back to indexBrendan Eich: JavaScript, Firefox, Mozilla, and Brave | Lex Fridman Podcast #160
link |
The following is a conversation with Brendan Eich,
link |
creator of the JavaScript programming language,
link |
cofounder of Mozilla, which created the Firefox browser,
link |
and now cofounder and CEO of Brave Software,
link |
which has created the Brave browser.
link |
Each of these are revolutionary technologies.
link |
JavaScript is one of the most widely used
link |
and impactful programming languages in the world.
link |
Firefox pioneered many browser ideas that we love today
link |
or even take for granted today.
link |
And Brave is looking to revolutionize not only the browser,
link |
but content creation online and the nature of the internet
link |
to make it fundamentally about respecting people's control
link |
Quick mention of our sponsors, the Jordan Harbinger Show,
link |
Sambaska Mule Delivery Service, BetterHelp Online Therapy,
link |
and Aidsleep Self Cooling Mattress.
link |
Click the sponsor links to get a discount
link |
and to support this podcast.
link |
As a side note, let me say that there's a tension
link |
between theory and engineering that I've
link |
been thinking a lot about.
link |
I tweeted something like, good execution
link |
is more important than a good idea, but one helps the other.
link |
I think the wording of that sucks,
link |
but what I mean is a good idea is a must.
link |
But in my experience, good ideas are in abundance.
link |
Good execution, on the other hand, is rare.
link |
I think some mix of good timing, good idea,
link |
and good execution is essential.
link |
Getting that mixed right is tough.
link |
And Brandon somehow, multiple times in his career,
link |
I'm starting to believe it's more art than science,
link |
like most interesting things in life.
link |
If you enjoy this thing, subscribe on YouTube,
link |
review it on Apple Podcast, follow on Spotify,
link |
support on Patreon, or connect with me on Twitter,
link |
And now, here's my conversation with Brandon Ike.
link |
When did you first fall in love with programming?
link |
I didn't program a lot when I was in high school,
link |
but I had a friend who had a Commodore pet.
link |
And after we saw Star Wars, he said,
link |
hey, let's make a basic program that
link |
does the Death Star trench run.
link |
And it was just simple 2D graphics.
link |
And I didn't know what I was doing,
link |
so I just talked them out on the math and stuff like that.
link |
I was a math and science kid.
link |
I was really into the HP calculators of the early mid 70s.
link |
These were the RPN.
link |
They were really strongly built.
link |
And all right, Gold Trigger, instead of gold, divinely heavy.
link |
There's probably some gold in them, too, gold metallization.
link |
But they were awesome calculators,
link |
and they had all the scientific functions.
link |
So I was really into that.
link |
So I aimed toward physics.
link |
I was a little late for the, I think,
link |
the 20th century golden age.
link |
And I read a lot of science fiction.
link |
So I was like, yeah, it's on the hyperdrives and warp drives.
link |
And physics was not going to get there quickly.
link |
And I started hacking on computers while I was studying physics
link |
as an undergraduate at Santa Clara University.
link |
And, you know, I dodged the Fortran bullet
link |
because I was in the science department
link |
instead of the engineering department,
link |
where they still did Fortran card decks.
link |
I think they had an auto collater.
link |
But we were using Pascal.
link |
And I got one of the first portable C compilers ports
link |
to the deck mini computers we were using.
link |
And I fell in love with programming
link |
just based on, you know, procedural abstraction, Pascal,
link |
just what now would be considered old school,
link |
like structured programming from the 70s.
link |
Niklaus Wirt, the creator of Pascal,
link |
was a good writer and a good pedagogue, right?
link |
He always at ETH would do these courses
link |
where it's like build your own computer,
link |
build your own compiler, build your own operating system.
link |
And I know some people who are grad students
link |
under him and said he would torture the students
link |
with things like this custom email system
link |
that had 25 word limit and things like that.
link |
I unfortunately dodged both the Pascal and the Fortran bullets.
link |
Could you maybe linger on the Pascal?
link |
Like what kind of programming language was it?
link |
What is the reminiscent of today?
link |
Because it sounds like it may have had an impact
link |
on your own trajectory.
link |
Yeah, it was in the Algoel family.
link |
And Algoel was the big successful language design
link |
and compiler project in the 60s.
link |
They had a successor called Algoel 68,
link |
which was ambitious but not as successful.
link |
But Pascal was kind of a wordy procedures
link |
and functions language.
link |
It distinguished between functions,
link |
which are a turn of value and procedures,
link |
which don't, which just compute.
link |
And you could say that whole Algoel family went into ADA.
link |
Pascal had a second life thanks to Borland with Turbo Pascal,
link |
which was hugely successful.
link |
I think in large part due to Anders Helzberg,
link |
who then went to Microsoft and did C sharp and done that
link |
with his team there and has done really well,
link |
doing TypeScript, Type JavaScript.
link |
So yeah, there's a lineage here.
link |
But I was also interested in C and Unix
link |
by the time I was an undergrad,
link |
because people were bringing Unix up on all sorts of hardware.
link |
I had some friends who were doing their own wire wrap computers,
link |
And I was wire wrapping for my engineering course,
link |
6809 or something simpler,
link |
building a computer on a board.
link |
And I wanted to build a more ambitious one
link |
and port Unix to it,
link |
but I picked the wrong processor.
link |
I picked the National Semiconductor NS 16032,
link |
which was this amazing, you know,
link |
SISC complex instructions that computer
link |
and not the reduced instructions that computers
link |
that were just being contemplated into the mid 80s.
link |
And risk ultimately went out.
link |
Risk one, in some ways, it dissolved into,
link |
Now you have these superscalar architectures
link |
where like Intel has kept
link |
probably too much backward compatibility
link |
at the instruction level,
link |
but that's just, there's a front end that parses that
link |
into these, you know, these wide internal instructions.
link |
So, you know, the very long instruction word research
link |
that was also interesting at the time
link |
kind of became the micro architecture inside
link |
the backward compatible Intel.
link |
But I picked the National Semi chip
link |
and it never got made successfully.
link |
It was full of bugs and I never could have brought it up.
link |
But I went on out of physics after three years
link |
into math computer science.
link |
And like I said, I did it because I saw,
link |
I was being sort of childlike and naive about physics.
link |
And I thought, meanwhile,
link |
the Valley is go, go for computers.
link |
The Apple to write the PC,
link |
8088 base PC, the IBM, you know,
link |
gave Microsoft the future for,
link |
you know, a somewhat fishy deal.
link |
So it was wide open in the computing space,
link |
but in physics, you were as optimistic about physics.
link |
No, I mean, I was one of three brothers
link |
who were all in the same grade.
link |
I have a twin and a younger brother
link |
who skipped second grade
link |
and was with us the whole time after that.
link |
And, you know, he went on,
link |
he actually studied under Kip Thorne at Caltech.
link |
But he also ended up in software.
link |
Does it make you sad that theoretical physics,
link |
even with strength theory,
link |
hasn't really had any foundation breakthroughs
link |
in the latter part of the 20th century?
link |
In fact, I'd say the problem is theory over experiment.
link |
I would say, you know,
link |
we need more Aristotle and less Plato.
link |
You know, mathematics is not all physical.
link |
There are lots of mathematics that cannot be realized
link |
as far as I know in this world.
link |
So to understand the world,
link |
you need to do experiments.
link |
You need to not just dream up inductive theories
link |
that could have lots of alternative theories
link |
competing with them,
link |
with no way to decide between them except aesthetics,
link |
which is not a good guide in my opinion.
link |
I don't know if you are friends
link |
or have a relationship with Elon Musk.
link |
Where's the, in terms of what you would love to see
link |
our society investing in, building up,
link |
is it closer to Elon or is it closer to Feynman
link |
and Einstein and those?
link |
Well, those gentlemen are no longer with us.
link |
And I think that's noticed.
link |
So like I said, the real glory day is of physics.
link |
The famous pictures from Germany before the Second War
link |
were just a fantastic assembly of brains,
link |
Schrodinger and Einstein.
link |
And physics, I think, took a wrong turn
link |
that maybe all of, I would say, western science took
link |
in going for models over reality, right?
link |
You see this in all sorts of fields.
link |
Now, we can build models that are very predictive and generative
link |
and then we build actual devices or, you know,
link |
semiconductors, things like that.
link |
That's good. I'm not dismissing that.
link |
We need good models.
link |
We need to experiment and prove them and test them.
link |
But the problem I've seen in physics, which you see,
link |
certainly in economics, the dismal science,
link |
and you see surprisingly in other so called hard sciences,
link |
is models that don't really have to be tested against reality.
link |
They can instead become policy tools or they can become,
link |
like I said, one of a large family of alternate theories
link |
that could be as predictive,
link |
but nobody's doing the winnowing out.
link |
That's such an interesting tension in society.
link |
You see this in even the software sciences,
link |
which have a deep love for psychology.
link |
You see this in epidemiology, now the virus.
link |
It's this tension of, you know,
link |
how much of the world can we understand
link |
through just a beautifully fit model?
link |
And then at the same time, my main work is a machine learning
link |
where it's like there is no provable thing usually.
link |
It's just kind of you, it's all about just getting the right
link |
data set and getting tricks and so on.
link |
And there's this tension, even in my own soul,
link |
of like I grew up on theoretical computer science,
link |
like I loved approximation algorithms,
link |
like all of that, like different complexity classes,
link |
just those little puzzles.
link |
I mean, I don't know.
link |
Do you, to you as somebody who was in math and computer science
link |
and then end up going into places where you engineer
link |
some of the most impactful things in this world,
link |
do you see the P versus NP,
link |
all that whole space as interesting at all?
link |
Yeah, it's not that useful in practice, right?
link |
People are using it with sort of crypto analysis
link |
or asymptotic arguments about, you know,
link |
can we have a quantum resistant crypto algorithm,
link |
things like that, which may not be practical, right?
link |
If you follow Mikhail Diakonov or Gil Kolay,
link |
there are big questions about how quantum computing
link |
will scale up, how practical it will be.
link |
Is that something that you think about quantum computing?
link |
Except for spare time.
link |
Like you said, I'm not using this kind of computer science
link |
in practice because almost everything now is engineering
link |
and finding ways to get computers to be more useful for people,
link |
which goes from, you know, design problems,
link |
which are really kind of an art.
link |
Like Newt said, anything you can't automate is an art.
link |
Well, we can have, you know, machine learning,
link |
compose music, and it can imitate, you can train it,
link |
and it can sound kind of decent,
link |
but maybe lacking that je ne sais quoi.
link |
But, you know, user interface still I think requires human art.
link |
So speaking of things that didn't follow a perfect theory
link |
and model, JavaScript, so there's two things.
link |
One had an impact on the world at a huge scale, obviously,
link |
and it also still is one of probably the most popular
link |
programming language in the world.
link |
So can we go back to the origin story?
link |
Can you tell the story of how JavaScript was created?
link |
I was at Silicon Graphics after graduate school for seven years,
link |
and it got to be big and successful and divisionalized
link |
and political, and I thought kind of boring.
link |
And a friend who'd been there went to one of the last
link |
of the super companies, the super startups in the early 90s.
link |
There were several, I suppose General Magic was a little after
link |
that around the same time, but Micro Unity was that company
link |
that I went to, and it was because my friend,
link |
Jeff Weinstein had gone there from Silicon Graphics.
link |
He recruited me, and Micro Unity was doing everything.
link |
So this was like the ultimate sort of pretend grad school.
link |
It was doing a new fab, new semiconductor process.
link |
It was doing new analog and digital circuits on the same,
link |
very large, but not way for scale chip.
link |
Originally it was five centimeters on the side.
link |
It was really hot too.
link |
So I needed a water cooler.
link |
It was a Craig Hiller, and then they shrunk it,
link |
and they tried to do a home sort of media processor
link |
that was essentially a barrel processor,
link |
but you could think of trying to do all the things that we now
link |
see in modern architectures with short vector instructions
link |
and sort of wide instructions or multiple issue.
link |
And doing a lot of the stuff in software,
link |
because the second iteration, the Set Top Box,
link |
was really for avoiding the cost to the cable company
link |
of rolling the trucks out to replace your garbage
link |
General Atlantic Set Top Box with a totally newer,
link |
less garbagey one.
link |
So if you could have software gradable Set Top Boxes,
link |
the cable companies thought they could save a lot of money
link |
Or which level of the software?
link |
We were using GCC.
link |
We were writing C++ and C.
link |
Today, I worked with their really very smart guy,
link |
hired from a sort of Wall Street hotshot programming consultancy,
link |
did his own hardware design as well as software.
link |
We were working on how to make not only short vector units,
link |
but general bit shufflers and promoters,
link |
so you could do things like crypto algorithms efficiently,
link |
and you could do demodulation of the cable complex quadrature
link |
amplitude modulated signal.
link |
You're basically taking A to D converters,
link |
dumping things in buffers,
link |
and then doing the rest in software.
link |
All the framing and the Reed Solomon and Viterbi
link |
and all that error correction.
link |
So that was really great learning experience,
link |
but it was not going to work.
link |
It was doing too many risky things at once, right?
link |
If you, as Jim Clark said to me when I hopped in Netscape
link |
after three years at Micro Unity, he said,
link |
oh yeah, you do 10 things each,
link |
one in 10 odds, it's going to be one in 10 billion, right?
link |
The multiplication principle.
link |
So Netscape was already a rocket,
link |
and I had passed the chance to go there in 1994.
link |
I knew the founders because I worked at SGI Clark's company.
link |
Could you pause for a second in Netscape?
link |
When was the launch of this rocket?
link |
1994 was the launch of Netscape?
link |
And I went there in early 95 in April.
link |
So you said you missed the launch?
link |
Well, I missed the first floor employment opportunity,
link |
but the IPO was August 1995, so I was there for that.
link |
How obvious was it that Netscape was like world changing?
link |
How was the layout?
link |
Was Netscape one of the first big browsers?
link |
So when I was at Micro Unity still in 93,
link |
we saw a browser called Mosaic.
link |
And up till then, we'd used email,
link |
and we'd used Usenet, the NNTP protocol.
link |
We'd used Newsreaders.
link |
We used all these old internet protocols,
link |
all relying on the DNS and TCPIP and UDP for that matter.
link |
When I was at Silicon Graphics,
link |
we brought up the whole stack, right?
link |
We had to discover how to find the ethernet,
link |
find the IP addresses on your network,
link |
and then find IP addresses for them,
link |
ARP protocol, all that stuff.
link |
And it was great because nobody knew in the 80s
link |
what was going to win.
link |
All the proprietary stacks like IBM,
link |
SNA, and DeckNet, and all these other protocols were saying,
link |
we're going to do it, or it's going to be heterogeneous future.
link |
Instead, it was Berkeley Unix and the TCPIP stack
link |
that dated back to the ARPANET that won.
link |
And I think we knew it.
link |
We all knew it at SGI, but the salespeople didn't.
link |
Multiple network stacks interoperating.
link |
But in the end, it won.
link |
And so that was the internet.
link |
And it was email and texty, and it was used as very texty.
link |
And then Tim Berners Lee did his thing,
link |
but I don't think I was paying attention.
link |
And I think the date when he first did it,
link |
or when he wrote the famous emails and pushed back to 89,
link |
but I noticed Mosaic in 93 because one of the things
link |
that Mark Andreessen and Eric Bina did at NCSA
link |
was they innovated on the early HTML standard.
link |
In particular, Mark sent this email saying, hey, everybody,
link |
we think you should be able to put an image in a page.
link |
And you know when he sent that, Eric Bina had already written
link |
And I talked to Tim Berners Lee more recently just a few years ago,
link |
and he was like, oh, we had another way of doing it.
link |
And it didn't work out because Mark shipped his in Mosaic.
link |
And this convinced me of several things.
link |
One, the internet meant there was a huge first mover advantage
link |
and being fast, getting on first mattered a lot.
link |
And so Richard Gabriel of Scheme and Poetry fame
link |
has written about this.
link |
Well, he's a poet.
link |
Oh, actual poetry.
link |
He thought it was something.
link |
I mean, he's the founder of Lucid,
link |
which is where Jamie Zewinski worked before in Netscape.
link |
And Lucid was doing compilers and Lucid Emacs,
link |
which was a fork of Emacs, famously Jamie fighting
link |
against Richard Stallman, Stallmax.
link |
And so Richard Gabriel, you know, very, very brainy computer guy,
link |
but also a poet, but he wrote a nice essay
link |
that gets abused all the time.
link |
In fact, Jamie's put a kind of warning
link |
in front of his version of it on his site,
link |
They were called worse is better.
link |
And this is about survival advantage of software
link |
in the network world, in my opinion.
link |
It started out being framed as Unix and Lisp,
link |
good news, bad news, because all the Lisp people,
link |
the MIT people were like, oh, you know,
link |
the crown jewel, right, Scheme, this Faberge egg
link |
or common Lisp, this giant cathedral.
link |
Of course we're going to win.
link |
This is civilization.
link |
And those, you know, those farmers in New Jersey
link |
to borrow from the Sopranos, those hicks down at Bell Labs,
link |
they're just, you know, there's nothing sound there.
link |
Well, guess what one.
link |
So you're saying this is a fundamental,
link |
like principle of the internet is moving fast winds.
link |
You could say in almost any network system,
link |
like biological evolution, you see successful,
link |
real sweet populations and they don't always have,
link |
you know, they aren't free of flaws.
link |
The heterozygous advantage, right?
link |
You can get both parents, give you the gene variant
link |
and you get sickle cell anemia, right?
link |
But if one of them does, you're more resistant to malaria.
link |
And so this isn't a beautiful process,
link |
except at large scale.
link |
And then you realize that because it moves fast
link |
and can adapt, it can win.
link |
And people still struggle with this.
link |
I used to struggle with this because JavaScript
link |
was done in such a hurry and the force of web compatibility
link |
meant early mistakes couldn't be fixed.
link |
And even the standards process injected new mistakes,
link |
But often standards bodies go back
link |
and making compatible changes.
link |
You can't do that with the web.
link |
It's more like, again, like biology,
link |
you preserve what still works.
link |
You don't want to break ATP metabolism or whatever.
link |
So you have to kind of resign yourself to the reality
link |
of versus better being enshrined
link |
in actual design points you might not like.
link |
And that happened with JavaScript.
link |
And I'm way over it.
link |
But it also, I think, was a huge advantage
link |
why JavaScript has kind of swept a lot of programming domains.
link |
People will say, oh, it's not because of merit.
link |
Well, you're right.
link |
But we also improved it over time in the standards body.
link |
I spent 20 years doing that.
link |
And you don't get that choice.
link |
It's like, I'm not saying that that was the best language.
link |
I'm just saying that was the right time to do it.
link |
And I like to say the alternative was not to do it.
link |
I could have told Netscape, I can't do this.
link |
And it would have been visual basic script.
link |
And it would have been bad.
link |
So that's a good way to present the alternative.
link |
But so it was a Netscape.
link |
And you have written it how many days
link |
and why was it only that many days?
link |
And what was the goal and the underlying principles
link |
So the whole, I'm sort of describing worse is better
link |
in a frenetic way because it fit the model of Netscape.
link |
When it was known that Jim Clark and Mark Andrius
link |
were founding Netscape and they did the first release in 1994,
link |
that browser took over from Mosaic.
link |
In fact, that's why Mozilla is called that.
link |
It's the Mosaic killer.
link |
It's like the giant monster that kills Mosaic.
link |
And they knew that it wasn't that, again,
link |
it's not like you're doing advanced scientific research
link |
that is changing the world.
link |
It's more like taking down the last iteration
link |
in the browser Mark did, which had images
link |
and other affordances before you stopped working on it.
link |
And you're making Netscape the new thing that has images,
link |
plug ins, which was the way to do video back in the day.
link |
It had something that's kind of died now for tiled windows
link |
called frames and frame sets.
link |
Oh, yeah, yeah, yeah, yeah.
link |
Eric Bina did tables in Netscape 11.
link |
So when I got there, they were heading toward IPO.
link |
Clark wanted to IPO early.
link |
I think his instinct was right.
link |
And that kicked off the whole dot com era.
link |
There was a recession in the US in 91.
link |
You can see old law and order reruns
link |
where they talk about the recession, how hard it's hitting
link |
And after that, Greenspan really goose things
link |
at the Federal Reserve.
link |
And technology had been sort of fermenting in a way that came
link |
together with the internet.
link |
And Netscape made it possible to do pets.com,
link |
to do eBay, to get people to recognize a URL on a billboard
link |
and then type it in when they get home.
link |
And that was huge.
link |
That was so fast moving a rocket that Mark and the
link |
engineering team there thought we need to make this a
link |
programmable browser, not just a document viewer,
link |
It was all HTML with images and tables.
link |
And also, like you said, frames.
link |
There's no dynamic element at all.
link |
The most dynamism you get was from a plug in,
link |
which there are a few of them then flash didn't exist
link |
Well, that's the thing.
link |
We did the deal with Sun.
link |
In fact, I was recruited to go do Scheme in the browser.
link |
Remember Guy Steele and Gerald Susman's beautiful
link |
I was going to do it in the browser because my friends
link |
from SGI thought, hey, we like Scheme.
link |
And I'm like, I hardly ever use Scheme.
link |
It's not really used in industry except in sort of silos.
link |
I'll come do Scheme in the browser.
link |
I have a slide from my 2017 talk where I have Bruce Willis
link |
crawling through the duct in diehard.
link |
He's like, come out to the coast.
link |
Have a lot of fun.
link |
Come on, do Scheme in the browser.
link |
But when I got there, there was no Scheme in the browser
link |
because they'd started a deal with Sun Microsystems.
link |
And my best contact there was Bill Joy,
link |
who I admired as a Berkeley UNIX founder and a Sun founder.
link |
And Bill got the idea of making the browser programmable too.
