If those input devices are so good, why don't we see people winning e-sports tournaments with them? The mouse (as in a high-end gaming mouse on a high-end gaming mouse pad) is the best positioning device yet invented because it allows very precise positioning with very small movements of the hand. And unlike a pen, you're not constantly fighting gravity.

My arm would get tired quickly if I tried to point to things by pen as fast as I can by mouse.

The classic paper about this topic is:

Buxton, W. (1983). Lexical and Pragmatic Considerations of Input Structures. Computer Graphics, 17 (1), 31-37.

https://www.billbuxton.com/lexical.html

TL;DR: "It's complicated."

TL;DR-but-not-illiterate:

Claim: How you spell a command (lexical) and how you move your body to do it (pragmatic -- device, gesture, layout) are different problems. Treating all "locators" as the same -- mouse, tablet, light pen, joystick -- is a design error.

Figure 1: Input devices sit in a grid by what they sense (position / motion / pressure) and how your hand moves. Light pens cluster with touch screens; mice with tablets; joysticks with trackballs. Different motor skills, not interchangeable.

Why mrob is wrong here: "Graphics need a mouse" assumes one cheap locator fits every graphics task. Buxton's point (and skydhash's): viewport pan/zoom wants isometric/spacemouse control; tracing wants a pen. A mouse is a compromise, not the default.

What PIXIE has to do with it: I already showed the right transducer: light pen on a vector display for tracing and radial lightbuttons. No 1000 Hz, no mouse. The 1969 demo is a pragmatic match; Buxton (1983) is the theory that says that match matters.

Buxton defines the terms Chunking: Good interfaces fuse actions into one gesture (pick up => drag => release). Pie menus and pen tracking are in that family -- syntax, tension, and closure enforced by the body, not remembered as steps.

Quieter computing isn't only slower phosphor or fewer Hz. It's also picking the right input device for the task instead of defaulting to mouse + glass.

Miyamoto closes the loop from the other end. He designs from the hands: start with the kinesthetics of the controller, how the thing feels before what it does.

Will Wright on Miyamoto: "Miyamoto starts from the kinesthetics of the controller. What is this thing going to feel like in my hands? Will I feel like I'm instinctively connected to this world? As opposed to, I've got sixteen buttons, and I'm trying to figure out which button does the super-thrust power-up, in which case it's very cerebral, kind of like learning to play the piano. Rather than, you know, just picking up a shovel and starting to dig."

Will Wright on Miyamoto, New Yorker profile "Master of Play":

https://www.newyorker.com/magazine/2010/12/20/master-of-play

And Miyamoto judges from the face. GDC 2007 keynote: "What I always try to envision is the face of the player when he or she is experiencing the game, not any individual part of the game."

I've posted about his two GDC keynotes before, with links to all the videos (1999 full, 2007 parts 1-6, the wife-o-meter) and my transcription of the face-of-the-player passage. They bracket an amazing time in game development history, 1999-2007, and every word is profound:

https://news.ycombinator.com/item?id=27918255

Shigeru Miyamoto GDC 1999 Keynote (Full):

https://www.youtube.com/watch?v=LC2Pf5F2acI

The grandfather and granddaughter moment, GDC 2007 Part 3 at 12m29s:

https://youtu.be/SY3a4dCBQYs?t=12m29s

Buxton is the theory, Miyamoto is the practice: pick the transducer with your hands, verify it with the player's face. Watch Heinz Lemke's hands and face in the 1969 PIXIE film and you can score the interface without reading a word of the thesis.

Flight of the PIXIE, 1969 light pen and radial menus on the PDP-7:

https://www.youtube.com/watch?v=jDrqR9XssJI

Heinz (the guy in the the video demonistrating the PIXIE system he wrote) recently wrote me:

>Hope it doesn’t surprise you, that the use of a light pen in combination with the radial menus of PIXIE are very much related to the activity of a typical technical draftsman of the 1950’s. During that time and in the beginning of the 1960’s, I was such a technical draftsman in a number of companies in Germany, see attached a typical drawing from 1960 (few days ago digitised from a 1m by 2,5m drawing).

>What you may observe in this drawing is the frequent use of special symbols such as circles, squares, arrows, valves, nuts and bolts, etc. These are usually entered into the drawing (on a vertically positioned but slightly tilted drawing board) with a pencil at one hand and a special template with symbols held at the other hand. In PIXIE, I simply translated this into the light pen with the templates being simulated by a radial menu of symbols (PIXIE’s control light buttons) that can be dynamically adapted to the technical content of the drawing. Both, the light pen and the template can then be operated with one hand, leaving the other hand free to enter data, for example on the TTY.

>If at the same time the drawing gestures and drawing components are also translated into a computational model on the small computer, which after model completion can then be analysed or simulated on a larger computer, it becomes a winning situation all round. This was the aim of the PIXIE-Rainbow (PDP7-Titan) framework in those times. What appears to be missing now in our times is the “vertically positioned but slightly tilted drawing board” that used to be in the 1950’s and 60’s a social meeting point for all the colleagues in the drawing departments. This may be the loosing part of the advances in computer graphics.