Covering SCO is a marathon, not a sprint, so after each big win, I tend to savor the moment, goofing off and enjoying a rest, knowing as I do that they never quit and there will be more awfulness to come.
I confess I have been cracking myself up this time by reading old Rob Enderle, Paul Murphy, and Maureen O'Gara articles I had saved from the early days, each predicting solemnly Linux's doom. I can't tell you how much *more* fun it is to read them now than when they first showed up in 2003 and 2004. Remember Why SCO Should Win? Or O'Gara publishing Dan Wallace's letter supporting SCO's then-theory that the GPL was unConstitutional? Here's a law professor's answer to Wallace, by the way, for those who are new.
Ah, those were the days, weren't they? It's so pleasurable to reread all the heady nonsense SCO and its supporters were spewing. My favorite was Murphy promising to quit if SCO lost. Of course, he hasn't. He's still at it. Would you like to see his new dire prediction? He seems to think Elliot Management is a front for Microsoft or some other player who can buy SUSE and the copyrights, and then go after Red Hat and you and me.
Oh goody. Some new FUD. Let's enter, temporarily, then Murphy's alternate universe, because he forgot one detail. Well. Two.
First, for those who just must give him hits, here is the url:
But here's the part that made me laugh:
Now, however, the case has effectively made them real enough that if a company like Microsoft or HP could grab Novell they’d be able to combine these copyrights with SuSe’s market position and corporate credibility to quickly erode the lets it pretend its not a license gag supporting both SuSe and Red Hat licensing -and that would enormously weaken not just Linux, but the whole open source community. He goes on to predict an alternative, that Novell could try to "unload" the copyrights on the open source community, then Elliot would counter somehow to "protect the company's value" to a potential buyer like Microsoft, leaving IBM wondering what to do, what to do about Linux.
As a result the very best outcome Linux advocates can hope for now is that the Judge orders the copyrights transferred to SCO. That moots a couple of the lawsuits, lets most of Novell’s directors off the collective hook, takes the copyright issue out of the underlying contract dispute, and leaves Elliot holding some speculative stock that simply didn’t pan out.
Here are the two things he forgot about.
so far as SCO has been able to put anything on the table, after seven long years, we haven't been shown infringed code to build a case on, so the UNIX copyrights' ownership is irrelevant. It doesn't matter who owns them if there is no infringing code in Linux. Elliot folks, or Microsoft, or whoever, would have to prove that there is code in Linux that infringes those UNIX copyrights. But here's the problem with that. It wouldn't be a money maker because it would be removed, and we'd all go our merry ways. SCO's problem was that it acted like copyrights are patents, where you can force people to license. With copyrights, even if you can prove infringement, which after seven years SCO hasn't been able to do, the infringer has the option of just removing it. Bingo, presto, shazaam. The scam is over. No free ride. And no one I know who has compared the code can find any infringement. So there you are.
Second, the GPL stands in their way even if there had been any infringing code. SCO itself, as Caldera and as SCO, distributed that code, before, during and long after making the infringement claims, and they did the distribution under the GPL. It doesn't matter if they own the copyrights or not. As distributors, they said it was OK to use the code they claimed they owned under the GPL. So that provided us a license to use the code, to modify it, and to pass it along to anyone we want to. And Novell has done distribution of SUSE under the GPL. No one can change that license now.
In Murphy's alternate universe -- which you can enter at http://talkback.zdnet.com/5208-11202-0.html?forumID=1&threadID=78216&messageID=1516427 -- here's more on the evil plan:
Imagine that Ms (or HP or Intel or any other leading wintel seller) owned both Novell's SuSe business and copyrights to IP in it that could reasonably be said to date back to 7.3. That company could then stop pretending to sell free licenses with paid support and instead sell real licenses with the option of paid support.
Why do people have these ugly fantasies about tying Linux to the railroad tracks?
Sorry, but they can't do it the way Murphy imagines, because the code is already licensed under the GPL, which forbids any other license on top. You can't sell the license for money. It's tied to the code and you get the license for free with the code. You can charge for the code and/or for support, but it's not allowed to sell the GPL license. Here. Read it for yourself.
They'd pretend all the other IP in the thing was still free, but it's hard to see how that would last very long once the SuSe market did what SCO's lawyers tried to do - value those copyrights for the courts - and allowed this vendor to go after Red Hat first and everybody else later.
I can't believe I'm still teaching classes in remedial GPL after seven years.
