Is the Open Web the Platform for Tomorrow

Is the Open Webthe Platform for Tomorrow?

Is ‘Web 3.0’ where we use something else instead?

So far, the web has won the platform war

Why?

4 Reasons:

Zero InstallLazy TextUI ModelOpenness

Benefits of Zero Install

•Saves time

•Builds trust

•Reduces clicks

•Saves disk space

Lazy Text means web pages are:-hackable-debuggable-learnable-Postel’s Law compliant

The Web UI discourages:

• Overlapping windows

• Complex dialogs

• Hidden options

• Deep menus

4 Reasons:

Zero InstallLazy TextUI ModelOpenness

Is the web good enough for tomorrow?

What the web is missing ...

Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

Javascript is currently slow

Which means:

• Slower everything

• Smaller page transforms

• Jerky annimation

• Weaker cryptography

• Less data manipulation

• Wasted user time

• Wasted developer time


Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

SVG/Canvas are not in IE

Which means:

• Pages are bigger

• Slower everything

• Maps are images

• More plug-ins

• Server rendering hacks

• Charts can’t interact



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

Available fonts are limited

Which means:

• Headlines are images

• Search is less effective

• More server load

• Less readable pages

• Things don’t look right

• Unhappy marketeers



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

Video isn’t part of HTML

Which means:

• More plugins

• Bad page/video interaction

• Codec confusion

• More security risks

• No video mash-ups

• Expensive tools



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

Local storage is a hack

Which means:

• Offline is hard

• Slower page load

• More complex servers

• More security risks

• Slower page use

• Everything is server based



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

3D is basically impossible

Which means:

• Poorer presentations

• Poorer graphics

• Worse visualization

• Less games

• Slower pages

• Bigger page downloads



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

Too many design mistakes

Which means:

• Lack of trust

• Users scared off the ‘Net

• ‘Identity theft’ is real

• The possibility of cyberwar

• Safety becomes a lottery

• Extortion is possible



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

It’s all too hard to program

Which means:

• Developers fail

• Projects fail

• Less reliability

• Lower security

• Libraries like Dojo/DWR/...

• Library selection issues



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

HTTP is limitted

Which means:

• Less interactivity

• Wasted network traffic

• More server complexity

• Everything is server based

• Peer to peer impossible

• Mega-hacks like comet



Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

There is still a shocking amount left to get right

Is the web good enough for tomorrow?

(Not yet)

Do the failings of the Web mean we need more options?

And if so,what are theoptions?

Zero InstallHard Easy

Lazy TextUnreadable Readable

UI ModelDesktop Web

OpennessClosed Open

Evolutionary SpeedSlow Fast

Is Openness the enemy of Evolutionary Speed?

JavaFX / Flex / Silverlight are all young

• It’s easy to move fast without the constraints of many users

• So current rapid evolution is no guarantee of future growth

Parallels between closed systems and monopolies:

• Both restrict choice

• Both give vendors more power

0

25

50

75

100

‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06 ‘07 ‘08

1.5

2.0

1.0

3.0

4.0

5.0

5.5

6.0

7.0

M2

M2.1

M3

M4.0

M4.5

M5.0

(U3)

U4.0

U5.0 M5.2 8.0

0

25

50

75

100

‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06 ‘07 ‘08

1.5

2.0

1.0

3.0

4.0

5.0

5.5

6.0

7.0

M2

M2.1

M3

M4.0

M4.5

M5.0

(U3)

U4.0

U5.0 M5.2 8.0

16 major releases of IE Across 4 platforms

2 major releases of IE? Now only on Windows

6 years

IE got to 90% market share in

2002

1996 20082002 6 years

0

25

50

75

100

‘93 ‘94 ‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06 ‘07 ‘08

2.0

1.0

3.04.0

Netscape Usage Share

4.5

6.0

5.0

4.8

7.0

If Netscape had not been a monopoly;

It might not have been sidetracked into servers

If Netscape had not been a monopoly;

Would it have chosen the risky rewrite path?

