Web History 101, or How the Future is Unwritten – Benjamin Young – Tech Forum 2017


– [Monique Mongeon] So our next speaker is Benjamin
Young, a web digital publishing and open-source advocate, here today to speak
about publishing and the World Wide Web. Benjamin is a solutions architect at John
Wiley & Sons, where he works on the Web Annotation Working Group and Digital
Publishing Interest Group at W3C. He’s currently exploring the edges of a
re-decentralized web using annotation, distributed identity, and offline-friendly
web apps and extensions. Please welcome Benjamin. – [Benjamin] Hi, everybody. Thanks for
coming. This hopefully will not be the history lesson that you were looking
forward to so you could take that nap. So, apologies if you were here for a nap.
Feel free to take one, I’m not going to stop you. So yeah, Web
History 101. There’s a document in the corner that we’ll be revisiting later. If
I were giving out bonus points, bonus points for knowing about that
already. This is me, I work for Sophia, who many of you already know. And I work
at Wiley. I’m BigBlueHat online. Much easier to find than “Benjamin Young,”
which is apparently in the top 100 most common names on the planet. So
the Web got its start a long time ago, we’re going to start with a reading list.
If you’re curious about how it started and the interesting anecdotes about CERN and
Tim’s computer, the one server that ran the whole thing for a while. This is a
fabulous book that he co-wrote with someone. His content, their writing. You
know how that works. And then there’s several links in here, it looks like I
missed one, that go to history pages from the W3C or from Tim that describe how it
all happened. So if you’re really curious about the interesting individuals involved
in the process, these are great places to start. But this is the overview of the
history. Pre-Web there were people thinking about hypermedia and the ability
to click around and move. Not like you do in a book, maybe like you
do in a Choose Your Own Adventure sort of thing. Vannevar Bush’sAs We May Think
article, which is available online, hence the hyperlink. From the Atlantic
Monthly in 1945, which is not when they put it online. Early by a couple months.
It outlines this beautiful future world where we can access information easily and
move from page to page and find all the content that we needed to find. Some days
I feel like that’s happened, some days I don’t. Douglas Engelbart
another 15 years later, if my math is working, did the mother of all demos where
he demoed this thing called oN-Line System and his newly-invented mouse. This is how
long ago that was. Which originally looked like a block of wood with a pie cutter in
it. And, anyway, it only went sideways initially, and then kept going. But he
also showed hyperlinking and cross-referencing and jumping
around. It’s all available online, you can watch the black and white
graininess, which is pretty fabulous. And then also that year, Ted Nelson coined
the word “hypertext.” How many of you know that word? Or “hypermedia,” which I’ve
already said. Yeah? Right room. We’re good. That hyperlink doesn’t go to
the actual document because you have to pay for it, I think it’s like $8, from
ACM. But that’s the big document in which he talks about a file structure and a
Docuverse that you can cross-reference and jump around. And this is 1960, so he’s
talking about a very webby sort of thing a good 20 years before it starts happening.
So behold the Web. 1980, I was born in 1980. It’s a good chunk of
years. I remember them well. Tim Berners-Lee was at CERN writing
notebook programs. So think Evernote-ish kind of things. Not nearly that graphical,
though. Think something more like that. We’ll come back to that slide in a second.
Where he was recording trees of information about the people at CERN and
giving them a way to organize their documentation, just like you may do on
GitHub or Evernote or anything like that. And it was out of this project, that had a
very complicated name,Enquire Within Upon Everything. Which is
how that Web history book starts out, with him doing this and exploring that and
what he learned from that, and then learned from hypertext
conferences, which were running in the ’80s. I had no idea. And then out of that
came this proposal in 1989, so he was at it for kind of a while. And
then createdInformation Management: AProposal, which is the document that was
in the corner that we’ll take a look at in a bit. And then from that became a program
called WorldWideWeb, all one word. That’s not camel case,
that’s studly caps, I think, right? I don’t know. There’s kebab case
and camel case. It’s important information. Which was the browser. So the
WorldWideWeb was the first World Wide Web browser when it wasn’t worldwide and it
was just on Tim’s desk. And Robert helped Tim at CERN and they
moved on to a second proposal that got practical. The first one was vague but
exciting. That’s what his boss said, anyway. So this is the World Wide Web. Not
as pretty as we know it. This is the line-mode reader because part
of his audience were using green screen terminals connected to, typically,
mainframes. And what he was really inventing was using the Internet, which
was quite new at the time, for jumping around between servers and
moving. Not always just dialling into one box and having just that access, but going
from that box to the next box and the next box and not really caring anymore what box
it was on, and doing that from contextual links. So it’s prettier in this version.
