Endgame For SCO

Gil Bates writes: "So, it looks like old SCO is finally on sale, and that provides an interesting chance to speculate again about the future of a company that had some good technologies, but couldn't respond successfully to challenges from Linux and Windows... Whoever acquires the rights to OpenServer and UnixWare needs to appreciate this fact and admit publicly that there will be no further work, besides bug fixes and sales of the existing versions, on these platforms. Instead, the focus will have to be on an immediate migration path to the next platform, almost definitely Linux. With this admission, the new vendor will certainly cannibalize short-term licensing revenue, but the longer-term benefits of acquiring a chunk of the SCO customer/reseller base will outweigh that (pretty damn small) opportunity loss. osOpinion has an extremely thought provoking editorial piece which expands upon this issue in detail."

Re:The future

You do realise that several old-school database companies use proprietary data and even partition formats to store their data more efficiently? That's part of what ReiserFS is about -- bring these efficiencies to a general purpose filesystem.

That aside, if you want an example, go visit Pick Systems [picksys.com]. Incidentally, they support Linux -- but require(1) [slashdot.org] the use of their own partitions for data storage.

--- (1) You can also use dd /datafile type files as data storage partitions if you wish.

The Novell connection

The author has pretty good ideas for Novell participation. Novell is going to die unless they jump on Linux soon. There is no such thing as a new Novell installation, only big businesses are buying the upgrades. Netware really doesn't offer much, it is far too overpriced. Yes, it probably does file and print better than anyone else, but that hasn't been enough for at least three years now. For as much pissing and moaning I see on slashdot about MS pricing, Novell aint exactly a bargain.

Now, some Novell evangelist is going to reply and tell me about all the wonderful apps that Novell can run. "Can run" is the operative phrase. Try finding anyone to support them. Big businesses do, but small businesses don't. One of the reasons small businesses cant run the apps that Novell is capable of is that they haven't traveled down the road of the upgrade path. One of Novell's problems is that Netware 3.xx was such a great f&p server that these small businesses never had a reason to upgrade to 4.xx or 5.xx. Netware is fundamentally graying, and I just don't see a lot of people who can support it past file and print, while on the other hand, there are tons of people learning Linux and NT.

Novell's best bet is an cluster capable enterprise ready Linux distro that does all the brilliant NDS stuff. They ought to position it as bringing application serving functions for the low price of Linux while remaining integratable with Netware. I have seen a lot of small Novell shops bring in Linux for email or web, bnut in a standard distro (read: Red Hat), there is a lot more functionality for making the network MS friendly (samba!) than there is to make it Netware friendly. Invariable email and web becomes mission critical (supplanting file and print, thus devaluing Netware's importance), and people question why they continue to deal with Novell client software (keeping up wtih it is a full time job), when they could use the MS stuff, and talk to either samba, or NT.

matt

Re:Was Linux the competitor?

I hate to be the one to point it out, because I like having karma, but you neglected to mention that some folks may have moved from SCO to say.... Windows NT?

I know no one likes to hear that, and neither do I, but some companies are going for Windows-Only networks. It hurts me to see my company moving to an Outlook/Exchange server, running our Intranet on IIS and running our file servers with file and printer sharing.

I don't want to be the devil's advocate here, but I think it's important to realize some companys are dumping SCO and *nix completely. Their loss I guess.

IS the traditional OS-sales business model dead?

A quote:Total revenue dropped by nearly 50% from Q3 1999 to Q3 2000, while sales of Tarantella and professional services have supposedly increased.

Linux really has introduced a new business model to computing. Almost overnight, without any warning, several different technology companies have chosen Linux or some other publicly available OS (think Mach kernel on MacOSX) to be their foundation for a variety of products. This begs the question: Can the concept of an Operating System now be considered to be that of a general-purpose, standardized factory part and not a specialized, proprietary addition to be sold separately for a premium price?

In the earliest days of personal computing the Altair (or S-100) bus was the rage because it allowed for a common set of peripherals to be constructed for PC's. That paved the way for the AT bus and the concept of a whole common PC architecture, which pretty much exists today. Because of this, the profit margins on PC's are small.

Shouldn't the next logical step be a common OS architecture? Traditionally, an OS was something on which you could make a huge profit. When you buy a copy of MS-Windows at the local computer store, what's the cost of the media and packaging compared to the price? Pretty small. Sure you can say that MS is trying to recover R&D, but seriously, most of that was probably recovered a LONG time ago. Besides, you're LUCKY if you get the media with a new computer these days. It's mostly profit. Now in walks Linux, the product of a Grad School student gone berserk. It's free, stable and encourages others to transport it to other platforms. Suddenly, the corporate powers-that-be realize that they no longer have to pay MS, SUN and SCO to get a good, general purpose OS. Furthermore, they themselves can add to the development of Linux, ensuring for themselves a common development platform thereby lowering their own proprietary development costs.

