Audio: enhanced by Adobe Podcast Enhance
Transcript by Whysper Transcription
Many people think the World Wide Web was my idea.
I want to clear that up.
The web is a dumbed-down of my idea, but I believe I created the web accidentally by two mistakes I made.
We'll be getting to that.
Where to begin a history?
Benzo, where do you want to emphasize?
I'd begin the story of hypertext with the Rosetta Stone about 196 BC, the Talmud after 70 AD, the Hexapla of Origen, Third century.
These are all parallel documents, side by side, connected.
The kind of document not possible yet in today's electronic formats, nearly 2,000 years later.
Jump to around year 1450, Gutenberg reinvents movable type and prints Bible's great impact on religion.
Shortly thereafter, an enterprising Venetian named Aldus Minucius invents the personal book, which moves the Renaissance from palaces to middle-class homes.
Note 500 years to 1945.
Vannevar Bush, that's his name is pronounced Vannevar Bush, President Roosevelt's science advisor, writes an article predicting a machine he calls the Mimics, storing all the world's writings on microfilm, all of this.
He misunderestimated, he he, the words of another Bush, the words of another Bush.
He misunderestimated how many writings there would be, as did everyone.
Vannevar Bush proposed that scholars make trails to the world's knowledge, a new form of publication, each author connecting things and adding content of his own.
Technical aside, Bush's article also profiled transclusion.
1951, West Coast, Doug's epiphany.
Douglas Engelbart is one of the greatest men of all time, a warm-hearted, soft-spoken, and saintly guy who invented much of our modern world.
Doug's name, Engelbart, means "angels' beard," didn't know angels had beards, did you?
Well, Doug read Bush's article in 1945 when he was stationed overseas.
Then when he got his PhD in California in 1951, Doug Engelbart had a great vision.
That interactive computer displays would make possible a new form of collaboration vital for the urgent complex problems of humanity.
And it would be rather like the mimics.
It was all about connection, connections between documents, but also connections between people and the dynamics of collaboration.
Step by step, Doug went through the painful departmental politics and funding to build a team that would change the world.
That was his intention, and that is what he did.
Nowhere near what he hoped to do.
Cut to the East Coast.
In 1960, I, Ted Nilsen, took a computer course and it hit me.
The interactive computer screen will be the new home of humanity, the next great media revolution, and I want to be in it.
A lot of people can't understand how I could figure this out without being an engineer, but that's the reason I could figure it out.
The engineers didn't have a clue.
It took a media person.
I'd been an actor on stage and TV.
I'd published a little magazine.
I'd produced an LP.
I'd shot a half-hour movie.
I'd seen TV developing from the inside, watching what I went on in my father's control rooms at NBC and CBF, starting when I was 10.
So the technicalities and art of interaction would be the next frontier.
I didn't actually see an interactive screen for five years, but I didn't need to.
I was a movie maker.
I could feel the screen.
In my mind and heart, I saw it and touched it.
I conjured it and caressed it.
It responded wonderfully.
Many people still don't recognize interactive software as an art form.
Many techies treat the interactive screen as a kind of blackboard.
But artists know it as a lumbar.
Here was my new credo.
There would be millions of readers and writers all over the world with their own computers and screens, and they would be able to publish to each other without publishing companies in the middle.
This meant new forms of writing, it meant a whole new form of literature, it would replace paper.
Five years later, when I started publishing on the subject, I called it "hypertext.
" I had not yet heard of Engelbart.
Okay, technical part.
Here was my initial tactical design.
Whatever the user typed in would be immediately saved for safety on tape, I thought at the time.
Each new version would consist of pointers to what was already there on the tape and what you were adding as you continued to write.
The method would be safe and efficient and have other benefits which I gradually figured out.
How to do the links was a separate problem that would take 19 long years to figure out, but I feared that if I didn't design the documents of the future, the techies would screw it up.
Which I believe is exactly what has happened.
