A Half-Century of Hypertext at Brown: Session 1


UGUR CETINTEMEL: I
think we are starting. Well, welcome everyone. It’s great to see this
very nice turnout. For those of you
who may not know me, I’m a professor in
computer science. My name is Ugur Cetintemel. I’m also the current
department chair. I’ve been serving for
the last five years. So we’re very excited
about this event. Some of you may know
that this is also the 14th anniversary of
the department, which is very exciting. And I really can’t imagine
a more meaningful event to celebrate our
14th anniversary. So when Norm first
mentioned this idea to me, I got really excited and I
thought that this is something that we will need to do this
year, which is wonderful. And especially given the long
history of hypertext work here in the department,
and all the seminal work that the department has done
and many of you have done, which is something that we
cherish and acknowledge as one of our main contributions
to the world, it’s great. So I’m really excited. So I thought that
I have a couple of minutes to talk a little
bit about the department. Many of you are alums and
friends of the department, but some of you may
not be following what has been happening recently. For the last four or
five years, especially the undergraduate program has
been growing significantly, which is awesome. So every year, we’ve been adding
10%, 15% to the enrollments, to the number of
concentrators we have. So this year on
Sunday, we’re going to be giving out about
250 diplomas, about 39%, 40% of our graduating class
is actually female students, which is awesome. [APPLAUSE] So Brown is the top
concentration, most popular concentration at Brown
by a large margin for the last couple of years. And the UTA program, I
know many of you have UTA. So that program has also grown
with growing enrollments. So in the fall
and in the spring, we had about 300 UTAs
working for the department. And I would like to thank
all of you who actually contribute to the UTA endowment,
which we completed last year. I would like to
especially acknowledge Norm Meyrowitz here, who took
a big leadership– thank you, Norm. [APPLAUSE] So in response to all these
sort of increased interest in computer science,
but also acknowledging the importance of computer
science as a discipline that empowers all
the other disciplines and that’s really critical
for a modern university, a modern society,
the administration allowed us to grow the
department, the faculty, by about 50%. So that’s going to happen
within the next five years. They are also very
excited about that. We started recruiting. So that’s going to help us
reduce the size of the classes, improve the student
to faculty ratio, but also start really designing
multidisciplinary research programs with a focus
on societal impacts. So all of these
things are happening. So you’ll hear
more communication about all of these things. So coming back to
this, I remember Norm sending me an email
about this event, I think, last December. I got excited. I asked him three questions. I said, how long
is it going to be? Who’s going to pay for it? [LAUGHTER] And who’s going to do it. So Norm said, it’s going
to be a low key event. Three hours max. [LAUGHTER] And then he said, he’s going
to find external sponsors to cover the cost. And then he said, and then
he will do the organization. Of course, it turns
out that, what? It’s nine hours? The program is nine hours long. If there are
students in the room, I’m sorry to say
that you won’t be able to get those new machines
in the sound lab next fall. [LAUGHTER] But his third answer
was accurate, in fact. And then he did
the organization. Each did what they do best. Andy picked the food,
and Norm did the rest. [APPLAUSE] AUDIENCE: And the wine. And the wine. UGUR CETINTEMEL: So
while lots of things have changed at Brown, you will
notice that some things remain the same. So HDAs still do the
professor’s work. [LAUGHTER] All right. Thanks again. Thanks for coming. I would like to thank, of
course, Andy and Norm, but also Lisa Manekofsky, who
is, I think, outside, who really did an amazing
job with the organization. But it’s important, I’d
like to thank all of you. I know how much time you
spend getting those software and hardware up and running. And of course, this
is a big team effort. So I’m really
excited to see what you guys have come up with. So I’m going to now
turn it to Andy, who’s going to introduce Norm. [APPLAUSE] ANDY: Thanks, Ugur. Ugur has been the best
chairman we’ve ever had. And I want to thank him for
his unwavering support of all of our many activities, And. This one in particular. He is paying the bills. So thank you. [APPLAUSE] A lot of people have helped with
this, but the man of the hour is Norm. So I’m going to say that this
is a five ring circus, one for each decade, and
the ringmaster is Norm, as we all know, in his
traditional role of, I’ll take it on and then I’ll
let mission creep dictate what happens next. Classic example. He’s also an amazing
digital archaeologist. He went through stuff
that people submitted and he had help– Greg Lloyd and others of you
pitched in– but he found this amazing stuff that most of us
did not know about and engaged in one of the most amazing
pieces of digital recovery restoration– I call it the Resurrection– and I’ll let him talk about
his particular Resurrection. We have a second
one, which is the– don’t give the lines
away, all right. You’ll see multiple
amazing resurrections, but I pronounce Norm as
the King of Resurrection across the spectrum
of all the things that could be resurrected. [LAUGHTER] And finally as a sort of catchy
title that doesn’t quite work, I also think of him as
the Houdini of Hypertext. Houdini, known for his
disappearing and reappearing acts, but more he was
a really good magician and his escapes were just
part of a much larger spectrum of things that he
knew how to do. Norm can do everything,
does everything, with great gusto, persistence,
chases things down that you didn’t even know
existed, and I’m so very, very proud of everything
he’s accomplished. I want to thank all
of you for coming. I want to thank you for all the
years of work that many of you here did– some
like Marty, almost two decades on these
various crazy projects that you’re going to hear about. And this has been, I hope
for you, a magnificent way of recreating some memories,
restoration of our memories– I really need that badly myself. And will now turn the
program over to Norm. Thank you so much, Norm. NORMAN MEYROWITZ: Thank you. [APPLAUSE] The mics off. ANDY: I forgot. Am I on still? AUDIENCE: Yep. ANDY: Very important
announcement. Norm, as you know,
among his many roles was also a superb
TA and head TA, and he was the spark plug
that started the UTA endowment program. It’s 300 UTAs per semester. You might have thought
it was per year. Per semester. Unbelievable. Anyhow, he’s the spark plug
that has made it possible. And in his honor,
tomorrow there will be a dedication of the
Norman Meyrowitz Fishbowl. The Fishbowl is our official
name of our major TA room. Karen did the appropriate, aw. Can we all do an, aw? One, two, three. AUDIENCE: Aw. ANDY: So he will be
forever memorialized as the UTA of UTAs. NORMAN MEYROWITZ:
Well, thank you. [APPLAUSE] That’s a surprise. Oh, there we go. Welcome to just– make you not
so nervous after Ugur’s talk, I will not be asking
anybody for money at any point during
this weekend. [LAUGHTER] But how did this start? Well, the folks at
the Computer Museum– Mark and David,
who are here, who also happened to be
Brown alums, they’re the curators of the web and
the internet collection– asked Andy and me to see if
we can get our systems running for a celebration of Doug
Engelbart’s 50th anniversary of his mother of all demos. So amazingly enough,
well, Dave Durant has been building an emulator
for the MLAG, for FRESS. And he and Andy and others, they
were able to get that working. And then I thought
intermedia was gone because we
had hard disks go, but one of our folks, Bern
Haan, had an old roadshow disk and I was able to find old Mac
2Cis, old cables, old monitors on eBay, and we
attached the disk– Tim Catlin and I– and amazingly enough,
it booted into Mac OS. When we tried booting
into AUX, it crashed. And somehow in the back
of my mind, I said, I remember something about
there being too much memory. So we had to go into the machine
and change the 32 megabytes to 8 megabytes, because
of 24-bit addressing. So it’s the first time I’ve had
to reduce the amount of memory. [LAUGHTER] So let me start kind of by
just doing a level setting. If you look into 50
years ago and you try to plot the growth
of computing linearly, it’s ridiculous. If you try to do
it by powers of 10, the bandwidth has gone up by
seven orders of magnitude. The memory has gone up by
four orders of magnitude. And I think the CPU and disk
is five orders of magnitude, which is ridiculous. If you think back
to 50 years ago, the number of people with the
understanding and knowledge of hypertext was less than 100. You know, that’s like
the beginning of TV with Philo Farnsworth
and Dworkin and Sarnoff. I mean, that’s really
early, because now there are 4.4 billion people
who understand hypertext. So that’s a pretty– three billion, 999 something. So where did this all start? There are pioneers of hypertext. One of them, Vannevar Bush,
is the one who’s cited. He wrote this article in
Atlantic Monthly in 1945 and then it was reprinted
in Life magazine in September of ’45. It was called “As We
May Think,” and it was a very visionary article. It talked all about how we
should have machines that can record our information and
create associative trails so that people can
travel through trails. And the machine looked
like that on the left. He talked about a vocoder
that would do voice recording. There were two multiple
monitors down there, multiple screens, where you
could use a pen to annotate, and it was all on microfilm. Paul Kahn has put together– they put together a
while ago an animation of what that machine
might look like. SPEAKER 2: The owner of
the memex, let us say, is interested in the properties
of the bow and arrow. He has dozens of possibly
pertinent books and articles in his memex. [CLANKING] First he runs through
an encyclopedia, finds an interesting,
but sketchy article, and leaves it projected. [CLANKING] Next, in a history, he
finds another pertinent item and ties the two together. [CLANKING] Thus he goes, building
a trail of many items. Occasionally, he inserts
a comment of his own– either linking it into the main
trail or joining it by a side trail to a particular item. [CLANKING] NORMAN MEYROWITZ: So it’s
sort of like the internet without ads. [LAUGHTER] Now I’ve had this
magazine in my office for years and years
and years, and I never noticed until I was