link |
And so the main idea was to put the Java VM,
link |
which at that point was not really easy to embed,
link |
into Netscape, including the Netscape version of Windows
link |
that was still most popular, which was the 16 bit Windows 3.1,
link |
which was going away.
link |
Microsoft was coming out with Windows 95.
link |
And everyone was afraid they were going to do Internet Explorer,
link |
I guess, two at that point, three the next year.
link |
They already bought or invested in somehow Spyglass,
link |
this other company that shot out from NCSA at University of Illinois.
link |
And in fact, Microsoft had tried to buy Netscape in late 94
link |
before I got there.
link |
And I heard about this later.
link |
I heard they offered way too little money.
link |
And so Jim Barksdale and Jim Clark said, get out of here,
link |
But then they realized, oh, this is going to hurt us
link |
because now they're going to copy us.
link |
It didn't happen right away.
link |
I'm not sure when Gates Internet Title Wave memo was written.
link |
That's the famous memo he wrote when Bill Gates realized
link |
that Microsoft was going down this old copy AOL path
link |
or copy compuser path, a project called Blackbird,
link |
presumably after the SSR 71, I don't know.
link |
But they were going to make a dial up service with a custom
link |
content language stack and custom rendering.
link |
It wasn't the web.
link |
They could have content partners.
link |
They have a lot of money.
link |
But it still wasn't to scale the web.
link |
It wasn't going to be compelling.
link |
And Gates realized this and he turned the company on at a dime
link |
and they couldn't buy Netscape.
link |
Again, I'm not sure the timing, so they decided to copy it.
link |
And once we realized that everybody inside Netscape
link |
felt even more urgency and more of a frenetic mood.
link |
And so my chance to do scheme disappeared when the Java deal
link |
But there was still a chance to do a companion language to Java
link |
because Java is a compiled language.
link |
It's evolved and improved quite a lot since then too.
link |
But it was for sort of serious advanced programmers
link |
that cost a certain salary or hourly rate.
link |
And people observed, Bill Joy observed, and Mark Andreessen
link |
and I observed that in a mature stack like Microsoft,
link |
you really benefit from having a scripting language
link |
like Visual Basic, which became Visual Basic Script in IE3
link |
but didn't take over and kill JavaScript,
link |
that you need two languages.
link |
One is for the component writers who are higher price
link |
And the other is for scripters, certified public accountants,
link |
designers, graphic designers with some programming inclination.
link |
Anybody, amateurs, doesn't matter.
link |
There's a much more demotic approach there
link |
for programming the components together, gluing them together.
link |
Some people will say ducked tape language,
link |
which I don't really like.
link |
But we saw Bill Joy and Mark Andreessen
link |
and we saw the need for companion language.
link |
And Gleaminar, I was to call it JavaScript.
link |
That was marketing's plan.
link |
Mark called it Mocha, which I liked.
link |
And Netscape marketing, I think, didn't like that.
link |
So they said, oh, there's some trademark and some software,
link |
somewhere that uses Mocha.
link |
So we can't use that.
link |
And they tried live script in August and that didn't last.
link |
And then finally, we got the trademark license in December,
link |
But the work I did to prove that it could be done was important
link |
because I came in in April and even then Netscape was growing
link |
so fast that they couldn't find an open hiring requisition
link |
in the client team for me.
link |
So they hired me into the server team.
link |
And I worked for a month on server team on what became HP11.
link |
So I was actually, I had done protocol work at Silicon
link |
Graphics with Greg Cheson, former Bell Labs intern,
link |
grad student intern who knew all the UNIX founders.
link |
And Greg was very interested in taking protocols to the next level
link |
with VLSI because he thought that CPUs wouldn't scale up.
link |
He was mistaken in that, unfortunately.
link |
Moore's law more than kept up.
link |
And you have gigabit ethernet running with conventional processors.
link |
But I worked on protocols at SGI as well as UNIX kernel hacking
link |
and NFS and things like that.
link |
So I came into Netscape to work on the server side for a month.
link |
But I was thinking the whole time, what should this language be like?
link |
Should it be easy to use?
link |
Might it syntax even be more like natural language?
link |
Like hyper talk, which is Bill Atkinson's language in hypercard,
link |
if you have ever used hypercard on an early Mac.
link |
And I thought, well, I'd like to do that.
link |
But my management is saying, make it look like Java,
link |
which looks like C from a distance.
link |
What does that mean?
link |
We're talking about visually.
link |
Does it mean like management?
link |
Do they understand what you think about it?
link |
No, marketing didn't know.
link |
But management did.
link |
Like Rick Shell, the VP of Engineering knew.
link |
And we had a plan even that was, if you have this companion language,
link |
you're going to glue things together between Java and JavaScript.
link |
So you're going to have commerce in memory in the heap with data types.
link |
So you're going to want some of the data types in Java
link |
to reflect in the JavaScript.
link |
You're going to want the primitive types that Java unfortunately
link |
severed it from objects.
link |
So at least some of them number double, let's call it in Java's terms,
link |
from the C term for double precision floating point or strings
link |
or Booleans and objects.
link |
And so right away, there was this constraint that looking like Java meant
link |
kind of the C curly brace syntax,
link |
but also some of the data types and objects.
link |
Like objects and so on and so forth.
link |
Yeah, I thought it called.
link |
Comparison operator.
link |
Garbage collection.
link |
Garbage collection.
link |
Even the bitwise operators and the shift operators,
link |
including the unsigned right shift,
link |
which Java had because it didn't have unsigned integer types.
link |
It said, if you want to do unsigned operations, use an operator.
link |
And that turned out to be important much later.
link |
I'll tell that story five times.
link |
But JavaScript inherited a set of operators,
link |
the expression grammar, the statement grammar up to a point from Java.
link |
But I wanted a functional language.
link |
I want to scheme a little bit of scheme,
link |
even though it wasn't as clean as scheme.
link |
You had a love search interrupt.
link |
You had a love for scheme and list,
link |
but that functional language last game.
link |
I wanted first class functions because I saw the need for callbacks
link |
in the browser where it's a single threaded program.
link |
All the early browsers were single threaded.
link |
And it's the right model for users.
link |
Most users weren't ready for mutual exclusion and threading.
link |
So in a single threaded world,
link |
you cannot block the user interface.
link |
So you have to use a callback and run later.
link |
And without getting too fancy and trying to capture the continuation
link |
like call CC does in scheme,
link |
I thought I'll just make it easy to have fun arcs.
link |
First class functions, you pass downward and they can call back.
link |
It'd be called back.
link |
And Java didn't have that at the time.
link |
It took forever to get proper first class functions.
link |
Lambda is now into Java.
link |
Java 7 or 8, I think.
link |
It did have concurrency, right?
link |
From the very beginning,
link |
but you were thinking that a JavaScript in the browser
link |
would not have the luxury of being concurrent.
link |
And the reason was Java was going to run in a plug in
link |
so it could fork threads and go to town.
link |
But the main action in the browser was in the single threaded program,
link |
single UNIX process on UNIX or Windows.
link |
And it was where you had to service the event loop
link |
and then go do things.
link |
Respond to the network, lay out some HTML, render it,
link |
turn widths into heights by filling containers,
link |
boxes, the early, what became the CSS box model.
link |
And run scripts to make the thing livelier,
link |
respond to user input.
link |
And all that event driven programming was in part like Hypercard
link |
because Hypercard had this on event name syntax.
link |
And so that's why you have in JavaScript on click
link |
running together as the name of the event handler.
link |
And there's some funny ones on mouse over and on mouse out.
link |
People still complain about those.
link |
But, you know, there are many more events now
link |
over the years standardized.
link |
But it was a mix of event driven single threaded programming
link |
because it had to run in the main thread of the browser
link |
where the action is.
link |
And Java never got there,
link |
which meant Java could not interact easily or quickly
link |
or in a nested way with the document,
link |
with the objects reflected from the HTML document,
link |
with the tables and forms and so on.
link |
And that is one of the reasons I think JavaScript survived
link |
and Java kind of died.
link |
Java was in this plugin prison.
link |
It essentially was confined to a rectangle,
link |
the Apple rectangle.
link |
And while we even built a next year,
link |
Nick Thompson, a friend from SGI,
link |
who was an intern grad student at CMU at the time,
link |
built the first version of LiveConnect to glue Java
link |
and JavaScript together to deliver on that vision
link |
where you do have commerce between the data types and the heap.
link |
But Java was in charge.
link |
JavaScript was in charge.
link |
And Java was just these components, these helper objects.
link |
You might as well do everything in JavaScript.
link |
And what happened over time,
link |
it's like an evolutionary filter.
link |
It's just kind of who needs the plugin.
link |
And in fact, Sun mismanaged Java as a plugin.
link |
They thought, oh, Netscape is giving us the distribution vehicle
link |
and we don't care about the browser.
link |
It's just about getting Java out there.
link |
And that was a big miscalculation.
link |
They then tried, because Microsoft's killing Netscape
link |
after a few years, they tried getting into Microsoft.
link |
And you may remember there was a Sun Microsoft deal,
link |
which famously blew up.
link |
And Microsoft kicked Java out of Windows.
link |
And that's when they really pulled the trigger.
link |
I think they'd already evaluated it
link |
and liked it on AndersHelsberg's.net and C Sharp
link |
and decided we're going to just not have Java.
link |
We don't want any of that Sun stuff.
link |
We don't want the patent risk.
link |
We don't want, I'm not sure what all fights were about.
link |
There was some patent angle to it, I think.
link |
And up till then, Microsoft had been using Java components,
link |
like in Outlook Web Access, which had a lot of JavaScript
link |
to be a webmail, like hotmail, like user interface.
link |
They had to call the mail server through HTTP
link |
and they used a Java object to do this.
link |
And when they gave the boot to Sun,
link |
suddenly, the left hand gave the boot and the right hand said,
link |
we better do something else in Outlook Web Access.
link |
What are we going to do?
link |
And they said, let's just add an ActiveX component,
link |
which is their own native way of embedding things in languages.
link |
And it'll be what became XML HTTP request,
link |
which is now a web standard for calling asynchronously.
link |
And it's been replaced by the Fetch API
link |
and HTML5 or HTML living document.
link |
But this whole lineage goes back to Java
link |
being successfully the loser and getting kicked out.
link |
And after Microsoft kicked it out, it was a plug in
link |
and you would find it required for smart card banking
link |
in Nordic countries where that was mandated by law
link |
but really didn't get used much.
link |
Or there were pilots who used it for flight information.
link |
But Flash, which Netscape could have bought
link |
but fortunately didn't.
link |
Fortunately didn't.
link |
We would have screwed it up.
link |
What year were we talking about with Flash?
link |
I think after that PO, so it was probably late 95.
link |
Flash was around. Was it Adobe? No, it wasn't.
link |
No, it was called Future Splash.
link |
And it was these brothers, Jonathan Gay, I think his name was.
link |
He came knocking and the marketing guy at Netscape
link |
who was screening the technology partners
link |
or wannabe acquisitions was brutal
link |
and just everybody wanted to get in on the Netscape stock gravy train
link |
and he sent them packing and they ended up selling to Macromedia.
link |
And Macromedia was where Flash was created.
link |
And the good thing about Macromedia was it was a tool company.
link |
So it invested in the best ideas I think
link |
which are still somewhat lost to us of Flash.
link |
The timeline animation has sort of been immutable function over time.
link |
They had the tooling around that too.
link |
Like the Dreamweaver, there's a Flash.
link |
Flash director, there were a bunch of them.
link |
Yeah, I mean, yeah, that was a great...
link |
Flash builder was one of the last ones.
link |
These tools were used by real artists and special effects people
link |
All the restaurant websites around 2005 were done in Flash
link |
which was, you know, we were trying to do HTML5 at the same time
link |
that was the Firefox era.
link |
We were trying to make the web capable enough.
link |
You didn't need Flash, but if you recall, you go to a restaurant
link |
and it's like, this is kind of like a game or something.
link |
It's like a Flash.
link |
All the font looks small.
link |
You didn't like Flash from the beginning.
link |
This doesn't feel right.
link |
I actually admire Flash's technology.
link |
I'm pretty pragmatic about these things.
link |
I realized that, you know, it doesn't matter if your delta bad hand
link |
like JavaScript was a rush job or if you have Flash as a plugin
link |
and you can invest in the tools and make it pretty good.
link |
You should make it better for your users and grow it as best you can.
link |
And what happened with the browser due to Microsoft's monopoly abuse
link |
for which they were convicted and, you know, even after that
link |
until I think Firefox and then Chrome was people kept saying,
link |
oh, the web can't do X.
link |
You have to have a plugin.
link |
You have to have a new approach.
link |
Well, we'll clean the slate and have a new web.
link |
And everyone who said that failed.
link |
And the reason they failed is because there's too much value in the web,
link |
this huge network, and the worst is better principle means that you can
link |
not only start bad, which they all sneer at, but get on first
link |
and get wide distribution and get sort of evolutionary advantage
link |
in the priority of place, but you can also improve it over time.
link |
So if you're going to improve flash and for some reason, flash is now
link |
Steve Jobs said you can't have flash on the iPhone.
link |
That was probably the death now.
link |
Put your energy into JavaScript.
link |
And that happened, right?
link |
So we did things at Mozilla with Adobe to improve,
link |
which bought my macro media, to improve flash and to improve
link |
the version of JavaScript that was in flash.
link |
We tried to standardize that.
link |
I'm getting ahead of myself.
link |
Can we just rewind to the magical, like, you know,
link |
it's a special moment in the history of all of computing.
link |
We'll talk about it later, but it's arguable.
link |
It's possible that the entirety of the world will run
link |
in JavaScript at some point.
link |
It's like those days, it would be interesting if you could just
link |
describe actually zooming in on the how the cake was baked
link |
from the several days that you were working on it.
link |
What was in your mind?
link |
How much coffee were you drinking?
link |
Why the freaking out?
link |
I'll try to remember it.
link |
I mean, you're right.
link |
There are these pregnant moments you see in hindsight.
link |
Maybe they're overrated, but like,
link |
Hegel sees Napoleon on horseback at Yena and says,
link |
there's the world spirit on horse.
link |
And I knew that there was a chance to do it.
link |
Mark knew and he was my, you know, executive sponsor
link |
and he was the one, you know, sort of brainstorming
link |
how the JavaScript should be right there in the page.
link |
That was important for him to say that because I thought so too.
link |
But a lot of people were like, well,
link |
you can't write programming language in the middle of the markup.
link |
And indeed there are problems if you did it naively,
link |
you'd see the code laid out as like random gibberish.
link |
So I had to figure out how to hide that.
link |
That was a challenge.
link |
Is that a breakthrough idea?
link |
I mean, so you and Mark thinking about this idea
link |
that you just inject code in the middle of the markup.
link |
Yeah, I was considered kind of heretical.
link |
There was an SGML guru, I forget his name,
link |
but he corresponded with me and at first he was angry.
link |
He's like, you should have used a marked section.
link |
Why didn't you use a marked section?
link |
And I said, well, SGML marked sections are not part of HTML,
link |
by the way, and they're not supported in the browser.
link |
And so I did some hack that was equivalent.
link |
And over time you could do the proper SGML thing.
link |
But eventually he came around and it was again,
link |
sort of a revolutionary necessity.
link |
It was almost like introgression, like, you know,
link |
the idea which Lynn Margulies, I think, helped get across
link |
that we have to consider mutualism and biology
link |
that maybe, you know, mitochondria were ancient prokaryotes
link |
that got into the cell and became beneficial.
link |
Somehow the same sort of thinking applies.
link |
You have to embed JavaScript in HTML.
link |
It's going to be a good virus.
link |
The code becomes data in a sense that just gets carried along.
link |
But is there the side of the,
link |
you were focusing on the Nescape at that time.
link |
Doesn't the browser have to support,
link |
interpret correctly this mix of HTML?
link |
And whatever code?
link |
I had to hide it from old browsers, including Netscape 1.1,
link |
which was predominant then.
link |
So I used an HTML comment.
link |
But inside the container that comment lived in,
link |
the script tag, which is a new element,
link |
I could make different semantics in Netscape 2,
link |
where those HTML comment delimiters,
link |
instead of being multi line brackets,
link |
became one line, or essentially one line.
link |
So you wrote, so JavaScript was written,
link |
the programming language was written as a comment?
link |
A comment for old browsers and a set of brackets
link |
that were ignored with real code for new.
link |
And it was this two way comment hiding hack,
link |
as I called it, that was absolutely necessary
link |
for us to get off the ground.
link |
We couldn't have bootstrapped JavaScript without it.
link |
We didn't have scripts that were loaded from a separate file.
link |
The only scripts in Netscape 2 were in line in the document.
link |
What were the challenges here?
link |
Like what, you know, typing,
link |
what were the choices you were thinking about?
link |
Garbage collection?
link |
I didn't have time to write a garbage collector.
link |
So I just, I didn't at first.
link |
So the thing was using essentially arenas,
link |
or what GNU calls object pools,
link |
and just would run out of memory eventually.
link |
And I added reference counting in a hurry
link |
after the 10 days in which I hacked.
link |
So after I was in the server team doing HTTP11
link |
and thinking about the language,
link |
I finally got transferred to the client team in early May.
link |
And that's when I, you know, I got the go sign from Mark,
link |
and it was like, we can't wait,
link |
because people inside Netscape are doubting,
link |
even people inside Sun are definitely doubting.
link |
Bill Joy was the champion, but he was like alone in that,
link |
in seeing there was a role for JavaScript as the,
link |
as I call it, the sidekick language,
link |
robbing the boy hostage.
link |
Frank Miller put it in The Dark Knight Returns,
link |
that there was this silly little language
link |
that would be the glue language,
link |
and it could become important over time.
link |
And you were better off having that complementarity,
link |
that pairing of languages, just like Microsoft stacked it
link |
with Visual C++ and Visual Basic.
link |
So what was the big moment of I'm done?
link |
So I had to do a demo.
link |
I forget the dates.
link |
I think I, for a history of programming languages,
link |
paper that Alan Worsbrock did with my help,
link |
he did a lot of the writing.
link |
I think it was the 10 days from like Thursday evening
link |
through to the following weeks, you know,
link |
the whole of that week and then into Monday.
link |
Did you get sleep?
link |
And I was really going fast,
link |
and I'd already used a lot of C compiler
link |
and front end compiler knowledge
link |
that I'd gained from undergraduate school.
link |
When I started getting into computing
link |
as a Renegade Physics major,
link |
people were formalizing more efficient
link |
bottom up grammars, parsers for bottom up languages.
link |
Really, LALR1 was the big thing.
link |
And I studied all this and learned how to parse them.
link |
And in the end, if you're doing C languages,
link |
you often do what Dennis Ritchie did anyway,
link |
which is the recursive descent parser.
link |
You can hand code it.
link |
And I did that for JavaScript in a blazing hurry.
link |
Mostly got it right.
link |
I didn't have precedence inversion problems or other bugs,
link |
but I copied a lot from Java NC.
link |
And I tried to keep things simple,
link |
like the equality operator in those 10 days sprint
link |
between two objects of different dynamic type said,
link |
no, they're not equal.
link |
Their types are different.
link |
They're not too different.
link |
And then after that, I had internal early adopters
link |
and they were using JavaScript to match a number
link |
against a database field that had been stringized.
link |
And they said, oh, can we just have implicit conversion?
link |
Like an idiot, I agreed.
link |
I gave them what they wanted.
link |
I was trying to please them and get adoption.
link |
And that broke what equivalence relation nature there was
link |
to the double equal.
link |
There's some edge cases with not a number
link |
that break that up too.
link |
But it really broke it.
link |
Having implicit conversions in the operator is something
link |
that people still roast me over.
link |
So let's talk about two things.
link |
One, it sounds like the comparison operator,
link |
the equality operator is the thing that you regret.
link |
So maybe can you...
link |
So what is the biggest thing you regret in those 10 days?
link |
And what is the biggest thing you're proud of?
link |
So that making it sloppy came after the 10 days.
link |
And my lesson there, which I've tweeted is,
link |
when people come to you saying,
link |
can you please make it sloppy or add this cute feature?
link |
The answer should be no.
link |
And I should have known that because I think,
link |
Nikolas, one of my heroes said,
link |
the essence of design is leaving things out.
link |
But during the 10 days, I also, like I said,
link |
I was in such a hurry, I left out garbage collection,
link |
came back to haunt me, but I got reference counting in in time
link |
that people weren't running out of memory right away
link |
on long live jobs.
link |
What happens when you don't have garbage collection
link |
and you have objects?
link |
Well, you just run out of memory.
link |
And, you know, at first you write a short script
link |
and the page doesn't last long or it doesn't do a lot
link |
But if you're writing a game or something
link |
and you're doing event based allocation,
link |
you run out of memory.
link |
And this was noticed in the summer of 1985
link |
and people were like, what's going on?
link |
I got to go back and do reference counting.
link |
And then the problem with reference counting is
link |
you're writing the language in the runtime in C,
link |
an unsafe language.
link |
And if you're reference counting
link |
and you overflow the counter, you mismanage it
link |
so it goes high, it gets stuck high,
link |
you leak memory again and you run out.
link |
If you underflow it, you pre memory that's still in use.
link |
And even then we knew what all the security hackers
link |
came to know that you therefore have a,
link |
potentially a remote code execution vulnerability.
link |
Because this was before things like non executable
link |
heap memory and stack defenses against taking over memory.
link |
So if you can, from the remote side,
link |
write some HTML and JavaScript that just happens
link |
to exploit a bug in memory safety,
link |
like it causes JavaScript to underflow reference counter.
link |
And the script still has its hands on that object.
link |
And it's trying to call a method on it.
link |
And there's some kind of lookup function table
link |
in the object, but you've managed to stuff the heap
link |
with strings that forge their own lookalike
link |
for the function table.
