People often have a fondness for relics, no matter how old and impractical they may be. The standard QWERTY keyboard was invented more than 125 years ago by Christopher Shole. At that time, jamming among the hammers of a typewriter was common. Therefore, Shole created a layout that made typing as difficult as possible. This slowed down the typist and reduced jamming. Thus, the inefficient QWERTY design was actually designed to slow down user input. Yet, the public still cling to it as a standard.

This standard is problematic to PDAs. They cannot have the same size keys as those of a typewriter or standard computer. Thus, their keyboards are reduced to sets of minuscule buttons crowded tightly together. This is evident in the PC110. This design adversely affects the userís input speed and comfort.

...

The other popular option is pen input. Apple bases its Newton 130 and 2000 on pen input and handwriting recognition. The company believes that a notepad metaphor is perfect for a everyday product like a personal digital assistant. Apple felt writing is more natural than typing and can more readily accepted by consumers. However, this belief has four main drawbacks.

1) Each individual user has very unique writing style and look. An advanced recognition architecture is required for accurate interpretation of inputted text. Apple only learned this after being thoroughly embarrassed by the poor recognition software in the original Newton.

2) An extensive recognition software and large vocabulary lists translate into slow interpretation of input. Users cannot afford to wait several seconds for each separate word to be processed and outputted.

3) The small form factor of PDAs makes for limited writing space. Users cannot write out long sentences as they would on regular paper. Instead, they can only input each word, wait for it to be processed, and then write the next word. This is both awkward and inefficient.

4) The public is not used to inputting information through writing. Even if developers solved the preceding obstacles, many people will still be drawn to their more familiar QWERTY keyboard. This mainstay product will be hard to displace no matter how viable alternative technologies are.

Many other companies have devised alternative solutions to Appleís handwriting recognition architecture. The most popular and interesting entry is Palm Computingís Graffiti. This software tackles three of Newtonís main problems with unique tactics.

1) Instead of having the PDA learn how the users write, users must learn how to write to the PDA. Graffiti uses a shorthand alphabet that simplifies each letter and makes them more distinguishable from one another. After some practice, users can quickly become accustomed to the new writing style.

2) Since each letter is easily identified, recognition is much quicker.

3) Users input text one letter at a time. Thus, there is no need to provide space to write sentences or even complete words.

Graffitiís speed and small space requirements have made it ideal for small PDAs such as the Pilot. However, its strange alphabet and peculiar letter-by-letter input make many customers grasp even more tightly to their well-known QWERTY keyboards.

Another trend for PDAs is to mix and match various input systems. For example, an optional keyboard can be purchased for the Newton. And, the Cassiopeia mainly relies on keyboard input, but is open to third party handwriting recognition software. However, this practice is not a long-term answer to the input problem. Rather, it is just combining two poor solutions into a single package.

...

The future of PDA input lies in speech recognition. The easiest form of human information transfer is speaking. Talking is even more natural than writing. But, speech recognition comes with its own set of problems. For starters, spoken language is more technologically difficult to interpret than writing. And, how would PDA identify its ownerís voice if the he is running the PDA in a room full of other PDA users? How will the public initially react to such a radical approach? Will early developers be able to reap enough economic benefit to have the resources to continue improving the technology? These questions show that the input problem is not an easily solvable one.