mo%lbl-csam@sri-unix.UUCP (08/06/84)
From: (Mike O'Dell[x-csam]) mo@lbl-csam I believe the Evans and Sutherland Knob Box was the first place I saw using the acceleration to select different bits of the counters to clock. I have long believed this behaviour is desirable. Another good place is in radio receivers with synthesized front-ends: you need 10 Hz resolution for some things, but having to push buttons is a pain to change speeds. Simply using the rate to determine the rate would make it soo much easier. To go fast, go fast. To go precisely, move slowly. Sounds like the way people work already! How come it doesn't seem to be more widely implemented? (I hav heard the Symbolics does it.) -Mike
eric@apollo.UUCP (08/09/84)
> To go fast, go fast. To go precisely, move slowly. Sounds like the way > people work already! How come it doesn't seem to be more widely implemented? Apollo certainly does it this way, and has for years. It works great. Eric Peters (...decvax!wivax!apollo!eric) Apollo Computer Inc., Chelmsford, MA 01824
ksbszabo@wateng.UUCP (Kevin Szabo) (08/12/84)
Could some people mail their favourite mouse 'gear ratio', to me? ( linear or non-linear ). I tried to implement a non-linear mousing transform ( dx * (dx+1)/2 ) but it didn't work as well as I hoped and thus I abandoned it. This implementation was for a general-purpose input package under UN*X. Running under UN*X introduced the problem of being swapped out now and then. The result was that the non-linear stuff would see large movements when you were swapped (this would occur for tablet sampling, or when the user briskly moved the mouse because of poor response) and the cursor would shoot to some side of the screen. Any comments appreciated, and of course I will summarize for the net. Kevin -- Kevin Szabo watmath!wateng!ksbszabo (Elec Eng, U of Waterloo)