Why AJAX doesn’t replace apps on the iPhone
Ok, so I spend a little time thinking about the iPhone, it’s almost launch time, isn’t it natural? I am firmly in the camp that the AJAX as SDK announcement was a load of horse nuggets that Steve knows and Steve knows we know, but Steve also knows we got nothing to do but take it. I was thinking about apps and thinking about Web2.0 and you know what I decided? It just doesn’t replace a native app. There’s all sorts of technical reasons why an online app is a different beast than a native one, but that’s been discussed. I want to talk about something I haven’t seen much talk of.
Let’s first assume a few things, let’s assume that you can make an app icon that links directly to your online app. Let’s further assume that when launched from this app icon, safari suppresses it’s normal window decorations so your online app can look just like a native one. Let’s further assume that 2.5g gives you good connectivity everywhere so you almost never want for a connection and the speed is fine. But these are all user facing hacks - these are there to make a web app look more like a regular app to the end user. The problem comes up if an app has any storage requirements it turns the developer from a product provider to a service provider in a very real way.
If the app has to store your information on it, then there’s a database with information that you probably don’t want to lose. The developer now has to go about ensuring that this data is kept around, that his service doesn’t go down, and that if something goes wrong he fixes it. If it does go wrong (as it inevitably will), he’s going to get all sorts of complaints (especially if anyone is paying money). This is a very different proposition and level of responsibility than testing an app, packaging it up and selling it. There are all sorts of other problems an online application can have that are, more or less, outside the developer’s hands.
That’s the bad news - it holds true even if there isn’t any storage necessary - it’s always a service that can go down if the ISP has a bad day, if too many people try to use it at once, etc… This is just made worse by having also to worry about managing people’s data, and I am not even going into privacy issues. Having this application is an ongoing cost that increases as it gets more popular. This will have the knock-on effect of making it less interesting to writing free apps. If you write one and release it and it’s popular, now you have all kinds of hosting hardware and broadband payments that are making your altruism a costly decision. It remains to be seen if advertising on the iPhone will be as lucrative for websites as it is on the desktop. I suspect that given bandwidth constraints and screen constraints, ad positions will be limited and click thru (and thus revenue) will be extremely low.
There’s obviously opportunities here, as well. Having an online app with an online database, gives people the opportunity to access that data from their regular computer with possibly an enhanced interface and functionality. The barrier to entry for building the app is much lower - more people can try to build things that they need (of course, the cost increases quickly as uptake does).
So, like I’ve said, AJAX on the iPhone is a good thing, I’ll never say it isn’t. All I’m saying that it isn’t the same thing as native apps, it can’t do the same things (link tightly with other local apps and the hardware machine), it doesn’t have the same strengths and it can’t replace them. I hope and I also believe that in the next couple years, whether or not the iPhone does as well as it is hyped, the platform will slowly open to a larger group of developers. Hopefully this will ultimately mean everyone.
