CNET también está disponible en español.

Ir a español

Don't show this again

Build it right and they will come

Action Engine CEO Craig Eisler has a suggestion for developers: Before thinking about the wireless Internet, build wireless applications that work.

No matter what you think about the wireless Internet, the mobile data industry, and the dozens of companies pushing their services, it's clear that businesses and consumers will continue to want more and better access to content and services on mobile devices.

Argue one analyst's predictions of adoption rates and timing over another's, but it seems irrefutable that widespread adoption of wireless Internet-capable devices is on the horizon.

Unfortunately, the predominant approach to wireless data solutions on mobile devices has been misguided. Although it may seem logical to port the browser-based wired Internet model found on the PC to mobile devices, the truth is that there's almost nothing in common between a PC and a PDA, a pager or a smart phone.

There is one similarity, and it's a critical one: Today's mobile devices are powerful computing platforms, even when viewed from the perspective of the recent past. Back in the mid-'90s, the Windows 95 platform was designed around a target PC with a 386SX Intel chip and 4MB of RAM. Need I say more about the success of that platform in enabling thousands of successful applications to transform the end-user experience and send adoption rates skyward?

If we look at the computational guts of the popular RIM BlackBerry 957 device, we find a 386 Intel chip and 5MB of RAM. And what are software developers doing with all that computing power? Enabling corporate e-mail users to fulfill their "crackberry" addiction.

If the industry looked at mobile devices as mobile computing devices, while taking into account the capabilities and limitations of the wireless networks they've built today, they could actually deliver a user experience that is compelling enough to drive massive adoption and drive bottom-line productivity and efficiency levels within enterprises.

"And," you ask, "What would the ideal solution look like?" Here are the must-haves:

• Move beyond the browser model.
Browsing was perfect for the wired desktop Internet, but it is far from perfect for the wireless Internet. Difficult data entry, small display screens, and high-latency wireless network connections all contribute to the challenges of relying on a server-centric browser model for delivering mobile data.

Not to mention the fact that people don't particularly want to browse and search when they're on the go; they want targeted, relevant results that they can access even when they're not connected to the network.

Emphasizing tapping instead of graffiti, data reuse instead of data re-entry, and local processing instead of network access could dramatically alter a person's experience on a smart device. A smart client/smart server approach could leverage client-side software to create a person's request by intelligently and efficiently gathering necessary information while offline. Only then would the wireless network be accessed to transmit the request to the server for processing and to access any necessary databases.

• Minimize dependence on network connectivity.
Carriers are investing billions in next-generation wireless networks. The promise of faster public networks (2.5G and 3G) sounds really exciting. But those networks are not here yet, and even when they are, building enough infrastructure to ensure complete wireless coverage around the globe isn't easy (think about tunnels, or elevators, or many office buildings) and is destined to be an ongoing challenge. Any useful wireless data technology should be designed with this limitation in mind.

A desirable solution must optimize the use of each data transmission by collecting more data up front from the user, thus minimizing the need to use the wireless network in the first place. That way, if a connection breaks, the software would remember where the user left off rather than restarting from square one when the connection resumes.

And if people want to access their data and modify it when there isn't a wireless connection, they should be able to do so. No more dependency on the network every step of the way.

• Make it worth the investment.
Compelling applications are a must. The same way the desktop PC needed spreadsheets and word processing applications to be truly useful, wireless Internet devices need applications that are relevant and truly useful in the mobile environment. Examples include applications that allow people to book business travel, order sales inventory, or manage customer-relationship data. In other words, activities and tasks that businesses and consumers find complementary to the current wireless killer app: e-mail.

Applications should work across multiple operating systems, from the Palm OS to the Pocket PC to RIM to J2ME, and they should take advantage of the separate devices' and OS' key abilities. A broad range of devices using a variety of operating systems will be prevalent in the marketplace for at least several years.

I wouldn't bet my livelihood on one winner in the short term, and it's conceivable that there won't ever be one winner. Some large and powerful companies lost in the desktop and browser wars; they are not going to give up the wireless battlefield without a fight.

There is a huge opportunity for companies building "middleware," the software layer that sits between the mobile device OS and the applications running on the device. This software layer will enable development of compelling applications that leverage the aforementioned points and are critical drivers of business and consumer adoption of the wireless Internet.

And I promise you, if you build it right this time, they will come.

Close
Drag
Autoplay: ON Autoplay: OFF