Yak-1 Hardware Usability Assessment

by Jani Patokallio
$Id: hw-usability.html,v 1.1 2000/10/25 10:30:42 jani Exp $
$Revision: 1.1 $

Hardware used

Block diagram of the Yak-1 hardware

Yak-1 is a prototype: hardware usability was sacrificed to allow it to be built rapidly with off-the-shelf non-wearable parts to serve as a fully functional testbed for the software. And it shows... see the hardware construction and software installation guide for full details about parts used and how they were fitted together.

Activating the Yak-1

Quite possibly the most annoying thing about the Yak-1 is the difficulty of starting the thing. Turning on the system requires the following sequence: Needless to say, this is extremely cumbersome: the only required action should be pressing the POWER key, and even that should be accessible from the controller. (Yes, I am sorely tempted to rip into the VAIO's internals and hotwire a new switch...)

One niggly little problem is where to place the Glasstron while hoisting the backpack, as this operation requires two hands. Hooking it into your shirt or a large pocket and being careful with the cable seems to be the only solution. (Yes, you can put the Glasstron on first, but since the glasses are attached to the backpack with a cable you'd better be extra careful...) A wireless connection between the display and the computer would be very, very nice (maybe Bluetooth?), a (poor) second best would be computer inside a clip-on belt.

Using the Yak-1

Once booted though, from a purely hardware point of view the usability of the system is quite tolerable.


The Yak-1 was carried around, mostly crushed at the bottom of an overstuffed backpack, for almost three weeks across some 3000 kilometers of northern Japan, through campgrounds, youth hostels, forests, rivers, swamps, rain, shine... and it survived entirely intact. So yes, in this respect it is definitely portable.

Battery life

With the internal display turned off the VAIO's lithium battery (size S) lasts at least 3 hours even under heavy load and the Glasstron's battery (infoLithium L) is at least as good, although the Glasstron will switch itself off automatically after 1.5 hours of continuous use. The major failure point is the scan converter, which requires non-rechargable 9V batteries and eats them at an alarming speed of one per hour. Once the scan converter battery starts to fail, first the display starts to flicker, then color information starts to warp until only black and white is left, and finally the video data itself disappears.

Scan converter aside, if the system could be turned on and off easily and recharged once a day, battery life would be no problem at all.

Input interfaces

Currently only mouse input is provided, which effectively limits usable software to Yakkey and intellectual pursuits like Solitaire and Mahjong. Two alternative mouse input devices were considered.

Victor Handy Mouse

It may be called a mouse, but the Handy Mouse is really a miniature joystick. Moving directly up/down/left/right is easy to do even with no training, moving diagonally and controlling acceleration is more difficult. A "real" mouse provides a wide range of possible pointer speeds, the Handy Mouse seems almost analog at times with only two possible states, "moving" and "not moving".

LuckyTech FinRing

The FinRing has a rather steep learning curve and its use seems quite unintuitive at first. The main problem is that the device recalibrates its directions with each press of the "move" button, so acceleration is determined by relative movement, not absolute movement. This is different from the traditional mouse/joystick paradigm and quite difficult to learn.

An example: The user wants to move the pointer up, select a menu item, then move the pointer down. To move up, the user holds the move button and tilts his wrist up until the pointer reaches the desired position, then lets go of the move button (the pointer stops) and presses left to select. Now, to move down, the user presses the move button again and tilts his wrist down -- but the FinRing just recalibrated itself so that "up" is "level", and a motion down now sends the pointer shooting to the bottom of the screen until the user tilts his wrist back up!

The FinRing is also difficult to move in a controlled diagonal, and while motion left-right is relatively easy (just turn the wrist), up-down motion is quite cumbersome and prone to sudden bursts of acceleration and drifting in the wrong direction. The move button has to be continuously pressed quite hard to transmit motion, letting go even for a split second will result in recalibration and much confusion. All in all, in terms of usability the FinRing is much worse than the Handy Mouse.

For wearable use, the FinRing's primary advantage is its wirelessness, it functions quite well when the receiver is in the backpack with the laptop. Unfortunately, from a psychological point of the view, any benefit from one cable less is outweighed by the bizarre sight of somebody rotating his wrist in all directions while trying to get the pointer to move in the desired direction.

Turning off the Yak-1

Turning off the Yak-1 involves the inverse of the cumbersome activation process, although the VAIO can at least be turned off with software by issuing a halt. apm -s can be used to put the laptop in standby mode, but waking up is not possible without opening the case. apm -S for suspend mode does not work. BIOS hibernation (which would otherwise be the best way to conserve power but allow rapid access on demand) works but cannot be triggered by Linux APM, a manual Fn-F12 is required, which (again) requires opening the case.


The following criteria need to be noted when designing the Yak-2: