commentary If you're one of those people who thinks a manufacturer actually should finish a gadget's software before putting it on sale, it's time to get a grip.
Because in the electronics world, those days are gone.
Today's exhibit: Amazon will update the Kindle Fire software in the next two weeks to address shortcomings in a product on the market for only a month.
I certainly don't like it when new products fall short. But it's time to be practical here, because the situation isn't going to get better anytime soon.
And, perhaps contrary to customer opinion, I think the system we have now actually works better for customers. That's because the fact that we've all become involuntary beta testers also means that manufacturers are listening--capturing our crash logs, seeing the features we use often or not at all, uncovering the baffling step where we abandon a process.
So yes, it's true that companies are often foisting version 1.0 on us when it ought to be labeled version 0.9. But--at least for the companies that are paying attention--today's process can deliver a better product sooner, regardless of its version number.
It's not unfair to expect a product to work when you buy it. That's the way commerce has worked for ages, with manufacturers' reputations taking a hit if they shipped goods that were faulty or that fell short of promises. And when you're buying something like a car or a pair of shoes, that's not unreasonable.
But in the age of networked electronics, things are different, for two reasons.
First, through the miracle of firmware and the Internet, a product can change after you bought it. Second, the competition is incredibly fierce. Manufacturers that wait lose out to those that don't--especially when they're selling not just a solitary product but something like a smartphone or tablet that's the foundation of an ecosystem of apps, content, and social networks.
It's taken a while for the electronics industry to adjust to the idea of firmware. First, of course, came software, a product that's famously mutable. But in earlier days, it was hard to distribute software, especially big packages that came on a CD. The Internet has gradually eroded that, introducing features such as Windows Update and antivirus definitions that can catch up to malware faster.
Electronics were left out for awhile. Now that's changing, too.
Canon turned my SLR into a much more powerful videocamera with a firmware update more than half a year after the 5D Mark II went on sale. And another major video change arrived nearly a year and a half after the camera's debut. But each improvement was a cumbersome manual process.
Now that Internet connections are coming to electronics--printers, smartphones, tablets, and more--you can expect upgrades to be easier.
And more frequent.
Some programs--Chrome, for example, or Windows Update if configured as Microsoft recommends--automatically download new software updates and install them without user input. I don't expect everything to move toward these "silent updates," but I think we'll see a general shift in that direction. I also think we'll see more a more continuous parade of point releases--bug fixes, performance tweaks, and modest new features.
People ridicule Google for having left Gmail in beta for years. In practice, it was silly, given how many people relied on it. But it was probably actually an honest assessment of the difficulties of building, testing, and operating a complex service at massive scale.
Another Google example is version 3 of its Android operating system, aka Honeycomb, which the company rushed to market so tablet makers could answer the success of Apple's iPad and so programmers could adapt their apps to bigger screens. Google initially declined to tag which Android source code patches constituted Honeycomb, making it harder for developers to see how to build it in hopes that they'd focus on the Ice Cream Sandwich version instead. Google apparently relented somewhat on the tagging issue, but the message is clear: Honeycomb was a stepping stone, not a destination, and those using it were guinea pigs, not just customers.
But what choice did Google have? Waiting for Ice Cream Sandwich--only barely arriving today on phones and not at all on tablets--would have given Apple even more time as the sole tablet seller.
As it stands, Apple remains dominant, but at least Android on tablets has now become an accepted idea. You can play Wind-up Knight, Cut the Rope, Riptide GP, and Fieldrunners; you can call your mom on Skype; and you can compose loops with RD3 HD. It's not a commercial triumph of the magnitude of the iPad, but it's not at the bottom of the barrel like HP's TouchPad or RIM's BlackBerry PlayBook, either.
Amazon's Kindle Fire also didn't have the luxury of waiting. The longer it takes the company to deliver the goods, the more time Apple has to get its cloud-based iTunes services in gear, or that Google has to make Google Music, Google Books, and YouTube into something worth paying for.
Even if you have as much lead time over competitors as Apple had with the iPad, there are limits.
Take the example of Siri, another interesting case of beta-labeling honesty to match Gmail. Apple's voice-controlled assistant, arguably the most notable difference between the iPhone 4 and the new 4S, is by many accounts imperfect or at least incomplete.
In a perfect world, Apple could have anticipated the use cases better. It could have refined its speech recognition to handle more dialects, accents, and slang. It could have wired it into more phone and online services.
But guess what? It's not a perfect world. Even Apple, arguably the company with the strongest perfectionism streak today and the most spare cash to attain it, decided it's not worth waiting. As Steve Jobs famously said, "Real artists ship."
For something as complex and as personal as Siri, there came a point at which Apple had to try it in the real world. There's only so much that can be done behind closed doors with a testing population sworn to secrecy.
Likewise, it's hard to predict the interaction of all possible hardware and software combinations. Microsoft Windows, to my mind, is unbelievably stable given the diversity of hardware and software it must accommodate. Apple has things easier with its Macs, iPads, and iPhones, but even with those relatively controlled circumstances, there are going to be surprises in the real world.
Another case in point is Google search. Google is constantly updating its algorithm for delivering results, often changing it several times a week. There is doubtless some behind-the-scenes testing, where Google runs queries that already have been run and checks the results, but at some point it has to try the real world, too. Thus, Google's fondness for split A/B testing, where it compares how one slice of the real world reacts to a change compared to a control group with the status quo.
There's good news for those who don't relish their role as guinea pigs.
Sure, you paid good money for that product whose rough edges you're filing down and whose bugs you're stomping. But at least if you got it from a company that listens--Apple, Google, Adobe Systems, Microsoft, for example--your pains are an investment in your future happiness.
The same pipeline that lets companies send fixes to your software lets you tell them what's going wrong. It all happens through the glorious technology called telemetry, or remote measurement. Mozilla has heat maps to quantify distinguish essential from optional in Firefox's interface. Windows gives Microsoft precise data on which video drivers bring computers down.
Just because everything electronic these days is a work in progress doensn't mean everybody gets a free pass. iOS 5 shouldn't have had battery-draining problems and Samsung's Galaxy Nexus shouldn't have had that volume bug, and both those problems should have been caught earlier. An IBM mainframe and your car's antilock brakes should be held to a much higher standard, too. And if Amazon doesn't get its Fire's act together soon, the market will punish it, because quality and reputation still matter.
If you're having a hard time reconciling yourself to your status as beta tester, I have some further advice:
Don't just get used your role fixing their products. Get revenge.
When that Android app dies, submit the bug report. When that Web page redesign rubs you the wrong way, leave a comment. When that sat-nav system steers you wrong, complain publicly on Twitter. Exercise your power through feedback.
Ultimately, the companies that listen and respond are the ones that deserve your money.