Tarantella Inc SCO’s Second Coming

Tarantella Inc SCO’s Second Coming

Unix veteran SCO sells off its Unix products and services to Linux major
Caldera, and attempts to reengineer itself as an application-server company

Every year, in the salubrious Santa Cruz campus of the
University of California, amidst fog covered pine trees, Unix veterans and users
used to gather at what used to be called the SCO forum. This year also they
gathered as usual. Only, the event was no longer the SCO forum. It had become a
plain and simple Forum. Why? SCO–Santa Cruz Operation–had just sold off its
operating systems and professional services division to Caldera.

For those who came in late, SCO is a twenty-year-old software
company that specialized in Unix for Intel-based servers. It had two products in
this space–Unixware at the higher end and OpenServer at the entry level. A
couple of years back, SCO ventured into Web-based application server software
with a product called Tarantella. Briefly, with Tarentella you could make
non-Web-enabled applications available over the Internet, on a Web browser. SCO
has a professional services division. Along with IBM and Intel, It had taken up
an ambitious project, code named Monterey, which initially atleast, was supposed
to provide a common direction to the various Unixes.

Sudden death

1999 produced excellent financial results for SCO, and every
quarter of the year, the company announced record-breaking revenues and profits.
Things were by all counts, on a roll. Come 2000, and the scenario changed
drastically. In the first quarter, though revenues where marginally higher than
the first quarter of 1999, profits declined a wee bit. SCO blamed it on
customers being preoccupied with Y2K concerns. It was in quarter two that the
bottom fell out. Revenues declined to two-thirds of the previous year’s level,
and to add to its woes, the company was saddled with a whopping $14 million
loss. As it is, profits at about 9% would have been a bit of a tightrope walk
for SCO in the high-tech industry. And a 40% loss figure would have been
crippling.

SCO continued to blame Y2K issues and "other market
factors" for the decline, even as it went for internal cost cutting and
restructuring. In March, operations were divided into three divisions: server
software, Tarantella and professional services.

The third quarter brought further bad news. Sales fell
further, as losses increased, wiping out profits from the five previous
profitable quarters.

How did a company that reported record-breaking quarter after
record-breaking quarter suddenly plunge into such a situation? The answer, many
feel, can be summed up in one word–Linux.

The Open Source challenge

1999 and early 2000 was when Linux, that upstart Unix clone
operating system, came sharply into focus. And that was the time when Linux
companies, including Caldera sharply upped the ante in the operating systems as
well as the stock markets. And all of them were playing in the same market that
provided SCO with their bread and butter–Intel-based servers.

Initially, SCO did not appear to take the Open Source
challenge seriously enough. And by the time they started being active in the
Open Source arena, it was perhaps too late to take leadership position there.
The bulk of Linux installations are on Intel platforms, and as more and more
corporates started evaluating and deploying Linux, it quite possibly came mostly
at the cost of SCO’s OpenServer, or even the higher end UnixWare.

There was also talk of a "SCO version" of Linux.
But that never happened.

One for the management books?

Phillip Evans, management consultant and author of the book
"Blown to Bits: The New Economics of Information", was one of the
speakers at the opening day of the Forum. He quoted the example of Sabre, an
airline reservation system. Sabre was owned by American Airlines, which because
of antitrust considerations, had to sell it off. Today, Sabre is much more
valuable in the stock market than the original parent. Evans’ point was that
when faced with the decision, the management should have sold off the airline
(read old economy) and held on to Sabre (new economy, information-oriented
business) instead.

By selling off its operating systems, and deciding to focus
on Tarentella, SCO is attempting to do something similar. Sell off the old (and
potentially less valuable in the future) where there is too much of competition,
and hold on to the new (and more valuable in the future).

If the move pays off, SCO could well become a classic,
earning its place in all management tomes of the future.

The deal

The deal that SCO struck with Caldera is as complex as they
come. SCO sold the server software and professional services division to Caldera,
but retained the intellectual property rights to the software.

In return, SCO will receive 28% of Caldera stock, plus $7
million and two seats on the Caldera board. Caldera will be the exclusive
distributor for SCO’s products. SCO will also continue to receive from Caldera,
revenues on SCO OpenServer, after expenses. As part of the deal, SCO has also
received a loan of $18 million from The Canopy Group, a major stockholder in
Caldera. Doug Michels, CEO, SCO, told Dataquest that this loan was more in the
nature of line of credit for the period when SCO undertakes the transition, and
would be used only if the need arose, and not otherwise.

