On Thu, Jan 19, 2023 at 7:44 AM Liam Proven <lproven@gmail.com> wrote:

> On Wed, 18 Jan 2023 at 15:56, Ralph Corderoy <ralph@inputplus.co.uk>
> wrote:
> >
> > The fine article ends with
>
> *Heh*
>
> >    ‘Which means that the last officially trademarked commercial UNIX™ is
> >     Apple's macOS 13, which underneath the proprietary GUI layer is
> mostly
> >     an open source OS called Darwin anyway.  The kernel, XNU, is based on
> >     Mach with an in-kernel "Unix server" derived from FreeBSD...’
> >         — https://www.theregister.com/2023/01/17/unix_is_dead/
>
> Indeed it does. I wrote it. Thanks to Arnold for posting it.
>
> I am getting some grief on Twitter too for "omitting" FreeBSD. I
> didn't, but the BSDs don't fit either definition of "Unix". The
> pre-1993 one being "based on AT&T code" -- after all, BSD (4.4 Lite r2
> was it? Before my time!) -- went to a lot of effort to eliminate AT&T
> code. The post-'93 definition, after Novell donated the trademark to
> the Open Group, is "passed Open Group testing". Nobody has the time or
> the inclination to pay, and indeed, why should they?
>

Yes. I think I tweeted that all the BSDs are derived from Unix. This is
very true: V7 was ported to the VAX by AT&T which produced 32V (which had a
lot of different versions, but Berkeley started with an early one, before
V7 was officially released). Berkeley added demand paging to it to create
3BSD, 4BSD, etc. While much of the original AT&T code was re-written, it
was replaced with functionally equivalent code so the system behaved the
same before / after the rewrite. In many cases the rewrite was an
improvement, in some it wasn't. More importantly, though, it retained much
of the structure of the original AT&T code, especailly in the kernel. So if
you understood the BSD code, you'd understand the System V code and vice
versa at least at a gross level. Ditto all the commercial Unixes which
also, btw, rewrote a substantial portion of the original AT&T code. All the
BSDs is derived from Unix, but can't use the Unix trademark... There have
been people that have run the compliance suite over the years against the
different BSD, producing patches that fixed issues (though the last one I
recall was in the early 2000s). But since nobody paid for official
certification, nobody knows how close things are thees days.

A similar level of rewrite and restructuring has happened to Linux over the
years. Outside of drivers, very little remains even from the 2.x days of
Linux, let alone the 1.x or 0.x lines. It too has been rewritten from
having strong assumptions about running on a single CPU to scaling to
thousands of CPUs these days. A similar evolution has happened in the BSDs,
though with different paths taken in the different forks.


> I hope to return to the subject of the BSDs in general on the Reg in
> future.
>

That would be cool...

Warner