SCO - IBM Lawsuit Takes on Movie-Like Proportions
By Bill Claybrook
November 17, 2003
Until the past week or so, the SCO - IBM lawsuit had been relatively quiet for a while. Then in the past few days, stories began to appear that SCO had subpoenaed Linus Torvalds and Stuart Cohen, CEO of OSDL, Torvalds's current employer. In addition, others that were subpoenaed by SCO include Richard Stallman of the Free Software Foundation and John Horsley, general counsel of Torvalds's former employer, Transmeta. In turn, IBM has subpoenaed folks from BayStar Capital, Deutsche Bank Group, Renaissance Ventures, and The Yankee Group. According to one news article [ http://news.com.com/2100-7344_3-5106450.html ], IBM claims to have 7,200 potential witnesses.
The initial lawsuit contained two basic parts: (1) SCO claimed that IBM had destroyed its Unix on Intel market via the Monterey 64 debacle, and (2) SCO claimed that IBM had infringed on SCO's IP by copying code from Unix System V into Linux as well as moving derived code into Linux. The first part of the lawsuit is likely going to be very difficult, if not impossible, for SCO to obtain a favorable ruling. SCO's Unix on Intel business was dropping prior to Monterey 64. With respect to the second part of the lawsuit, only Torvalds has anything of value to add in terms of testimony. The rest of those folks subpoenaed by both sides have no relevant testimony per the second part. I signed a NDA and viewed the code printed on two sets of paper and my conclusion was that I had no idea if code had been copied from System V to Linux. Anyone can print anything on paper. The source trees from which builds are done, if they are still available from years back, are relevant evidence, not code printed on sheets of paper.
The relevant people to testify in this trial are the developers who worked with the code. There are lots to choose from. Developers from Sequent should be subpoenaed because they were working with both BSD 4.2, 4.3, and System V. At some point in the trial, someone has to determine the actual source of the code that SCO says has been copied into Linux from System V. I would not be surprised if some of the code in Linux came from the BSD Unixes, and I would not be surprised if some of that same code is in System V. These things need to be determined. The best way to do this is to subpoena some of the developers and hire independent Linux/Unix knowledgeable folks to examine the source code, the build trees, etc, to try to determine the source of the code in Linux.