Contents: The developer: in the trenches The CEO: the view from the mountain The analyst: following the action Coming up May 15: Resources About the author Rate this article Related content: Keep it simple for wireless Subscriptions: dW newsletters dW Subscription (CDs and downloads)

Deborah Durham-Vichr (deborah@nemix.com)
Freelance writer
01 May 2001

Where do developers find the guidance they need to make intelligent choices within a giddying expanse of possibilities and obstacles? Somewhere, an answer falls in the spectrum of professionals who are intimately acquainted with wireless technology, from the developer to the CEO to the analyst, and beyond. Here's what they have to say about pinning hopes to -- and leaping between -- platforms.

The developer: in the trenches
no mVivek Malhotra, director of wireless practice for Etensity, Inc., an e-business consulting firm in the hi-tech corridor of northern Virginia, points his team, reasonably enough, in a paying direction. What's the future of development to him?

"More than anything, it's what our clients are requesting," says Malhotra. And, more than anything, they are seeking to know what it is that a wireless application can do for them, and -- note this is a big 'and' -- what their return on investment will be, he says.

Then and only then does a developer ask, "I want to make this wireless application [for Client X]; what new technology is out there, and how easy is it to use?" All based on the device that best suits the client, which determines platform to a large extent. The flavor among senior execs this year is Palm and Rim Pagers, says Malhotra.

WAP-enabled cell phones are gaining ground, however. One of Etensity's recent projects was a WAP-enabled version of a B2B desktop site (on WAP-enabled cell phones) they created for a client that allows car dealerships to trade their vehicles with one another in the same area. The wireless version of the site delivered real-time information that allowed users to instantly place bids on other dealers' cars and view offers placed on their vehicles.

The company benefited so much from onsite information that they then wanted a voice application that gave users much of the same functionality as the desktop and WAP sites did, but was accessible via any regular phone. A budget shortcut to wireless voice-recognition software, true, but it worked. Among other commands, employees could enter 16-digit vehicle identification numbers by speaking them out loud, dramatically cutting down on errors.

The experience lead Malhotra to surmise, "In the future you'll see wireless and voice are very complementary technologies. Where wireless lags, voice is pretty much able to pick up the slack."

He's already assuming that further convergence, that of cHTML used in i-mode and future WML or WAP 2.0 into XHTML will dictate the next greatest decision developers need to make. "XHTML will become the standard de facto platform of choice," Malhotra predicts, despite the fact that no browser runs on it yet.

It's an exciting prospect. "Wireless development is just getting its initial kick-off, and it still has a huge amount of unused potential," Malhotra says. "The single most important piece of advice to give to someone who is starting out in wireless development, is to think outside the lines."

The CEO: the view from the mountain
Ian Cullimore, CEO and founder of Informal Software, which straddles Santa Clara, California and the green hills of England, makes a line of software that enhances how people put pen and paper to computing. He takes thinking outside the lines one step farther: Erase the lines.

To him, it's all in the human interface with the computer. Cullimore has been working on how to break down formal barriers to computing for the last ten years, hence the name of his company. He was one of the first engineers with Psion, working with the likes of future Symbian CEO, Colly Meyers. He left to form his own company, D.I.P., in 1986 and created a prototype Pocket PC. That started him on the road to a Ph.D. in Human Interface Computing (HIC).

Cullimore asks, "Why do we have to conform to the way a computer works?" An obvious example of the market leap in the human direction, he points out, is the stylus. His company makes enotate(TM) software, a real-time, natural interface for sketching and annotating on documents and images. As a user writes on a Palmtop, they can can see their writing on a PC.

"If you want to ubiquitously enable technology, you want to do it anytime, any place, anywhere, so that's how we [at Informal Software] got to wireless," Cullimore explains.

His advice to developers: "Architecture is first -- Think way, way forward, make sure you're well-suited to change, and be able to be added to completely. Abstract that to the highest level you can, [but not too high, he adds, because fundamental issues in small devices will come up]. Above all, take care of the quality of your code."

Though a fan of WAP and one who carries an Ericsson R380 in his pocket, Cullimore's products remain platform- and device-agnostic. He is convinced that such a standpoint shelters his products from technical vulnerabilities, whether they be carrier- or network-inspired. It also helps in avoiding obsolescence, of course.

