Beyond Nerds Bearing Gifts - 
The Future of the Open Source Economy
-------------------------------------

Thank you all for having me here, 
this is my first time in Australia, 
and it has been an amazing experience so far. 
I was originally expecting more of a culture shock, 
coming from the northern hemisphere, 
and I even prepared <X> some custom Australian slides, 
but it turns out I was misled, 
the equipment is actually compatible.

Anyhow, I have a problem 
which I'd like to share with you, 
and it's not an uncommon problem. 
It's probably one many of you share. 
My problem is, <X>
I have a mother in law. 

And when we sit down to Christmas dinner,
which is coming up, 
she has this question she asks, 
and it is a perfectly reasonable question 
for a mother-in-law to ask a son-in-law. 
She asks me, <X>
what it is that I "do". 
What do I "do"?

The subtext being, 
am I doing something that can rationally be expected <X>
to help support her perfect daughter 
and two darling grandchildren? 

And this is where things get difficult. 
Because what I "do" is, 
<X> I sit in my office 
at the back of the house, 
and I work on software 
that people are encouraged 
to download and deploy and use 
for free. 
For free. 

She understands the programming part of it well enough, 
at least insofar as I sit in front of my computer 
and I type things on it, 
but the "for free" part causes serious discomfort. 
Because how do the perfect daughter
 and the two darling grandchildren get supported "for free". 

Why am I confident 
that I can earn a living in my chosen field, 
when my work product 
is free? 
Am I <X> crazy, perhaps? 
Or some kind <X> of utopian hippy?
Maybe not good son-in-law material.

The proximate answer -- 
and on a good night this will cool her off -- 
is that I'm employed by 
a company, <X> OpenGeo, that 
pays me to work on this 
software that we give away for free. 
Voila! 
But it doesn't take much to see past that dodge, 
and by her second glass of wine she'll come around again, 
"so, <X> who pays OpenGeo for this free software?"

Who indeed? 
In the competitive marketplace, 
<X> full of beasts like Oracle, Apple and Microsoft 
(that picture of Steve Balmer freaks me out) 
how, 
<X> how can something 
as warm 
and fluffy 
as an open source company 
survive?
What is it, 
exactly, <X> that we are selling?

If you want to understand 
what open source companies are selling, 
it helps to understand what the 
existing proprietary vendors are selling, 
and -- 
here's the surprising part --
<X> they aren't selling software.

<X> They are selling *products*. 

Let me explain what I mean by that.

<X> In the business classic "Crossing the Chasm"
(which I highly recommend), 
Geoffrey Moore says that in the
technology adoption life cycle, <X>
(which is traditionally understood 
as a smooth passage 
from the small early market of
visionaries and early adopters 
to the large mainstream market
of pragmatists and conservatives), 
there is 
a little understood gap, 
in fact a huge *chasm*, 
<X> between the small early markets,
and the big mainstream markets

And this chasm is there
because the personalities
of customers in the 
early market
are very different
from the personalities
of customer
in the mainstream market. 

Early adopters and visionaries 
have a high tolerance for risk
They like to learn things themselves, 
and don't need a lot of support. <X>
Here's an early adopter 
with the iPhone he bought on the 
first. 
day. 
they were available. 

The <X> mainstream customers are exactly the opposite. 

In order to prosper, growing software companies <X>
must cross the market chasm, 
to gain access to the 
big mainstream markets,
and to do so, 
Moore says they must transition 
from just selling software 
into selling a what he calls <X> 
a "whole product".

Now, 
a young naive technology company might say 
"but we have a product! <X> it's on this CDROM right here!" 
But they don't have a product, 
they have <X> 
*software*. 
What they have is 
salable in the 
early markets, 
but *not* to the 
majority markets.

The "whole product" has software at its core, 
but it adds in a critical layer 
of extra services and infrastructure around the outside, 
Things that reduce the risk 
(or perceived risk)
associated with adopting the product. 
<X> Training courses, 
support infrastructure, 
re-sellers and consulting networks, 
update mechanisms, and so on. 

And it is the combination of the software 
with the added layer of valuable extras
that make a <X> compelling "whole product".  

Only a "whole product" can move 
from the adventurous early adopter market 
into the risk-averse mainstream market.

For example, 
<X> as a piece of database *software*, 
Oracle is not all that compelling. 
It's kind of bulky, it's very hard to learn, 
and it's pretty easy to screw up. 

But, add in 
<X> a 300lb shelf of professionally written documentation, 
training to build a population of developers and administrators, 
<X> a rich ecosystem of third party tools, 
<X> a reputable (if mercenary) company providing support, 
<X> an evil genius, 
and <X> deep integration with the 
other elements of the Oracle software portfolio, 
and you have <X> a very compelling *whole product*. 

