It's flexible, and there are LEDs and linear optical sensors set up with an occluder in between that casts a shadow (or passes a light through a slit, according to the patent). The sensor detect where along the sensor the shadow or light falls, which indicates how much you have translated or rotated the control in a particular axis.
I don't know exactly, but using one, it has the sensation of using the trackpoint on a Thinkpad. The puck you push, pull, and twist to make the motion is a bit springiness. The harder to push it, the faster the motion happens in the software on screen. When you release, it springs back to neutral. The old ones (late 90's vintage) had trouble coming back to center, and there was a hot key to rezero it if it was starting to drift. The newer ones don't seem to have the issue.
It’s a knob that spring back to its origin when let go, that can move ~45 degrees or ~1/4” each axes(subjectively). It feels a bit like those old bobbing dashboard toy in hand.
Mechanically, it’s something like an Rx/Rz two-axis joystick with sliding knob part on top that can be slid in X/Y, pushed/pulled in Z and twisted in Ry as well. All axes are spring centered.
Imagine a 3-axis joystick that can also sense translation. So aside from pitch/roll/yaw you also get the linear up-down/left-rigth/forward-back axises.
I can see how pitch roll yaw would work, but I can't imagine how linear movements are made on the same joystick? Maybe up and down is push and pull but what about the other directions? Do you push the device itself?
The trick in comparing it to a joystick is that it can distinguish between tilting the control to a side, and pushing the control to the side. Does that help? Similarly, it can detect pushing straight down and pulling straight up.
Looks like current incarnation is Stewart platform lookalike optical setup for all axes[1]. I misspoke in a different comment: Mine was older SpaceBall 5000 model, and it was more like two joysticks joined at the stem.
There is an opensource driver package for Unix type systems[1]. I find them to be pretty reliable. The stock Windows drivers do leave something to be desired.
I've always found it fascinating that companies that makes specialized hardware devices like this lean so hard into proprietary software, as if that's the thing that gives them the edge on the competition.
If anything the hardware is the hard part to copy and open source drivers could do nothing but benefit them and ensure wide platform compatibility and longevity, and yet so many companies like this insist on closing things up and forcing hackers to reverse engineer everything. It's bizarre logic to me.
Would it help with scripting macro software such as AutoHotKey? I imagine it wouldn't pick up some inputs due to proprietary driver.
It have the same issue for my Logitech G600 that it can't pick up some inputs. I set up the macros in LGHUB with keystrokes (multiple modifiers with keys) and use AHK to grab those and expands the capabilities of my G600 than what LGHUB offers.
Could it help if the driver have those axes assigned to a different key to allow AHK to pick that up and AHK can change it as axes? I done that with my XP-PEN AC-19, I assigned the dial wheel to different keys in XP-PEN software and AutoHotInterception with AHK to pick that up. Then assigned it with wheelup/down depending on what active window I am on.
It does not expose it as a key either. You must link their proprietary driver (and sign some legal document) that then gives you an API you can use to read the position of the spaceball.