taking photographs– so he wrote this article
originally in 1939 before the war, and then it
was published after the war. He was Roosevelt’s
science advisor. But in 1939, this
was pretty visionary because that ad there, nylon
had not been perfected yet. And so that’s an ad
for men’s sock garters, because socks
couldn’t be held up. So that’s kind of
a juxtaposition of how forward looking he was. Doug Engelbart is
the second person who’s cited as a pioneer. He did a system starting in 1962
for augmenting human intellect. He wanted people to actually
be in a cockpit of information. He wasn’t all about– even
though he invented the mouse, which is there on the right– it was all about
having so much power that you can do anything
with information. And NLS had all sorts of things. It had linking,
direct manipulation. It had outlining. It had audio and
video conferencing, view specs, keywords. I can go on and on forever. But he was an
inspiration and he did this demo called the
Mother of All Demos that was a 90 minute video
conference between Menlo Park and San Francisco. And it turned out it
was at a conference that Andy attended and spoke at. This is Bill English who
was down in Menlo Park at the other end. And you can see over
there on the slides, it says jump to entity, jump
to name, jump ahead and return. So these are the linking
things they had by 1968. Well, the third
pioneer is Ted Nelson, and Ted went to Swarthmore,
graduated in 1959. He was a philosophy major. He went off, he took a
computer course at Harvard, and he found them intriguing. so he had read Vannevar Bush’s
article when he was a kid. Doug Engelbart had also read it
when he was stationed somewhere in the Navy. And that was an inspiration. And in 1963 in Miami, Ted said
he coined the word hypertext. And this is the first
published paper in 1965. And that’s the first published
use of the word hypertext. And these are some
of the drawings in that original paper, and
they look like the drawings that we all do of hypertext now. Well, Ted has invented lots
of words besides hypertext and hypermedia. All of these words– docuverse, which I think
is really important, because that was his view of
what we were trying to create– a universe of documents. If you want more info,
he wrote a great book called Computer Lib. Actually it was two books. If you turned it upside
down and turned it around, on the other side
you got that book, and then Literary Machines
was more in detail about the docuverse and Xanadu. And this was my favorite
coinage and kind of the philosophy
of the whole thing. Everything is deeply
intertwingled. So how did hypertext
get to Brown and how did we end up
with a fourth pioneer? Well, that’s not Brown. That’s Swarthmore. And not only did Ted
go to Swarthmore, but Andy went to Swarthmore. [LAUGHTER] And he graduated
a year after Ted, but he knew Ted from
his freshman week. They had a chance meeting at a
conference I believe in 1967, and Ted told Andy about
all these ideas he had. And Andy said, wow,
that’s really interesting. I have this graphics
computer and I have all these undergraduate
and graduate programmers and I think it would be
way more edifying for them to build a system than
to sleep at night. [LAUGHTER] So that’s how
hypertext got to Brown. Ted came, Andy came, and
many of you did not sleep. [LAUGHTER] So it’s the early ’70s– I mean, the early ’60s– late ’60s and early ’70s. If you went to school, you
might have looked like that. [LAUGHTER] Your calculator
looked like that. 1970, the pocket
calculator didn’t really become commercial until 1972. Your word processor
looked like that. Your campus computing
looked like this. This is just the panel of the
360, which took up a room. And that had 512k. So if an elephant steps on your
watch and you pick up a piece, it will be more than 512k. [LAUGHTER] That was the terminal
that you had. No screens. It was a Selectric typewriter. But given that state, or
sorry state of computing, in the castle,
stuff was brewing. And so now I’d like to invite
Andy and his sleepless crew to give the next step. [APPLAUSE] ANDY: So a little bit
about my own background. I got my PhD at University
of Pennsylvania, starting off as a
hardware engineer and switched after I took
my first optional course on digital computers, a
word that I had never heard, having graduated with an
engineering sciences degree. I know about analog
computers, but never heard of digital computers. My office mate talked me
into taking this course, and I fell in love. Now in this course,
we had assignments. They were writing 0s and 1s
in binary on a piece of paper, because the university’s
only computer– 25,000 people, one computer– was a UNIVAC 1, and it
was too scarce a resource to be used by mere students– a theme you might hear
about a little bit later. So we wrote 0s and 1s on paper
and had them hand-corrected. Well, that was an interesting
environment, but at some point, I did get to see the new IBM
mainframe that was rolled in and I was able to get some
time on it occasionally. I was, at that time, very
interested in information retrieval. This artifact is a
IBM card or punch card with a little piece
of microfilm in there, which had terrific
storage density, so you could store a lot
of engineering drawing on that little
piece of microfilm and still use the computer to do
Boolean retrieval of keywords. And those of you who did the IR
assignment in the early days– that’s where that came from. Well what– second thing
that changed my life, besides that first course,
was seeing Sketchpad, which again, blew me away. Communicating in picture
language in real time, as opposed to
dealing with cards, which was the best medium we
had available at that time, was a box of cards. For those of you
interested in souvenirs, I will part with a few
precious remains at the end. I’m moving on a
little too fast here. Let me go back here. So I decided to switch
from information retrieval to computer graphics. We didn’t have a
graphics console, so my thesis actually printed
out things on the line printer with little asterisks and
other marks to simulate lines. And that’s all we
thought about– line graphics, in those days. Computer science, just to take
you back, wasn’t even a thing. Nobody knew what
computer science was. It was not a name
in common usage. And if people heard
it, they said, what? Is that like political science,
i.e. not a real science? [LAUGHTER] And that was the
attitude that everybody had to this upstart young field
that didn’t have departments and didn’t have
degrees or anything. So that idea of having
sort of to fight for our legitimacy continued
through the ’60s and ’70s, and these arguments about
what is computer science continued for decades. Now we don’t worry
about it anymore. Hey, we own the campus. Why do we have to
defend ourselves? All right. Computer Science. Computing resources,
however, remained scarce. And there was this continual
fight for funny money. So there was a central
administrator, who allocated it per course, per person. There was micro accounting
that IBM did on their operating system so that you could be
billed for the number of disk sectors and bytes that you
used, the number of microseconds of computation. Everything was micro tome
so that you can get a bill, and you had to find money to
pay for this stuff somewhere. It was an attempt
to try to pay back the monstrous monthly
rental fees, which ranged in the tens of thousands
of dollars for that computer. So this fight for funny
money was very real. Many stories about it. I’ll just tell you one of the
ones where I had a big fight to get money for using
computers for word processing and hypertext experiments. I had this big
fight with somebody named Paul Mader,
who was in charge of allocating this money. And at one point, he
said, the computer is for scientists
and engineers who have real problems to solve,
and let the humanities users use typewriters– the equivalent of let
them eat cake, in short. And then I pointed out to
him that, in the brochure about the computing center,
it said that the computer was a resource for general use. And I said, it
wouldn’t look very good if there was an article
in the BDH or elsewhere saying that the computer was
for everybody, as long as they were in engineering
and the sciences. So that’s one of
the fights I won. Rod Chisholm, a famous professor
that Joe Strandberg will mention briefly, got his money. We got money for courses. But it was always a fight. All right. We heard about Ted Nelson. I met Ted at the spring
joint computer conference and we decided, yeah,
let’s put on a show. I had the garage. I had the equipment. I had the people. He had many of the ideas. A bunch of you were involved. You were all not just schleps,
you were idea generators. You were treated as
equals most of the time. [LAUGHTER] And I’m very proud of that
idea of using undergraduates in research, because again,
part of the culture that was absolutely unheard of. I took shit from my colleagues
for having undergraduates tromping up and down the
halls of applied math all hours of day and night. And I was told famously,
you can’t do research with undergraduates. And then we published
papers in which– one of the very first one, Steve
Carmody was the first author. And that was a pattern
that we continued. Well, it had a dual nature. It was both for testing
out Ted’s hypertext ideas, but I also wanted it for
producing documentation so that we didn’t have to
depend on Selectric typewriters anymore. And it was kind of a constant
struggle between the two of us which set of features
would dominate. I want to give a big thank
you to Sam [INAUDIBLE],, who was the director of the
IBM New York Scientific Center, whom I told about this project– because it was really a bootleg
project and I hadn’t told him we were doing it until I
could show him a demo of it, and then he said, wow,
that’s really great. And then he helped me try to
sell it to a number of people that I’ll mention later. I’m now going to
turn it over to one of the very first undergraduates
to work on that project, Terry Gross, who is a IP
attorney and civil rights attorney in San Francisco. You’re up, Terry. [APPLAUSE] TERRY: So basically
what happened at this point is Andy
brings Ted down to Brown and we’re all working our late
night hours in the computer lab and stuff. And they start telling us– basically they start
visioneering to us and telling us there’s this– now that we have
digital computers and we have displays
screens, we’re able to look at information
in a different way. And information can all be– all information in the universe
can all be interconnected. And we can peer at it
and we can manipulate it using display screens. But then they said, but
then there’s a kicker. This already exists, but you’ve
been too blind to see it. It’s like the blind man
