Bradley M. Kuhn

Two Blog Posts Disguised as Mailing List Posts

Bradley M. Kuhn at

There are plenty of mailing list threads to read, and I don't actually recommend the one that I'm talking about. I think it went on too long, was far too “ad hominem” rather than real policy. Somewhere beneath the surface there was a policy discussion being shouted down; if you look close, you can find find it underneath.

As he always does, Jon Corbet did an excellent job finding the real policy details in the “GPL defence” ksummit-discuss thread, and telling us all about it. I am very hard on tech journalism, but when it comes to reporting on Linux specifically, Jon and his colleagues at have been, for nearly two decades, always been real, detailed, and balanced (and not in the Fox News way) tech journalism.

The main reason I made this blog post about it, though, is that I actually spent as much time on a few of my posts on the list as I would on any blog post, and I thought readers of my blog might want the content here. So I link to two posts in the thread that I encourage you to read. I also encourage you to read these two posts that my boss at my day job, Karen Sandler, made, which I think are very good as well.

And, to quote the fictional Forrest Gump: "That's all I have to say about that."

der.hans, Jason Self, Charles Stanhope, Claes Wallin (韋嘉誠) and 3 others likes this.

der.hans, Sarah Elkins, Stephen Michael Kellat, Claes Wallin (韋嘉誠) and 4 others shared this.

Goodness is that a messed up flame war. Nobody hear of estoppel much?

Stephen Michael Kellat at 2016-09-02T23:10:41Z

Claes Wallin (韋嘉誠) likes this.

... or basic incentive structures.

Claes Wallin (韋嘉誠) at 2016-09-03T09:48:19Z

I really think the biggest impact of the GPL have been almost entirely outside any legal issues, and that it's a great document not because it's a great legal piece of writing, but because of much bigger issues. That may be why I reacted so negatively to seeing it argued that it's pointless without enforcement. Almost none of the successes of the GPL have ever been about the legal side.
-- Linus,

So basically: "[The release of the hostages was all our doing as brilliant negotiators, and not in any way related to the fact that the perpetrators  knew there were snipers outside the building, watching their every move.]"
Let's talk about the lies spread by Bradley Kuhn, and about the projects where the SFLC and SFC killed them and salted the earth with their "help".
-- Linus,

Assuming this assessment of the Busybox outcome is correct (a big assumption), "salting the earth" is not necessarily a bad thing. It makes me think of the recent Bayesian Conspiracy episode on game theory.

In the Hawk-Dove game*, in a field of Hawks, Dove is the winning strategy, because Hawks are likely to fight other Hawks and get nothing. We don't want a field where all the Hawks are GPL violators and none of the Hawks are GPL enforcers. We want potential violators to see Dove as the winning strategy. All Doves would be nice, but we know from Brad's and Harald's enforcement backlogs that's not where we are.

Evolved cooperative species spend enormous amounts of effort punishing bad behavior, even when the outcome for the enforcing individual is a net negative. @evan and/or @mlinksva have also talked about how social movements need the full "radical fringe" to "pragmatic moderate" spectrum to win.**


 * Hawks vs Doves:
H vs D, H gets +5, D gets +1
H vs H both get 0
D vs D both get +3

The Violator/Cooperator-Enforcer/Cooperator game is different of course, because two Violators or two Enforcers never fight (real-life Violators may fight of course, but that's outside the boundaries of this game, and maybe Cooperator vs Enforcer doesn't happen, but let's include it as Enforcers may make mistakes, and proto-Doves may look like Hawks). Also Violator vs Enforcer action has externalities on Cooperators etc. But I think people can still follow the analogy. :-D

** Of course, the list thread may be the display of a theory that "pragmatic moderates" need to publicly denounce the "radical fringe" in order for the dynamic to be maximally effective.

Claes Wallin (韋嘉誠) at 2016-09-04T07:15:00Z

AJ Jordan, likes this.

I haven't read everything, but now I got to this part. I think this is the killer argument:
One reason (among many) we should bring lawsuits in those rare cases is to show everyone the 99.43% that they are much better off working with the community. We have plenty of evidence from company representatives who say clearly: "I can't get my company to comply unless the threat of lawsuit is realistic; we agree it's totally reasonable to sue the very few bad actors". Greg, representatives of some the same companies that I know you've worked with to improve compliance have told me and Karen that directly.
-- @bkuhn,

Everyone in the discussion seem to agree that there needs to be a last resort, and the disagreement is when it should come, but less than a handful of "last resort" instances in the 25 years of Linux history really isn't in any way a disproportional response, given the endless stream of violations.

I guess the difference is whether we trust you when you say that you tried really hard over the course of years to be the "good cop". I don't know if it's possible to convince Linus and Greg that you did. On the other hand, I'm not sure it matters.

I guess that means after a long flamewar, status quo reigns. Some people think this was a "last resort" moment, some people don't, almost every think they know one when they see one, but most people weren't even there.

Still, I'm glad you see it as your responsibility to be Kibo when the community talks about your actions. Otherwise, who would?

Claes Wallin (韋嘉誠) at 2016-09-04T11:03:32Z