> A group will never admit they were wrong. A group will never admit, “We made a mistake,” because a group that tries to change its mind falls apart. I’m hard pressed to find examples in history of large groups that said, “We thought A, but the answer’s actually B.”

Is this supposed to be satire?

The Linux kernel mailing list is full of admissions from most of the subsystem maintainers talking about past failures. In fact, their strongest arguments for avoiding new commits is essentially that the changes smell like some of their past failures. Look at the kdbus thread. I'm certain I've also seen some ranting about untangling code related to CORBA as well.

Frankly, if an organization were more critical of ongoing fuckups and failures than the Linux maintainers, you'd have to question their core competency.

And Debian! What about the Debian dev who removed entropy from their key generator? Again, if Debian had a more egregious admission of systemic failure it probably wouldn't exist as a distro today.

Even cranks on HN don't attempt a steel man that fucking up the GPG key generation somehow wasn't a failure. Instead, cranks speculate about whether some super secret spy was recruited to introduce a "subtle" change that would have been laughed out of the room if anyone had bothered to directly ask an openssl dev about it.

Now let's commence HN cranks attempting to steel man a novel argument that two of the most prominently discussed groups on HN aren't actually groups.

Edit: also notice that Debian did not fork after the GPG key generation/valgrind patch debacle. AFAICT they released a statement saying, "All your Base from the past several years belong to script kiddies." One could argue that they didn't take the right steps to fix their process going forward. But one cannot argue that they didn't admit failure.