Introducing Android Wireless Application Development

This chapter is from the book 

The mobile development community is at a tipping point. Mobile users demand more choice, more opportunities to customize their phones, and more functionality. Mobile operators want to provide value-added content to their subscribers in a manageable and lucrative way. Mobile developers want the freedom to develop the powerful mobile applications users demand with minimal roadblocks to success. Finally, handset manufacturers want a stable, secure, and affordable platform to power their devices. Up until now a single mobile platform has adequately addressed the needs of all the parties.

Enter Android, which is a potential game-changer for the mobile development community. An innovative and open platform, Android is well positioned to address the growing needs of the mobile marketplace.

This chapter explains what Android is, how and why it was developed, and where the platform fits in to the established mobile marketplace.

A Brief History of Mobile Software Development

To understand what makes Android so compelling, we must examine how mobile development has evolved and how Android differs from competing platforms.

Way Back When

Remember way back when a phone was just a phone? When we relied on fixed landlines? When we ran for the phone instead of pulling it out of our pocket? When we lost our friends at a crowded ballgame and waited around for hours hoping to reunite? When we forgot the grocery list (see Figure 1.1) and had to find a payphone or drive back home again?

Those days are long gone. Today, commonplace problems such as these are easily solved with a one-button speed dial or a simple text message like "WRU?" or "20?" or "Milk and?"

Our mobile phones keep us safe and connected. Now we roam around freely, relying on our phones not only to keep in touch with friends, family, and coworkers, but also to tell us where to go, what to do, and how to do it. Even the most domestic of events seem to revolve around my mobile phone.

Consider the following true story, which has been slightly enhanced for effect:

• Once upon a time, on a warm summer evening, I was happily minding my own business cooking dinner in my new house in rural New Hampshire when a bat swooped over my head, scaring me to death.
• The first thing I did—while ducking—was to pull out my cell phone and send a text message to my husband, who was across the country at the time. I typed, "There's a bat in the house!"
• My husband did not immediately respond (a divorce-worthy incident, I thought at the time), so I called my dad and asked him for suggestions on how to get rid of the bat.
• He just laughed.
• Annoyed, I snapped a picture of the bat with my phone and sent it to my husband and my blog, simultaneously guilt-tripping him and informing the world of my treacherous domestic wildlife encounter.
• Finally, I googled "get rid of a bat" and then I followed the helpful do-it-yourself instructions provided on the Web for people in my situation. I also learned that late August is when baby bats often leave the roost for the first time and learn to fly. Newly aware that I had a baby bat on my hands, I calmly got a broom and managed to herd the bat out of the house.
• Problem solved—and I did it all with the help of my trusty cell phone, the old LG VX9800.

My point here? Mobile phones can solve just about anything—and we rely on them for everything these days.

You notice that I used half a dozen different mobile applications over the course of this story. Each application was developed by a different company and had a different user interface. Some were well designed; others not so much. I paid for some of the applications, and others came on my phone.

As a user, I found the experience functional, but not terribly inspiring. As a mobile developer, I wished for an opportunity to create a more seamless and powerful application that could handle all I'd done and more. I wanted to build a better bat trap, if you will.

Before Android, mobile developers faced many roadblocks when it came to writing applications. Building the better application, the unique application, the competing application, the hybrid application, and incorporating many common tasks such as messaging and calling in a familiar way were often unrealistic goals.

To understand why, let's take a brief look at the history of mobile software development.

"The Brick"

The Motorola DynaTAC 8000X was the first commercially available cell phone. First marketed in 1983, it was 13 x 1.75 x 3.5 inches in dimension, weighed about 2.5 pounds, and allowed you to talk for a little more than half an hour. It retailed for $3,995, plus hefty monthly service fees and per-minute charges.

We called it "The Brick," and the nickname stuck for many of those early mobile phones we alternatively loved and hated. About the size of a brick, with a battery power just long enough for half a conversation, these early mobile handsets were mostly seen in the hands of traveling business execs, security personnel, and the wealthy. First-generation mobile phones were just too expensive. The service charges alone would bankrupt the average person, especially when roaming.

Early mobile phones were not particularly full featured. (Although, even the Motorola DynaTAC, shown in Figure 1.2, had many of the buttons we've come to know well, such as the SEND, END, and CLR buttons.) These early phones did little more than make and receive calls and, if you were lucky, there was a simple contacts application that wasn't impossible to use.

