J2ME: Life after WAP?

The first incarnation of the wireless Internet – commonly known as WAP (Wireless Application Protocol) – was largely scorned by consumers. While WAP as a technology got the blame, the real culprit was the underling slow, clunky second-generation networks combined with inexperienced mobile operators that forced users into "walled gardens" of inferiorly designed applications and menu trees.

So, how will Java 2 Micro Edition help? Given the limitations on network speed and connectivity for mobile networks, the key benefit of J2ME is the opportunity to write client-side applications that run in the device independent of network connectivity and speed.

Apart from the obvious end-user benefits, the ability to run J2ME applications on cheaper 2.5 G networks also means operators will not have to spend billions of dollars to immediately create the much-hyped 3G infrastructure.

Qualcomm’s BREW is similar to J2ME technology. But whereas BREW is very focused on particular chipsets, J2ME promises the true "write once and run everywhere" ability on mobile handsets.

How far along are we with J2ME? At this point, experiences are promising. In Japan, NTT DoCoMo launched a Java implementation in some handset called "i-appli" that enables, for example, game providers to build fast and interactive games that run as a seamless part of the handset. In the United States, we already have seen the first commercially available J2ME handset from Nextel Communications Inc.

Problem with J2ME

However, the problem with J2ME today is it is good for little else than games and other consumer applications. Because the more powerful aspects of Java have been stripped away with J2ME, it becomes extremely difficult to connect to heavyweight, back-end resources, such as corporate databases, or interface with more complex aspects of mobile networks, such as messaging systems.

In effect, J2ME is in danger of becoming a mostly device-side technology, creating an unmanageable mess of disconnected islands of data and applications reminiscent of the PC. And while circumstances have turned most people into skilled "PC operators," it is unlikely that time-crunched mobile users will have similar patience.

How can we save J2ME from this danger? I found an inspirational answer to this question while watching a space adventure movie recently. In the movie, a very large and powerful starship was equipped with small, one-man ships that enabled the crew to make quick and inexpensive expeditions from the mother ship. These small ships would leverage the mother ship for any complex, resource-intensive task, such as firing a laser gun at a threatening object. In effect, the smaller ships had the power to remotely control the mother ship’s resources, giving them the virtual power of that ship, while being infinitely smaller and nimbler. These small ships would be relatively simple and inexpensive to make and maintain, yet they were as powerful as the biggest ship.

Answer to J2ME

What is the answer to J2ME? At Oracle, we believe the answer lies in an applications server that acts as the mother ship and lends its virtual power to the tiny, nimble J2ME devices. Oracle has done this with its Oracle 9i application server. We also have created a J2ME software developers’ kit so programmers can create tiny J2ME applications that use the mobile network to effect dramatic processing on the backend as seamlessly as if it were done on the device. Examples include carrier- and device-independent messaging, complex rules-based alert generation, searching through corporate databases, and even synchronization of targeted sets of data down to the tiny device for local real-time use.

I’d like to think that the possibilities enabled by this approach are as endless as the universe. Accordingly, as the power of corporate servers grows, mobile devices will be able to leverage this power seamlessly without gradually turning into mini-PCs – or space monsters.

* * *

The author is Oracle Corp.’s chief technology officer and vice president for product development of the company’s mobile products and services division, where he is responsible for product development of the wireless option of Oracle 9i application server (Oracle9i AS) and other mobile online services.