Ars Technica posted new browser market share numbers courtesy of Net Applications and it looks like Firefox has gained ground at Internet Explorer's expense. According to the data, we're seeing more and more browser fragmentation. Currently it is fairly small, with Opera and Safari making only cursory gains, but as a percentage of previous market share, Opera, Safari and Firefox continue to grow by leaps and bounds.
When you look at the graph (or look at a graph that hasn't been modified), Internet Explorer still rules the roost. It has a whopping 82% market share and when people build applications for the browser, IE tends to be the one they do first. People are willing to hack around IE to gain that market share. But look at the growth of the "up and coming" browsers. What happens as the market share levels out and we see IE with 45%, Firefox with 40%, Safari with 10% and Opera with 5%? Do people still build browser-based applications for all browsers, or are they content to target their favorite, or the one that supports what they want to do best? Will browsers, fighting for that market share, cater to different sets of developers by offering "advanced" features just like Microsoft did with Java?
I don't know the answer, and I'm not predicting doomsday for the browser. Displaying text and HTML is one thing, but building an entire application that runs inside that browser is another thing entirely. As developers, if the goal is to hit the mass market, it becomes awfully enticing to use runtimes that are cross platform and independent of the various browsers in use. And moving away from the browser doesn't have to mean moving away from the web. Technologies like Apollo and Windows Presentation Foundation are great, and they provide a way to "take the web with you", but if you are looking to build web-only applications, Flash, OpenLaszlo and Microsoft's XBAP provide that as well.
It's great to see Firefox adoption continue to grow, and it's good to see Safari improve its standing. I think competition is a good thing because it helps foster innovation. But the browsers will mature, and there will be differences in how they support the applications that we're building in them today. That's a tough thing for a development shop to keep up with, and I think that has to be in the back of the minds of people making the business decisions. There are trade offs to both, but I think the trade offs are less when you go outside the browser and treat the web as a universal desktop.