link |
You can call some other code.
link |
And this was a problem right away.
link |
So security, JavaScript up the ante.
link |
Java had this problem too, but in its own VM.
link |
And it just was, you know, a separate headache
link |
for son to worry about.
link |
We had this problem in Netscape right away.
link |
So Netscape 2 came out after my 10 days
link |
and after these, you know, follow along work
link |
to embed JavaScript better in the browser
link |
and to add garbage collection through reference counting.
link |
Really, I call it reference counting.
link |
And get it shipped.
link |
We had a bunch of dot releases where we fixed security bugs
link |
But what is the thing you're, you know, when you sit back
link |
on a porch and just look out into the sunset,
link |
what are you most proud of from those 10 days?
link |
I think the first class functions shines.
link |
I think especially since Java didn't have it
link |
and it was somewhat unusual.
link |
The scheme made it in somehow at the end of the day.
link |
I mean, people complain because scheme has, you know,
link |
It has, you know, six or seven special forms.
link |
It has hygienic macros.
link |
It has sort of a beautiful, complete set of forms
link |
to make the Lambda Calculus pleasant to use in practice.
link |
And JavaScript is, you know, kind of a multi paradigm
link |
Just in a small tangent, you mentioned Mark Andreessen.
link |
It sounds like in Bill Joy, but staying on Mark,
link |
it sounds like he had an impact on you
link |
in that he sort of believed in what you were doing there.
link |
Can you talk about, like, what role Mark had in your life?
link |
We would meet at the Pencila Creamery in downtown Palo Alto.
link |
And Mark was just fresh out of, you know, grad school
link |
or whatever he was doing.
link |
And he was a big dude and he got fitter later.
link |
He would order giant milkshakes and burgers.
link |
And we would meet there in brainstorm about what to do.
link |
And it was very direct because we didn't have much time.
link |
The sort of, we didn't talk about it.
link |
The implication was Microsoft was coming after us.
link |
Mark was saying things boldly, pre IPO,
link |
like Netscape plus Java kills Windows, right?
link |
Make a browser programmable.
link |
It becomes the new runtime for programs.
link |
It's the meta OS or it's the replacement OS.
link |
But he still saw value in JavaScript.
link |
Even though he was saying that and Java was the big name,
link |
hence the trademark license,
link |
he saw JavaScript as important.
link |
And he even thought, what if we got,
link |
I told us another interviews, I can say,
link |
he thought, what if we had my friend Kip Hickman,
link |
who'd been at Netscape from the beginning
link |
and who was a kernel hacker at SGI when I joined,
link |
he'd started writing his own JVM before we consummated
link |
the Sun deal and got our hands on their code.
link |
And the Java compiler, Java C,
link |
which Arthur Van Hoff had written, very nice code,
link |
was all written in Java.
link |
It was self hosted or so called bootstrap.
link |
And so we could use that as soon as Kip's Java VM
link |
would run the bytecode from the Sun JVM
link |
running the self hosted compiler to emit the bytecode.
link |
So once we could bootstrap into Kip's VM,
link |
we wouldn't need Sun and Mark was like,
link |
well, maybe we can just ditch Sun with Kip's Java VM
link |
or if you're at JavaScript VM, we,
link |
now we need graphics.
link |
So Mark was thinking far ahead because he knew
link |
you could do things with HTML and images,
link |
but at some point you really want this.
link |
Like 10 NEMA graphics or three dimensional.
link |
Like even SGI had already started its downfall
link |
because the first four to VLSI team there
link |
had gone off to do 3DFX and all these other companies
link |
that made the graphics card on your PC, right?
link |
Doom was big and quake.
link |
And so we were all playing quake.
link |
I was old, so I was terrible.
link |
But why not put that graphics capability on the web?
link |
And in fact, it finally happened at Mozilla
link |
with Firefox era with Vlad Vukicovic
link |
taking OpenGLES and reflecting it as WebGL.
link |
But OpenGLES is the mobile version of OpenGL,
link |
which is a standard based on SGI GL.
link |
So there's this whole lineage of graphics libraries
link |
or really graphics languages for what became the GPU.
link |
And Mark was thinking ahead.
link |
It's like, we need graphics too.
link |
And I thought, okay, I can try to get somebody I knew
link |
at SGI, but he's a grad student at MIT.
link |
He was studying at the Barbara Liskoff.
link |
He laughed when he heard about this later.
link |
Andrew Myers, he's at Cornell long time.
link |
I think he's a full professor.
link |
And Mark said, great, well, get him.
link |
I'm not sure he's going to come.
link |
We don't have any real money, you know, stock options.
link |
And they did the Sun Deal.
link |
So Kip Nobley put aside his own JVM
link |
and we used the Sun JVM.
link |
So that was an ambitious period.
link |
And Mark was very generative because he was pushing hard.
link |
He was ambitious and he wanted to have Netscape
link |
possibly be in control of the ball.
link |
Maybe you can speak to the dance of Netscape
link |
versus Internet Explorer.
link |
You've thrown some loving words towards Microsoft
link |
throughout this conversation.
link |
But that's a theme with, I mean, Steve Jobs
link |
had a similar sort of commentary.
link |
From a big sort of philosophical principle perspective,
link |
can you comment on like the approach
link |
that Microsoft has taken with Internet Explorer
link |
from IE1 to Edge today?
link |
Is there something that you see as valuable
link |
that they're doing in the occasional copying
link |
that kind of stuff or is it the world worse off
link |
because Internet Explorer exists?
link |
So I'm going to segment this into historical errors
link |
because I think Microsoft is today with Satya.
link |
It's quite a different company and what they're doing
link |
with Edge is different.
link |
But back then, Gates, you know, aggressive character,
link |
not really original in my view, not an originator.
link |
Steve Jobs famously said once, he doesn't have any taste
link |
and I don't mean this in a small way.
link |
Apple at the time had beautiful typography
link |
and ligatures and kerning and the fonts looked great.
link |
And Windows had this sort of ugly system font
link |
that was carefully aligned with Pixel
link |
so it didn't get into anything.
link |
Why is it, I'm sorry to keep interrupting,
link |
but why was Internet Explorer winning
link |
throughout the history of these competitions?
link |
Distribution matters more than anything.
link |
And this is why, you know, even now
link |
we're seeing in the browser worse Edge doing better
link |
because it's being foisted on people of Windows.
link |
We have Windows 10 boxes at home.
link |
Windows 7 boxes or laptops we keep running to
link |
because we don't connect them to the Internet generally.
link |
But once you have that operating system to hold,
link |
you can force, you know, Edge.
link |
And Apple did it with Safari too.
link |
It's not unique to Microsoft.
link |
Distribution matters.
link |
And that's why I think IE was going to win.
link |
That's why everybody at Netscape felt we're doomed.
link |
This was something Michael Toy and Jamie Woodson were doomed.
link |
But for a while there, we had a chance
link |
and we innovated in Netscape too.
link |
We did a big platform push, Java and JavaScript
link |
and plugins, more plugins and, you know,
link |
more HTML table features and really started making
link |
a comparable stack out of what were pretty static web languages.
link |
And even in the beta release of Netscape 2,
link |
people were using JavaScript to build what you would call
link |
single page applications like Gmail.
link |
And they were using JavaScript locally to compute things
link |
and to call the server on a hidden frame in the background.
link |
So it was prefiguring a lot of what came later as AJAX
link |
or Dynamic JavaScript, Dynamic HTML.
link |
So people saw that.
link |
Even when they saw it.
link |
That's kind of, I don't know,
link |
but from my perspective, that seems quite brilliant.
link |
It seems like really innovative
link |
that you would have code running in the browser.
link |
It did impress me with something,
link |
which I learned later about from Eric von Hippel of MIT,
link |
which is user innovation networks, lead user effects,
link |
that throwing out JavaScript,
link |
even though we weren't doing open source,
link |
we were doing beta releases early and permissively with Netscape.
link |
Getting early developer feedback, absolutely critical.
link |
I did some of that with SGI with some of the products I worked on,
link |
but it really came to the fore in Netscape.
link |
And that, you know, culminated in Mozilla,
link |
where you're dealing with developers all the time
link |
and early adopters, lead users.
link |
But lead users helped improve JavaScript,
link |
even in those last few betas where I could hardly change things.
link |
I was under pretty rigid change control.
link |
So we're talking about just a small collection of individuals
link |
that are just like upfront.
link |
My name is Bill Dorch. You can find his work in the Web Archive,
link |
It's a single page application.
link |
It's an artist gallery of mountain art.
link |
He uses JavaScript.
link |
It doesn't quite work. He uses JavaScript locally.
link |
He uses a local database.
link |
What you would think of now is JSON,
link |
but it's all pure JavaScript code,
link |
a bunch of objects being constructed.
link |
So how is, if you can do sort of a big sweeping
link |
progress of JavaScript,
link |
how has JavaScript changed over the years,
link |
nearly 10 days, with a quick addition of garbage collection
link |
and fixes around security?
link |
How is this evolution that now it's taken over the world?
link |
It's been a bumpy ride because the standards body
link |
got shut down after Microsoft, I think,
link |
took over the web and then felt punished
link |
by the USV Microsoft.
link |
Can you speak to the standard body?
link |
That was a fun ride, too, because Netscape had taken
link |
the lead with the web and HTML innovations
link |
like frames and frame sets, tables.
link |
And the W3C was sort of off even then
link |
in SGML land heading toward XML la la land.
link |
I'm going to be a little harsh on it.
link |
What's SGML? I'm sorry.
link |
SGML was the precursor of markup language to HTML,
link |
or it was sort of the more extensible
link |
standard generalized markup language.
link |
It was a pointy brackets, but it had
link |
all sorts of elaborate syntax for doing
link |
different semantics.
link |
And this is why, I think, TBL and others
link |
who wanted to do the semantic web then
link |
took XML forward, but they had this,
link |
or some of them anyway, had this strange idea
link |
they could replace the web with XML
link |
or that they would upgrade the web to be XML.
link |
And it couldn't be done.
link |
Worse is better had concrete meaning.
link |
The web was very forgiving of HTML,
link |
including sort of minor syntax errors
link |
that could be error corrected.
link |
Error correction isn't generally done
link |
in programming languages because...
link |
That's another amazing thing about HTML.
link |
It's more like biology than programming.
link |
And so, XML was, in its standard form,
link |
super strict and could never have admitted
link |
the kind of users who were committing these errors.
link |
And the funniest part was Microsoft said,
link |
hey, we're doing XML, but the way they put it
link |
in Internet Explorer under the default media type,
link |
put it through the HTML error corrector.
link |
So they kind of bastardized it to make it
link |
popular and usable and accessible.
link |
And so, XML as a pure, you know,
link |
thing was never going to take over.
link |
And what W3C was kind of not fully functional
link |
because Nescape wasn't cooperating with them.
link |
We thought about where to take JavaScript
link |
and we realized our standards guru Carl Cargo
link |
realized there was a European standards body
link |
that had already given Microsoft fits
link |
by standardizing parts of the Windows 3.1 API,
link |
which European governments insisted on.
link |
They said, Microsoft, we can't use your operating system
link |
without some standards.
link |
And Microsoft said, you know, here's our docs.
link |
And the government said, no, we need a European standard.
link |
So this body called the European Computer Manufacturers Association,
link |
ECMA, which eventually became global
link |
and became a proper noun instead of an acronym.
link |
It's just one capital E now with a lowercase CMA.
link |
And as one of the early Microsoft guys I met
link |
when we first convened a working group to talk
link |
about JavaScript said, it sounds like a skin disease.
link |
But it gave, I mean, maybe you'll speak to that,
link |
but it gave the name to JavaScript of ECMA script.
link |
That was the standard name because Java was a trademark of Suns.
link |
They were so aggressive, they were sending cease and desist letters
link |
to people whose middle European heritage meant their surname was Javonco
link |
and they called their website Javonco.com
link |
and Sun would send them a letter saying, you're using JAVA
link |
at the start of your domain name.
link |
You must cease and desist.
link |
I love marketing more than anything else in this world.
link |
So ECMA script and now is popularly named as ES plus version.
link |
I would say people use JS more than anything.
link |
People still say JavaScript.
link |
JavaScript is in all the books.
link |
So I mean, when you're referring to, it's usually JavaScript
link |
and when you want to refer to a version of JavaScript,
link |
you'll say ES6, ES5.
link |
Or now they've gone to years, which is kind of confusing
link |
because it's an offset of 2009, ES6, ES 2016.
link |
Yeah, it doesn't match the years perfectly.
link |
Yeah, so what were the choices made
link |
and how did JavaScript evolve here?
link |
So we took this new standards body,
link |
which we thought was sort of a proven record of standing up to Microsoft.
link |
But Microsoft sent a lot of people.
link |
They sent some people who were pretty good
link |
and then when they realized that I was there and Netscape
link |
was not going to just bend over and do whatever they wanted,
link |
they sent somebody really good.
link |
And this smart guy, he did a lot of the work on the first draft of the spec.
link |
Sean Katzenberger, he's left Microsoft.
link |
He even did what I sort of did.
link |
He told his bosses, stop bugging me to do other things.
link |
I'm focused on this because it took a lot of focused work
link |
to create the first draft of the spec.
link |
And I was still holding, I was spending almost all the plates.
link |
I had like part time help in certain areas
link |
and on the front end integrations, I had the front end guys.
link |
But I couldn't take as much time as Sean was to write the draft spec,
link |
but I had to participate because I was essentially helping write down
link |
what the language did.
link |
And in areas where we didn't like what it did
link |
and Microsoft didn't agree, we sometimes got away with slight changes.
link |
And that's the story of standards.
link |
You have different implementations.
link |
And depending on their market power, they interoperate
link |
where you have agreement and where they don't,
link |
the dominant one usually sets the de facto standard.
link |
And then you should probably reflect that into the de jure standard.
link |
And this happened with JavaScript.
link |
Over time as Netscape went down and Microsoft went up,
link |
we did the first edition of the standard codified in 1997 in France.
link |
We had a trip to Nice, which was very memorable.
link |
For any interesting reason or just because it's Nice?
link |
And ECMAS European and IBM and others were, you know, there.
link |
Mike Cowleshawn, IBM fellow was a Britisher.
link |
And the guy who ran ECMAS at the time,
link |
Jan van den Belb, was quite a raconteur and a very fun guy.
link |
And he had us out for, you know, the great, you know,
link |
the Fuitimer, the Bouillabaisse.
link |
Was the standardization process beautiful or painful
link |
that those early days you as a designer of the language?
link |
It was painful because it was rushed.
link |
Now Guy Steele was contributed by Sun.
link |
So even more than Sean, you had this giant brain Guy Steele
link |
helping bringing some of that scheme magic.
link |
He even brought Richard Gabriel for fun.
link |
And Richard wrote the fourth clause of the ECMAS standard,
link |
which was kind of an intro to what JavaScript is all about.
link |
So we had some really good people and we didn't fight too much.
link |
There was a lot of tension where I was fixing bugs
link |
and I was late to a meeting and Sean Katzenberger
link |
or Microsoft was actually mad, like where is he?
link |
And when I got there, I saw that only he saw this sort of
link |
off by one bug in somewhere in the spec.
link |
And then I saw it too.
link |
And I said, there's a fence post bug there.
link |
And then we kind of locked eyes and we realized
link |
we were on the same page and we kind of, he wasn't mad anymore.
link |
What were the features that are being, like,
link |
struggled over and debated and thought about?
link |
It was mainly writing down what worked
link |
and what we thought should work in the edge cases
link |
that didn't interoperate or that seemed wrong.
link |
But we were already laying the groundwork
link |
for the future editions that I was already implementing.
link |
I was still trying to lead the standard
link |
by using the dominant market power to write the code
link |
that actually shipped.
link |
So the de facto standard would lead the deserter standard.
link |
And I was putting in the missing, you know,
link |
function forms that I didn't have time for in the 10 days.
link |
So this is the engineering mindset versus the theoretician.
link |
So you didn't want to create the perfect language,
link |
but one that was the popular and shipped and all that kind of stuff.
link |
And you could say there was, I was standing out of the shoulders of giants.
link |
So there was a staged process where I had to hold back things
link |
that were well designed by others in other languages
link |
that I could imitate, but I couldn't do them all in the 10 days.
link |
So they came in in 1996 and 97.
link |
And they came into the third edition of the standard,
link |
which was finalized in 1999.
link |
But at that point, Netscape had been sold to AOL
link |
and was, which was a decent exit considering
link |
and, you know, had previously been mercilessly crushed.
link |
Netscape was selling the browser along with server software
link |
that it had acquired after its IPO.
link |
And Microsoft was just underpricing it.
link |
So there was no way to compete with that.
link |
Microsoft was also, you know,
link |
making Internet Explorer the default browser in Windows,
link |
which is called tying in antitrust law.
link |
And they were doing even more brutal things.
link |
There was a famous investor.
link |
He did very well on Google.
link |
So he's a billionaire, Ram Shri Ram.
link |
And he was sales guy or head of sales at Netscape.
link |
And he got off the phone looking ashen faced
link |
after a compact call and said,
link |
Microsoft just told us they're going to pull
link |
our Windows license if we ship Netscape
link |
as the default browser.
link |
So there is some bullying going on.
link |
It was totally material in the antitrust case.
link |
But JavaScript escaped into the standard setting
link |
where there was fairly good cooperation.
link |
Microsoft had a really good guy on it
link |
and, you know, Guy Steele was there for a time
link |
and there was some good work.
link |
But after the antitrust case and
link |
Netscape kind of dissolving into AOL
link |
and not really going anywhere quickly,
link |
Mozilla took years to really bring up.
link |
The standard froze.
link |
And by 2003, even though they'd been sort of
link |
noodling around with advanced versions,
link |
JavaScript 2, I'd given the keys to the kingdom
link |
to another MIT grad, Baltimore Horwatt.
link |
It's still at Google, I think.
link |
He won the Putnam in 86.
link |
So he's, you know, very mathematical.
link |
He designed this successor language,
link |
JavaScript 2, but it only showed up in mutated form
link |
in Microsoft's ASP.NET server side.
link |
And it didn't last there.
link |
And it showed up in Flash.
link |
And that's what became ActionScript 3.
link |
And then Flash was declined.
link |
And so how did we arrive at ES6 where it's like,
link |
there's so many, where everyone,
link |
okay, there's this history of JavaScript
link |
that people were just like, cool when you're like
link |
having beers to talk crap about JavaScript.
link |
Everyone loves to hate, like people who are married
link |
say, ah, marriage sucks, is they just want to get,
link |
let off some steam, even though everyone uses the language.
link |
But ES6, it's become this like reputed,
link |
like it fixed major pain points, I think.
link |
It added things to the language and added something
link |
that was already ES5 strict mode,
link |
but made it implicit in class bodies and module bodies.
link |
It was a big jump, but it accumulated some of the ES4 designs
link |
that we'd done with Adobe for what we hoped would be
link |
the fourth edition of ECMAScript,
link |
that were supposed to fold in some of these old JavaScript
link |
two ideas that had come into ActionScript 3.
link |
So you look at the family tree and you see these forks
link |
and the main ones are the ones that go into Adobe Flash,
link |
acquired from Acromedia, and the one that went into the
link |
server side of Microsoft's stack, which kind of died.
link |
And then trying to bring them back into the standard
link |
and not quite succeeding ES4 was mothballed,
link |
but all the good parts that everyone liked made it into ES6.
link |
And so that was a success.
link |
And I said earlier, I had the wrong year.
link |
I think it's 2015, so it's off by.
link |
Yeah, it was done, finalized in 2015.
link |
It took a little longer than we hoped,
link |
but because ES5 was 2009 and that was a smaller increment
link |
We skipped four again.
link |
And we had a split in the committee where some people said,
link |
We're going to work on incremental improvements,
link |
no new syntax in particular, they promised.
link |
Not quite true, but they added a bunch of interesting APIs.
link |
Alan Weir Sprock, my coauthor of that Hubble paper.
link |
And he was at Microsoft at the time.
link |
I ended up hiring Mozilla.
link |
He wanted to get to Mozilla and keep doing the sort of
link |
editor job of the JavaScript standard ECMAScript.
link |
And when we got ES6 done, it was a little late 2015,
link |
and we switched to year numbers.
link |
People still call it ES6.
link |
But if you remember, off by nine plus 2000.
link |
Yeah, ES6 is such a big job.
link |
Like you said, there's a thread that connects all of it,
link |
but ES6 is when it became this language that almost feels ready
link |
to take over the world completely.
link |
More programming in the large features,
link |
more features you need for larger teams.
link |
Software engineering.
link |
Microsoft did something smart, too.
link |
They, Anders and company, Luke Hoban,
link |
who's left Microsoft, also did TypeScript.
link |
And they realized something,
link |
I think that Galat Braka's also popularized,
link |
and he was involved in Darted.
link |
Google, if you don't worry about soundness in the type system,
link |
you don't try to enforce the type checks at runtime in particular,
link |
just use it as sort of a warning system, a tool time.
link |
You can still have a lot of value for developers,
link |
especially in large projects.
link |
So TypeScript has been a roaring success for Microsoft.
link |
What do you think about TypeScript?
link |
Is it adding confusion, or is it ultimately beneficial?
link |
I think it's beneficial.
link |
Now, it's technically a super set of JavaScript.
link |
So of course I love it, right?
link |
The shortest JavaScript program is still a TypeScript program.
link |
Any JavaScript program is a TypeScript program,
link |
which is brilliant because then you can start incrementally adding
link |
type annotations, getting warnings, learning how to use them.
link |
Microsoft had to kind of look around corners at the standard's body
link |
and guess how their version of modules or decorators should work.
link |
And the standard's body then may change things a bit.
