The Mobile Computing Revolution: The Computer in the Pocket

Zusammenfassung

This article traces the history of mobile computing from its origins in the early personal digital assistants of the 1990s, through the smartphone revolution that began with the iPhone in 2007, to the Android ecosystem that brought that revolution to the mass market. It is the story of how computing escaped the desk — and how the device most people now use for the majority of their digital lives emerged from a decade of failed experiments, one theatrical product launch, and a search company that gave its operating system away for free.

Before the Phone: The PDA Era

The idea of a pocket-sized computer predates the smartphone by nearly two decades. Its intellectual origin is usually traced to Alan Kay’s 1972 concept of the Dynabook — a personal, portable computer the size of a notebook, accessible to children, capable of running any program. Kay’s vision was two decades ahead of the hardware required to realize it, but it seeded the imagination of an industry. His role at Xerox PARC and the Dynabook concept are explored in The Xerox PARC Revolution.

The first commercial attempt at a portable personal computer came with the Apple Newton MessagePad (1993). Newton was ambitious: it offered handwriting recognition, a calendar, contacts, and the ability to exchange data with desktop computers. It was also slow, expensive ($700, equivalent to over $1,400 today), larger than anyone wanted to carry, and its handwriting recognition — the feature most prominently marketed — was unreliable enough to become a running joke. Doonesbury ran a week-long strip mocking it. Newton sold modestly for five years before Steve Jobs cancelled it upon returning to Apple in 1997.

Palm Computing succeeded where Apple had failed, by doing far less. The Palm Pilot (1996), designed by Jeff Hawkins, was smaller than Newton, faster, and made no attempt at handwriting recognition in the conventional sense. Instead, Hawkins invented Graffiti — a simplified alphabet that users learned to write, adapted to what the hardware could reliably recognize rather than forcing the hardware to recognize natural writing. It was a compromise. It worked.

The Palm Pilot was the first PDA that people actually carried. By 2000, Palm had sold ten million devices and commanded over 70% of the handheld market. Hawkins and co-founder Donna Dubinsky left Palm to found Handspring, which produced the Treo — the first device to meaningfully combine a PDA with a mobile phone, anticipating the smartphone category before the word existed.

The Convergence Problem

Throughout the 1990s and early 2000s, the mobile industry circled around a question it could not answer: what should a portable device do? Phones were for calls. PDAs were for contacts and calendars. Cameras were for photos. Music players were for music. Each device did its job well; attempts to combine them produced devices that did everything poorly and sold to no one. The technology press called this the “convergence problem,” implying it was a product design challenge. It was actually a software challenge: no one had yet built an operating system that could run all these functions on hardware with a small battery, a small screen, and no physical keyboard.

Nokia and BlackBerry dominated the mobile phone market of the early 2000s. Nokia sold more phones than anyone and had its own smart-device platform, Symbian, which ran on tens of millions of handsets. Research In Motion’s BlackBerry (1999) had captured the enterprise market with physical keyboards, push email, and a security architecture that corporate IT departments trusted. By 2007, BlackBerry had 35% of the US smartphone market. Its executives felt secure.

Steve Jobs and the iPhone

On January 9, 2007, Steve Jobs took the stage at the Macworld Expo in San Francisco and opened with a claim:

“Every once in a while, a revolutionary product comes along that changes everything. And Apple has been very fortunate — it’s been able to introduce a few of these into the world… Today, we’re introducing three revolutionary products of this class. The first one is a widescreen iPod with touch controls. The second is a revolutionary mobile phone. And the third is a breakthrough internet communications device. So, three things: a widescreen iPod, a revolutionary mobile phone, and a breakthrough internet communications device. Are you getting it? These are not three separate devices. This is one device.”

The audience understood. They had been watching him hold up the same object each time.

The iPhone was not the first smartphone. It was not the first touchscreen phone. It was not the first phone that could browse the internet or play music. What it was — and what made it consequential — was the first phone that did all of these things through a software architecture designed from first principles for the constraints of mobile hardware, rather than adapting existing desktop or embedded systems to fit a smaller device.

The key decisions were:

No physical keyboard. Every smartphone before the iPhone allocated a significant portion of its face to physical keys. Jobs bet that a touch-sensitive screen that could reconfigure itself for any task — a phone keypad for calls, a QWERTY keyboard for email, a musical keyboard for a piano app — was worth the loss of tactile feedback. The bet required a capacitive touchscreen (which responded to the electrical properties of a fingertip rather than pressure) combined with software fast enough to make the on-screen keyboard feel responsive. Both were new in a consumer device.

No stylus. The PDA tradition had relied on styluses, which required a dedicated slot, could be lost, and created a barrier between user and device. Jobs’ instruction to his team was simple and absolute: if you need a stylus, you’ve failed.

A desktop-class operating system in a phone. The iPhone ran a version of Mac OS X — based on the same Unix kernel (Darwin) as the desktop Mac. This was the architectural decision that separated the iPhone from every competitor. Symbian, Windows Mobile, and Palm OS were all purpose-built embedded operating systems, designed for the constraints of 1990s hardware and carrying years of compromises. OS X had none of those constraints; it assumed capable hardware and provided a full application framework. The iPhone’s hardware — an ARM 11 processor clocked at 412 MHz (underclocked from its rated maximum for battery and thermal reasons), 128 MB RAM — was modest by desktop standards but sufficient for OS X running at reduced capability.

