[conspire] (forw) Re: Someone needs help in SF

[Rick Moen]

Forwarded with identifying information stripped, since my correspondent 
was writing privately.

On the other hand, if y'all know of any SCO consultants in this area,
please let me know off-list, and I'll pass it along.


----- Forwarded message from Rick Moen <rick at linuxmafia.com> -----

Date: Tue, 1 Nov 2005 11:40:03 -0800
From: Rick Moen <rick at linuxmafia.com>
To: [redacted]
Subject: Re: Someone needs help in SF

[quoted text redacted]


Being a bit too busy with other work, I cannot take on this job, and I'm
not sure I could give you assistance of professional quality.
Explaining why requires some background.


There have been two main SCO OS-product lines.  The older is a kind of
Unix clone originally written for the 80286, and which was for most of
its life named XENIX -- as a sideline of Microsoft Corp., which then
spun that division off as a new firm called Santa Cruz Operation aka SCO
(of Santa Cruz, CA).  Recent XENIX releases are called "SCO Open
Server".  It was always a very peculiar (i.e, non-standard, i.e.,
something you have to specialise in before you can deal with it in
confidence) offering.  A lot of low-cost business management /
accounting / invoicing systems were written for XENIX in the 1980s, in
part because everything was cheap and it ran on low-end hardware:
Unlike most other Unix-like OSes of the day, it ran on commodity x86
clones.

The other SCO-marketed OS is "SCO Open UNIX", which started out life as
AT&T UNIX System V.  AT&T unloaded that declining property to Novell,
which renamed it Novell UnixWare, failed to make much of it, and
unloaded certain rights -- the nature of those rights being thrashed out
in an ongoing copyright and contract lawsuit -- to Santa Cruz Operation,
Inc., publishers of Open Server ne XENIX.  SCO of California offered
"SCO UnixWare" for a while, and then decided to exit the OS business
entirely, shrink itself radically in head count, and concentrate on Web
services and remote application access software.  The firm renamed
itself to Tarantella, Inc., which earlier this year was bought and
devoured by Sun Microsystems.

SCO of California, when it decided to exit the OS business,
sold off Open Server (ne XENIX) and whatever rights it possessed to
UnixWare to a _Utah_ firm that was then called Caldera Systems, Inc.,
who at the time published a Linux distribution called Caldera OpenLinux.
Caldera Systems then renamed itself to "SCO Group, Inc."  Thus, one must
distingusih between the earlier, California SCO company with the more
recent Utah one.  

SCO of Utah has offered since then three OSes:  SCO Linux (formerly
Caldera Open Linux), SCO Open Server (ex-XENIX), and SCO Open Unix
(ex-AT&T UNIX System V) -- and has been famously unsuccessful in its
remaining OS business, in which it tried and mostly failed to strong-arm
its legacy Open Server customer base into upgrading to current SCO Open
Unix offerings.  It was approximately when that tactic failed that the
Utah company underwent a radical changeover of management.  New
management refocussed the company on trying to coerce licence fees from
Linux-using companies through litigation.  They also suspended offerings
of SCO Linux indefinitely as of 2003-05-15.

Which brings us up to date.


When someone talks about a potential client "running SCO", that leaves
ambiguous whether they're talking about the ex-XENIX product or the
ex-System V product.  Most often, it'll turn out that the customer
himself/herself doesn't know:  To those folks, it's just the black-box
business-management system.  They typically don't have documentation of
the setup; it was just something sold as a turnkey system by some
long-vanished guy who offered a complete system to run (e.g.) a
car-service garage and filling station.

I'm betting that 90%+ of the potential clients described as "running
SCO" will turn out, upon examination, to be running some decrepit
installation of the ex-XENIX OS.  10%, you might luck out and it's 
System V, which would inherently be less of a horror -- but then there's
the fact that historical Unix printing subsystems were always pretty
awful messes, right up until the spread of CUPS -- pioneered on Linux
and BSD systems, and then picked up Apple for OS X -- fixed the problem
and made them somewhat reasonable.  

You -could- end up finding a reasonable working situation and a setup
you can study and understand, and give the customer good value -- but
(going solely by how you described it) it's exactly the sort of job 
I would not want to get involved in at all.  One of my bywords about
accepting or not accepting jobs is the "You touch it; you buy it" rule:  
If you accept and start work on a job, you're ethically obliged to
pursue it until you and the customer are happy with your work (or it has
to be ended for other reasons).  And another rule is that you will tend
to get associated, in customers' minds, with the sort of technology
you're willing to work on.  If it's failing, decrepit technology, and
your work props it up and makes it less sucky, you'll end up being 
associated in the client's mind with failing, decrepit technology, even
though that's not fair.

SCO's ex-XENIX installations, and even probably the minority of
ex-System V ones, can I think fairly be written off as _in generl_ 
failing, decrepit techology in 2005.  I wouldn't want to touch one, 
specifically because then I would "own" it in the ethical-obligation
sense, and I'd simply rather not go there.  And, alas, I know of nobody 
I could refer you to, for that job.


----- End forwarded message -----