link |
So I think they're obligated with TypeScript,
link |
either to carry their own version or to bring it back
link |
with incompatible changes toward the standard over time.
link |
And I think they've played generally fair there.
link |
There's some sentiment that why don't they standardize TypeScript?
link |
Well, they've been clear they don't want to.
link |
They have a proprietary investment.
link |
They've controlled the ball.
link |
And in some ways, you can say the same thing to any of the other
link |
big companies in the standard's body.
link |
Why doesn't Google standardize its stuff?
link |
So you think it'll continue being like a kind of a dance partner
link |
to JavaScript, to the base JavaScript?
link |
There's a hope that at some point, if they keep reconverging it
link |
and the standard doesn't break them and goes in a good direction,
link |
we will get at least the annotations syntax
link |
and some semantics around them.
link |
Because when you're talking about type annotations,
link |
they're generally on parameters and return values
link |
and variable declarations.
link |
They're cast operators.
link |
You want that syntax to be reserved and you want it to work
link |
the same in all engines.
link |
And this is where ideas like Gilad's Plugable Type Systems
link |
might be good, though then you could create the same problem
link |
you have with Blisp and Scheme where there's a bunch of macro libraries
link |
and they don't agree and you have conflicts between them.
link |
But the Plugable Type Systems could be one way to standardize this.
link |
What do you think about the giant ecosystem of frameworks in JavaScript?
link |
It feels like, because I mean, this is a side effect
link |
of how many people use JavaScript.
link |
A lot of entrepreneurial spirit, like create their own JavaScript frameworks
link |
and they're actually awesome in all different ways.
link |
And this is an interesting question about almost like philosophically
link |
about biological system and evolution and all that kind of stuff.
link |
Do you see that as good or should some of them die out quicker?
link |
I think maybe they should.
link |
Now, jQuery was a very clever thing.
link |
John Resick made this library that was sort of query and do
link |
and blended sort of CSS selector syntax with JavaScript
link |
sort of object graph or DOM querying and made it very easy
link |
for people to do things almost like they were learning jQuery
link |
as its own language, domain specific language.
link |
And that I think reflected in part the difficulty of using
link |
the document object model, these APIs that were originally designed
link |
in the 90s for Java as well as JavaScript.
link |
They're a very object oriented or even procedural.
link |
They're very kind of verbose.
link |
And it took like a constructor column,
link |
three different, you know, hokey pokey dances to do something
link |
whereas in jQuery, it's just one line, right?
link |
So that fed back finally into the standards.
link |
It didn't mean we standardized jQuery.
link |
It wasn't quite that concise.
link |
You find now with the modern standards
link |
that we were working on in the HTML5 sort of effort
link |
that things became simpler, the fetch API
link |
and the query selector API, document query selector.
link |
A lot of things can be done now in raw JavaScript
link |
that you would make more concise and terse in jQuery,
link |
but it's not bad. It's pretty good.
link |
Whereas in the old DOM of 15 years ago, it was just too verbose.
link |
So maybe the frameworks were born kind of because JavaScript
link |
lacked some of the features of jQuery.
link |
And so, like, now that JavaScript is swallowing
link |
what jQuery was, then the frameworks will,
link |
only the ones that truly add value will stick around
link |
and the other ones will die out.
link |
And that highlights also this division between the core language
link |
JavaScript, which can show up in other places
link |
like Node.js on the server side
link |
and the browser specific APIs
link |
or the document object model APIs,
link |
which are even managed by the W3C,
link |
the standards body that was off in XML,
link |
and we were doing real JavaScript standards in ECMA.
link |
And you have this division of labor,
link |
division of responsibility and division of style
link |
and sort of aesthetics and also speed.
link |
So the document object model really stagnated
link |
after Microsoft kind of deinvested in the web.
link |
And Microsoft did something in their haste
link |
in the spirit of Netscape, doing things quickly
link |
and getting on first called DHTML
link |
and some of their innovations that were like
link |
an alternative document object model
link |
didn't really get standardized until HTML5
link |
when we pragmatists set up for at the time
link |
Ian Hixon, he went to Google,
link |
Apple and Mozilla said,
link |
let's XML is not going to replace HTML.
link |
Let's standardize HTML5 based on all this good stuff
link |
including that DHTML variant, dynamic HTML5.
link |
HTML5, it feels like to me, maybe you can correct me
link |
like a beautiful piece of design work.
link |
It's not often with web stuff,
link |
you have this breath of just like,
link |
oh, whoever did this, they just feels good.
link |
What are your thoughts about HTML?
link |
Am I being too romantic?
link |
Are there fundamental flaws to it that I'm just not aware of?
link |
My old friend Hix, he did a great job.
link |
He was another renegade physics student
link |
and he was basically a QA guy at Opera,
link |
but he obviously trained physics student
link |
and someone who could write a Britisher.
link |
He developed test suites
link |
and he started thinking about them more axiomatically.
link |
Now, this can be good because you can sort of
link |
systematize in a way that makes a better HTML
link |
or you can get caught in the pragmatism of saying,
link |
well, we have to handle all these edge cases
link |
so we're just going to have sort of a test matrix.
link |
And if the matrix is large, it will not be beautiful
link |
by many people's likes.
link |
Everyone likes to minimize along their preferred dimensions,
link |
the seven special forms in scheme or whatever.
link |
But reality is HTML needs to be big.
link |
It's kind of shambolic.
link |
It's a creative, multi paradigm.
link |
And Hix, he did a good job, I would say, with a bunch of it.
link |
Other people came in, in the spirit of Ian Hix,
link |
and to do HTML5 work.
link |
And they've carried on that effort.
link |
So it's a mix of pragmatism,
link |
de facto standards from the past being sort of
link |
combined or written down for the first time
link |
and then rethought in a way that has a simpler syntax,
link |
like the Fetch API instead of XML HTTP request.
link |
This video, too, as well.
link |
It ultimately, it feels like maybe you can correct me.
link |
It feels like it was the nail in the coffin of Flash.
link |
Steve Jobs saying no Flash on the iPhone, in my opinion,
link |
was the actual stake to the heart.
link |
But, well, I'm not sure what trope you want to use.
link |
Flash was a zombie until just this year, right?
link |
I think last year was the end of Flash and main browsers.
link |
But Jobs really did the death blow.
link |
And yet, you're right.
link |
We had to make HTML5 competitive.
link |
I still don't think we got that beautiful timeline animation.
link |
The timeline things.
link |
You like the time.
link |
I mean, me, from, you know, I used to animate
link |
all kinds of stuff inside Flash.
link |
Plus, there's a programming element.
link |
It was a little bit, I don't know if you commented on that,
link |
but to me, it was a little bit like go to statement,
link |
like in a sense that there's a little bit too chaotic.
link |
Like it didn't, that OCD part of me as a programmer
link |
wasn't satisfied by Flash.
link |
It feels like there's bugs that were introduced
link |
through the animation process that I couldn't debug easily.
link |
Yes, I heard that too.
link |
So I'm doing the grasses greener thing here.
link |
The thing I liked about the animation model
link |
was that it was this immutable function of time.
link |
So you could time warp and you could,
link |
if you dodged these bugs or worked carefully,
link |
you could really make it sing in ways
link |
that I think still a little challenging
link |
with the web animation standards,
link |
or just using Rock Canvas and WebGL.
link |
But there's so many tools now that maybe it doesn't matter.
link |
And yet we had to, you know, do video.
link |
We had to do WebGL and then evolve it.
link |
We had to do web audio.
link |
But once we did all these things that helped Flash die
link |
thanks to Steve Gobs, we had something that
link |
people didn't realize.
link |
We had that vision that Mark and Jason had,
link |
this graphics capable to the metal portable runtime.
link |
And we at Mozilla realized this,
link |
and we saw JavaScript was something that you could compile to.
link |
Adobe had somebody in the Adobe Labs doing this too.
link |
He had a project called Alchemy.
link |
We had somebody who's now at Google,
link |
Alon Zakai, who did his own LLVN based compiler
link |
that would take C or C++ and it would emit JavaScript.
link |
I think this is crazy going from this sort of machine types,
link |
low level, you know, controlled memory allocation language
link |
to this garbage collected, dynamically typed,
link |
high level, higher level language.
link |
But Alon sort of just phenomenologically carved nature of the joint
link |
and found the forms that were fast in JavaScript.
link |
And then with Dave Herman, who I'd recruited from
link |
Northeastern University, who was a type theorist,
link |
and Luke Wagner, who's still at Mozilla,
link |
who was the compiler guy and the JIT guy,
link |
they figured out how to codify what Alon had done
link |
into a typed subset of JavaScript called Asm.js.
link |
And this is a strange thing to think about
link |
because it doesn't have new syntax.
link |
The types are casts that occur in
link |
dominator positions in the control flow graph.
link |
So it's like a hack on JavaScript and it's a subset.
link |
It uses those bitwise operators that I talked about
link |
copying from Java to basically cast
link |
numeric types, which are double precision flowing point,
link |
And so inside JavaScript, in the kernel semantics,
link |
And if you use these operators, if a compiler
link |
emits them in the right places, you can then treat them
link |
as typed values, typed memory locations.
link |
And you can type check your program.
link |
You can not only type check it, you can compile it.
link |
And this is all in sort of linear time.
link |
Oh, and you can compile it to have deterministic performance.
link |
It doesn't touch the garbage collector.
link |
It calls a bunch of functions that come from the C functions
link |
or C++ code that you're compiling.
link |
And you can make the Epic Unreal Engine go
link |
in 30 frames a second.
link |
And when we did this in 2013 in the fall,
link |
Tim Sweeney, I met, didn't think it could be done quickly.
link |
He thought it would take years.
link |
And the team went to Raleigh to Epic and in four days
link |
they had Unreal Engine ported by pressing a compile button.
link |
But they had to have WebGL, which came from OpenGL.
link |
Yes, it came from OpenGL, which came from Silicon GraphicsGL.
link |
They had to have WebAudio so they could map OpenAL, which
link |
was another audio library standard, to WebAudio, which
link |
was kind of a Chrome idiosyncratic thing.
link |
But they could make it work.
link |
And they had to have Asmjs for fast C++ to JavaScript.
link |
And if you didn't have that fast compiler step,
link |
the JavaScript you'd write by hand trying to do
link |
an Unreal game would be too big and too slow.
link |
It would touch the garbage collector.
link |
It would not keep up with 30 frames a second on the hardware,
link |
So we demoed that at, this must have been fall 2012
link |
now that I think about it.
link |
Because we demoed it at GDC Game Developer Conference 2013.
link |
And people were stunned.
link |
It's like Unreal Engine, Unreal Tournament running in my browser
link |
No plug in, no flash, no Java, no.
link |
So where does the early days of, because JavaScript now
link |
is able to run basically on par with a lot of the C++?
link |
And even before then, you had the fast JavaScript VMs.
link |
In 2008, when Chrome came out, just before it came out,
link |
Mozilla, my friend Andreas Gallo and I, and others
link |
hacked out Trace Monkey, our trace based JIT.
link |
The Squirrel Fish Extreme team at Apple did their JIT.
link |
And we were all competing on these crazy performance
link |
It was a little bit too much tuning to the benchmark.
link |
But JavaScript started getting fast.
link |
And developers started noticing it.
link |
But it was still kind of its own high level language
link |
with garbage collection.
link |
The Asm.js step helped us go further.
link |
Because until we really proved the concept, people were still
link |
saying, well, JavaScript's OK.
link |
It's getting faster thanks to V8.
link |
Everybody gave Google credit, especially Google.
link |
But we need something to kill flash.
link |
Let's use the portable native client code that Google
link |
had acquired, native client, which is a separate lineage
link |
that we're taking basically C code, compiling it into a
link |
software fault isolated container or some sort of using
link |
some kind of virtualization technique.
link |
And maybe it can even be in process and still be memory
link |
That would be awesome.
link |
But they ended up using process isolation too.
link |
And that kind of weakened it.
link |
And in the end, it was like, portable native client, OK,
link |
meet the new boss.
link |
Name is the old boss.
link |
This is the Google Flash, right?
link |
But when we did Asm.js and we showed Unreal Engine working,
link |
I think it was only a matter of time before Google threw
link |
Everybody agreed in the spring of 2015, we're going to take
link |
what was proven by Asm.js and make a new syntax, a binary
link |
syntax that's efficient, that loads into the same JavaScript
link |
VM that JavaScript loads into.
link |
So there'll be two source languages, one VM, very important,
link |
one garbage collector, one memory manager, one set of
link |
And that's called WebAssembly.
link |
And that's the successor to Asm.js.
link |
And it's important that it have binary syntax because at the
link |
end of the day, especially on mobile, if you're downloading
link |
JavaScript, even if you're using LZ compression on the
link |
wire, that's cool.
link |
But you've got to blow it out into memory and then parse the
link |
silly eight character function keyword that I picked when I
link |
could have used something shorter.
link |
I picked it because of AUK, the Unix tool.
link |
So anyway, I want to, but I'm not following the AUK thread.
link |
I said surprising to you that how damn fast JavaScript is
link |
I mean, you've been through the whole journey.
link |
I know every step of the way, but is it like, I mean, it's how
link |
it feels incredible.
link |
But I knew, so the funny thing is computer science is this big
link |
karmic wheel, right?
link |
And in the, it's not the 97, I was loaned by Netscape to do
link |
due diligence for Sun in their acquisition of Anamorphic,
link |
which was David Unger and friends, people, Craig, I'm
link |
forgetting his name, he went to Microsoft.
link |
These Stanford language buffs who had taken small talk and
link |
then David created itself as a simpler sort of small talk
link |
language and made really fast just in time compiling VMs for
link |
And they, you know, well ahead of Java hotspot or JavaScript V8
link |
or any of these modern VMs, figure out how to make dynamic
link |
code fast because small talk is dynamic language, right?
link |
It has classes that has, I think, more lockdown declarative
link |
syntax than JavaScript, but it's fundamentally dynamic.
link |
You don't declare the types.
link |
But you could infer the types as the program runs and you
link |
start to form these ideas about what types are actually
link |
flowing through key operations and you form little so called
link |
polymorphic inline caches that are optimized machine code.
link |
The cache is the machine code that assumes it does a quick
link |
check to make sure the type is right.
link |
If it's not right, it bails to the interpreter.
link |
If it is right, you go pretty fast.
link |
And that short test is a predicted branch.
link |
So things are pretty quick.
link |
All that amazing stuff I knew about in the 90s and I didn't
link |
have time to do it and anamorphic got bought by son and
link |
they did hotspot and you needed that even in Java because at
link |
scale Java has some dynamic aspects due to invoke interface.
link |
You can have basically collections of Java code where you
link |
don't know at the time each module or package is compiled
link |
exactly what's being called, what subclass or what
link |
implementation of an interface is being called.
link |
And so you want to optimize using the sort of dynamic
link |
polymorphic caching there too.
link |
And they did that and hotspot is an amazing beast.
link |
I've met like 13 people who all claim they created it.
link |
I think one of them may deserve credit more than others.
link |
But I didn't get to do that in JavaScript.
link |
And when we knew that Google was going to do their own browser
link |
which we knew it was around 2006.
link |
I also met the team that did v8.
link |
And it turns out it was Lars Bach who was one of the young
link |
engineers from anamorphic got acquired by son.
link |
And so Lars is like the one of the world's expert on these
link |
kinds of virtual machines.
link |
And he picked my brains up JavaScript.
link |
I could tell he didn't like it at the time but he had to do it.
link |
And really interesting.
link |
Yeah, 2006 lunch at Google's campus.
link |
And then I had another friend who was Deverell at Chrome and he
link |
said, yeah, we don't know what they're doing.
link |
This is getting 2007 to fall, getting toward 2008.
link |
We're trying to get Chrome out and we don't know what's going
link |
They're often all who's Denmark, you know, rewriting their engine
link |
four times, which is good.
link |
That's the right way to do this kind of development.
link |
You know, they were learning JavaScript including all its
link |
quirks which they came to hate the fire of a thousand sons,
link |
which is one of the reasons that Lars and company did dart
link |
their own language.
link |
But they also made the language fast.
link |
And meanwhile we knew this was happening.
link |
So we got our act together with Trace Monkey, our tracing jet
link |
at Mozilla and Apple I think was also aware.
link |
And so they were doing their own jet.
link |
So the era of jitted fast JavaScript in 2008 had this
link |
prehistory going back to small talk itself and anamorphic.
link |
And again, the lineage is interesting because you had Lars
link |
and anamorphic and then he ends up at Google.
link |
And today we have an incredibly fast language that like you
link |
said, still, you know, without hate, you can't have love.
link |
So I think there's both love and hate for this dance, this rich
link |
complex dance of JavaScript throughout its history.
link |
There's a dialectic for sure.
link |
Today JavaScript is the most popular language in the world.
link |
Why by many measures?
link |
Why do you think that is?
link |
Is there some fundamental ideas that you've already
link |
spoke a little bit broader that you think is the most
link |
popular language in the world?
link |
So I think I did, you know, by doing first class functions
link |
and taking the good parts of the C operator hierarchy and
link |
just keeping things simple enough.
link |
Maybe it could have been simpler, but I had to make it
link |
look like Java and interoperate with Java.
link |
That there was, you know, inherent goodness,
link |
you know, Aristotelian quality there and people perceive
link |
that even through all the quirks and warts and then over time
link |
working on it with the standards body, working on it,
link |
not only as a core language, but in the context of HTML5
link |
and making the browser better, listening to developers,
link |
thinking about this is something that Nick Thompson
link |
wrote nicely about on the hacker news.
link |
I was very flattered.
link |
He said Java was this thing where the experts were writing
link |
the code and it was compiled and you had to declare all
link |
your types and some didn't really give a damn about,
link |
you know, the average programmer who wanted to build
link |
real web apps, dynamic things.
link |
And I was in there meanwhile doing, you know,
link |
a bunch of people's jobs, making JavaScript survive
link |
those early years when it was kind of touch and go, right?
link |
JavaScript was considered Mickey Mouse language.
link |
It was for annoyances like the scrolling text at the bottom
link |
of the browser in the status bar.
link |
But I kept listening to developers working with them
link |
and trying to make it run in that single threaded event loop
link |
And I think that forged something that people have come to love.
link |
Now, you don't always love, you know, the best thing, right?
link |
I talked about Shakespeare, sonnet about,
link |
I'm Mr. Seizer, nothing like the sun or, you know,
link |
the scene from Josh Whedon's film Serenity at the end
link |
where the actual piece in the score by David Newman is called
link |
Love where Captain Mal is teaching River Tam
link |
about how to pilot the ship.
link |
And she's a super genius, super soldier.
link |
She knows how to do it already.
link |
And he's basically talking about how you have to love the ship
link |
because if you don't, it's going to kill you.
link |
And then the piece falls off the ship.
link |
It's kind of like JavaScript.
link |
You have to love it.
link |
You have to love it because now people will say,
link |
we're stuck with it because it got this priority of place.
link |
But there's love underpinning that and actually the listening
link |
to developers, that's kind of beautiful.
link |
There's most successful products in this world
link |
with all the messes, with all the flaws.
link |
Perhaps the flaws themselves are actual features,
link |
but that's a whole other, that's a discussion about love.
link |
But underneath it, there's something
link |
that just connects with people.
link |
And it has to keep connecting.
link |
If JavaScript kind of went off in this,
link |
people sometimes complain about ES6.
link |
Oh, you put classes in JavaScript.
link |
You know, you've ruined it.
link |
But it's not true.
link |
It's a dynamic language.
link |
Smalltalk had classes.
link |
Python has classes.
link |
There are lots of Lisp variants that have classy systems.
link |
So, you know, people who don't reject it based on some sort
link |
of fashion judgment do use it and do interact
link |
with the standards body.
link |
The standards body is competing browser vendors mainly,
link |
but also now big companies that use JavaScript heavily,
link |
the PayPal's and other such companies, Salesforce.
link |
And they have to cater to web developers.
link |
They have to hire developers who know JavaScript.
link |
They have to keep their engines up to the latest standard.
link |
And this creates all this sort of social structure
link |
around JavaScript that is unusual.
link |
I mean, you get C++ buffs to follow the inner workings
link |
of, you know, C++.
link |
But it's a more rarefied group.
link |
It's more like the old language, you know, gray hairs.
link |
Whereas JavaScript is a younger and more, you know,
link |
vibrant large crowd.
link |
There's a community feel to it.
link |
Perhaps I don't want to draw too many similarities.
link |
Maybe you can comment on it.
link |
There's a C++ is like Wall Street and the JavaScript is
link |
like Wall Street bets from the recent events.
link |
It's like there's a chaotic community of all and there's
link |
some power from that distributed crowd of people.
link |
It's more demonic.
link |
It's more of the people.
link |
It lets people in without requiring these credentials.
link |
I remember in the late 90s into the 90s, people were all
link |
getting Java credentials and I knew people and friends knew
link |
people who became Java programmers and you knew they
link |
really should have been like nature guides or pilots.
link |
They hated programming, but they thought, I got to make money.
link |
I'm going to become a Java programmer.
link |
Do you have some, because it's such a monumental moment in
link |
our current history as a quick aside, do you have thoughts
link |
about this huge distributed crowd sourced financial
link |
happenings with Wall Street bets?
link |
That's like nobody could have, well, you could have predicted
link |
but the scale and the impact of this kind of emergent
link |
behavior from independent parties that could happen.
link |
Like I said, my own experience with the dismal science as
link |
with physics led me to reject a lot of bad models and
link |
economics was always compromised by politics,
link |
political economy.
link |
You could also argue that it used to be a branch of moral
link |
philosophy, so it was concerned with the good and it became
link |
divorced and became sort of in this quasi Newtonian way
link |
just about everything's just running by itself.
link |
Don't worry about it.
