appreciative
of everything they Danny has done for me. One thing I do I want to add– and I’m going to make this very
quick, maybe make up for some of the others. Why am I wearing a tie today? VINT CERF: Yeah, why are you
wearing a tie today? LARRY MILLER: Except for Vint,
who I don’t think has ever been without a tie. I’m the only one here
with a neck tie. And the instructions were
keep it casual. And the reason I’m wearing a
tie today is because when I was at USC and went off to
defend my dissertation with my committee, the one word of
advice Danny had before I left was wear a tie. And so I have tremendous
respect for Danny and everything that he’s
done for me. That’s why I’m wearing
a tie today. Thank you, Danny. [APPLAUSE] DANNY COHEN: Thank you. STEPHEN CASNER: So I met Danny
in fall of 1973 when we both joined ISI. I was just beginning USC
as a grad student. So I was there as a research
assistant. And I’ll get to that
slide in a moment. The first thing that I did as
part of the project when I joined, I got assigned to
work with Danny on the packet voice project. And the first thing I did was
work with Bob Parker, actually, on the ARPANET
interface, which you saw blinking at the very beginning
of the movie there. Right now, of course, if you get
a computer and get on the net, it’s part of the
way it’s built. But at that time, you had to
build the interface that you were going to get on
the network with. So we had to build the interface
to the computer, build the software in the
computer, and build the compression device. We were able to put all
of that together. And the timeline of those
developments you can see here. The first communication with
packet voice across the network that we did was
in August of ’74. That was using CVSD. The blue boxes that you
saw in the movie were CVSD encoder boxes. But I think this was actually
running CVSD implemented in the SPS-41. And then, because we had not
yet implemented the linear predictive coating to give a
better compression rate for the audio to run across
the network. The second bullet here does not
involve ISI and Danny and I individually, but the folks
at Lincoln Laboratory and [INAUDIBLE] Harrison in Santa Barbara. But the point in time when it
got really interesting for me was in ’76 when we actually
had multiple sites working together. And the protocols were really
having to do their job from the control side as well
as the data side. As I may have mentioned, this
is Danny’s slide from review that we had in 1982. And the last two points are
one which has been a particular interest to him
and to me as well. After we did all these things,
then in 1977, James Flannegan from Bell Telephone Laboratory
filed a patent on the notion of doing packet voice. VINT CERF: Finnegan. LARRY MILLER: Indeed, indeed. And that patent was granted
in 15 months. So these days, you can’t even
get your patent viewed by the examiner in 15 months. So there’s two parts of this
problem, at least two– but two significant parts
of the problem. And the first one is doing
the compression. And I couldn’t find a picture
of an SPS-41. But it’s similar in size, at
least, to this picture of an FPS AP-120B, the white
box in the top half of the PDP 11 rack. So at that time, in order to
compress voice you needed half of a six-foot rack. Instead now, we have these
things that do a much better job. Thousands of time capability
fits in your pocket. So an aspect that I remember of
doing the LPC on the SPS-41 is a piece of work
that Danny did. And I’m sorry, Danny, I don’t
remember exactly which part of the algorithm. This is a diagram of the
analysis side, that is the compression side as opposed to
the decompression side, of the LPC algorithm. And so it has a number
of calculation steps that go there. But I remember Danny was
particularly proud of achieving implementation of an
inner loop for one part of this algorithm getting it into
six instructions where every component of the multi-processor
SPS-41 was working in every instruction. So that’s how you’ve gotten
it to go as fast as it possibly can. The second part was the network
voice protocols, which have been referenced earlier. This is, again, Danny’s cover
slide from that same 1982 presentation. The network voice protocol that
bears Danny’s name as the author for both the first and
second versions was relatively simple compared to many of the
protocols we have nowadays. But it did have a number of
important components in it. We were working out
things that hadn’t been figured out before. And it also has direct
lineage to RTP– the real time transport
protocol– which is one of the fundamental
protocols being used for voiceover IP. We didn’t have IP when NVP was
first done, but we did by the time RTP came around. And I’ve noted here on the
bottom, one of the key concepts was that in order
to properly do the synchronization of the voice
data, we needed to have both a sequence number and a time stamp
so that we can represent periods of time when there
wasn’t any data. So these are, as I say, the
first two components of the packet voice problem. Like others, I want to shift now
to really talk more about what it was like to work with
Danny, advice that we got from Danny and such. And we’re back to this same
picture that you’ve seen a couple of times already. We didn’t all coordinate
as to who was going to use which pictures. But I want to take a little tour
first of Danny’s office. So here, he explained to me that
to the visual acuity is supposed to be better
when looking down. And that’s why the terminal
was mounted that way. But it looks to me, Danny,
like your neck would get really sore after a
period of time. But I suppose not. So if from this position,
looking at, that would have been the north wall
of the office. If we rotate around,
we see the wall going out to the Marina. And on a pillar of that wall is
a piece of paper with this message on it– one of the saying that
we should remember in practicing our work. “Indecision is the key
to flexibility.” Then if we turn around again
to the next wall, Danny had his framed membership
in the Flat Earth Society posted there. You still have that? AUDIENCE: Denied membership. STEPHEN CASNER: I’m sorry. AUDIENCE: It was denied
membership. AUDIENCE: He was accepted. And then he was thrown
out very dramatically a few months later. We have both letters together
on the wall. STEPHEN CASNER: Good. And then back to the desk again,
Bob mentioned Danny’s interest in hooking the phone
and the computer together. So there’s a couple of
aspects of that. The STNI card is a way to hook
up to a phone line to bring voice data in and out from the
circuit switch network, the telephone network, into
the packet network. But just as interesting
was work on controlling the telephone. And so what Danny had done was
he managed to get a hold of a Motorola 68,000 development
board and got Bob Parker to mount it on a box with
a power supply. And that’s what’s
in green here. And as was mentioned, the
speakerphone was kind of tucked underneath the
edge of the desk. What this would do, it was in
series with the character stream going from the terminal
to the computer. And that allowed receiving an
email that had a phone number in it that you wanted
to call back. You could select that
phone number. And it would come down and cause
the board to dial out on the telephone and the
speakerphone. For me, that was an interesting
addition, and led to one project that I did some
years later at ISI to reverse engineer the way a roam digital
phone worked, so then I could hook a computer
up to it. And I built a server for all
of ISI to use that provided the same facilities for any
of their digital phones. Danny taught me a number of
things, one of the most important was emphasizing
content over form. And an aspect of that was not
using fancy word processing programs when you’re trying
to get an idea down. Get the words down on paper. Get the notion there. And I follow that to this day. I eschew using Microsoft Word or
anything else, I prefer to write things out in text. An aspect of writing in text is
you still need to have some reasonable shape to the
paragraphs and such. And so Danny had made a program
called pretty, or pre-tty, as in a sense of
preparing the text for appropriate display. And I believe that was
implemented as TECO macros. Is that correct? DANNY COHEN: [INAUDIBLE]. STEPHEN CASNER: I think so. Danny was also quite
proficient in TECO. He and I both graduated to Emacs
in the years since then. But both of those are skills
that are held by, well, a fraction of the population,
but not a large fraction. But in addition to all of
these computing related aspects of his advice,
there’s another one that I always remember. And that is “to wear socks
with your sandals when visiting the East Coast
in winter.” Thank you. [APPLAUSE] BOB BRADEN: I’m Bob Braden,
another denizen of ISI. I fortunately never flew
with Danny, never went sailing with Danny. I guess I escaped, luckily. So I think the breadth of people
in this room speaks to the breadth of Danny’s
accomplishments. Danny, I often noticed that you
seemed to come in with new ideas every day. And some of these ideas stuck. Some faded. Some deserved to fade. But you continuously
generated them. I had the good fortune to be
taught computer science by Al Perlis at Carnegie Tech. He was the same kind of guy. Every day, he couldn’t make
progress on a project because every day he’d come in with
a different idea. And that’s not your problem. You’ve somehow managed
to get things done. I wanted to echo what Larry
started off with. One of Danny’s great
contributions to computing was the recognition of the
importance of the datagram. You’re Mr. Datagram
in my mind. I remember when DARPA, and then
you, wanted BBN to turn on the type 3 datagram packets
in the ARPANET. And VPN essentially refused. And I believe that Barry Liner,
who was the program manager then, put a great
deal of pressure on. I see Bob Kahn was the one. And later, on of course, Danny
was one of the instrumental people in separating
IP from TCP. And that was very,
very important. There’s an interesting article
by Alex McKenzie. And I guess it’s the recent
“Annals of the History of Computing.” This was
worth reading. It gives an alternate
view a history of OSI versus the internet. And one wonders if Alex’s
prognostication had happened and OSI protocols became
real whether ever would have had datagrams. So just a quick comment
about Danny as a satirist, as a humorist. Before we came over here, I
was explaining to my wife about the Little-Endian
and Big-Endian. And she said, was he serious? So Danny, I want to pose
that question to you. Were you serious? DANNY COHEN: Just
think about it. BOB BRADEN: So you and John
were close friends. You worked together. You always seemed to have the
same opinions about the architectural issues. And I remember when the network
working group, or internet working group, or
whatever it was, was trying to decide what an IP address
should look like. You and John were in the back of
the room, very vociferously objecting to the imposition of
fixed length IP addresses. You believed in variable
length IP addresses. Now one can speculate whether
Vint was right or not. I’m interested to ask
Vint whether he thinks now he was right. But you were very clear. You and John were very clear
that the flexibility of variable length addresses
was important. Through some weirdness in
the world, a warp in the space-time, you became
director of Division Seven for a while. And John was assistant
director. So none of us really believed
that you were [INAUDIBLE]. You were clearly the
intellectual force behind a lot of it. But as far as writing reports
and doing all the things that directors do, I’m sure
John did it all. Which leads me to a personal
note and a sad note. The phone rang one
day in 1989. And it was you. You said, John has died. I think it may not be true. But I like to believe
I was one of the first persons you called. And if it’s not true, don’t
disabuse me of that, because that’s very important to me. So thank you. [APPLAUSE] VINT CERF: Oh, I’m up. Well, I’m Vint Cerf. It looks like I’m on my side. There we go. And I’m standing between
