Sunday, August 12, 2007
Disregarding handwriting, counting on your fingers and even typewriters as being just methods of indicating, the first method of selecting information for use or further processing is the abacus. This is a direct selection method. The selector and the indicator are joined, or in this case, the actual same object (the bead). The slide rule, to include various manual rotary computers, are of course very similar. There is no further removal of the user from the interaction than by moving the relationship bars or grabbing the edge of the indicator slider. The magnetic compass similarly has a manual dial which allows establishing relationships of position and azimuth, and when coupled with constantly updating magnetic information allows monitoring of bearing and heading during travel. Such interactivity continued thru to relatively technical, computerized systems such as the 1960s-era flight simulator control panel shown here. The radio frequency selector (a similar method was used on aircraft of the era) is a single dial for megacycles (the 100s, 10s and 1s place from a limited list) and a dial below it for the tenths place. An arguably similar method is most early adding machines (and some early cash registers). In the pre 10-key days the were of the direct-entry type. A typical layout is an entire row of all 10 possible digits for each place in the total number. Selecting a value per place leaves the key depressed, serving as an indicator. Eventually, registers (lists of the numbers selected) started appearing. This led to the 10-key pad allowing an indicator separate from a selector. The adding machine evolved into the current form with a single set of entry buttons, and a register of entered values (or a printer). Really all 10-key devices (even aviation radios these days) use this model. The desktop computer is more or less the ultimate extension of this, with the keyboard and mouse often not even attached to the display device. And now, for some time really, there seems to be a push to move back to directly connecting the input and display, or action. One good example is that bastion of selectors and indicators, the light-up elevator button for each floor. Which is being reconsidered as a 10-key system, with its predictably unpredictable results. Touch screens are the most obvious and dynamic of these though, from Cintiq-like products, to of course mobile devices. I have seen two things that bug me about touch screens. One is that the electronic tying of the functions is rather tenuous; display-thickness induced parallax and processing delays means the pointer is near where you are indicating, not at the tip of the stylus (or finger). The other problem is the tendency of practically everyone to forget their history, so any good designs (or bad) from the past will not be applied. As I've touched on before, the iPhone has this issue in several regards. See a video of how the user's finger (okay, its me) covers the indicator/selector, so you have to try, then move out of the way to make sure it worked, or fix it. For all my whining, a solution might emerge over time. Eventually some good design standards will be developed for these products, mergers and product failures will cause consolidation and with any luck the good ideas will rise to the top, relatively universally.