Guild Companies

Bruce Perens, a defender and evangelist of the open-source community, has gotten his hands on the presentation The SCO Group made last week at SCO Forum in Las Vegas. Any lawsuit is a gamble, especially one pitting two headstrong technology companies against each other in a jury trial. So "Lost Wages" was probably as good a setting as any to divulge what SCO believes are some of the smoking guns in its lawsuit against IBM.

Perens, as you can see from his analyis of SCO's Las Vegas slide show, is not impressed by the code snippets SCO rolled out to support its claims in the court of public opinion, which is the only place in which the SCO-IBM suit will be heard before the April 11, 2005, date set by the U.S. District Court in Utah.

By the way, according to Perens, SCO leaked the slideshow to the IDG News Service without an non-disclosure agreement, which calls into question the whole idea that SCO has about doing due diligence with Unix code by not showing the objectionable code snippets to the public at large.

But all of this is beside the point. The real issue is the code snippets themselves, and Perens says that two of the allegedly copied code segments do not belong to SCO at all. Perens says, in his analysis of the presentation, that he expected that SCO would put the best examples forward to illustrate what intellectual property was stolen. "But I was easily able to determine that of the two examples, one isn't SCO's property at all, and the other is used in Linux under a valid license," Perens said. "If this is the best SCO has to offer, they will lose."

The first example dealt with Berkeley Packet Filtering, which is an implementation of Internet firewall software that was created at Lawrence Berkeley Laboratory using American tax dollars. He says further that Berkeley Packet Filtering is a derivative work based on a program called enet, developed by Stanford University and Carnegie Mellon University, two of the biggies in the Unix movement.

He says Berkeley Packet Filtering was first shipped in BSD Unix 4.3, which was created by the University of California at Berkeley (which gives BSD its name), and that SCO later copied this code into Unix System V. The BSD license, Perens says, allows SCO to do this. But, he says, SCO doesn't own Berkeley Packet Filtering, and has no right to prevent anyone, including the Linux community, from using it. Moreover, Perens says that the Linux implementation of Berkeley Packet Filtering is not even BSD code, but a clean-room reimplementation of Berkeley Packet Filtering by Linux developers lead by programmer Jay Schulist, based on the documentation provided by Berkeley Packet Filtering.

The other SCO code snippet Perens walks through has to do with memory-allocation functions in Unix System V and Linux. He says there was "an error in the Linux developer's process," specifically a programmer at Silicon Graphics, and he says that while the Linux community had the legal right to this code, it didn't belong in Linux and was therefore removed.

According to the SCO slides, two-thirds of the new code was added to Linux between the 2.2 kernel and 2.4 kernel releases, comprising 1.1 million lines of code in 1,549 files and touched by many of the Unix licensees (including IBM, Silicon Graphics, and others). This is a pretty big claim, obviously. And Perens brings that point home by explaining the similarities between the SCO-IBM case and a lawsuit between AT&T Unix Systems Labs and the University of California and BSDi (which sold BSD Unix, in the early 1990s).

"Under SCO's theory," Perens says, "if any code created by a Unix licensee ever touches Unix, SCO owns that code from then on, and can deny its creator the right to make use of it for any other purpose. SCO's legal theory fails, because they ignore the fact that if a work doesn't contain some portion of SCO's copyrighted code, it is not a derived work."