Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!ns-mx!pyrite.cs.uiowa.edu From: jones@pyrite.cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: comp.human-factors Subject: Knob on a keyboard? Message-ID: <6449@ns-mx.uiowa.edu> Date: 12 Jun 91 17:08:18 GMT References: <2909@sumax.seattleu.edu> Sender: news@ns-mx.uiowa.edu Lines: 35 The discussion of track balls leads me to remember the keyboard on my old HP 9836 (RIP). That machine had a knob on the keyboard, easily reached by the left pinky without a great leap from the home row. The knob was a circle about 3 cm in diameter with a raised edge, set flush with the keyboard near the upper left corner. It would spin freely in either direction with only a slight amount of friction, and it gave input to the computer with a rotary shaft encoder of some kind. The knob moved the cursor in two dimensions, with the shift key selecting the direction. If I remember correctly, un-shifted knob rotation moved the cursor horizontally, shift-knob was vertical. Prior to the domination of HP by UNIX, HP had a screen editor that was tightly integrated with the knob, and I still hold that that editor was one of the best I've encountered from the point of view of ease of learning without a manual. From my experience using the knob, it was better than a mouse for text editing because it was constrained to move in only one dimension at a time. You'd home in on the right line, then home in on a character, without the awkward problems mice cause with slipping up or down a line as you seek a character on a particular line. In addition, the knob let me switch between typing and cursor positioning much faster than a mouse simply because I didn't have to move my hands as far from the home row. For coarse graphics, I'd vote in favor of a mouse, but for fine graphics, where you're trying to get precise alignment of items on the screen, I think the knob was again a winning technology. I conclude that I'd like something like HP's knob on my keyboard as well as a mouse off to the side, and for most purposes, I'd like them to be completely equivalent, so that motion of either would move the cursor. Doug Jones jones@cs.uiowa.edu