How Apple Made the Web Tiny

The system of hyperlinked digital documents called the World Wide Web running on top of the global network of computer networks called the Internet is huge. Before the web arrived in 1991, something like the web had been an esoteric topic in academia and a common technology in science fiction for years. It was easy to imagine access to all information via some device, but it turned out to be a bit harder to turn that dream into a reality. Even so, the web from fantasy, the web in reality, and the next-generation web in development all share the same scale: huge and getting bigger. The web is also a mass medium rapidly subsuming other mass media like radio, television, newspapers and books. It is just that versatile and it continues to gain new capabilities that allow it to reach further than ever before.

Even huge things can be made tiny when placed next to something of an entirely different scale. While Tim Berners-Lee, the creator of the web, was trying to bring the web down to size for academic purposes, Apple Inc. was preparing to confine the web to a digital box 57 by 57 pixels in size.

Web Science

The system of hyperlinked digital documents called the World Wide Web running on top of the global network of computer networks called the Internet had been available to the public for over fourteen years when “A Framework for Web Science” was published in January 2006. During those years, the web had evolved. It was now referred to as Web 2.0 and it featured social networking, mass distribution of digital media, improved graphic design and layouts, and desktop application-like capabilities. With the Semantic Web (SW), Berners-Lee was promising to make it even more mind-boggling in scope. In his vision, the SW would become more than just a system of hyperlinked digital documents: it would become a system of hyperlinked data. About the SW, Berners-Lee and his co-authors would write in “A Framework for Web Science”:

“But the design of the SW and its associated formalisms and tools is intended to extend the Web to cover linked data, not, as is often assumed, to improve search or get greater power from annotated text (which is another, separate, type of extension of the Web)” [17].

If the web was all about convincing people to share their documents publicly “by adding links to make them accessible by link following” then the SW was “an attempt to extend the potency of the Web with an analogous extension of people’s behaviour. The SW tries to get people to make their data available to others, and to add links to make them accessible by link following. So the vision of the SW is as an extension of Web principles from documents to data. [...] The aim of the SW is to facilitate the use of data as well as their discovery, to go beyond Google in this respect” [17-18].

The SW would be a web of data reaching into the past and future. Legacy data, stored in ancient formats, might be revitalized and made accessible via the new web. Future data, stored in future formats that were SW-aware, could quickly be added. Humans would continue to view these data and the information gained from them in human ways, but machines would also have optimized access to these data. The result would be a vast ocean of data. Into this ocean web agents like web browsers, spiders, intelligent agents, and those yet to be invented would dive, returning to the surface with information that might be pieced together into knowledge (perhaps even automatically).

The SW would be huge, but with the introduction of Web Science, Berners-Lee and colleagues were hoping to explore, constrain, and model the web as well as other networks that exhibited emergent behaviors as they scaled. Web science would take a multidisciplinary approach to this research and it would also examine the legal, social, and ethical implications of these networks. In “A Framework for Web Science” Berners-Lee, et al. merely set out a plan for better understanding the web and related phenomena. In the August 11, 2006 Science Perspectives paper “Creating a Science of the Web”, Berners-Lee and his co-authors stated that Web Science would also be about “more than modeling the current Web. It is about engineering new infrastructure protocols and understanding the society that uses them, and it is about the creation of beneficial new systems” [770 - 771].

Yet in an article in the October 2008 issue of Scientific American, Berners-Lee was still trying to make the case for Web Sciences, despite it having been launched “as a formal discipline in November 2006, when the two of us and our colleagues at the Massachusetts Institute of Technology and the University of Southampton in England announced the beginning of a Web Science Research Initiative. Leading researchers from 16 of the world’s top universities have since expanded on that effort.” Nearly two more years would pass before Web Science took another step forward. In March 2010 the United Kingdom’s Prime Minister Gordon Brown announced funding for the “Institute of Web Science”.

Two months later, the funding was revoked.

Web Science remains a fledgling discipline while the web becomes a daily destination for billions of people. By 2013, Forrester Research expects 2.2 billion people to be online, a significant fraction of the nearly 7 billion-strong human population. An even higher fraction will have access to a mobile phone, which will drive even more people online over the coming decade as these devices become more sophisticated, web-enabled, and cheaper. The web is huge and getting bigger.