Which is not quite contemporary with the first one, but pretty close. This runs on
a NeXT box. Well, I think it still runs, actually, at the Computer Science Museum
in Mountain View. Obviously, more graphical, we even have some 3D, something
going on in there, which is fascinating. But this is what Steve Jobs was doing, was
building this operating system, because he’d left Apple. And so these are
the newfangled toys in the computing world. And Tim wrote in Objective-C.
Anybody here write in Objective-C, doing Apple apps or anything? Don’t, don’t
bother. He wrote the first web browser in that for the NeXT, and NeXT is where we
got that wonderful language. This is Tim and he’s sitting at, what it
probably looked like first, a black and white NeXT Computer
that he was running that on. And also, I will say “Tim” throughout
this. That’s Tim. Sir Tim to his friends. The Web begat a lot of companies and
conferences immediately as soon as more people found out about it. It started
solving problems for CERN, but then, as it grew, because anyone could
make it grow through creating more pages, it got more and more interest. And
eventually, by 1994, we have the W3C, which is where Liam, hi, Liam, works and I
collaborate on web publication documents and stuff like that. But they
were already predating all this, there were already conferences on
hypertext, which still boggled my mind. And then about five years later, there are
conferences about the World Wide Web, and quite large ones. There were several
browsers spun out. Erwise, Viola, Mosaic. Mosaic later becomes
Netscape, then AOL buys Netscape and we don’t know about Netscape anymore.
I’m not sure why that happened. Mozilla was born out of Netscape. Mozilla
makes Firefox. It’s all this family history, and I’m sure there’d be a
wonderful chart somewhere that shows that. There is a place online that keeps browser
archives where you can download the binaries. Evolt or something. I’m seeing
some head nods. Something like Evolt browser history, if you look that up. It’s
a massive list of things you’ve never heard of. Some of them were around for
just a few months, sometimes days maybe, but some of them got
some traction, like Mosaic. So then we’re here in the present. So we’ve got all this history, Tim made
this thing, people started using it and it’s more and more places. We all get
these AOL free disks, right? We get online. We’re browsing what we
think is the Web. It gets better and better, and then we start going back to
AOL disks where we only dial into a thing called Facebook and we only read that
data, right? So we’ve got these social networks that don’t like hyperlinks
because they do exactly what Tim meant them to do, which is take you other places
besides Facebook. Instagram, if you’ve ever typed in a URL, it goes no
place, right? They have to highlight it, copy it, go to a different thing. Some of
these people on the Web don’t like the Web, which is sad. It’s ad-ridden, I
forget this. I use somebody else’s browser without AdBlock or a VPN and it’s a scary
place, I don’t know how you use it. It’s sad. Fake news, I’m going to skip
that one. Just because. Privacy loss. Tim’s been talking a lot
about this one for the last year. We’re all connected to a public Wi-Fi
right now. Do some homework on that. I’m not going to talk about that one
either, but it’s worth digging into. This is one I care about a lot, actually,
is the next one, the “body, script, body” thing. How many of you write
HTML or JavaScript apps? Have you seen this in guides about how to
build web apps? Right? That’s not a web page. Yeah? – [Man] [Inaudible] who shall
remain nameless [Inaudible]. – Yeah. This is how the kids are
doing it. Please educate them, please help them. That’s not a document,
that’s just a runtime for a thing that’s meaningless, right? Because what does that
do? Anybody have a guess? It doesn’t do anything. Right now it
doesn’t point to any JavaScript. But if it did, it’s whatever that
JavaScript does, and it’s a black box. And etc., you can fill that in. But let’s
back up. There’s this document. Something happened to my pretty swooshies.
Oh, well. This is the proposal that Tim did in 1989. And it started with this
wonderful chart and quote. But this chart outlines all the pieces of software that
they were using at CERN. And there’s circles and arrows about what
Tim’s doing. And smack in the middle there’s the “mesh” proposal, which was
just a working title. And it points to its relationship within CERN to all of these
different hierarchies or things like IBM GroupTalk. I worked for IBM briefly and
did all kinds of internal research to find out what that was. It was three things in
the course of IBM’s history. And I couldn’t find enough information on
that first version to know what it was. It was called, like, the first groupware,
I think it was maybe a mailing list server. I don’t know. But most of these things are
gone and they’re in history lessons, right? Like this one. But the Web has now
subsumed all of that, and that’s precisely what this first proposal about it was in
1989. It discusses the problem of loss of information. Anybody have this problem?
Yeah? About complex evolving systems, and derives a solution based on distributed
hypertext. So hypertext was a thing, as I’ve said too many times already, with
conferences that people were using. They’re talking about HyperCard and these
new things called CD-ROMs. And you were doing these multimedia
hypertext experiences where you bounced around a CD-ROM. It’s amazing. Then what
happened was Tim stuck the word “distributed” in front of it and started
jumping across servers, and that’s when the world blew up. Ouch.