link |
This monopoly is crushing your Netscape company,
link |
but that's just nature.
link |
An economist couldn't or doesn't really have good models
link |
for the Wall Street bets subreddit.
link |
They know how to squeeze the short, right?
link |
So the amazing thing is you have Robinhood app,
link |
which was again, supposedly for the demos for the people
link |
and eliminated the fee through various kinds of straddles
link |
or some kind of spread operation that helped them
link |
eliminate the fee or eat the fee.
link |
And in fact, as a broker in these days,
link |
because it takes two days to settle,
link |
there's counterparty risk as they found out.
link |
And so the Wall Street bets people,
link |
the memes are like the Terminator robot with the $600
link |
Stimmy check and the hedge funds,
link |
the little girl hiding under the desk.
link |
There is a problem which I talked about in the recent podcast,
link |
which I'm conscious of from the history of the web.
link |
And that is, you could say it's monopoly,
link |
which antitrust wasn't enforced after USV Microsoft
link |
And a lot of this was due to the money interests buying
link |
control of politicians.
link |
And in Plato's five regimes, that's oligarchy.
link |
That's where we are.
link |
And now we're seeing a fight against the oligarchs.
link |
I don't know if it'll work, but you're definitely seeing it.
link |
And it's also kind of hackerish, right?
link |
It's got a hacker ethos.
link |
Hey, Robinhood, no fees.
link |
Hey, buy a fraction of a share in this thing
link |
or I can keep buying with my stimulus check.
link |
So I mentioned Hegel seeing Napoleon on the horse.
link |
Hegel also talked about the cunning of reason
link |
to have this sort of God's history in full.
link |
And if you believe in God, or we don't know the future,
link |
but there's always this sort of fly in the ointment,
link |
this unintended consequence that confounds
link |
the best plans of the powers that be.
link |
And we're living through it.
link |
I'm glad it's not street warfare or mechanized warfare
link |
because it has been in the past.
link |
It's more like soft power.
link |
And people are fighting back.
link |
Do you think it's possible?
link |
So the JavaScript used to be for the front end of the web.
link |
It's now increasingly so being used for back end,
link |
like running stuff that's like behind the scenes.
link |
And it's also starting to be used quite a bit
link |
for things like TensorFlow.js.
link |
So starting to actually use like these heavy duty applications
link |
that are using your networks when you're learning
link |
and so on on the browser.
link |
Is it possible in 10, 20, 30 years
link |
that basically most of the world runs on JavaScript?
link |
This is a dystopia and a nightmare to some people.
link |
When we did Asim Jassi WebAssembly,
link |
I would joke and meme people with scenes
link |
like Neo waking up in his pod in the matrix
link |
and he's all skinny and weak and hairless.
link |
And you know, you realize in the future
link |
that you're living in some simulation
link |
that it's all running on JavaScript
link |
and you just scream forever.
link |
Gary Bernhard does these funny talks.
link |
He did WhatJS and then he did this life and death of JavaScript
link |
I think it's called where he took some clever ideas
link |
that actually have a thread of credibility to them.
link |
But I mentioned software fault isolation.
link |
In the old days when we were using computers,
link |
we said we're going to use the Unix monolithic monitor
link |
and it's the privilege program.
link |
This is before you even had hardware rings of protection,
link |
some of the early 60s operating systems
link |
used hardware protection zones.
link |
But Unix is privileged and the program
link |
that runs user code in a process is hosted.
link |
It's the guest in the host and you get to suspend it.
link |
You get to kill it.
link |
If it crashes, it doesn't take down the whole wonderful idea.
link |
But the call into the kernel is expensive.
link |
The system calls, so called.
link |
And this has even been optimized now
link |
for things like getting the time of day
link |
so it doesn't actually enter the kernel.
link |
And meanwhile, hardware architectures
link |
and virtualization techniques have gone in a different direction
link |
even to the point where you can do software fault isolation
link |
very cheaply without entering the operating system kernel.
link |
And so you get unicernals and exo kernels
link |
and very lightweight VMs.
link |
And so Gary took this idea and said,
link |
JavaScript will take over computing
link |
because the system call boundaries too expensive.
link |
So everything ends up in JavaScript
link |
with these lighter weight isolation enforcement mechanisms.
link |
It's not totally beyond belief.
link |
People have assembly too.
link |
It's nice to ask you sort of for advice
link |
to there's so many people that are interested
link |
in starting to learning about programming,
link |
getting into this world.
link |
Is there some number of languages,
link |
programming languages that you would recommend people learn
link |
or maybe a broader advice on how to get started in programming?
link |
Well, so you asked about machine learning
link |
and JavaScript is a general purpose language
link |
and it's a language that's not that great
link |
for doing matrix operations
link |
or doing parallel programming, I would say,
link |
without using some extensions or some libraries
link |
that have some magic in them.
link |
So if someone wanted to learn,
link |
there are amazing languages in sort of the APL family
link |
that are very useful for, I would say, linear algebra
link |
which gets to a lot of the kernels in machine learning.
link |
And so APL had J and then K
link |
and there are K variants
link |
because the guy that did K is still going
link |
and they're proprietary but he's still innovating there.
link |
So people talk about TensorFlow.js.
link |
That's surprising because Python was heavily used
link |
for machine learning.
link |
And Python was always, you know,
link |
they didn't have this fast just in time compiler tradition.
link |
There were some projects that tried this
link |
and some of them were interesting.
link |
PyPy was interesting, but the philosophy with Python was,
link |
oh, you need to go fast or write a C plug in
link |
and drop into C code.
link |
So I think people should look at multiple languages
link |
because there are different tools in the belt.
link |
If you're trying to do supervision or rapid prototyping,
link |
you want a dynamic language, you want to throw things together
link |
and see what works.
link |
If you're trying to go down to the metal very fast,
link |
well, I'm an old C hacker,
link |
but I was also the sponsor, executive sponsor of Rust
link |
at Mozilla and Rust has now escaped, you know,
link |
from that sort of nest where it was born
link |
to be adopted by a bunch of companies
link |
and have a foundation in the works.
link |
Some of the key core team members are working now
link |
at Amazon and other places.
link |
So it looks like Rust has reached escape velocity,
link |
and Rust is an interesting language
link |
because one of our goals there,
link |
one of the reasons I sponsored it was,
link |
we were all tired of seeing those
link |
remote code execution vulnerabilities
link |
due to C and C++, and we thought,
link |
can we have a sort of safety property
link |
through a type and effect system
link |
or an ownership system, and Rust has that.
link |
And that ownership system is interesting
link |
because it doesn't just give you memory safety.
link |
There's a sort of theorem for free,
link |
a dual that falls out for protection against data races.
link |
So Rust is better for low level programming,
link |
you delimit your unsafe code where you do have to be unsafe,
link |
and you can prove certain facts about memory safety
link |
and race condition avoidance.
link |
And so I think people should learn these new languages.
link |
I think Go is a great language I admire,
link |
you know, the Unix, people who did that,
link |
Ken Stoll was involved, Rob Pike, of course,
link |
David, what's his name, and other people.
link |
Go is a huge success,
link |
really on the server side, anywhere you have,
link |
you know, a lot of networking to do,
link |
and it's garbage collected,
link |
but it's also very pragmatic, it has that sort of C flavor.
link |
As an old C hacker, I can't get used to the fact
link |
that they swapped the type and declarator
link |
in the declaration order.
link |
I haven't used Rust, but this is one of the most respected
link |
and loved languages currently.
link |
And it's still young, you look at these things,
link |
JavaScript is now considered old.
link |
It's gone through so many versions
link |
that you can fall in love with it all over again.
link |
25 plus years, you know, it's an adult.
link |
It should be out of the house.
link |
But it could be around another 25 years.
link |
Cannot rule it out.
link |
So Rust will be around for a long time.
link |
The longer you're around, the more likely you're
link |
Lindy, and you're around.
link |
A lot of people ask me,
link |
I'm often torn between recommending either Python
link |
or JavaScript as the first language to play with,
link |
because it's difficult,
link |
because it's so easy to do JavaScript incorrectly.
link |
It's much easier to do it correctly these days.
link |
Or like, well, like learn about programming.
link |
But the cool thing about JavaScript
link |
is that you can create stuff
link |
that will put a smile on your face.
link |
Like as a developer, you can create stuff
link |
and it'll visually look like something
link |
and it'll do stuff and it makes you feel good.
link |
It makes you fall in love with programming.
link |
With Python, you could do the same.
link |
And with C++, it takes 5 to 10 years to write a program
link |
that actually does something.
link |
So there's that tension between
link |
is JavaScript the right first step
link |
And I've been going back and forth on those two.
link |
I had my Python right. It came from a lineage of ABC,
link |
which was a pedagogical language in the Netherlands.
link |
And it was a good teaching language too.
link |
I think it is a good teaching language
link |
and it's more restrictive
link |
that if you misspell something
link |
in a way that JavaScript might let run,
link |
let reach runtime, it'll get stopped
link |
at syntax check in Python.
link |
That's good for beginners.
link |
I think the sloppiness that some people object to,
link |
like people were just tweeting at me,
link |
having just learned JavaScript, they said,
link |
I can take a number and I can index into it
link |
and get undefined out of it as a property.
link |
And why is that? A number is not an object.
link |
And I explained why it is.
link |
The primitive types, which unfortunately are not objects,
link |
can be automatically boxed or wrapped by an object.
link |
And I made that implicit.
link |
In Java, it's typed and you have to declare things
link |
and you'll get type errors.
link |
But there are cases in Java where you get auto boxing
link |
or auto wrapping because you've declared that you want it.
link |
In JavaScript, it just happens.
link |
So once I explain it, I'm like, oh, wow, I get it.
link |
But it also means that you can commit a blunder that just...
link |
You don't get punished for, you don't detect.
link |
You get an undefined value and you don't know where it came from.
link |
I've been reading a lot about military history recently.
link |
And one way to paint the picture of browsers,
link |
Internet browsers, is through the various wars
link |
throughout its history.
link |
I don't know if that's a useful way to look at it,
link |
but we've already talked a little bit about Netscape
link |
and Internet Explorer in the early days.
link |
Can you tell the story of the different wars,
link |
if that's at all an interesting way to look at it,
link |
of the 90s and to today?
link |
Yeah. So I mentioned that Microsoft,
link |
which was convicted for it, did abuse its monopoly,
link |
but they had a pretty good team by the time they did IE4.
link |
And Netscape, unfortunately, I was like the second floor
link |
and I was friends with all the first floor people,
link |
the front end guys who did the JavaScript event hookup
link |
and things like that,
link |
that that team was fairly burnt out
link |
and I think having gone public,
link |
the upper management wanted to buy a bunch of companies
link |
to try to go head to head with Microsoft.
link |
Didn't work, but buying a bunch of companies
link |
usually doesn't work.
link |
I think the modern sort of approach,
link |
roughly is like Mark Zuckerberg took,
link |
which is to keep them at arm's length and let them do their thing.
link |
And now that he's pulling WhatsApp in and people are fleeing it
link |
because it's tied into the ad surveillance.
link |
But for a while there, keeping it separate really does work
link |
because you bought it for its value.
link |
And you're not messing with it.
link |
With Netscape, when they bought a bunch of companies,
link |
they had some of the first floor people
link |
or the founders burned out.
link |
They had newcomers who wanted their turn to do the browser
link |
and they hadn't really done browsers or understood them.
link |
And so Netscape 4 was originally supposed to be 3
link |
and it was so late they renumbered it.
link |
We did a 3 release, Jamie and a few others
link |
put some extra effort into SecureMime
link |
was supported in the mail, built in mail program.
link |
And Netscape 4 was late and it was only on Windows at first
link |
and Microsoft had really started doing better.
link |
Like they do, they copy in the first version's trash
link |
and the second one, you're starting to feel a little threatened.
link |
The third one, you can tell what's going to happen
link |
and the fourth one's good.
link |
And plus there's the benefit, like you said,
link |
that it comes as a default browser.
link |
Yes, and yet Netscape's screwing it up
link |
and Microsoft really putting some quality people on it.
link |
I4 was good, on Windows it was good.
link |
And they did the dynamic HTML innovations.
link |
Scott Isaacs, my old buddy, a former accountant
link |
who programmed in basic and became what Microsoft calls
link |
a program manager, which is kind of an elevated position.
link |
You can be a programmer and engineer in track,
link |
but you switch to it and you sort of lead
link |
a lot of design and standards efforts.
link |
And so Scott Isaac put in a lot of those funky DHTML APIs
link |
that didn't quite have the same flavor as the stuff that I did
link |
and neither of them was like the later sort of verbose Java
link |
like DOM W3C standardized.
link |
But I4 was pretty darn good.
link |
I remember a friend, Scott Furman, and I got invited by Scott Isaacs
link |
to Gordon Beersch and San Jose.
link |
They were doing a preview of I4.
link |
This must have been 1997.
link |
And Scott said, yeah, we've got here's the new graphic stuff we're doing.
link |
We've got something like your Netscape layers.
link |
We've got VML, a vector markup language.
link |
We can do virtual reality.
link |
And Scott and I looked at each other and said, we're doomed, right?
link |
Microsoft was starting to fire in all cylinders.
link |
So I have to give them credit for that,
link |
even though they abused their market power.
link |
And maybe, you know, I shouldn't give them credit
link |
for having the resources to hire talented people,
link |
but they did a credible job on I4.
link |
What really was bad was that phase of the browser wars
link |
ended with monopoly and perhaps due to the antitrust case,
link |
perhaps due to regulation in Europe,
link |
perhaps just due to Microsoft not liking dealing with standardization,
link |
they would let it rot.
link |
They just abandoned it I5, I5, I6 later,
link |
but these were not well maintained.
link |
They had a lot of security bugs.
link |
Browsers really closed and outdated too,
link |
even though it was getting updated.
link |
Browsers like Mozilla and Firefox were adding tabs,
link |
Operahead, a version of tabs.
link |
And they didn't add tabs.
link |
And they pop up blocking,
link |
something I should have done from the start.
link |
People realized that you can tell when the user clicks something
link |
and it goes in JavaScript to open a little window
link |
that you can sort of inspect the stack
link |
and see that the click originated that.
link |
And it's probably okay.
link |
Whereas if you're just loading a script
link |
and it opens a new window, that's a spam technique
link |
and you should block it.
link |
Tabs were brilliant innovation.
link |
Like you said, Operahead, but I remember
link |
I fully switched to Firefox the moment.
link |
I remember the moments of first using tabs in Firefox
link |
and not liking it for the first few minutes
link |
and like, wait a minute.
link |
You get the groove.
link |
You get the groove and you understand.
link |
So that timing, what year was this?
link |
Because also as a aspiring web designer,
link |
So we didn't mention layout or CSS much.
link |
There's also a change in the way,
link |
like the frames were going away.
link |
So there's a change in the way websites looked
link |
and behaved and all that kind of stuff.
link |
CSS finally, which Microsoft embraced with IE4
link |
and Netscape never really did right.
link |
CSS became a better standard over time
link |
for doing table layout that relieved you of the need
link |
to use what are called spacer gifts.
link |
The spacer gifts, right?
link |
You would throw in the space up tables.
link |
The typographic power of the web has gotten better
link |
but it's still not on the level of PDF
link |
and you can't do advanced typography.
link |
But it's gotten really better
link |
and even then tables were getting better.
link |
If you were using Firefox, that would have been 2004
link |
and it was called Firebird until early that year.
link |
No, yeah, I think it wasn't.
link |
It was a Firebird which had tabs.
link |
We had tabs the whole way.
link |
So it started out as Zola slash browser in 2002,
link |
There's a BIOS that has an embedded version of IE
link |
and they said, we're called Phoenix technologies.
link |
You can't use Phoenix.
link |
And so we said, okay, we'll call it Firebird.
link |
And then this Australian centered open source database project
link |
started really like in the true Mad Max style
link |
where we were streaming at us saying you can't use Firebird.
link |
And I had to sort of be the ambassador
link |
and say, okay, we're going to rename
link |
and I'm like, we don't believe you,
link |
you shouldn't have used it, we hate you.
link |
And then we renamed it to Firefox
link |
and they're like, we love you.
link |
And then I haven't heard of them ever since.
link |
But Firefox was a clever name.
link |
We had to think of something distinctive.
link |
We wanted to keep the fire going
link |
and it turns out there's a red panda.
link |
Right, it's a nickname for.
link |
So that's the second set of browser wars.
link |
Second browser wars.
link |
How did you, how was Firefox born?
link |
How was Mozilla born?
link |
Is there a, there's a long story there too.
link |
So Netscape got acquired by AOL,
link |
which as I say was a reasonable happy ending for a lot of people
link |
because Netscape otherwise was going to go out of business
link |
because Microsoft was just killing its market.
link |
There was no way to charge for a browser.
link |
Windows came with IE, IE four was pretty good
link |
and Netscape four wasn't that good
link |
and took a while to get better.
link |
But the Netscape executive said,
link |
let's do an open source escape pod, you know,
link |
and like in Star Wars and New Hope,
link |
the gunner won't shoot it
link |
because there's no life forms on board, right?
link |
It's not a threat.
link |
And so we did Mozilla in 1998
link |
and it looked like it was going to, you know,
link |
initially just give the world an open source browser,
link |
but it's really hard to get people to work on this sort of hairball
link |
that had been hacked up over by that point four years.
link |
It also couldn't have the crypto module for secure sockets,
link |
so called or now transport layer security.
link |
That was an electronic munition.
link |
We were not allowed to release that in the full 1024 bit key strength.
link |
And yet people, one of whom I happened to meet previously at SGI
link |
when I went on a sales support engineering trip,
link |
Tim Hudson in Brisbane, Australia and Eric A. Young
link |
did what became open SSL.
link |
It was called SSL EAY after Eric's initials
link |
and Tim and Eric took their open SSL
link |
outside of the purview of the NSA
link |
and the Department of Commerce
link |
and they stuck it into Mozilla's code.
link |
And that was perhaps the best hack that was done
link |
in the first few months after we open sourced the browser.
link |
We had other problems.
link |
The politics inside Netscape were driven by these acquisitions.
link |
So the one acquisition that kind of messed up Netscape 4
link |
also wanted to keep doing a proprietary mail and groupware program
link |
not Jamie Zewinski's mail program that was in Netscape 2 and 3
link |
and they held it back from open source.
link |
We didn't have a mail program, it was just a browser.
link |
We didn't know what AL will do to us.
link |
Turns out they didn't interfere with us for a long time.
link |
But Netscape wasn't the best steward of Mozilla.
link |
We were operating Mozilla as a pirate ship without a legal entity.
link |
So most of us worked for Netscape under a separate organization
link |
and initially the first engineering manager, Tom Pekwin
link |
of Netscape was the Mozilla Founding Manager.
link |
But he left pretty quickly and he left me as the acting manager
link |
which is more like method acting in my case.
link |
And that was my first management stint.
link |
But then someone who'd written the licenses, Mitchell Baker,
link |
she was a lawyer at Netscape, she was involved in the open source
link |
license decision making and the actual writing and construction
link |
of those licenses, that was Mitchell's job.
link |
Netscape Public License and the truly open Mozilla Public License.
link |
And there were two because Netscape needed, because of some encumbered code,
link |
needed some special rights, but that went away over time.
link |
Mitchell was always interested in Mozilla and she came back
link |
from maternity leave and she said, I'll be the manager if you want
link |
and Jamie and I said sure.
link |
And then Jamie quit.
link |
He quit after a year and he said, this didn't work, I'm sorry.
link |
He acted like it was a total failure because Mozilla didn't restart
link |
the browser market.
link |
But there's no way it could have, right?
link |
Netscape was still shipping variants of Netscape 4,
link |
which was based on the old code.
link |
Mozilla was trying to rearchitect the code to make green field for developers.
link |
So it was one of my big goals.
link |
It wasn't a technical goal so much as, again, a social goal.
link |
People wanted a more standard spaced browser.
link |
They wanted a less of a hairball that had been hacked on by
link |
X grad students starting four years prior.
link |
So we said we're going to make a modular code base.
link |
We're going to use a variant or an open source version of Microsoft's
link |
component object model.
link |
It has reference counting and standardized vTables,
link |
virtual calls and C++.
link |
And we're going to use JavaScript.
link |
We're going to have a bridge between those two so you can script
link |
those components just like Java components.
link |
We're going to make a portable front end with a markup language
link |
for the user interface.
link |
Not tables, not HTML, but custom menus and dropdowns and toolbars.
link |
And that was called Zool, XML user interface language.
link |
And some real talent on the Netscape side delivered that Dave Hyatt,
link |
who was instrumental in Zool, Chris Watterson, Joe Hewitt,
link |
And Blake was an intern.
link |
He was like a high school aged intern in Netscape.
link |
And at some point we were innovating rapidly in the Mozilla world.
link |
And Netscape was still caught up in this management mess from these
link |
acquisitions and it wasn't delivering.
link |
And every year they were wondering if AOL was going to come and start
link |
beheading the executive because it didn't do anything useful.
link |
And there was a thought you should take the Netscape browser engine
link |
and put it in the Windows AOL client, which was the dial up client
link |
that all the increasingly aging users of AOL were using.
link |
It would have been too big a change.
link |
So it wasn't clear why AOL bought Netscape, but as I said,
link |
they left it alone.
link |
But Netscape didn't leave Mozilla alone.
link |
And so in 2001, Mitchell called me up and said,
link |
I'm no longer employed.
link |
And I was like, what? You quit? No, no, this wasn't my choice.
link |
And there was a layoff, which maybe accidentally or on purpose,
link |
got rid of Mitchell.
link |
But the funny thing was we had an open source project.
link |
We had a lot of the engineers on staff on our side and we had
link |
people we'd hired through the Mozilla community who were top notch.
link |
They'd risen, you know, they came in high quality.
