Given that, we've found that the PalmPilot is a strong tool for field applications, and that it's getting stronger. As it exists today, the PalmPilot is portable (Palm says "unconsciously portable"Ñmore on that later), programmable, and integrates well with back-end data systems. Symbol Technologies has worked with Palm to develop ruggedized systems better able to withstand hostile environments, and units that are integrated with scanners, allowing easier data collection in some applications. Clearly, field data collection is emerging as an important direction for the PalmPilot platform.

Having said that, though, you also have to consider the platform's limitations before developing your application. As we all know, the device doesn't appreciate being dropped (though mine has survived more than its share of abuse). Also, the standard development environment is oriented toward trained developers familiar with the C language, which can be difficult to learn. Further, Graffiti itself can be an impediment, especially when your application is aimed at a user community that has neither the time nor the inclination to train itself on pen-based input. In addition, there is only limited ability to attach additional equipment to the PalmPilot. Finally, there are some special issues such as alarm volume and screen readability that affect some applications. Although this may seem like a long list, you can usually address these issues during application definition and development.

As a specific example of PalmPilot-based field data collection, next month we'll talk about an application we just finished developing for the Smoking Research Group of the University of Pittsburgh.

Steve Niles is Senior Editor for ZATZ Publishing. He is also an independent filmmaker and freelance writer. For more information, visit http://www.sn-films.com.