Here's an odd thing.
Even though it is easy to see 
that the "whole product" offers 
tremendous value beyond the software itself, 
our <X> mental model of technology acquisition 
is still one where we pay money for the software,
and the vendor throws in 
the rest of the whole product 
for free.

All the great extra value, for free. 
And this is where our cognitive dissonance 
about open source software companies comes from.
Because what happens 
to our mental model 
when the cost of the software <X> 
goes to zero? 
No company 
can give away the software for free, 
*and* also provide the rest of the 
whole product for free! 
Something has to give!

For the early adopters and 
technology visionaries,
the people in the early market,
(and lets be honest, 
this audience is full of 
you people) 
there's no problem -- <X>
they just use the software as is, 
and take advantage of the thinner layer of 
support provided for free 
by the open source community. 
They don't need the whole product, 
and they probably never will.

But what <X> about that 
huge early majority and late majority? 
What will it take to get 
them to adopt open source software? 
The low price of the software alone is 
not sufficient to seal the deal, 
because the 
rest of the whole product 
is missing. 

<X> The long term open source business model, 
as a general proposition, is about 
providing a whole product
suitable for mainstream customers,
<X> but changing 
the point 
of monetization. 

Instead of companies selling software and 
cross-subsidizing access to a 
free network of services for customers, 
we will have companies selling 
access to a network of services and c
ross-subsidizing the development of free software.

This is precisely the Red Hat Linux model.

<X> Say you've got some open source software, 
being developed on the internet. 
You've got Red Hat.
And a risk-averse mainstream customer.
He gives <X> Red Hat money for Red Hat Enterprise Linux.
Red Hat uses that money to build a whole product 
and to fund open source development. 
They take a copy of the raw software, 
wrap it in value added services, 
and give it to the customer as a whole product.

The customer could simply 
<X> download the software directly, 
but then he wouldn't get support, 
automatic upgrades, testing, and so on. 

<X> And for a mainstream customer, 
that feels like a risky situation to be in.

That's how an open source company is 
supposed to work 
in an ideal situation.

However, 
it's not like the old proprietary model is fading away. 
<X> Microsoft is a $223B company and 
<X> Oracle is a 103B company and 
<X> SAP is a $58B company.  
The *biggest* open source vendor, 
Red Hat, <X> is only a $5B company. 

So, shouldn't I be worried? 
Polishing my resume? <X>
Both open source 
*and* open source companies 
sound like helpless fluffy bunnies. 
Cuddly and fun to play with, 
but way overmatched in the marketplace. 

Why am I so confident?

I'm confident because there is <X>
more to this story than 
just the market for shrink-wrapped software. <X>
And there is a lot more 
to the fluffy bunny than meets the eye. 
The fluffy bunny is busy 
transforming the field 
of information technology 
in profound ways, 
leaving carnage in its wake. 

<X> For example,
the Economist magazine, 
arbiter of free market orthodoxy, 
has already taken in the situation and 
declared open source a serious player: 

<X> "Open-source software has won the argument."

<X> "It is now generally accepted 
that the future will involve 
a blend of both proprietary 
and open-source software."

How can this be? 
The leading open source company 
has a market capitalization less than 
3% of the leading proprietary company. 

How can open source "win the argument", 
when it is so manifestly overmatched? 

Here's how.

<X> First, understand that
you can't comprehend 
the success of open source 
exclusively by looking at the 
marketplace. 

<X> In the marketplace 
the unit of competition is a company, 
and the measure of success is profit. 
<X> The more dollars you take in, 
the more successful you are, 
and if you take in too few dollars <X> 
you go extinct.

Lots of people have noticed that
there are parallels between the 
market and the 
field of evolutionary biology.
Taken to an extreme, you get 
theories like social Darwinism,
but for for our case the metaphor
is instructive.
<X>

Only the strong survive.
<X> Those companies (organisms) 
that best <X> adapt to their markets (environments) 
<X> take in the most money (food / energy) 
and those that do not 
<X> go extinct (they end up bankrupt).

Also as with evolutionary biology, 
it's <X> easy to be distracted 
by the big lumbering beasts that 
*appear* 
to be directly engaged in competition. 
And that's a bad thing! 
<X> Because 
the really *important* competition is 
not at the level of the organism, 
but at a lower level, much farther down, 
in the realm 
of the gene.

The competition between genes is 
described by Richard Dawkins 
in his book "The Selfish Gene", 
<X> an exploration of how simple selfish
reproductive behaviour at the level of genes can lead
to apparently altruistic behaviours
at the level of the organism.