I’m not sure where the black box came from, sorry. Not my laptop. “Firstly, it
discusses the problem of…” Yeah, that was the other quote, wasn’t it?
No. Yeah, so it’s focused on CERN initially, but in this quote is the idea
of linked information systems. So when you think of hyperlinks, like most
people, you’re thinking of moving from a page to a page. This document didn’t stop
there. Tim’s initial idea was thinking of computers moving from thing to thing and
these links talking about more than just other pages and more than you moving from
page A to page 75, you know, it doesn’t matter how you catalogue them. That you could jump
between them. That’s amazing, but then he went farther. And he was
talking about talking about real things in the data, using this same graph-y, linky
thing. So a linked information system. CERN was losing data, using lots of
different systems. Because they had all kinds of scientists coming and these
people knew their thing, just like happens in our companies, and they wanted to use
their thing to organize their notes. It’s all fine, whatever. But there was no way to
see across all of them or to expose them to each other, and that was increasingly
needed. So information was getting lost, including in things Tim built, like
ENQUIRE. Where it did fine for that department and the people who had access
to it, but it couldn’t talk about stuff in a different department and they couldn’t
talk about that, nor could they see each other’s stuff. And this is integral to why
the Web took off. CERN is a model and miniature for the rest of the world. It
was open, it was collaborative, it was multicultural, it was located in
lots of different buildings, and yet it was small enough and networked
that he could build this model and make it work. So this one still bugs me, and we’re
only at page five. The method of storage must not place its own restraints on the
information. Has that ever happened to anybody? Like maybe when you
were making an e-book and you think, “I can’t present this information because
X, Y, Z,” whatever. Or, “This is not spreadsheet data, but I’m in
a spreadsheet or a Word doc or wherever I’m sadly wandering.” And the system we
need is like a diagram of circles and arrows, where circles point to things,
right? So he goes on to talk about this. We can call the circles “nodes” and the
arrows “links.” So he says that and we do this. Right? Web page, link, web page. But
he also meant this to be possible. So human being plays video game, whatever
those identifiers on, on each side. They might not bring up a web page, or
they might bring up a web page if you’re browsing about that thing, but they’re
still actually talking about Benjamin playsMad Max. If you were in my talk
yesterday, you know that. And these are some examples he gave
because he didn’t playMad Maxat the time. I don’t know if he does now, either.
He talks about people being in a family tree graph where someone is the parent of.
Or all the way back in my first slide, I had these little “works at Wiley” and
“works for Sophia.” They’re not more complex than that, conceptually. That it’s
really just a sentence structure of a subject, a predicate, and an object, and
you’re building this graph of things. So let’s take a concept break, har har.
Get really think-y for a minute. There’s a term that is sometimes
derided. Often derided, Liam? Maybe often derided. The Semantic
Web. Do any of you know what that is, or have heard about it? How many of you
have heard positive things about the Semantic Web? All right, Liam. Oh,
okay, a few more people. Yeah, so the Semantic Web is really what Tim was
talking about inInformation Management: AProposal. So he doesn’t have any software
written here, he’s talking about this ability of real things pointing to real
things, expressions of graph data on a multi-node network system across the
entire planet. It’s a bit nuts. The Semantic Web conversation, that
happened in ’98. So we’re only, what, eight years into the actual
practical Web, maybe nine? Tim begins talking about the Semantic Web as like,
“This is really where I wanted this thing to go.” It gets called Web 3.0 in 2014 because 2.0
had happened in-between. And it gets kind of tangled up with
artificial intelligence, like a lot of other things. And it could certainly be
the foundation of that. And in Tim’s wilder moments, he described
this feature where it would do that, it would deduce things for you
and explain things back to you, which was actually the kind of stuff he
wanted ENQUIRE to do even before he was doing the Web. And those things are maybe
still possible. But in recent days, we’ve started talking about linked data, which
is like Semantic Web Lite. It’s like the ability to say those graph-y
terms and to use words, not just words, use URLs and identifiers
that universally talk about a thing. So an identifier for me that anywhere on
the Web is always talking about the actual me, not a web page that talks about me.
