This post was originally featured on the Pristine blog.
The first PCs could do anything. There were no restrictions. They were sold as disparate hardware components that had to be assembled and programmed by the customer. They were completely and totally free to be modified in any way at the hardware and software layers.
Over time, computers have become increasingly locked down. Through the 90s, it was still relatively easy to open a computer and change out just about any hardware component. Similarly, it was really easy to wipe the OS and load any software stack. But mobile computing has changed things. I remember that as a child, I hated laptops because I couldn't open them up and change out the parts myself. I remember thinking how incredibly stupid it was that every OEM was designing custom, non-rectangular mother boards. 10 years later, I've come to learn that they had good reasons.
Mobile devices are fundamentally different from desktop bound computers. Desktop bound computers physically separate the computer from the human computer interaction (HCI) mechanisms - they keyboard, mouse, speakers, and display - and are inherently limited by battery. This separation allows for modular, standardized innovation within each major hardware and software component. But this doesn't hold true for mobile devices. Mobile devices are forced to make thousands of compromises and design decisions because the physical, power, heat, mobility, battery, configurability, repairability, and software design parameters cannot be isolated. Each of these features pulls against the others. Although it's not a 100% accurate or complete representation, these compromises can be thought of along a spectrum, with power and size at one extreme, and mobility and efficiency at the other.
Over the past 10 years, as we've witnessed the proliferation of mobile devices, convention has triumphed over configuration. There're no signs that this trend will slow. Computers are getting smaller, and smaller computers simply don't accommodate the level of configurability that larger computers do for myriad reasons: smaller components, design and engineering compromises, extremely complex assembly driven by aesthetic design needs, and smaller screens to name a few.
Google Glass and eyeware computing represent the next frontier in mobile computing. These computers are even smaller and lighter than smartphones because they have to be - humans cannot rest more than a few ounces the nose. Glass has been designed to be feather-light so that it can be worn all day without any discomfort.
The Glass Mirror API also reflects the movement towards convention over configuration. The Mirror API is quite limited in what it allows for. Luckily, Google is opening up a native Android development framework - the Glass Development Kit (GDK) - that will be far more powerful than the Mirror API. Even still, the GDK will be limited by Glass's inherent hardware shortcomings. Users will not be willing to configure anything on Glass because configuration requires usernames, passwords, accounts, addresses, selecting items out of lists, etc. Sure, there are passable work arounds for some of these problems via social network integration and other tricks, but even navigating social media menus will be a pain on Glass. Glass apps simply don't accommodate on-device configuration. Google realized that sometime ago, hence the MyGlass app. There will be opportunities to configure Glass apps via smartphones and web interfaces, but the scope of those configurations will pale in comparison to modern smartphone and desktops OSes, which offer hundreds of settings and switches.
What does all of this mean? There will be right and wrong design decisions. And they will be delightfully (or painfully) clear. Apps that won't work really won't work, and apps that work will really shine.
Glass devs, you better use your best judgement. Don't let users configure much of anything, especially on the device. Luckily, small screens don't require nearly as much design and thought expertise as smartphones, tablets, or PCs, simply because the artistic canvas is dramatically smaller to work with. Even still, I'm sure we'll see a hoard of consumer facing apps that fail to account for the fact that people don't actually want to interact with Glass. It's a pain.