[TUHS] Is there a good, even definitive, list of reimplementations of the Unix kernel? What would good cut-off criteria be?
Warner Losh via TUHS
tuhs at tuhs.org
Fri Apr 24 23:46:58 AEST 2026
On Fri, Apr 24, 2026 at 4:02 AM Steve Jenkin via TUHS <tuhs at tuhs.org> wrote:
> Adam,
>
> Great catch...
>
> It appears Bill Jolitz claimed he rewrote 386BSD 'from scratch' after
> deleting all UCB/BSD code and the final 386BSD was unencumbered, ie not
> from USL or UCB/BSDI.
> Perhaps he wrote it in a year, unclear.
>
> Notably, the claim 386BSD was 'not encumbered' ( a reimplementation )
> wasn't refute in the next issue by BSDi,
> who claimed ownership of the 1st version of 386BSD by Jolitz, leading to
> his departure.
>
A while later, FreeBSD and NetBSD based their initial trees on 386BSD +
patchkit. As part of a settlement with AT&T,
they switched to 4.4BSD-Lite that had a number of files removed (or stubbed
out, like physread), and several more
with AT&T copyrights. The missing files were then re-written. There's no
doubt Jolitz wrote a bunch of stuff, but it
wasn't sufficient to keep the AT&T lawyers from calling. While largely a
moot point today, 386BSD has the same legal
problems that FreeBSD 1 had. From a quick grep, it was a few dozen
functions.
> Jolitz details what code he wrote, for whom and when, creating two
> software artefacts designed by him & named '386BSD'.
Yes. The dispute with BSDi and the USL encomberance were different things
though. IIRC, he rewrote all the 386 support code so that it didn't fall
under the terms of his BSDi contracts (as well as adding the missing bits
from net/2).
Warner
> Sorry I can't give you details on SMP...
> Wasn't that done by a group at Bell Labs before V6?
> Thought I read a comment on-list, perhaps in this thread.
>
> cheers!
> steve
>
> > On 22 Apr 2026, at 03:29, Adam Koszek via TUHS <tuhs at tuhs.org> wrote:
> >
> > The Wikipedia says that Jolix could be another. Looks like they had the
> copyright on it?
> >
> > I’d be very interested in hearing more about commercial efforts - AIX,
> IRIX, Solaris.
> >
> > The part that always felt interesting was what cool things happened
> behind the corporate door - which UNIX got to run on 2 CPUs first, and then
> how it all got scaled up, who got SMP networking first.
> >
> > I also remember from MeetBSD that folks had good words about Solaris
> efforts to refactor things around VFS, buffer cache. Perhaps many more
> ideas.
> >
> > Adam
>
> ===============
>
> Wikipedia source:
>
> Unix Labs' Berkeley Software Design Suit Finds Berkeley University
> in Disarray
> By Maureen O'Gara
> Computergram
> August 6 1992
> <
> https://web.archive.org/web/20210903181408/https://www.tech-insider.org/usl-v-bsdi-ucb/research/1992/0806.html
> >
>
> [ points to Unigram 'interview the previous week', text near
> identical to Unigram 396 below ]
>
> ===============
>
> Unigram / X Number 396
> August 3-7 1992
> Page 4
>
> 100,000 USERS HAVE YET ANOTHER BERKELEY VARIANT - 386BSD...
> by Maureen O’ Gara
>
> <
> https://archive.org/details/UnigramX1992366-416/page/n213/mode/1up>
>
> [ selected extracts ]
>
> The man who says he named both pieces of software is former 386BSD
> project leader and principal developer of BSD 2.8 and 2.9, Bill Jolitz.
> Jolitz reportedly mortgaged his house to start the initial 386BSD
> project and subsequently finished it in his own time.
>
> The code and its rationale were published over the course of a
> year in Dr Dobb’s Journal beginning in January of 1991.
> It was also picked up by Dr Dobbs’ sister publication Unix Magazin
> in Germany.
>
> The full code has been available on InterNet for the last two
> months and was to go on CompuServe last week,
> according to Dr Dobbs’ editor Jonathan Erickson.
>
> Jolitz, interviewed by Unigram.X last week, says that his 386BSD,
> at least in its initial versions,
> was encumbered.
> [ no interview with Jolitz is in the 1992 copies of
> Unigram.X that are saved. O'Gara only said 'an interview', not that it was
> published. ]
>
> He also says that 386BSD is the basis of BSDI’s BSD/386 which he
> worked on in 1991 at CSRG
> initially under the financial sponsorship of UUNet
> Technologies.
>
> 386BSD was Originally intended to be "a university curiosity,”
> Jolitz said, a non-commercial, non-industrial strength way
> for students, facility and researchers to have access to
> Berkeley code on inexpensive machines.
>
> Increasingly through last year it became apparent that what CSRG
> wanted was “basically the same thing as BSDI":
> an unencumbered commercial system.
>
> He broke with BSDI in November, he says, ...
>
> He subsequently received letters from CSRG and university counsel
> claiming that all the work he had contributed to Berkeley
> since NET2 was “University proprietary,"
> a phrase he had never heard before.
>
> In November he was asked to destroy all his own work and anything
> in his possession having to do with Berkeley or 386.
>
> He says he complied and rewrote the current 386BSD Release 0.0
> from scratch.
>
> He says he receives no money from BSDI for his code
> though he alleges BSDI has told its customers that he
> does.
>
> Jolitz does not believe NET2 is encumbered.
>
> ===============
>
> Unigram / X Number 397
> August 10-14 1992
>
> Page 6 [ last page, unnumbered ]
> <https://archive.org/details/UnigramX1992366-416/page/n223/mode/1up>
>
> Unix System Labs is apparently looking at the Tiny 386BSD boot disk
> (the one that’s been circulating in the academia, which derives
> from Bill Jolitz’s work and is not from Berkeley System Design Inc)
> checking that it doesn’t infringe their property rights.
>
> Meanwhile, BSDI’s only reaction so far to our story last week is to claim
> Bill Jolitz was a founder of BSDI.
>
> ===============
> --
> Steve Jenkin, IT Systems and Design
> 0412 786 915 (+61 412 786 915)
> PO Box 38, Kippax ACT 2615, AUSTRALIA
>
> mailto:sjenkin at canb.auug.org.au http://members.tip.net.au/~sjenkin
>
>
More information about the TUHS
mailing list