Listen to this article
Every year I install the latest version of the GNU/Linux operating system on an Intel- compatible computer. Each year it is little bit easier to install and use.
Easier, but unfortunately not easy enough, even for someone who’s fairly accustomed to – and usually able to handle – standard snafus with technology. And every year I conclude: maybe next year.
It’s not that I want Linux to fail my personal test; quite the opposite, given the well-known flaws in Windows and relatively high price of Macintosh hardware. More choices for customers are always better than fewer.
No, the issue is more of style than substance. The open source software community, which has done wonders in so many ways, has not adequately tackled fit and finish.
Imagine that big open source projects such as Linux are buildings. They’ve been designed by well-schooled engineers, erected by expert carpenters, steel workers, brick layers, electricians and plumbers.
A warehouse doesn’t have to impress the crates and other items stored in it; but an office or residence needs alluring fit and finish.
For many open source projects, the look and feel are unimportant. The typical user will never need to know the inner workings of server operating systems, databases and various items of internet plumbing.
It’s important to recognise the genuine progress in ease of use and installation that we’ve seen with some of the bigger projects.
But on my last Linux installation on an IBM ThinkPad, I kept running into little quirks that were not present with the pre-loaded Windows operating system.
To get the “sleep” feature working at all, for example, I had to make manual changes to a system file that no normal user would attempt. OpenOffice still has some compatibility problems with Microsoft applications. Much other open source software is similarly solid in its plumbing and not so great in its user interface.
It’s not a mystery why this is so. For the smartest programmers, elegance is more about the solidity of the foundation, floors and walls – whether the software does its essential tasks quickly, reliably and safely – than the veneer.
For information technology professionals who install and maintain servers, the veneer is close to irrelevant.
It’s relevant to at least one company that sells to consumers, and to the customers. The central code in Apple Computer’s Macintosh operating system is BSD, which like Linux is an open source variant of Unix. On top of that base is easily the most elegant, easy to use and ultimately most powerful end-user computer operating system on the planet.
For the moment, despite its somewhat higher cost, the Mac is the answer to the question: “Which version of Unix is best on the desktop or laptop?”
Companies that embed open source software in limited-purpose commercial devices, moreover, can do much the same.
The TiVo hard-disk video recording system uses Linux, but boasts what is widely considered the best user interface of any such system.
As mobile phone makers adopt Linux and other open source software, you can bet that ease of use will be at the top of the list, because a phone that’s hard to use won’t sell.
Another common lapse in open source software is documentation. To be fair, almost no commercial software comes with helpful instructions.
Apple’s manuals, like those of many others, are so lacking that book publishers do serious business selling what should have come in the box in the first place.
Cost, of course, helps explain why many open source projects achieve solid underpinnings but frustrate non-technical users. When a project depends on volunteers, they will work on what’s elegant to each other as opposed to what’s elegant to end-users.
I’d like the open source community to spend more time reaching out to design and usability stars who want to make the same kind of contributions their technical friends have already made.
There’s been remarkable progress. It needs to continue.
Dan Gillmor’s website is bayosphere.com
For more articles written by Dan Gillmor, click here