Robert Clayton Miller has done an excellent thing: he’s re-imagined the personal computer GUI and associated input devices in a practical, achievable, realistic way. It’s no small achievement, this: we’ve essentially been using the same lineage of human-computer interface metaphors and technologies for more than 25 years. Originally developed at Xerox PARC, Steve Jobs brought the windowing operating system and the mouse to consumer computing in the early 80s. While Moore’s Law has grown computing power exponentially and we now have multitasking and colour screens, nothing much has changed when it comes to our input devices. (If you’re at all interested in the history and impact of Xerox PARC – and you should be – this book is a fantastic read.)
Here’s the 10/GUI video. (Note this is a designer’s demo, not an actual operating system.)
When I first saw this, I found myself appreciating the immense amount of work this must have been to produce. I’m not a designer but I’ve worked with enough designers on enough design projects to know that this can’t have been easy. It’s thoughtful, comprehensive, compelling and just realistic enough to make you believe.
So what do I think about 10/GUI and the associated CON10UUM windowing metaphor? First, here are the current UI/interaction challenges it addresses well:
- It proposes to combine the ten-finger multi-touch interaction pad with a keyboard (watch closely). This addresses the somewhat ‘lost’ feeling users have when faced with the Microsoft Surface or other, multi-touch only devices. There’s a keyboard, you can still enter text. (If you’re in the market for a humorous aside, watch this.)
- It distills the essence of the mouse’s success (being apart from the screen so as to not obstruct the user’s view of the screen).
- It addresses the ‘arm fatigue’ issue that straight-ahead multi-touch screens would undoubtedly cause.
- It involves as many or as few fingers as required. The potential for finely manipulating digital objects is pretty self-evident. If the multi-touch trackpad on my Mac is anything to go by, this will be a fairly natural user experience and easy to learn.
- It doesn’t propose a gratuitous 3D computing environment. In the end, the key insight for me is that 3D is justified when you’re working on things that benefit from 3D. When you’re working on things that don’t (documents, web pages, email, video), a 3D UI would cause cognitive friction. 10/GUI with CON10UUM respects our basic sense of the appropriateness of a metaphor.
- It’s evolutionary rather than revolutionary – recognizable – and therefore implementable.
Here are some areas where I have questions about the 10/GUI vision:
- I find myself skeptical of the straight line of windows in CON10UUM’s proposed UI. I can understand why this is being proposed: it essentially extends Firefox’s tabs to the entire OS. And, truth be told, I sometimes find myself in a deep computing mode where I’m momentarily surprised that all apps don’t act like Firefox. But I struggle to see how CON10UUM would facilitate certain operations that we’ve come to expect as pretty basic features of operating systems: dragging and dropping of files, for example. I didn’t get a sense that the linear arrangement – grouped by originating application – would lend itself to that. Of course, there could be work-arounds – perhaps a file manager like the old Norton Commander.
- I also think the desktop, shown at the beginning of the demo, is a missed opportunity. I sense that you would never see it, and that it would have no purpose. Then again, that’s not unlike either Windows or OS X – the desktops there don’t necessarily have a function either.
Chris Clark points out that he really likes the multi-touch gestures. I second that whole-heartedly. And I think it’s alarming that – like crop seeds – multi-touch gestures are already being patented by tech corporations. In the interest of computer users everywhere, we should resist the notion that things we do can be proprietary.
If I could give one more bit of feedback to Clayton Miller, it would be the lack of community infrastructure around 10/GUI. The website doesn’t have a forum or comments. There’s no way to leave feedback centrally, and – as a result – it’s hard to see how the open source community might go about embracing this vision in a practical sense. But that may not have been his intent.
On the whole, I’m curious to see where 10/GUI ends up. Will someone create an open source windowing environment for Linux once the hardware is available (and I’m certain Synaptics is working on something like this already)? Will Apple or Sony have a go at this (as Clayton suggests, only a vertically integrated hardware/software manufacturer could really do it justice today)? Time will tell. In the meantime, we can use up to 4 fingers on our Apple trackpads and wait patiently for the arrival of the ten-finger version.