elsewhere, could you please check what you can build if you remove all glibc and kernel headers containing the BSD license notice?
Also, please remove sudo, ssh, mandoc, the groff mandoc, me, ms, and other BSD-derived macros, mdoc, tcl & tk, ldap, libtirpc, and any other permissively-licensed packages containing code copyright by "the Regents of the University of California."
Also post the output of these commands:
;-)Code:head -n 30 /usr/include/linux/quota.h modinfo aes_generic
2) BSD is permissive, but not unrestricted like PD. Removing or changing the license and copyright notice is not permitted.
3) In theory, I'd be inclined to agree with the "never said if it's bad, but it's sure bad for a company" (to distribute their own software under).
In practice, I'm more dubious...was it RHEL+Fedora and RHEL clones, OpenMoko, or Android that had more systems deployed?
OK, that may be unfair. But an observation of all FOSS indicates more that success is sporadic than that BSD is worse than GPL in practical terms.
And a look at the features of Linux vs BSD seems to me to indicate that Linux has a superiority in some areas to BSD, but I don't see anything earthshaking and I have on occasion seen areas where BSD seemed better. Not positive what all of them were, but a few I can think of...NetBSD appeared more responsive under load than Linux (single core PIII, 2 packages building, desktop still seems to be perfectly responsive); the state of ZFS (some time ago); BSD had pf jit on x86 and amd64 long before Linux BPF got a JIT for amd64 (afaict, there is no bpf jit for 32bit x86 linux); crunchgen; more POSIX-conformant tty support (glibc on linux still doesn't support one of the flags mandated by POSIX, which will reset state to sane defaults)--sorry I don't have more details, I read about this a few months back; and kernel ABI stability.
4) What do you mean "BSD is not a stable spot to build upon"?