[Info-vax] What is a "real" Unix ?

[Dan Cross]

In article <klmuo1F5o2aU2 at mid.individual.net>,
Bob Eager  <news0009 at eager.cx> wrote:
>On Mon, 04 Sep 2023 12:15:22 +0000, Simon Clubley wrote:
>> On 2023-09-02, Bob Eager <news0009 at eager.cx> wrote:
>>> "Jumped uo UNIX wannabe"
>> 
>> In that case, what is a "real" Unix ?
>
>I was talking about the legal definition; those with an actual UNIX 
>licence. 

That's not the real measure anymore either, though.  Unix
licenses haven't really been a driving force in a long, long
time.  As commercial Unix continues its slide from relevance,
the licensed code just doesn't matter.

The thing that matters now is SUS and POSIX conformance, as
measured by the Open Group, who own the Unix trademark and
certification program.  To be properly called Unix, one needs
to meet the certification criteria; whether the source code
is licensed from one entity or another is irrelevant.

>You can also argue that some implementations using the Linux kernel ahave 
>now diverged a long way.

One could make that argument about any of the traditional Unix
distributions from the commercial vendors.  AIX didn't look very
much like 6th Edition, for example, the same with NeXTStep back
in the late 80s/early 90s, as well, even though that contained
licensed code from e.g. BSD.

>I would argue that systemd does not conform to 
>the conventional UNIX model.

I would concur in that judgement, but I don't think it much
matters.  Quite frankly, Linux has grown large and important
enough to not have to care.  They are the descendents for all
intents and purposes, and the "No true Scotsman" arguments about
what is, or is not, Unix are not worth squabbling about for
them.

	- Dan C.