And to take that even farther so that I could build a graph that says, “This is
where I work, this is who I know, this is who they know.” And then through
some mapping figure out who I know from what company, like LinkedIn does for you,
or whoever. A lot of what you hear described as knowledge graphs at big
companies is using something like this, some sort of graph-based thinking. And
these are great reads if you want to understand the bigger vision. So a tree has…oh, yes. So from the Semantic Web,
this is what he’s thinking, the graph-y bits. And then he goes into
describing. Oh, he doesn’t use the word “Semantic Web,” by the way, inInformation
Management: A Proposal
. He doesn’t…hardly used the word “Web,”
actually. But he does describe all the things that had been tried inInformation
Management
, writ large, that have or haven’t worked and why
they haven’t worked. First, he picks on trees
and talks about how they’re great, but then you’ve got to
always find it in a tree system. And if it doesn’t fit there, then it ends
up in the General category in ONIX, right? It’s a tree, you’ve got to stick it
someplace. So there’s this bucket where everything else goes that doesn’t go
there. And then if you change the tree, what happens to the whole structure when
you want to move something and all your references break and bad things happen?
File systems still do this. Anyone like organizing on a file system?
Yeah, me either. So then it goes into keywords, which is a little more hackable,
right? It’s like the folksonomies and tags and stuff where you can just stick
keywords on the things. And if you didn’t limitations of number of keywords, you
could maybe categorize everything with keywords. And keywords, for the
most part, are kind of graph-y, or can be. But they’re also really not
fault-tolerant and you and I might use the same keyword to mean something else, and
matching our data together could cause serious problems. So he proposes that
hypertext is the solution to this because it gives you the ability to go from A to
72 and to also describe that relationship between A and 72 as
you jump across the Web. And you can click through these
slides and read all these quotes, if you like. So references
in this document all being associated to the network address.
So this is just the description of a URL or a link that you use all the time and
take you to the thing that it referred to, but then it can also refer to other stuff
with just a click of a mouse, which, again, is not an old thing at this
point. Can you image doing that? It’s amazing. And the thing that I had to
keep remembering when I read this the first time, which was a decade ago, was
that none of this existed, right? This ability to jump around was
brand-new. Like you dialled into a place, sometimes on a really ugly green screen
terminal. And just that experience that is in Engelbart’s demo when he starts doing
stuff like that, or even using a mouse, that that was mind-blowing
at the time is easy to forget. And CERN had requirements
that, again, made it a good miniature for the future Web.
It had access needs that were distributed among buildings
and among geographic locations. Had heterogeneity. These are all
the terms from his document. I liked “non-centralization.” We mostly say
“decentralization” now. But it didn’t have, because it was like a bunch of
departments, there was some, probably guiding authority, making some
shot calls. But they had to collaborate in a decentralized or non-centralization
fashion. And Tim’s original concepts mapped to that. And things that we’ve
built into the Web since then, or along the way, like DNS, have removed a
lot of that. So when you name things on the Web, you have to have a domain name
before you can do that. And you rent that thing. So that seems
like a bug, right? Because if you quit paying rental, then you can no longer have
those names. That’s bad. So we’re looking into ways with
recentralization to name things irrespective of where they’re located or
in addition to where they’re located, or something blockchain-y where it’s the
same thing. Right? But you don’t have to always pay monthly to rent them. Or
yearly, rather. And then access to existing data was something he spent a lot
of time on. And that really changed the future of what the Web became. Because he didn’t say, “This is
a whole new system where we’re going to put all our stuff and it’s
in a new place. And then if everybody uses the new thing, it’ll work.” Which is a
lot of what you hear from people selling you software, right? Like, “Your old
pipeline is broken. Screw it. Come over here and we’re going to start
over.” And it sounds so nice and it never works, right? Or it takes, you know, just way too
long. So the Web was dumb enough to sit on top of all of that and point to stuff and
give you a way to extract it and present it in maybe a dumber fashion, but a thing
that had this superpower of jumping around places. So you could expose databases,
you could expose those trees and those keyword maps into this hypertext space and
move between them. The private links thing is probably my favourite section. He also
says the word “annotation” in here a couple times. It talks about this ability
that none of us still have to make your own links and keep stuff. Now, in this
document, he’s simply referring to you making your own web pages. But at CERN, you were