Mobile Safari

The system of hyperlinked digital documents called the World Wide Web running on top of the global network of computer networks called the Internet had been available to the public for over fifteen years when the iPhone was announced on January 9, 2007. Until then, the web experience on the cellphones, feature phones, and smartphones with small screens and physical keyboards that supported it was deplorable. People used their mobile phones primarily for talking, secondarily for texting, and rarely for other purposes.

In a lighthearted moment during Apple CEO Steve Jobs’ keynote address, he displayed a humorous rendering of an iPod crossed with a rotary phone [Macworld San Francisco 2007 Keynote Address podcast, 29:44]:

Jobs kept teasing the audience about three new products he was going to introduce: “A widescreen iPod with touch controls, a revolutionary mobile phone, and a breakthrough internet communications device” [27:40]. It turned out the three new products were all integrated into just one device: the iPhone. With its expansive 3.5-inch LCD screen, multi-touch interface, a set of useful preloaded apps, a virtual keyboard that adapted to the needs of open apps, and a graphical user interface that was unquestionably intuitive and attractive, the iPhone was a consumer electronics marvel.

It was also a hit and remains so to this day. The latest fourth generation iPhone is the fastest selling consumer electronic in history, a position it stole from Apple’s iPad which was launched just a few months before, and a position held by each of the previous three generations of iPhone upon their release. More impressively, all iOS (the new name for the operating system that runs in iPhone, iPod touch, and iPad) devices in active use will soon number 100 million units, a figure that will be reached in only a little over three years. With sales increasing and developers flocking to the iOS platform, growth rates continue to exceed every expectation, to the point that Apple is struggling to keep up with demand.

Even so, the number of iOS devices sold is just a fraction of the total number of mobile devices sold every year. In 2006 nearly 1 billion mobile phones were sold. These sales now out pace those for desktops and laptop computers. In 2010, the mobile space has diversified to include smartphones, tablets, and dedicated eBook Readers. As the category “smartphones” implies, many of these mobile devices are increasingly powerful computers in their own right. They are converging with other devices to become multipurpose, portable, and ubiquitous.

These features demand access to the web. But during the 2007 keynote, while Jobs further broke down the “breakthrough Internet communicator” aspect of the iPhone into email, a “real” web browser, Google Maps, and weather and stock widgets, he spent a surprisingly short time on Mobile Safari. True, Mobile Safari provided the full web (minus Flash, which would later become a top critique of iPhone and other iOS devices) to users, but Jobs specifically stated that the iPhone provided “the Internet in your pocket”. This focus on the internet rather than the web would come to define this and future Apple keynotes and press conferences. In fact, it appeared that Apple was going out of its way in minimizing the web browsing capabilities of the iPhone. When the web was necessary, it was there, but Apple promoted other features and capabilities much more prominently. This practice continues to this day, including the recent iPad and iPhone 4 introductions.

At first this apparent oversight seems to be merely convenient. After all, how else would you group together web browsing, email, and widgets that grab their content from the web? In presentations, a common design suggestion is to hit the audience with no more than three bullet points at a time. The capabilities of the iPhone group together naturally into iPod, phone, and internet communicator. But the relative lack of focus on the web browsing component of internet communications stands out in every Apple presentation since the iPhone announcement. It is just there, a passing mention, while email, calendars, and widgets get relatively more focus.

The web, the same web that Berners-Lee was trying to both extend and quantify, had been confined by Apple to a Mobile Safari icon 57 by 57 pixels in size. At the March 6, 2008 iPhone Software Roadmap press conference, Apple’s positioning of the web became even more obvious.

The original iPhone did not include any third-party apps. All of the included apps were developed in house by Apple. Developers began to complain that they could not build apps for this extraordinary device. Apple provided information about developing iPhone-compatible web apps that would open in Mobile Safari, and they even added the capability to save icons for these web apps for quick launching from iPhone’s home pages.

With the launch of iPhone SDK, the web development platform was relegated to a footnote. Apple provided developers with the same sophisticated APIs and tools that Apple used to create its own iPhone apps. Jobs said at the event “Apple is a platform company” and it was clear they would be focusing on their own platform rather than the web. The components built into iPhone OS included internet networking and web view, so that browser capabilities could be built easily into third party apps. But the web was not the platform; instead it was a component to be built into apps if needed.