and the elephant’s trunk, that that’s what text
has always been like. You’re limited by the
tools that you have. And the tools that
we’ve had so far have been either stone
tablets, or paper and pencil, typewriters. That sequential. That’s the only way that
you can look at text. But there have been
these work-arounds that have allowed
us mere mortals to peer into this
intertwined, interconnected multiverse of documents–
what Ted called the docuverse. And the way that you could– these limited tools
that us blind people had were footnotes
and bibliographies and table of
contents, which were different ways that took you out
of this two dimensional thread. And then they said, and
there’s also the bibliography. And that’s links. That’s how all of
information in the universe is linked together. And this has already existed,
but now we finally have– with these digital tools
and with the display screen as the device that allows us,
and now you can finally see it. And this was, I guess I
would say it was sort of, I think to us, it
was like explaining to an early physicist
string theory, or given that it
was the ’60s, it was really a psychedelic
view that the insight that they gave us
of what information was like that we had. But then Andy said, Andy
and Ted were saying to us, but now the next
task is, how are we going to access this and
create this initial thing? How are we going to be
able to manipulate and use this information? And that’s when
they started talking about an author’s console. I mean, we finally
have some tools here to be able to do this. And so the initial concept is
let’s have an author’s console where we can sit
there, as you’ve seen, at this display screen. And we want to be able
to do two things– we want to be able
to manipulate text. I mean, word processing
at that point– Norm showed you the typewriter–
word processing was basically printing something out
on a piece of paper, marking it up with
red ink with all these arrows and cross outs,
giving it to a secretary and have them manipulate it. And it was revolutionary
to think, wait a minute, you could do all of this
by sitting at this terminal and poking at things. The only input device we
had then was a light pen. We didn’t even have mice yet. You know, it’s just– but
to be able to push it, this thing, push some
buttons and manipulate text. So I mean, that was
the first thing. Let’s manipulate text. But an essential part of what
we were trying to develop also had the links and the
hyperlinks and branching. And so it was basically– the assignment that we
were given at this point was basically using
a computer that has less computing power than
the chip in your refrigerator. How are we going to be able to
manipulate all of this text, create all of these links,
store all of this information, be able to swap it in and
out so that we can see it? And so that was essentially
our first assignments. But I’ll leave you, I mean,
there were two things about it. I mean, the first thing is that
the initial visioning thing that when Ted started
talking about the docuverse and everything, his
descriptions of it– which I don’t have time to
read, though I was going to– mimic exactly what we now
know as the hypertext universe that we live in– almost every element of it. What you see up
here and what you’re going to hear after this is all
the technical details that it took to get there. This is from a 1966 memo
proposal of the Xanadu system that Ted did. And all of the elements
that you see in here– an undo, multiple screens, all
the linking indexing and stuff like that– all of the
elements that we have now were envisioned back then
before any of this existed. But I guess the second
cautionary thing that I will leave you
with before this is just to understand that Ted, with
everything that his vision, he is deeply upset with the fact
that where this has led to is the World Wide Web, is
that what we have right now is we are still blind
men in Ted’s universe, and there are still many multi
dimensions that are still for us to be defined
and to find out later. And now I’ll turn it
over to the people who actually started developing it. SPEAKER 3: Thanks, Terry. [APPLAUSE] So here’s some things about HES. I won’t take time to explain
them to you, because you know all this stuff already anyhow. It’s just kind of a summary
of the major features. Uni-directional
links only, at first. Ted’s back button,
hugely important. I think it’s the most important
function in any system, be able to undo, in some sense. We had from graphics,
master instances, where you could use
blocks of text repeatedly. And I decided I wanted
this really fancy paging system, where pages would
grow and shrink dynamically so that we could modify
in place basically. Turned out to be a very
bad idea, but it worked. And FRESS reset that idea. There’s our friend
Steve Carmody, who went to Holland with Marty
and a bunch of other people where we started
a computer science department and a computer
graphics research group. And he’s going to tell
you a little bit more about the details of HES. That, by the way, is the
traditional Dutch herring. And that’s how you eat it. [LAUGHTER] STEVE: My daughter just recently
took a trip to Copenhagen and I was using
this to inspire her, but she didn’t take the bait. AUDIENCE: I don’t
think the mics on. There you go. STEVE: So this was
our personal computer. There’s the person using it. And I believe this is
Chris Braun, we’re told. And as you can see,
the only thing I regret is we had apparently you could
get 2250 that had a foot pedal, so when you pointed the
light pen at the screen, you stomped the pedal. But ours lacked the pedal. So there’s– So the initial one,
had the 2250 mod one on the top there,
and later on we had this wonderful
1130 computer and we were doing remote computing. We had an 80 foot
long by Sync line that connected the
1130 and several people wrote an emulator on
the 1130 for the 2250, and this was actually
demoed by IBM at some point. AUDIENCE: At a computer
conference in Boston. We were in their booth. STEVE: We were at the
center of the IBM booth– I’m sorry, our code was at
the center of the IBM booth. They kept throwing
me out of the booth. They thought I was
spoiling the show somehow. On the right hand side,
you see the connections for some of the low end devices
that FRESS initially supported, so 2741s are type– literally Selectrics. 2260 was a small display
with a keyboard, no light pen or anything. So can I give you this? So I’m going to
demonstrate I’m old. I have notes. But in this kind
of a group, that’s nothing to be ashamed of. AUDIENCE: He is so old that he
has just retired from Brown. [LAUGHTER] A zillion years of service
as a networking guru. STEVE: So Norm, to
my immense regret, actually found the
source for HES online and then unfortunately
David Doran took the bait and he broke the files apart so
you can actually look at them. So several things about that. It’s amazing how
few files there are. It’s a small system compared
to the systems we write today– that the undergraduates
write today. It’s almost trivial,
looking at it. And it’s shocking how close
it is to the hardware, that scattered through the
code are all these things where it’s embedding these hardware
commands in the stream of data. And it was written– you know, today people write
programs, applications, and they’re built on
these huge frameworks. And that idea hadn’t
come into existence yet. So you’ve got this
giant mess of code. Oddly, it did– there was
enough sense to break it apart, and it actually was, to my
mind, one of the first instances in this MVC model. So sitting over there
is Mark [INAUDIBLE],, who wrote the display
portion of it early on. And it did actually break
apart using that model. But at this point, I think
code is useful primarily as a milestone, showing us how
far the concepts of programming have evolved. Next, please? The other thing that I
think is really worth noting is the process
that was used here. So lost in the history is– neither Mark nor I can
remember how exactly we got involved here. It was, we were taking– SPEAKER 3: My course. STEVE: We were taking
applied math 101, 102. Somewhere in the
spring, the two of us got drafted into
working on this vision. How that happened,
we don’t remember. I assume we got kidnapped
in the middle of the night and we were introduced
to third shift. I worked on HES across
the summer of ’68 and Terry was there. I remember him
ducking in and out. I remember him needing
to take a couple of weeks for a vacation in Chicago
in the summer of ’68. And by Labor Day,
there was a proof of concept for this system. And then from that point
on, I remember the sort of collaborative process. People would use the system. There was constant feedback. There was constant evolution of
the facility, the functionality the program offered. Things were added. Sometimes things
were taken back out. And this is the process that the
commercial world has discovered and they call it Agile,
but we were doing this since ’68 and ’69. The last thing I’d
like to note is that when I got involved
in this, I was– I think I was
probably 17 years old when I started working on this. I was not much more
than a child actually. That first summer in ’68,
I was living at home. My parents were having a little
difficulty with the third shift routine. And what didn’t occur–
it wasn’t until years later that I look back
on this and realized that what I was
doing that summer was building something that
had never been imagined. But while I was in the middle of
it, this was perfectly normal. This was what you did. This is what you’re supposed
to do with your life. You’re supposed to think. When someone presents
you with a problem, you’re supposed to
not worry about what they put in front of you. You’re supposed to think
about the real problem, the big problem, and
then you’re supposed to solve the big problem. I didn’t have time to
think, so Ted and Andy say, they have this vision
and they say, go program it and you trust them, right? If they say you
go program it, it must be because it’s possible. It’s doable. [LAUGHTER] And off you go. But even more than that
was their trust in us. And I think you’re going to
hear this many times today, that they allowed us to do this. They brought us in, they
let us work on this stuff, and then they actually
exposed us to IBM and I remember a trip
to the AP in New York, which was kind of mind-blowing. There was the
devil agency stuff. They trusted us to be
able to take these ideas and present them and grow
them and build on them. And that was my major–
it wasn’t until years later I realized that
was my major takeaway from all of this. And I spent the
last 20 years kind of building this infrastructure
across the globe that supports distributed research. And the software team I
led, the software they built is used at 4,000 campuses