probably running your own web server and adding to this. Or you had direct access
to his server and you’re storing it on his hard drive because everybody was very
trusting there and they could be. So you’re making your own HTML pages about
what you know about this department and you’re linking to the people that you’ve
met maybe, and it’s nothing more complicated than that. What has happened
since then is there’s this strong desire, since we mostly just stay in our browsers
and aren’t always publishing out to a web page or a blog, sometimes we do, but
that’s a separate space. That there’s the potential that the browser could get smart
enough to take care of its user and let that person annotate and let that person
keep stuff. How many of you use the bookmarks in your browser? You still do
that? That’s awesome. Okay, that’s good to know. I thought a lot of
people are like, “I use Pocket now that Mozilla bought it,” or whatever. Okay,
cool. That’s great to know. But how many of you go there to find that
thing that you found when you need it versus going to Google to find that thing
you found? Yeah? A couple people? Yeah? The Awesome Bar is actually pretty
awesome because it does bring up search history stuff sometimes, things that you
have found. Okay, cool. As a co-editor of the web annotation spec,
this one’s kind of important to me. And there’s a version of a future where
you do a bit more than just click a little star and put it in a tree someplace in
your browser. He says, “This project isn’t going to have a whole lot of bells and
whistles, we’re only going to worry about ASCII text on a 24 by 80 screen,” so 24
lines by 80 characters. Yeah, I wish that was our target. Oh, but he did start
saying the word “hypermedia” in this, that, in future, we could link
to documents that were videos or embed them or 3D stuff. So he’s not limiting,
“This is the Web, it’s 24 by 80.” He’s saying, “This is our target and it could
go who knows where.” These were things that he expected would happen at CERN, and
these are things that we do with the Web all the time. Data analysis, the ability
to link into a database. This hadn’t been done yet, right? PHP was
not a thing. There were no databases exposed as web pages that
you could click around at this time, but that’s exactly what he describes.
We’re going back to that chart and saying, “All this craziness that doesn’t talk to
each other, I want to put a layer on top of that and expose them so they all link
to each other.” And then the non-requirements are interesting, too. He
doesn’t worry about copyright enforcement or data security. Because he doesn’t need
to because they all work at the same place. But he points out very accurately
that all of this stuff can be layered on top later. So by leaving it out, he
actually accelerated past the one well-known contender in this space at the
time, which would have been Ted Nelson, who I mentioned earlier. His Xanadu solved
all these problems and it was kind of conceptualized before all this happened.
There’s a great article calledIn Praiseof Evolvable Systemsabout Tim’s Web and
Ted’s Xanadu, and the comparison between which one made it out the door first and
why. Ted’s system is technologically better. You can quote my work and pay me
for my quotation when you sell your document. So you sell Liam a viewing of
your document and I get a cut of that based on my quotation that he quoted. So
it’s like it’s all built in and it’s all secure and it’s all amazing and sometimes
hard to get your head around. Whereas Tim’s was really dumb, it was like
text that turned blue and you went to these known network locations, and that
was pretty much it. The first version of HTTP was GET, like, “Go get me this thing
and bring it back and I want to look at it and I’m going to go get something
else.” There was no POST, no PUT, no DELETE, it wasn’t complicated. But it
was, as Clay Shirky says, “It was the best worst thing out there.”
It’s not great, it’s really stupid, but it was good enough at what it did that
we could all build on top of it. And we’re still doing that and we’re not
going to stop, most likely. So they go from here into specific
applications at CERN, what would they build. And this is where his collaborator
comes in later to really get down to, “Here’s the software we need to write to
make this a success.” “Personal skills inventory” sound like any
website you used this week? I am Benjamin Young on LinkedIn,
if you want to find me there, not BigBlueHat.
Although you could probably find me there with that name. “Project
documentation,” anyone doing this every day? Right? On the Web, but not quite on
the Web, right? We’re all using stuff tucked inside of the Web, plus/minus links
out of it, right? But now we’re using whole interfaces built inside of the Web.
And document retrieval, which, of course, is what it does
natively. But then we’re also not just retrieving webby documents, which
is what he talks about as well, where you’re pulling in things like PDFs.
Which are not quite webby, right? They kind of have links. But you
could be getting videos that have zero links or whatever. And it allotted for
that and it’s strange to think about, but a lot of other systems didn’t, right?
Well, even his ENQUIRE system or software we use now that’s just note-taking stuff,
but you can’t drag a video into it and have that sit in line with your text.
