When my wife points out all the problem spots in our yard, her instructions will amount to "let a hundred flowers bloom" -but with due deference to that particular mass murderer's speech writers, there will be only one school of thought permitted: hers.
And isn't that exactly what happens when big players like IBM and Microsoft intervene in the open source process to force the market to standardize on one type of hardware or one piece of software in a segment?
Visualize Freshmeat and Sourceforge as display gardens for open source and you'll see well over a hundred thousand points of light struggling to be stars. That's amazing, and what's most amazing of all is that well over a quarter of these are genuinely active, and several thousand have a real chance of getting the kind of user support needed to generate new ways of thinking about the problems they address.
Two percent may not sound like much of a survival rate, but it compares to the success rate of internally generated software ideas at companies like IBM or Microsoft the way the Pentium IV does to the i80386.
Part of what's going on is that big organizations tend to resist change. New ideas, management books to the contrary, are taboo: it's marginal change, not creativity, that generates promotions and budget increases for middle management -- and top management generally knows that only acquisitions and bankruptcies can disrupt the middle's hold on bigger organizations.
Mao got survivors down to one school of thought at the point of a gun, Microsoft does it with a check book, but one of the great questions facing companies like IBM and Red Hat as they try to profit from open source innovation is how to keep acceptance of one thing, like Linux, from becoming the Medusa -freezing middle management in place to prevent adoption of the next big thing, whatever that will turn out to be.
It saddens me to know that the smoke bush I'll be planting while you're reading this will die, just as its predecessor did, but it's not a battle worth fighting -- and my attitude is what happens to a lot of corporate IT projects too. We do what we think the boss wants, not what we think might make sense for the project or the company. No one really knows what the numbers are, but my guess is that something like nine out of ten really good software ideas are discarded before ever being presented to management -- real victims of perceived organizational barriers to innovation that may not, themselves, actually be real.
Thus one of the things that make Sourceforge and Freshmeat so great is that they take these barriers away, establishing a free market in ideas by exposing them to a very diverse audience. If your idea resonates with anyone, you've found your market and success then depends mainly on your ability and willingness to work -not on someone else's ability to stick handle your project through the hoops in some corporate new product development process.
The result is an open source project maturation process that is far simpler, faster, and more effective than the typical corporate process and therefore a large part of the reason the success rate for open source projects is so much higher than for proprietary software.
It's that success rate which appeals to big players, because it offers them the ability to pick up new ideas that have already been linked to proven markets. Unfortunately when companies like IBM pick winners, they distort the market by implicitly applying their corporate agenda to directly affect the viability of the project's contributors, competitors, spin-offs, and successors. In effect, when they point the finger at a winner, their actions shade other projects, reducing overall success rates and threatening the very process they seek to benefit from.
Personally I'm going out today to plant perennials where they won't grow, but I know, and companies like IBM know, that this can't continue. What nobody knows is what will replace it -- but I'm pretty sure IBM doesn't have a secret store of Russian Sage.