you and a break. So I’ll try to be thoughtful
about that. First of all, Mems. It turns out that Steve
Crocker must have been infected by the Mem disease. Or maybe he was in the
smoke filled room. But he and I started a company
we called capital C Squared we bought a gross of Mem games. And then we proceeded
to try to sell them. And after a couple years, we
actually got rid of all of them, except for the few that
I have in my basement now. So I didn’t realize that you
were part of that whole story. But I remember trying to show
somebody how to play Mem, only to discover later that he
actually was colorblind, and none of it made any
sense at all. I can’t even read my
own handwriting. What was it I was– oh, yes. So one day Danny shows up
and says, let’s get together for lunch. So I meet him for lunch. And this is probably in the,
I would guess, early ’80s. And we sit down. And he hauls out this giant
gray thing with an antenna on it. It’s the brick from Motorola. It’s one of the first mobiles. And of course Danny
has that, right? So I started asking a whole
bunch of questions about how in the hell does it work,
and blah, blah, blah. And Danny says, why don’t we
call the guy that designed it? So sitting at the table with
everybody staring, the dials in the number. And I get to talk to Marty
Cooper to find out more about the details of that brick. So talk about Danny flying
around in airplanes, I never had the pleasure of flying in an
airplane with Danny, thank goodness, after listening to
some of the other stories. But I was 3,000 miles away when
Danny was testing his own version of packet radio in the
airplane, with packet-voice running on top of it. And they were flying
around in ISI. I don’t know whether you
were circling or not. But it sounded to me like
you were, based on the conversation we were having. So simultaneously while we were
talking, we have instant messaging going. And I figure, OK, so I can
get in on this, too. So the question is, what are
you seeing out there? And he’s telling me that there
are nuclear detonations out in the center of Los Angeles. And there we go through this
whole Dr. Strangelove. And I’m sorry to say that
I don’t have the transcript of that. If I did, it would
be a classic. Now several of you have already
gone through all the things that Danny has
been involved in. So he’s smart. He’s funny. And he’s got this incredible
scope. So like Bob and others, I’m not
going to try to go through all of them. But I do have to tell you the
story about packet-voice. By the time I got to DARPA, it
was DARPA at the time, 1976. And so we are well into some
of the experiments that you heard Steve and others
talk about. And by this time, the linear
predictive code mechanism had been implemented. So we had people involved in the
packet-voice experiment, not only in the United States,
but the Norwegian Defense Research establishment guys
were also part of this experimental stuff. And it turns out that when you
take 64 kilobit voice and you crank it down to 1,800 bits per
second, you really do lose a certain amount of quality. So what happens is basically
anyone talking through the system sounds like a
drunken Norwegian. I don’t even know if you
remember this, Bob. But I had an opportunity to
demonstrate the system to a bunch of guys over at
the Pentagon, some fairly senior people. And I remember thinking,
OK, how the hell am I going to do this? And then I had this brilliant
idea, because Yngvar Lundh was at the Norwegian Defense
Research establishment. So the way we set it up is at
first we demonstrated what it would sound like through
the regular voice network, the VON system. Then we had them talk through
the packet radio system. And it sounded exactly
the same. What we didn’t tell
them is everybody would sound that way. To come back to the TCP and IP
split, it turns out there were three people who lobbied in
favor of splitting off this internet protocol so it could
be fast, and not necessarily reliable but be very,
very responsive. It was Danny. It was Jon Postel. And it was David Reed. I don’t know if you remember,
David was at MIT at the time. And the three them
were adamant that we had to do this. And Danny, in his just
inimitable way, underscored this with a Finnegan story that
involved milk and wine. And then the whole point was
that it was OK for wine to take a long time, because
it got better. But the milk didn’t. And so we needed faster
delivery for the milk. And so I’ve always
remembered that. If anybody ever asked me what
was the justification, I said Danny told me that the
milk would spoil if we didn’t do IP. Another thing, this variable
length addressing that Bob Braden brought up, I remember
this very well. When Bob and I started the
design of IP, we really didn’t know how many terminations
we would need. After all, it was
an experiment. And we weren’t even sure
it would work. So I remember going through
the following loop. First of all, the question was
how many networks will there be per country? And we had just gotten
the ARPANET running. And it was not exactly
inexpensive. So I think we concluded that
you wouldn’t need more than two nationwide networks. And so there’d be some
competition. So that established
that parameter. Then we said, how many
countries are there? We didn’t know the
answer to that. And we didn’t have any Google
to use to look it up. So we guessed 128, because
that was a power of 2. So that’s 256 terminations. That’s eight bits. And we thought, now what
about the computers? And we thought, well, computers
at that time– this is 1973– mostly were big machines that
served anywhere from 1,000 to 10,000 people, large scale
time sharing machines. And they didn’t exactly
get up and run around. They had to sit in air
conditioned rooms. So we thought, well, trying to
be futuristic about it, we said, how about 16 million
machines per country, or per network. I’m sorry. So that’s 24-bit. So that got us to 32. That was 4.3 billion
terminations. So we thought, surely that’s
enough for an experiment. Then a few years later, this
debate came along again, just as we’re getting very close to
finalizing TCP/IP stuff. I didn’t remember that Jon was
in favor of variable length, to be frank with you, Bob. But thank you for bringing
that up. But Danny I remember vocally
arguing this has to be variable length. We have to be prepared for
bazillions in networks. In particular, he was saying
that local area networks were going to be a big thing. This is 1977. And so I’m listening. And the programmers are all
levitating out of their chairs, because they’re
saying, wait a minute. You’ll never be able to find
any of the fields in the packet because you have to go
parse things because of the variable length in
the addresses. And they were just raving. So we said, OK, that’s
really a big problem. So we’ll stick with
fixed length. Then the question was should
it be 32 or 128. And I remember thinking, 128
is 3.4 times 10 to the 38th addresses, a number only the
Congress can appreciate. And I thought, it didn’t pass
the red face test because most of the data that was passing
around in the network was one byte of data, and this
giant header, because everything was Echoplex. And so I thought, well,
128 bits or something. It was overkill. So we stuck with 32. And as luck would have it,
it actually lasted almost 30 some odd years. And it was only in February
of 2011 we ran out. So we turned on the IPv6
128 bit network on June 6 of last year. But Danny was probably right. If we had had better crystal
ball gazing, we would have done variable length. But we didn’t. Now some of you have heard about
the Strategic Defense Initiative. I didn’t have much involvement
in that personally. I remember speculating about
whether this was even feasible or not. But Danny got me involved
in a very different way. In his inimitable way, anything
that’s important has to be done in minus 2 seconds. A phone rings. It’s 6 o’clock at night. And Danny says, you have
to come over to my hotel right away. I said why? And he says, because I can’t
figure out how to put this black tie on. He says, you wear them
all the time. So I came over. And the tux you saw, I think,
was probably in Geneva for the Internet Hall of Fame. But I came over. And I think I put a black tie
and jacket on Danny, an outfit that he had never, ever worn
in his life, and probably never wore after that until
March of last year. So I hope you appreciated
my responsive service as your valet. And finally, Danny’s wonderful
style of writing the Oceanview Tales from Professor Finnegan. But I remember at least one line
somewhere in one of them where Danny is explaining how
important inventions are. And this particular invention
was the invention of oxygen. He said Lavoisier invented
oxygen and it made it a lot easier after that for everybody
to breathe. And so Danny, I hope you keep on
inventing things for all of us, because it’ll make
it a lot easier for all of us to breathe. Thank you for everything
you’ve done, Danny. AUDIENCE: Thank you. IVAN SUTHERLAND: I’d like to
take 30 seconds on the subject of Big-Endian and
Little-Endian. If you read the Big-Endian,
Little-Endian paper carefully, you will find some magnificently
crafted footnotes describing the
references to Jonathan Swift. And people on this side of the
Atlantic don’t appreciate the original source of the
Big-Endian, Little-Endian controversy. I face the Big-Endian,
Little-Endian controversy at breakfast with my wife. And here is the reason. She’s from the Netherlands. I was brought up in a
very British family. And we both like to eat
eggs out of the egg shell in and egg cup. And Jonathan Swift wrote a note
about this, from which Danny quotes and used as his
reference for the Big-Endian, Little-Endian controversy. I remember when I was a boy, my
father would whack the top off an egg and set it
in the egg cup. And then we would
spoon it out. And my wife and I both
like to do that. But she always puts her
egg in upside down. So the original question is, is
the Big end of the egg or the little end of
the egg the top? [APPLAUSE] DEBORAH ESTRIN: So first I have
to make a confession to those of you in the room. So I first knew of Danny– this is something I’ve
never spoken to him about face to face. I knew of Danny from my grad
school days at MIT, because I was one of those uninformed,
not very intelligent progressive folks who were
protesting against what you guys were doing with
Star Wars. I was involved with CPSR
in its early days. And Danny at that time
was on the dark side. And after I met Danny at ISI,
I never had the guts to take it up with him. It was only many years later
after we became friends I used to revel in hearing David, his
son, argue politics with him, because really, he was the only
person who had the guts and the intelligence and the
conviction to argue politics with Danny and win. So it’s only in David’s presence
am I willing to acknowledge. If I could have been able to
find it, I would have worn my shirt with that mushroom
cloud on it that said, “it’s 11:00 PM. Do what your expert system just inferred?” I have it someplace. Well, there’s an ironic end to
the story which is appropriate to you as well in several ways,
which is that many years later, and just a few months
ago, I was sitting in Tel Aviv under Iron Dome, the
great-grandson of Star Wars. And all I could think of
was the irony of life. A couple of other
small remarks– many people have mentioned
the network voice and the separation of TCP and IP. And every year when I teach
networking, that’s always part of my introductory lecture,
talking about that early work and that really– AUDIENCE: Will we be
tested on this? DEBORAH ESTRIN: Excuse me. AUDIENCE: Are we going
to be tested? DEBORAH ESTRIN: That’s right. That’s right. And the essential forcing
function that that was and that Danny was. And I really love it as an early
and really important example of how that concrete
application as driver is such an important process of our
discipline, and try to make a point to the students who don’t
