re: Java 2 Micro Edition for Handsets, Handhelds & WIDs
Java 2 Micro Edition promises new functionality for handhelds. Does this client-side software deliver the beautiful music it promises?
>> J2ME: A New Tune for Wireless
Phil Carson
Bryan Morgan
Wireless Internet
May/June 2001
Software that raises the IQ of a handheld offers a pretty catchy tune, especially when there’s a backbeat. In this case, Java software has enjoyed years of success on the server side of the equation. And it enjoys a large, devoted developer base to expand applications.
Small wonder that observers are asking whether Sun Microsystems’ Java 2 Micro Edition, or J2ME, is just another commercial jingle or a bona fide melody of lasting value.
As handsets sporting J2ME roll out in Japan and the United States and an array of carriers, vendors and content providers announce partnerships to bring J2ME-enabled services to market, Sun is strutting. Analysts, however, are watching actual deployments to define the issues. Among them: How much functionality does the software provide? At what cost? Who in the value chain will benefit and how? Where does this development fit in the evolving wireless data landscape?
Sun officials predictably refer to J2ME’s debut as a "giant leap" and point to a long list of advantages. Observers emphasize a handful: superior user interface control; the ability to work with applications off-network; device-to-device networking; increased support for graphical interfaces; and improved security and consistency of applications across platforms and devices. Over the air, or OTA, downloads of new applications should provide convenience for enterprise and instant gratification for consumers.
These advantages have found a willing audience, according to Sun.
"All the major handset vendors have committed to roll out handsets supporting this technology," says Eric Chu, group manager for marketing in Sun’s software systems group. And he points out that the industry’s de facto agreement on J2ME specifications allows diverse partnering. "That’s the beauty of this open standard: anybody can play with anybody."
Anyone Seen ROI?
The business case for implementing J2ME is simple, Chu suggests. Carriers face one or more of three basic scenarios: those with 2G networks need to offer profitable services now without further network investments; those incurring the cost of building 2.5G services need to offer a differentiator; and those laden with 3G license debt are pressed for return on investment.
Clearly, wireless carriers and the handset manufacturers who supply them both stand to profit if J2ME devices are successful. For Sun, J2ME is another product that enables the company to monetize Java. Third-party developers will benefit from J2ME through the sale of software applications and services. The initial impetus behind developer interest in J2ME is the positive experiences they’ve had with server-side Java technologies - the work of 2 million-plus developers says a lot.
Actual Rollouts
Perhaps not surprisingly, data market leader NTT DoCoMo rolled out the first J2ME-enabled handsets from Panasonic/Matsushita and Fujitsu in January in the Japanese market. By April, despite an unrelated software problem with the first 230,000 handsets from Panasonic, the Japanese carrier had sold a million units to consumers. The handsets are relatively high priced, but the added interactivity for entertainment applications apparently sells.
In the United States, Motorola’s i85 and i50sx handsets for iDEN networks from Nextel and other carriers were released in April. Initially, they offer specialized business calculators and expense report generators and PDA-like features, such as a datebook.
Beyond DoCoMo and Nextel, however, one plunges into the pool of "me-too" announcements. Sprint PCS and Cingular say they’ll develop J2ME-enabled handsets for the U.S. consumer market, and a litany of vendors and content providers have chimed in as well. Canadian-based Research In Motion Ltd. has announced an impending trial with the U.K.’s BT Cellnet for GPRS-enabled BlackBerry devices, predicated on 2.5G rollouts across the Continent. Nokia, Sony, Ericsson and Philips echo their intent. The list of carriers and handset vendors piling on is long. Content providers developing applications include Disney and Sega.
"Given J2ME’s client-side functionality, it holds the promise of making mobile data services a much easier sell," says Bryan Prohm, analyst with Gartner/Dataquest. "First-generation J2ME handsets are not a panacea for the wireless data market. They’re a niche play for early adopters."
Prohm says carriers should partner with handset manufacturers in two areas: the pursuit of mass market applications such as mobile gaming and entertainment; and targeting corporate vertical markets.
In the United States, many observers believe that Nextel is on the right track and that enterprise will be the early adopter.
"Is there a business ROI?" Keith Bigelow, director of product management at Lutris Technologies, asks rhetorically. "Absolutely." He enumerates a handful of vertical apps:
* field service (service record updates, parts ordering);
* field dispatch (UPS, FedEx);
* transportation (asset tracking of rail cars and trucks);
* inventory management (warehouse inventory);
* real estate (calling up listings while on the move);
* construction (ordering building materials to site).
Bigelow says that software solution firms such as Lutris first will try to solve existing enterprise problems for their current customer base.
