Tuesday, May 29, 2018

Pie Menus

Don's NeWS Pie Menu
IIRC it is 1988, and James Gosling and I are in the Sun Microsystems booth at SIGGRAPH demo-ing the NeWS window system. Don Hopkins walks up with a tape cartridge in his hand and says "load this". Knowing Don, we do, and all of a sudden all the menus in the system are transformed from the conventional pull-right rectangles to circles divided into pie-slices. And Don, at that time the most caffeinated person I'd ever met, is blazing through the menus faster than we've ever seen before.

Why am I writing this thirty years later? Follow me below the fold.

Don's CHI88 demo
Two weeks ago Don posted Pie Menus: A 30 Year Retrospective. It starts:
Today (May 15, 2018) is the 30 year anniversary of CHI’88 (May 15–19, 1988), where Jack Callahan, Ben Shneiderman, Mark Weiser and I (Don Hopkins) presented our paper “An Empirical Comparison of Pie vs. Linear Menus”. We found pie menus to be about 15% faster and with a significantly lower error rate than linear menus!
This should not have been a surprising result. As Don writes, it is a consequence of Fitts' Law:
the bigger and nearer by the target, the faster and more accurately you can hit it. Pie menus maximize the target size, and minimize the target distance!
Don's post is fascinating, full of examples and links to valuable human interface work. You should read and enjoy the whole thing. Even if you don't want to read Don's post, you absolutely have to watch the video of Simon Schneegans' Gnome-Pi, a beautiful implementation of pie menus for Gnome.

But I can't resist this snippet. Don recalls that:
On October 25, 1988, I gave Steve Jobs a demo of pie menus, NeWS, UniPress Emacs and HyperTIES at the Educom conference in Washington DC. His reaction was to jump up and down, point at the screen, and yell “That sucks! That sucks! Wow, that’s neat! That sucks!”

I tried explaining how we’d performed an experiment proving pie menus were faster than linear menus, but he insisted the linear menus in NeXT Step were the best possible menus ever.
Jobs' expertise in the design of menus can be illustrated by the following story, which I hope I recall correctly. The then (and still) canonical way sub-menus work is that they pop up with their top aligned with their parent menu item, so that the most frequent choice at the top of the sub-menu is closest. But Jobs thought that this looked messy, so on the NeXT they popped up with their top aligned with the top of the parent menu. Everyone else at NeXT knew this was bogus, so one time when Jobs was on the road they fixed the menus to work right. When he got back he went ballistic and made them unfix the menus.


Chris Rusbridge said...

Makes sense once you're not working on a character-based computer!

Slightly related, my favourite note taking and document planning tool since early in my working life was the mind map. You put the concept in the centre and elaborate radially. It's great virtue for me was eliminating top-left and prior though bias; you fit a new idea where it best fits rather than (by default) at the end. It was particularly useful with later version of mind map software (better than white boards, invaluable though they were) which move stuff around to accommodate new, and allow you to drag ideas around the place. Then (after retirement) Apple dumped backwards compatibility, my software stopped working, and the company had radically changed (worsened) the design and changed to an unaffordable rental model. So now I'm back to paper when I need to plan anything. Oh well.... :-)

David. said...

Don Hopkins' research into the pre-history of pie menus traced them back to a system called PIXIE, developed in 1967 at Cambridge by Neil Wiseman and his students on the PDP-7 + DEC 340 display system that I used as an undergraduate.

Now, Don has worked with a librarian at Cambridge to unearth and digitize two movies of PIXIE in action, and has set excerpts to music here.

David. said...

Permanent links to the two PIXIE movies are one and two.