necessarily appreciate it at that time, at that first
lecture in networking. But it’s always important
to me. I also just want to make a
remark of gratitude, because I know that it was actually
Danny who got me to ISI. Because Jon told me at one point
that Danny had told Jon, you know, you should invite
Debbie Estrin to come over to ISI. And I then spent a lot of really
wonderful years there. I was an assistant and then
associate professor, then a full professor at
USC at the time. And I think it’s the thing
that kept me from leaving academia, was being at ISI,
because you guys were all there building things. And it provided me with that
model for the rest of my career that I’ve just sought
after, which is to try to find a place to do innovative work
that could have real impact by really building things. And so when I just moved from
UCLA to Cornell Tech, many people have asked me, how many
graduate students did you bring with you? And while I love my graduate
students, I didn’t bring any one of them. But I did bring my three
software developers. So there are many other
stories that I’m not going into. They all, or most of them, have
to do with time spent with then Cody and my son Josh,
who became very good friends after they met at an ISI
holiday party at the age of three, close to three. AUDIENCE: If that. DEBORAH ESTRIN: Yes, if that. And they’re wonderful stories,
trips to Italy, treks in Topanga, many dinners in Pacific
Palisades where we both lived. And interestingly enough, as
you’ve heard in every talk here, or most of them, all of
those voyages had lots of humor, and usually had something
to do with gelato. AUDIENCE: The green spoon. DEBORAH ESTRIN: The
green spoon. Danny took a group
of a dozen– how old were you guys? Eight. AUDIENCE: Something like that. DEBORAH ESTRIN: For Cody’s
eighth birthday, trekking up in Topanga. Cody didn’t have the most
coordinated group of friends, my son among them, trekking
up on these treacherous hikes in Topanga. When we all get there, finally,
and no one has fallen off yet, out comes the gelato
from his backpack. AUDIENCE: At the top
of the rock. DEBORAH ESTRIN: Yes. Anyway, Danny, wonderful years
of friendship, for me and for Joshua, who had to
leave early. Thank you. DANNY COHEN: Thank you. DEBORAH ESTRIN: I get a hug. [APPLAUSE] DANNY COHEN: [INAUDIBLE]. VINT CERF: Patrice Lyons,
Bob Kahn’s wife. PATRICE LYONS: Thank
you, Vint. Greetings, Danny. And I’m so pleased to be here. I met you, you may recall, back
in the mid ’70s when I met this gentleman, Bob Kahn. And we had many wonderful
talks. I remember you’d come to
visit us in our home. And we have lots of good
times together. At the time, I was an attorney
in the Library of Congress. And I had been tasked to work on
what had been defined as an issue having to do with
protection of semiconductor chips. And I had learned about the
MOSIS project from Bob, chapter and verse, and
your role in it, and ISI’s role in it. And he gave me the
Mead/Conway book. I looked at that. But then I felt I needed more. I needed to get my fingers
in it to see what this was all about. And so I’m a lawyer. So here I went out. And Danny was so kind. You took me by the hand. And you brought me
around to the different places, the foundry. You gave me great help in all
that, and followed up eventually to help me, too, to
understand how to translate that into words. And I remember we did one draft
which was going to look ahead to try to speculate for
years to come, so that it would last more than
just a few years. Unfortunately, they did
something that was addressed just the moment. And so it didn’t
last that long. Well, it’s still in effect. But it’s not as effective
as it might have been. That was my point. So then Bob and I get
married in 1980. And maybe this is too much
information, excuse me. But he did us a great honor. I noticed he had the tux on. And you had a new pair
of jeans you wore. I was very pleased, you see. That was a big honor. And then I got to meet
lovely Delia. And I’ve heard a lot of the
stories about the plane and the motorcycle. There’s no ski stories. Now [INAUDIBLE] Utah there in. And somehow they had their
meetings in February in Salt Lake City. And then we went up
skiing afterwards. And I remember Danny has a
certain way of skiing. Delia, you know about his. Most people go like this. Danny went, whoosh. So more Danny stories. Then I lived through
various different episodes, the SDI saga. And that had its moments. Today oftentimes young
people think they invented sliced bread. And they talk about VoIP. Oh, goodness. And I tell them, well, there
had been voice in the early internet days. No. I say, yes. So I’m really pleased that
this is honoring your work back then. And they had the early program
in the year on the voice, which was really good. Now last year, I was in Geneva
representing Bob at the Internet Hall of Fame. And Cody, I hadn’t seen him
in a number of years. And it was like, oh my goodness,
you’ve really grown. And Danny was honored as
an internet pioneer. And I was really pleased to be
there to see you having proper honor for your work. So thank you, Danny. DANNY COHEN: Thank you. [APPLAUSE] DANNY COHEN: Thank you. EVE SCHOOLER: Couple of comments
before my slides. I have to thank all of you,
because my face hurts so much from smiling. This is so enjoyable. So I’m Eve Schooler. And I first met Danny when
I joined the high speed networking division of
ISI in February 1988. My first slide, please. When the internet had, at this
point, 30,000 hosts, FTP was the main tool for remote
document sharing. The IETF was on the
brink of meeting number nine in San Diego. And the numbering of the RFC
document series had recently crested 1,000. Long time ago. It’s at almost 7,000 now. It seems quite fitting to– I thought I was going to be
following a speaker who was coming to us via video, because
by entree into the world of Danny Cohen and ISI
and DIV7 was via the ISI multimedia conferencing
project. I was excited to join this
project because, of course, it gave me the opportunity to work
with really smart people and to do cutting edge research
on the real time internet architecture, and to
dabble a bit music under the guise of showcasing
state of the art packet audio and video. One of my earliest recollections
of Danny, of interacting professionally
with him, occurred on a Saturday morning when I dropped
into the office to put the finishing touches on some
code needed for a Monday demonstration to sponsors
at DARPA. And when I ran into Danny in
the hallway outside of his corner office, he immediately
got to the point, with his Israeli directness. And he asked me if I could
help him debug his program for his demo. So what’s a recent college
graduate student to do but to say yes to his boss’s boss? It sounds like something’s
working now with audio. So I found myself helping
Danny with his own coding issues. But I don’t remember ultimately
if I resolved the problem or made it worse. And it must not have gone too
badly, because after that, Danny was much nicer to me. Up until that point, I had
wondered if Danny knew I was in his organization. During the time that I was at
ISI, the research playground for packet audio and video
transitioned from the satellite wideband network,
DARPA’s high speed descendant of the ARPANET with a remarkable
data rate of 3 megabits per second, to the
Terrestrial Wideband Network. And I’ve always loved this
photo, because it makes me nostalgic for that beautiful
seminar room at ISI. And I’ve always wondered, who
are those crazy people silhouetted in this photo
atop the roof, Danny? Meanwhile, the real time data
that was once carried by the stream protocol through a BBN
butterfly eventually was taught to behave over IP
networks and was routed over the DARTnet, where sun
workstations began to replace specialized hardware, and for
the first time supported integrated microphone, speakers,
and cameras. Although I feel honored to share
a technical lineage with Danny and with Steve, I find it
hard not to think of them when I encounter a video
conferencing system, whether in a fancy Cisco TelePresence
room or via software running on one’s device of choice. What I value most from that era
goes beyond the technical details of the internet’s first
radio stations, its multi-site teleconferences,
or even distributed music performances. Along the way, I enjoyed
discovering that Danny had a mischievous sense of humor,
and discovering that he is well read, that he is interested
in all things clever and elegant and fast, all
the things that were fast. And in fact, that bottom
picture is of a Myricom switch. And of course, that he doesn’t
suffer fools lightly. I confess that I have struggled
to understand Danny fully, and I mean that
literally, in part because his creative ideas tumble out at
breakneck speeds, and in part because I simply wasn’t capable
of keeping up with his mental prowess, and in part
because of Danny’s signature Israeli accent. He and I recently joked that,
after all these years, I can finally understand him a little
better, an odd but nonetheless silver lining
to his illness. It wasn’t until Danny left ISI
to help start Myricom that I really came to know him,
especially outside of work where he was no longer in my
managerial food chain, and where he worked very closely
with my husband, Bob, along with others at Myricom
who would become like extended family. There I discovered that Danny
was very much devoted to his friends and his family
and to being the progressive single dad. His humor was in rare form
from Myricom holiday gatherings. And he was like a little kid,
giddy to share newly discovered recipes that
needed taste testers. He gave us the sage advice to
document watching our kids grow up in the pre-digital
camera era. And he encouraged us to purchase
a video camera and videotape our eldest son, Sean,
who is now 16, believe it or not, on each of his
monthly birthdays. And indeed, during those early
years of Sean’s life, we were dedicated to this ritual of
pulling out the video camera on the seventh day
of every month. Later we came to be the happy
recipients of toys and books and train sets from
David, AKA Cody. And what was perhaps most
touching to discover about this tremendously cerebral
person named Danny Cohen was the pride that he took in
David’s creativity, sharing David’s home grown puzzles,
games, and most notably, David’s stories that we read
at Passover gatherings. Most recently, I’ve come to
appreciate that special window of time at ISI even more, given
the pivotal role that Danny’s ideas played in making
the internet hospitable to real time traffic at a time that
predates the arrival of the worldwide web. It wasn’t until Danny and Steve
allowed me to hang out with them as they put together
their keynote on the history of voiceover IP that I learned
from Danny that the separation of TCP and IP was influenced
as much by real time considerations as it was by the design elegance of layering. There were other historical
nuggets that surfaced, that were debated. And all were interesting
and enlightening. But sadly, many I have already
forgotten, which underscores, Danny, why you really need to
capture your recollections in writing and archive them
for safekeeping. And I know I said
this last year. But let’s make that
a goal for 2013. So in summary, thank you, Danny,
for sharing your clever wit, your rare intellect,
and your wisdom. I have always treasured your
creativity, your sense of humor, and your courage
to be yourself. And most of all, thank you for
entrusting me with your friendship. [APPLAUSE] DANNY COHEN: Thank you. That was beautiful. VINT CERF: Go ahead, Paul. PAUL LOSLEBEN: Neil assures me
that the downlink is working. So if for any reason, I stop
seeing you there, I’m going to just forge on. I’m coming to you from a small
island in Puget Sound, about 50 miles northwest of Seattle. Now since Neil feels strongly
that the downlink is working, I think that that confirms a
suspicion that packets flow downhill much better than
they flow up hill. So moving right along, where
to start about Danny? Danny is one of my
