Re: Unified BSD?

[steve%prd.co.uk@localhost (Steve Blinkhorn)]

I would just like to add my voice to what Ignatios has written, having
never even heard of MirBSD until today (but having a history of using
BSD releases since the 1970s - on PDP11 and VAX hardware - through the
original BSDI commercial release and then up to BSD/OS 4.3, and then
NetBSD since release 2.0.2.

I too value stability in an OS - they are the platforms on which I
earn a living.   I'm not a hobbyist, I'm not interested in twiddling a
kernel (though I'm grateful to those who are), but I do have a body of
software that I have written or acquired over decades.

I tried Linux - RedHat, Ubuntu and Debian to name but several - but
found the need to constantly update kernels a distraction from doing
paywork.   I also tried FreeBSD (and for that matter an iMac) and
found a similar constant pressure to change kernels (or else some
feature I made regular use of would no longer function properly.
It's not that I think there is no place for any of these; on the
contrary, they have their niches.   But it is of the nature of a niche
to be limited.

I think the ambition to keep NetBSD available on the widest range of
hardware keeps it honest.   It's not that I have no gripes - I have
found major release upgrades problematic from time to time, and as I
recently aired on this list I have had problems with pkgsrc and shared
library versioning.   But a BSD variant like FreeBSD that exploits
just a single architecture is bound to have different priorities from
one like NetBSD which aspires to catholicity.

Others have commented on the licensing issue, which really can affect
one's ability to run a business using software as a platform for
delivery of added value, rather than dealing in software per se.   I
think I understand why the BSD variants split - in a way it's a
tribute to the universality of the concepts underpinning the
general design of Unix-like operating systems.   And there is surely
benefit in not making variants differ for the sake just of trying to
be divergent.   But whatever the surface politics and personalities,
systems mostly diverge in the way species do in the natural world, and
for the same sorts of reasons.   Some live, some die, some thrive, some
starve.

Would the world be a better place if there were just one breed of
cattle or sheep, or more ambitiously just one species of mammal?   If
we end up with the latter, it will be us, and that may not be an
improvement.

--
Steve Blinkhorn <steve%prd.co.uk@localhost>
You wrote:
> 
> On Tue, Nov 13, 2012 at 01:04:27PM +0100, Lars Engels wrote:
> 
> > MirBSD / MirOS is dead:
> > 
> > http://www.freshbsd.org/search?project=mirbsd
> > 
> > Last commit:  2011-08-29 23:00:00
> 
> I'm no Mir* co-worker, so take this with a grain of salt. But on
> general principles:
> 
> a) I question the date itself - that's the last commit to whatever
> freshbsd.org watches, not necessarily the last thing the developers 
> did.
> 
> (In fact, I've heard from Thorsten at FrosCon that he does definitely
> not consider his project abandoned.)
> 
> b) Besides - I question the notion of "unchanging" == "dead". In
> fact, as somebody who *uses* software, and who administeres computers
> for others who want to *use* the software, I consider changing
> software - e.g. the fortnightly changes of Firefox-Current's user
> interface - a nuisance. (That's why Mozilla has their "extended
> support release", currently 10.0.9.) People want to use software for
> some work, not spend half of their time rewriting configuration
> files or relearn key bidings or menu entry positions.
> 
> (Now, nobody being there who looks at bug reports etc... thats
> something different. But you only see changes through this activity
> if there really *are* bugs.)
> 
>       -is
>