More on Microsoft's need to show (if it can) Linux's patent infringement

My blogging colleague (and ZDNet editorial director) David Berlind sent me an interesting follow-up note on my call for Microsoft to show proof that Linux is infringing on Windows patents -- as Microsoft CEO Steve Ballmer alleged last week.

Berlind -- who has written more than his fair share of stories on intellectual property and patents -- explained that what I really should be asking Microsoft to do is demonstrate that one or more Linux varients 'reads' on its patents.

Here's Berlind's clarification: "Microsoft doesn't have to show infringing code. It has to demonstrate that some implementation "reads" on its patent. That implementation could be in grapefruit, tiddly-winks, or a final executable program.

"The 'show-me the code' thing worked with IBM SCO because SCO claims that IBM is infringing on its copyrights.

"You can connect specific snippets of source code to copyright violation. But patent violation doesn't necessarily have to show code (although a copyright violation would make it a sealed and shut case). It has to show likeness. For example, if you have one-click buying on your web site, the code behind it is irrelevant. As long as works as 'one-click,' in the user interface, it's Amazon's patent."

What I felt Friday still holds: I think Ballmer & Co. must do more than continue to make saber-rattling FUD-filled claims regarding Linux's supposed infringement on Windows patents. The form that proof takes is up for debate. But the requirement for it remains.