and sites around the world. But I was presented–
at one point, we had a picture of a napkin
on our website that started it. And that was the problem
that was put in front of us. And we ended up with something
that was much larger. Thank you, Andy. [APPLAUSE] ANDY: Thank you, Steve. Great. Well, as part of Norm’s
enormous resurrection effort, he managed to find
a piece of really hard to look at
videotape and then used his incredible
multimedia skills, honed at places like
Macromedia, where he was the CTO, to put
bits and pieces together. It’s not easy to
watch, but we’re going to get a live
narration from Diane, who was the woman in the
video that you’ll be seeing. And that will give you
a little bit of an intro to the feature set. DIANE: So I’ll just– ANDY: I should have
mentioned, Diane next became a fellow at
IBM in networking and then she had a second
career as undergraduates czarina at UNC, from which
she just retired, but she reassured me she’s
going to continue teaching. Good. [APPLAUSE] SPEAKER 4: And Diane, we are– as fast as you can. DIANE: You think we’re
running a little late? SPEAKER 4: Yes. DIANE: The good news is it’ll
be as long as this video, maybe. Here we go. So that’s the screen. Over on the left there,
you could see the board. And you can tell we had
high tech graphics there. Basically going to create a link
by just clicking on the button there. In a second, you’ll get
an asterisk that shows up. We skipped a few
pieces in this video, so you don’t see us
actually creating a label. Here’s a description
of where it’s going and you’ll see it says,
type in explainer or travel. Travel meant navigate to another
part of the system, we think. We’re having a hard time
remembering that part. But in fact, this
was the pointer to the biblio, bibliography. And magically comes
up an asterisk. There we go. And that’s the link. Yay. And then we can follow the link. Here we go. [LAUGHTER] And if you wanted
to do it, you can click on the button
that says Get Label. And here you go. It’ll give us a list of all of
the labels, which we magically created somewhere. There are pieces of
it that some of us have been having a hard time
remembering about like– you could just see
that one there. And the view specs is actually
the text formatting part of it. This is where we all,
where predecessor to Word and all of those good things. You can show the links or
not, and we can create titles. I’m trying to remember
where we are here. You can get justification. Yay. you can do that and
hypertext, in HTML too, these days. Different widths,
different formats, how wide the columns are going to be. And I’m not speeding this. This is just on speed. And indent. We’re going to do
hanging indents. That’s actually kind of
risque for the internet still these days. Negative indent. You know, it was
kind of exciting. And there we go. And we’ve got the
hypertext hanging indent. And coming up, do I
need to click again? I think I need to click. This is just some stills in
case you couldn’t read them when you were watching them. And we’ll talk about how
it was actually used. ANDY: Thanks. So I need to apologize to Greg
Lloyd, who did a 16 millimeter movie. I had completely
forgotten about that. And this is from that lost– my apologies–
movie that he made. You’ll hear from him
in just a moment. So where was this all used? Ted did an incredible design
on a patent database, electro plating patents. Everything is intertwingled,
and that was our demo vehicle. We showed it at Time
Life, AP, New York Times. And we said, this is the future. And they said, wow, really? And we predicted that
people like journalists would be sitting and
creating and editing their stories online. And they said, sure, sure, kid. In about 20 years maybe. It happened way
sooner than that, and Joe Strandberg was
one of the people who made that happen. Another thing we did
is we contributed to the shareware library. And Norm, incredible Norm,
found the actual bloody source on the web. And you can’t quite
read it, but that is the code, the assembly
language code for branching. That’s the header comment. And there is the nicely
documented, nicely formatted assembly
language code. Here is stuff he
found reporting bugs. Notice the PSWs. Yeah, that’s my handwriting
on top and Ted’s there. I skipped– I’m going– haven’t quite mastered
the user interface. One of the really
cool things is, because it was distributed
on the type 4 library tape, Fred Brooks at UNC did a
color terminal version of it, and their department apparently
used this for multiple years. Years later, I got a thank
you from a NASA administer saying they used the
system after IBM sold it– this was freely contributed– to produce Apollo documentation. And apparently microfilm that
we had typeset the documents for went to the moon,
which is pretty neat. So we proved that, yes, you
could have online hypertext. Yes, you could have
online document creation. I forgot to mention that
in my first year at Brown, I commuted once a month to
Philadelphia, Pennsylvania by plane with
carrying schlepping two decks of these things
with my program on it so I could get half an hour on
the computer at 2:00 o’clock in the morning. I had a personal
computer, my mainframe. Compare and contrast
to this idea, walking in and being online
and expressing yourself in arbitrary ways. That was amazing. IBM had never seen
anything like this. They instantly took it and
helped us trying to sell this, but we were a little
ahead of ourselves. We did get a lot of press
and one memorable TV show, and Greg Lloyd, who’s
also a master archivist, has been part of
all five decades, going to tell you about it. GREG: Thank you. So I– there we go. I started my CS career at
Brown as a hypertext schlep in September of 1968, two weeks
into the AM 101-103 semester and was immediately put to work. In October of the following
year, Andy, Beth [INAUDIBLE],, Al [INAUDIBLE],,
Chris Brown, and I were starring in a WGBH
live TV show, 30 minutes, broadcast at 7:30
in the evening, whose premise was after dinner,
a professor and students talk in a ’50s Ozzie and
Harriet style living room actually adjoining Julia Child’s
WGBH kitchen and the PBS News Hour set. So I won’t read the slide. It was loads of fun, but
the heart of the discussion was why we all cared
about hypertext and why we thought it was
going to be important. We saw great prospects for news
and newspapers and journalism, great prospects for the
arts, literature, music, great prospects for a
platform for building new classes of applications– and my particular hobby horse– blazing trails across
a wealth of knowledge. This is all based on the
example of Brown’s simple– relatively– hypertext editing
system, which we all believed can and would change the world. That’s what Ted
certainly wanted. That’s what intrigued Andy. That’s why he did it. And that vision and belief is
why I believe that HES actually made a difference. So from that screenshot
of the after dinner TV show to this photo,
which I used to print that photo for the
after dinner TV show, to this original print that
was Andy was pointing to in that show, in that video. Thank you, Andy. ANDY: Thanks, Greg. [APPLAUSE] Did I eat the pointer or
does somebody have it? Great. SPEAKER 4: We’re going
to the next topic, right? ANDY: Yes, we are. Just in case you didn’t
realize it, Karen is our gong. She’s been very kind so far,
but I want to stay with her. So we had, I think,
some early influence, but we realized this 2250
thing, which costs something like $150,000 in 1967 money–
it’d be a million bucks just for the display today. That’s never going to work. We need a multi
terminal version. And there was a lot
of functionality we never got to implement. So let’s build the next system. And that became FRESS. FRESS is an acronym I
coined from the Yiddish, to eat ravenously, because
it was a memory hog. It used an entire 128k,
which was a quarter of the memory of the mainframe. We ran a complete time-sharing
system in that 128k. So not too shabby, but
still, it was a memory hog. And as you can see, we wanted
to support a bunch of stuff. So I’m going to set the
stage, but David Durand, whom you will see
in just a moment, is going to show you an actual
demo of 50 to 40-year-old code. He got it to run. He’ll tell you how
he managed that feat of most impressive magic. So what do we want to do? Well, a little bit of history– we developed HES essentially
completely independently of NLS. I didn’t know about
Engelbart and NLS. Hadn’t properly done my homework
because he had published documents which were available. Nothing was online
in those days. I just was ignorant. In any case, because I
was running a session at the Fall joint computer
conference in December of ’68, I happened in on this demo. And it was the most amazing
demo I have ever seen, and I predict it’s the most
amazing demo that anybody will ever have seen. They spent, today’s
money, something like a million dollars. It was a complete rock musical,
whatever you want to call it. It had an army of people. They threw a boatload
of money at it. It was professionally produced. It was truly mind-blowing
in the literal sense, and I staggered
when I saw what they were doing, which was way
ahead of what we were doing. So I came back and
I said, guys, we are working on this new system. I don’t think I coined the
name even by that time. Whatever we were thinking,
it’s off the table. We need to steal the best ideas. So FRESS is an amalgam of
what I learned about NLS. Doug very kindly invited me
to spend a couple of days in his lab really
learning about the system. I came back with lots of
ideas and that is FRESS. What we learned from HES, what
I learned from Doug, and NLS. And here are some things. We still believe in a Nelson
notion of text arbitrarily long strings. Don’t break it up in statements. And we added device independence
on a variety of terminals, which Doug didn’t have for NLS. And so see it as the
next generation system. Like NLS, it had a
prefix command language, so you said MV for
move and then specified the scope of the string that you
wanted to move as the source, where you wanted to move
it to as the destination. You could do it with a light pen
or you could type text strings. LP, light pen, or
location pointer. We had various
kinds of hypertext. Dave will show them to you. We did introduce NLS’s hierarchy
for people who wanted it. We had sophisticated
viewing controls, which David will talk about. So I think the best
thing to do is, instead of talking about this
stuff, actually demoing it. And so with great pleasure, I
turn it over to David Durand. [APPLAUSE] DAVID: I don’t need that. ANDY: 50-year-old code. DAVID: I got nothing
up my sleeve. Hang on. So unlike Norm, I could not
go on eBay and buy a 360. [LAUGHTER] Or even a 370, which