favorite people. And I am so happy to be able
to participate in this. Going back to the beginning is
probably the best thing. I met Danny probably in 1979. It might have been 1980,
memory fades. It was at a VLSI meeting that
Bob Kahn invited me do attend. I was working at NSA
at the time. And I had heard about the
VLSI community and what was going on. I was pretty excited
about getting an invitation to attend. And when I showed up, I wanted
to make sure that I captured everything. So I sat down and set up a
little pocket recorder so that I wouldn’t miss a word. Danny noticed that right away. And he walked over and explained
to me that somebody coming from NSA and hoping to
have a future with this community, I probably
wasn’t starting off on the right foot. I determined at that point that
this was someone who I should pay attention to. Now remember the time, this was
1979 or 1980, within the Department of Defense,
the [INAUDIBLE] program was starting. And it was sopping up every bit
of money that had anything to do with semiconductors. At the same time, the Japanese
Fifth Generation program was getting started. And we were concerned that
research in computer architecture was becoming
installed in the United States. Now fortunately, people like Bob
Kahn and Duane Adams and a lot of the people in this
room saw something that could be done. And we saw the beginnings of
that in what came to be called the MPC adventures, the work at
Caltech, work that was then carried to MIT. Chuck mentioned some
of that earlier. Now you need to understand,
within the Department of Defense, we felt that we needed
to do was to make sure that there would be a source of
semiconductor manufacturing available to researchers so
that we could do more than just talk about ideas, new
concepts, that we could, in fact, start building some things
and learning from that experience. Now what we proposed to do,
and what the community was encouraging us to do, was to
provide free access to semiconductor manufacturing,
to any researcher working in the field. Now you can imagine how that
would have gone over in the Department of Defense. We needed to start an
unusual program. And I’m happy to say that Danny
was in a position to take charge and run what became
the MOSIS service, the MOS fabrication service. That was something that only
an unusual person could do. And I’ll tell you right now,
we were darn lucky that Senator Proxmire didn’t give
us a Golden Fleece Award [INAUDIBLE]. Danny I didn’t always
see eye to eye. And I gather that some of the
rest of you had that same experience with Danny. But I learned early on that
yelling at him over the phone didn’t help a bit. As the VLSI program matured,
MOSIS grew like topsy. It expanded. It seemed like every run that we
put out had more people who wanted to get on it. And instead of just something
that was a haphazard thing, that maybe we would get and
maybe not get, it became like a well-scheduled airline,
that those runs came out like clockwork. They didn’t always turn
around as fast as we wanted them to do. In fact, the slogan at MOSIS
there for a while was fast turnover, no matter
how long it takes. So Danny was the right guy. And we started expanding. In particular, despite all
the talk about systems [INAUDIBLE], we learned if
we’re going to build new computer architectures and, in
fact, all sorts of things, that we needed to be able to
build systems that contain standard components and printed
circuit boards. And Danny borrowed an idea that
he had originally from Ivan Sutherland, something
called electronic commerce. Could we buy components over the
internet and provide them to researchers? And yes, we could. And then there came the problem
of, OK, we’ve got a bunch of researchers out there
who can design chips. But how in the world do we get
them to the point where they can design printed
circuit boards? Well, Danny had this cockamamie
idea that we should use the same interface, that
they could submit [INAUDIBLE] And we could plot it on the
[INAUDIBLE] machine, plot a bunch of them on the same
radical and use appropriate magnification to actually
print the circuits. Well, that was kind of a
crazy idea, I thought. But Danny persevered. He dragged me to this trade
conference in Anaheim. You know, the kind of conference
where you have a lot of people who are selling
components and selling printed circuit board services
and so on. And we went from booth to booth,
Danny and I. And Danny explained his idea, or attempted
to explain his idea to the sales people
in this booth. I was so embarrassed. He dragged me from
booth to booth. But you know what? He found someone who understood
optics and lenses, and knew that lenses work
in both directions. And that began a new service. It worked. Now Danny also introduced me to
a Professor Finnegan, who I understand is going to
be talking later– Professor Finnegan from
Oceanview University, Oceanview, Kansas. One of my prized possessions, in
fact, that I still hang on to is a collection
of works by the esteemed Professor Finnegan. But Danny was distressed that
Professor Finnegan didn’t have the appropriate identification
to get into meetings where everybody showed up
with a badge– you know, the kinds of badges
that they wear in the industry and the defense industry
and the government. But Danny set out to make
Professor Finnegan a badge, a full-fledged, laminated plastic
badge with a very distinguished looking picture
of Professor Finnegan on it. And it had various pins
in it that illustrated accomplishments in that
particular organization. And [INAUDIBLE] an American
flag pin in it. But you know, the thing that
really made it look official was a bar code that Danny
clipped out of “Motorcycle Racing” magazine and included
on the badge. Oh, we’re getting a
look at the badge. I was going to ask if
you all had it. It’s so great. There it is. More seriously, Chuck had
mentioned earlier that Danny had led a research effort which
attempted to evaluate issues and feasibility
of the software portions of Star Wars– I’m sorry, the space
defense initiative. I was not there. And I didn’t hear it. But it’s been reported to me
that Danny made a very penetrating observation. Danny looked at computer
science and made the observation that computer
scientists tend to build finally engineered systems,
where every part of those system is designed to work
precisely with every other part, that there’s nothing
extra included in there. It’s kind of like building
a fine watch. Now if you want to build a
weapons system, I don’t think you want to build it
like a fine watch. And Danny continued with the
observation that it’s difficult to anticipate the
grain of sand that might screw up the whole thing. In fact, he said that we should
consider the discipline of the communication engineers
who acknowledge the failures that are sure to occur and build
systems that not only compensate for errors but are
built on a science around maximizing the performance and
the preservation in the presence of errors. That’s a lesson we can
all learn from. Thank you, Danny. It’s been a pleasure
knowing you. I wish you all the best. And I thank you for all the
contributions that you’ve made to our [INAUDIBLE]. [APPLAUSE] BOB FELDERMAN: I met Danny,
actually, in about 1991. I was a grad student
with Len Kleinrock. And as you heard,
Len knew Danny. And he suggested, as I was
looking for various jobs, that I should go talk to his buddy
Danny who was at ISI, because he was always doing interesting
things. Now that might have been a
little bit unnerving for me. But since my wife was already
working there, I thought, oh, this is great. No problem. AUDIENCE: Can you bring
the microphone up. BOB FELDERMAN: Oh, sorry. And so Danny and I met somewhere
down near ISI, one of the copy shops down there. I don’t know. It was Marie Callender’s,
perhaps. And immediately after we
introduced ourselves, he pulled out the napkin. It’s not the napkin. But it’s a napkin. And he pulled out the napkin
and started scribbling pictures of a graph. And I had the same experience
that everybody did, that I actually wasn’t quite sure
what he was telling me. But we had the picture. And so what he was actually
talking about was that he had been working with Chuck Seitz
for many, many years. And as Chuck pointed out, he
was actually building some useful components. And he was building
multi-computer components. And what Danny had learned or
understood was that there was a potential there with those
chips to do networking with those chips, which were
ostensibly computer chips, really, for building
multi-computers. And so he took the chips and put
them together in all sorts of strange ways. This looks better
than the napkin. But what he was talking to me
about was the problem of taking a graph like that,
a strange graph that was connected in all sorts of
strange ways, and figuring out what the graph looks like. But the problem is is that you
actually have to do that without knowing much
about the graph. And you have to sit at the edge
of the graph and push messages into it and try to
understand when they come back to you where they went, and then
try to pick up the bread crumb trail. It was actually a
big challenge. So we took the napkin, and we
made it look much prettier and gave fancy names to it. But it was a very exciting
project. But another thing that people
have mentioned here which was very intriguing to me– so I
interviewed all over the place, Bell Labs, lots
of other places. And I ended up at ISI,
particularly because of this project, and particular
because of Danny. And what I could see was it
was a great environment, wonderful people, very bright. But they were also
building things. And I had been in a research
program with Len, which was fabulous. But we did a lot of
theoretical work. And it had been long since the
old days when he actually helped to build the ARPANET. We weren’t building
a lot of things. And so at ISI, what I could see
was people were doing very innovative things. But they were also putting
them together. And that was very
exciting to me. So that’s what we did. And in fact, as a young just
Ph.D. student, it was very exciting to me to actually
write one of these 110% acceptance ARPA contracts
with Danny. And I think we wrote a contract
for this not long after I showed up at ISI for– I seem to remember a huge
amount of money. It was $1 million or $3
million or something. And we got that contract. And so we put the chips together
in this network. At the time, Danny was very
politically astute. We called it ATM over Mosaic. At that point in time, 1991,
ATM was the networking technology that was
all the rage. And so instead of avoiding
it, we say, well, great. We’ll co-opt it and
do what we did. So we actually initially sent
our 53 byte packets around this little network, and
then realized that was too much overhead. So we made them a lot bigger. But it was a lot of fun. So we made this thing that
was called Atomic. It was very high speed
at the time. It was hundreds of megabits when
people were barely doing anything in the single
megabit range. After a few years at ISI,
where we worked on this project, Chuck and Danny
decided to try to commercialize that technology. And as I’m listening to people
talk before me, I was sort of struck– and I only just
realized it here– that this is yet another, let’s call it
a fork or a twig of Danny’s life, where was really
there at the beginning of cluster computing. So what Myrinet enabled people
to do was to take PCs, cheap computers at the time, network
them together with something much faster and more capable
than what you could do with a boring network that people
were using at the time. And you were able to beat out
the supercomputers of the day that were being built by
Cray and other folks. And so you could use commodity
technology, lash it together, and do really interesting
things. And in fact, one of our largest
customers early on was a company called Inktomi. Thos of you who have been around
may remember that. That was really one
of the first very, very good search engines. And it was doing it because
