Here’s a link to something called pad++, a “zoomable interface” that was put together sometime in the late 1990s. The gist: With infinite zooming power, one can have documents at different scales and browse amongst them.

Jef Raskin, the user interface specialist who did early work on the Macintosh UI (though apparently, not involved in the actual final product) is doing some work on his humane interface. There’s a demo of one concept from it, which bears resemblance to the Pad++ concept.

Hmm. I’m curious to see what this turns into. After a cursory stumble through the documentation and demos on the raskin project homepage, one can see that the concept is simple enough, but execution is what will really make the difference.

The adoption of of a new user interface depends at least three things:

• You have to get something done on the computer
• You’re forced to deal with the GUI
• Positive progress overrides frustration level with the GUI

There are extenuating factors too:

• Does it look “nice”? (this is a loaded term)
• Does it jibe with my expectations? (aka it’s “intuitive”, another loaded term)
• Do I have a choice to use this GUI or not?
• Perception: Is it worth my time to learn this GUI, or go do something else?

The zoomable interface is one of those solutions looking for a problem. Neat idea, but it’s only part of an entire solution. Personally, I’m not that convinced it would be as usable as I want without introducing overlays of some kind, like a Heads-Up Display providing additional navigational cues as your explore your mega desktop. Orienting and referencing other landmarks in your document is a pain in the butt, something well understood by graphic designers using Photoshop and Illustrator at zoomed-in magnifications. It sucks. Making zoom control more accessible has not solve the problem.

The ZoomDemo also suggests that using the mouse buttons as your primary “zoomin, zoomout” controls would be far easier and more natural. Yah, if your primary task was exploring. When you’re actually working on something, you want those mouse buttons to be doing something more important. The mousewheel might be a better substitute.

On another note, Apple’s Expose works rather well by implementing the idea in reverse…you start from your window focus, and expand into a zoomed out overview of your desktop.

There’s a kind of computer use that I haven’t seen addressed: piloting versus potato digging. By piloting, I mean performing multiple subtasks as part of the overall task. By potato digging, I mean the focus on getting one of those single tasks done. The focus on singular tasking and simple intuitive commands is fine for people who are just trying to do something for the first time, but quickly we want something better, faster, and more powerful as defined within the context of the specific task we’re trying to get done. I am doubtful, given the myriad of different tasks that are out there, that a single “human environment” can productively address them all. Which is why we still have Operating Systems, not OpenDoc… this is the structure that supports differing approaches to problems without imposing a lot of simplifying-but-ultimately-time-wasting conventions.