The interesting lines in the large graph are the faint ones showing the adoption level of specific versions of each browser. With Firefox, each new version quic

The interesting lines in the large graph are the faint ones showing the adoption level of specific versions of each browser. With Firefox, each new version quickly replaces the previous one in terms of adoption, so when you see that Firefox market share is around 22%, that's almost entirely the most recent release, installed over the last 12 months. Safari has similar adoption curves for each version, but releases are more frequent so the sum is smoother. Now look at the peak of each curve for IE versions. IE6 held over 80% market share for at least 2 years, by this graph, and probably longer. (IE5's peak would have been even higher, if memory serves.) IE7 had a strong initial uptake, and then gradually grew up to a plateau around 47% -- and now, with the release of IE8, it's dropping again. Given that IE7 was a substantial improvement over IE6 (tabs!), I don't see any convincing reason why IE8 would have a higher peak adoption level than IE7 did. That's an interesting trend, to me. If IE can't reverse it in subsequent releases (IE9+), and Firefox stays the course, then the peak adoption levels of FF3.5 and IE8 will be close, and as IE6 and IE7 inevitably drop off, it will become increasingly difficult for IE to maintain its overall lead.