link |
They knew the code and they actually were better than the average
link |
or median higher of Netscape.
link |
And so the funny thing was the executive who thought they'd gotten
link |
rid of Mitchell in the layoff on the next week's community call
link |
around Mozilla and what to do.
link |
And so this show, you can kind of transcend your, you know,
link |
boundaries of corporate open source.
link |
If you get a project that has enough loyalty, even among the paid
link |
staff, because we had outside people contributing with people
link |
at Red Hat and a few other places.
link |
But the majority of the hackers were employed by Netscape.
link |
But a lot of them at that point had come from the community
link |
and others got the community and wanted to work with it.
link |
And it was really the weakest engineers at Netscape who didn't
link |
like Mozilla and didn't like the crucible of competing with the
link |
better programmers.
link |
So if the project is good enough, it will rise.
link |
The Phoenix will rise out of the...
link |
That's exactly right.
link |
And so we had this Mozilla code base that was getting better.
link |
In fact, I think at some point in 2002 when we declared Mozilla 1.0,
link |
I engineered a roadmap that successively through similar sort
link |
of six week, five week releases like we all do with browser
link |
releases nowadays, Chrome does and Firefox braves us three weeks.
link |
We got to a point where we said, you know what?
link |
This is like the 1.0 that you want to release because if you hold it back
link |
any longer to polish it, you're denying others the ability to use it.
link |
It's like pro engineer, the mechanical CAD tool embedded the code.
link |
They embedded the layout engine.
link |
And Mozilla 1.0 was like a Netscape communication suite.
link |
We had at that point gotten male people to reintegrate male news.
link |
And we had an editor for HTML.
link |
And it felt like a 90s suite, a suiteware.
link |
And it felt kind of bloated.
link |
And the people who were taking that Mozilla open source and then adding
link |
Netscape flavor to it were not calling the shots right.
link |
And they were also under ALL's thumb a little bit and that they said,
link |
we should probably put the ALL instant messenger chicklet on the toolbar.
link |
And we should put the ICQ, the other messaging system that ALL had acquired.
link |
We should put the ICQ button on the toolbar.
link |
And pretty soon Netscape looked like a bit of a NASCAR badged version of Mozilla.
link |
And that also made Mozilla more popular.
link |
And yet they had contrived to fire or lay off the leader.
link |
And we carried on with an open source structure where Mozilla was still,
link |
Mitchell was calling management or project level shots.
link |
And I was calling technical shots.
link |
And we had a popular suite.
link |
But we thought, why not make it just a browser?
link |
Because it'll be simpler.
link |
It'll do one job well.
link |
And even then we can strip it down by having extensions.
link |
So Dave Hyatt and Blake Ross, the high school aged intern, did the first
link |
version, which was called Mozilla slash browser.
link |
It was a group of us, Ian Hicks and Asa Dotsler, me, Joe Hewitt and Hyatt and Blake.
link |
And Hyatt was really the senior hacker.
link |
He'd done all these things like amazing cross platform menus
link |
through the user interface market language.
link |
And he knew how to do tab browsing.
link |
He'd implemented it natively on macOS at the time in Camino, originally called Camara.
link |
He'd written multiple limitations, which was a thing programmers should do.
link |
It's like the VA team did for those missing years when the rest of the Chrome teams
link |
In fact, Dave's wife, Rebecca, told me a story about when they were at UIUC.
link |
They were also University of Illinois grad students.
link |
There was an assignment.
link |
It was a programming assignment.
link |
It was supposed to be due at the end of the semester.
link |
And Dave's friend was this, I'm going to go think and I'm going to design and I'm going
link |
to make this platonic perfect form of the program.
link |
And then I'm going to write it at the end when it's due.
link |
And Hyatt just went there and started hacking.
link |
He wrote one version, he wrote a second version, a third version.
link |
The end of the semester comes around.
link |
The friend's not doing too well.
link |
It wasn't perfect and it wasn't written.
link |
I'm not sure how that story ended, but for him,
link |
but Dave's version was a fifth iteration.
link |
And so he'd done that with everything you needed in a tabbed browser.
link |
And this really showed well in Phoenix, what we called Phoenix,
link |
and I had to rename two more times.
link |
And Blake went to Stanford.
link |
He became a Stanford student and couldn't work on it.
link |
Dave Hyatt went to Apple in 2001.
link |
He was one of the founding Safari team members.
link |
But he was still blogging about tabbed browsing.
link |
I think Apple at some point said,
link |
Safari have tabbed browsing?
link |
But it was because of Hyatt.
link |
Hyatt was quite a feather in their cap.
link |
Don Melton, who had been the engineering manager at
link |
for Safari from the beginning, had been in Netscape also.
link |
And so there's just a diaspora of talent.
link |
And yet Hyatt was still kind of writing blog posts
link |
about how to do tabs, right?
link |
And at some point Apple said, don't blog about that.
link |
That's our proprietary tab technology.
link |
And I was like, no, it's not.
link |
So we had to replace people.
link |
And we had Ben Goodger, a New Zealander.
link |
We hired at Netscape.
link |
And he stepped in to be the Firefox lead.
link |
And we also had this weird circumstance where AOL finally
link |
did notice that Netscape was kind of an albatross,
link |
that they bought it for no particular benefit.
link |
And even then the AOL politics were also heinous,
link |
sort of East Coast politics.
link |
And we were taking two trips there because I was a principal
link |
And so as principal engineers got trotted out to do dog and pony
link |
shows in Dallas, Virginia, and the AOL for management was
link |
very East Coast in flavor.
link |
And they were at that time merging with Time Warner,
link |
which would not go well.
link |
So one of these years we went out there and we were all doing
link |
dog and pony shows.
link |
And there were these characters that were sort of like
link |
marketing guys, one of the most wearing a cravat.
link |
And one of them was named Reggie.
link |
And they were very you rather than non you,
link |
or they were like what's still in metropolitan film,
link |
UHB, urban, oh, bourgeoisie.
link |
They were funny and they were kind of useless and kind of
link |
And then the next year we went back and I said,
link |
And it's like, oh, Reggie's not here anymore because Time Warner
link |
realized that the merger wasn't in their interest either.
link |
And then the sort of knives came out.
link |
And these mergers rarely work, right?
link |
This is very difficult.
link |
You get these giant companies and they think there's going
link |
That was the late 90s watchword.
link |
And there wasn't synergy with AOL buying Netscape.
link |
And there wasn't synergy with Time Warner and AOL.
link |
But did AOL ever really work?
link |
Was it ever really cool?
link |
Like the same kind of fire and excitement that Firefox
link |
eventually created, was that ever there in AOL?
link |
AOL was the right time to do a dial up service that got
link |
distribution by basically leaflet bombing compact discs on
link |
And they beat out CompuServe and the other ones,
link |
And then the web happened.
link |
And so you had almost like this isolated continent.
link |
Like some of the evolutionary biologists I follow make fun
link |
of the funny large marsupial mammals of Australia,
link |
how silly they are.
link |
So AOL is like Australia.
link |
And you saw it over time because they kept aging and they
link |
were using AOL to get online and they couldn't really use
link |
And it became sort of a valued cohort because they still
link |
have relatively high socioeconomic status and they
link |
have grandchildren.
link |
But it's going away.
link |
It's dying at some point.
link |
Towards the end of the aughts that decade and then to the
link |
decade 2010 plus that Firefox became this incredible.
link |
I forget when Chrome came out.
link |
But Firefox was the sexy, cool thing that represented a
link |
lot of the cutting edge technologies and all that kind
link |
Tim O'Malley and John Battelle did the first Web 2
link |
conference, which eventually became huge and they split
link |
But that was in 2004.
link |
It was right when Firefox was out.
link |
Craigslist was huge.
link |
It was killing classified revenue for newspapers.
link |
But there was just this ferment.
link |
People starting Wikipedia along there somewhere.
link |
Gmail was already done and it was an impressive Web mail.
link |
There were others before it like Hotmail, but Gmail was
link |
really impressive from Google and Google Maps.
link |
People started seeing what could be done.
link |
They thought, how can you drag the map around and how does
link |
And it was all JavaScript and images.
link |
So Gmail was 2003, 2004?
link |
Yeah, it actually started quite early.
link |
It might have been 2002 or three.
link |
But by the time we started dealing with Google and Firefox
link |
to get the search deal, which was the main revenue source
link |
for Mozilla and still is 2004.
link |
Early Sergey Brins, one of his trusted engineer guys,
link |
Fritz Schneider, made contact with me at Mozilla.
link |
And we started talking and we realized search and browser
link |
And this is deeply true, right?
link |
This is still true.
link |
This is why a lot of the search engines have their own
link |
Yeah, so in case people don't know, the main revenue source
link |
for the browser is the default search engine, which is kind
link |
of incredible to think about.
link |
That is a revenue source.
link |
It's a little bit sad.
link |
Yeah, it leads to this capture or kill effect where you
link |
have the search engine own its own browser.
link |
And other browsers may struggle to get this distribution
link |
we talked about earlier.
link |
So where, and you said you've figured out that Google is
link |
working on its own browser at some point.
link |
So would you say Firefox versus, was Internet Explorer
link |
part of the war here or was the Firefox versus Chrome?
link |
So Firefox didn't quite cause Microsoft to reconvene IE.
link |
And I remember being on a plane back from the standards
link |
meeting, the JavaScript standards meeting from Seattle
link |
And there was some Microsoft guy in front of me.
link |
It turns out my wife knew him from her past life before
link |
And he was just this bearded big guy.
link |
And he was like, we should have just killed Firefox
link |
All we needed to do was add pop up walking in and tabs.
link |
And we could have made Internet Explorer kill Firefox.
link |
And it's like, should have cut a wood up, pal.
link |
And I was right behind him hearing this.
link |
And IE7 wasn't that great.
link |
And what really got them started, I think, was Chrome.
link |
And I talked to Larry Page in 2005.
link |
I think I said, we're talking about the Firefox relationship.
link |
But he was also saying, what a web kit.
link |
This was Apple's version of the old K HTML engine
link |
from Linux, the KDE side of Linux that was used in the Conqueror
link |
browser, all spelled Ks, that Apple had forked.
link |
And in 2005 was when Apple's principles, including Dave Hyatt,
link |
Maches de Covex, some of my friends who were still there,
link |
said, we must stop patch bombing this poor K HTML project.
link |
We should make a proper Mozilla like organization WebKit.org.
link |
Now, it wasn't a separate nonprofit or anything.
link |
It was still Apple.
link |
It was Apple controlled.
link |
But they made their fork first class.
link |
And they made it be something that they all worked in
link |
And that was before Chrome.
link |
And then Chrome, Larry Page said, what about WebKit?
link |
I said, yeah, it's nice.
link |
I have friends who work on it.
link |
You might use that if you do your own browser.
link |
Why don't you do your own browser?
link |
Don't worry about Firefox.
link |
You should do your own browser.
link |
You can have your own opinion of how it should work.
link |
And sure enough, they did.
link |
So by 2006, we knew they'd been working on it.
link |
Some of my friends who'd been at Netscape did the original demo.
link |
And the demo wasn't what you thought.
link |
It didn't have the fast JavaScript yet.
link |
That was still often Denmark in the farm.
link |
It had tabs because all browsers had tabs at this point.
link |
And it had this software fault isolation I mentioned.
link |
It was through process isolation.
link |
So in theory, each tab has its own operating system process.
link |
And so what's going to take your tab down?
link |
Well, WebKit has bugs that can crash it.
link |
But Flash was still big then.
link |
All the restaurant sites, remember?
link |
And Flash crashed a lot.
link |
So the demo that I heard about my friends at Netscape,
link |
as a lot of people did, inside Google was the sad tab.
link |
They showed an early version of Chrome,
link |
which is this bare bones tab browser.
link |
They loaded a site with a known Flash bone,
link |
and then suddenly Flash crashes.
link |
And everyone expected the whole browser to go down.
link |
But instead, you got this little sad face in the tab.
link |
And you could reload it.
link |
And there it is again.
link |
So this was an improvement.
link |
It was a real move for security.
link |
It was based on a company they acquired called Green Border.
link |
They had some really big brains, like Olfar Erelingson,
link |
I think was involved.
link |
And they had done some exotic security stuff.
link |
But they ended up simplifying it to this process isolation.
link |
And Firefox didn't have it at the time.
link |
So we were still struggling with security bugs.
link |
So we knew Chrome was coming,
link |
but it took two more years to come out.
link |
And we were still getting the Google search revenue.
link |
We were still making Google the default engine.
link |
And Firefox was still growing.
link |
Firefox grew, I think, until 2011.
link |
That was when it peaked.
link |
And as it started falling, it was because of Chrome.
link |
Chrome came out in 2008.
link |
And it had a comic book that leaked accidentally
link |
that showed some of the people who worked on it.
link |
Lars Bock was in there and so on.
link |
It was kind of soft launch because they didn't market it heavily.
link |
They didn't push distribution.
link |
But Google had reason to worry about distribution
link |
because Microsoft was, you know, doing a search engine,
link |
In fact, when they came out with Bing,
link |
Google was worried that Microsoft would just brute force
link |
switch the default browser and everyone's Internet Explorer
link |
or even Firefox on Windows to Bing from Google.
link |
And Microsoft wasn't, I think,
link |
ready to dare the antitrust cops that way,
link |
even though they'd gone to sleep.
link |
And I don't think Bing was ready either.
link |
But just in case it happened,
link |
Sundar Pichai, who rose very well based on this work,
link |
was sort of in charge of getting distribution deals.
link |
And he got Google toolbar and Google desktop search distribution.
link |
And remember those pieces of software?
link |
Those were like desktop extensions,
link |
toolbars or operating system extensions
link |
for doing desktop search, searching your local files,
link |
kind of like macOS spotlight, right?
link |
Sadly, died of doubt.
link |
And there were some features that we still missed
link |
that didn't make it into Chrome.
link |
But Sundar got OEMs to bundle those.
link |
And then he got enough of those deals
link |
that by 2007 or 2008, Google felt,
link |
well, if Microsoft does the worst and tries to force Bing,
link |
we can reach in and reset it with that point of presence.
link |
So that was good for Sundar's career.
link |
And it was good for Google,
link |
but it never came to pass that they had to defend.
link |
Microsoft was still slow.
link |
And by the time they saw Chrome come out,
link |
then they did what would have been IE9.
link |
And then they said,
link |
we're going to have a fast JavaScript engine to Chakra,
link |
And they did okay.
link |
So that was another process isolated fast JavaScript browser,
link |
So it sounds like there's a deep fundamental coupling
link |
of search engine and browser
link |
that's mixing this whole thing up.
link |
And obviously Firefox doesn't have a search engine.
link |
That's like, I mean, you're partnering with somebody
link |
with a search engine.
link |
With Yahoo or with Google or so on.
link |
That was unfortunate because I think,
link |
even though Marissa Mayer talked about it,
link |
she never pulled it off and never restored the search team
link |
that had been laid off.
link |
I believe Carol Bartz was running Yahoo when Carol said,
link |
I've got to get rid of one of the three expensive things.
link |
I'm going to get rid of search.
link |
And those researchers went to Google and Microsoft.
link |
And there was no way to put Yahoo search back together.
link |
So when Firefox tried switching all their users who'd stuck
link |
with a default from Google to Yahoo,
link |
like mid December 2014, a bunch of users said,
link |
what just happened to my Firefox?
link |
And others didn't notice right away,
link |
but over time they did.
link |
And so over the next year,
link |
the traffic just went away for Yahoo.
link |
And yet they were obliged.
link |
I don't have inside knowledge,
link |
but this is leaked out.
link |
And Danny Sullivan's written about it, search engine land.
link |
I think the deal was like fixed payments to Mozilla.
link |
So Mozilla was getting a bunch of money for traffic
link |
that wasn't staying because users were resetting their default.
link |
And this shows how defaults are important,
link |
but they have to be good enough that the user doesn't override them.
link |
And a lot of the commercial value in popular apps
link |
is one of the default settings.
link |
What is the default search?
link |
But oftentimes there's something just like you said.
link |
I mean, if there's something compelling,
link |
that's also can beat out the default like tab browser and so on.
link |
And that's where, I mean, we'll talk about Brave browser.
link |
It feels like now we're in this third stage
link |
where there's Chrome, Firefox, Edge, I guess it's called,
link |
and Brave, and these all seem like really exciting,
link |
I don't know, innovative browsers.
link |
They're all kind of copying off of each other,
link |
picking up the good stuff.
link |
There's evolution again, especially on tracking protection.
link |
Privacy is this sort of global wave that's rising.
link |
I like to call it a wave because it's a large,
link |
somewhat chaotic structure.
link |
It's not a unitary good.
link |
You can't say, I'm buying privacy for $3.
link |
I'm paying $3 privacy.
link |
Some people think a VPN does this and are disappointed when it fails them,
link |
but often people use VPNs for region unlocking video
link |
or getting the US Netflix catalog.
link |
But privacy is not a unitary good.
link |
It's complex and people are understanding it only over time
link |
and as they get burned.
link |
But there's a genie that's not going back in the bottle there.
link |
People are fed up.
link |
Apple has responded to this.
link |
Apple was always making Safari, I think,
link |
more of a privacy branded browser from the very beginning.
link |
I think this was probably Steve Jobs.
link |
Safari had private Windows, private tabs before Firefox did.
link |
And these are only private in the sense that they don't leave local traces,
link |
if you don't want them to, turns out Safari does keep them around
link |
But the canonical model is no local traces after you close the private window.
link |
No leftover traces that you went to some site that you were embarrassed by
link |
or bought a gift for somebody you wanted to keep secret.
link |
But there's still some level of tracking.
link |
There's network tracking.
link |
Network privacy is not guaranteed at all
link |
because you're using the same internet and ISP as a public window,
link |
a non private window.
link |
But Safari had that early on.
link |
They also had a cookie blocking policy that might take a little explaining.
link |
When you, if you know what a cookie is,
link |
it's a little bit of storage in the browser indexed by the name of the site.
link |
And it's really only the main name of the site, like bfa.com or, you know,
link |
something like npr.org.
link |
Every site can store some information in a cookie.
link |
Every time it's contacted by the browser,
link |
the previous version is sent back and in the response from the server,
link |
the cookies updated.
link |
So it's this little bit of storage in the browser that the site can keep updating
link |
and it can store an encrypted version of your login credentials with a time stamp
link |
so you can stay logged in without having to retype your password every time you navigate,
link |
which is how it would be if you didn't have cookies.
link |
The web protocols, especially in the 90s, are so called stateless protocols.
link |
So you go to your bank, you log in, you go from your login confirmed page
link |
to your account view.
link |
If you didn't have a cookie, you'd be logging in again.
link |
Every time you type in a cell.
link |
So that was a great thing about cookies.
link |
Lumontually did it in a hurry in 1994 before I joined that escape
link |
and he did it for really holding that kind of credential.
link |
But even then, there was the image element embedded in the page
link |
and the image gets fetched possibly from a different server
link |
and that request carries the last cookie, which could be empty at first
link |
and the response carries the updated cookie.
link |
So just by having images and cookies yet tracking
link |
because that image server can be serving a little one by one pixel
link |
and they still use the word pixel on ad tech
link |
and that pixel can be served from the same server embedded differently
link |
with different URL spellings in the New York Times and ESPN.
link |
And as you go from one to the other, the image server can say,
link |
I haven't got a cookie for you.
link |
It's empty initially.
link |
I'm going to assign you user number one, two, three, four.
link |
I'm going to put a database entry in.
link |
And I see by the way, I always fetch the name of the path part of the URL
link |
that I was in the New York Times.
link |
So you're a New York Times reader.
link |
And then you hit ESPN, same thing.
link |
And the database gets updated and the number user one, two, three, four indexes
link |
in the database to a profile of you.
link |
You've been tracked.
link |
This was not intended.
link |
And it was too late to undo by the time I got the Netscape.
link |
I think Lou wanted to do Twinkies, he called them in.
link |
He was trying to solve several problems.
link |
He wanted them to be bigger because initially cookies had a short size limit.
link |
I think he wanted to solve the third party problem,
link |
but Tom Paiklin, the engineering manager said, no, no Twinkies, just cookies.
link |
And that's how a lot of that stuff was.
link |
That's how JavaScript, you know, got frozen like a flying amber in some ways
link |
with that sloppy equality operator that I made because the early adopters.
link |
And the cookie got stuck with this tracking hazard.
link |
And then because JavaScripts can be like images, they're embedded in the page.
link |
By the time of Netscape 3, I made that work.
link |
You can get a request with the last cookie value and the response updates it.
link |
That's a tracking mechanism.
link |
And that's why you don't even need images to track.
link |
Now you just use scripts.
link |
So this whole tracking economy evolved.
link |
And it depended on these accidents of the 90s, these unintended consequences.
link |
Well, it created some of the richest companies in the world, right?
link |
I mean, it's the social media.
link |
All I got was t shirts.
link |
All I got is this crappy t shirt.
link |
I mean, so that's, that's the fundamental problem the world is facing now.
link |
They're looking at what social media has created and they're looking at,
link |
and like a world is looking itself in the mirror and seeing that privacy is actually something.
link |
It's supposed to like a nice thing to have.
link |
It's something that is actually should be fundamental to the way we interact with the world as part of our tooling.
link |
And that's where the brave browser comes in.
link |
And I suppose others as well will play with this idea, but brave is at the forefront of that.
link |
So maybe can you like describe what brave is and what are its key principles and what's broken
link |
and what is it brave trying to fix?
link |
So when I realized that these accidents, like the third party cookie, the image or script that's tracking you,
link |
or the java scripts that can do it invisibly now, that all this stuff wasn't intended
link |
and that Firefox had supported extensions that block some of these things,
link |
I thought probably we should have browsers just block some of these things by default.
