The Two Towers

We are not ready. And it scares me. We’re meddling with major

forces, armed with some hopes and dreams, and we think it’s enough.

No one can deny the astonishing jump in attention that Linux and

OpenSource have made. It’s incredible, and all due to the hard work and diligence of

some very cool individuals. But I’m here to doomsay, though it’s not my nature,

because I think it’s vital that some important thoughts get thunk.

First, Microsoft is irrelevant. Totally. Ignore the man behind the

curtain. Gates is just a ring-wraith to the Sauron of unethical capitalism. We run the

real risk that by concentrating too hard on killing Windows, we’ll become just like

them. People are calling for "feature parity" with the Office (TM, Pat

Pend)

applications to ease the transition of newbie users, without once questioning whether

that’s a good thing.

Personally, I like the idea of giving all those potential new users

a nasty shock to the complacency. (Which is just above the navel, next to the spleen.) I

don’t want Linux to be "Just like Windows only Better!", I want it to

challenge the basic assumptions that Microsoft engineers have made. Windows is a poor copy

of the Macintosh is a poor copy of the Xerox Star, now 20 years old. We haven’t

challenged the basic WIMP (windows, icons, mouse, pointer) paradigm in 20 years? Come on!

And have you ever considered what might happen if the stated Linux

goal of "Total world domination" is actually achieved? And are you really

prepared to accept the consequences?

I don’t think you are.

Consider: Red Hat used to release a new version every few months.

That has now slipped. The reason is rather involved, but essentially the reseller channel

doesn’t like version releases too often. Even once a year is too frequent, in

reseller land. Two kernel updates a day wouldn’t be believed.

To get Linux onto everyone’s desktop, the release frequency

will have to drop down to that kind of time scale. Even if we perfect the Debian

auto-update, people won’t use it, even if we get the bandwidth. They don’t like

their software changing unpredictably from day to day. The hacker/early-adopter/mainstream

model acts like a filter to keep bad or incomplete software off computers. (And consider

carefully the possibility of a single buggy kernel module, sent out through the

auto-update chain, taking down the Net.)

First, accept that the commercialization of OpenSource products is

inevitable and necessary. Some people only attach value to things they have to pay for.

This is the businessman’s mindset. He believes in money, with a capital M. It defines

him, and his relationship with the world is interpreted in terms of the flow of cash. This

is the work of Adam Smith’s Invisible Hand, which has long been wedged up Adam

Smith’s Invisible Butt.

And as far as commercial Linux vendors go, there Can Be Only One.

Microsoft has a monopoly. they just happened to be the one ruthless

enough to get it for the moment. If Linux "wins", then it will become the

monopoly OS. And it will be a plodding Red Hat once-a-year release. Is that what you

wanted? Because sooner or later, Red Hat, the company, will be owned and run

by an economic rationalist with a legal obligation to increase shareholder value (as all

publicly traded companies are required to do, or they get sued) using any and every means

at their disposal. It happened to Netscape. Ask Jamie. Linux, in the wrong hands, is a big

gun, which someone will use to shoot Microsoft, probably in order to replace them. That is

simply what will happen. If Microsoft is clever, then they’ll buy that gun to shoot

themselves in the foot, rather than wait for someone else to aim it at their head.

That’s the consequence of Total World Domination. This is called "Commercial

Reality", and the only way to win is not to play the game.

Don’t let the fact that 90 percent of the population wants to

be spoon-fed their software, distress you. People are weird. Accept it. Just remember that

while a thriving OpenSource community exists, the other 10 percent can get what they need.

And if you care, then you’re in that group. And the way to live is to do it on your

own terms. Great works can be done outside the corporate environment. We’ve proved

that. We don’t need them anymore. Independence is the key.

Along the way, there are many traps. First, though copyright and

patent laws are flawed, it’s still the law. Be very, very sure that the code you put

into an OpenSource program infringes no patents or copyrights, even though you disagree

with the concepts. Don’t use LZW compression, don’t call your program

"Excel", and don’t share code between your day job and your project unless

you’ve got a piece of paper to counteract the other piece of paper they usually make

you sign. If it becomes a crucial part of the system and suddenly gets yanked because of a

legal dispute, that’s a very bad thing for the project, for whoever gets the blame,

for all the people down the track who may have used your code in something else, and for

the image of OpenSource. Maybe we need a "100 percent pure OpenSource" campaign

or something, I don’t know.

