Enough of intellectual property (IP) issues, and who actually
wrote what in Linux. Let's get down to brass tacks. You're
selling or working with Linux, SCO
OpenServer, or UnixWare.
What the heck do you do now?
First, if you're using Linux, should you be seriously concerned
about having SCO's lawyers coming to your door? The answer
is no. Right now, all the analysts, not to mention SCO, are
telling you to get your attorney's opinion.
It's hard to argue with that idea, but really unless you've
got top-flight lawyers at your beck and call, will it do you
any good? Does your law firm or in-house counsel know IP law?
If you're with a big company, they probably do. But, most
business attorneys know contract or tax law, not IP law.
In my opinion, though, you should stick with Linux. Let's
ignore SCO's threatening to sue Linus Torvalds, Novell's sword-rattling,
and the screams of outrage from Linux fans everywhere. There's
only one major law suit now. What does SCO really need to
do to prove its case against IBM?
I am no lawyer, but the core of the case to me is not whether
there is Unix code inside of Linux. I'm sure there are hundreds
of lines of identical code in each. I mean, how many efficient
ways can you write 'hello world' or open a file in C anyway?
No, the real heart of the matter, as far as Linux goes, is
that IBM stole code from Unix. And, while a jury may not understand
that, there still remains the vital question of who put the
code in there in the fist place. I simply don't see how IBM
can be found guilty of that beyond a reasonable doubt
by any court.
SCO, as I explained last month, both before and after the merger,
was itself blending Unix and Linux. The record is clear.
For years, SCO was trying to make a best-of-breed Linux operating
system with contributions from Unix. Thus even if there has been direct copying, which remains
without a shred of publicly available evidence, trying to
prove that IBM put
the "illegal" code in is impossible. It's like trying to decide
the fate of a single strawberry after dozens have been thrown
into the blender and you've turned it on puree.
On top of this, as late as April, SCO itself was still releasing
the public under the GPL, including by its arguments its own Unix code. Would a court find it reasonable
that a company releasing its own Unix secrets into the world
under a free software license had lost these same IP secrets
because of IBM?
No, just no. Even if you accept all of SCO's arguments, its
history muddies the IP waters to the point that I can't
see the company winning the case. I also think that SCO must know that.
So it came as no surprise to me when CEO Darl McBride indicated
in an interview
that he'd be open to the idea of IBM buying out SCO. I've
thought for a long time now that a buy-out was SCO's only
viable exit strategy.
But what does that mean for you? Well, if you're using Linux,
stick with it. It's not going away.
Using SCO operating systems?
If you're using SCO Linux, I'm sorry but you're hosed. SCO
has promised continued support for its customers. Indeed,
as of June 2, the company
says, "SCO continues to honor our contractual relationships
with customers; and will continue to support our Linux customers."
But, what can that mean, when the company is campaigning against
Linux? That they'll release bug fixes for code which they
claim was stolen from them? It makes my head hurt just thinking
Some SCO resellers have told me that they're being encouraged
by SCO to get SCO Linux customers to switch over to UnixWare.
SCO's corporate communications manager Paul Hatch
tells me that it's not that SCO is encouraging SCO Linux customers
to switch now, it's that SCO is telling them that in the future
their upgrade path will be UnixWare. Neither customers nor
resellers are thrilled. Many resellers have been down this
road before with SCO trying to force OpenServer resellers
into pushing UnixWare. It didn't work then. It won't work
now. UnixWare, while a fine middle-of-the-road server operating
system, has never been popular.
Besides, few identical applications run on both UnixWare
and Linux. Unless, of course, you count the ones that require
the Linux Kernel Personality (LKP), and then we're back in
the same SCO Linux support mess we just left.
What should SCO Linux users do then? Most of the people I've
spoken with are planning to move to Red Hat. Personally, while
I don't think there's anything wrong with Red Hat, I think
anyone who has invested in SCO Linux as a server should
look to SuSE. After all, both SCO Linux and SuSE Enterprise
Linux are based on UnitedLinux, so there's almost no trouble
moving settings and applications between platforms. In short,
it's simply a lot less trouble to move from SCO Linux 4 to
Enterprise Server 8 than it is to move to Red
Hat Enterprise Linux (RHEL) AS or ES.
Personally, I moved my two SCO Linux 4 servers to SLES and
I was done with the transfer, including applications, user
and group settings, and customizations, in one morning.
If you're using OpenServer, especially if you're using OpenServer
in a vertical application, you don't have anything to worry
about for now. The real question is: what do you do in the
years ahead? SCO Linux was always this group's fallback plan
for the day when OpenServer finally ran out of steam. The
OpenServer resellers and users I've spoken to are really worried
about what happens to them in 2004 and beyond. They've got
If SCO goes out of business or is sold to a buyer, say Microsoft,
with no interest whatsoever in OpenServer's continued existence,
they're in a lot of trouble. If SCO ends
up in the hands of IBM, I doubt IBM has any interest in continuing
the product, but I could see it selling OpenServer to another company
or letting it linger in the background the way it did with
OS/2. The latter isn't the greatest of fates, but at least
OpenServer would keep going and Linux would open up again
as the long-term operating system insurance.
As for UnixWare, I don't know. I suspect it may be the operating
system in the most trouble. It never took OpenServer's place
in vertical applications and small business. It also never
really took off as an enterprise operating system. I can see
UnixWare being killed off by any future buyer or, at best,
being left to wither on the vine.
That being the case, if I had UnixWare now, I'd be planning
on migrating sooner than later. The $64,000 question is to
what? Given UnixWare's strong points, I think you'd be best
off with SuSE Enterprise Server, RHEL, or even Windows 2000
Server. Windows 2003 Server? Please! Even if you kind of like Microsoft operating systems, 2003
no server application support.
Unfortunately, none of the operating systems I've mentioned
have one thing going for them: a strong reseller channel.
Oh, Microsoft's can be very, very good, but you really need
to be a big-time dealer or have a very long history with Microsoft
before you reap the benefits. Besides, you're only competiting
with thousands of other integrators and resellers. Finally,
6 can really get in the way of both resellers and end-users.
As someone who's used, liked and recommended in print many
SCO products and services over the years, I hate having come
to these conclusions. But, SCO isn't the company I used to
know, and if your business depends on SCO, you can't afford
to think that it will ever be business as usual with SCO again.