Whee! AOL Leaps Into the Whirling Saw Blades

You know that cliché about horror movies where the ingénue ends up off by herself in creepy surroundings, and the audience is saying "What are you thinking? Don't go in there!" ?

Well, I'd like to start a MST3K takeoff where the "film" is a tech show off of G4TV or a clip reel from the vlogosphere, and we have, say, a VC, a computer scientist, Ted(!), and a couple o' engineers who have done their time at Initech in the audience.

They could all yell "sure, why not go in there... looks like fun" when someone (like AOL this week) does something colossally foolish, like try and build yet another abstraction platform so everyone can write little apps that run on every cellphone, everywhere, like magic.

AOL says:

"The new open platform will help stimulate innovation by providing developers with ready access to the tools and source code they need to build and distribute applications across all major mobile device platforms and operating systems including BREW, Java, Linux, RIM, Symbian, and Windows Mobile."

Anyone who's ever done anything with cross-platform mobile knows that this is a really hard problem, and not the software kind (where solid R&D might yield results), but the business/social kind, where even the resources of megacorp like AOL are a drop in the ocean.

That's ok, because they have a plan for exactly how it's gonna work. If you're playing along at home, stop reading. Close your eyes and guess the innovative architecture for pulling this rabbit out of the hat. Ok, now you can look:

"The platform will consist of three components:

• an XML-based, next-generation markup language;
• an ultra-lightweight mobile device client;
• and an application server."

I think the secret sauce must be the next generation markup language. See, the next generation markup language will become sentient, take what you write, process it into extraordinarily cogent Powerpoint, email it to all the VPs of all the mobile carriers, device manufacturers, and mobile OS vendors in the world, and then wait on a queue until they all agree to change their businesses so your stuff will work.

I've written lots of times before about the mobile app ecosystem and why, for now anyway, this kind of idea is just a complete non-starter. At least startups fail fast, since they have vaguely limited resources. Then you have players like Yahoo! that find themselves in a deep hole, and say "hmm... this is so big and so deep, we can roll an excavator down here and start the real digging!" They do things like deploy an implementation of J2ME in .net bytecode, so that people can run a mediocre (I'm being polite here, guys) MIDP app on their brand new Blackjack or HTC Touch.

In my new MST3K show, I'll keep the video clips short, so we can watch, laugh, and get on with our lives.