To date nearly a quarter million apps had been created and launched in the App Store. Some are directly web-related, with web browsing capabilities built into them, repurposed web content, or access to the web as a primary source of their data. Most, however, are not specifically web-related. They are instead games, digital media, and other experiences in which the web plays a minor if any role. More apps are internet-aware than web-aware. At best, the web is a component. At worst, it is replaced by developers own digital gardens.

The activities that Apple demonstrates on its iOS devices are more often internet-capable than web-enabled. The focus is on the internet and the web is just one of a multitude of services that take advantage of the internet.

iPhone and related devices are internet devices, hardware and software bundles that connects to the massive hardware and software bundle that is the internet. In doing so, these devices gains access to a variety of services, of which the web is only one. For example, Augmented Reality apps, like Layar provide a window into the digital world around you through the camera on mobile devices. The digital information is provided as layers superimposed on the video feed. These digital data, however, are not always from the web. Through apps like this, data does not have to be web-based to be accessible.
Games offer another example of the internet versus web focus of iPhone. Many of these games participate in social gaming networks that make use of the internet but do not necessarily have a web-facing presence. These networks are not built using web technologies but using in-house software development platforms. Apple is building its own Game Center as a component developers can utilize in their own games. While the data eventually gathered by Game Center might be accessible in a web browser (Apple has not announced many details about the service), it will not be primarily a web service. It is a service built using more sophisticated tools than are currently available to the web.

Mobile Safari (and the APIs that integrate web browsing capabilities into apps) confined the web into specific user experiences, but there are many other user experiences that do not require the web at all. The icon for Mobile Safari just might serve as the coffin for the web and the native app development platform is the first nail.

The Web Versus Native App Development Platforms

The web, despite important progress, is nowhere near as capable a development platform as those based on programming languages like Objective-C. This is the result of constraints that have been in place since the web was first publicly launched in 1991 and further imposed by Semantic Web development in the 21st century. “The WorldWideWeb (WWW) project aims to allow links to be made to any information anywhere,” Berners-Lee said on a newsgroup on August 2, 1991. In a follow-up post he stated “The WWW world consists of documents, and links.” This focus on digital documents led to a focus on text and pages. As the web matured it began to support the addition of graphics, images, audio, video, and other multimedia content, but always these digital media were integrated into webpages. This is a highly successful metaphor, and it remains to this day.

But software developers create tools that make use of other metaphors. They are not limited to pages. The tools they use are more complex than the tools used to create webpages. Even as web services begin to add the same capabilities as desktop applications, they are generally not as sophisticated. Likewise, when it comes to 3D gaming and virtual reality, vector graphics, sensing, and many other capabilities, web apps cannot yet compete with native apps. It is doubtful they ever will, if only because the web itself constrains their potential. The web with its focus on digital documents simply does not allow for the richness in experiences other platforms enable.

The Semantic Web continues this trend. With a focus on XML-based markup languages, the SW stores data in digital documents that are shared between web services. While this allows for a base metaphor by which legacy and future data stores can be connected, the performance of this metaphor compared to other metaphors is often poor, and it does nothing to enable other metaphors that may be useful in the future. As devices add multi-touch, haptics, gesture and voice recognition, and brain-machine interfaces, the web lags in taking advantage of these.

By opening up APIs to device hardware and other capabilities, supporting apps through iTunes, taking care of sales and distribution, and taking a small 30% cut of proceeds for any paid apps, Apple created a unrivaled developers’ platform. Developers responded immediately and the App Store became the crowning jewel of the iOS platform. Updates to the iPhone operating system and hardware kept Apple ahead of the pack of emerging smartphone platforms, including Google’s Android and Palm’s WebOS. With the successful introduction of iPad and iPhone 4 in 2010, the ecosystem of devices and software based on the iOS platform continues to grow rapidly. In the future, the platform will likely extend to television-related consumer electronics.

The web is coming along for the ride, but native apps retain key advantages in speed, power, and capabilities. Will web browsers catch up? Will web apps someday rival native apps? It is possible, but the web is running out of time. Other emerging technologies could arrive before the web has time to mature in this way.

One immediate advantage iOS development has over web development is its proprietary nature. The team of engineers and developers pushing out a major new iOS upgrade ever year are able to use non-standard APIs to take advantage of hardware advances at a far quicker pace than standards bodies and open source web browser developers can do so. Three years after the introduction of iPhone, browsers like Firefox and Chrome are just now gaining native multi-touch support. While they plan future browser support for hardware like accelerometers, iPhone 4 has introduced gyroscopes to mobile devices for the first time.

