The Web Commons:
Past, Present and Future

 

@briankardell

Developer Advocate At Igalia

Open JS Foundation

W3C
I'll tell you more as we go..

"The Web is Amazing."

- Me: just now.

 
 

And yet...

Question...

Will we be talking about the Web in 2049?

You are here

2019

I'd like to convince you that it very much is

Chaos 

Unprepared

Revolutions

 Ever Better

Figuring stuff out

2004

The second half was better than the first....

Chaos 

Unprepared

Revolutions

 Ever Better

Figuring stuff out

2004

The second half of that was better than the first half too

Chaos 

Unprepared

Revolutions

 Ever Better

Figuring stuff out

2004

And a whole bunch of that value is increased in the latter half of that.

Chaos 

Unprepared

Revolutions

 Ever Better

Figuring stuff out

2004

It's getting better, faster.

Learning and Correcting inertia of:

  • Our ideas about the future
  • Our ideas about standardization
  • Our ideas about competition
  • Our ideas about licensing 

 

All established very early.

The First Half

The good parts, and the not so good parts

May 1989

The web was born into chaos.

 

I mean computers,

hardware,

the internet,

and software,

and standardization...

All this stuff seems somehow different, and we're just not sure how to handle any of it.

 

It wasn't the internet, that was decades earlier Vint Cerf

It wasn't the idea of hypertext, that was this guy Ted Nelson.

hypermedia already existed.

what is hyperstuff?

It wasn't documents marked up with tags: That was Charles Goldfarb

It wasn't putting several of those things together.

Hypertext Markup Language
(.hml)

Ian Richie's Guide called those documents...

European Conference on Hypertext in Versailles...

Imagine if this were all universally addressable and referencable?

?

I guess I have to build it...
- tim

A shrewd move...

Tim's pitch was a little like this.

Web Content

existed before the web.

<a href="procedures">

   CERN Safety Procedures
</a>

?

Meh, it kind of almost exists already?

CERNDOC 

tens of thousands of these!

Actually the first browser... Read Write!

parsing...whoops.

happy accident?

grave misfortune?

The first several browsers were

'for the commons'

Those people want more from HTML, and why not, really?

Keep in mind that this was HTML

  • <title>
  • <p>
  • <ul><ol><li>
  • <h1>...<h6>
  • <a href>

VM

Programs

Browser

document

moar programs*

It's hard to overstate the importance of this

NCSA: Hey, let me show you this thing..

Enter: Mosaic

The licensing terms for NCSA Mosaic were generous for a proprietary software program.

No browser you know of today exists here

 

~ 1.9 m households have internet access

a tiny few of those using the web

hardware is wildly divergent

It was slow.

view-source

1994

  • IETF WG / HTML2?
  • Netscape
  • W3C*

1995:

  • New tags
  • IETF WG / "HTML 3"?
  • JavaScript/DOM 0
  • Java
  • Internet Explorer
  • Internet Explorer 2

1996:

  • W3C Publishes CSS
    • IE3 Supports it
    • Netscape doesn't

 

IE on windows and mac!

IE supported VRML and sound

I feel like we need standards? Something to prevent this fracturing?

How do you do that, exactly?

Like I said..

Why are there so many standards bodies?

 

ANSI

ISO

Kronos

MPEG

IETF

W3C

ECMA

WHATWG

OASIS

OSGI

OMG

SMPTE

OIF

IEEE

IEC

ASME

And why do they all work differently?

VM

Programs

Browser

moar programs*

You know... at least for 'real applications'.

HTML... for applications?

Don't we need... something else?

<input>

<a href>

<table>

<p bgcolor color>

<hr>

<center>

<img>

Many "Real" Designers and Engineers...

Really nice user experience

Works offline... 'online' was the special case

Written with programming languages

which were efficient

and had abstractions about applications and GUI

None of that

1996

And yet...

1997-1998

8-9 years in...

A pivotal time.

First JS Specification

HTML 4

CSS 2

4.0 Browsers

<xml> + DOM

1998: HTML is Dead

The future is coming, and it is XML based

Oh and...

9 Years in - the return of legit

Open Source / Free Software Browsers:

Netscape

KHTML

End of 2001: IE6 </browserwars>

Bye this team.

Hello this team.

HTML... for applications?

Like... for real ones we need something else, right?

  • A virtual machine.
  • A markup language about application ui
  • Data services
  • A stylesheet language
  • Some kind of imperative programming language,but UI designed to need much less 

You know... for the real applications.

Meanwhile.... Developers...

2004

The Second Half
 

 

Two revolutions, and stopping power.

Web 2.0

The Web as a Platform

XML

SOAP?  WDSL?  BPEL?  XSD?

"REST"

{ "what-about": "this" }

Meh, it kind of almost exists already?

JSON I CHOOSE YOU

??????????????????????????

??????????????????????????

June 2004 and everyone is like...

Web 3.0?

HTML... for medium level application requirements?

June 2004, W3C workshop...

Ian Hickson, Opera

"Hixie"

Nope.

