FreeBSD 13.0 released

A lot of this is just the inherent fragility of systems too complex and with far too many code paths to test, of course. Heck I just tried to build a kernel on a (Linux) system I thought I hadn't changed at all and, oops! cordumps all over the place, because of a change I was sure couldn't affect anything and thus had forgotten about. This has been routine for *decades*, so I suppose I shouldn't be surprised that systems that date from long ago, with the cavalier error handling typical of back then (and here I include C itself), are a source of considerable trouble at this date.

(I'd say "Lisp Machines were never this painful", but oh yes they were whenever anything went wrong enough. Just in different ways.)