It's not rocket science. The GPL doesn't work like a Microsoft EULA. Trust me. Buying Novell and SUSE doesn't let you do what you could do if they were the same, because they are not the same. Here's the bottom line of the GPL:
Greedy people can't misuse GPL'd code with impunity. That's chapter one in GPL for Dummies, if someone ever gets around to writing it.
It's a feature of the license, planned for long before Microsoft or Murphy or SCO tried to find devious ways around it. The author of the GPL figured this day would come, and the prohibitions on bad actors were baked in. I have a question though. Why does ZDNet publish this kind of misinformation?
But, you may ask, didn't MySQL dual license, and sell proprietary as well as under the GPL? Yes, MySQL dual-licensed, but Microsoft/Elliot/whoever can't do that with Linux, even if it bought Novell and SUSE, because they still wouldn't be able to change the GPL license terms, because Novell doesn't own all the copyrights to all the code in SUSE. Not even close. All they could achieve, at most, would be the death of SUSE and the inspiration for some lawsuits against them for violating the GPL. MySQL did own all the copyrights, so it's apples and oranges. You have to be the sole owner or get all the authors to agree before you can dual license like MySQL. I don't think Linus and the gang will agree.
How is it possible that an entity like Elliot, which presumably has lawyers and advisers, can't grasp the GPL? I don't know. Because it interferes with their hopes and dreams, I guess. That's if Murphy is speaking for them or even understands them, which isn't a given.
Novell doesn't own the Linux kernel, period, and so Murphy's plan, while it seems to make him drool, can't come to fruition in real life. I tried to explain all this to SCO, but they never paid attention in class, so they'll have to learn the hard way, which they will, I trust, in the IBM case.
Some in the media have been asking me what's next in the SCO saga. Here's what I tell them. I have no idea what creativity SCO will come up with next. Here's what I do know: it doesn't matter what they try, or even what the judge does next in the Novell case, because the GPL protects Linux. There is no legal way to do what they want to do that I can see.
But SCO lawyers are nothing if not creative. One of their strategies we just saw play out to the end, regarding the slander of title allegation SCO raised against Novell. Some have pondered why SCO brought that claim in the first place. Remember how SCO tried to move to state court instead of federal? There was a reason for all that. And when I was reviewing the old Enderle materials, I came across
the piece Steven J. Vaughan-Nichols did, which included a remarkably insightful explanation from an attorney, who explained why SCO wanted to be in state court. Here's how he explains the impact of that decision to compel SCO to stay in federal court:
Let me introduce you to Michael R. Graham, intellectual property attorney and partner with the Chicago-based law firm Marshall, Gerstein & Borun LLP.
And that is exactly what we just saw play out in Utah in the jury trial, where the jury found that in fact the APA and Amendment 2 did not transfer the copyrights. SCO is now telling the media that the issue in the IBM litigation is contract based, so copyright ownership doesn't matter. But it did and it does, as Mr. Graham pointed out. But the real issue in my mind is, why did SCO back away from a straightforward look at copyright ownership? Was it because they already knew what the jury in Utah has just told us, that SCO didn't own the copyrights?
Graham tells me, "Judge Kimball's decision is a serious loss for SCO. Not only in its slander of title case against Novell, but in SCO's case against IBM. The threshold issue in both cases is whether SCO owns copyright in the Unix software code."
"SCO wanted the case remanded so that the only issue would be contractual: whether the APA [Asset Purchase Agreement] and Amendment No. 2 transferred ownership of the Unix code," Graham says.
"But Judge Kimball concluded that a more fundamental issue is whether the APA and Amendment No. 2 constitute the type of 'writing' required under the federal copyright law to effect a transfer of copyright. This federal analysis could prove fatal to SCO's claim."
Why, you ask? Graham explains, "In simplest terms, a state contract analysis would presume that the writing transferred copyright, making the only issue—albeit a complex issue itself—what copyrights were included in a transfer of copyright. This could include testimony and evidence of what the parties thought the agreement transferred."
But that isn't what happened. Instead, Graham continues, "The copyright analysis questions whether the form of the writings and the terms of Amendment No. 2 could even be a 'writing' as required by the Copyright Act and analyzed in case law."
"SCO would have preferred—and could have a better chance of success—if the only issue were the analysis of the terms of the writing. The federal court held, however, that the sufficiency of the writing itself must first be determined," Graham adds.