An attorney friend recently explained that problems aren't patentable, only solutions are. That is, you can't patent the idea of a cure for cancer, and then claim the rewards when someone finally accomplishes that task. Clearly, that philosophy is outdated, as this patent has apparently thrived.
The inventor, Dr. Doyle, has a PhD and was employed at UCBerkeley, seems more legitimate than your average troll. Nevertheless, he didn't describe any technical implementation e.g. a scripting language, dynamic libraries, or even IO redirection of an external app. Obviously, he must have intentionally avoided patenting a specific invention in order to cast a wider net.
Ironically, none of the defendants actually sells a product that infringes. The objectionable products are all provided for free. AJAX and the rest are neither critical nor valuable. Damages? In Microsoft's case, I'd have to guess the $500M award was entirely putative. The only defendant that even indirectly benefits from the technology is Adobe.
All of the described functionality existed prior to 1994. The supposed "innovation" is describing the browser as a virtual platform. So '906 patents the browser. Mosaic was already available. The premise must be that, prior to this invention, the browser was limited to reading HTML documents.
There are so many defendants, I don't expect this case to be settled. In spite of the recent decisions, which only seem to have heightened Eolas's hubris, the legal merits of this case are much flimsier than most— even patents that seem painfully obvious. I expect this case to be a real watershed for software patent disputes.
"Take that, you hostile sons-of-bitches!" -- James Coburn, in the finale of _The_President's_Analyst_