Ideal position:

• Multiple implementations

• At least one Open Source

• None with a monopoly position

The web has seen 3 browser monopolies

• Mosaic

• Netscape

• Internet Explorer

We may soon have the first sustainablenon-monopoly situation

How can we encourage the web to evolve faster?

More browsers are generally a good thing

A browser can’t all compete on all fronts at all times

There are now 5 good browsers, but they share components

Browsers should be encouraged to take risks

Browser Wars 1.0 took place before we had the basics right

Now the basics are much better, we can risk another browser war without spilling users blood

Is the web attacking the issues we logged earlier?


Speed

Vector Graphics

Fonts

Video

Local Storage

3D

Security

API Simplicity

Flexible Comms

TraceMonkey/Chrome/Squirrelfish

SVG+Canvas in most (not IE)

Firefox+Safari working on it

HTML5 and Minefield

Gears, HTML5, IE8

Minefield

Content restrictions / SOO

Dojo / DWR / jQuery / etc

WebSocket

Boat anchors of the web ...• Netscape 4• Internet Explorer 6

TL;DR

Should the update process be transparent?

• The inputs to the question “Should I upgrade” are usually too complex

• Asking too often is annoying

• The upgrade is inevitable given time

• Sometimes a “Delay” option makes sense

The web has scalability built in like nothing else

But there are still many missing features

There are options to add features to the system

Closed systems are not long term answers

Ideally we need multiple competing solutions

The web is getting better; auto-upgrade helps

Is the Open Web the Platform for Tomorrow?

Questions?

Who decides if it carries on winning?

What if there is only one effective browser vendor?

Do the failings of the Web mean we need more options?

And if so,what are theoptions?

Monopolies: Good / Bad?

Monopolies: Good / Bad?

Monopolies are a form of Absolute Power (and that’s not good)

A mono-culture is easier for the developers

A mono-culture is easier for the developers

(but the goal is to be better for the users)

Plug-ins: Good / Bad?

Plug-ins: Good / Bad?A plugin is just a ‘browser’ in a browser

The History of the Web

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

© Jo

e W

alke

r. 20

08. A

ll R

ight

s R

eser

ved

4 things the web needs to carry on winning ...

Features

Reliability

Trust

Momentum

Features

Signs of Disfunction


Zillions of toolkits are a sign that something is wrong.

If browsers were perfect we wouldn’t need Dojo or DWR or any of the other toolkits


What do the toolkits do for us?

• Abstraction

• Bug fixing

• Simplification

• Harmonization


How do toolkits harm us?

• Downloading

• Integration

Browser Features

Browser Developer Todo List:

• Make ajax toolkits irrelevant

Browser Features

The down side of browser evolution

• Do new features break old features?

• It’s slow

• Ubiquity is vital

Browser Features

How far can we evolve without vendor support?

• Comet

• 3D using VML / SVG / Canvas / etc

• Storage and offline

Reliability

Reliability

Local or Global?

• Depending on small scale reliability can hurt

• Large scale reliability requires anticipation of failure

• Markup defines the Open Web through reliability

Reliability

The web has some great features borrowed from Unix

• Small, independent units

• Linked through debuggable text based protocols

It adds features from living things

• The ability to cope with failure

• That can adapt when the world changes

Trust

Trust

What do you worry about when using the web?

• Web worms

• Privacy invasion

• Data in transit

• Identity Theft

• Browser Security

• How to build something trustable

Trust

What do users need to build trust

• Lack of spelling mistakes

• Reliable patterns

• Principle of least astonishment

• Time

Trust

What does the web need?

• Content Restrictions

• An anti CSRF marker for cookies

• Sandboxing

• Less brittle models of building relationships

• SMASH, XIP, etc.

Momentum

Momentum

What people know today defineshow they think about the future

Momentum

Web Standards:

• Not all have been transparent

• Working groups break down when the participants don’t play ball

• Adding IQ to a large group reduces the IQ of the group

• Working groups don’t test for fitness

What if we don’t get the tools we need?

ToolkitsFlashGears

Business

Is the Open Web the Platform for Tomorrow