What I really wanted was to make movies, but I thought it was more important to be the new Gutenberg.
Little did I know that Gutenberg went bankrupt.
1964-ish, meanwhile in California, step-by-step Doug Engelbart builds his lab.
He gets sponsorship from Bob Taylor, the defense department, and assembles a team at Stanford Research to not stand for you.
In this project, Doug proceeds to invent multiple windows on a screen, screen sharing between collaborators, text editing on screen at the first word crossing, links between texts, and much more.
Technical point.
Doug's links weren't embedded.
The whole of Doug's vision was greater than these parts.
The augmentation of human intellect, especially through collaborative networks of people, constantly improving their networks with methods of collaboration.
Those innumerable parts that I just enumerated of Doug's system were just the hockey sticks for a much greater game of collaboration he hoped to create.
Okay, 1965, East Coast.
After four years of study in design, I submit papers to four conferences and everyone is accepted.
The biggie is the ACM National Conference in, of all places, Pittsburgh.
I think that most of the computer scientists in the world were in that room.
It was possible in those days, and most of them heard my paper.
It was very well received, and I could see that all my work was just it.
The only question confronting me was how to get leverage, convincing somebody that I was right so I could get machines and a team together.
But how could I make people believe I knew what I knew, let alone what I could do about it?
Then somebody came along that appeared to believe me and had the bucks to make it all happen.
A few days after my Pittsburgh talk, I got a call from a guy identifying himself as the Director of Information Processing Research for the Central Intelligence Agency.
He came to my house and asked how I'd like to be funded.
People asked me why would I take money from the CIA?
Hey, our country needs good intelligence, and I needed the money to do good things.
So I began writing proposals for the CIA, which went on for five years, till I gradually realized they were just stringing me along, as most prospective backers do.
1962.
1966.
Meanwhile in California, Douglas Engelbart invents what everybody credits him for.
A mouse.
Oops, it dropped it.
Is that it?
In the excitement it tried to escape.
Again Engelbart invents the mouse, which is a big improvement over the light pen.
But somehow the process of myth-making has dropped the more important stuff from his reputation one liner, leaving only in my dog's overall vision was and is so much more than that.
East Coast, 1967.
The Hess Project, HES.
I am asked to join a project at Brown University that will be a chance to "try out my hypertext ideas" at no sale, right?
It ends up costing me all my savings.
On the first day, I did not get along with the nasty guy who ran the project, but I am not a quitter.
I should have quit.
As I wrote in New Scientist magazine in 2006, I believe that humanity went down the wrong road with that project.
Dumbing down interactive documents through one-way jumps among paper simulations.
A great and fundamental loss to civilization.
The project was called HES, Hypertext Editing System, and it turned out the guy didn't really want hypertext.
What he wanted was a system to prepare paper documents for printing, what we now call word processing.
I consider this unworthy and entirely the wrong path.
You have to choose which will be.
Are you going have real electronic documents or are they going to simulate paper?
That project chose appearance over connection, dumbed down hypertext one-way jumps among paper simulations, and this would later become, alas, the structure of the World Wide Web.
1968, West Coast, by which time there was still no personal computer, Doug Engelbart's great demo at the Fall Joint Computer Conference, where he shows cooperative work by people at separate screens, amounts, hypertext links, collaboration over the phone while pointing at the same text on different screens, and much more.
At this point, computer screens were still considered exotic and crazy by most of the computer world.
Alas, this was Doug's high point, because politics and disagreements took over as his lab gradually fell apart.
Big skip to 1974.
Again on the west coast, Xerox Palo Alto Research Center in Xerox PARC starts hiring people away from Doug Engelbart and his lab collapsed.
I was at dinner with Doug a couple of years ago when he actually wept at the memory of how his people had deserted him and how Bob Taylor, his earlier supporter, betrayed him and the vision he had tried to betray him and the vision he had carried forward so faithfully.
