The web spread out so fast because HTML was so lame even a monkey could understand it. That simplicity created a low barrier to entry. Monkey see, monkey do learning was the key. A million monkeys at a million keyboards and a few years later … boom!
Even before the war buffs could turn around in their lawn chairs the long running standards war over document formats was transformed. It took Microsoft years to displace Word Perfect from the rich farm lands. It’s took Adobe years carve out an encampment on the high ground above Microsoft Word. Suddenly HTML was in control of the seas.
Bill Gates onces said of Netscape that they owned all the river front property. But, the HTTP and HTML were the water and owning those was harder. Not that they both didn’t try to own them.
Spreading fast creates very fragmented markets. The entire middleware industry is a side effect of this. Building a custom niche oriented authoring tool, say a bridge from your real time control system into the web, was the work of a weekend. Happy monkey.
Diffuse big markets call out to be condensed. They call out to capitalists to be owned. They call out to engineers to be made safe and efficent. They call out to the monkeys to come join in the fun, and those monkeys start demanding regulations, police, schools, etc.
One road to condensation is to roll up the big monkey firms, the ones that have found important bits of real estate in the new jungle. This is, for example, what Adobe is doing in buying Macromedia. Adobe likes the high ground, the sophisticated elegant page layouts and such, and Macromedia is it’s closest competitor for that ground. Flash is more widely deployed than the Adobe reader; end of story. Buying up the big monkeys isn’t an effective tactic if the new ones appear faster than you can buy them.
Another way to condense a difuse market is to raise barriers to entry. But, how do you raise the barriers to entry around HTTP and HTML? Easy. Make them more complex.
For HTTP the answer is SOAP, WS, et. al. You create a specification that is sufficently complex it scares off little monkeys. Frustrate their desire to whip off an implementation over a week end. It’s not clear this tactic is working; the value add of these stacks just isn’t as overwhelming at it needs to be.
The idea is to make raise the barriers so the monkeys will wander off and find something else to play with. View source used to be a lot more fun for a lot more monkeys than it is today.
This is a story about standards. Simple low barrier to entry standards can spread like wild fire over an ecology. After the burn there begins a progression toward increasing complexity. Some players in the forest society will work to increase the complexity. Their motives vary. Some do it to for fun or to address real needs. Sometimes, particularly inside proffesional standards bodies, the complexity rises because the process tends to compromise to the sum of all features.
For some players the rising complexity is a conscious attempt to prepare the forest for something more suitable for their big agriculture of monoculture crops.