April 24, 2014


Lessons from the SCO/IBM Dispute

Conventional wisdom about the SCO/IBM dustup is that it demonstrates a serious flaw in the open-source model – an asserted lack of “quality control” on open-source code that leaves end users open to potential copyright and patent infringement suits. If any pimply-faced teenager can contribute code to open-source projects, how can you be sure that that code isn’t copyrighted or patented by somebody?

SCO charges that IBM took code from a SCO-owned version of Unix and copied it into the open-source Linux operating system, in violation of a contract between IBM and SCO. There is also some ambiguous evidence that SCO may own copyrights on some of the allegedly-copied code, in which case IBM might be liable for copyright infringement.

It may well turn out that SCO’s claims are hooey, in which case the only lesson to be learned is that we shouldn’t take the claims of desperate companies too seriously. But let’s assume, just for the sake of argument, that SCO is right, and that IBM, in violation of contracts and copyrights, did copy code without permission into Linux. What lesson do these (hypothetical) facts have to teach?

Assuming that SCO’s charges are correct, the moral of the story is not, as the conventional wisdom would have it, to avoid software that comes from pimply-faced teenagers. Quite the contrary. The moral is to be wary of software from big, established companies like IBM. In SCO’s story, the pimply-faced teenagers are bystanders – the gray-haired guys in expensive suits are the crooks.

More likely, though, the fact that SCO’s story involves their code ending up in an open-source IBM product, rather than a closed-source one, is just a red herring. IBM would have had just as large an incentive to copy code into a closed-source product, and doing so would have reduced the chance of getting caught. Nobody has offered a plausible reason why the open-source nature of the end product matters.

Now let’s turn to SCO’s argument that ordinary Linux users might be liable for infringing SCO’s copyrights, even if they didn’t know that Linux contained SCO’s code. It’s hard to see how the merits of this argument depend on the fact that Linux is open-source. SCO’s arguments would seem to apply just as well to customers who made copies of closed-source IBM products (presumably, with IBM’s permission but without SCO’s). Once again, the open-source issue seems to be irrelevant.

Now it may well be that open-source products are more prone to copyright infringement or patent infringement than closed-source products. That’s an important question; but I don’t see how the SCO/IBM dispute will help us answer it.