Taking IBM to Task on Patents: What's Useful and What's Not?
IBM has made lots of friends in the open source community by pledging a fraction of its patent portfolio to the defense of open source projects, and by pouring billions of dollars into development of open source and marketing Linux and other open source solutions. Does that give IBM a free pass to attack an open source project, or are FOSS advocates justified in turning on Big Blue for one action against a history of open source support?
IBM is one of a a handful of companies that have committed patents to the Open Invention Network (OIN) and promised not to go after open source projects that may be infringing on those patents. Out of more than 40,000 patents, IBM has committed 500. Now it's going after an open source project (the Hercules mainframe emulator) using patent claims to attempt to defend its own mainframe business.
Simon Phipps asks if this is a sign that IBM has gotten tired of FOSS:
Regardless of the merits of IBM's case against TurboHercules, the fact the incident has happened at all is an important signal. I can’t for a moment believe this is the first time since IBM's patent pledge that any part of the company has wanted to act against a community participant. We can see the tension between the statement Dan Frye makes through the Linux Foundation and the statement of another IBM spokesperson in the WSJ attempting to say the Pledge doesn't apply to everyone. To hazard a guess, the competition is now characterised by Google -- a huge user of and contributor to open source software -- instead of IBM's old foes, Microsoft and Solaris.
IBM doesn’t seem to need the FOSS community as a stick to beat its foes any more. This action tells us that there is now no FOSS advocacy function at IBM with the authority to veto actions against open source. I hope IBM will categorically say I am wrong about this, but until they do all of us need to take note of this development.
Phipps is absolutely right that this is something to take note of, but it seems a bit over the top to declare it shows there's no FOSS advocacy that can veto actions against open source. Perhaps that's true, but this is (so far) a solitary action by IBM and not necessarily a predictor for how IBM will act in the future in other situations.
The larger question is whether it's reasonable to expect a company to go against its own interests in the support of FOSS at all times in order for the company to receive credit as a supporter of FOSS. The open source community is very quick to notice when a company breaks ranks, even if the bulk of the company's actions are supportive. That stance can make it much harder to advocate good FOSS citizenship within an organization, if all it takes is one wrong move to lose the organization's community goodwill.
Perhaps, given IBM's longer track record of supporting FOSS, it would be more appropriate to reserve judgment while indicating that this sort of thing isn't acceptable. Give IBM a bit of room to revise its position rather than taking a hostile position right away, and a bit of credit for what it has done right. It's a two-way street, and expecting IBM to be loyal to open source unconditionally even when it's against its business interests doesn't seem in line with casting the company as the villain as soon as it takes an action against a single project. This is not to say IBM should get a "free pass" for using patents against open source projects, but it's not a black and white issue.