it was able to scale. And I remember going up there to
work with them and deliver some technology and fix
the problems that we have on our network. And I was just watching there
the machines just coming. And they were using
Sun systems. But the boxes of machines
were just coming in. And they would stack them up,
put them in the racks, connect them up. And their system was
just growing. They were competing against
Alta Vista at the time. That was also the other good
search engine at the time. And DEC had done a very good job
of building a good system. But they built it out of their
huge SMP machines. And so it didn’t scale as
incrementally and as easily. And Inktomi eventually
crushed Alta Vista. As Eric Brewer had mentioned,
then Inktomi got distracted from their search business to
look at other kinds of things, like caching and so on. And they took their
eye off the ball. And that’s when Google managed
to become the big search engine company. But Google today still builds
systems in a way very similar to what we were proposing
there. So I think this is an unheralded
part of Danny’s life that people don’t really
talk about, because Danny wasn’t the face of Myricom. But Danny was the only
networking person at Myricom. So we took the computer chips. We did net working at ISI. Danny taught me everything
I knew about networking. And then when we first started
Myricom with seven or eight of us, everyone else was from
Chuck’s research group, incredibly bright people. But they were mostly computer
architects. And so the networking side
really came from Danny. So Myricom actually had a very,
very successful run. The company’s still
in existence. A lot of the original founders
are still there. This is a hard graph
to read, perhaps. But this is what the top 500
supercomputing organization puts out every six months, which
is the fastest computers in the world that are willing
to register themselves with the top 500. Probably the NSA and other
places don’t even do that. But what this is showing is
what the interconnect technology is. And Myrinet shows up
there in 2002. Nearly 50% of the fastest
computers in the world were using this technology. And again, it was really during
that time, as Myricom started to show up in the top
500, that people really realized that you could do very
serious computing with effectively rack and
stack computers. And this is really
Danny’s idea. So one of the things that other
people have mentioned that was really true for me is
that I spent a lot of time with students who
come to Google. I happen to be at Google now. But students who come to Google
are grad students. We have interns come. And they’re often asking
career questions. What’s the right
way to do this? Should I go to do is startup? Should I go teach? Should I go do something else? And I try to give advice that
I didn’t plan my life. What I did was I went to
places where there were interesting people working on
on exciting projects, really smart people. And the rest kind of takes
care of itself. So I met Danny. I worked on an exciting
project. He and Chuck started
a company. We went off to do that. From there, I went and worked
with Steve Casner and Judy Estrin and others at places. And then that got me to
Google, and so on. It’s not about trying to
find the right thing. It’s trying to find
the right people. And Danny, thank you. You’ve been in a big hub in my
network of the things that I’ve done in my life,
so thank you. DANNY COHEN: Thank you. [INAUDIBLE]. VINT CERF: I don’t know about
you, but I have the feeling that nothing is official unless
it’s written on the back of a napkin. Almost all the really good
designs start that way. Neil, who’s next? It’s you. You’re on. NEIL GERSHENFELD: So this is
my part of the Danny story. And it’s a piece of Danny that
I think is widely known for impact, but less known
for Danny’s role. When I met Danny, it was a phase
when I was interested in the device physics of
user interfaces. And so that was furniture for
the Museum of Modern Art. That was health care for the
White House Millennium, a stage for the Karamazov
brothers, building as computing architecture
in Barcelona. And I thought this stuff
was really great. And I met Danny. And then I had an experience
that I’ve come to learn today is a common one. He explained to me all of
this is not interesting. And what is interesting is
what I wasn’t doing. And in fact, I should do what
is interesting and don’t do what isn’t interesting. And I’ve come to understand
that’s kind of what Danny does. And to do those projects, we
were making these things. It’s a $1 web server. And this was decades ago. And that wasn’t research. That was just an engineering
hack. And Danny thought, IP for $1 you
can paste into anything. That’s interesting, not
so much this stuff. And that grew into this paper. And I think this paper is the
first thing called “Internet of Things,” or the first thing
I’m aware of written articulating the notion of
there’s a reason for things to be on the internet. And when things are on the
internet, it’s IP. But it’s not exactly everything
the same. Constraints are different
for light bulbs. And so really starting to think
about the architecture of an internet of things. So in turn, from that paper,
that since a lot of people have credited with help seeding
the idea of “The Internet of Things,” but didn’t
really realize Danny’s key role in it. Because it was Danny who said
the projects were boring, the little engineering hacks were
interesting, which maybe is a summary of his life. The hack we were doing was
a really simple one. Since light bulbs don’t watch
broadband video but we need heterogeneity in transport, we
threw away almost everything. And the only good parameter
we trusted was time of transient events. So we’d just encode a packet
as transient events. And we’d encode it
the same way in whatever physics you had. And so Danny then introduced
me to a number of you. And we ran this meeting. And what came out of that
meeting was this interesting notion still percolating, and
I think still promising, of inter-networking in Layer 1 not
Layer 3, encoding in the physics in the same way IP isn’t
good anything, but it’s just good enough, doing the same
thing in device physics. And internet working in the
physics was the idea that came from that. And I think that one still has
decades to ripple out. And then from there, that’s
when it brought up to the present of my collaboration
with Danny. And many things radiated
from that. So that led to a lot of early
work on intelligent infrastructure for energy
efficiency where we show buildings use 3/4 of the
electricity in the US. And the third is wasted, not
from inefficiency but just from stupidity. They just don’t work right,
that heating systems fight cooling systems, and all
this pathological data. And so a lot of projects were
then seeded from IP for $1 that traced back from Danny’s
surgical intervention. Now there’s a postlude to the
story, which is I described all of that, an event
with Vint at the World Science Festival. That’s when Vint and I started
talking about doing what we’re doing today. And then coming out of that,
we talked about a shared interest in great apes. And so a day or two ago at TED,
Vint and I, with Peter Gabriel and Diana Reiss
announced an interspecies internet. And so you can think about the
internet of people to the internet of things. And then quite literally, it’s
interspecies internet. At TED, we had a video
conference orangutans watching great apes watching elephants
watching dolphins. And Vint gave a wonderful
presentation on all the technology for interspecies
communicating– people to animal, animal to
animal, and eventually someday animals to extraterrestrials– and launching this very
interesting project that you can really trace as a descendant
of, again, Danny’s intervention. So that’s my story with Danny. And thank you. [APPLAUSE] DANNY COHEN: Thank you. VINT CERF: We haven’t figured
out exactly which species Danny is yet. But eventually we’ll
figure out how to communicate with him. That’s what we’re working on. JIM MITCHELL: I knew that by
this time in the day, pretty much every story would
have been told. And I thought what could
I possibly add? And it’s a little bit of a
refrain on things you’ve heard, which is how Danny talks
people into things. And what I’ve actually seen,
though, is it’s how Danny gets a huge amount of leverage by
seeing what is a fundamental idea and then pushing it with
all of his friends. And the great thing is,
they stay his friends through all of this. So this is a little short story
about how Danny taught me about leverage. The other one. I got it. So I had heard about Danny for
years as a wild and crazy guy, I think from Bob Balzer, who
himself was not that calm a person, and from Ivan. I think I first heard the
story about the tenured position of adviser from Ivan. And finally, I met
Danny at Sun. He was a researcher. I had moved out of the JavaSoft
world to have a stint at Sun Labs. But we were both pilots. So that meant we always had
something to talk about. But I want you to know
I’m the sane one. It’s safe to fly with me. And Danny and I have flown
together, actually. Then in about 2002, when I’m
running the labs, Danny kept coming into my office and
saying, you know, I am sure that optics really matters
with computers. And DARPA has this BAA out
for doing supercomputers. And we should do that. And I said, well, Danny,
Sun doesn’t really do supercomputers. I’m not sure about this. But by about the 10th time, he
began to wear me down and convince me that we could
really apply for this. And it was only phase zero. And it was like a half a million
dollars or something. What could go wrong? And so we did. It ended up that not only did
we go through the first few phases of this, we only lost out
at the very end to IBM and Cray, for which I’m eternally
grateful, because we would have had to build the thing
that IBM almost delivered. But somewhere in there I got
talked into stepping aside from running the labs, which
is a pretty cushy job, and actually being the PI on this,
because we were feeling it wasn’t moving along
well enough. So that was lesson number one
in leverage from Danny. I made him pay for this. One of my friends, Butler
Lampson, and Chuck Thacker they were being made fellows at
the Computer History Museum down the road. And so I invited Danny along,
but told him he had to wear a tux. And so I got him
to wear a tux. So you see the leverage,
right? He gets me to work
for five years. I get him to wear a
tux one evening. AUDIENCE: Who tied the
tie that time? AUDIENCE: Yeah, it’s
a great tie. JIM MITCHELL: I think that– pre-tied, yeah. I certainly don’t know
how to do that. So we didn’t get that. But then about 2008, Danny’s
in my office again saying optics are really important
for computers. And we’ve got this great
optics team that we assembled partly. You’re going to hear from one
of them in just a moment. And we could really
do something here. We should apply for this. We did. And thanks to the great team
that now Oracle has and Sun had, we were awarded this DARPA
contract to work on silicon photonics. So he talked me into managing
that program. I’m still doing that and
microelectronics packaging. I used to do software sometime
in the distant past. I don’t do it. And that is all Danny’s doing. But I made him pay. He had to go to the Computer
History Museum once again. And I got him to wear a tux
for the second time. He looks pretty good
there, doesn’t he? So the lesson I got is he talked
me into spending 12 years on interconnect
and optics. I talked to him into two
evenings in a tux where he got free food and entertainment. And I figured out he’s much
better at leveraging others than I am. And then this made me think
of, so is this like native with Danny? Is it nature or nurture? You can’t answer that
question in general. But I thought, well, he
had Ivan as a tutor. And Ivan has talked me into
things, too, over the years. But after what I’ve
heard today, I’m really glad I came late. Because it is clear from all
these things, it’s like 90% nature with Danny, with