was a little later. So I’m going to start by saying
that we have done FRESS demos before and way post FRESS. So in 1989, we did one at
the hypertext conference, and in 1996, we did
it again because we knew Brown’s mainframe
was about to go away. And we had been able to
run FRESS continuously on the IBM hardware
because IBM really likes backward compatibility. But we figured at that
point it was gone. I did these demos with
Steve DeRose, who’s here. Steve wrote a
beautiful FRESS file exporter that turns it into
SGML, so we actually have– and I’ve kept that up to date. Syd Bauman helped out
by actually spinning the backup tapes that had been
made by the computer science department when the
mainframe was going away. So I had these bits in my desk. And then at some point, I found
out about a marvelous program called Hercules, which lets
you run a 370 in your computer. You notice from the
little bars up there, it’s not really taking
that much of my laptop– although it’s not
doing anything. That’s the operator’s
console, so you know, I could finally have
control over that computer. Step back a second. My background, since I
did not work on FRESS, why am I here doing this? I was– my microphone is down– I was a faculty brat
and the first one to have a computer account. So I started using the
computer at Brown in 1972, so FRESS actually became a
part of my high school career because I used it
to do my papers. So the way FRESS worked was
it was a satellite computer. We’re looking at– we’re
about to be looking at– the virtual MLAG, and
when I’m using the mouse and so forth, this is
replacing those light pen hits when you would pick
stuff on the screen. And when I drag, that actually
sends two light pen hits. And the way FRESS worked
is you typed the command and then it built the
command up in the buffer so you can change your mind
and not submit the command. And then you hit
character return. So I’m now going to
start typing and doing some FRESS things for you. I’m going to start off
with the leach file. This was a textbook,
a linguistic guide to English poetry. It was one of the many resources
available to the students in the poetry course,
which will be talked about in the next session. Joe Strandberg typed in this
entire book from, I believe, a Penguin. You’ll tell us. And of course, you
could scroll around. There’s a little italics here
on the page, so we had fonts. FRESS supported Greek
and special symbols for bibliography. ANDY: David, just
mention Tyler’s emulator. They’re not looking at FRESS. DAVID: Yes. So one person that I didn’t
mention is that this emulator– I wrote the first two emulators
for the first two demos. I was not going to write
another one for the Engelbart Symposium. Andy offered Tyler
[INAUDIBLE] help as a student. He wrote this really
nice emulator, and working
together– you know, I helped him with specs and
fixing bugs and so forth. And we even managed to get
the light pen hits to work, which I had never done before. There will be some demo
opportunities later if you want to see more. Computers were slow, as Norm
and Andy have reminded us, so you couldn’t scroll
through a 350 page document. You couldn’t even search
the whole document unless you used the slow search
that would take a minute. So I’m going to do Get Label,
a feature that was also in HES. And I know that there is
an index in this book, and when I Get Label
to the index, bam, this is the entire index
of the print book converted to hypertext links–
again, by Joe, who was one of the master FRESS users. One of the great
things about FRESS was how dynamic all
the displays were, and we’re going to see a
couple of examples of that. But here just quickly I’m going
to switch to edit view specs. So you can see all
the codes he had to type to get the indentation
and everything to look nice. I’m going to run
a command macro. FRESS let you make
your own commands out of little FRESS commands. You could make your own. This MLAG command just
resets the view specs to a good browsing situation. You’re only seeing one
window, so what’s with that? Well, you had to use the Set
Window command to pick what shapes of windows you wanted. There were seven
different shapes, up to four windows at a time. So now I’m just going to do the
light pen click here for a jump and follow to the chapter
on the irrational in poetry. You may have noticed
these numbers, 15. There were chapter numbers. These were decimal
labeled blocks that automatically tracked
and it could be 2.3.5.2. In this section on the
irrational, we see %T, 6.1. % told you you had hypertext. That’s a decimal label
tag to section 6.1. And if I rearrange the
sections, the tags all update. So you always had up
to date references. I’m now going to switch to
a slightly different setup and I’m going to show a
little bit about the editing. And I think I can stick– I’m going to go to two
window layout here. And I’m going to
get the file test. I’m going to get
it into window two. You could select which
window would be affected. Get file. Oh, sorry demo. Put it into window two. Just a quick double check. All right. We’re going to take a moment
while the virtual machine reboots behind the curtain. [LAUGHTER] It is a real demo, yes. This is now proof. ANDY: We have to shorten, David. DAVID: I know. We’re getting a little late. ANDY: We probably
want to skip editing. DAVID: Yeah. All right. It’s too bad. So the thing that I was
going to show, which I will– oh, I hope it didn’t crash again
while I was in the background. No. I think it’s just my bad typing. The thing that I was
going to show was there was a structure space,
so a special view that showed you all the
links and all the blocks. If you edited in
structure space, it also edited the tech space. So I was going to
rearrange sections. You can see it afterwards
on the official demo. I want to show a little bit of
the experience of the poetry class, so I’m going
to start with Eliot. And the way this
worked was there were several different
kinds of assignments. Now wait a minute, that
doesn’t look like T.S. Eliot. FRESS had, and I
think this is one of the most important
features of FRESS, a thing called display keywords. So you could make
a Boolean request of keywords that turned
the visibility of things on and off. And this is what Joe Strandberg
was the master of configuring. And so I’m going to
use the View All macro and I’m going to get
the instructor TA view of this poetry file. So the first thing, each poem
had three different passes through it. I’m going to use the label. I know that the responses
to the first item on Eliot are in resp1. And here the instructor can see
all of the different student comments. Kate does not want to click on
her comment, but I’m going to– hit the wrong one. It shows up in the next window. It’s the response
of KMA to Eliot. You go down, she
says what she does, then we get a comment from
the instructor saying, I agree that the opening second
person pronoun calls for. This was the experience
in the class. They came in. They interacted. They interacted with each
other and the professor. I believe they wrote a lot. And the final one that
I just want to show, and we don’t have to do
anything more with it, is this is the directory
file of all the resources that were available–
every poem indexed. This was one file. It was linking to 10
other files on the disk. So even though the file
didn’t have a file size limit, people tended practically
to want to organize it. So this gives you a sense
of how much was available. ANDY: Just take
one minute and talk about your dad’s use of it. DAVID: Yep. Sorry. So hypertext was a family
affair in the Durand household and computing was generally– so I’m just going to switch
to a single window view. So my father was
associate provost for faculty for a while
working in University Hall. This, again, is a file that
has a lot of view specs. It was also done
by Joe Strandberg. So I just set a
keyword display string. I said anything that’s
tagged with the tag all. And we see all of
the faculty here. So that’s nice. If I switch to the
Edit View for a second, you can see that there
is a lot of coding here and all of these things– visiting assistant
professor– the secretaries used to have to type that
every single time they did the roster. There was a code here. They could just say it’s a
VRA, it’s a visiting research associate, and it all came out. If we look at the
structure space view, we see that everything is
keyworded with blocks– gender, age, retirement
status, department. So they use this as
kind of a database before there was a database. And with that, I’ll
just say that the best thing about doing this demo
has been how much fun it is to play with FRESS. And so come take a look and
I’ll show you some stuff. [APPLAUSE] ANDY: Those of you went
through applied math in the early days will
recognize many of those. Pardon? SPEAKER 5: You tried
to erase it before. [LAUGHTER] ANDY: We need to take
you off the screen. Yeah. So you heard a little bit of
a sneak preview of the poetry experiment. And I’m going to
tell you about that and then I’m gonna
have the people who really did it speak to it. Before that, let’s just say it
was used in a number of sites, unlike HES. Carol is going to talk about its
use in a course in the physics department. And there were actually
commercial spin-offs. Craig Mathias worked on the one
for Raytheon called Ray Text. For a while, National CSS, which
was the time-sharing service startup that came out of
the Cambridge Scientific Center that built CPCMS
created an office in Providence and we were going to be
their online documentation and hypermedia provider,
and we had a 4,800 bought private line
to the huge mainframe running CPCMS in
Norwalk, Connecticut. It was divine until they
started going south financially and they laid off 30 of us. But it was great
while it lasted. HES itself was shown to
a large number of people. And I just want you to
look at these names. Licklider, Knuth, and Brooks in
the space of a couple of weeks. Pretty amazing. And we have– we don’t
know who kept these notes. I don’t know we had it
till Norm found this stuff. And we had a lot of visitors. Here is a little bit of FRES. I can’t read it, maybe you– Make Jump. Again, nicely annotated code. And now I want to talk about
one particular use, which is this landmark course taught
in the English department using our system. Because of the earlier
course and the good thoughts that a lot of us put
together in a proposal that Craig Mathias
brought to us today, so you can all look at the
proposal that NEH accepted, they decided to fund us. And this is the control card for
the royal McBee card retrieval system with needles– old technology with which
you can find things. Now this card is
really important because I had a call two
years ago from a program manager at NEH who