Opera, Mozilla, Apple

Europe usage share data from ADTECH's press releases; this is an ad serving company
Period
Internet
Explorer

Firefox

Safari

Chrome

Opera

Other Mozilla

Netscape

July 2004 93.08% 1.64% 0.99% 2.62% 0.97%
January–April 2004 94.72% 0.73% 2.50% 1.49%

BUT REMEMBER...

  • Microsoft isn't onboard
    • They dominate
    • No browser team
    • A 'next web' competitor
  • ​Apple has no browser
  • Google has no browser
  • Mozilla's power is fractured, small

JavaScript Core

A Primer for Busy People

Webkit

KHTML in 10 Seconds

Webkit

Mozilla + Webkit

Open source, with some backing, to implement open standards.

Is Google building a browser?

Rapidly experimenting in open source commons and we begin coalescing..

First draft of... the HTML Parser Specification?

Holy standards reality check Batman!

no tests?

To give you an idea what a change this was, here's what Opera would end up calling their new parser...

Ragnarok

2007: OK Let's do HTML5

... you know... for medium-sized applications.

IE7

2007

September 2008 Surprise...

Browser

"Chrome" & Internals

OS Implementations

Webkit

A Primer for Busy People

Webkit in 10 Seconds

Webkit based Chrome

v8

A Primer for Busy People

Webkit in 10 Seconds

Webkit based Chrome

JavaScript Core

Spec compliant, but.. things break O_0

This, as it turns out, is one of those things standards organizations are really good at managing.

Apple + Google

vs

Mozilla

vs

Opera

vs

Microsoft?

...vs Adobe?

open

proprietary

Apple + Google

vs

Mozilla

vs

Opera

vs

Microsoft?

...vs Adobe?

HTML 5

2008: Adobe Air

You know, for real applications...

1989-2009:  20 Years in we finally have 

decided this web is, in fact, not dead.

Harmony

Let's get to the last 10...

Web developers are competing, adapting and connecting and improving things in open-source

May 2009

v8

HTML5 Looks great, but can we use it?

The economics don't look great.

IE is holding us back.

Polyfills

Feb 13, 2013

"In a surprise move, the Opera Web browser is moving from using its Presto Web rendering engine to using the popular open-source Webkit engine."

The commons gets BIGGER

 Apr 3, 2013

Google forks Apple’s Webkit to create Blink rendering engine for Chrome

 Apr 3, 2013

Opera confirms it will follow Google and ditch WebKit for Blink, as part of its commitment to Chromium

Privately, a lot of people started talking...

A lot of us begin to find common ground.

Does this seem kind of broken to you?

Doesn't it seem like a lot of the problems are more about how we get there?

The Economics of Web Standards: Incomplete.

Maybe we can fix this....?

June 10, 2013

  • We have a lot of greatness buried here

  • The way we do this is not great... we can do better.

  • Prioritize explaining the existing architecture in APIs

  • Give devs ways to plug into and extend these, and new powers, and experiment

  • Developers are important to the process

  • Let's find a standardization model that actually works like this...

Turn!

and also...

Houdini

In the next couple of years.

Sydney 2015: Houdini

Everyone:

So much magic explained and connected... increasingly better

All the loady bits:
Fetch

Request

Response

Streams

Promises

URL

Cache

 

ElementInternals

AOM

ShadowDOM

:focus-visible

inert

formData/Blob

ArrayBuffers

DataView

WASM

Modules

Classes

Public/Private/Static

async/await

 

 

 

Experiments are experiments

 

Developers can seek consensus/champions/manage IP

 

Official incubation is more serious experimentation

 

 

Amazing Progress!

Meanwhile...

2017: CSS Grid

2005?

1996

Because, open...

But more on that at the end

By the end of 2018...

Our commons is rich, deep and multi-layered.

content + open data

All of the standards implementations are open

Not just new browsers - whole new ideas... 

This Part

Embedded

We all move forward together.

Lots of common interests,

Lots of contributions,

Increased involvement,

Increasing ways to be involved at many levels

The same basic tech, standards, skills 

and code are increasingly useful everywhere.

 

Standards are starting to click...

2019-*

Possible futures articulated

Science!

More collaboration and outreach

Sure... But what about the politics?

Why does browser X think Y is more important than Z?

We'd like to think of process like this...

Here's a thing that doesn't get talked about a lot...

A lot isn't stuck over silly politics...

They're stuck because of organizational problems and priorities

Browser vendors have...

  • Departments
  • Budgets
  • Managers
  • People with special skills
  • and...

this problem...

It's only a standard if everyone makes it through

This is a feedback loop.

The pipeline gets jammed.

Imagine if we could fix this problem?

We love the commons

  •  V8, JavaScriptCore, SpiderMonkey, ChakraCore
  • Chromium, Gecko/Servo, Webkit

ResizeObserver

Responsive image preloading

Web Packaging

and...

MathML in Chromium!

RSVP Now

Abstractions: Celebrating The Web Commons - past present and future

By Brian Kardell

Abstractions: Celebrating The Web Commons - past present and future

  • 1,069