some maybe fine edges put on by Ivan. But it’s been a great amount
of fun, Danny. And thank you. [APPLAUSE] DANNY COHEN: Thank you. RON HO: So my name is Ron Ho. And unlike most of you, I
haven’t a known Danny for all that long. I met him about 10 years
ago, about 10 years ago this month, in fact. It was 2003. I had just joined Sun
Microsystems. I had been at Intel for the
previous decade and decided I want to work at a
research lab. And I decided to become a
student of Ivan Sutherland’s at Sun Microsystems. And so I showed up in 2003. My very first week
on the job– well, the very first day, I had
to push start Ivan on his motorcycle. That’s a different story. I think it was the second
day at Sun. And Danny walks into
my office. Now I didn’t know Danny. I’d read this papers. I’d read the Big-Endian
paper as a grad student and as an undergrad. And so I knew who this
Danny Cohen guy was. I didn’t know what
he looked like. And I didn’t recognize him when
he walked into my office. So in he walks. And he says, I’m Danny. I think you should read this. And I said, excuse me? And he said it again. Oh, OK. And I got what he meant. So I took the paper. And so I knew who
Danny Cohen was. But I hadn’t met
James Finnegan. I had no idea who James
Finnegan was. And the paper was “The VLSI
Approach to Computational Complexity.” Now I came from Intel. I was a big VLSI guy. I had done microprocessors. And so I started reading this
paper from this guy from down the hall. Didn’t know who he was. And I got more and more outraged
by this paper. What is this? And it wasn’t until I got to the
very end, where it said, “the more processors, the better
the paper,” that I realized it’s a joke. So that was my first
encounter, in fact. So I actually claim that I
have known James Finnegan longer than I’ve known Danny. So my second anecdote takes
place about a month later. And so then I’d sort of
met this guy, Danny. Oh, that’s the Danny Cohen. OK, great, great, great. And about a month later, I was
talking to Mary Holzer in the lab, another person who
happened to be at Sun Microsystems. And she had this video of Danny
doing this VoIP demo that we all saw earlier today. And I watched that. And I was kind of surprised. And then the next day, Danny
mentions, I didn’t develop my accent until many years later. [LAUGHTER] RON HO: It was probably two
months before I realized he was joking. And I think I told many other
people that same story. Could you believe that he
developed that accent later? Those are my two encounters
with Danny and his sense of humor. So those are my two anecdotes. I do want to talk about a great
honor I’ve had over the last several months. And that is Danny and I have
been working together to collect all of the Finnegan
papers into a book. And so the Festschrift should
have been a book, right? But if you ask any engineer to
give a timeline of when he will be done with something,
like in my case, you can’t trust him. Because I thought I’d
be done months ago. And it’s still dragging out. But I’d thought I’d
share with you– this is left and right,
correct– the book. So it’s right now 0.1 beta. Danny and I are picking fonts. And I think he likes Bitstream
right now. So we’re trying to figure out
how this all goes together. Hopefully by the summer? We’ll see. But it’s got Professor
Finnegan’s essays, plus some of his best friends’ essays,
like Danny Cohen. Because as he says, “their
general style matches mine so closely that I could have
written them myself.” The content, we’ve got a plea
for peace, AI, the VLSI approach, dating– that’s the important one– transportation, productivity. We’ve got the memoranda,
his forecast for 2036 as made in 1986. Those are fun. So I’ve gotten a chance to read
all these, as I’ve edited them and put them
into [INAUDIBLE] into a book form. We’ve got new ones– the high
tech applications of light and the time dimension. And somebody mentioned “The
Ethics and the Pursuit of Ethics.” That’s a great paper,
and certainly worth putting into this volume. So we’re going to
have all these papers collected together. It’s got the inimitable Danny
Cohen ASCII diagrams, because why have line art when you
can have text art? And it’s got even more recent
essay, such as his 2013 essay on “Evidence-based predictions
on IP.” So if you take the evidence of IP and extrapolate,
it becomes obvious that IPv12 is going to
come out in the year 2387 and have 8K address bits and will be
good for another 432 years. And that’s proven
in this paper. So as I said before, this is
a great honor of mine to actually go through these old
papers, edit them, put them into book form. And so let me know if
you’d like a copy. I’ll be talking with Neil. So you can ping him if you don’t
remember my address. And hopefully if I can get IEEE
to release the copyrights on all these, they’ll be on
Amazon for books on demand. So you can collect many
copies if you’d like. So thank you, Danny. It’s been an honor being able
to read all these papers and to have worked with you over
the last 10 years. Thank you. [APPLAUSE] DANNY COHEN: Thank you. Beautifully done. Let me try. I have some problems with
my knees, as you can probably see. VINT CERF: You’re actually
wired here. You may not need this. Try it. Attention. DANNY COHEN: Can you hear me? VINT CERF: Yes. We can’t make any sense of it. But as long as we
can hear you. DANNY COHEN: If you don’t,
please raise your hand. Raise your hand if you
don’t hear me. VINT CERF: Here we go. We even have the director’s
chair for you. DANNY COHEN: I’m from
Los Angeles. AUDIENCE: I’ll hold the chair
while you get on. AUDIENCE: Wait. Put it in the white marks. AUDIENCE: Oh, yeah. Thank you. VINT CERF: That’s good. Perfect. DANNY COHEN: Are we on? VINT CERF: Do you
need the slide? DANNY COHEN: For the past few
weeks or months, I’m thinking what to say here. And by postponing it one day
at a time, I managed to get all the way here without
thinking whether I could talk about. As some of you probably know, I
have a medical problem which makes me move much slower and
converse in a funny way. Like when I start talking, it
takes a longer time for the words to go through the mouth,
such that by the time I’m ready to say something
its place in the conversation is lost. So I apologize. Now you’re all kind to me and
don’t talk when I talk. One person at a time, so there’s
a chance it will work. What I wanted to tell you
today is about a new technology that we invented and
some application about it. We talked with the Army people,
and they explained to us that basically they have
two important problems. One problem is to see
the enemy at night. Today it’s easy. You raise your head, and you
see where the bullets hit. And the other problem is seeing
at night, which is a problem, and not being seen
during the day by the enemy. So those are dual problems. And they remind us of active
noise cancelling devices that most of you probably have. And the idea is that the
machines estimate the backround, and according
to that give you mindless noise signaling. So what you hear is
the [INAUDIBLE] sounds and the static sounds. They all get together and
cancel each other. So you get a great system. Next time you’re in a plane,
you can try it. I’m sure you’ll enjoy it. So the important part that we
did is about the day vision, how to hide from the enemy at
full sunlight, such that the enemy who will look at
you cannot see you. And it turns out not to be as
difficult as we thought. And the key to it is
the technology of projecting darkness. During the day, we project
light on the enemy so we can see. During the day, we
project darkness. So it sounds like an easy
challenge to do, like [INAUDIBLE] some bit somewhere. But it’s much more complicated
than that. So we built some devices. And we’ll demonstrate
some of them. And I think [INAUDIBLE]. This is Professor Finnegan. PROFESSOR J. FINNEGAN:
Do you have my badge? I lost it at a [INAUDIBLE]
conference a few years ago. I’ve been wondering
where it was. Thank you very much. DANNY COHEN: Don’t you
dare losing it. PROFESSOR J. FINNEGAN: It’s got
the right glasses, though. Thank you. Thank you. He’s my very close
collaborator. We’re looking for– not these. These are baby slides. DANNY COHEN: Let me steal two
minutes from you and describe to people what that is. NEIL GERSHENFELD: [INAUDIBLE] Professor Finnegan by not
responding to me in any way whatsoever. I understand what you need. But they don’t understand
what you need. DANNY COHEN: I wanted to share
with you my badge. In DoD, you must have a badge
to get any place. So this badge, which I have to
put myself, because I couldn’t find anyone else to do it for
me, it has my name obviously. The motorcycle universal
code mode, and red and green buttons. It must have some meaning
somewhere. The best stuff is magnetic
stripe on the back. Here it says in English that
it is a sample badge. With this badge, I managed to
get to places like NSA. VINT CERF: It’s OK. You’re actually wired. So you can talk. And we’ll hear you, and
save one hand free. DANNY COHEN: Thank you. With this badge, I go
to many places. The only place that I have
a little problem was the White House. The White House is guarded
by the Secret Service. And those guys do a much better
job and are trained much better than people
in most places. So this is my prized
possession. Oh, it also have gray hair that
you can see here, because I think that men look much more distinguished with gray hair. AUDIENCE: Well, talk about
helping at the White House. AUDIENCE: What did the
Secret Service do? DANNY COHEN: Oh, yeah. In the White House,
I showed it. And they looked and asked
me who issued it. And I managed to say, without
laughter, that it’s the [INAUDIBLE]. If you cannot read it, you’re
not supposed to. And they said, oh. [LAUGHTER] DANNY COHEN: And the guy
say, oh, one of these. VINT CERF: This isn’t the droid
you’re looking for. DANNY COHEN: I’m not sorry. PROFESSOR J. FINNEGAN:
Thank you, sir. Excuse me. DANNY COHEN: I did not introduce
We’ll cover that. DANNY COHEN: Thank you. PROFESSOR J. FINNEGAN: I’m not
sure this audience will understand. My close collaborator asked me
to come today, take time out of my busy schedule, to talk to
you about some of our new discoveries in optics. Very exciting. So as Danny mentioned, there
are two problems that the military’s talk to us about– seeing the enemy, daytime or
nighttime, and then our troops trying to hide from the enemy. Very, very serious and
important problems. Seeing the enemy was
solved long ago. Lights, or specifically
lasers, can be used to illuminate the enemy. The important thing that I’m
going to talk about today, we’re giving you a brief
glimpse into some new technology that we have that
we’ve been testing. We call it the daser. The daser we use on ourselves
to project darkness or invisibility. It’s your own choice to
hide from the enemy. It’s very, very, important. Remember LASER– Light Amplification by Stimulated Emission of Radiation. The DASER is darkness
application. It’s incredible. Now we don’t always need
to use the daser. In smaller systems, you want
to make your children invisible, we can use a DED. That’s a darkness emitting
diode for smaller things. But generally, for military
applications, we find that the daser works much better. And as my collaborator
mentioned, the idea is very similar to what you can do with
those canceling headsets, slightly different though. So camouflage is the
classic mechanism that people use today. But theoretical problems– you can see here camouflage
working well. But we can still see something’s
going on there. And then, of course, for certain
people, it looks a little silly. The key insight, as all of us
have known over the years, is the Fast Fourier transform. It solves every non-problem. But we were able to build an
FFT control card to do some very, very special
kinds of things. Now I know this is a fairly
learned audience. But for those of you who don’t
really know much about the FFT, the F– well, one of the Fs– stands for frequency. One of the other Fs is some
French guy, I think. But FFT is about a
frequency domain. And for those who are not that
