X

Commentary: Java has no future on cell phones

The software works on back-end servers, not on cell phones and today's personal digital assistants, which lack the processing power and memory requirements to handle Java's large overhead.

3 min read

Sun Microsystems has announced it can now put Java software into cell phones. This technology, however, will not become an important force in mobile devices.

The dream of many companies, such as Sun and Microsoft, is that a single technology--each one has its own--will emerge as the common language of application delivery.

To maintain this dream, every company has to describe its technology in universal terms that indicate it can be used ubiquitously. Sun's newly announced mobile information device profile (MIDP) "profile specification" is essentially Sun's attempt to project Java onto mobile devices.

See news story:
Sun serves up Java on mobile phones

No one should believe it. Java belongs on back-end servers, not on cell phones and today's personal digital assistants (PDAs), which lack the processing power and memory requirements to handle Java's large overhead. Even if someone built a cell phone with a 200-MHz processor, cellular phones do not have the bandwidth to download huge Java applets.

Cell phones also lack even minimal screen size for displaying the graphics, animation and multimedia presentations that Java is designed to handle. Some cell phone manufacturers have experimented with heads-up displays and other high-tech answers to the tiny-screen problem, but these technologies remain too expensive to be practical.

Entering information is another basic problem with the cell phone, which must depend on its keypad--a clumsy way to enter anything more than a few letters at a time. The cell phone is a specialized device, and companies developing services for cell phones must design those services to its limitations. There is no silver bullet that will make cell phones look like PCs. Certainly, Java cannot do that.

It is true that cell phone limitations are shifting. Within two years, we expect most cell phones will incorporate Bluetooth short-range wireless technology. We also expect increasingly high-powered PDAs to aggressively take on the market. Compaq's newest PDAs sport a 200-MHz StrongARM processor, for example.

Bluetooth would enable cell phones to act as wireless modems for PDAs and laptops, allowing companies to deliver advanced wireless services. However, the bandwidth restrictions of cellular technology will not change soon, and this will remain a gating factor on downloading Java applets and other large files to wireless devices of any kind.

Ultimately, although some wireless manufacturers may attempt to implement MIDP, those attempts will be doomed to failure. Just because Nokia, Ericsson and others have said they support MIDP does not mean that they will actually use it.

Meanwhile, several other technologies already are working effectively on cell phones and wireless PDAs. These technologies include wireless application protocol, AvantGo and similar Web clipping services, as well as services developed by several of the large financial companies.

Whether Java will eventually supplant them remains to be seen.

Companies developing services for cell phones or wireless PDAs need to design to the limits of these specialized devices, including their display size, input capabilities, processing power, memory and connection speed. Even the popular Palm PDA lacks the resources to run Java applets at a reasonable speed. And although Windows CE devices are starting to come with the capabilities of running more advanced services, they are not incorporating Java in their technologies.

Ultimately, the low bandwidth of cellular technology will limit most to alphanumeric information with little multimedia content. Java should be treated as a tool and should be used where appropriate. At this juncture, Java is appropriate on servers, not on wireless phones.

Sun's latest Java marketing pronouncements should just be ignored.

Meta Group analysts Dale Kutnick, Peter Burris, David Cearley, Val Sribar, William Zachmann, Jack Gold and Jean-Louis Previdi contributed to this report.

Entire contents, Copyright © 2000 Meta Group, Inc. All rights reserved.