Wireless Looms Large

While wireless awaits the killer app that will make it a solid business proposition for the enterprise, here’s what to watch.

Heard this one before? "Wireless will play a huge role in the future of IT." Since that's a given, what's holding back wireless development? The fact is, the more wireless matures, the more promise it offers. But corporate wireless adoption is facing a classic chicken-and-egg quandary.

It's undoubtedly a tremendously useful technology waiting for the killer applications that will make its adoption a no-brainer business decision for large enterprises. But killer apps have a hard time appearing until more large enterprises adopt wireless technologies. In the absence of concrete IT demand, wireless hardware and software developers are left looking for a problem to solve.

Uses for wireless in the enterprise are still up for grabs. Advocates for wireless technology fall into the "it's already here" camp (mostly around the 802.11b standard) and the "it's coming and will surely have a huge impact" camp (pretty much everything else).

Handheld communications devices, probably a variety of telephony/PDA/laptop amalgamations, are building on the massive popularity of the mobile phone. The next iterations of these devices will have a profound impact on how we conduct our business and our lives. Which technologies will drive them?

First, there's 3G wireless, which is likely to provide the kind of pipe that robust mobile applications will need. 3G is a sort of catch-all term, incorporating a couple of actual technical specifications that range from 307kbps to 4.8mbps of transfer speed, at least here in the U.S. In Europe, 3G will offer similar performance but be based on a different underlying technical specification (GSM versus CDMA), so speeds won't be exactly the same. Already deployed in Japan, 3G technologies won't see widespread availability in the U.S. for at least a year or two.

The big news with 3G is bandwidth—driving more and more bits down to mobile devices. We're now talking about far more bandwidth for mobile devices than typical office computers had just a few years ago.

What nobody seems sure of is what we're going to do with all those bits. Mobile computing platforms are clearly more flexible. Every business traveler can appreciate the difference that just having a laptop in the hotel room can make.

The problem lies in the specifics. Nobody is likely to compose lengthy documents on a cell phone-size device. Between the input and display problems, it's just not feasible (oh, we could get off into the promised land of perfect voice recognition and direct optical nerve stimulation, but this is an emerging tech column, not a science fiction story).

So what will users of 3G handsets need? If Motorola, Nokia and Ericsson are going to be building the devices, they'll need some idea of what applications users will want, right?

Wireless LAN
Microprocessor Sales

2000: $252 million
2001: $240 million
2006: $890 million

Source: Aberdeen Group

Natural Stratification
Here's where the natural stratification of a maturing industry kicks in. When IBM built the first PCs, they didn't worry too much about writing every possible application. They relied on outside developers to deal with the whole application issue. IBM released a next-to-unusable word processor (DisplayWrite), which showed that the machine was a capable word-processing platform. More importantly, it spurred several others to write competing word processors because clearly anything would be an improvement on DisplayWrite.

Intentionally or not, 3G companies look to be relying on that same model: Build the hardware, offer adequate software and look to third parties and end users to provide the killer apps.

Growth in Wireless
E-business

2000: $75 billion
2001: $110 billion

Source: Gartner Inc.

But now there's another issue—if 3G wireless devices are more of a platform than a turnkey remote terminal, how do you develop for the platform? What, exactly, is the platform?

How about Microsoft versus Java, for starters? There are zealots on both sides, of course, and both company's technologies have their ups and downs. For Microsoft, this is a chance to lock up edge applications: If a user has PocketAccess on a remote terminal, it's a pretty good bet you can leverage that user into Access on the desktop, and pressure the IT staff to support SQL Server on the back-end.

Java's Shot at Corporate Computing
For Java proponents, it's yet another chance—perhaps the last one—to demonstrate real and deep relevance to corporate computing. Don't get me wrong—Java is great stuff, a fantastic idea that could've gone a lot further if it didn't have a boat anchor called Sun dragging it down each step of the way. However, with the exception of specialized applications, there really isn't anything yet that makes Java business-critical end-to-end. That's as opposed to Microsoft, where everything from Word to Internet Explorer to SQL Server is well entrenched.

Wireless Tools for the Enterprise
CleverPath Portal
Computer Associates International Inc.
(800) 225-5224
www.cai.com
and
Blue Moon
AirTuit Inc.
(865) 673-9600
www.airtuit.com
Development and portal environments for content delivery to wireless devices.

Unicenter TNG (ShipIT, AimIT, eTrust)
Computer Associates International Inc.
(800) 225-5224
www.cai.com
Modules for PalmPilot management for CA's enterprise management suite.

SNAPP
IBM Corp.
(888) 746-7426
www.alphaworks.ibm.com/tech/ snapp
Allows administrators to monitor and manage RS/6000 systems (AIX and Linux) via PDA. Free download.

ZENworks for Handhelds
Novell Corp.
(888) 321-4272
www.novell.com
and
Afaria Mobile Infrastructure Tools
XcelleNet Inc.
(800) 322-3366
www.xcellenet.com
Lets IT managers monitor, control and synchronize PDAs and other wireless devices in use on a network.

What the Java camp has come up with is J2ME—Java 2 Micro Edition. J2ME is, as you'd expect from the Java crowd, a very well-thought-out set of APIs for embedded applications. As you'd also expect from the Java crowd, there are already two versions ("configurations") of J2ME; one aimed at devices with less than 512K of RAM, and one aimed at devices with more than 512K of RAM.

J2ME is an interesting construct—it's a platform API with no intrinsic user interface functionality. Instead, it relies on multiple "profiles," each of which offers the UI functionality needed on a specific device. What you've got there is a very flexible, adaptable platform. But is it too over-architected to achieve widespread commercial success? What happens if or when mobile devices converge on a single UI model? Like the Java we know and love, J2ME offers compatibility and flexibility, but maybe at the expense of performance and specialization.

Like the Java offering, Microsoft's approach is true to form: Develop software standards that are intimately tied to hardware, and applications that are intimately tied to the underlying platform, bundle them all together and link them to Microsoft applications at the home base. The Microsoft evangelist says, "That way everything will have been developed and tested together and will offer high performance." The anti-Microsoft camp, of course, isn't so thrilled.

Microsoft already had a mobile computing platform: the PocketPC (originally and unfortunately named WinCE—ouch). Tack on a few APIs here, trim a few there, and PocketPC becomes a remote terminal platform. The fact is, 3G will probably be a hit no matter which platform wins the battle (or in the unlikely event of a draw). My advice is not to be caught flat-footed. These are technologies that it will pay to have advance knowledge of. When the CXO brings in a clipping from some airline magazine on wireless, you want to be prepared with a strategy. Besides, how often do you get to expense cool, first-generation products like these?