technically savvy, just think about it as squiggly lines. Squiggly lines works
much better. So the key to this technology
is one, sense our own image. Take a video. We have lots of cameras here. Reverse the coefficients
in the FFT. Integrate it together with
the original one. And then project that new
image back on yourself. And voila, darkness,
invisibility, your choice. These are probably too
detailed for you. So we built our system. And in fact, we even
used some Mems that was referred to earlier. Lasers, rooms, block diagrams,
all very important. And always very important is
trying to find the minimal of the function. And here it happened
to be a 0.5. And that was really one
of the key things. So we built a few systems. Here is our prototype imager
and FFT matter with our lab assistant who, unfortunately, is
now deceased, but was quite a clever fellow. And our darkest projector, that
was an initial prototype. Yeah, it had to change. But we took it to the field for
our first demonstration. And it was incredibly
successful, except for one part. The transmit power was
a little too high. Some unfortunate things. But nevertheless, we were able
to go back to the laboratory, back to the drawing board,
working with a few graduate students who actually understood
more about electronics and things. And we built something
like this. And we were able to have our
successful demonstration with far less things. The other key insight that
really got us to success is, if you don’t have a Windows
implementation of your technology, it won’t work. As you can see, we have
beautiful, noisy FFTs on our Windows box. So we were able to have
successful demonstration. And I would show you the
successful demonstration. But it’s dark and invisible. So there’s not really an
exciting slide I can show you from that. But we have two sizes that ARPA
had asked us to develop. So we have the handheld
version. Looks very similar to
laser pointers. But it has very sophisticated
technology on the inside. It does come with batteries
when we ship it to the military. We also have, unfortunately,
the tankheld version. And down here, you can’t
see that 25– it is classified, actually. But in this audience, since
most of you have high clearance, you may know. This is 25 million AA batteries,
not included. But they can be towed
along later. So we’ve delivered this
to the military. They’re using it in Afghanistan
and other places. It’s been very successful. But you don’t hear much about
it in the newspaper because it’s very secret,
very classified. So what we’re now able to do is
our soldiers don’t have to worry about camouflage. During the day, we project
the darkness. They can just take a nap. Then at night, they can go
out on their sorties. So it’s very successful
technology. Keep it very confidential,
though. We’re not really talking to
the enemy about this. So just keep it all
in this room. Thank you. And I’d like to thank my
close collaborator over many years, Dr. Cohen. I really couldn’t have
done it without him. Thank you. [APPLAUSE] DANNY COHEN: You did it
so well [INAUDIBLE] and lack of smiling. That was terrific. PROFESSOR J. FINNEGAN: I’m
heading back to Washington. So I won’t be able to
talk to you today. VINT CERF: I want to know
where he got that coat. PROFESSOR J. FINNEGAN:
Top secret. VINT CERF: I also want to know
what it’s going to look like on the video. Who’s up next? ASHOK KRISHNAMOORTHY: My
name is yhtroomanhsirK. My name is yhtroomanhsirK,
kohsA yhtroomanhsirK. It’s not my real name. It was actually a clever alias
devised by none other than Danny himself to protect
my true identity. But I’m here to talk a little
bit about how to explain perhaps a little bit about the
DASER that you just heard from Professor Finnegan. So this is really to try to
explain the complexity that you heard, and the brilliance
of his accomplishments. DANNY COHEN: What did it say? AUDIENCE: Confidential,
do not look. ASHOK KRISHNAMOORTHY: So I have
to say that I first met Danny about 10 years ago. And at the time, I was living
on the East Coast of the United States. And I actually had relatively
moderately productive, happy and modestly successful
career. And all of that changed. All of that changed when I met
Danny and Professor Finnegan. I say this to you today, of
course, because I was told that I was in the presence of
trusted friends who want to know nothing but the
absolute truth. So that is what you’ll
hear today. They told me about this concept
called a DASER, the darkness laser. And that immediately enamored
me, enamored me to the point where I dropped everything
and decided to come work with Danny. Now the reason I became enamored
with it, because I actually thought that while
mathematically, quite complex as you saw, conceptually it was
possible given the right frame of reference. So let’s imagine a frame
of reference where a DASER is possible. Imagine a life form. And we call him IP. And IP is a two dimensional life
form that is short-lived, born in vacuum in the presence
of monochromatic laser light. IP only understands this. He’s short-lived. So he’s presumably
less evolved. And since he’s in the presence
of monochromatic light, he cannot distinguish if
you turn or modulate that light on or off. It doesn’t matter
which end it is. If you turn on and off, he
cannot distinguish between the projection of darkness or
the absence of light. And so in IP’s framework,
a DASER exists. However, it’s not a very useful
framework, because he is a two dimensional
life form. And we wanted to do something
a little more interestig. So let’s take another example. The other example is a
simple DASER pointer. So you know what a
laser pointer is. So we can make a data pointer. And we thought, well,
maybe we can make a DASER pointer in software. We can make a DASER pointer
in software. So if I were to just point, and
I can make it point, can go all the way around. And I can come back. Of course, it’s not a very
useful DASER pointer, because you kind of have to know where
you’re pointing ahead of time and program it in. But that’s just a matter of
building a better computer. And that is why my conversation
with Danny was so intriguing. Because he told me in 2003 that
he had just met a chap and convinced him. I think the chap was
named Jim Mitchell. And you had convinced him to
design a really fast computer that could actually take this
and project it into three dimensional light form and
calculate what the projection of your DASER would do by
sensing you, by measuring it, and by doing the appropriate
graphic display in three dimensions. But you needed a very
fast computer. It didn’t exist. But we had a DARPA contract
to build and design a very fast computer. So that then became the answer
to the catch-22. The catch-22 was we needed a
much faster computer to do the DASER optics. But we needed more optics to
make the faster computer. And that really told the
story of my last 10 years with Danny. So let me fast forward, because
we ran into some setbacks in actually making the device, budget and schedule. And this is where I like
to give an anecdote. Since this was not funded by
DARPA or ARPA, I actually asked Danny whether we could
raise some money to actually do this experiment. And we were at Sun
at the time. And Danny, of course, and his
foundation, said, yes. Price is no concern. So this was very encouraging. But after about two or three
phone calls, I realized that the reason why price was no
concern was that they had no intention of paying anything. So there was the budget part. And of course, the schedule
was a little bit tight for this event today. But nevertheless, we tried
to do our best. And we particularly have to
actually thank the actual sponsors of this demonstration
that you’re going to see, which is Holospex for the optics
of the DASER, and IEEE for the gain material. So let me then let you behold,
for the first time, Danny– this is the world’s first
prototype of a projection device which projects
not a DASER yet. [INAUDIBLE] of that. But it does project something
which no longer exists. So if you don’t mind, we’d like
you to point it to that and press anywhere on the box. It projects something that
no longer exists. DANNY COHEN: One,
two, come down. [LAUGHTER] VINT CERF: Oh my god. You’re right. It doesn’t exit anymore. ASHOK KRISHNAMOORTHY:
That’s yours. DANNY COHEN: Thank you
for the present. VINT CERF: What happened
to the microphone? Did you run away? DANNY COHEN: [INAUDIBLE]. VINT CERF: They’ll be a brief
pause while we swap computers to get the slides up. AUDIENCE: Point that at
somebody, and they go over to the dark side. VINT CERF: They go over
to the dark side. That’s minus 2. BARBARA TVERSKY: We’re going to
go way back to prehistory, before skis, before motorcycles,
before airplanes, before Finnegan, back to 1937. And we’ll bring you up to MIT. So I wasn’t there then. But I vetted everything
with Finnegan. So Danny was born in
Haifa in 1937. And Israel at the time was the
perfect laboratory for think different, to borrow a slogan
from another company. The people in charge
were the British. They were the people that
wouldn’t let Holocaust survivors or refugees into the
country who blocked their entrance by land and sea. It was settlers against
the British, David against Goliath. They made the rules. And the settlers figured out
ways to go around them. Does that sound familiar? The settlers had little in the
way of material goods. That situation only got worse
after the declaration of the state in ’48 and the
ensuing war. People had very little. And they had to improvise. To improvise, you first have to
figure out how they did it. And then you have to figure
out how to do it better. Sound familiar? If people were poor in goods,
they were rich in each other. And helping was a way of life. So the kids, like Danny,
were born in Palestine. But the parents came
from elsewhere. And they spoke a cacophony
of languages. They spoke Hebrew with
heavy accents. Some of them didn’t
speak well at all. They worked hard. Being a worker or farmer had
great prestige because these things built the country. And they’d been forbidden
to the Jews in Europe. The kids basically grew
up on their own. They raised themselves. They knew what was going on. Their parents were struggling. So Danny’s father David, on whom
Cody is named, came to Israel at 15 with his parents. His schooling ended then. He turned out to be a
good electrician. And that became his
profession. Before ’48, he was, like most
young men in the Palmach, the army that fought the British
for independence, and later became the Israeli Army. He worked actually with maps. And one of those maps would turn
out to be crucial in a later escapade of Danny’s,
one of his more reckless. His mother, [INAUDIBLE], came
from Hungary, also in her teens, to join a sister. The rest of her family
that stayed behind in Hungary perished. His mother worked as a nurse
during the War of Independence and later for the Army. They had a one bedroom
apartment. Danny slept in the bedroom
with a renter who helped make ends meet. And his parents slept
in a pull-out bed in the living room. You will not be surprised to
learn that Danny was a lopsided student. He was brilliant in math. And he wasn’t very much
interested in anything else except being outdoors
and running around. You also not be surprised to
learn that he was a daredevil, even before motorcycles
and skis. And inevitably there
were injuries. Here is one of them. At about 10, he fell and
sprained his thumb. And he needed a splint. And so since he was wearing a
splint, he could write it couldn’t write, he couldn’t
do his homework. That splint got worn for weeks
and weeks and weeks, until his parents came to school,
something they did quite frequently. So many of the teachers
were European PhDs. But the only jobs that
were open to them more were teaching. So here is Danny there. And that’s his math teacher
from elementary school. Danny’s around 10. And the math teacher, you
can see, is in uniform. He’s fighting the war in ’48. So Danny is 11. And Danny adored him and
is clutching him so that he won’t leave. So this center of life for the