link |
These were not intended and they're now unsafe.
link |
They're tracking you.
link |
There could be data breaches, malware distribution, you know, bullying and PSI ops and other attacks on people.
link |
Block that JavaScript.
link |
I'm Dr. Frankenstein.
link |
I've got to deal with the monster here.
link |
But obviously you go to Gmail, there's a bunch of script there to make that amazing web client.
link |
That's first party JavaScript.
link |
So how do you tell the first from the third party?
link |
And it's not easy.
link |
It's not a matter of just what's embedded from a different server because a lot of publishers use benign scripts
link |
from unrelated domains or apparently unrelated domains.
link |
So you end up having to develop a sort of human and machine learning practice around blocking.
link |
And at Brave, we did that from the start and built a research team to help drive it and automate it.
link |
We realized that protecting people needed machine learning.
link |
And around 2017 spring, I talked to my friends at Apple about this too.
link |
And they were also doing what they call intelligent tracking prevention,
link |
which uses local machine learning in the browser.
link |
And the funny thing is, you know, great minds think alike,
link |
they were taking their third party cookie blocker that was in Safari from the old days
link |
and making it not have a big loophole.
link |
Because what they did was in 2003, when I came out, they said,
link |
we're going to block cookies that are from those third party embedded elements
link |
where you've never visited that site before.
link |
So I'm going to pick an ad company that got sold to AT&T.
link |
So I'm not picking on anybody on fairly appnexus.com.
link |
Have you ever been to appnexus.com?
link |
I've never been there.
link |
But I guarantee you 10 years ago, you probably had, if you're using Firefox,
link |
you had a cookie third party cookie because you're being tracked by them.
link |
And they were using that cookie to build up a profile of you.
link |
In Safari, as long as the user never went to appnexus,
link |
that cookie would not be set.
link |
And that was a real move for privacy early on when jobs were still around in Safari.
link |
But it had this loophole that if you do go to appnexus,
link |
then why it's okay to be a third party cookie?
link |
And so appnexus did something very naughty.
link |
They took their ad partners to put the actual ad you click on.
link |
And they said, hey, add a little script so that when somebody clicks on the ad,
link |
before it goes to your landing page, redirect to appnexus
link |
and we'll redirect to the landing page.
link |
And by doing that, they set a first party cookie and they got whitelisted.
link |
So it was a loophole they exploited.
link |
Intelligent tracking prevention in Safari was sophisticated enough to counteract this
link |
and it did other things and it's evolved since they did it.
link |
And we've evolved brave too.
link |
And so when I say machine and human learning, there's a real set of techniques here.
link |
They have to fight.
link |
It's a fascinating problem.
link |
Finger printing, right?
link |
Anytime you have a little bit of storage in the browser associated with a website,
link |
if the bad guy can get 32 websites, each one has a bit of storage.
link |
You can turn the bit on or off.
link |
You can make four billion numbers.
link |
You can make an identifier.
link |
It's called a super cookie sometimes.
link |
There are weaker ways that are statistical.
link |
They're called fingerprinting.
link |
You have to block all of them and you have to not only automate,
link |
you want to work in the web standards body to put privacy in by default,
link |
by design, from the get go, not added as an afterthought
link |
or go hog while with new web APIs that add a bunch more local storage
link |
or fingerprint surface area.
link |
And that's been a struggle too because guess who's the new Microsoft in the standards body?
link |
And they're not in favor of privacy first.
link |
They want to do privacy their way only under, I would say, market pressure.
link |
But with Apple and with Brave leading the way,
link |
we block third party cookies almost without exception.
link |
So we just block them and that gives us a very strong privacy benefit.
link |
But it also means some sites just don't work right.
link |
Embedded YouTube videos might not work right.
link |
So we're adapting in similar way to Apple's done with ITP to make third party cookies blocked
link |
but to sort of simulate what looks like a working third party cookie for the site.
link |
It essentially tries to partition each site and its third parties into its own sort of cookie jar.
link |
So like you said, is this both like a human fine tuning issue and a machine learning problem?
link |
And as humans learn, then they train the machine learning.
link |
But maybe Google aside or including Google, there is millions of dollars,
link |
if not be billions of dollars to be made from fighting the ways of Brave.
link |
And it's been an interesting change from when we started in 2015.
link |
Adblocking extensions, Ablock Plus was one of the big ones that started on Firefox in 2006,
link |
I believe, had gotten to a certain level of use around the world.
link |
And browsers like UcWeb, UcBrowser in Asia had some amount of ablocking built in and on by default.
link |
So a page fair was a startup and they measured ablocking adoption and they tried to say,
link |
hey, publishers, you're 30% of the visitors to pitchfork or wire to kind of NAS properties
link |
are using ablockers.
link |
If we can somehow convince them to lower their ablocking for your site,
link |
that could be like a 43% lift, right?
link |
And three sevenths.
link |
Well, that's easier said than done.
link |
And page fair and other source point that many others tried to either smuggle ads through
link |
or cajole the user into letting ads appear.
link |
And it didn't really work.
link |
And meanwhile, the ablocking adoption has just continued intelligent tracking prevention in Safari in 2017.
link |
Brave from 2016 on with very strong cookie blocking and other protections.
link |
And this is not going away.
link |
The publishers used to rage against it.
link |
We would try to say, we can help you.
link |
You're dealing with users who are already blocking all your ads.
link |
We can try to put back some economics that helped the user and you that led to the basic attention
link |
token that we started with Bitcoin.
link |
We can be your friend.
link |
Don't just fingerprint us as an ad blocker and treat us as an enemy.
link |
But in 2015 or 16, it was like, nah, you're an ablocker.
link |
And by 2017 or 18, it's like something's happening.
link |
The ablocking is not stopping.
link |
And we're all getting sort of pulled on the Google's plantation through AMP, AMP.
link |
Or we're getting killed by the Google ad system we use because it's taking all the revenue.
link |
Or some other vendors we use are permitting ad fraud.
link |
So a fake New York Times is getting paid by the marketer running an ad that a bot clicks on.
link |
And the real New York Times, it's supposed to get the ad, doesn't get it.
link |
And there's something really broken about that kind of system.
link |
And that fraud is mediated through Google's ad exchange, which is the biggest of them all.
link |
And Google takes a fee.
link |
There's a flip side of that, which is malware distribution, malvertising, where fake advertisers
link |
put malware payloads or exploit kit loaders in JavaScript and they smuggle them in ads
link |
onto real publisher pages.
link |
The ad exchange takes the fee.
link |
Now, I'm not a lawyer.
link |
I'm not going to say this is a RICO predicate.
link |
But why is the ad exchange facilitating fraud and malware distribution and taking a fee?
link |
As opposed to just fighting, this is the really interesting thing about Brave.
link |
As opposed to just fighting and then being treated like an ad blocker, you're providing an alternate.
link |
There's a philosophical idea here that might change the nature of the internet with the basic attention token.
link |
Maybe what is basic attention token, BAT, and how does it work?
link |
I'll tell the story first by saying how I came to it.
link |
I realized for a long time at Firefox, we were dependent on this Google search deal.
link |
And I thought, now that Chrome's out, maybe that's going to go away.
link |
At some point, Google will say, Firefox, like old Yeller, you saved me from the rabid beast.
link |
Now, I have to shoot you in the head.
link |
And what can we do?
link |
And I think this, Mozilla doesn't know what to do.
link |
This is something that I couldn't solve there and I don't think they can solve.
link |
But I thought, why is the browser the sort of passive servant of these big tech companies?
link |
I said a blind runtime for ad tech JavaScripts, including from Google.
link |
Why doesn't it block some?
link |
And if it blocks some, why can't it reconnect users, readers, fans with publishers, creators, websites?
link |
Why can't it help people make direct payments or even possibly get an ad revenue share for private ads that are placed in the browser?
link |
The ads are all placed in the browser.
link |
Some people have this sort of model that the server is painting the ad into some flash combined package or into some giant image and then it all gets sent down.
link |
That's not how it works.
link |
All the ads you see on the web are placed in your browser by calling out to various ad tech partners and Google's among them.
link |
And so if you block those scripts, you break the advertising flow of money from the brands and their agencies to the publishers.
link |
And if you want to reconnect it directly with the user, you have limited choices.
link |
The user generally isn't going to sign up with a ACH bank connection or a credit card.
link |
The publisher isn't going to sign up the user except as a subscriber.
link |
And then they're going to overcharge you because they want you to cross subsidize all the content and buy more than you read and all that stuff.
link |
And people are doing great who are big names like New York Times and The Washington Post.
link |
But how many subscriptions are you as a user going to pay for?
link |
This is why startups like Tony Hale's Skrull are trying to do a portable subscription system.
link |
By the way, just on a small tangent there, even in New York Times, it's really annoying how difficult it is to subscribe.
link |
There's way too many clicks.
link |
They don't make it easy.
link |
I had friends a few years ago, I think they fixed this, who would pay for the paper and then they'd go online and they get upcharged for the digital.
link |
And there was no break.
link |
There was no connection between them.
link |
But publishers are not that technical and they can't all get you to subscribe.
link |
You can't have a thousand subscriptions.
link |
So for a long time, people talked about micropayments.
link |
There was Blendol and other ones which came to the US but didn't grow.
link |
And I thought if you have just a browser and it's protecting you by blocking all this ad tech tracking junk,
link |
it can provide you an option that uses cryptocurrency to let you support your favorite sites and even YouTube channels.
link |
And that we prototyped with Bitcoin.
link |
And that meant the user had to be of means to contribute and willing to contribute.
link |
But it could be done on the Bitcoin blockchain and it could be fairly efficient even though Bitcoin went through a period
link |
when we had this prototype running in 2016 into 2017 where Bitcoin was very congested and very slow to confirm and the fees got very high.
link |
And a lot of users who were not Bitcoin maximalists or even experienced,
link |
we helped them out by embedding a Coinbase buy widget and they had the income to buy but it was hard.
link |
It was like, do I buy $5 a month?
link |
But the fee is like $4.50.
link |
I better buy in larger batches, right?
link |
Then they're like, I don't want to own that much Bitcoin.
link |
So it became this painful thing.
link |
And the real idea that I had of private ads that pay the user a rev share couldn't be realized alone in that kind of system.
link |
In these cryptocurrency systems, especially with the blockchain we switched to Ethereum, you can have smart contracts.
link |
The Bitcoin system is not turn complete so what you can do with the script is more limited.
link |
But you can still do sort of clever things even with Bitcoin script.
link |
What we wanted to do was sort of a three sided ecosystem.
link |
We wanted users, creators, publishers and advertisers.
link |
And we wanted the advertisers to put money in just like they do today.
link |
But without going through the Googles and the app nexuses and all these other ad tech companies,
link |
because those companies take out a huge cut.
link |
The Guardian in the UK once did an experiment for a month.
link |
They bought out their own ad space.
link |
They put in a pound and they were paid 30 pence.
link |
70% was coming out to the intermediary vendors they were using.
link |
And that's like the opposite of what the App Store does.
link |
The App Store takes 30% and gives the publisher 70%.
link |
In the old days of the Superstation TBS, the media owner would get 85%.
link |
So these splits have become really unbalanced.
link |
And the middle players, the ad tech vendors are taking out way too much money.
link |
And they're doing something worse, which has been noticed.
link |
They're letting not just the malware vendors, but also the ad fraud side, which fakes the publishers,
link |
and clickbait merchants come in and steal traffic from good sites.
link |
Because once you have a certain audience identified at one site,
link |
Jason Calconas told me this about his experience with, I guess it was in Gadget.
link |
Once he started using an ad partner that was sharing his audience information across multiple sites,
link |
he saw his competitors stealing all his traffic.
link |
And then what's worse is the clickbait sites that just have much cheaper rates steal all that traffic.
link |
And that facilitates fraud, facilitates fake news, all sorts of problems.
link |
So Grave blocks it and then we give users the ability to give back.
link |
And because we invented the basic attention token in Ethereum, we can do this three way split.
link |
And we can give users a share of the revenue.
link |
And if they want to take it out, they can.
link |
Now, unfortunately for us and for all blockchain, the regulators are saying,
link |
we're going to have to know who you are.
link |
There's the Treasury Department's FinCEN agency.
link |
There's the Office of Foreign Asset Controls, OFAC.
link |
There's the other regulators in the federal government that take a very dark look at things like money laundering
link |
and sending money to someone named Osama bin Laden.
link |
So compliance starts to come in.
link |
And even now they're threatening for pure Bitcoin sending to some address.
link |
If you're a Coinbase, you're going to have to know who's at that address.
link |
Make the actual identities of people involved.
link |
Now with Coinbase members, you sign up and they know you and they comply with the regulations.
link |
They're a regulated money services business.
link |
But if somebody's using their own self custodial wallet where they have the hardware private key
link |
and they're not named and they want to send to that address,
link |
our friends in the federal government are talking about requiring at some threshold knowing who that is.
link |
Some threshold that's unreasonable.
link |
It's not that big.
link |
I don't know how this will play out.
link |
I think crypto is here to stay.
link |
I think the beauty of being able to send peer to peer without any bank in the middle without any, you know,
link |
huge wire charge and two day delay and all that nonsense.
link |
It's beautiful and I've used it and I love it.
link |
But we're pragmatists that brave about crypto and we realize that anything like a revenue split,
link |
we can't facilitate without being licensed in a certain way and it requires knowing who the user is.
link |
So our default mode doesn't know who the user is.
link |
It instead imputes to the user's browser some of the revenue and allows that browser to steer it back to the creators.
link |
And we do have to identify the creators.
link |
But as things improve and, you know, who knows how it'll play out,
link |
there should become a day when this full vision can be done more fully on a blockchain.
link |
But regulations and the practicalities of today's blockchains, which are not that fast and not anonymous over time,
link |
you fingerprint yourself over time, we do some of this with the browser.
link |
So one of the ideas of the basic attention token is to make a hybrid system that's stronger than blockchain alone.
link |
It's the browser and the blockchain.
link |
And the browser is this trusted endpoints software.
link |
It's this universal app.
link |
Everyone uses browsers.
link |
The bigger the screen, the more you're in the browser and the less you install, you know, fat clients for things.
link |
I use Slack on Mac OS and it's like a browser.
link |
It's based on an electron framework we used to use and it's just, it's not that great.
link |
Some people at Brave use Slack in Brave as a browser.
link |
And I noticed on the iPad, I use apps less.
link |
The smaller the screen, you know, the browser got handicapped by Apple and Android both.
link |
And it also, you know, can be slower or not have the right, you know, affordances,
link |
the interface with the security limited APIs.
link |
But in principle, with the right permissioning, you can make the web browser just as good as any app.
link |
You make it be a super app.
link |
And that's part of our mission at Brave.
link |
So we want to have the economics that got captured by these big tech companies through tracking and through social networks.
link |
We want to block that for your own safety and then let you opt into a cleaner world where you keep your data defended in your browser and you can actually realize value from it.
link |
So the way our ad system works, I mentioned it being private, but how does that work?
link |
We don't see your data at all.
link |
All browsers are sort of the mother of all data feeds, your history, all your searches at all engines.
link |
Each engine sees the queries you send to it, but it doesn't see the others.
link |
But the browser sees them all.
link |
Machine learning in the browser that you can opt into can study all that in a very complete way and do a better job than Google does.
link |
Google has, you know, cookie and scripts across the web from, according to DoubleClick, they have YouTube, they have Android, they have search, which is still their big revenue link.
link |
But they don't see everything.
link |
The browser sees everything.
link |
And if it can do a good job locally, and this is not advanced machine learning, this is not TensorFlow.
link |
This is like SVMs now, Naive Bayes.
link |
Then you can match intense signals, intense signals from those data feeds, searches, the queries, the history, how much you're scrolling down a page, how much you redid a search.
link |
It's all blind browser algorithm.
link |
We don't see that data.
link |
And then pick the best ad from a fixed catalog per day.
link |
And the catalog is fixed across a large population per day and it only updates once a day because new offers come in and old ones inspire.
link |
Sometimes every week or every month.
link |
And that catalog, and there can be many such catalogs, is sold by our direct sales team.
link |
And so we're making an anonymous audience available to advertisers without the advertisers tracking them.
link |
Instead, each browser is a little machine learning system that's picking the best catalog entry.
link |
Now, the catalog is not the ads.
link |
Those are big, right?
link |
It's not the video or web page.
link |
It's just the link to an edge cache.
link |
And there are many such edge caches.
link |
We're not trying to protect them from seeing your IP address.
link |
It's not really feasible.
link |
We could use Tor, but we don't yet.
link |
And some keywords about the ad.
link |
So it's basically like metadata and a link.
link |
And that's what the catalog consists of.
link |
And that's what the machine learning pick.
link |
And the machine learning is learning about the use specifically locally in order to choose from the catalog of different ads.
link |
Couldn't this possibly be like a multi billion dollar?
link |
Isn't this taking on the Google ad ad?
link |
So like what, I mean, one question to ask that there seems to be some really profound ideas here that are different than what the internet has grown up to be.
link |
If brave or something like brave, the idea is the fundamental philosophical ideas underlying brave win out and runs 95% of the internet.
link |
How does that change the, what, what are the major things that changes about the internet?
link |
So social networks and then the creatives like YouTube creators and all that kind of stuff.
link |
So let's talk about that.
link |
First of all, if brave gets 95%, I'm going to demand a recount because I won't believe it.
link |
I think we're trying to put things into web standards that can be standardized across browsers.
link |
So the main value of brave will be the trust users having us and our ability to give the best deal to users.
link |
70% of the gross ad revenue we give to the user.
link |
And if they go through that KYC process I mentioned, they can take it out.
link |
They can also give it back.
link |
They can take some out, give the rest back.
link |
They can add basic attention tokens to give back.
link |
Some of them turn off the ads because you just don't like ads, but they put in $20 a month.
link |
So I believe Zuko of ZCast fame does that.
link |
And that's very generous because the browser is just anonymously based on his browsing sort of keeping score on how much time he spent on this video on that website.
link |
And if those sites verify in sort of like getting a domain certificate fashion, they can get paid.
link |
They can get part of his $20 a month.
link |
So that vision could go big.
link |
And if it does, I hope it's across multiple browsers.
link |
I don't know that they'll all compete well on the quality of the ads, the quality of the ad blocking and tracking protection.
link |
Those are subject to competition.
link |
It'll take a while to standardize them.
link |
But I think that would be a better world.
link |
It would have less counterparty risk.
link |
The fee takers in the middle, really just the browser.
link |
It's sort of the app store split.
link |
And if we get bigger, maybe we can take even less.
link |
Social networks, creators.
link |
If you look at YouTubers, a lot of them are the indies that are getting some size are getting sponsorship deals.
link |
They're using Patreon.
link |
They're encouraging people to subscribe and give them regular money for Patreon.
link |
But that's centralized your Patreon.
link |
There's censorship hazards.
link |
What if that were a web standard?
link |
What if Brave pioneered it first and we took 3% and we did it in a way that was through your browser so we couldn't censor it?
link |
Do you think it could be standardized to cross browsers?
link |
Can Internet Explorer come in again?
link |
Protocols are easy to copy and they're meant to be interoperable.
link |
So there's a risk there.
link |
And the loyal users might be tricked into leaving you.
link |
Because of that distribution power, you might end up getting stomped.
link |
I don't know. I can't predict the future.
link |
I think antitrust is back on the case finally in the U.S.
link |
and certainly in Europe, DG Comp is doing its thing.
link |
So I'm hopeful that we'll have a period of innovation.
link |
You know, people were talking like Elizabeth Warren was talking about breaking up the tech companies very clearly.
link |
Now, she didn't win and I suspect that won't happen.
link |
But I also suspect that Google might be smart enough to see they should do something more than just put privacy perfume on Chrome.
link |
They should maybe get rid of Double Click or something, divest something.
link |
I don't know. It might happen.
link |
So Brave might inspire Google to completely change the way they're doing things?
link |
They're already doing something you may have read about called the Privacy Sandbox or Flock,
link |
which they have this bird metaphor going, Turtledove, Fledge.
link |
But these systems have been very googly, kind of overengineered.
link |
And yet, depending on differential privacy, which has weakness over time,
link |
if you know how that works, it's kind of injecting noise to hide you in a crowd.
link |
But over time, an adversary can pull you out of the crowd.
link |
This doesn't look like it's going to become a standard like Apple, Brave, Mozilla.
link |
We're not going to just say, oh, Google, you saved us.
link |
You've invented the Privacy Sandbox, so I'll just adopt it.
link |
Not going to be that easy.
link |
It's going to be more like pieces of what we do in Brave,
link |
the synonymous ad matching or the blind signature cryptography we used to confirm the ad impressions.
link |
That's David Chum's invention.
link |
That could get standardized.
link |
In fact, some of that is being standardized.
link |
Even Google's in favor of so called trust tokens,
link |
which are Xiaomi and blind signature certs.
link |
But they're not using them for ad confirmations,
link |
because they don't want to blow up their own business.
link |
And they need to let some of the publishers they serve have other ad tech scripts on the page.
link |
And so they're kind of caught.
link |
And this is something I realized doing Brave.
link |
I thought, what's Google's innovators dilemma apart from just being mature
link |
and having trouble innovating?
link |
It's that they have come to depend on this ad tech system that has all these vendors
link |
that publishers rely on, because publishers aren't technical enough.
link |
And I feel for the publishers, but I realized the users have to come first.
link |
And if you give the users a better browser that's faster,
link |
then you'll get enough users to give back or support publishers.
link |
The speed and the battery savings and the data plan savings are significant.
link |
There's so much bad JavaScript involved in ad tech that if you block it
link |
and you sort of chop off the what's called the programmatic waterfall,
link |
which chains a bunch of requests.
link |
Yeah, that's one of the incredible things about Brave.