The Software Advantage

The iPhone’s hardware in 2007 was not dramatically more capable than a contemporary Nokia or BlackBerry. The ARM processor in the first iPhone was similar to what Nokia was using; the screen resolution was comparable. The difference was entirely in the software. Apple had spent three years writing an operating system that treated the phone as a small computer rather than a telephone with extra features. The first iPhone review by Walt Mossberg in the Wall Street Journal noted: “It’s the internet in your pocket for the first time ever.” Nokia and BlackBerry had put the internet in phones years earlier. Nobody had said that about them.

The original iPhone had one significant omission: it did not support third-party applications. Jobs’ position, initially, was that web applications — accessed through Safari — were sufficient. The developer community disagreed loudly. In 2008, Apple launched the App Store and the iOS SDK, allowing third parties to write native applications for the iPhone. The App Store generated $1 billion in its first year. It created an economy.

Android and the Open Counterattack

While Apple was building the iPhone in secret, Google had quietly acquired a small startup in 2005. Android, Inc. had been founded by Andy Rubin — a former Apple and Danger employee — with the goal of building a modern operating system for mobile devices. Google paid $50 million for it.

When the iPhone launched in January 2007, Andy Rubin’s team at Google watched the announcement and went back to the drawing board. The operating system they had been designing assumed a physical keyboard and a trackball. Jobs had just demonstrated that neither was necessary or desirable. The team redesigned Android from scratch.

The first Android device, the HTC Dream (T-Mobile G1 in the US), launched in October 2008 — fifteen months after the first iPhone. It was clunkier than the iPhone, with a physical keyboard that slid out from beneath the screen. But it ran Android, and Android had a property the iPhone did not: it was free.

Google’s strategy was to give Android to any hardware manufacturer that wanted it, license-free. The reasoning was not altruistic. Google’s business was advertising, and advertising revenue depended on users searching the web. A world where most people browsed the internet through an Apple device — with Apple controlling the default search engine — was a world where Google’s influence was mediated by a competitor. Android was Google’s insurance policy against an Apple monopoly on mobile.

The strategy worked beyond any reasonable projection. By 2010, Android device sales exceeded iPhone sales. By 2013, Android commanded over 75% of the global smartphone market. By 2020, over three billion Android devices were in active use worldwide.

The Linux Phone

Android is built on a modified Linux kernel — the same open-source kernel that Linus Torvalds posted to Usenet in 1991. Every Android phone is, at its foundation, a Linux computer. This is one of the more remarkable examples of how open-source infrastructure became the substrate for a commercial ecosystem worth trillions of dollars. The path from Torvalds’ “just a hobby” kernel to the operating system running three billion devices is traced in The Open Source Revolution.

Dead End: The Windows Phone and the Platform Trap

Microsoft had been in the mobile market longer than either Apple or Google. Windows Mobile — a descendant of Windows CE, first released in 1996 — ran on a wide range of devices from multiple manufacturers and had a respectable enterprise market share by the mid-2000s. Microsoft had done the convergence work. It had the partnerships. It had the brand.

The iPhone made Windows Mobile obsolete overnight. Microsoft’s response — Windows Phone 7 (2010) — was a genuine reimagining of the mobile interface, with a distinctive “Live Tiles” design that many reviewers considered more innovative than either iOS or Android.

The Platform Trap

Windows Phone failed for reasons that had nothing to do with the quality of the operating system. By 2010, iOS and Android had established app ecosystems — hundreds of thousands of applications written specifically for their platforms. Developers wrote for iOS because iPhone users were there; iPhone users were there because developers had written for iOS. The same dynamic held for Android.

Windows Phone launched with a fraction of the apps of either competitor. Microsoft offered incentives for developers to port; most calculated that the Windows Phone user base was too small to justify the effort. The user base stayed small because the apps weren’t there. This is the platform trap — a network effect so powerful that a technically competitive product, launching into an established duopoly, cannot acquire enough users to attract enough developers to attract more users. Microsoft spent billions on Windows Phone and Windows 10 Mobile before abandoning the market in 2017. BlackBerry, Symbian, and Palm all fell to the same mechanism.

The Consequences: A New Platform Shifts Everything

The smartphone did not merely add a new device category. It restructured the entire computing industry.

Computing moved to the edge. The majority of internet traffic — over 60% globally by the late 2010s — originated from mobile devices. Web services redesigned themselves around mobile-first assumptions. Entire industries — navigation (GPS in every pocket), photography (compact camera sales fell 90% between 2010 and 2020), music (downloads replaced by streaming), retail (shopping while standing in a store) — were disrupted by a combination of mobile connectivity and location awareness.

The App Store model restructured software distribution. Before 2008, software was sold in boxes or downloaded from websites. After the App Store, the dominant model was a curated marketplace controlled by the platform owner, taking a 30% cut of every transaction. This created new businesses (Instagram, Uber, WhatsApp existed only because smartphones existed) and new conflicts (Apple’s control over its App Store became the subject of antitrust investigations in multiple countries by the 2020s).

ARM became the dominant processor architecture. The iPhone’s success validated ARM’s power-efficient design for mainstream computing. When Apple launched its M1 chip in 2020 — an ARM-based processor for Mac computers that outperformed Intel x86 chips in both speed and power consumption — it was a direct consequence of a decade of ARM optimization for mobile. The architecture of the world’s most powerful consumer laptops traces back to the need to make a phone battery last a day. This transition is explored in The Microprocessor Revolution.

Two companies came to control the access layer to most of the world’s computing. By the 2020s, Apple’s App Store and Google’s Play Store were the gatekeepers through which software reached mobile users — over five billion people. The implications for competition, privacy, and the concentration of technological power remain among the most contested questions in technology policy.

For the network infrastructure that made mobile computing meaningful, see The Connected World. For the AI capabilities that smartphones eventually enabled, see The Rise of Artificial Intelligence.