kids was the youth movements. And they were of varying
degrees of socialism. And those degrees were
fiercely argued. The youth movements worked in
associated kibbutzim in the summers and on vacations,
bringing in the harvest, whatever needed to be done. They later both built kibbutzim
and served in the Army as units. So those summers and vacations
working on the kibbutzim were also pre-military training. And here you see Danny before
this growth spurt. And he was in his second or
third year in high school. And he was given the honorable
job of guarding the kibbutz at night. And for that, he was
given a rifle. And everybody was jealous. NEIL GERSHENFELD: No shoes? BARBARA TVERSKY: No shoes. Right, thank you, Neil. No shoes. NEIL GERSHENFELD: Same barber. [LAUGHTER] BARBARA TVERSKY: Danny was in
high school shortly after the War for Independence that
founded the state and established borders that were
life threatening to cross. So after ’48, if you went into
Lebanon or into Jordan, you were shot if you were seen. So at the same time, there had
been many stories of bravery and bravado that we’re
oft told from ’48. And crossing the borders became
irresistible to certain destinations to young men,
including Danny. So one of the coveted places
to go was the Litani in Lebanon, which was one of the
sources of the Jordan River. Another was Petra in Jordan. So they worked on a kibbutz,
this is a high school, near the Lebanese border. And this was quite close to
the source of the Litani. So one night Danny set
out with two friends. One of them was also
in that photograph with the math teacher. Another one who was
[INAUDIBLE]. They set out to go
to the Litani. It was, again, irresistible,
a few hours walk at night. They did come back obviously. But you sort of wonder what
motivated them to do it. Was it thrill seeking,
self-delusion, recklessness, thoughtlessness? Your guess is as good as mine. Many years later, or several
years later, he made a pact with another friend
to go to Petra. And at some point, they were
both in the [INAUDIBLE] of serving in the Army. And that friend called
Danny to come. Danny didn’t make it in time. And that trend was shot
on his way to Petra. So another high school story. Danny happened to I read an
article about an electronic brain and decided that
was what he wanted to do in high school. So the youth movement went
to the Army as a unit. They did basic training
together along with building a kibbutz. This is more Danny, the class
picture you can see. This is his [INAUDIBLE], the
group that he went into the youth movement with and built
the kibbutz with. And Danny’s right there. And I have a blow up
of that picture. Again, the same barber. The kibbutzim that did
agriculture were key to the country’s survival
at that point. They were on the border. And in fact, the kibbutz that
they built [INAUDIBLE] was on the narrowest part of
then Israeli on the Jordanian border at a point where Israel
was seven miles across. So you could basically walk
the country in two. That’s where their
kibbutz was. And I have some photographs
of building it. That might be Danny. We can’t figure out. But this is what they
lived in, tents. And they built this kibbutz
from scratch. These are some more
pictures of Danny. Here he is. And here this fellow here is
the one who was responsible for the pictures, [INAUDIBLE]. So in the middle of
his Army service was the 56 Suez campaign. Danny was, at that time, in
an elite movement, the paratroopers. And he fought in the hardest
battle of the war, the bloody incursion of the [INAUDIBLE] Pass. There’s the incursion
into the [INAUDIBLE] Pass. Again, we don’t know
if Danny’s there. He might be. But he was in a Jeep
like that. So that incursion into Mitla was
led by Ariel Sharon, whose name will be familiar to you,
against the explicit orders of his commander, who
was Moshe Dayan. Sharon did that repeatedly, did
what he thought was right and against orders. So again, does this
sound familiar? This is on the training. So after the Army, the youth
movement returned to the kibbutz they had built. Danny stayed a year or two. I think I have another. Yeah. This is really hard. What it says in the
corner is then. So it’s from that kibbutz,
but in those years. So he was working on
the kibbutz. he stayed on the kibbutz. This was your patriotic
duty and a pleasure. He worked in the orchards. And he also worked,
you will not be surprised, keeping the books. And to do this, he worked with
punch cards that he brought to Tel Aviv to an IBM calculator. So at some point, his
father got ill. He had a heart attack. And Danny took a leave
from the kibbutz and went home to help. While he was there he, since it
was close by, enrolled in the Technion to study math. And this, again, is
I think typical. [INAUDIBLE] said it a little bit. Danny didn’t plan his life and
work up a CV and decide between options. Things fell in his lap. And that fell in his lap. And he did it. So he never got back
to the kibbutz. He kept those friendships. And those friendships
run deep. They’re still his friends. The people that were there
then are amazing people. They’ve had amazing lives,
as scholars, researchers, business people, artists, poets,
and some of the many of those things. So Israelis also
love to travel. So this is some pictures
that we got from there. This is looking near the Dead
Sea in the Judaean Desert. These are hikes that
they took. When they went on these hikes,
they carried everything they needed, food and water
for a week. Again, Danny’s there with that
keffiyeh on his head, which he firmly denied that they’d ever
done that until we showed him the evidence of it. And again, those people are
still friends, very close and in fact, I know some of
the stories from them. So did I skip anything? No. So after he finished his
bachelor’s the Technion sent him to MIT. And that’s where the rest of you
picked up the story, and where the rest of you know. You can see the Danny
that you know from the earlier pictures. [APPLAUSE] DANNY COHEN: [INAUDIBLE]. DAVID COHEN: Many of you have
thanked my dad for many different things. And I wanted to thank all of you
for coming, and especially for talking and telling
me about his life. My dad is, as many of you said,
a remarkably modest man. And it wasn’t until the last few
years that I realized what an extraordinary range of things
he’s worked on and been involved with. Of course I have memories
of ISI. I guess I had a knee’s eye
view of that part of the development of the internet. But it’s really only the last
couple of years that I’ve learned about how many different
things he was involved in, and how many
sort of pies my dad had a finger in. I remember a few years ago,
one of the first times I visited my dad after he moved
up here to Palo Alto, I went to lunch at his favorites
falafel restaurant. And a colleague of his from Sun
Labs at the time, Puneet came and joined us. And I asked him, my dad,
he never really talks about his work. And I wonder if you could
tell me what he does. And I’d asked my dad. And he said, well, I’m doing
some things with lights, and like some switching. And Puneet said, I think
everything we do here your dad’s involved with somehow. Whenever we have a question,
we go and ask him. I wanted, when I started
thinking of what to say here, I knew that I wanted to talk
about my dad some teaching me math when I was very young. I don’t think I learned
any math in school. I guess I did multiplication
in first grade, and then started doing algebra in fourth
grade, and did calculus in seventh, all under my
dad’s [INAUDIBLE], of course on napkins. There’s an Italian restaurant
in the Pacific Palisades in Los Angeles where they give
you paper place mats. So we always went there. I’ve realized now because it
provided a very large surface to draw on. VINT CERF: That’s called
home cooked schooling. DAVID COHEN: I’m
very grateful. [INAUDIBLE] a lot of memories of being
young, but I hadn’t quite realized until I was sitting
here the point that I think this makes. I haven’t done math
for a long time. I’ve gotten into writing,
something my dad also has done and has shaped me a lot. But I think I understand. I was sitting here. I thought of something which
happened a few months ago which I think explains why my
dad never talks about his work, which is I was
sitting around. And for some reason I wanted to
know one of those formulas you learned in algebra, a quadratic equation or something. And I said, dad, I did all this
math when I was young. And I had forgotten this
very basic formula. And he said, well, why do
you need to know it when you can drive it? I’ve realized now that I think
that explains his attitude to talking about his work. He’s told me a lot about how to
switch packets and why you need fast and slow protocols. And I think that the projects
he’s done, he thinks why do you need to know about them if
you can do them yourself? That’s most of what
I want to say. But I have a couple of stories
to leave you with. I was going to tell the story
about the badge and sneaking into the White House. But he took the words
out of my mouth. So I’ll give you another story
that’s about my dad in the military, which he’s told me
many times and I think illustrates two themes that have
come up today, both his modesty and the impenetrability
of his accent, as you’ve told him many times. When I was young, I don’t quite
know what this project was, no one seems to have
turned up to talk about. He was doing something with
putting networks on large ships in the Navy so that all
the various computers that manage the different parts
could be integrated. And he tells me he give a
presentation at one point to some admirals also about
networking radars on ships. And he goes on about
20 minutes. And at the end he says,
any questions? And an admiral raises his
hand and says, just one. Why would we want to put
radars on sheep? [LAUGHTER] DANNY COHEN: [INAUDIBLE]. VINT CERF: That’s right. Now we know. DANNY COHEN: [INAUDIBLE]. DAVID COHEN: One more story,
and one more of my dad’s breakthroughs in networking,
which has not been brought up anyone. It’s a story my mom has told me
many times, which is how my father invented online dating. This predates the internet
by a few years. It would have been– what year would it
have been, mom? AUDIENCE: ’73. DAVID COHEN: ’73. And my mother at the time had
a job working at ISI. She was an undergraduate at
UCLA, an art student. And through friends
of her family– I think Shelley’s here. Oh, she left. She found this job as a night
computer operator attending the machines in the top floor
of ISI where they were, I think because they were
closest to the air conditioning, and
it saved money. She tells me that her job at
the time was to back up the network every night, and then
turn it off for a few hours, download it down to these
huge magnetic reels. And as such, she had access
to the ARPANET. And my dad had posted
a puzzle. What game was it? A Gomoku puzzle, which my mom
found and solved and sent him the solution. And he wanted to meet her. And given the size of the
ARPANET at the time, it’s unsurprisingly that they worked
in the same building. And that was how
my parents met. I’m pretty sure that
that’s the first prototype of online dating. VINT CERF: It worked. DAVID COHEN: Yeah, I was the
first physical product of it. I think that’s all
I have to say. [APPLAUSE] NEIL GERSHENFELD: Do you
want to say something? DANNY COHEN: No. DAVID COHEN: Thanks, Dad. NEIL GERSHENFELD: You
don’t have to. DANNY COHEN: No, I’m too
excited, emotional. NEIL GERSHENFELD: Danny is too
excited to say anything. But I think the day has
said everything. So I think the best way for
Danny to say something is to break and have a reception. VINT CERF: We need to make sure
the reception is actually there, since we originally
scheduled this for 6:00. NEIL GERSHENFELD: I think what
we’ll have is we’ll have a bio and email break. And the words really don’t do
justice after a day like this. But thank you, Danny. [APPLAUSE]

Add a Comment

Your email address will not be published. Required fields are marked *