The first-generation mobile phones were designed and developed by the handset manufacturers. Competition was fierce and trade secrets were closely guarded. Manufacturers didn't want to expose the internal workings of their handsets, so they usually developed the phone software in-house. As a developer, if you weren't part of this inner circle, you had no opportunity to write applications for the phones.

It was during this period that we saw the first "time-waster" games begin to appear. Nokia was famous for putting the 1970s video game Snake on some of its earliest monochrome phones. Other manufacturers followed suit, adding games such as Pong, Tetris, and Tic-Tac-Toe.

These early phones were flawed, but they did something important—they changed the way people thought about communication. As mobile phone prices dropped, batteries improved, and reception areas grew, more and more people began carrying these handy devices. Soon mobile phones were more than just a novelty.

Customers began pushing for more features and more games. But there was a problem. The handset manufacturers didn't have the motivation or the resources to build every application users wanted. They needed some way to provide a portal for entertainment and information services without allowing direct access to the handset.

What better way to provide these services than the Internet?

Wireless Application Protocol (WAP)

As it turned out, allowing direct phone access to the Internet didn't scale well for mobile.

By this time, professional websites were full color and chock full of text, images, and other sorts of media. These sites relied on JavaScript, Flash, and other technologies to enhance the user experience, and they were often designed with a target resolution of 800x600 pixels and higher.

When the first clamshell phone, the Motorola StarTAC, was released in 1996, it merely had an LCD 10-digit segmented display. (Later models would add a dot-matrix type display.) Meanwhile, Nokia released one of the first slider phones, the 8110—fondly referred to as "The Matrix Phone" because the phone was heavily used in films. The 8110 could display four lines of text with 13 characters per line. Figure 1.3 shows some of the common phone form factors.

With their postage stamp-sized low-resolution screens and limited storage and processing power, these phones couldn't handle the data-intensive operations required by traditional web browsers. The bandwidth requirements for data transmission were also costly to the user.

The Wireless Application Protocol (WAP) standard emerged to address these concerns. Simply put, WAP was a stripped-down version of HTTP, which is the backbone protocol of the Internet. Unlike traditional web browsers, WAP browsers were designed to run within the memory and bandwidth constraints of the phone. Third-party WAP sites served up pages written in a markup language called Wireless Markup Language (WML). These pages were then displayed on the phone's WAP browser. Users navigated as they would on the Web, but the pages were much simpler in design.

The WAP solution was great for handset manufacturers. The pressure was off—they could write one WAP browser to ship with the handset and rely on developers to come up with the content users wanted.

The WAP solution was great for mobile operators. They could provide a custom WAP portal, directing their subscribers to the content they wanted to provide, and rake in the data charges associated with browsing, which were often high.

Developers and content providers didn't deliver. For the first time, developers had a chance to develop content for phone users, and some did so, with limited success.

Most of the early WAP sites were extensions of popular branded websites, such as CNN.com and ESPN.com, which were looking for new ways to extend their readership. Suddenly phone users accessed the news, stock market quotes, and sports scores on their phones.

Commercializing WAP applications was difficult, and there was no built-in billing mechanism. Some of the most popular commercial WAP applications that emerged during this time were simple wallpaper and ringtone catalogues that enabled users to personalize their phones for the first time. For example, a user browsed a WAP site and requested a specific item. He filled out a simple order form with his phone number and his handset model. It was up to the content provider to deliver an image or audio file compatible with the given phone. Payment and verification were handled through various premium-priced delivery mechanisms such as Short Message Service (SMS), Enhanced Messaging Service (EMS), Multimedia Messaging Service (MMS), and WAP Push.

WAP browsers, especially in the early days, were slow and frustrating. Typing long URLs with the numeric keypad was onerous. WAP pages were often difficult to navigate. Most WAP sites were written one time for all phones and did not account for individual phone specifications. It didn't matter if the end user's phone had a big color screen or a postage stamp-sized monochrome screen; the developer couldn't tailor the user's experience. The result was a mediocre and not very compelling experience for everyone involved.

Content providers often didn't bother with a WAP site and instead just advertised SMS short codes on TV and in magazines. In this case, the user sent a premium SMS message with a request for a specific wallpaper or ringtone, and the content provider sent it back. Mobile operators generally liked these delivery mechanisms because they received a large portion of each messaging fee.