The second trap is commercial interests. Look at Silicon Valley.

It’s a case in point of what to avoid. The defining characteristic of the Executive

Suit is the search for power. They look for the biggest game in town and muscle in at the

top using the forceful application of money and the law. Can you say "Venture

Capital"? I knew you could. As soon as you turn a technology into a company, or a

patent, or something that can be owned, then it will be bought up by the kinds of bottom

feeders who like to own things, because they’re very, very good at it.

OpenSource has got the attention of the world, now. And that means a

million Suits are looking at Linux and asking "How can I use this". And they

don’t mean "How can I use this to make things cool and froody" but

"How can I use this to increase my grip on power and crush my enemies like the bugs

they are!" and then they laugh maniacally and stroke their fluffy white cats while

pushing evil buttons.

Well...maybe I’m being a little melodramatic, but you get the

idea. Effective executives are users who treat every thing and person as a

"resource". And some of them do have cats.

There’s a saving grace, if you choose to believe it: Linux is a

byproduct. A created artifact. The important thing is the community which created it. If

every copy of Linux (source and binaries) could be magically removed from everyone’s

hard drive, (possibly the result of some evil Microsoft virus delivered by nuke-detonation

EMP pulse and Tantric chant) we’d scream blue murder, and then re-write the whole

thing in two years. Some would even count it a blessing, like the Great Fire of London,

and how it made possible some proper town planning for once. So, let them "own" Linux, if we must. Let them gloat over

the binaries, like some rare and sparkly gem. Just don’t tell them that we can make

another anytime we wish. Because then they might be tempted to try to own the community,

and the process. And that would be awful.

But the worst trap of all would be for the OpenSource community to

deny what makes it so powerful: a global collection of individualistic, ornery, talented,

and opinionated people who make things happen, knit together by a new communications

medium that we built ourselves, which somehow keeps us moving in the same unified

direction.

We have dangerously re-focused from being a hacker collective to

taking aim at a product. The first can be described as an open-ended search for self. The

second is a closed contest with a single winner, and then Game Over. Then what?

Sun Tzu said "Know yourself and know your enemy, and you will

come through a thousand battles without harm". We know the enemy well enough...but.

Saruman looked too deep into the palantir. His wisdom gave way to

the need for power, and in grappling with the Enemy on his terms, he became like him. His

mind was warped by what he saw. Insengard was remade into a poor imitation of Minas

Morgul. The Two towers. One a sad copy of the other. Both failed in the end.

Addendum: Writing for an online forum can be interesting.

Within hours of the article being posted to www.slashdot.org,

upwards of 180 people had posted public (sometime lengthy, but usually thoughtful)

replies, and several sent me direct mail. One ex-Microsoft product manager mailed about

his experiences trying to push OpenSource while still there. One person asked for help

installing Linux. (And got it.) The majority of replies were positive, speaking of a

general feeling that yes, perhaps things had gone a little too far. Writing for an online forum can be interesting.

Within hours of the article being posted to www.slashdot.org,

upwards of 180 people had posted public (sometime lengthy, but usually thoughtful)

replies, and several sent me direct mail. One ex-Microsoft product manager mailed about

his experiences trying to push OpenSource while still there. One person asked for help

installing Linux. (And got it.) The majority of replies were positive, speaking of a

general feeling that yes, perhaps things had gone a little too far.

Another general thread was "the GPL will protect us".

I’m not so sure. It’s just a document. The real protection comes from the daily

efforts of all in the community. I hold a deep belief that things will work out for the

best when driven by people of good faith and good intentions.

Last, my mention of Red Hat in a possible worst-case scenario means

nothing. Red Hat have performed a sterling service for the community. They’re simply

the most visible example of exposure to the commercial world. I wish them the best of luck

in walking their fine line. Special kudos to the guys in the Red Hat labs.

Jeremy Lee is a programmer/analyst living in Australia. He is currently designing a new user interface which will totally revolutionize computing as we know it. He does not have a cat. This article was originally posted at www.slashdot.org. Reprinted with the author’s permission