In Dawkins formulation, 
behaviours that maximize the chances of 
*genetic* survival are 
passed to future generations, 
even when those behaviours 
endanger the survival of 
a particular organism.

Here's <X> a behaviour that makes 
no sense in a organism-centric model. 

This is a Killdeer on Vancouver Island, where I live, 
engaging in a 'distraction display'. 
The parent bird puts on an elaborate 
(and totally fake) 
display of being injured, 
to draw an approaching predator 
(or, in this case, videographer) 
away from its young.  

The parent organism is placing itself at great risk. 
Why? 
Because the strategy is a good way to 
preserve the genetic heritage of the 
*children in the nest*.

<X> "The Selfish Gene" came out in 1976, 
and in the opening pages, 
Dawkins has this to say about 
some of his contemporaries, 
(and you can see right away where he gets his 
reputation as a gentle and self-effacing man),

<X> "The trouble with these [other] books 
is that their authors got it 
totally and utterly wrong. 
<X> They got it wrong because they
 misunderstood how evolution works. 
<X> They made the erroneous assumption 
that the important thing in evolution 
is the good of the species (or the group) 
rather than the good 
of the individual (or the gene)"

The same criticism 
applies to people 
who attempt to understand 
the success of open source 
through an analysis of the marketplace. 

They misunderstand how software lives and dies, 
confusing the host with what it carries. 
<X> Because the unit of competition 
in the world of software is 
NOT the corporation, 
<X> it is the PROGRAM, 
it is software,
it is code.

<X> In the biosphere, 
<X> organisms feed on other organisms, 
which feed on plants, 
which feed on light from Mr. Sun. 
So, in the end the competition is for 
sources of *energy*, 
either direct (in the form of sunlight) 
or stored (in the bodies of plants, and other animals).

<X> In the cybersphere, 
<X> programs also compete for resources.
The resource that programs compete for 
is developer time -- 
<X> commonly known as human attention. 
So programs "feed" on developers,
<X> who in turn feed on caffeinated beverages 
and try to stay away from Mr. Sun.

Programs need 
programmer attention 
 <X> to survive.

A program that is no longer 
being maintained and updated 
is a program that is dying. 

<X> A program that no one has a use for, is dead. 
First it will be abandoned, un-run, 
then it will become un-runnable, 
and then it will be deleted.

Programs need programmer attention to survive.

Don't believe me? 
<X> When Oracle gobbles up 
yet another enterprise software company, 
do the customers bemoan the death of the *company*? <X>
No, primarily they worry about the *software*: 
Will bugs be fixed? 
Will we get that next release with the new features? 
Will the developers flee to greener pastures?
<X> *Will the new owner continue to feed the software?*

And what does the software worry about? 
All that matters to the software is that it 
continues to receive a steady supply of developer time.  

Understanding the competition between 
proprietary and open source 
as a competition at the 
program level 
clears away a lot of distractions. 
Now, we can directly evaluate 
which strategy is the most 
adaptable strategy for survival: 
the open source model; 
or the proprietary model?

A proprietary program 
can best be understood <X>
as a form of parasite. 
It resides in symbiosis <X> with a host organism, 
the corporation that owns it, 
and draws its sustenance 
exclusively from the 
developers provided by the corporation. 
The amount of sustenance provided to the 
program pretty directly correlates 
to the success of the corporation selling the program 
(though sales success may or may not 
correlate with the actual quality of the program). 
When the corporation dies, 
the program usually dies too. 
If the corporation is subsumed by another corporation, 
<X> the new host may continue to feed the program, 
starve it to death, 
or terminate it immediately 
in favour of some other program.

When you think about it that way, 
it is easy to feel a little sorry <X>
for a proprietary program. 
It is very much at the mercy of its host. 
Its success or failure may have 
nothing to do with its intrinsic quality. 
It may have <X> only a small team of developers to love it, 
feed it, and carry its memory forward if it should die.

In contrast to the sheltered, 
monastic existence of proprietary software, 
<X> the lifestyle of a successful
open source program is 
incredibly promiscuous. 
Any developer with a nice smile and 
a good patch is welcome to join the party. <X> 
Open source programs can draw sustenance 
in the form of long term, 
stable commitments from corporations 
who sell services or products around the software, 
<X> from devoted contractors 
who derive income from contracts 
for features development or bug fixes, 
or from <X> quick relationships with casual developers 
who just drop off a patch and run away. 

In contrast, proprietary programs are 
embedded within the <X> institutional framework 
of a corporation, 
so it is much harder for them
to form relationships with 
new sources of development -- 
people can't just stroll in the door a
and add a new feature to Microsoft Word. 
<X> The relationship between 
developer and proprietary software 
is formal, 
contractual and 
exclusive.