Subsequently, in keeping with its new mission and product
line, SCO announced plans to rename itself as Tarantella Inc. Just as we were
about to go to press came the announcement that SCO would be reducing its staff
by 190 in preparation for the sale of its two divisions to Caldera.

A complex Web

Competition for the newly formed Tarentella comes mostly from
Citrix, with its Winframe and Metaframe. Is it going to be competition all the
way? Not always, it seems.

In early January 2000, SCO, along with Citrix, Sun, Novell
and two other investment firms, invested $30 million in Caldera. Subsequently,
Ed Iacobucci, chairman and CTO, Citrix, joined the Caldera board. Now, as a
result of the acquisition of SCO’s divisions by Caldera, SCO gets two seats on
the Caldera board, with one of them being Michels. To complicate things further,
Caldera will be the sole distributor for Tarentella. Wonder whether they will
discuss Tarentalla plans and figures at the Caldera board, with Iacobucci
present!

That is not all. During CeBIT 2000 in February, SCO had
signed a deal with SuSE, another Linux vendor, to provide professional services
to SuSE’s customers. Ransom Love, CEO, Caldera, told Dataquest that the
agreement would continue to hold good.

A Tarantella future

In selling off the server software division, SCO seems to
have done the right thing, exiting a declining market, instead of clinging on to
it like many others would have. The question is whether Tarantella can pilot the
company to new heights. The application server market is yet to attain large
volumes, and Tarantella revenues in the third quarter has been a paltry $2.5
million. How this market will develop will hinge to a large extent on how and
how fast the ASP market will develop, and how crowded the market would get.
Right now the ASP market is considered to be hot, with Gartner Group predicting
that the worldwide ASP market would be worth $25 billion by 2004 ($1 billion in
1999). While Tarantella could have the first mover advantage, they need to get
more aggressive to build up revenues and more importantly, profits.

OpenServer revenues ploughed back from Caldera should give
the newly formed company some breathing space, but if the assumption of a
declining market is right, then that breathing space cannot be held for long.

One obvious option would be to add more products to the
existing portfolio. But as of now, the only product being talked about is
Tarantella for NT. And acquisitions, at least at this stage, seems to be absent
from Doug Michel’s ‘things to do’ list.

What does Caldera get?

Other than the two operating systems, Caldera gets instant
access to the infrastructure and facilities that SCO has set up across the
globe. More than the technical team and infrastructure, what could prove more
valuable to Caldera would be the SCO sales setup. At last count, SCO had sales
representatives in eighty-odd countries. Now these are people with experience
selling Unix on Intel platforms, surely a prize catch from Caldera’s point of
view.

Also, SCO’s team of Unix experts would be a welcome
addition to Caldera’s scheme of things. These could bring in immediate revenue
streams that could make Caldera different than the rest of the crop of bleeding
start-ups that dot the stock markets. This could make it more attractive to
investors, and could have a positive effect on its stock, something that Caldera
could do with.

But the prize catch could well turn out to be something else
altogether–a small piece of software that was demonstrated during Forum 2000–the
Linux Kernel Personality.

Linux Kernel Personality

The Linux kernel Personality (LKP) essentially makes UnixWare
"Linux like" in that UnixWare with LKP installed would be able to run
Linux applications without any modifications. In fact, LKP was demonstrated at
Forum 2000 by running a Star Office-based presentation on a UnixWare 7 machine
with the LKP installed.

What does the Linux Kernel Personality achieve? It works both
ways. To the UnixWare market, it brings the ability to run the ever-increasing
crop of applications being created for Linux. At the same time, it could bring
to the Linux market the strengths of UnixWare–scalability and Clustering
capabilities.

If Caldera can successfully leverage the LKP in the
traditional Unix markets, then it has a potential killer on its hands. Picture
this–the ability to run Linux applications in a Linux environment, on your
existing "high end" Unix installation. It could also open up new
markets for Linux application developers, markets at the higher end of the
enterprise that were traditionally not open to them.

KRISHNA KUMAR

in Santa Cruz, California The author is executive editor

of PC Quest