Alas, Doug Engelbart's system never got to the public.
A company called Timeshare tried to productize it but with no real interest in Doug's greater teaching.
What if they put NLS out to a broad public?
What if they productized Doug's software for distribution?
Wilminton.
Now the page turned.
1975.
Personal computing begins with kit computers, soon also modem kits.
1977.
Fully assembled personal computers at last, notably the Apple II.
Now that some people have personal computers and slow over the phone modems, soon there are file sharing methods, built bulletin boards, message boards, forums, discussions, hosting.
The yearning for online publishing by individuals is at last fully revealed.
Though most people, as the American company, still don't get it.
And it's still a small subculture.
East Coast, 1979, the big year for my Xanadu project.
Six of us come together, Roger Marks, Stuart Rowland, Eric and me, to design our worldwide hypertext network.
Note that as of that moment, set in 1979, there are still very few personal computers in the world, and the only interactive screens most people have seen are airline ticket counters.
but we knew what was coming.
My team figures out how to do the links brilliantly, making links of reuse work together.
This is described in my book, Literary Machines.
What it would do.
Deep hypertext, that is, connected systems of documents with any connectivity and complexity.
Every quotation connected to its origin by my method from 1960.
Detailed side-by-side inter comparison with between any two documents of versions, showing which parts are the same and and other connections as well.
Links in profusion, not embedded but overlaid, a system of publication involving unrestricted reuse of content, but sale of the content could be required by the publisher, separate from the document container.
Our plan was to put stabilized content on servers with every character retaining its permanent address, stabilized XANA links on servers individually published, nothing at all like web links.
Finally, Mark Stewart and Roger Market Stewart work out a brilliant unifying addressing scheme with a new kind of mathematics for compactly referring to content and structures throughout this universe and creating new documents and versions as permutations in this space.
The boys start programming and Roger looks for backers.
He and I agree that he can do the technical develop.
I'll flounder the publishing business.
But we're all agreed on certain ideals that would later become popular.
Freedom of publication for everyone everywhere.
Dodging government controls over free publication.
Privacy for everyone that wants it.
We see this as a new sweeping mechanism for democracy, citizen awareness, truth telling, honoring minority views in documentation of the world.
Eight more long years to the West Coast, 1988.
Autodesk backs Zalazak in 1988.
Nine years after, eight or nine years after that what I just told you, my colleague Roger Gregory gets backing for his development company from Autodesk Inc.
and his visionary founder, John Walker.
Roger and I had agreed that he would develop a Xanadu system that would work on interfaces and plan the publishing network.
Unfortunately, that puts us in different departments at Autodesk, a recipe for danger, in business, your enemy, or your suspected savagery is all in the other department.
You may know the story of the appointment in Samara where a man brings on his doom by trying to avoid it.
That happened to me.
I got a call from one of the Xanadu programmers saying Roger was out of control and throwing things, and they were going to quit.
So while I had no official status in Roger's development company, I called John Walker, the head of Autodesk, and told him the situation.
And Roger was relieved and con- [pause] Turned out that Roger was right.
He was throwing these because the team wanted to redesign, which would obviously delay the project six months or more.
With Roger no longer in charge, it was much worse than that.
The program was redesigned for four years, till Autodesk shut the project down.
At which time, a very different animal slouched into our other- our ecological niche.
So what were the two mistakes I made that brought about the World Wide Web?
Mistake one, participating in and endorsing the 1967 S-Project that Brown popularized hypertext links as one-way jumps between simulations of paper.
I thought it would be fixed later.
Famous last words.
Mistake two, urging John Walker in 1988 to demote Roger Gregory from his own Zanadu development company, unwittingly giving up Zanadu's pole position as the world's hypertext system.
What if Roger had not been demoted?
I believe he would have kept his promise to John Walker and delivered the server network and a workable client within the one year he promised, definitely two.
