Retrograding

An early version of Microsoft Word for the Mac.

Technophiles have always been a nostalgic group: witness the guy who converted his spare room into an Apple museum (much to his wife's dismay, no doubt), or any of the retro design elements that have been making their way into our new electronics. But lately there are signs that there's more than just nostalgia at work; that as technology consumers, we often prefer the good old days, and we want them back.

Exhibit one: Windows Vista. There's been enough said about Vista's failure to live up to the hype to fill a thousand blogs, much of it a rancor mix of late-adopter justification and some pure and simple schadenfreude. Still, Microsoft's widely publicized agreement to extend Windows XP support and continue to offer the old OS on new machines is a powerful statement in itself. It's not enough to come out with a new product, even if it's arguably the most successful in all of computing history, to generate sales and widespread adoption.

Vista is hardly the only example. Integral pieces of desktop software, such as AOL Instant Messenger, have grown from small and nimble programs into memory-hogging monstrosities loaded with features we hardly need nor want. Perhaps that was one of the motivations behind the aptly-named website oldversion.com, which features a full back catalog of software across a range of apps, including AIM. Upon installing version 4.2, I was shocked at how little I missed the pretty tab features of the newest AIM client, and how nimble the software, written in 2001, performed on my Core 2 Duo laptop.

All of this has important design implications. It's hardly a new position to be against the often-derided feature-glut. But renewed interest in old software serves as a potent reminder of real user needs. Familiarity is a powerful thing; software upgrades need to actively justify new interfaces when users already know their way around the old one. We should be feeling the wondrous effects of Moore's law, in that our computing experience should be getting more responsive, not just keeping pace (frustrations here have run rampant). Most importantly, it raises a simple question of economics: are consumers willing to pay for smaller software with fewer features if they are better executed? Or is the old logic still correct - that we need ever more features to justify a spend on new software?

It's hard to say. I don't think consumers truly want to go backward (I don't forsee an impending renaissance of Microsoft Bob, for example). But particularly as software categories mature, we need to work harder at justifying new upgrades, and think hard about why we're asking consumers to upgrade at all. Otherwise, the prevailing wisdom may be right: if it ain't broke, don't fix it.