said, as you may know and you may have read
about it in Time magazine, we’re celebrating the
50th anniversary for NEH. They featured some of
the projects we funded. Yours was one of them. And I want to read back to you,
in case you don’t remember it, a documentary film about the
project is being produced. Did you guys ever
produce this film? And I said, amazingly
enough, a, we did, and b, I actually have it as a print. So cut to the chase, we
organized the symposium at the University of Maryland. It was the first public
showing of the film. The entire top hierarchy
of NEH came out for it. Director, sub directors,
program managers. They brought the tray with a
bunch of cards and the needle to search it and drop this card. It was a fabulous
celebration, and you’ll get a little piece of
it from the folks who actually did the work. So here are some of the
people who worked on it. That’s Marty’s wife. The KMA that David talked
about whose trails we followed, we had David follow
that particular trail because it was really good,
without knowing who KMA was. Turns out, it’s Marty’s wife. Sorry? ANDY: No, Norm’s wife. Sorry. [LAUGHTER] AUDIENCE: [INAUDIBLE] ANDY: Small editorial
correction there. And she famously said to
him, that hypertext stuff, I knew about that years before
you even heard the term. And so she did. Very nice. So take it away, guys. Who’s up next for the
poetry experiment? Carol? AUDIENCE: The video
is going right now. We just saw a little bit– ANDY: Oh, there you go. You’re going to– who’s first? I’d forgotten. AUDIENCE: Carol is
actually first for slide. SPEAKER 3: Just get the link. CAROL: The mic is what I want. ANDY: Carol is a
distinguished professor of law at the University of Minnesota. CAROL: So you had an
intro to the poetry course and we’re going to spend
most of this time talking about the poetry course, but
I want to back up one stage. There was also a
reference in one of Andy’s slides
to the course done. And he mentioned there was a
course done before the poetry course. Funded by Exxon
Education Foundation, Professor George Seidel,
physics professor, taught an
interdisciplinary course called Man, Energy and
the Environment, which my memory says was a modes
of thought freshman seminar. Could be right. Could be wrong. I don’t remember. Craig says, yes,
so I think it was. And we had– it was for a year. It was ’73, ’74. 12 students. A few of them with computer
background, mostly not. It was the first
effort to put hypertext into teaching in this
format with people who were not computer connected. And the experiment
ran for the year. Just brief conclusions
were it worked. I mean, people used it. The students used it. Early data mining, before
institutional review boards, the report which I
just saw this morning has the names of the
students who were involved and it tells how many
times they clicked and what they clicked on. It’s like, you wouldn’t have
a report in quite that format now. But they enjoyed using it. They tolerated a change
in color of the text in the middle of a semester
and dealt with that. I mean, but these
were students who had never dealt with computers
before at all, many of them. But it was an information
delivery system really, more than anything else. It was not an
interactive environment. It was, but students didn’t
do very much with that, so it was successful. Students followed
links, but it was a passive clicking on something,
going reading something else, using multiple windows. But it didn’t test really the
strength and the complexities that were envisioned
by Andy and others in using hypertext,
which led to the proposal to do the poetry course. SPEAKER 6: So we’ll get into
talking about the poetry course a little bit after we
show you a couple of– we’re gonna use the
word text a lot– a couple of texts that
related to the poetry project, as you probably get that
playing– oh, there we go. Basically the first delivery
system, if you will, for information on this
project was, yes, print media. On the left is the NEH
report, which came out in ’76. On the right is
an article that I was tasked with writing, and
through the vagaries of print production, that
came out in ’79. What we’ve been
talking about a lot is how we really didn’t know
what we were precursors to. But if you look down
at the second column– first full column,
second column– pardon me, first full
paragraph, second column– you will note that we blithely
said that a hypertext is thus best seen as a web of
interconnected materials. And so there we are. There’s that word web. Where it came from, none of
us can actually remember. We’re happy that we even
remember the course. But at any rate, we had a
another delivery system, if you will. And this is the film that
Andy was talking about. It’s about three minutes long. The whole film was somewhere
around 17 minutes long, but this is about
a three minute clip that we’ve trimmed out of it. And you will see all of
us in our youthful glory. It’s a little distressing
for us to watch, but we hope you enjoy it. So let’s see if we
can get this going. [VIDEO PLAYBACK] SPEAKER 5: Using
the system, and we hope we can take some
of the fear of poetry out of the student and
also that we can solve some of our own problems of
communication and innovation in the classroom, and we’ll be
able to put students in contact with one another
through the system about the poetic materials, and
that instructors and students will also achieve a higher
level of communication and a greater understanding
of the materials involved. CAROL: In preparing for
the computer-based course, the first step is to
gather the raw materials– the primary and secondary texts. The instructor selects
manuscripts and portions of manuscripts that relate
best to the central themes of the course, then the
material is organized, giving the initial structure. [LAUGHTER] The poetry course
material is broken into 12 units, each
of which concentrates on a particular poem. The students will have three
sessions with each unit. At the first session, they
will view only the poem which is central to the unit. At the second session,
they will view material to place the poem in context. This material includes
biographical and genre information, other
poems by the same poet, and poems by other
poets which influence, or have been influenced,
by the central poem. At the third session, secondary
source material will be added. After reading some or all
the material in each session, the student will add
his or her comments to the information available
at the next session. The instructors and
students will all have an opportunity to
add comments and questions to evaluate and critique
each other’s work. Each new arrangement
of a text provides the opportunity for
something greater than simple transmission,
if various texts can be interactively
arranged, juxtaposed, paired with each
other and played off against each other, the process
of reading the text can be made concrete and the text itself
may no longer be isolated, pristine and inviolate, but
manifold and pluralistic. this web of inter-related
text is called a hypertext. The computer system we use to
create and manipulate hypertext is called FRESS. ANDRIES VAN DAM: FRESS, which is
the computer system being used in the poetry course, takes
no particular training in order to operate it
beyond an ability to type. And we use the keyboard
to specify commands to the system, which
set up questions, insert comments, and browse
through the system in a manner to set up dialogues– a dialogue
not only with poetry, but also with professors, with other
students in the course. This dialogue feature can
be used as little or as much as possible at the
student’s own behest, and he can use initials of
himself or other students to browse selectively through
a particular set of comments or a particular
set of questions. And furthermore what
is really exciting is he can blaze a new trail
by looking at themes or images in a set of poems and
entering his thoughts into the hypertext for others. This kind of selective
browsing and adding to the creative graffiti,
you might call it, we hope to use to develop a
different kind of community with professors and students
working together to develop a true process of learning. [END PLAYBACK] [APPLAUSE] JIM: So, yes, that would be me. And that’s Josiah
and the other person would be Carol, for those of you
who didn’t know who Carol was. I mean, it’s pretty obvious I
think from everything that’s gone before and also I
guess from our youthfulness that you mentioned that
the Exxon project was more of a delivery system. And we never did see this
project as a delivery system. FRESS was not, in our
mind, a delivery system, and English 16 was
not a poetry course. We were thinking much larger,
perhaps overblown thoughts. But effectively
we felt that what we were doing was
we were we we’re developing a
philosophy of a concept that, for lack of a better
term, we would call textuality. And textuality itself in
the building of the project that Joe was deeply involved
in it, what we were doing, we were modeling in our own
minds a form of understanding. That’s how far afield
we had sort of gone without thinking about this. And in terms of the
project, we were also modeling a form of pedagogy. And then when we actually put
the course together and did it, we were performing– for lack of better terms, we
were performing an epistemology and we were
performing a pedagogy. It was a process that we felt
we were actually engaged in. It might have been overblown. I don’t know. But that’s how we
felt at the time. I’ll give you a
quick example of it, and then Carol and
Josiah have some time that they want to speak to you. Example of it was
that we were talking– we’ve been talking
about hypertext and how this is all a web
of interconnected texts, but interestingly
enough, when we designed the course the
first thing a student would see in the first session
is only the bare poem. Why we made that decision
is lost in the ether. But the next time,
they would come in and the poem would basically
light up like a Christmas tree. You would have the structure,
the poem, the tree, and then all of a sudden, bing,
you turn the lights on and you have all the
links that students had built while they’re
reading in their first session. If the tree metaphor
doesn’t work for you, then it was like a night sky. You’re in the city and
you can see the moon because it’s bright enough. Get out in the country,
all of a sudden all the stars are
appearing as well. Those are two
metaphors, which I think are appropriate for
a course in poetry. But I’m going to stop
now and I’ll hand it over to people who are working with
the machine side, if you will. JOSIAH: So one thing– so my involvement at the
time, I was a graduate student in philosophy. My advisor was Roger
Chisholm, and Chisholm had the Andrew W.
