Originally, this article was going to be about the OpenMoko, and why I was likely to buy it. However, I changed my thoughts a little and decided to write more about what’s wrong with Apple’s stance on third-party development, as well as my feelings on the mobile platforms of the future.
Symbian gets it. Sun gets it. Hell, even Microsoft gets it… and until recently I thought Apple got it. Third party development drives platforms, not the other way around. Can you think of a single platform that survived to the present day in which all of the operating system and applications were “first-party” i.e. produced by a single manufacturer? I think Commodore tried it once… it was called the Plus/4 and it was NOT a great success.
It used to be that closed platforms were the way to go. Even in the early days of personal computers in the ’70’s, it wasn’t unusual to find a single company making the computers, writing the operating system and writing all the software. Of course, in those days it was because there were a multitude of incompatible platforms which forced manufacturers to produce their own software or else perish because their computers could do nothing.
All of this changed even before the PC came on the scene. Third parties started writing software for some of those platforms, and the software industry was born. By the time the PC rolled around, the idea of a completely independent third party producing an operating system and yet another third party producing software for a platform was old news.
Yet here we are, 30 years later with a new and promising platform about to arrive… and it’s locked out the third parties. I mean, what? When you have an extremely small developer base working on a platform, the platform will stagnate. It’s just that simple. With no influx of new independent development, there’s no reason for the developers of the existing software to actually improve what they’ve already written. I know, I’ve watched it happen.
The problem runs deeper and in my opinion has the ability to really destroy the iPhone as a platform. I have worked with the disabled, and organizations to help the disabled. I know from experience that “one size fits all” is a model that definitely doesn’t fit. What works for one’s disabilities will be a problem for another. As a result, when working with computer systems they often must be customized and tailored to the specific needs of the individual.
Take for example the Windows Mobile / Smartphone model. They allow third-party development and have done for years. Contrary to comments from Steve Jobs, Cingular and T-Mobile’s networks have yet to come crashing down due to a rogue application… and generally these devices perform relatively well. The strength of the platform is that custom applications can be written and installed, and thus a mobile device running WM can be tailored to the individual just like a desktop or laptop computer. Believe me, no-one wants mobile technologies they can use more than the disabled. They can get far more use out of any enabling technology than able-bodied people ever will, and mobile enabling technology frees them from so many shackles that the able-bodied will never truly understand.
I’m not going to claim the WM model is perfect; it’s not. Yes, a rogue application can cause problems in the operating system, but it’s still a relatively closed environment. It’s not like spyware on a PC that installs itself; you have to actually manually install software to a WM device. I’ve used one for years, and I like it.
The poster-child for an enabling technology among the disabled in the last few years has in fact been the Blackberry. It’s a good technology with a decent implementation based on a relatively open Java-based platform (therefore developing for it is relatively simple). It’s also an aging platform that is really out of its depth in today’s world, but hangs on because of inertia. Windows Mobile 5 has exceeded it in every respect except market penetration, but that’s starting to change. You see, the cost of Blackberry is high. It requires its own server in addition to your Exchange (or similar) email server. It has licensing costs, it has per device costs… it’s incredible. WM5 can hook up to Exchange as another Outlook client and in that it works really well. The cost is significantly lower, both in initial investment and ongoing cost. As a result, organizations relying on these devices are already looking for alternatives.
When I first watched the iPhone presentation at Macworld, I hoped this device could be one that could supplant the Blackberry. But the more I learned, the more I realized that this device is merely a toy at an horribly inflated price. When I heard about the closed development I switched off, and the Steve Jobs Reality Distortion Field released its grip.
Being open to development is a huge deal. You may think you have the best platform in the world, you may think you have the next big thing. But you’re one person, or one company with a relatively myopic vision of what the future will bring. You can’t predict every use your device may be put to, no matter how many analysts you hire. Because of this, you can’t possibly develop all possible applications for a platform, unless you intend the platform to only have very limited use. You artificially limit your own growth potential because you are unwilling to allow others to help your platform succeed.
I had this discussion with a friend of mine about the iPhone and he said “Well, then Apple will allow third-parties to develop, but you’ll have to go through Apple to install them. Like you do with games on the iPod.” Fair enough, but this is the viewpoint of someone who’s never developed for a platform before.
