Stallman is a good example of the old "a stopped clock is right twice a day" thing. More interestingly to me, a stopped clock is precisely right in a way that a

Stallman is a good example of the old "a stopped clock is right twice a day" thing. More interestingly to me, a stopped clock is precisely right in a way that a running clock never will be, for those two moments. But it isn't usefully right. A clock that is off by half a minute is useful all day for most applications. A stopped clock is useful only, maybe, for calibration but even then you have to wait half a day to get the chance. Stallman is precisely right in a few very specific ways, but it's almost impossible to apply that precision to any real world problems we face. We can calibrate against it, and there's some value in that. But not that much I find.