So these people stop buying OS's. SCO, which had called themselves "an undifferentiated solution with price as the selling point" dies first. Price IS a way of differentiating your product; Linux, on that front, apparently does it better than SCO (it's free). Linux has taken over the low to medium end in server computing. How long before it encroaches on the high end? The surviving Linux companies today differentiate themselves by offering different services to different markets. The Linux companies don't sell Linux the OS; they sell Linux solutions to people. John Zedlewski's article talks about SCO selling of Tarantella, even though it's making money for the company. I think SCO should realize that there's far less money to be made in the general purpose OS market than there used to be. The money right now is to be made in special-purpose server/middleware and other application software.

But what of Microsoft? Aren't they suffering? Not in the desktop market. Linux doesn't have a simple, relatively fool-proof desktop anyone can use - yet. Additionally, MS's monopoly power will generate cash for the forseeable future. The server market is another story. Most low-end boxes that run Linux are boxes that don't run NT/2000. Sales of NT/2000 may be good, but not as good as they can be. Furthermore, for the sophisticated user, Linux flexibility far outstrips NT/2000's. Shell scripting, background apps and stablility are 3 big advantages that immediately come to mind. With the coming breakup of Microsoft, I wonder if, in ten years, it will be viewed more of as a spin-off than a break-up. Windows currently contributes 1/2 of MS's revenues, but would this still be the case 10 years from now if no breakup were to occur? The break-up will hurt MS now, but certainly won't kill it. And without Windows to worry about, what's to stop it from pushing into the UNIX/Linux markets?

I realize that there are niche products/markets where a proprietary OS is the best solution. But for the general purpose market, is this the future?

Stop Speculating: Here's An Article On Their Plans

Here's a Forbes article [yahoo.com] on SCO's [yahoo.com] Linux plans.

Quote: So, the company is hitching its wagon to--what else?--Linux. SCO has been selling support services for Caldera (Nasdaq: CALD - news) and TurboLinux for about 6 months, and it has already given some intellectual property to the Linux open source community. But SCO hasn't yet taken the big step--distributing Linux.

But that's coming. Sources say it's working out an arrangement with France's MandrakeSoft to distribute its Linux-Mandrake operating system. SCO will use Linux-Mandrake as the base OS and add some features like clustering, which is a complex way to improve the performance and expansion of servers. MandrakeSoft also has offices in Altadena, Calif.
--

Re:Yes, Linux toasted them

There are currently more Linux seats than there have ever been "UNIX" seats. While companies like SCO, who made their money selling a commercial UNIX for i386, are certainly getting hit hard, there is no underestimating the damage that Linux is doing to Windows NT.

The only reason that Windows NT is even considered, in most cases, is price, and Linux gives you UNIX features at an even lower price. Besides, Linux doesn't have to make a profit to survive. In other words, while the Linux firestorm may be especially hard on Microsoft's competitors this does not necessarily mean that this benefits Windows. After all, Linux is replacing Windows too, and Linux is going to be very resistant to Microsoft's usual tactics. Microsoft isn't going to be able to put economic pressure on Linux, GPL software is very resistant to embracing and extending, and they certainly aren't going to be able to undersell Linux. Nor is Microsoft going to be able to count on Intel to close the performance gap between Windows and Linux. After all, Linux runs on the same platform as Windows (and a whole lot more).

Microsoft was competing quite successfully against commercial Unix, but they have yet to come up with a tactic that is even somewhat useful against the Free Unixes. I would bet that Microsoft would give anything to be able to compete with the likes of SCO instead of Linux.

What will become of the Monterey Alliance now?

Monterey [ibm.com], as you may or may not know, is the recently formed alliance between several Unix distributors, who having faced the prospect of rewriting their operating systems for the 64-bit Itanium processor decided to pool their resources. Monterey was to have been a combination of IBM's AIX, Sequent's Dynix and SCO's Unixware with technical support from other companies such as Compaq and would run on the IA-64 and PowerPC chips (support for AMD's 64-bit chips have yet to be announced).

Monterey was to have a signified a shot in the arm for Unix vendors who are being beset by not only the growing market share of Linux and Windows NT. Industry momentum is building and commitment is growing for Monterey on IA-64, which will be "a leading, high volume, channel-ready, shrink-wrapped, UNIX operating system" as trumpeted by its developers.

Now that SCO is up for sale, will the momentum on the Monterey project be severely hampered? In addition, will the purchasers of the Santa Cruz Operation continue to work on the Monterey project or will they break off from the alliance and take a solo path such as Hewitt Packard and Sun in developing their next generation Unix distributions? Are the prospects of profit arising from the Monterey Project be enough incentive for another company to buy SCO?

Nevertheless, it is sad to see SCO leaving the BSD/*nix scene after their pioneering work in the earlier days and their philanthropy in providing free/lost-cost licenses to students long before Linux became prominent.

Re:Yes, Linux toasted them

I will grant you that Windows NT still has a distinct advantage in application availability, and I will even grant you that there are a certain number of shops that are going to side with Microsoft no matter what they do.

This does not take away from the fact that Linux has become a force to be reckoned with in the server realm, and that it is even gaining acceptance in Microsoft's own backyard, the desktop operating system.

The reason for this is quite simple. Linux is less expensive, and it comes complete with an impressive array of programs (especially programming tools).