"We mix our proprietary code -- sometimes you need to invent your own, sometimes you don't." Take what you can, he says, don't reinvent the wheel. TCP/IP, for example, serves as a buffer for him. "[For wireless], we don't care what kind it is -- Bluetooth or 802.11. That's your starting point."

With regard to Bluetooth, Cullimore adds, "It's small, it's light, and it will get there...we're just going through a frustrating period. It's definitely the early days [for Bluetooth]. 802.11 is more mature -- that's a good platform; but there are issues of power consumption."

Cullimore's team writes software that runs on Windows CE and Palm OS, because that's his market. "The key parts of it you can make portable and recopy, at the lowest level you have to tailor the solutions."

Because of their built-in insulation, Informal Software has no problem adapting products to wireless, development-wise. It's the market, Cullimore says, that's the problem. "If a new standard came up, we'd be happy about it. You don't want to be doing fully fledged [products] through a low data rate. That's not the way to get a good product at the end of the day."

He hopes that "wireless, son of Bluetooth, becomes really ubiquitous. It's potentially, obviously huge." A simple, effective approach is what Cullimore preaches to make good on its 'anywhere, anytime' promise. "I do believe in pragmatic solutions today, and architecting for tomorrow."

The analyst: following the action
Dan Downey is a wireless and mobile technology analyst for The Yankee Group, a strategic research and consulting firm based in Boston. His take -- and that of his ilk -- on what devices and technology bear the most promise serves to drive the direction of development at many a corporate meeting. They are the talking heads spouting figures upward of $18 billion in market demand by 2004 for wireless devices that access the Internet.

Where do they send developers? Downey says to good ol' Joe. Java code, that is. The magic cure-all to WAP interoperability problems. The WAP that was never meant to be a substitute for Web-browsing PCs.

Considering that the promising Kyocera smartphone -- a combination cell phone / PDA -- that runs on Palm OS has a price point of $500, "about five times the price people are willing to pay." Downey looks to a lower-end smartphone taking its place that would spark the ubiquity Cullimore speaks of. That's where Java code comes in.

Well, not Java code, exactly. One of the most promising developments in wireless Internet to Downey is BREW by Qualcomm, and J2ME, a mobile version by Sun Microsystems.

"Both of these technologies are going to enable people to download new technologies from the Web and have it reside in the phone," he says. Essentially, quality of service, a major issue in wireless, will no longer be a problem.

"Because connections aren't all that stable, a user can just download a game, for example, directly to their phone, and not worry about the connection," he explains. Another plus: Consumers don't have to pay for airtime, either.

"It's a big boon for developers," Downey says. J2ME's significant advantage lies in the fact that there are many Java developers out there. "J2ME is something worth paying attention to. As far as BREW goes, it works with both Java and C apps. It seems like it would give some advantages, but Sun has first-mover advantage." Other companies, IBM included, are starting to offer similar Java versions, such as J9.

The first phones with J2ME built into the chipset will be out soon from Motorola. Palm and Symbian already have an agreement with Sun to get J2ME as well. All major manufacturers will follow suit.

Now back to the market lament. "The biggest obstacle is that people don't feel like they need it. There hasn't been a demonstrated demand for wireless data," says Downey. "That has a lot to do with the fact that people are used to a wire line to the Internet. These technologies are trying to give people another way to access. And, it's going to be a lot cheaper to implement. It's a lot easier to put J2ME on a phone than to build a PDA into a phone."

Whether or not J2ME is the great white hope of wireless or not is going to depend on the applications that come out of it. "If you want to talk about killer apps, J2ME is not a killer app... This is a good start to get people to buy smart phones when they realize they will save them time and money." The magic combination.

After all, the 14 million users of DoCoMo can't be wrong, right? For the U.S. and its developers, says Downey, "The big question is, 'Is this worth paying for?'"

Coming up May 15:
The Faces of Wireless series will shine a light on the underside of wireless development -- its security and secrets -- without which no application is successful.

Resources

• The WebSphere Everyplace Suite SDK allows you to create and test wireless applications.
  
  
• Visit the IBM pervasive/wireless home page for links to wireless development solutions.
  

Rate this article This content was helpful to me: Strongly disagree (1) Disagree (2) Neutral (3) Agree (4) Strongly agree (5) Comments?

developerWorks > Wireless >