The open source browser projects are also trying to push out HTML5 and CSS3 support even before these languages have been formally approved. While this is not in any way uncommon in the history of web browser development, it suggests the tension between compatible standards support versus rapid development will only increase, while proprietary projects woo developers to platforms with low costs, high percentage of any returns, regular platform updates and accelerated device obsolescence.

While it is difficult for open source projects to adapt to the rapid pace of proprietary development, Apple can open source technology when it suits is needs. A key example: the open sourcing of the video technology behind FaceTime. Not only did Apple introduce a compelling new video conferencing capability in iPhone 4, but it also promised the technology would be available to other platforms, and that tens of millions of people would have FaceTime capabilities over Wi-Fi by the end of 2010. Jobs said that by 2011 Apple would begin rolling out FaceTime across AT&T’s 3G network. Just like that FaceTime became the compelling new must-have feature of iPhone 4 and developers began contemplating how they could make use of the technology in apps and on other platforms.

Mobile Safari on iPhone was a much improved mobile web experience but when websites decried the omission of support for Adobe Flash, Apple defended its decision by suggesting Flash was old technology, slow and battery killing. Whether or not this is true, HTML5 is an open web standard undergoing rapidly developing (although progress had previously been sluggish for many years.) Mobile Safari is now an HTML5 powerhouse. Websites that depend on Flash are transitioning to HTML5 for video and audio streaming. Games in Flash are being recreated in Objective-C as iPhone apps.
Apple can adapt quickly through proprietary and open source means, while open source web browser development struggles to catch up. As a consequence, Mobile Safari is just one app among many. Web browsing had been cornered into a square icon on iPhone’s dock, competing for attention with other experiences that are not necessarily limited to the web. Other apps can have web features, or they can make use of far more powerful tools than HTML, CSS, and Javascript. An app can be internet-enabled without being related to the web, or it can be web-enabled but so much more in addition. 250,000 apps later, Apple has succeeded in making the web tiny.

With each hardware refresh, app developers can take advantage of even more APIs developed by Apple. In iPhone 4, apps can now take advantage of a back camera that takes still shots at 5MP or video at 720p in 30 fps. There is also a front camera for video conferencing and lower resolution stills. In addition to GPS and a digital compass introduced in past generations, iPhone 4 gains a gyroscope, allowing higher resolution sampling of the iPhone’s position in 3-D space. The custom chip that powers iPhone 4 has more memory, the battery is bigger, the screen resolution has been increased to near-print quality, and the device is thinner.

Apple made the web tiny, and now it is tiny on ever competitor platform. The Android platform is not nearly as polished as the iPhone platform, but thousands of apps have been created that add new capabilities to Android smartphones and other devices. Games have flourished, as have GPS navigators, digital media consumption apps, etc.

While it is true that web browsers are just one of many applications on laptops and desktops, for many people web browsers serve as the primary and most prominent application. In fact, Google is developing Chrome OS for netbooks and tablets in which everything is contained within the web browser. The web browser in this example has become the operating system.

This is not the case of mobile devices, however. The web browser is just one of many apps. One could browse to Twitter using Mobile Safari, but instead most people use a dedicated app for accessing Twitter. These apps wrap the twitter experience in new software, specifically tailored for multi-touch, location detection, and other capabilities that are just not common on desktops and laptops.
But it does not have to stop there. Continuing to use Twitter as an example, app developers are experimenting with new ways to display tweets. Using Augmented Reality, the phone can “point” to where tweets are being posted. The video feed through the camera is augmented with flags. As the user turns around, they can see tweets in various directions. Location-based apps can also use Tweets as reviews for restaurants and other physical destinations. Twitter then becomes a database which provides data that can be displayed in unique ways on capable smartphones.
This suggests a radical direction internet-enabled technologies are heading: consumers do not need to depend only on the web. In fact, in some of these cases, the web would be too limiting. Apps can be written that treat the web as a database but use other tools for displaying this information. Each app, then, becomes its own mass medium, limited only by the experience of developers on increasingly capable hardware and mobile OS platforms.

In other words, each app is its own experience, its own space. It does not have to be a web browser while making use of data from the web. It can project these data in new ways. The web is just one of many data projections. The web is now tiny, and the space that includes all data projections is enormous.

If the web is only one data projection, then what are all of these data projections together? In the next chapter the Platform Wars and fragmentation are unleashed.