31 Aug
2024
31 Aug
'24
10:54 a.m.
On Fri, Aug 30, 2024 at 05:44:20PM -0600, Warner Losh wrote:
On Fri, Aug 30, 2024 at 5:40???PM Kevin Bowling
<kevin.bowling(a)kev009.com>
wrote:
Pretty sure this was BSDi and had a bunch of the CSRG people trying to
make money off of BSD. There was much drama and it didn't go anywhere.
Sun was pretty much the only company to make money off of BSD and it
was because SunOS was famously known as "a bug fixed, and completed,
BSD". wnj had the mmap docs somewhere, he had the idea, Sun actually
implemented it and realized that a page cache and a buffer cache are
two caches of the same thing (read/write went through the buffer cache,
mmap went through the page cache) and unified them. Killed the buffer
cache except for directories and maybe inodes. Took the other Unix
implementations at least a decade to catch up. Ask me about HP-UX.
BSDi never did anything like what Sun did but they wanted the money,
hence the drama.
I have been playing around a bit with this in
VirtualBox.
Maybe due to the backing company, I had assumed it was a commercial
FreeBSD variant. But looking a bit harder, it seems like it was a distinct
strain of 386bsd like NetBSD and FreeBSD. There seems to be scant
information about it online. Does anyone know if its story is told
somewhere?
Wasn't this from BSDi? https://en.wikipedia.org/wiki/BSD/OS It wasn't a
commercial version of 386bsd, per se, but based on a porting to the PC work
Jolitz had done prior to 386BSD. They shared a common (very recent)
ancestor. And there was much drama around it all.