For any CIO who's ever been caught on the wrong side of technology obsolescence, it's worth reading Charles Babcock's ode to a 30-year-old operating system that's still kicking.Babcock offers this bit of perspective on the survival instincts of the VMS operating system:

VMS has few peers in its age bracket. The hoary software of the IBM mainframe also is over 30 years old, but that's mainly because it's embedded in a kind of castle that won't fall. VMS never had a castle.

Is sticking with VMS a fool's game, despite its hearty tenure? Art Wittmann took on that question a month ago:

There's no doubt about the wisdom of "if it ain't broke, don't fix it," but one wonders about the wisdom of not finding a way to migrate OpenVMS or Alpha users, or PA-RISC users for that matter, to newer platforms and operating systems. The Alpha hasn't been updated since 2003 and the PA-RISC hasn't been updated since 2005.

Those chips will soon be unsupported, but HP continues to support OpenVMS. Sure, OpenVMS has a great history, and sure, it even has some unique clustering -- but you gotta believe that at some point, when the revenue from OpenVMS drops below, say, 10% of its revenue from yellow ink, HP will rethink OpenVMS support.