Editing Pointing stick (section)

== Design challenges ==
The IBM TrackPoint III and the TrackPoint IV have a feature called Negative Inertia that causes the pointer's velocity to "overreact" when it is accelerated or decelerated. Negative Inertia is intended to avoid the feeling of inertia or sluggishness when starting or stopping movement.<ref>{{cite patent |country=US |number=5570111 |title=Graphical user interface cursor positioning device having a negative inertia ... |fdate=1994-10-03 |gdate=1996-10-29 |invent1=Robert C. Barrett |invent2=Robert S. Olyha, Jr. |invent3=Joseph D. Rutledge}}</ref> Usability tests at IBM have shown that it is easier for users to position the pointer with Negative Inertia, and performance is 7.8% better.<ref>{{cite journal |url=http://www.sigchi.org/chi95/proceedings/shortppr/rcb_bdy.htm |title=Negative Inertia: A Dynamic Pointing Function |publisher=[[SIGCHI]] |author1=R. C. Barrett |author2=E. J. Selker |author3=J. D. Rutledge* |author4=R. S. Olyha* |year=1995 |access-date=2012-12-03 |archive-url=https://web.archive.org/web/20170220093752/https://sigchi.org/chi95/proceedings/shortppr/rcb_bdy.htm |archive-date=2017-02-20 |url-status=dead}}</ref>

Another challenge with pointing stick design is identification of the zero position (the position where no motion is desired). Because the amount of motion is small, the sensitivity of the sensors must be high, and they are subject to noise interference.

A typical solution, which assumes that pointing sticks frequently go out of calibration, is to interpret a variation below a certain threshold (over a given interval, perhaps one or several seconds) as being a neutral stick. However, the recalibration can also allow brief periods of 'drifting' (movement of the pointer while the user is not moving the pointing stick).<ref>{{Cite web | url=https://www.reddit.com/r/thinkpad/comments/2hvpzj/track_point_drift/ | title=R/Thinkpad - Track point drift| date=30 September 2014}}</ref>

In practice, if the re-calibration interval is set too short and if the user applies moderately consistent pressure to the stick for such an interval, this method results in an incorrect zero point. Additional pressure again moves the pointer, but the calibration may occur again, requiring even more force. If the user releases pressure at this point, the change will be interpreted as an instruction to move the opposite direction. In time, the software will re-calibrate and stop the motion.

Additionally, if "press-to-select" is enabled, the software may generate unexpected click events by touching the pointing stick during typing.