Except GP wasn't making any of the points you hand-waved away with "who gives a crap" (hint: all the people who complain about those things "give a crap"). They were saying that the fact we even care the GitHub went down at all is a problem with how much we've built our workflows around GitHub when Git was specifically designed to avoid this problem.
Linux kernel development doesn't shut down if kernel.org is down (heck, when kernel.org was completely pwned it only delayed the kernel release by a week or two). People still send each other emails and even if a maintainer is AWOL, other maintainers will pick patches and route them through their own trees. The only problem with the kernel development system is that it isn't easy to on-board people. But if we had a better UX for this system it would be far superior in every respect. There was a recent post by a kernel dev on how this might be improved by building on top of Secure Scuttlebutt and having a nice implementation on top[1].
Linux kernel development doesn't shut down if kernel.org is down (heck, when kernel.org was completely pwned it only delayed the kernel release by a week or two). People still send each other emails and even if a maintainer is AWOL, other maintainers will pick patches and route them through their own trees. The only problem with the kernel development system is that it isn't easy to on-board people. But if we had a better UX for this system it would be far superior in every respect. There was a recent post by a kernel dev on how this might be improved by building on top of Secure Scuttlebutt and having a nice implementation on top[1].
[1]: https://people.kernel.org/monsieuricon/patches-carved-into-d...