Re:Hysteresis (1997)
by DeathArrow on Saturday July 09, 2022 @06:32AM
(#32033681)
This is a delay or added latency, not hysteresis.
[ Parent
| on HN ]
Re:Hysteresis (1997)
by overlisted on Saturday July 09, 2022 @07:38AM
(#32034073)
It's neither of those terms. I'm not sure if anyone knows this word outside of web dev, but I would call that "debouncing", because all that effect does is ignore input states that haven't stuck for long enough. There's no latency, no lag and no "behind". Only a threshold.
Re:Hysteresis (1997)
by spacemanmatt on Saturday July 09, 2022 @08:35AM
(#32034459)
Debouncing is exactly hysteresis. We often implement it with a very poor digital method but the analog inspiration is a capacitor that must fill before current flows through a closed switch, effecting hysteresis.
Debouncing using capacitor+schmitt trigger(which even has hysteresis response graph drawn on top) is no doubts a hysteresis. But I wouldn't call all debouncing strategies a hysteresis. One (easy to implement in code) debouncing strategy is lockout - after the first input transition perform the output change immediately and then stop reading any new inputs for some time. This has the benefit of having no delay, but downside is that it amplifies random noise caused by other factors when the switch isn't moving.
Re:Hysteresis (1997)
by elromulous on Saturday July 09, 2022 @12:29PM
(#32036665)
Fyi debouncing predates web development by many decades E.g. your keyboard almost certainly has debounce logic implemented on the microcontroller in it. And long before digital circuits existed, it was implemented with capacitors.
To me, it seems consistent with the definition/origin of the word, ie lagging behind
https://en.m.wikipedia.org/wiki/Hysteresis#Etymology_and_his...
Maybe you are deriving your sense of the word from examples in physics and engineering where hysteresis is referred to that do not look quite like the situation described in UIs here?