On Jan 4, 4:23 pm, "ness...@xxxxxxxxxxxxxxxxxxx"
<ness...@xxxxxxxxxxxxxxxxxxx> wrote:
> <Quote>
> It's going to happen. Some time in 2008, SCO will finally stop
> thrashing on the floor and die.... I'd bet on Novell doing SCO in at
> the District Court with a lead pipe.
Actually, it could even get uglier than that. Remember, IBM is
looking for deeper pockets. They are looking at Bay Star Capital,
Microsoft, and other backers who might have used SCO as a puppet to
avoid liability.
It looks like IBM intends to prove that SCO knowingly filed a
fraudulent lawsuit, that they owned the intellectual property in
question and can prove that the intellectual property belonged to IBM,
and that SCO knew that it belonged to IBM and that they had received
it from IBM.
If IBM can also proved that those backing SCO knew the lawsuit was
fraudulent, they can go after all of those backers as co-conspirators
in a fraud scheme.
> ...SCO's death by suicide--what else would you call suing IBM and other
> Linux-using companies without proof,
Even worse, suing IBM and Linux using companies when you know that
your claims that the only way Linux could have possibly gotten the
contested software was from SCO, when in fact, the testimony shows
that Daryl McBride had been told by his top technical advisors that
SCO had gotten the code from BSD (OSS), and that Linux could have
legally obtained the same software through the same sources. This
means that the charts used by SCO to show that it owned ALL UNIX was
actually willfully fraudulent.
Second, it was SCO that defaulted on their contract with IBM, and it
was only after SCO had failed to deliver on their Unix for Merced
project (Monterey), that IBM started supporting Linux by offering
their scheduler, which dates back to early versions of MVS and CICS.
> or as Novell has shown, without
> even actually owning Unix's IP (intellectual property)?
SCO was hoping that Novell couldn't prove that Novell still owned the
UNIX copyrights. They seem to have "lost" the documentation in which
Novell told them that Novell would only offer SCO unlimited
distribution rights, source code, and derivative rights, but would not
relinquish the UNIX copyrights themselves.
Worse, SCO hadn't even paid their bill. Again, SCO was in default of
their contract, and yet they were claiming that they still owned all
of the rights, including rights explicitly excluded in the contract,
which means that there was additional proof that SCO was engaging in
several counts of fraud and extortion, including securities fraud.
There is a pretty good chance that some top SCO executives could be
going to jail for a "pump and dump" scheme, an extortion scheme, and a
fraud scheme. Their only way to avoid a very long prison sentence
will be to name all of their co-conspirators. If McBride names top
Microsoft executives, and/or corroborates Goldfarb's sworn testimony,
then Ballmer and Gates, along with a few other top Microsoft
executives might be forced to personally cough up a few $billion in
restitution and could even be facing some prison time.
> [Questions:]
> "What will Novell do with Unix?" I don't know, but what I'd like them
> to do is to open-source as much of the code as they can.
It looks like they are leaning in this direction. As has been shown
by the SCO vs IBM case, there really isn't much code that is actually
not already available as Open Source. Novell has already given away
the trademark. Ironically, the big asset might actually be Caldera
Linux. SCO still has contracts with most of the major fast-food
franchises, as well as NASD brokerages, for both servers and cash
registers. These customers still need support, and are still willing
to pay for it.
> There's still some goodness left in Unix that hasn't been duplicated in Linux.
> For example, even generic Unix System 5 Release 5 can handle up to 32
> processors and terabyte-sized files and does extremely well at multi-
> path I/O.
Linux had a problem with that until they started using queues instead
of holding spinlocks. I think it was corrected in Linux kernel
version 2.4.13 or thereabouts, about the time of Red Hat 7.2, and SUSE
8.1.
These days, Linux does very well on large SMP and NUMA systems, such
as HP Superdome servers, Solaris StarFire, and IBM Regatta servers.
The irony is that Linux developers had been supporting grids, and
grids are much more scalable, because you don't memory contention
issues. Modern blade servers running Linux can hold as many as 256
processor cores in a 19 inch rack, and at very low power consumption.
In addition, when these servers are connected to storage array
controllers using Fibre-Channel, Infiniband, or iSCS 10 gigabyte
ethernet interconnects, they can often outperform SMP systems with 1/2
the number of processors, thanks to wide striping, deep caching, and
read-ahead/write-behind.
UNIX is still widely used for database servers, but even then Oracle
now prefers to use Linux, especially Red Hat Linux as their Oracle
server platform, because it's easier to do deep root-cause analysis if
there are performance or reliability problems.
> [But hard to open source all of unix, contains copyrighted code...]
There are very few pieces that are available exclusively as
proprietary copyrighted code.
AT&T did have a few patents, but these have also expired (regex,
streams).
In many cases, the original authors, who wrote the code as class
projects, have also released the same code that Berkeley gave AT&T,
under GPL or BSD licenses. In other cases, functionally correct
implementations, based on public standards published by X/Open and
POSIX, have been implemented as Open Source. Even the authors don't
know if their implementation is the same or similar, but there are 10
other functionally correct implementations available, all different.