Mellon chair, which was the senior chair in humanities. And it gave him the right to
have a secretary part time, and I ended up being
the secretary part time. I was nearing the
end of my studies. So through finagling by
Andy and Professor Chisholm, they got money from me
and Chisholm said to me, you can use as much
money as you want, but you’ve got to teach
me how to use the system. And I did. And he became–
and we’ll discuss that a little bit– but he
became phenomenally productive and said later that he did two
years of work in six months. And he finished
two books, which he wouldn’t have done otherwise,
and numerous articles and publications. I was involved in a
lot of that stuff. So we did a final
report about NEH project and there were some
conclusions that we got. One of them was the phenomenal
productivity of the students. I recognize this wasn’t
strictly hypertext because we were competing with
typewriters and pens and paper. And so just basic text editing
gave people a lot of power. A Typical Course would have– I guess there were 12 sessions,
so you would expect 24 pages. The average number of
pages in this course was 84, ranging from 50
something to a one student did 137 pages of comments and
text and thoughts and papers. So the other thing that struck
me was the technology that was so great to us was really
very primitive from today’s standards. We had one terminal that
the students accessed. And so we had to allocate hours. Each person had
two hour segments, which we had to sign up for
and plan my role at that time. Well, I’ve already– David mentioned that I
input a lot of the material. We’re very proud
of the fact that we had 1,000 pages of material
that the students had access to. And it was very useful. But all that material– you know, the
cutting and pasting we just saw– all that ended
up with somebody typing that material into the system. I did have the joy of
inputting the leech book. So just access to material,
that was sort of crude– that we had one
terminal, and right now we take for granted all access. And the third thing, which maybe
somebody else could talk about, was the community that
we developed offline. Do you want to talk about that? CAROL: So we’re
running out of time, so we’ll say some things here. Happy to talk about
it, any of us, afterwards and through
the rest of the day. So just a little bit more
about the technical thing. So one computer, one room,
two hours slots, and it had to be across the street
from the computer lab so that the text would
come in fast enough. It was 1,200 baud. It was 120 characters. It was faster than
reading speed. That was terrific. But by today’s standards,
I mean, the slowness of it was amazing. So I got the numbers of this
looking at the final report. The 360 through the
semester crashed 123 times. [LAUGHTER] And Jim said earlier, and
every time that happened, he called me. So I’m the technical person. And FRESS had some issues. We had to build new things. So the fact that it
worked as well as it did and that the students were as
successful in their work, given the technical crudity,
is quite amazing. But on the other side of
it, I think for the years since, in terms of takeaways,
I have said for years– and I think I will
continue to say, although the rest
of the day and some of the stuff in the
afternoon may show me things I don’t know were going on– I’m not in computers anymore. I went to law school. I’m a law teacher, but I use
computer systems in teaching. But it is so still crude in
terms of the conceptions of how to use these tools. Finally, I mean, we started– you know, it’s still
very sequential. It’s still very delivery,
because whether you’re talking about traditional
computer-assisted instruction, giving people information,
testing them, letting them follow links like that,
or what we’re doing now in sort of classroom
teaching– interactive teaching in the classroom, cooperative
learning– we haven’t yet figured out a way I
think, by and large, to use these kinds of tools. I think that this was still–
the poetry project was still incredibly innovative and better
than most uses of computers in teaching. Today in the classroom, we need
more information spread out so that people can
use new systems. We had a couple of
slides that just showed using Moodle, which was
Jim’s, and using Canvas, which finally we’re progressing to
at the University of Minnesota, which gives some tools,
but we haven’t figured out how to use them yet. And my only– I have two regrets in
looking back at the project. One is that I didn’t take the
poetry course, because I still don’t get poetry and
I think if I took the course I would
have learned more, and the other is
that I didn’t stay. I sort of left this
behind and went on to law and now belatedly
realized that I should have kept my mind on
how to use these kinds of tools on the teaching side. But I think we can
still get there. ANDY: You want to
finish up, Jim? Do you want to make
any concluding remarks, as this is scrolling by? JIM: Yeah, it’ll go
around and then you get to hear Orson Welles. Maybe he won’t play– [VIDEO PLAYBACK] SPEAKER 6: Did intuition
also tell you about my wife? SPEAKER 7: Wife? SPEAKER 6: She was accosted in
the street a little while ago and lead across to some guy
on your side of the border. [END PLAYBACK] JIM: Yeah. There we are. Yeah, I hate to cut out him
laughing, but that was Moodle. As you can tell, LSU
was behind the times. We were still running an
ancient version of Moodle. This is Canvas, which
was Carol’s system– is Carol’s system. We wanted to show you where
we’ve come, and in some ways, how all we’re doing
is reinventing the wheel in lots of instances. But that’s a sense of
what we’re trying to do, and I think we’re
trying to evidence how innovative, we actually–
in looking back– feel that we were
in this project. CAROL: I’ll say one more. I know. So I mentioned the idea
of this being innovative in terms of delivery,
but also that notion of community that Joe
ended and sort of Josiah ended and handed off. There are so many tools now. I mean, there’s track change
documents and Google Docs and creating wikis together. And we have still
not harnessed that for creating the
online community. It happens in pieces. I think it happens for many
of us with our colleagues, but trying to figure
out how to have it happen with our students–
one of the ideas mentioned in the final report, which
when I read it, I thought, why didn’t we think of this? Why don’t do this? Is students, in order
to use this tool to really talk with
each other, they need to create their own
hypertext first and then share it with others
and then they’ll start to use the system
more effectively. And I think that’s an idea
that we can bring forward so that we can take what we
learned from this project and really colonize
the university. [LAUGHTER] ANDY: Thanks. [APPLAUSE] The poetry course was
the world’s first example of an online
scholarly community. There’s no doubt about that– even though they weren’t
online simultaneously. All right, Norm. Back to you. NORMAN MEYROWITZ: Well, I’m
gonna go through this quickly. The late ’70s, because we’re
switching now to the late– am I on? To the late ’70s. If you went to school,
that’s what you looked like. Your calculator was this, and
if you were a regular person it cost $355 in today’s dollars. If you were a cool
computer scientist, you got one that did HEX
and you spent $15 more. Your word processor
was a Smith Corona, and if you were cool like me,
you had the Report Deluxe. And then that was the state of
the world and the next thing we’re going to hear
about is Steve Feiner. That’s a picture of
the graphics lab, and while people were using
Smith Corona typewriters, the graphics group
was using a ram tack, which was a very high end
bitmap color graphics display. So with that–
Steve can’t be here because he had
another engagement, but I believe he’s on Zoom. Steve? STEVE: Hello? NORMAN MEYROWITZ:
There you go, Steve. You’re on. STEVE: Good. So you can actually
hear me as well, right? NORMAN MEYROWITZ: Yes. STEVE: That’s great. So I guess it’s
my turn to start. So what you’re
looking at over here is a printed
screenshot of a page from the interactive graphical
document system, which is being shown in context of
a beautiful plastic mockup that Andy had commissioned to
show the maintenance and repair workstation of the future in a
picture courtesy of Greg Lloyd. NORMAN MEYROWITZ: I’m just
showing it off, Steve. STEVE: Can you see it? NORMAN MEYROWITZ: Oh, yes. 1982. [LAUGHTER] STEVE: Yes. So let’s move a little bit. So what’s the
paradigm behind IGD? The idea was to move from
the text with some graphics of the systems that you’ve
just heard about to more graphic-centric,
graphics with some text– to also move from the scrolling
systems, like HES and FRESS, to ones in which hypertext was
linking together screen size laid out in advance pages that
contained pictures, buttons that could be
pressed, and actions that were invoked by displaying
a picture, pressing a button, or even entering a page
in the directed graph. So those pages were embedded
in the chapter hierarchy. They were linked together
in a directed graph, and the pictures were
defined as hierarchies of 2D primitives, courtesy of
all of the neat vector graphics that preceded this work and the
standardization efforts that were going on at the time. So the machine, or
rather the system, actually could be thought
of as being a state machine. All the pictures buttons and
links could be keyworded, and if their keywords
were matching those in a keyword tool,
that would be updated by those actions that
I just mentioned, then they would actually
be active or displayed. So the domain behind
most of the work that was getting done over
here in the funding was maintenance and
repair assistance with lots of O and R funding. So what are the key features? The document structure, all
the page contents, the chapter hierarchy, the directed
graph of links, all of that was created with
a graphical editor so you could actually
show and point, draw on the screen, a link
from one place to another. And that would actually
create the link in the system. All of this was maintained in
the mighty ARIES relational database system,
courtesy of Steve Rice. Among the things that IGD
could do is sketchy layout. You could sketch in a
placeholder picture. You could draw a so-called
fuzzy link saying, I’d like to have a link from
this chapter to that page without really going
into the details. All of those things would
be maintained the database and could later be refined with
real pictures and real links. We also had automatic generation
to certain kinds of pages that you could show when
you were presenting. And I will show you some
examples of those very shortly. Another thing that
we’re very proud of is the idea that we
could selectively control the level of
detail that was displayed for pages, for the
chapter hierarchy, for the links on the
part of the person actually doing the
editing of the document. And the idea was to address
the kind of disorientation and clutter in large graphs that
people would later talk about as being lost in
hypertext problem. So to give you a
high level example of what you’re going to
see when you actually look at the screenshots,
here we have electric pages, or representations of them,
linked together with links. Already, there’s not a lot
of pages, not a lot of links, but it’s looking
kind of complicated. So we can embed these within
this chapter hierarchy. And then with the
chapter hierarchy, we can play some tricks. We can look at a
link, for example, follow its parentage
up until on both ends of the link we ended
up with parents that had the same ancestor and
then only draw the links between that set of parents. If we did that, a lot of
those links disappear. And these links are now
more like metal links, showing that something in
this links to something in that chapter, for example. Of course, when