Well, the Web does that, right? You can have an HTML document that
does that. But Evernote probably lets you do that now. But most note-taking software
doesn’t, sometimes deliberately. Yeah, so these are some conferences that
he had been to and suggested they go to the next one. He wanted to present this
new idea at the 1990 one, and I believe he did. There’s all kinds of
great talks that he’s recorded online, mostly his notes from them that are
wonderfully rambly. And the browsing techniques. “Lost in hyperspace” was
already a phrase. This is pre-Web, right? You’re just on a CD-ROM and you’re
already lost. How is that even possible? But, you know, we still are stuck with a
back and forward button. That’s it, right? And how many tabs do you have open
today? Right? No less than 20, at least. And then I all too frequently
tweet about my windows inside of tabs inside of windows and it’s just
mind-numbing, and at some point you just have to delete them all. The Onion even
did a piece on something about “woman closes tab,” tab-browsing experience,
and it has this moment of enlightenment. And she’s just so elated and happy, and
then finds the “Restore session” button and is doomed forever because now she’s
back to that really complicated space again. So there’s still a couple problems
to solve. And one of them is this “lost in hyperspace” feeling, which I’m certain we
feel a little more acutely than they did back then. “Interconnection or
publication?” This is still a question, probably especially here, about what
we’re doing. Are we connecting data across boundaries, or are we going to close it up
and package it and publish it within boundaries? That’s an ongoing discussion.
But Web technology does give us the way to build on top of proprietary stuff and
expose that conditionally. Because now we have built some of that
security stuff Tim left out and we have built some of those copyright things that
Tim left out. So most of the options are quite possible here. So yeah, this one
talks about us. There is also much support from the publishing industry and from
libraries, whose job it is to organize information. Do you have roaring support
for web publication at your publisher? You should come work for Wiley. Yeah, it’s
still a thing that’s not quite gelled yet. I was talking to our videographer in the
back about broadcasting, right? They didn’t care for the Internet.
Maybe still don’t, right? As Netflix generates all kinds of its own
content now. The Web is changing everything, that’s not new to anybody,
it’s 25 years old now, 26 years old, 25 last year. Publishers do what the Web
does, right? We distribute publication. The Web, it’s called web publishing, it
always has been called web publishing. And we can either be scared of it or we
can adopt it and run fast after making it awesome, maybe on our terms. Or at least
providing the intelligence of 209 years, or however Wiley is old, and experience,
and bringing that into new distribution models on this new medium. But I
come from the Web to publishing, so I’m happy to say I’m naive. And,
“Only way in which sufficient flexibility can be incorporated is to separate the
information storage software from the information display software.” We don’t
have this anymore in the web apps that we consume, right? We sort of do, right? The
browser is still providing the experience of clicking a link and going to a place,
but that “body, script, body” nastiness that I showed you earlier,
that gets rid of a lot of this. By allowing for that flexibility, and that
is not a bad thing that we allowed for it, we have given up some of that separation
of information storage and information display and now those come together in a
single package, which might be good. There are reasons for the developer, even
the publisher, to want to own that entire experience. But then there are also
reasons, like accessibility, where it should be separated enough that
the person can come up with or be provided some other way to experience the content
that is different from how you stored it and sent it to them. Because they need to
because they can’t get your content any other way. So, yes, he’s absolutely right.
It’s amazing and we don’t do it enough. Here’s what it looks like, this is the
Web. Pretty amazing. I really wanted that gradient to put in my slides, but I wasn’t
sure how he made it. Probably on a green screen someplace, or a
NeXT box. So browsers talking to hypertext servers, which then talk to other
hypertext servers. This is what REST was meant to do, representational state
transfer, where two servers basically browse links and fill out forms to each
other. And that’s that conference I run, you can come, too, we’ll talk later. And
the incentives to make this happen. Separating the information is amazing
because it’s an important phase in designing this world of how you move
across all this. Is how when you don’t necessarily own the browsing bit but you
do own the data, there’s a whole gap of design in here for moving the person
around. And we got as far as tabs and kind of stopped. So there’s a lot more design
to be done for the display program, and also for scaling these
database server things. Yeah, so exposing existing
information was a huge incentive at CERN,
still. And here’s an example of that. I just love his charts, so I put them all
in. They all say about the same thing. In this case, okay, this case back here,
these are just maybe file systems, I guess? I don’t know. They all have that
wonderful hash pattern. And then this one, because it’s got a box,
a table spreadsheet thing, is a database. So over here he’s browsed
to a database, and over there he’s browsed to a file system, just raw HTML. These are
the fabulous things he was exposing to the world, UUCP news, super decentralized and
overrun awesome tech that got not awesome because of the users on it. Let’s not let
that happen to the Web any more than it already has. VAXnotes, anyone still using
VAXnotes? I don’t even know what that is. CERNDOC, that sounds proprietary to CERN.