Developer programs set up by carriers and vendors also will tackle enterprise apps and the more speculative work on appealing consumer apps. Interest is keen. In January, Evans Data Corp., which specializes in market research based on developers’ attitudes, surveyed 500 developers active in wireless applications. Fully 33 percent said they would focus on Java and J2ME, 25 percent selected Palm OS; 22 percent opted for Windows CE.
"What I read into that report is that far greater than 50 percent of developers are moving to J2ME because Palm and RIM are both moving to J2ME," Bigelow says. "Look one step further to what will be the default API for those devices in the future. Surprise, it’s Java."
At last year’s JavaOne Conference, developers glimpsed next-gen devices enabled by J2ME that included phones with games, pagers running SOHO apps and PDAs fielding remote database needs. Undoubtedly, options at the 2001 JavaOne conference in San Francisco on June 4-8 will be even broader, with a rosy tint.
The Hurdles
Barney Dewey, a partner in the Andy Seybold Group, agrees that an improved user interface, the ability to work offline and OTA provisioning of new applications bode well for J2ME.
"J2ME is very interesting for the higher-end phones with decent-sized displays," Dewey says. "And there aren’t a lot of those in the United States yet. The Motorola-Nextel units aren’t bad, but bigger displays, like those in Japan, would be even better."
The analyst suggests that with the "right" handsets - those with large displays, easy navigation and sufficient memory - enterprise applications may proliferate. If memory is an issue, he suggests, vendors might provide slots for plug-in memory so the user determines the investment in that typically costly element. He predicts that phones with Palm OS for data-centric uses will most successfully leverage advantages
of J2ME.
Price point could be a hurdle in Dewey’s view. "The downside for J2ME is the higher cost for the device in a market driven by low-cost devices for the masses. That’s why uptake will come in enterprise first. You’re not going to get teenagers plunking down hundreds of dollars for a phone."
The biggest pitfall, he suggests, could be the same mistake WAP encountered in Europe and the United States: Operators failed to specify a minimum screen size and graphics to handset manufacturers. "If operators don’t learn this lesson and just let vendors build whatever they want to build, they’re going to have the same kinds of problems," he concludes. "That’s the big risk."
Sun’s Chu responds that the DoCoMo experience this spring shows price point is not a hurdle to consumer adoption. On the memory issue, WAP requires a half-megabyte to function; J2ME only half that. As for memory required for the application itself, on the consumer side, full color, interactive apps require less than 10 megabytes apiece and no one in Japan has complained, Chu adds. As for enterprise apps that might require more, that’s not a technological issue, he suggests. "If you are smart in client-server computing and only fetch the information that’s relevant, that shouldn’t be an issue." Prudence from a security standpoint points to small database downloads, anyway, he says. Though Cahners In-Stat Group’s research on J2ME uptake by early adopters points to concerns about the trade-off between increased functionality and battery life (see survey, pages 32-33), Chu again points to the early DoCoMo experience; Sun has heard no complaints on the issue.
On the Horizon
One issue critical to the success of J2ME is the deployment of bug-free virtual machines to a wide variety of mobile devices. Sun is in the precarious position familiar to parents of teenagers: They’ve designed a solid technology, explained to vendors how to successfully implement it, and now must watch as vendors take their first steps into the real world.
Despite Sun’s high hopes, some developers wonder if such an ambitious technology will work flawlessly across the dozens of device hardware architectures and operating systems currently in use. A single, major implementation snarl by one vendor could damage developer confidence in the technology, hurting all who support it.
Down the Road
"We think of Java as ‘G-independent,’" Chu says of the future. "This industry is too technology-centric, always talking about 2G, 2.5G, 3G. In reality, look at the marketplace and the health of the telecom industry in terms of what’s available. We’re focusing on helping them make money today with what they have today - rather than tell them they have to spend hundreds of millions of dollars before they can
make more money."
The key to successfully selling these capabilities, however, will lie in convincing end-users that applications residing on the device can somehow improve their quality of life or their ability to conduct business while away from the office.
"It’ll be interesting to see how this all plays out," Bigelow says. "Towards the end of the year, when you have general availability from every carrier, it’ll be easier to judge adoption rates."
Bigelow is looking to see how larger devices that support J2ME work out. "When Palm actively endorses J2ME," he says, "that’ll be a big deal."
For his part, Prohm sees the software as an accelerant. "Everyone is trying to start a mobile data fire," he says. "Chances are it won’t be any one thing, but a well-planned, well-executed aggregate of many factors that appeal to diverse user segments. And anything you do to enhance handset functionality is a selling point for carriers."
If carrier sales are the point, it shouldn’t matter whether J2ME is melodic or cloying, as long as it’s a catchy tune. <<
- Eric - |