WAP fell short of commercial expectations. In some markets, such as Japan, it flourished, whereas in others, such as the United States, it failed to take off. Handset screens were too small for surfing. Reading a sentence fragment at a time, and then waiting seconds for the next segment to download, ruined the user experience, especially because every second of downloading was often charged to the user. Critics began to call WAP "Wait and Pay."

Finally, the mobile operators who provided the WAP portal (the default home page loaded when you started your WAP browser) often restricted which WAP sites were accessible. The portal enabled the operator to restrict the number of sites users could browse and to funnel subscribers to the operator's preferred content providers and exclude competing sites. This kind of walled garden approach further discouraged third-party developers, who already faced difficulties in monetizing applications, from writing applications.

Proprietary Mobile Platforms

It came as no surprise that users wanted more—they will always want more.

Writing robust applications with WAP, such as graphic-intensive video games, was nearly impossible. The 18-year-old to 25-year-old sweet-spot demographic—the kids with the disposable income most likely to personalize their phones with wallpapers and ringtones—looked at their portable gaming systems and asked for a device that was both a phone and a gaming device or a phone and a music player. They argued that if devices such as Nintendo's Game Boy could provide hours of entertainment with only five buttons, why not just add phone capabilities? Others looked to their digital cameras, Palms, BlackBerries, iPods, and even their laptops and asked the same question. The market seemed to be teetering on the edge of device convergence.

Memory was getting cheaper, batteries were getting better, and PDAs and other embedded devices were beginning to run compact versions of common operating systems such as Linux and Windows. The traditional desktop application developer was suddenly a player in the embedded device market, especially with smartphone technologies such as Windows Mobile, which they found familiar.

Handset manufacturers realized that if they wanted to continue to sell traditional handsets, they needed to change their protectionist policies pertaining to handset design and expose their internal frameworks to some extent.

A variety of different proprietary platforms emerged—and developers are still actively creating applications for them. Some smartphone devices ran Palm OS (now Garnet OS) and RIM BlackBerry OS. Sun Microsystems took its popular Java platform and J2ME emerged (now known as Java Micro Edition [Java ME]). Chipset maker Qualcomm developed and licensed its Binary Runtime Environment for Wireless (BREW). Other platforms, such as Symbian OS, were developed by handset manufacturers such as Nokia, Sony Ericsson, Motorola, and Samsung. The Apple iPhone OS (OS X iPhone) joined the ranks in 2008. Figure 1.4 shows several different phones, all of which have different development platforms.

Many of these platforms have associated developer programs. These programs keep the developer communities small, vetted, and under contractual agreements on what they can and cannot do. These programs are often required and developers must pay for them.

Each platform has benefits and drawbacks. Of course, developers love to debate about which platform is "the best." (Hint: It's usually the platform we're currently developing for.)

The truth is that no one platform has emerged victorious. Some platforms are best suited for commercializing games and making millions—if your company has brand backing. Other platforms are more open and suitable for the hobbyist or vertical market applications. No mobile platform is best suited for all possible applications. As a result, the mobile phone has become increasingly fragmented, with all platforms sharing part of the pie.

For manufacturers and mobile operators, handset product lines quickly became complicated. Platform market penetration varies greatly by region and user demographic. Instead of choosing just one platform, manufacturers and operators have been forced to sell phones for all the different platforms to compete in the market. We've even seen some handsets supporting multiple platforms. (For instance, Symbian phones often also support J2ME.)

The mobile developer community has become as fragmented as the market. It's nearly impossible to keep track of all the changes in the market. Developer specialty niches have formed. The platform development requirements vary greatly. Mobile software developers work with distinctly different programming environments, different tools, and different programming languages. Porting among the platforms is often costly and not straightforward. Keeping track of handset configurations and testing requirements, signing and certification programs, carrier relationships, and application marketplaces have become complex spin-off businesses of their own.

It's a nightmare for the ACME Company that wants a mobile application. Should it develop a J2ME application? BREW? iPhone? Windows Mobile? Everyone has a different kind of phone. ACME is forced to choose one or, worse, all of the platforms. Some platforms allow for free applications, whereas others do not. Vertical market application opportunities are limited and expensive.

As a result, many wonderful applications have not reached their desired users, and many other great ideas have not been developed at all.