File systems, that one can die in a fire, I’m so tired of that. Telephone book, I
think he means a digital one, but he might not. And Unix manuals, I use
those all the time. Right? Like Maya and Python, whatever. They work,
still. And databases, which are insanely diverse, right? This one is fun. In a news
article, he goes through each of these and has ways they can do it. In news
articles, for example, one could use, in the text, a standard format to
reference to another article, or maybe inject emojis, which he left out because
they weren’t a thing. Or make mentions or use hashtags, right?
So it’s just text that the localized system, or potentially your
browser, could have said, “Oh, wow, I know what this reference looks like.”
Right? It’s to a religious text or it’s to a person because it has an at sign in it,
or it’s to a keyword because it has a hash in it, or it’s to a smiley face, that
one’s the most important in that list. So, “We lose out if we try to constrain
them, as we will exclude systems and hamper the evolution of hypertext in
general.” So he didn’t want to say, “This is just for tree systems,” but it’s
a new way to browse tree systems. Or, “This is just for databases.” And he
could have and it would have been a much smaller project. The other thing he did
not constrain, and this is debatably the main reason that it won, he didn’t
constrain the license of this. So CERN does not get a nickel every time
you browse a web page. They maybe should, but they don’t. It was
not patented, it was not restricted, and now we all built on top of it, and
will continue to for the foreseeable future. In ways that Xanadu and lots of
others never saw the light of day, or did briefly, like IBM GroupTalk,
whatever it was, because it did the thing and it did it in a proprietary fashion. It
couldn’t reach out to other people, or companies, or things. It couldn’t link
across those things. And even if it could, it still came with all kinds of patents
and restrictions and costs and things that hampered it. So Tim’s system
was really stupid and really free. And from there, we built all kinds of
great stuff. But the scary part is, we are killing off the bits that make it scale
and make it grow. And we need to keep that evolvability in mind. And including
licensing, not necessarily of content but of the things that we build on top of
that via open source and open standards, we can raise all the boats to continue to
make webby things that will be here in the next 100 years. And maybe there’s another
Web coming, I don’t know. But restraining it didn’t work then and
it’s not going to work now. That’s my theory. We should work toward a
universal, again, linked information system, he keeps saying this. It’s not
about just the pages or these constrained little apps like a Facebook thing that
does one thing. Or, you know, it’s huge, right? But it’s still inside of itself,
navel-gazing. So this is linking information across everywhere. And not
just information that is encoded in the system, but being able to talk about a
book by its ISBN, or being able to talk about a person by probably something
besides their Social Security number. And then taking that next, I hope, into a
portable thing that moves beyond the network, which is not
something that he touched on. In conclusion,
in his text, at least, “The aim would be to
allow a place to be found for any information or reference which one felt
was important, and a way of finding it afterwards.” That sums up the whole desire
that he had and I have every freaking day, when I sit down on my computer
and I still can’t do it. Right? We have all kinds of standards, good ones
and bad ones, that do not create this plugability into a place where you could
find and put everything to find it again. It just doesn’t work that way. I was super
happy about all the bookmark usage because that’s about as close as the browser has
gotten and they kind of stopped innovating there. There’s more to that, right?
