I’m not sure what Stephen is driving at with the competition argument, actually. Success has little to do with avoiding competition altogether (it simply can’t be done, for one thing — anything you do that proves successful is going to inspire competition, no matter what the underlying platform might be).

All platforms disappear in time. I’ve been working for CTER since 2002. Since then nearly every instructional technology we use has been replaced, in some cases multiple times (sole exception: email).

The half-life of technology is very short. If you want to build something for the ages, you’d going to have to go into a different line of work.

Facebook
Flickr
Twitter
iPhone
SecondLife
Android

Although I think your breakdown of apps is broadly a good one, Stephen doesn’t do anything to make this distinction as you have – that is he doesn’t define what he means by an ‘application’.

I’d also still say it is about cost/benefit and audience. To break two of the commandments above, if I build an iPhone app to browse Flickr collections, I accept that I’m at risk of the things that Stephen says (platform decline, being outcompeted) – but if the the cost is low and benefit high then it seems rational to build it (especially if the benefits can be realised in a short timescale)

To not build a platform-specific app, I think it’s necessary to not think of apps as atomic units, but to separate out (1) the actions/activites performed by the app, (2) the information/data that the app uses or creates, and (3) where the human interaction points of the app are. 1 and 2 must be platform independent to ensure survival, and open source is usually an expected component to achieving this. 3 can be (must be?) platform-specific, to fit the user interaction models and unique capabilities of each platform.

When 1 and 2 are independent, a failed platform demolishing 3 is not a major concern. 1 and 2’s independence also leads to the triviality of building 3. The trouble with building 1 and 2 away from 3 is that framework mentality tends to creep in. But, Stephen also warns against that two points down.

If it is trivial to build (e.g.) a Facebook app, then I think if yoiu have an audience who will appreciate it.

I wouldn’t say don’t build for iPhone as Stephen does, although I would say the cost is likely to be on the higher end, so you’d want to know it was worth it.

I’d go a bit further and say ‘use technology that makes building apps easy’ so that not only is the cost low for you, but also your community can look at building apps on their desired platforms – be that MySpace, Facebook, iPhone, Wordpress Widgets etc. etc.

Many people I know and respect, like Stephen, are adamantly against closed platforms, to me at almost a religious level. I am less sure (and I think there is also a level of difference between apps and platforms…) – I would say, “don’t build proprietary BAD apps” – I have yet to find a need, a restriction in my use of the iPhone, or “closed” Apple products that prevented me from doing something or making me need to hack it– ahem, maybe not– I don;t like the lock in to AT&T.

Good point too on Gartner not being specific to a tech like the iPhone.

I am enjoying being corrected, believe me; keep ‘em coming.

The issue I see here is that the iPhone is a proprietary platform – and Stephen advice is “don’t build platform specific apps”. Stephen is clear that alternative interfaces are an interesting area, as is portability.

In terms of the Gartner hype cycle, I don’t think the ‘iPhone’ as such would be something they would plot (I may be wrong). However, if we were to look at ‘mobile internet’ I think we can say that this has now passed through the trough of disillusionment, and is now moving along the slope of enlightenment to the plateau of productivity. That the iPhone has contributed to this there is no doubt in my mind – whether the iPhone as a product lasts or not is not really relevant though.

It does seem, that the word “fad” has a bit of a condescending negative connotation moreso then your dictionary definition.

I think what defines something as a fad is how quickly it moves from “technology trigger” to “peak of inflated expectations” based on Gartner’s hype cycle. Move quickly and move high: fad. What happens after the peak, determines its value and game changing power. If it falls low and falls quickly, then there probably wasn’t any actual substance to the fad. If it doesn’t fall much and moves quickly through the remaining phases, it’s a game changer.

For me, iPhone, iPod touch, and virtual worlds are all fads. I think iPhone/iPod touch have passed their peaks and become game changers. Virtual worlds has also passed the peak, but slid down and isn’t coming back up.