link |
I guess you're saying you should attribute it to the fact that the message JavaScript,
link |
no offense, it's not my exclusive.
link |
I mean, Brave just feels faster.
link |
I mean, Chrome was fast.
link |
One of the things that it was like impressive is it showed that browsers could be really fast.
link |
And Brave is even faster than that, which is incredible.
link |
And it saves the network, which means data plan, it saves battery
link |
because the radio consumes your battery when it's running more to do those requests.
link |
And it's just stunning how many there are.
link |
Some of my Google friends are like, oh, that's just that bad site.
link |
And you actually do a survey of web pages.
link |
They're mostly like that.
link |
I know Google engineers could make everything super efficient,
link |
but they can't, especially in antifrust court, do it.
link |
They cannot take over all the publishers and do that.
link |
They're trying with accelerated mobile profile AMP.
link |
They're trying to pull publishers who are like, oh, you poor publishers,
link |
don't know how to make your pages fast.
link |
Put them on our AMP system and we'll give you extra placement in the search carousel.
link |
That's an antifrust problem for one.
link |
But it's also publishers we talk to hate it because it degrades their brand.
link |
Now, they look like a gig writer wrote a piece that's got Google's framing an AMP URL on top of it.
link |
And they're trying to fix that too.
link |
But it just looks like Google's borgifying all these publishers
link |
and they don't want to be plugged into the borg cube.
link |
They want to build up their own brand and have loyal readers.
link |
So, you know, I'm in favor of giving the users power to help all the publishers
link |
and the little platoons and the creators.
link |
And so we talked about Patreon.
link |
What about social networks?
link |
Well, they're inherently like search a global algorithm.
link |
You're trying to find friends of friends.
link |
You're doing the transitive closure of a graph induced by this friend of relation.
link |
But you should own your friend relation.
link |
You should own your posts.
link |
They shouldn't be owned by somebody else who can take them down or send to them.
link |
And your friend relations, you should be able to find those friends on other networks.
link |
And that's why I've tweeted about this.
link |
I haven't built it yet.
link |
What if the browser could keep track of those for you?
link |
What if the browser could maybe combine Facebook and Twitter
link |
and you could find your friends on both and you could have a sort of...
link |
So that relationship is not owned by Facebook or Twitter.
link |
It's owned by you through the browser.
link |
They don't have terms of use and they'll say they own it.
link |
But if they zap you on one and you're still on the other, your friends find you
link |
and the browser could preserve a combined view.
link |
You could resurrect almost across networks.
link |
It's something I want to maybe quickly ask you about.
link |
On that front, there's been quite a lot of centralized...
link |
We talked about Wall Street Bets and then Robinhood.
link |
There's been centralized banning of different accounts
link |
and removing like Parler, for example, from AWS and this kind of overreach of centralized control.
link |
Is your hope that it's possible to...
link |
What are your thoughts about that in general?
link |
Is it possible to create tools that give individual people the power
link |
to fight back against overreach of such control?
link |
So we're talking about oligarchy, I do think.
link |
And that if it controls a nation state, that's formidable.
link |
It's the tax and the police power, the military power.
link |
It means that you may have the great firewall of China.
link |
You may have people in China who are jailed because of their tweets.
link |
This is a serious threat.
link |
I can't minimize it or say that we'll win.
link |
I don't know how it's going to go.
link |
But I do think, like I said earlier about the cutting of reason,
link |
people find ways around things, the internet routes around censorship.
link |
And this is not to endorse any particular bad faction.
link |
One of the things that happens when you try to wave the free speech flag too much,
link |
I'm not going to censor anything and you get colonized by terrible, terrible people.
link |
I don't care if you call them neo Nazis, some of them could be doing illegal things.
link |
And you don't want them colonizing because it'll ruin your reputation and destroy your business.
link |
So what you really want is that kind of user first subsidiarity, that subjectivity.
link |
I want my social networks to be composited in some multi social user interface
link |
where I don't lose track of people across networks.
link |
And if they leave one or they get banned from one, I can find them on another.
link |
I can still sort of thread them together.
link |
And this didn't happen because browsers got captured by the central powers.
link |
Why did they get captured mostly because of search?
link |
And search is a central algorithm.
link |
So Larry Page said this too many years ago.
link |
He said, with search, you're giving up a little privacy by handing the query over to us.
link |
And we'll error correct it.
link |
Alan used to be a Google executive.
link |
He said, oh yeah, we used to laugh.
link |
They'd all be doing typos and they'd be typing the wrong word.
link |
And we're like, no, dummy, type that query.
link |
And it's like, okay, Google might want to dial back that ego a little bit.
link |
But yes, you do see all the queries and you can improve them
link |
and you can find the best results.
link |
And that was Google's forte.
link |
When we did the Firefox deal in 2004, Google was really good.
link |
And over time, SEO, which is an adversarial game,
link |
and Google itself buying all these companies and crowding its own results page
link |
with its own tied in stuff, the YouTube.
link |
The slipper slope that happens when you have control over these kinds
link |
of really important mechanisms.
link |
Yeah, monopoly capitalism or cartel.
link |
You get this with the Robin Hoods and the hedge funds.
link |
You get sort of the money interest takeover and kind of abuse their power
link |
and wear out their welcome.
link |
So how do you get around that?
link |
You have to have either new land to go to, which some people's ancestors,
link |
not mine, did to found the country, mostly Irish, German.
link |
You have new virtual space people go to.
link |
And that requires an ISP or a Colo Center or Amazon to host you.
link |
It requires domain name registrar who will not strike you.
link |
And so when Parler was taken down, I thought that was egregious.
link |
Parler, it was not well designed.
link |
And I tried it out because I tried all these things, but I didn't use it.
link |
And I also felt they were being unfairly scored for not moderating
link |
because you can find tweets to this day that are horrendous
link |
and threaten all sorts of violence.
link |
Whereas Twitter, why isn't Twitter being taken down?
link |
But so it was very selective.
link |
It was the insiders who have the power are going to take out the newcomer.
link |
And it looked bad, sort of like the hedge funds sort of in GameStop.
link |
You know, you're seeing a piece in Time Magazine this week that's like
link |
basically saying, yeah, we interfere with the election,
link |
I don't know if you've seen this piece yet.
link |
If you tried to say that as a Trump supporter in November after the election,
link |
you get banned from Twitter.
link |
But now time in its Twitter account is saying, you know, we saved the day.
link |
It's AFL CIO and big business.
link |
The Better Business Bureau got together, kept Trump from spreading fake news.
link |
So the country's kind of broken.
link |
I don't know how to fix that.
link |
The oligarchs have run wild in my opinion.
link |
And big tech is in the antitrust doc.
link |
What's going to happen?
link |
I don't think they get out.
link |
I think some of the DOJ and certainly the state cases,
link |
because they're separate cases, are not going to go away
link |
just because somebody got elected differently.
link |
And these are career prosecutors and they have a strong case.
link |
And Google's smart.
link |
Microsoft almost got split up, right?
link |
The judge, Thomas Penfield Jackson, he overreached.
link |
He didn't hold a hearing about the remedy.
link |
He just said, I'm going to break you up.
link |
And Microsoft appealed on the higher level court said, go back and figure this out.
link |
You're not breaking them up.
link |
You didn't even hold a hearing.
link |
And when they got back, Microsoft said, let's settle.
link |
We don't want to get broken up because Jackson was going to make the
link |
Opsco, the operating system company and the Opsco office,
link |
And that would have been a huge blow to Microsoft.
link |
But ultimately, I don't know if you're optimistic or cynical
link |
about the possibility of breaking up big tech.
link |
To me, I'm optimistic that tools like Brave,
link |
I love the idea of owning your friendships.
link |
And users more and more owning the stuff is the only real way.
link |
Unfortunately, it's like the Wall Street bet subred is the only real
link |
way to fight the centralized power.
link |
You can't break them up with the regulation.
link |
It's very difficult.
link |
Certainly, I don't want to wait for the law.
link |
Netscape was long dead or acquired by AOL and effectively dead.
link |
It was only most of the return in Firefox to the market by the time
link |
that the USV Microsoft case was finally settled and the penalties were put
link |
And yet, Antitrust has a role to play.
link |
Those penalties caused Microsoft to kind of turn away from the web.
link |
They did Windows Vista and they thought the web's too painful.
link |
We got punished in court and we had to standardize things with those
link |
So they ran back to proprietary lock in and Windows Vista flocked.
link |
Longhorn, remember.
link |
Now, what Google, I was going to say, but Google is smart enough,
link |
they won't get split up.
link |
They'll split something out to get off the hook, I think.
link |
This is a complicated subject, but I myself was so I decided to journey
link |
from the world of being a researcher at MIT and potentially doing a
link |
And I've been thinking of, you know, I wanted to come to Silicon Valley
link |
It's the land of the entrepreneur.
link |
And there's a lot of my friends.
link |
A lot of them are successfully have been entrepreneurs themselves have
link |
said, do not come to Silicon Valley.
link |
You've started, you ran amazing teams of engineers.
link |
You started a lot of successful businesses.
link |
I wondered if you could comment on why a lot of people are leaving
link |
Is there something that could be fixed about California?
link |
If you were starting a business today, would you consider somewhere else
link |
like Austin or some other place or is Silicon Valley still?
link |
Is it just a little lull?
link |
Everybody's being overdramatic during this particular year of the
link |
coronavirus and so on.
link |
I think, you know, even Austin's getting overheated here.
link |
I've had relatives and friends move to Texas within the last few months.
link |
So Texas as a whole is a big place.
link |
And, you know, people are moving to Florida.
link |
There's a big movement toward Miami.
link |
Peter Tielke, these people, the mayor has been very business friendly
link |
about it, which I think is just good politics.
link |
America is fundamentally a commercial republic.
link |
So you would think this would be what's happening.
link |
For a long time, California was the golden state.
link |
I came here in late 76 when I was a teenager.
link |
It's in crushing debt to the lockdowns.
link |
It's got the highest taxes.
link |
That's got to matter.
link |
People will do a fleet at high taxes.
link |
It's got likely fires every year because of the deadfall.
link |
It's not global warming.
link |
It's because the forest weren't managed like they had been in the first
link |
part of the 20th century.
link |
Just, I would say, corruption at all levels, especially up to the governor
link |
who, you know, famously was eating at the French Laundry and claimed
link |
the outside was inside and they were out in masks off and it was great.
link |
Do what I say, not what I do.
link |
Rules for thee, but not for me.
link |
When you see that in leadership, people either run or they get rid
link |
of the leadership.
link |
So there's a recall drive, which is about to reach the threshold.
link |
Or in the old days, they get their guns, right?
link |
You don't put up with this junk.
link |
But ultimately, the thing that made Silicon Valley a special place,
link |
it gave freedom to like young kids, entrepreneurs, young minds,
link |
brave minds to think bold, to try different stuff.
link |
I mean, even if the taxes are high, so outside of financial stuff,
link |
outside of all of that.
link |
Housing is super expensive.
link |
Housing is super expensive.
link |
Okay, everything about startups is hard.
link |
And it was narrow and they didn't plan the roads, right?
link |
They got rid of public transportation in LA like the Who Framed Roger
link |
Rabbit cartoon show that used to have trolley cars in Portland too.
link |
The oil companies and the DOD conspired to build highways
link |
and make cars dominant.
link |
And the rights of way are long gone.
link |
Like Elon's going to go underground.
link |
And I wish him well.
link |
That's probably the only way to do it now.
link |
But is it still a place?
link |
Do you think it's possible that Silicon Valley is still a place
link |
where magic happens?
link |
Where the next Google's built?
link |
Where the next, I mean, Brave is built where?
link |
I think all good things come to an end.
link |
I think the problem is Silicon Valley had strong network effects
link |
through the angel investor networks and the wealth effect.
link |
And originally you have to give the federal government credit.
link |
Like the ARPANET was a government project.
link |
Let's not kid ourselves.
link |
This wasn't wild free market, you know, libertarian capitalism.
link |
This was all Cold War stuff.
link |
You had out of the academia, you had Shockley and then the
link |
the Traders Aid and Fairchild and Intel.
link |
But now, you know, when's the last fab that was built in the valley?
link |
Micro Unity might have been the last.
link |
We built a fab in Sunnyvale and in Micro Unity in starting early
link |
in the 90s and now the fabs are overseas.
link |
And the one thing that I would say that the oligarchs have
link |
intentionally done in both parties is sort of labor and environmental
link |
protection law arbitrage by going where the labor is cheaper
link |
and the environmental laws aren't as strict.
link |
And, you know, that's polluted the hell out of parts of China,
link |
but it's made things, you can make cheaper junk.
link |
And this is not a story that's over yet.
link |
So what is Silicon Valley for now?
link |
It's for the network effect, the brain trust of who you know,
link |
the parties, the Stanford sort of network.
link |
That's fragile too over time.
link |
I'm afraid, right?
link |
Stanford, a lot of good professors are like they still filter,
link |
you know, mainly based on socioeconomic status,
link |
but it's kind of a skate school.
link |
I had a friend hired out of Harvard 20 years ago at Netscape
link |
and we talked about Harvard and he said, yeah,
link |
there's still professors who are great on the curve.
link |
And I said, oh, yeah, I don't,
link |
they're any doing that at Stanford.
link |
And he said, oh, it was shocking.
link |
Some of the students got C's and D's and they were crying.
link |
It's like, yes, that's right.
link |
The precious deers can't take that at Stanford.
link |
So they get A's and B's.
link |
Now you look at China and say what you all about China,
link |
they prove Russia to a lot of math science training,
link |
a lot of engineering, a lot of people who are doing their coursework
link |
to get the A's and B's.
link |
So I'm an American.
link |
I'm born on the 4th of July.
link |
And America, as I say, fundamentally is a commercial republic.
link |
You can try to make it something else.
link |
You can see it's the new Atlantis and mystify it.
link |
You could talk about it in a more, I think, correct way,
link |
which is 13 colonies that grew.
link |
And then there's a lot of local or original design anyway.
link |
The Federalist Papers talk about this.
link |
There's just a lot of subsidiarity.
link |
But that's been eroded over time.
link |
And like I say, a lot of the offshoring is hurt.
link |
So what happened with coronavirus?
link |
People working from home.
link |
At first it was funny because I have friends at Google who used to grumble
link |
that not only did they have to come into the office,
link |
if they joined a different team that was sent to a different office,
link |
Or if the VA team was reconstituted in Munich,
link |
which it was after Lars Bock just got tired of JavaScript,
link |
that they hired in Munich,
link |
or they hired PhDs in Germany and moved into Munich.
link |
With coronavirus, everyone's working from home,
link |
and it's like, what a relief I can work for Google from home.
link |
But then the next shoe dropped,
link |
and people started asking Mark Zuckerberg,
link |
hey, can I move to my hometown in the Midwest?
link |
And he said, okay.
link |
And they said, I'll kind of keep getting myself in Valley Pay.
link |
We're going to figure out what your cost of living there is,
link |
and we're going to adjust your pay accordingly.
link |
And these colonies and these little mini experiments
link |
that all combine to the big giant experiment,
link |
I have a, I don't know,
link |
I have this vision of America,
link |
which is a country I was born in Russia,
link |
and this is truly a wonderful country.
link |
I wasn't born on the 4th of July,
link |
but I might as well be.
link |
People still flee here.
link |
I still, and I'm a red blooded American at this point,
link |
and I have a sense that we've figured it out somehow.
link |
If Silicon Valley burns,
link |
another place will come up in this place
link |
that even more innovation and people will move.
link |
And the remote work might change fundamentally
link |
how we work or might not.
link |
It might just give you the freedom to then create
link |
many other small Silicon Valley throughout the place,
link |
like Austin included, but other places as well.
link |
And we somehow figured it out.
link |
And I think that's true,
link |
that there will be more mobility
link |
and maybe new places that come up.
link |
I don't know if Silicon Valley has, you know,
link |
passed some sell by date because it did hurt.
link |
The coronavirus hurt.
link |
The lockdowns hurt in the sense that
link |
part of what keeps things going is social.
link |
And so a lot of young people,
link |
even before coronavirus, moved to San Francisco.
link |
It was very strange to watch
link |
because in the 80s, we all lived in the valley
link |
and it was less populated.
link |
And San Francisco was grungier.
link |
It was more like dirty hairy in the 70s.
link |
But by the 90s, and Jamie runs a nightclub there,
link |
and he's talked about this.
link |
You had sort of wealthy tech people moving in,
link |
south of market, fancy townhouses being built.
link |
And that's continued in such a point
link |
that it's almost like, what's the movie
link |
by the South African director Nils Jody Foster
link |
up in the space colony.
link |
Matt Damon is the guy on the earth who has to go up.
link |
And anyway, it's about the stratification.
link |
It's about the great inequality.
link |
The people in the space station have like amazing
link |
medical autodocs that can extend their life
link |
or save them cure cancer.
link |
People on earth are all suffering,
link |
ground down in poverty.
link |
And, you know, that sort of happened.
link |
While I was here, you saw a lot of money
link |
drive prices up along the narrow peninsula,
link |
and the single people wanted nightlife
link |
so they were in the city,
link |
and the condos in the city got super expensive.
link |
And I know even Google friends who are, you know,
link |
socially responsible say,
link |
we should have more housing built.
link |
We should have yes in my backyard, not in my backyard.
link |
But that's not happening, as far as I can tell.
link |
And from the government to the incumbent,
link |
you know, landowners and renters,
link |
it's just not happening.
link |
And that has to drive people away.
link |
And I appreciate that people come here
link |
and you should wait for the prices to moderate.
link |
But a lot of people are going to go
link |
where the prices are lower.
link |
You, and sorry for silly questions here,
link |
but just looking back,
link |
you have created things,
link |
have been part of creating things
link |
that have transformed this world,
link |
the world of technology,
link |
perhaps more than almost anything else.
link |
But you're still a human being.
link |
And unfortunately, this ride ends.
link |
Do you ever think about your own mortality?
link |
I mean, I'm Roman Catholic,
link |
so I am not afraid of death.
link |
I think a lot of people who have problems with death
link |
are suffering from some lack of either faith
link |
in their transcending death,
link |
maybe they don't have children
link |
or they feel like, you know,
link |
they get later in life
link |
and they feel like they've missed opportunities
link |
to do something that endures.
link |
And I sympathize a lot because I'm old.
link |
I got married fairly old,
link |
so I understand all that.
link |
Nothing human is alien to me, as Terence said.
link |
But I don't fear it, no.
link |
What do you hope your legacy is?
link |
It's going to be JavaScript.
link |
I think, you know,
link |
my legacy has more to do with my children
link |
and their children.
link |
I think it also has to do with web standards.
link |
It has to do with things like Brave.
link |
The things we did with Firefox,
link |
when we did, you know,
link |
I'm not going to over solve Brave,
link |
but I think Brave is important
link |
and we will continue to prove this in a way
link |
that counts for many decades to come.
link |
whatever its future fortune,
link |
showed you can restart the browser market.
link |
This thing you said about people opting out
link |
and routing around,
link |
you don't need everybody to do that.
link |
It's more like Talib's stubborn minorities
link |
It's the lead users,
link |
Airfront Hippos lead users.
link |
You can be a few percent,
link |
you can tilt the market.
link |
And that can be done in spite of the incumbents
link |
of money interests not being in favor of what you're doing.
link |
So I think what we do with Firefox won't be forgotten
link |
and it needs to be done more
link |
and we're doing it with Brave
link |
and you could argue that other projects are doing it
link |
in some ways blockchain is doing it.
link |
The Robinhood take down the use of Robinhood
link |
by the Wall Street Bets kids, similar.
link |
So yeah, that kind of spirit endures.
link |
And I think it in some ways it's American, right?
link |
It's not hard revolutionary.
link |
It's not trying to burn the past and destroy everything.
link |
It's more like we have these certain, let's say, rights.
link |
We have duties too.
link |
So there's some debate about which comes first in American jurisprudence
link |
and the founding documents.
link |
But as long as things are working,
link |
we'll be like pragmatic Americans,
link |
like the Tocqueville described in his writings.
link |
But if things get too out of whack for one reason or another,
link |
too unequal, too oligarchic and abusive,
link |
we're going to start our rights
link |
and even a few of us can do it.
link |
And even in the American Revolution,
link |
it was a minority who fought
link |
and put their lives, treasure and sacred honor at stake.
link |
A bunch of people went to Upper Canada.
link |
I think it was called Ontario.
link |
Yeah, that's the beautiful thing.
link |
I mean, that is at the core where your work stands for
link |
is that a few people can have the power to transform society
link |
with just a few radical ideas
link |
with just a little bit of code to change the world.
link |
And that's empowering.
link |
And that is the American way.
link |
That's why this country is, I believe,
link |
the greatest country in the world.
link |
Romance says it too much,
link |
but I think some special things have already happened
link |
in this country and will continue to happen.
link |
And that spirit can continue.
link |
No matter who comes here,
link |
they can adopt those folk ways and that spirit.
link |
Brendan, as I can't tell you how much I was freaking out,
link |
how much of an honor it is to talk to you.
link |
You're an incredible human being.
link |
It's one of my favorite conversations ever.
link |
Thank you so much for wasting all this time with me.
link |
I really appreciate it.
link |
Oh, it seems like you have a breeze.
link |
Thank you for listening to this conversation with Brendan Nike.
link |
And thank you to our sponsors,
link |
Jordan Harbinger Show,
link |
Sunbasket Meal Delivery Service,
link |
BetterHelp Online Therapy,
link |
8Sleep Self Cooling Mattress.
link |
Click the sponsor links to get a discount
link |
and to support this podcast.
link |
And now, let me leave you with some words from Jeff Atwood.
link |
Any app that can be written in JavaScript
link |
will eventually be written in JavaScript.
link |
Thank you for listening and hope to see you next time.