There’s more to the experience of building your own web, maybe on your hard drive
somehow where it doesn’t look like saving HTML pages all the time, but where you’re
creating this internally browsable graph-y information system for yourself that also
connects to the Web. But it’s yours and it’s private and you
didn’t have to publish it on the Internet just to read it back. So one thing that’s
important to me in the stuff that I build is this concept of egocentric
architecture, where the person that is creating the data keeps the data, and it
starts there and it goes out from them later. Right now, you type in a comment
box and if you don’t click “save,” right, you lose it. That’s annoying. But if you
save it, it’s not on your hard drive, it’s not in that device, it goes away from
you. And then you got to find it again if you ever want it. And your browser forgot
that you even filled it out, it could at least keep the text for you
and say, “On this date, you commented this text on this web page on this
topic,” and let you find it again. Rant over. “The passing of this threshold
accelerated by allowing large existing databases…” I can’t stress this enough,
that this was an amazing piece. That he made a system in which you could
map all kinds of dead weight and make it look not dead with the new ones. And we
need to not forget that as we try and modernize publishing pipelines and deal
with old metadata. Exposing these things to something as organic as the Web is not
a bad thing and is super liberating for you and your data. “The system we need can
be combined from various sources. For example, a browser from one source and a
database from another.” So pick your browser, right? You probably got three of
them installed right now, I hope, at least three. Because they’re
not all great. And the databases are from all over the world, right? This is just
common now, but it wasn’t then, right? You would buy from, I don’t know
who the killer companies were at the day, most of them are gone. You would buy a
mainframe and you would buy a terminal. And that’s what it did, right? You dialled
into company terminal into company mainframe, and, yay, wow. And the Internet
was amazing because now you could dial into Joe’s hard drive from your
hard drive, and that was it, right? Like, you used a BBS or something. The
Web, on top of the Internet, lets you jump all over the place. So we get this experience
of that. But it was accelerated in part by open source and the ability that these
browsers came from all those sources, all the way back to those history slides
where two years in we’ve got like 10 browsers. That’s amazing. He had this core
concept that you could build all kinds of stuff on top of. And those people were, at
that time, sharing and talking a lot, then they didn’t for a dark
period we call the Browser Wars, and then they started again recently. So
we’re in the good part right now, right? The third age of man where we kill
Mordor and whatever, I don’t know. And databases, who knows
what they are underneath the Web, right? Who knows what Facebook is using
today? Let’s hope they open-source it, right? They do often. So, concluding fast,
this is a picture of the 2012 Olympic Games. “This is for everyone,” is what
that says. You can’t see the “this,” it’s off the thing. Search that because I don’t
have time to tell you about it. So is it for everyone? This is the mess
we’ve made. It’s mostly English, vastly English. Most people don’t speak
English on the planet. Sorry if that’s a newsflash. It’s tied to
dirt. I keep talking about this with lots of people. The Internet and Web live on
this weird plane above dirt, right? Above geopolitical boundaries. But
you still trade licenses for content based on dirt, but it doesn’t move on dirt. You
might have gone through an e-mail server to have that conversation that wasn’t even
in your country about dirt-based stuff and copyright, but we need new models that
actually acknowledge that we don’t work that way anymore. All kinds of models,
maybe political. Technological, need to be accessible. Better people here
to talk about that than me. And then copyright and licensing, you
can’t always find what the copyright or license is of a thing once you find it
digitally. It’s a bug in the web. And we can fix that, there are specs for
that. Right, Liam? Attempts at it, probably several. Author and attribution,
I’ll go to blogs and read the blog, I can’t find the date, I can’t find who
wrote it. Maybe it’s just got some nickname, which is fine, I’m okay with
nicknames. You can be a dog on the Internet, that’s okay. But I at
least want it to go somewhere, I want to find like, “Okay, if I’m going
to talk about this entity, I now know where they are or what they are
or that they have an identifier.” Publication dates, authority, ownership,
they’re all missing. Tim proposed a link data system, not just documents. A
graph-based data presented of relationships between the things. A
borderless information space to wander, not dirt-based. On-click takes you across
the Web to other servers, countries, it doesn’t have boundaries, the
Moon. InterPlanetary File System, look that up, too. That’s a thing. It’s
fabulous. And a worldwide note-taking system for all your knowledge to go into.
Right? Amazing. Okay. We have the technology, we can fix this.
Clay Shirky, talked about him. I’m wrapping up quickly. So these things
are all terrible, it’s the whoopee cushion and joy buzzer of technology, is his kind
words about it. And, “Except, of course, for all the others.”
It’s still a heck of a lot better than all the contenders were. So the opportunity is
publishing at the W3C. The IDPF and W3C are now one thing called
the W3C, just trying to make it easy. EPUB was built out of bits for the Web. If
you were in my talk yesterday, you know this. But it’s not webby. Right?
And the web is built out of all kinds of things that aren’t pub-y, right? The
distribution model is only networked and humans aren’t always networked. If you
live in Cuba, somebody shows up at your house with a hard drive and you pay them
per megabyte to download “the Internet” from last week on your hard drive and
browse it until they show up again with the hard drive later. We should build technology that
works for that. That’s okay. I’m sorry they have to do it that way, but
that’s okay. They’re moving data, that’s a network, it exists, it doesn’t
use TCP/IP. Sorry, Vint, but it doesn’t. Vint Cerf, TCP/IP.
Combining these into web publication and then, by extension, portable web
publications. There’s that Hugh McGuire quote again, it’s awesome. Thanks.

Leave a Reply

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