@justine I dunno.

AmiTCP (derived from BSD/CSRG) was something I ran at home when I wasn't using the SunOS, IRIX, (also CSRG/BSD derived) UNIX systems at nps.navy.mil & elsewhere in my youth.

To me Linux always felt as if it was a bad imitation of UNIX. UNIX already felt as if it was long in the tooth, but if you were going to use it, use IT, not a knock off?

Andrew S. Tanenbaum chastising Torvalds in the early 1990s on Usenet, from my vantage, seemed sound too. Similar to John Gilmore, decrying the use of SHA-1 in git, when SHA-1 already had known failings.

I don't think BSDs are forbidden fruit? They never were & it's difficult for me to wrap my head around that perspective.

BSDs are pretty well established & tend to have a higher attention to detail in my experience & less of a haphazard "throw it.sh at the wall & see what sticks" but I think if we, as a society, were really trying to move the needle forward in OS design, we would probably be using @sel4 with nixpkgs as a userland or something to get a svelte microkernel with a gargantuan ports/contrib realm?

IMHO, if it weren't for the USL/AT&T vs BSDi lawsuit, Linux probably never would have become popular.

A lot of shops that were using BSD derived UNIX variants, got scared that they had put all their eggs into some basket that a convicted monopoly was trying to smash. GCC on Linux could compile sufficient amounts of code in the early 1990s when that was happening that it seemed as if it might be a viable escape route.

SGI for example, devolved from IRIX (their own BSD derived UNIX variant) to becoming a RHEL VAR essentially. ;( IMHO, XFS is still more mature than ext* & Linux FSes.

To me, forbidden fruit would probably be more in the realms of exokernels such as xok. Fascinating research! While writing different filesystems on contiguous blocks on disk, seems kind of "gee whiz" fascinating & might have some interesting anti-forensics properties, disks fail, a LOT, nightmare recoverability.