Browser design

I’ve been thinking about the design of web browsers in the light of making comparisons between Firefox and Internet Explorer, and rapidly came to the conclusion that simply making check-lists of features is a banal exercise. So, back to first principles…

The basic problem is that for various reasons (including Microsoft’s operating systems monopoly and the elimination of competition in the browser market after the demise of Netscape) there was no serious competition in this area. That doesn’t mean that better browsers didn’t emerge (they did — in the shape of Opera, Safari, Omniweb, etc.); but none of them gained real traction in the face of the monopolistic bundling of Explorer with Windows. Which meant that Microsoft had no real incentive to enhance or improve Explorer, so the program wound up being the end of a particular line of browser development. To take it any further would have meant rethinking the program completely — re-architecting it in computerspeak. And there was no commercial imperative to do that.

For once, I’m not making an anti-Microsoft point. As far as they were concerned, Explorer wasn’t broke, so why should they bother fixing it? But this has had a serious downside for the rest of us, because the web browser is the program people use most: it’s their window onto the Web, and (to develop that metaphor) if a window is distorting and dysfunctional and inefficient, then it distorts one’s view and makes it more difficult to see things.

So browser design matters, and this thought led to the question of what kinds of principles should guide browser design. At this point I stumbled on this terrific essay by Scott Berkun, who in an earlier life worked for Microsoft and indeed was a program manager on the Internet Explorer project. So he knows a lot about the thinking that went into the design of Explorer. He examines the things people want to do while browsing and then ponders their design implications. The only point where I found myself disagreeing with him concerned the way he appears to lump RSS technology with the discredited ‘push’ technology of aeons ago.

The main conclusion I draw from Scott’s essay, though, is that although Firefox is undoubtedly a step forward from Explorer, browser design has an awful long way to go. Software’s wonderful, but it’s still terribly crude compared with the sophistication and adaptability of the human beings who have to use it.