Way ahead of HTML and the Mosaic browsers that made it possible.
Unseating Roger lost us the world, and no one knows what the world lost.
I believe that without these two mistakes, humanity would have a far better tech system.
This is a very heavy burden, but I fight on.
To be continued.
Thank you.
I am Ted Nelson with another disclosure of Computers for Cynics.
Today's undertaking is the real story of the world wide web.
In the preceding discussion of projects added, I skipped over the one-way hybrid systems, one-way general-between-paper simulations based on Hess et al.
1967.
But in the 1980s, there were lots of one-way hybrid systems, off the top of my head, on the West Coast, HyperCard, SuperCard, NoteCards.
On the East Let's go with Cowles from the DoD, the Department of Defense, Fress, Intermedia, the Eastgate School, Hypertext, and New Literary Style.
Based on one way jumps with surprises, making literary hay out of the invisibility of the next item till after you select it.
Again, that's the Hess zone.
Oh yeah, and in Scotland, Guide.
In 1984, the year of the Macintosh, the CIA finally got their hypertext system, 19 years after they started teasing me with plausible backing.
It was called Notecards.
I have this from Kathy Marshall who says she worked on the contract in Xerox PARC, and she says it's not secret, just not well known.
Would my system with the CIA have been better years and years earlier?
Of course I'd think so, but who knows?
1986-7, Neil Larson publishes Houdini, a hypertext system with lots of links to documents out on the net.
His hypertext filers also include HyperRes, PC Hypertext, and Transtech.
1988, as mentioned in the previous talk, the Xanadu group stumbled and alas got out of control with the unseeking of Roger Bricker.
Now to 1989, Europe, the first public initiative for what will become the World Wide Web.
Tim Berners-Lee, a physicist at CERN, the huge European nuclear research lab, circulates a document entitled "Information Management Proposal.
" "There are few fuzzets which take Ted Nelson's idea of a wide docuverse literally.
" Actually, as already mentioned, there were several such products, though the people who made them didn't want to admit it.
But in that proposal, Tim also said, "There seems to be a general consensus about the abstract data model which a hypertext system should use.
" Hey, wrong out of that one.
He didn't ask the senator, guys.
Our abstract model was off the charts a bit, but we didn't care.
Next year, 1990, Tim and Robert Cailliau write an official proposal, a worldwide web proposal for Hypertech, with a capital "D" in the middle, and they get support from CERN.
Next year, 1991, the web starts up locally.
Tim's text format and server software are working and deployed at CERN.
It's a preliminary package of text pages, jump format, page structure, directory structure, but it does not reach any general public.
Two years later, 1993, mosaic.
The web browser as we know it makes history.
1993 is when the World Wide Web really happened.
The World Wide Web as we know it wasn't created by Tim Berners-Lee, but by two university students who Poo-ified the net.
That is, they put the Parc user interface, or Poo-ee, around Tim Berners-Lee's document format, repackaging Tim's phage layout in an application frame with Parc user interface, or Poo-ee, around it.
This created not just jumpable pages, as in Tim's design, a new hybrid, a puppet theater and shop window, now called the web browser and Mosaic was its name.
The Mosaic vision, that is Andrisa's vision, was to create a user platform essentially competing with Windows and the Macintosh, in which not only could text appear, but interactive applications could be presented.
New application cattle pens, not boxed inside your computer, but managed from servers on the net.
Far away.
Andriessen and Bina could easily have chosen to frame, glamorize, and foolify some other format.
Certainly, if they had built the poetic, bless Neil Larson's 1986 Houdini format, the Houdini browser, it wouldn't have been called the Houdini browser.
That would have had the impact of the web.
Arguably, it was sheer chance that Tim's format best suited their purpose.
Without Mosaic, Tim's format probably wouldn't have gotten any further than Larson's.
That's why I think the real creators of the World Wide Web were Andriessen and Bina.
Now who exactly should get credit for Mosaic?