For years the Macintosh was a much more sophisticated machine than the IBM PC, and yet DOS reigned supreme because of one simple fact. Macintoshes were more expensive than commodity PCs running DOS, and DOS was essentially "good enough." In the end DOS + Windows not only surpassed the Macintosh in revenues, but they surpassed it in technology as well.

The same thing is happening again. Only this time the commodity platform is not Windows, it is Linux. For an ever increasing number of people Linux is "good enough," and the price is certainly right. Windows advocates have been explaining to me how Linux was doomed since 1995, and yet it continues to grow at an exponential pace. Even worse, for Windows anyhow, Linux is getting to the point where it is much harder to find Windows applications that don't have a functional equivalent Linux application. This is especially true when you are considering using Linux as a server.

Not to mention the fact that Microsoft's biggest customers (the hardware OEMs) would absolutely love to see Linux become the standard for the new Millenium. After all, why should Dell or Compaq have to pay Microsoft a tax to sell their own hardware. Especially when Linux is free.

Was Linux the competitor?

While Linux-based systems are infiltrating the server world, I question whether they are replacing current commercial unix systems or creating a new niche. While SCO's apparent demise (and possible upcoming resurrection?) may be due to a surge of Linux use, I would doubt it.

More than likely, their business was merely not as profitable anymore in a developing industry. I'd like to see statistics, to know whether people were replacing SCO systems with other commercial unix systems or Linux/*BSD.

It may be a bit pretensious to assume it was Linux pushing SCO out of the market.

SCO's next venture

SCO has hired Chase H&Q to explore future directions, including "strategic combinations". Those strategic combinations will almost certainly be with a Linux firm, and almost certainly TurboLinux at that. Here's what I expect SCO to do:

SCO has seemed to have finally discovered the Linux kernel. At the same time, SCO owns a wealth of proprietary technologies and licenses a bunch more. What to expect from this combination is something along the lines of "TurboLinux Enterprise Edition", with SCO's proprietary technologies and licensed technologies (including CDE) making their way into a Linux distro.

At the same time, the number of non-Linux members of The Open Group is shrinking. Expect to see The Open Group forced to open CDE in the same way that it opened Motif soon, and the UNIX trademark being "donated" to certain Linux distrobutions.

Note: This is wholly speculation and is for entertainment purposes only.

SCO is still around?

Although I will admit not having much recent experience with SCO. My nightmares still harken back to the very late 80's and early 90's with that frankenstein of a *NIX, Xenix.

I find it amusing that they have their hands on the child of Sys V, and wish luck to whomever ends up with it.

All in all, SCO has been missing the ball the entire time. First with their clunky and inept implementations of *NIX, and then with their inability to see the trends (right before their eyes) of the ever-encroaching open source community.

Score 1 for our team. And so long, Santa Cruz. Thanks for all the fish.

gitm

How many chickens was it?

Is Gil Bates a new Slashdot columnist? He's got the length AND the leaping to conclusions down...

We go from "SCO is on sale" to "SCO didn't/doesn't have a viable business model" to "new owners have to kill the product" (but not the business model?) to "migrate to a different product". From there we conclude, somehow, that there will be "longer-term benefits of acquiring a chunk of the SCO customer/reseller base", whatever that means.

How about we just have the summaries spell out the NEWS and leave the theorizing to the comment area?
--

reports of SCO's demise...

So, it looks like old SCO is finally on sale,

Funny, but I must have missed that press statement, and you don't link to one here, nor does the opinion column that you linked to.

and that provides an interesting chance to speculate again about the future of a company that had some good technologies, but couldn't respond successfully to challenges from Linux and Windows...
(Cough) FUD (Cough)

Whoever acquires the rights to OpenServer and UnixWare needs to appreciate this fact and admit publicly that there will be no further work, besides bug fixes and sales of the existing versions, on these platforms. Instead, the focus will have to be on an immediate migration path to the next platform, almost definitely Linux. With this admission, the new vendor will certainly cannibalize short-term licensing revenue, but the longer-term benefits of acquiring a chunk of the SCO customer/reseller base will outweigh that (pretty damn small) opportunity loss.

That has got to be one of the dumbest things I have ever read. Everybody here is a little dumber for having read it.

There is one reason, and one reason only, that anybody would want to buy SCO, and that is for their products... specifically, SVR4, UNIXWare, and Tarantella.

Nobody who wants to be YALD (Yet Another Linux Distributer) would have any reason to buy a non-Linux company. A friendly take-over of one of the smaller players (read: not Red Hat) would make a hell of a lot more sense.

As for the loopy idea of buying SCO's customer base... If, as you imply, the Linux and NT vendors are stealing these customers away so easilly that SCO is on its death-bed, how much could those relationships possibly be worth?

osOpinion has an extremely thought provoking editorial piece which expands upon this issue in detail.

No they don't. They have an extremely speculative rant that reminds me of the "Apple is dead" stories that ZDNet used to run every other week a couple years ago.

Linux is very cool, but we gotta tone this kind of zealotry down a little. One bad quarter != a dead company.