One of the fundamental differences between System V UNIX and Linux is
that Linux makes intensive use of the memory management unit. UNIX
uses a shared memory model, where applications write to shared memory
assigned by the kernel and then have the kernel copy it to the
destination application, and then return the buffer to the original
application. Linux often uses the MMU where an application tries to
do memory to memory copy.
> Since Novell is a Linux company, it makes perfect sense to me if they
> were to cherry-pick Unix for its best code and release it to the
> public. If they elect to go this route, I'd expect to see the first
> code appearing within a few months of SCO kicking the bucket.
There is a very good chance that Noorda and others have already
contacted most of the original authors, to get them to release OSS
versions. I'm curious what UNIX would have that Linux didn't already
have, or hadn't already replaced with something better.
The main advantage of OSS is that there is some pretty intense
competition, which means that you frequently get to choose from
several very good and creative implementations. The chance that a
UNIX implementation would be the preferred implementation is probably
not that great. And once you have decided on an implementation,
there are lots of patches and upgrades to improve it.
> As for OpenServer and UnixWare customers, well, good luck, guys...
That's probably the one thing SCO has that is worth owning.
> Where will they go? ... They'll be moving to Linux.
Many of them had already adopted Linux. Remember, Caldera was winning
a bunch of franchise business from SCO, so much that they needed to
purchase the SCO service organization. SCO managed and maintained the
servers remotely, and Caldara added Linux Cash Registers to the
classic UNIX style display in the kitchen.
> That's because
> SCO's resellers are already working with Linux or getting ready to.
SCO has been remarketing Linux as "OpenUnix". Remember, Caldera was
trying to pass the X/Open certification tests in order to be able to
use the Unix trademark. It's possible that they passed, but decided
that the Linux brand name was more valuable. Linux got a lot of press
in 1999, and early 2000, while UNIX was starting to look more like a
legacy system.
> For example, DTR Business Systems, one of SCO's best resellers,
> is still offering SCO Unix products, but it's also a Red Hat partner
> now...
SCO's biggest business is the franchises. Pizza Hut, Burger King,
Taco Bell, and so on. There is a possibility that Pepsico has decided
to get out of the mess by switching to another Linux vendor. It seems
that they were happy with Caldera, but the SCO mess, and bankruptcy
may have soured them on SCO.
> As for SCO's people? Well, I think they still have a few good engineers.
They have a very good and reliable international support
organization. It's the main reason that Caldera tried to buy their
service organization. Caldera deliberatly did not purchase the
software organization, because they didn't want to trigger the clause
in the Microsoft contract that would allow them to renegotiate the
contract.
The problem was that Caldera sold stock and did a stock swap to fund
the purchase, and even though it was impossible to gain controlling
interest in SCO, it was very easy to gain controlling interest in
Caldera. In a proxy fight, Ransom Love was fired, McBride was hired,
and McBride purchased the SCO organization, which meant that Microsoft
was no longer bound by their Xenix agreement, which prevented
Microsoft from reentering the UNIX market.
> Perhaps Novell, which still maintains a presence in Utah,
> would make a good home for them.
> And if any of their sales people were
> actually able to sell OpenServer or UnixWare to a new customer,
> someone should snatch them up immediately.
The sales organization also has value, since they seem to have good
relationships with companies like Pepsico. On the other hand, Novell
has to be very careful not to over-extend itself. As it is, it looks
like there has already been a bit of a proxy war, and several
executives have already been rolled out.
> Anyone who could sell SCO
> Unix to customers in the last couple of years could sell Linux to
> Steve Ballmer.
Keep in mind that SCO already sold UNIX to Ballmer, for a paltry $7
million. Of course, the terms of the agreement were undisclosed, and
it might have been $7 million per year for some unknown period of
time. Perhaps IBM will force disclosure of the details of that little
contract, along with all of the memos and documentation related to the
contract, from the approach to the termination terms.
> SCO executives? Well, as the pages of Groklaw detail, they'll come out
> of it with plump wallets. For example, the
> day before SCO filed for bankruptcy, the company gave Ryan Tibbitts,
> its general counsel, a $50,000 raise and a $50,000 bonus.
That's peanuts. Remember, McBride got about 1 million shares at 75
cents per share. Darcy Mott had about 1/2 million shares at the same
price. Both sold most of the shares when the stock price jumped to
$14 per share, but before it dropped below $7 per share. That's
around $10 million for McBride, and about $5 million for Mott.
> Nice work if you can get it.
>
> I wouldn't worry about SCO's top brass suffering from the company's
> failure. Their golden parachutes were packed and ready to go long
> before the company found itself on its death bed.
Are you kidding, the parachutes were already fully deployed by the
time the public discovered that the charges were without merit.
McBride only got a salary of $30,000 per year, but he got a $10
million paycheck by cashing his options.
> It's a pity, really.
> Caldera and SCO were both grand companies, but their legacy was
> squandered on anti-Linux lawsuits that, in the end, came to nothing
> but tens of millions of wasted dollars.
It wasn't Caldera management, or SCO management that ran the company
into the ground. All of the people who built the company were fired
right before McBride took over. It looks like IBM may be digging up
some financial records of SCO investors, including major holding
companies and institutional investors, to see how they might have
played a role in the proxy fight.
> </Quote>
>
> http://www.linux-watch.com/news/NS6243163353.html
|
|
