White Lights Lead To Red Lights

Mark Pilgrim

December 7, 2005

Where the hell have I been?

Writing books

[Ethan with Dive Into Python]

And having kids

[Wesley with Greasemonkey Hacks]

How to write

No shortage of advice
“How to write a novel in 30 days”
“How to write a resume”
“How to write a better weblog”

What to write

No shortage of advice
“10 great blog post ideas!”
“50 great short story plots!”
“100 books waiting to be written!”

When to write

How to become an expert

Write code
Make mistakes
Get yelled at
Fix your code
Write about what you learned in step 4

Why aren't there more experts?

Most people give up before step 4
The rest give up before step 5

Dive Into Python

Little-known fact:

I learned Python in August, 2000
It filled me with wonder
I started writing Dive Into Python in September, 2000
The first version was riddled with errors
...which people kindly (and not so kindly) pointed out
...and I promptly fixed

Dive Into Python, 3 years later

September 2003: Apress
It was much harder than I expected
diveintopython.org was wildly popular
Professional development
Became an expert
Lost the sense of wonder

Greasemonkey

Little-known fact:

I'd never written a line of JavaScript until this spring
Then I found Greasemonkey
It filled me with wonder

Dive Into Greasemonkey

First Greasemonkey script in March, 2005
Dive Into Greasemonkey in April
Greasemonkey Hacks in May
First security hole in July
- ...and learned all about XPCNativeWrappers
I may never write a line of JavaScript again

When to write

Write when you’re filled with wonder
If you wait until you’re an expert, it’s too late
Write when you can’t not write

When to stop writing

Just before you lose your sense of wonder
“Quit while you’re ahead”
Burnout is what happens to experts who don’t know when to quit
Rebirth

The future of the web

What’s going on?

Greasemonkey
Atom
Microformats
WHATWG
- ...a.k.a. “HTML 5”

Greasemonkey

[Wesley with Greasemonkey Hacks]

Greasemonkey

A reaction to an exceptionally poor redesign
Scripts are content-focused
Unofficial web services
“My DOM is my passport; verify me”
Brittle, backwards, broken
- ...and brilliant

Atom

A reaction to the ever-shifting RSS spec
RFC 4287
RSS without the sociopaths
Real content model
- ...primarily for transporting HTML

The Zen of Microformats

A reaction to the “Semantic Web” (whatever that is)
Invisible data sucks
Humans first, machines second
Humans browse the web
- ...the HTML web

What's wrong with RDF?

RDF has a great data model
Rich semantics
Infinitely extensible
Optimized for machines
Machines first, humans second

HTML

HTML has a crappy data model
Poor semantics
No apparent extensibility
And a long, sordid history of abuse

Hacking HTML

You know those weird-looking elements in the appendix of your HTML book?
The ones that aren’t <TABLE> or <EMBED>
<ADDRESS>, <CITE>, <DFN>, <DL>, <Q>, ...
Some crazy dreamers got together and decided to resurrect them
They call it the “small-s semantic web”

Hacking HTML extensibility

Despite first appearances, HTML is extensible
<head profile="...">
New link relationships with rel
New semantics with class
class is a space-separated list of keywords → infinite extensibility
Cry if you want, but it works
- ...and it validates

Microformats so far

rel=nofollow
rel=tag
rel=license
XFN
hCard

WHATWG

A reaction to XHTML 2
HTML forms suck
Nobody understands XForms
Why can’t we just have better forms?
And while we’re at it, let’s document all the crap that browsers actually support
- ...like XMLHttpRequest
- ...and document.write

HTML 5

I look at XHTML 2 and all I see is politics
I look at HTML 5 and all I see is passion

Common themes

Pushback against bad ideas run amok
Greasemonkey: Adrian Holovaty, Aaron Boodman (allmusic.com)
Atom: Sam Ruby, Joe Gregorio (RSS, SOAP)
Microformats: Tantek Çelik (RDF)
WHATWG: Ian Hickson (XHTML 2)
...and many others

Common themes

Expertise
Burn out
At the right moment, a rebirth
New-found passion
A sense of wonder

The future of the web

...is HTML

References