Larry Smarr, who was their boss at the National Center for Supercomputer Applications, tells me I'm supposed to say the NCSA browser, National Center for Supercomputer Applications browser, and credit Bina and Andreesen equally.
However, it's important to point out that Bina was being paid.
Andreesen was doing it on sheer guts and sleeplessness, and many say he should get the main credit.
Probably Larry Smarr too had something to do with the initiative.
So why not call it the NCSA Smar Bina Andreessen Browser, or just PuyoNet?
1994, the impact of Mosaic.
One browser to rule them all, one browser to bind them in the land of Mordor where the shadows lie.
Thousands, powered by the Mosaic, the web's expansion was now following exactly the expansion rate I'd predicted presented.
Thousands, soon hundreds of thousands of users were using this mosaic, and thousands of servers were firing up in companies and apartments, even in secret hiding places.
Meanwhile, on the West Coast, Microsoft jumps in with a version of web browser PUI called Internet Explorer, which gets off the ground fast because Bill licensed his mosaic from Larry Swarmer, technically Spyglass from NCSA.
And also on the West Coast, Mark Andreessen, who has just graduated from college, is taken on as a partner by Jim Clark, just dethroned from presidency of Silicon Graphics.
Clark takes young Andreesen on as co-founder of Netscape Communications, Inc.
For a kid just out of college, to be brought in as equals in a new enterprise by a multi-millionaire shows that Andreesen was a pretty impressive character.
In 1995, Netscape goes public and brings in lots of money.
However, Microsoft gives away Internet Explorer and that kills the Netscape business home.
By 1999 Netscape manages to get acquired by AOL just in time, and Andreessen and Clark managed to get away with their money by the scale of 18.
So these those were the important events of web history as far as I'm concerned.
Younger people would have a different look.
As a free-form standard, the web come around wildly and the public thinks it's "technology" rather than an accidental packaging of a lot of stuff by a bunch of ambition gut.
So whose idea was it?
Who quote "invented the web" unquote?
I'm not in this race, I'm neutral on this issue, my team was doing something entirely different, but I want to see people like Neil Larson get the credit they deserve.
The notion that Tim Berners-Lee quote "invented the web" shows the process of myth-making.
It's like the invention of the telephone.
The same was Alexander Graham Bell who invented the telephone.
Way dumb dumb down the truth.
Many inventors of the telephone don't make it into the STEM history.
Tim Berners-Lee would have won the credit lottery for things a lot of guys were already doing, creating viewers and document formats for jump access among paper-like documents on the net.
But Tim's document format was blessed, nay consecrated and sanctified by Eric, Bina, and Mark Andreessen when they boobified it by putting Tim's text format into a frame on the Internet, the Fooey frame.
Bina and Andreessen could as well have blessed some other Internet text program at that time, like Houdini or Silversmith, or whatever even Microsoft Word as a World Hybrid X format by putting it in a networked POUYA and making it the Houdini browser or the Silversmith browser or whatever.
Did Tim have some transcendent supernatural understanding?
Nobody else did.
He would never make that claim.
To repeat what he said in 1989, he built on what he considered "a general consensus about the abstract data model which a hypertext system should use.
" He put together a clean and professional package around what he saw as a consensual objective, nothing supernatural about that.
What he has undergone in deification process that makes people want to believe in his transcendent supernatural understanding.
I'm sure credit is never claimed.
Credit Lunduli.
He's an extremely decent fellow.
This is just how the myth-making process works.
Where did the parts of Tim's format come from?
Embedded format was from Xerox PARC, among other places, as it moved to the Microsoft Word.
in angle brackets were from Neil Larson's MaxThink in 1984.
SGML, which Tim renamed HTML was from Yuri Rubinsky and it had firstly used in a browser by Bottoms in 1987.
What I think what Tim should get the most credit for is the URL, whereby he harmonized and standardized network addressing a rational organization across the diverse and divergent file structures of Unix, Windows and the Macintosh and indeed the whole world.
