I'm weighing in on

http://www.randomhacks.net/stories/the-missing-future.html

It seems to me that the strength of the ideas comes from the fact that the article is well-written. This is mixed praise. If you pinpoint your article between two poles, and position yourself and your reader with accuracy, it will lend an authority to your interpretations that a bad writer might not be able to deliver. Of course, a bad writer might have a better grasp on the separation of facts and opinion.

The answer is clear: folks are hungry to turn the next page in technology, and they're tired of turning to authorities to make it happen. Each of us has to turn the page ourselves: fine, but the tricky part is just finding the corner and getting a grip. Will it be Longhorn? Mozilla? P2P? .NET? Killer apps sit on top of platforms. There are lots of platforms out there and emerging, it seems like fertile territory. Find a underutilized platform, and anything you write for it may turn out to be a killer app. This is a common strategy for coming up with something new. You feel for the corner, grab it and turn, but the next page is blank.

The other strategy is to come up with the idea, the "killer app" first, and then moosh it onto the platform that supports *most* of it. Trouble is, this is not historically how these things happen. E-mail was invented along with its own protocol at the same time. Same with HTML and browsers. So with this strategy you get "Idea in Space" with no execution plan or supporting technology unless you write it yourself, or embark on an extensive mooshing process. Nevertheless, "Idea in Space" seems like a brave and happy place to those on the other side, swimming in the capabilities of their luscious platforms and staring at a blank page in terms of apps.

The thing I want to comment on is the killer app mentality. It's like baiting your hook for the big marlin and missing all the salmon that are out there. Often we don't know a killer app when we see one. These things best happen organically, when the environment is set up to best support autocratic nonconformity by individual developers and users. For example, you could make the statement that you want to find the "killer app for longhorn." This is like saying you want to find a killer app for your house. A house does not have a killer app. It is an environment, you live there. Killer apps for living there happen to be eating and sleeping, but you can do those anywhere to a variety of success. Really you want a house as a great environment to make things happen. Same with a platform.

This posting begs the question: how do we go about developing successful V1 software concepts. Is there a methodology that works, that has worked in the past, and will take us from concept to killer app via the straightest path? I'd like to see others weigh in, particularly software methodology specialists.

comment []9:40:50 AM