we’ve got chapters, we can then do some
suppression so we don’t see the things inside
of a particular chapter or the pictures in a page. And of course, we can select
one of those chapters or pages and make it take up
the whole screen, or as we’ll see in a
window-based document layout system, the entirety
of a window. This is literally scanned
from our transactions on graphics volume
one number one paper. It shows the architecture
at the top level. Three components,
picture layout system for making pictures, document
layout system for making documents, document
presentation system for actually presenting
those documents. All the pictures were
maintained in a modeling system, so the pictures
where hierarchical, and as I said
before, were composed of individual primitives. And all this is running on
top of a graphics package that was roughly designed to be
like some of the standards that were being created at the time. And the document layout
and presentation systems both relied on all the structure
of the document maintained in a database system. So this is a cast of thousands–
well, maybe not quite thousands– but software
developed by a bunch of folks, documents created by a bunch
of folks, our Dutch uncle overseeing everything
and making sure that stuff is going to work,
and funding coming in from 0 and R and NSS. A page from– or two pages
from one of those documents– about the system that we wrote
back in 1981, ’82, and a page from some source code. Unfortunately, this wasn’t
from the core of the system but rather from an animation
language I’d written, which was used for some of
those actions that I mentioned. Here was the mighty
Ramtek, which still lives in the Brown museum right now. And I’m not sure the
status of the VAX. To be fair, this was–
this was Nancy, not Sluggo, so I don’t
have the actual VAX, but I have it’s sister. So now for a demo. First I’m going to show you
a bunch of screenshots, then because of the low quality
video that we’ll have over Zoom, we’re going to
switch to actually having the video that was
recorded that is played live. So these are all screenshots
captured directly digitally, so these will even look better
than they did back then. First image over here,
which I hope you’re seeing, is a page from the document
presentation system, from obviously a
maintenance repair manual. Pictures at the top, buttons
at the bottom, a little of where you are, a list
at the bottom showing you the name of the page,
chapters all the way up to the top level, and the time– unfortunately not with
the year actually showing. So we’ll now go from the
document presentation system to the document layout system. And here you’re seeing a set
of non-overlapping windows. Each contains a
black-bordered chapter, like the one at the lower
right or white-bordered pages. The one at the
lower right actually is the entirety of
this particular manual with a lot of detail removed. So you can see at
this level, this is basically a
three element graph of chapters, fully connected. Every one of the chapters
connects to and from every one of the other ones. At this point, we’re going
inside to show you some detail. And so you’ll see that
little cursor in the lower left of that big window, and
when we press the plus detail button a couple of times, you’re
now seeing lots of extra detail inside. And so now we’re going to move
down into one of the chapters. And so that cursor’s
on top of it. We go down into it, and
it takes up that window. And at this point, we’ll
actually make a button. So this kind of button is going
to be a upright rectangle. So we’re going to determine
two points by clicking along one of its diagonals. And you’re seeing the little
help documentation at the lower right hand corner there. Choose the first point. So we choose two
points, and then through the miracle
of color lookup table animation with the 8-bits deep
worth of display that we had, we dim the screen, put up a set
of nice bright white lit names of the different
kinds of actions. We’re going to choose
a get new page action. We’re gonna go and select
the page from another list or from what’s on the
screen to be able to show the destination of that link. And then when we do that,
we see a little arrow going from that page, that
white-bordered page on the right, to–
turns out– the chapter that’s inside of the
destination that is. And now at this point, you might
want to know a little bit more where that really is. And instead of using this kind
of detail suppressed version, we can say, gee, we really
want to see what this is linked to, even if the arrows are going
to go across chapter boundaries and across the
window boundaries. And when we do that, you see two
links coming out of that page. And now we just made one
link, but it turns out that those two pages and the
two little windows at the left are actually the
same page, but being seen in context of
different keywords. So because of the
difference in the keywords, that page now shows
different pictures. So let’s see. At this point, we
decide we would like to go up in that
window at the lower right to see the entirety of that page
that the link is coming from. And now it occupies the
entirety of that window. And now we might
want to see what about the parents of this thing? Let’s actually make this
page go out of window mode into the full screen mode. And when we do
that, we are now– because we also did
some other stuff– we decide we want to go
up within the hierarchy. The screen dims again. We’re now seeing the
complete set of parents of that page on the right. The little arrow is next to the
one that says reference manual. That’s the very top level. When we select that, we go
up to the very top level of the document with that
added amount of information. So very quickly, I
will show you a couple of views of the
automatically generated pages that I mentioned. We go from document layout
to document presentation. Now you’re seeing the links
page, whatever the current page was when we were actually
traversing the documents, is going to be at the center
in its first level of chapter. At the left, you’ll see all of
the pages that reference it. If there were more,
you’d see a little arrow at the bottom that let’s
you scroll through the list. Scroll is really aspirational. It would very chunkily wipe
the left part of the page and display the other pages. And on the right, you see
all the pages you can go to. If I select that page at the
bottom of the column of pages at the right, it
becomes at the center and then we can see the
pages you went to and from, and the ones you could go to. And now he can look at our
timeline page, where you’re seeing where if you
went into this from, the history of what you
did during that session, timestamps across
the bottom, each page inside of its nesting
chapter going back to the beginning of the session. And then finally, I’ll
show you our index page, in which you can
type in key phrase and you’ll see all of the
pages in the document sorted by their containing chapter,
all little miniatures drawn courtesy of that modeling
system on the screen. ANDY: Steve, can
we go to the demo. STEVE: At this point, we’re
about to go to the demo. So you’re going to go and
put it up locally over there and we’re gonna let
Nicole take over. One quick thing to
mention when you start is that any point to which
you see the screen smoothly cross fade, any point
to which you see a cut, all that is done in the video
camera and video editing, and any place where
it slowly gets drawn and you see rectangles
slowly drawn, that’s the speed at
which the Ramtek went. Two seconds to draw
a single rectangle the size of the full screen. So take it away, Nicole. [VIDEO PLAYBACK] NICOLE: We have designed
a prototype document as an example of a system that
might run on our maintenance repair kit of the future. This is the first page of
an electronic maintenance and repair manual
for a sonar system. When I log onto the system, I am
identified as an expert repair technician. Another user might be classified
as a novice or an intermediate. A user status dictates the– ANDY: Oops. Hold on. My fault. NICOLE: We have designed
a prototype document as an example of a system that
might run on our maintenance or repair kit of the future. This is the first page of
an electronic maintenance and repair manual
for a sonar system. When I log onto the system, I am
identified as an expert repair technician. Another user might be classified
as a novice or an intermediate. A user status dictates
the path he or she travels in this dynamic
non-sequential document. This next page
shows the location of the ship and
the sonar cabinets. We are actually looking
at a conglomerate of pictures drawn using our
graphical editor, the picture layout system. Using one of the buttons
in the menu at the bottom, we can step through the
document page by page. This next page gives me a set
of options to choose from. In this case, we
assume that I have been assigned to fix a problem
with the sonar display. I will opt to run the diagnostic
test built into the sonar equipment by picking the test
button in the main picture area. Before I can actually
run the test, however, I have to
perform a setup procedure. I am told how to connect
the maintenance and repair kit to the sonar cabinet. I will now go on
to run the test, but notice that I could
have asked for extra help if I needed more information
or could have used the back button, a miniature
of the previous page, to review the information
I have already seen. To run the test, I pick
the start test button. You’ll notice that
we use animation to indicate the passage of time. [LAUGHTER] The built in diagnostic
system indicates to us that there are problems
with the system hardware. This page tells me how to
proceed to run further tests and how to fix the problem. If I now want to
travel non-sequentially through the document, I can
pick a number of options. If I hit the manual
button, I’m given a group of non-sequential
travel facilities. The timeline button
will display my progress through the document,
providing me with an easy way to go back in time
to any of the pages I’ve already seen
during the session. The date and time
I visited each page are indicated just below
the miniatures of the pages. [END PLAYBACK] ANDY: Steve, are you still on? I guess we lost him. I just want to
mention about Steve– CAROL: He’s here. STEVE: Still there. ANDY: Oh, you’re still there. STEVE Yes. You can hear me, right? ANDY: In the world of
maintenance and repair. He’s a distinguished professor
at University of Columbia– Columbia University, rather. STEVE: Yes. [LAUGHTER] ANDY: Very different. Sorry about that. And he’s a world
expert in AR and VR. Thanks very much
for helping us out. [APPLAUSE] STEVE: Thank you. Very welcome. Yes, I wish I could be there. Very frustrating
not being there. Goodbye. ANDY: Bye, bye. Back to you, Norm. NORMAN MEYROWITZ: Yeah so now
we waited till the end to put up the slides to congratulate
and honor the people who did all this work. So if you did this
and you’d like, you can stand up and
get your applause. [LAUGHTER] [APPLAUSE] ANDY: Over the past couple
of weeks, thanks to Norm and others, we’ve been doing
this archaeological dig. Part of it is trying
to, in our memories, reconstruct who might have
worked on these projects. We probably left out people. If you remember
them, come up to us so that we can enrich our
history of these projects. So here are the FRESS folks. [APPLAUSE] Here are the poetry folks. [APPLAUSE] And before lunch,
here’s a tribute to those who aren’t
with us anymore. One of the original
11 at Microsoft. Well, it’s nice to
see all those faces, even though they’re not with us. So we’ll give them
a round of applause. [APPLAUSE] Now we have lunch in the lobby. You can eat outside
or in room 115. And in room 115 is some
demos of Intermedia and FRESS and Hypertext Hotel VR.

Leave a Reply

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