I say this because first of all if you can’t manually install the applications on an iPhone how can you develop? The best you can do is develop in a “Virtual iPhone” environment, but you have no way of knowing how a REAL iPhone will behave once your application is installed. Virtual environments are never exactly like the real platform, no matter how well developed.
Secondly, what happens even if your application is developed perfectly? Well, if Apple is the gatekeeper, then first of all there’s the certification process. This process will inevitably incur cost. This cost can be covered in one of two ways; either that cost must be paid by the developer or it must be paid by the end user.
Let’s examine the second one first. The user pays. Fine and dandy, but first there’s the issue that the user may not want to pay for the application. Especially if a company already hired a consultant to develop an application for the platform, they don’t want to pay a “per user” license fee to download that application to all their devices. Secondly, there’s the issue of economies of scale. I said the certification process will incur cost; there’s time, there’s people, there’s testing. This can be extremely costly even for a simple application, and this must be recouped. If the application is of limited use (such as the examples of the disabled I mentioned above), they will only have maybe a few dozen installs of an application during the application’s entire life-span. In that case, the per-user license must be more costly in order to recoup that cost.
What about the developer paying? Well, many developers for small applications like I’m referring to can’t realistically afford it. Many of them operate on a financial razor’s edge and will come out with only a small profit at the end of a large application deployment. Any increase in their costs must be passed onto the customers. This will inevitably decrease the number of customers for that coder and will eventually drive then out of business as a vicious cycle ensues.
Even the “Manufacturer as the gatekeeper” model inevitably leads to the destruction of third-party developers as it becomes untenable to actually develop for a platform. If third-party developers can’t develop for a platform, then it becomes an also-ran in the race for platform deployments. Sure, the average Joe Consumer doesn’t worry so much about this, but what about when that aforementioned consumer finds that their friend’s device does something theirs doesn’t? Do they go to the Apple App Store and download a new application? Where does this app come from? Well, if it’s of limited interest then it won’t come from Apple. Apple will only develop applications that interest them and have a wide appeal. That’s because development of niche applications is not profitable for all the reasons mentioned above and then some.
But didn’t Apple create its current market out of niche applications like graphics, design and desktop publishing? Why yes, they did. As a result, their core market for Macs today are those self-same “niche-users” who bought Macs in the past for their specific application requirements. These same users are the exact same ones who by definition often have unique requirements of their devices. Sure, they’d love a slick phone… but what about the CAD viewer? What about the reference dictionary containing all they need to know about design aspects or art? What about a drink mixing database? These are all niche applications that either Apple will have to develop or a third-party will have to get certified by Apple. Any of these applications I can download for my phone today, often for free.
Thus we have a quandry. We have a device that’s been marketed by a de-facto niche company, but it’s a device that will fail to meet the needs of many of the users who are most likely to buy the iPhone. The extensibility of the platform is hobbled by the fact that Apple wants to control the complete “experience”. The usability of the platform is only good for the able bodied or those broadly defined “disabled” who will be able to use the accessibility functions; but my experience says the realistic number of disabled who can make do with those generic functions will be less than 20% of those who even want to buy the iPhone.
Joe Consumer might buy it, but the price point is too high for most people to care. Its price point pits it against smartphone devices like the Treo, Windows Mobile devices and others including the Blackberry. Buyers of these devices are looking for something that’s an application platform as well as a phone.
So who will buy it? The yuppie who cares more about style than functionality. The same people who rushed out to buy a RAZR on day one at $500 a pop because it was trendy, then lost interest. Witness how quickly the RAZR dropped in price; now I can pick one up for less than $100 without a contract! And you know what? I see a hell of a lot more SLVRs and ROKRs around these days than I do RAZRs. Fashion is finnicky. Will Apple sell some? Yes. Will they capture the market? I doubt it.
Let me put it this way. I own a Mac. I look at my desktop right now and I see 6 applications running, only one of which is produced by Apple (Safari). The rest (Adium, Ecto, Yahoo! Messenger, Gnome-Terminal) are ALL third-party apps. Would the Mac be anywhere today if you couldn’t run Microsoft Office on it? I may not like it much (I prefer OpenOffice), but software is what makes a platform. And Apple may create a great platform, but most of its software is easily surpassed by third-party offerings.
So what am I going to buy when I replace my MPX220? Well, quite honestly it’ll probably be another Windows Mobile device. Must admit I have a soft spot for the GloFiish M700. Anyone wanna buy me one for my birthday? 🙂