A hell of an achievement.
But if people ask me what I think of Tim Berners-Lee, what can I say?
First of all, I likened him.
I would compare Tim Berners-Lee to Ronald Reagan.
Perfect figurehead.
Very nice guy.
Very idealistic.
Upright.
Totally focused.
Totally enclosed.
I've argued with him for hours at his home in Massachusetts, in bars in Tokyo.
One bar, anyway.
with no sense of communication.
But as with Reagan, that could just be a pose of political stance.
He needs to throttle initiatives as much as possible.
He is trying to defend the formats of the browser.
Tim has a very tough political job.
Head of a juggernaut, the World Wide Web Consortium, a political tiger by the tail, with running a system that's getting worse and worse, messier and messier, with lots of politics, people want to add more and more junk to the web, i.
e.
into the web browser.
And he wants to prevent that joke as much as possible, or at least organize it in the most rational way he can.
Despite Tim's best efforts, the web internals get worse and worse, full of special effects and special cases.
The web browser format is still misleadingly called HTML.
So when people say HTML, they may think they're referring in some sense to Tim's original clean design, but more and more has been crammed into the ever-gnarlier and large definition of HTML a tangled, mangled, reprehensible mess.
HTML is now piled high with hierarchy, the document object model, cascading style sheets from [inaudible] for which we can thank Tim Bray, tabs, and probably more hierarchies I don't know about.
Why?
Are documents hierarchical?
Let me paraphrase Einstein, who may or may not have said, everything should be as simple as possible but no sin.
I propose this paraphrase, every document should be as hierarchical as necessary but no more so.
We must not impose false templates in the pretence of orderliness.
Great writing and reporting as in the New York or other key magazines is not hierarchical as it's trying to represent all the connections of the story.
For web formats to impose false hierarchical templates is highly questionable, if not a cultural atrocity.
I have to say something about JavaScript.
Okay, JavaScript is the atrocious language of the web browser.
It got its name for strange political reasons.
It has nothing to do with Java.
I'll just say one thing about JavaScript.
I was afraid that telling this would be violating my non-disclosure agreement with Google, but I cleared this with Peter Norvig.
In men's rooms at the Googleplex, Google headquarters, they post the day's special JavaScript advice over each urinal, Thus reaching the programmer through sensory motor channels that conventional documentation cannot pursue.
Don't expect HTML to get any better.
Just an ever-growing salad of special case effects.
They will add garbage to HTML forever, especially when Tim retires, and don't expect it ever to be as smooth as Flash because the JavaScript will always have to pause at garbage.
Anyway, some negative thoughts about the web from the user's point of view.
To begin with, the web is about appearance rather than structure.
Any structure you wish to represent must be matched into the one-way jumps established in 1967.
The very concept of a website is grotesque and should have nothing to do with darkness.
What I think of as today's typical web page is a flapping, screaming, unprincipled mess with text lines a mile wide that you can't get on the screen all at once because they're locked against rearrangement and they're in pale tiny sans serif type you can't read anyway.
The WYSIWYG model of paper simulation imputes false value to the parts of the rectangle, desperately valued real estate that must be allocated carefully.
Articles are broken into individual paragraphs, embedded one by one in bedsheet-sized pages.
And these isolated individual paragraphs, surrounded by flapping, screaming, jumping animations, are divided into smaller and smaller thoughts, perhaps for the smaller and smaller minds of the Twitter generation.
But hey, make it sit.
And of course, we have the social web with its new cattle pens.
And finally, of course, we have the two new super monopolies, both based on cunning uses of the PUI.
The Facebook Tsunami versus the Google Octopus.
Hey, the word Google was originally a number, so we can call it the Google-pus.
Do you wilking?
First, a brief look at something you can come back to.
[ Silence ]
Ted Nelson Computer for Cynics - the WWW