In response to the SCO lawsuit, Eric (with consultant Rob Landley) wrote the "OSI Position Paper on the SCO-vs.-IBM Complaint." This position paper addresses in detail SCO's claims of intellectual property ownership over Linux. The paper has been widely read and is considered by many to be the best analysis of the topic available. In short, the paper addresses the question, "Who owns Unix?".

LWM was able to catch up with Eric on the day of the Novell announcement that SCO did not own the patents or copyrights to Unix.

LWM: In a nutshell, what exactly is SCO trying to do?
esr: What they were trying to do, I think, was shake IBM down for a payoff or a buyout offer. That has blown up in their face, especially now that Novell has made a public statement that all but accuses SCO of lying about the disposition of the IP. But now they have to play this losing hand out to the end - because admitting that they knew they didn't have a real case to begin with might land their management in jail for fraud and harassment.

LWM: So tell me about the position paper you developed. Why did you and Rob Landley write it?
esr: I was trying to do two things really. One, I was trying to give IBM ammunition. Two, I knew the open source community would have to respond to SCO's attack sooner or later, and that it would be better if it was sooner - before SCO's propaganda (if any) had time to take hold.

But part of why I was upset didn't have anything to do with Linux. I'm actually an old Unix developer - back to 1982. I wasn't one of the original developers of Unix (though I've contributed code to Linux and the BSD Unixes), but I know those guys and they know me. The SCO complaint was insulting. It was SCO claiming that they owned all the code that we wrote - and then using that claim to harm Linux.

LWM: What's happening with your "No Secrets" effort?
esr: I'm trying to prove that the proprietary Unix vendors don't have any trade secrets. Right now I have enough people willing to sign affidavits about having uncontrolled read access to Unix source code that I can show there's been a pervasive failure to enforce even the minimum level of nondisclosure required to maintain trade secrecy. Thousands of people who have seen the Unix source code were never under nondisclosure. This is the kind of evidence that destroys trade-secrecy status.

If SCO continues, I'll get enough signed affidavits to prove that they have no trade secrets.

This is also an attempt to send a powerful message to potential future litigants: it's not safe to mess with the open source community because we can bite back.

LWM: And what is IBM's position on all this?
esr: You'll have to ask IBM that. I'm their ally in this, not their spokesperson.

LWM: For readers who may be unfamiliar with your work in this area, can you share some of your background with open source and Linux?
esr: I wrote the foundational paper on open source development, ran the meeting where the term "open source" was invented, and have been one of the community's principal ambassadors to the rest of the world for the last five years. I am the president of the Open Source Initiative, one of the community's two leading advocacy organizations.

LWM: What is the position of the Open Source Initiative on this issue?
esr: We believe SCO's claims are utterly without merit. In much of their complaint they seem to be, plainly and simply, lying through their teeth. We have published a detailed rebuttal at www.opensource.org/sco-vs-ibm.html. It looks even stronger than it did in light of Novell publicly announcing that they, not SCO, own the Unix patents.

LWM: So, who owns Unix?
esr: Legally, it's very unclear. Novell holds the patents. The OpenGroup owns the trademark. The copyrights are in some weird limbo - first Novell came out and said they owned them, but SCO now claims to own them under the terms of Caldera's deal with Novell and Novell is keeping mum. The one thing we do know is that the transfer of the copyrights (if any) was never recorded with the U.S. Patent and Trademark Office. That has interesting legal implications, and may be the reason SCO hasn't come out and made an explicit copyright-infringement claim in the lawsuit.

Ethically, OSI's position is that Unix belongs to the distributed development community that wrote it. SCO's threats broke the tacit understanding that kept us from asserting this for 30 years. It used to be that we agreed not to fuss over the fact that AT&T or Unix Systems Labs or Novell or SCO were claiming to own the code as long as they agreed not to fuss over the fact that every senior Unix developer had a technically illicit copy of the source code in his hip pocket.

Everybody took code from everybody. AT&T used Berkeley and Xenix code and got called on it during a 1993 lawsuit. Truth is, the rights picture is so tangled that nobody's theory of ownership would stand close scrutiny of the source code's history. The law of intellectual property doesn't handle this kind of situation well. The equitable thing to do would be to just give up, throw it open, and admit it belongs to the hackers.

LWM: What do you see as the potential downside risk for companies using Linux? Will SCO try to sue everybody?
esr: The risk dropped to zero last May 28 with Novell's announcements. SCO's response (http://biz.yahoo.com/prnews/030528/law059_1.html) basically admitted they've got no grounds to sue anybody but IBM.

SCO have since changed their minds, but I think this is just bluster. Furthermore, the various lawyers I've talked with agree that it's just bluster. When you think you have a strong case in court, you don't fight it in the media. SCO would scare me worse of they weren't huffing and puffing.

LWM: If you were a manager in a company considering using Linux for a first project, would this lawsuit impact your decision to give Linux a try?
esr: Not at all. Ignoring the occasional FUD storm is part of the job.

LWM: In your book The Cathedral and the Bazaar you describe the Linux development process as being like a bazaar, where all kinds of people with all kinds of interests are developing different pieces. Is Linux development still that way? How has it changed?
esr: If it has changed, it has changed by becoming more conscious and better organized. I played a part in that by giving people language with which to reflect on what they're doing.

LWM: What do you think will happen with this suit? Any idea how long it might be before it becomes clear what's going to happen?
esr: They can't win, not in front of a judge with any brain cells operating - and the word on His Honor Dale Kimball is that he's a sharp guy. Timeframe? Who knows. These things can drag on for years.

LWM: How can the Linux community ensure that Linux stays free of IP claims in the future? Can there be a process instituted that ensures this doesn't happen again?
esr: See my "No Secrets" page for an example of what network activism can do (www.catb.org/~esr). I've collected nearly 100 responses, with at least 40 people willing to sign affidavits. I think we can prove that there are no trade secrets in Linux. I think we can use the same methods to turn up prior art in patents cases.

LWM: Switching gears a bit, in the IBM -v- SCO analysis on the OSI Web site (www.opensource.org/sco-vs-ibm.html), you referred to a "seismic shift" occuring right now in the software industry. Can you explain what you meant?
esr: I already have. Readers should go to www.opensource.org/sco-vs-ibm.html#seismic for the story.

LWM: Will all applications eventually be open sourced? Which kinds might not?
esr: I don't think it will be all - there are economic circumstances in which closed source makes sense, though they're not common. I think "most" is a fairly safe bet, though. I've discussed this at length in my paper "The Magic Cauldron."

LWM: What will the software industry look like in five years?
esr: A lot like the legal profession does now, I think. Independent software firms will be like law firms, partnership organizations of professionals. Other programmers will work in-house at corporations the way that corporate lawyers do now. Programmers in general will be operating from a common open source base; secrecy will be a feature mainly of legacy software.

Regarding outsourcing and offshore development - one thing you can't outsource is getting inside a customer's mind. You can't move face-to-face, person-to-person communications and design offshore. You can outsource cookie-cutter code, but I predict a lot of companies are going to discover they're paying for large portions of code that don't match their requirements.

One of the things we know is that the most effective ways of writing software involve a series of interactions - a succession of prototypes - using continuous feedback. You can't do that if your customer's in Teaneck, New Jersey, and your developers are in Bangalore.

About Eric S. Raymond
Eric S. Raymond is an observer-participant anthropologist in the Internet hacker culture. His research has helped explain the decentralized open source model of software development that has proven so effective in the evolution of the Internet. His own software projects include one of the Internet's most widely used e-mail transport programs.