Open source's usability challenge

commentary The iPhone has been out for a year, and known about in detail for considerably longer. Yet the very latest crop of state-of-the-art Windows Mobile phones, clearly designed as head-on competitors to that phone, miss the mark by miles.

They all have the same feature list--indeed, they capitalize on the many aspects of the iPhone that are well below par--but they all feel cruder and more frustrating to use. You can't just bolt this stuff on.

Usability is extremely important. It's also very hard to do well and requires extreme corporate self-discipline.

If you ever find yourself asking which of Apple and Microsoft has the more effective management, compare their products. Unless an engineer knows with total clarity that their part of the project will be thrown back for failing to meet usability standards, then the urge to cut a corner or half-bake a feature can be irresistible. Functionality is easy to specify and test: usability far less so.

That should make usability supremely challenging for open-source projects, especially ones of the scale and complexity of Ubuntu. But then, open source is not known for following corporate theory: witness the fact that the Linux desktop has evolved to the point where it's beyond merely usable but a fully-featured and viable alternative in daily work to paid-for software. That should have been impossible--it turned out to be anything but.

So when Mark Shuttleworth calls for usability to be a primary concern in future Ubuntu developments, it's fair to say that he expects it to be achievable. That he's employing a team to add usability to X, OpenGL, Gtk, Qt, Gnome and KDE shows that he's prepared to back that expectation.

But it needs more. Open source needs its Steve Jobs--which is admittedly like saying America needs an emperor. It needs someone with an extremely clear and uncompromising vision of usability, and the charisma and energy to make people want to follow.

As many in the movement have a visceral reaction to autocrats, the primary role of such a figurehead won't be to tell, but to teach--to show people how to want to write usable software more than anything else.

Any glance at open-source discussion groups will show just how far out of mind such considerations stand. Unless this starts to change, open source will remain limited in its power and reach--and its competitors will carry on getting a free ride.