Win95 development head Silverberg leaves MS

So it's finally farewell again, Brad Silverberg, combative street fighter and Microsoft executive, who as a senior VP led the development of Windows 95. He will be resigning tomorrow after a very extended leave. He had been a consultant to Microsoft, advising on its consumer strategy. The arrival of Rick Belluzzo from SGI has apparently convinced him that "his help was no longer needed" although there is no evidence that he is overly sore about this, since some time back he had turned down the job of heading the interactive media group, i.e. what turned into Belluzzo's job. Silverberg often was the front man when Microsoft was being criticised. He was in charge of the Windows 95 delays ("If we're not perfect at scheduling, we apologise" and OS/2 "was a cheap imitation of [Windows 95]"). He disliked OS/2 with a passion, once claiming that Sprint was deploying Windows 95 beta in a mission-critical production environment, but this turned out to be a three-PC deployment where there were 1,300 PCs running OS/2. Intel also had a somewhat negative view of him, it came out in the trial: he was "incredibly arrogant... dangerous... extremely hostile... the guy hates us... pent-up anger... impossible to deal with." Silverberg came from Borland and joined Microsoft in 1989. He later spearheaded Microsoft's first efforts at luring away key Borland staff to Microsoft, offering a $3 million signing-on bonus for Anders Hejlsberg, the chief architect of Delphi, plus $200,000 salary and 75,000 shares. Paul Gross, then VP of research at Borland, was only given a $1 million signing-on bonus, apparently. When Microsoft included the DR-DOS trap in Windows 3.1, it was Silverberg who tried to imply that the compatibility problems had been with DR-DOS, rather than deliberately engineered by Microsoft. It was also Silverberg who announced Windows 95 sales to much incredulity in late 1995: his figures did not square with Microsoft's supposed accounting principles. When Microsoft was caught red-handed with Apple QuickTime code in its Video for Windows developers kit, he claimed that "The licensed code is low-level driver code" but did not explain why it just so happened that it improved Microsoft's video performance dramatically. Nor was the code licensed from Apple, of course. Clearly, he's vested. ®