Open source programs 
can form relationships with 
<X> multiple developers and 
multiple organizations simultaneously, 
because open source is 
not trapped inside a single organization. 
The rules of participation are 
cultural, not contractual, 
and broad community participation is 
the WHOLE POINT.

Take the most successful open source example, <X> Linux. 
The Linux software has gone 
from drawing development effort from 
a single Finnish graduate student, 
<X> to receiving the attentions of 
hundreds of fully funded developers
in multiple fortune 500 corporations, 
government agencies and 
academic institutions. 

Even organizations that are in 
direct market competition -- 
IBM and Oracle, or
Red Hat and Novell -- 
provide code for Linux, 
as do thousands of other developers with 
institutional affiliations ranging from
top secret government agencies 
to academia, 
to individual developers 
whose only real affiliation <X> is to 
their cat and 
their coffee vendor.

In fact, 
kernel developer Greg Kroah-Hartman 
did a study of the kernel source code 
in 2008 and found that 
the number one developer affiliation <X>
was "unaffiliated", 
accounting for 17% of kernel. 
Red Hat was #2 at 11%.

<X> The example of Linux 
shows that open source programs 
<X> are not limited to feeding off of 
pure open source companies. 
They can feed off of 
<X> any company that 
derives competitive advantage 
from using open source.

One of the most perceptive 
observers and commentators 
on the entrance of open source ideas 
into the marketplace <X> 
is Matt Asay. 
He recently wrote 
<X> "We are all open-source companies now. 
Which also means that none of us are." 

What he means 
is that every company in the marketplace 
is now deriving competitive advantage 
from open source in one way or another, 
even deeply, 
deeply proprietary companies.  

<X> IBM was once so proprietary that 
Microsoft looked open by comparison. 
But in 2000, <X> IBM was the 
first major company to adopt a Linux strategy. 
They invest directly 
in Linux kernel development 
to ensure it runs on <X> their CPUs and systems. 
They are a founder 
<X> of the Eclipse Java framework project, 
and a number of their proprietary products, 
like Rational <X>, are built on top of Eclipse libraries. 

<X> Oracle 
has purchased 
several open source companies, 
over the last few years, 
<X> database companies InnoBase <X> and Sleepycat, 
and this year purchased Sun, <X>
which netted them some 
very <X> well-known open source names.
And they aren't just sitting on them.
At Oracle OpenWorld last week, 
Oracle's evil genius <X> promised to 
invest even more money in MySQL R&D than
Sun is currently spending.

Even <X> Microsoft, 
which practically invented the idea of 
proprietary shrink-wrap software back in the 70s, 
now has an active open source strategy. 
They have an open source 
code hosting repository, <X> CodePlex. 
They are a sponsor of the <X> Apache Foundation. 
They invest in the development of 
Windows-compatible open source, 
<X> like IronPython and now <X> even PHP. 
They have even <X> contributed patches to the 
Linux kernel 
under the GNU GPL. 

In our own industry too, 
the momentum is towards 
more and more <X> open source use. 
ESRI uses the GDAL raster library in ArcExplorer. 
So does Google Earth. 
<X> PostGIS is becoming an 
industry standard spatial database, 
supported even by old guard companies 
like ESRI and MapInfo. 

When even 
the proprietary companies are 
investing in open source, 
what does it mean to be an 
"open source company"? 
Everyone is doing it! 

People like to talk 
about the change from 
proprietary <X> to open source as an 
"open source revolution".
But revolutions are quick, turbulent affairs <X>

Is it a revolution if it takes 25 years?

I think that what we are experiencing 
is not an "open source revolution", 
<X> it's an "open source evolution". 

The progress is slow and incremental, but 
the movement 
is always in the same direction, 
month by month, and
year by year. 

<X> We are just at the start 
of a transformation in the software market, 
<X> where purchasers recognize 
that they have the option to buy the whole product 
and get the software for free.

And we are in the middle 
of a transformation in how we build software, 
moving very quickly <X> 
from a closed corporate model, 
 where source code is private; 
<X> to an open collaborative model, 
 where source code is a commons.

And it is the 
combination of those two trends 
that 
fills me with confidence. 
Because the two trends 
are re-enforcing each other. 

And that's why 
I can look my mother-in-law in the eye and say 
"don't worry, it'll all work out". 

<X> I'm on the side of history, 
on the ground floor or a growing market, 
riding a wave that is just picking up. 

And so are all of you. 
Make the most of it! <X>

Thank you.