advertising information
   personal technology

 custom news
 Headline News brief
 daily almanac
 CNN networks
 CNN programs
 on-air transcripts
 news quiz

CNN Websites
 video on demand
 video archive
 audio on demand
 news email services
 free email accounts
 desktop headlines

 message boards




Linux' success is remarkable, but nothing is guaranteed

January 4, 1999
Web posted at: 10:00 AM EST

by Nicholas Petreley, Editorial Director


(IDG) -- Many Linux fans were upset at a remark I made in a recent InfoWorld column. The quote that caused this concern goes as follows:

"Someone will undoubtedly label 1998 as the year Linux took the market by storm. This conclusion is premature. Linux is enjoying remarkable growth only because vendors are scrambling to make the most of the brief window of opportunity they have while Microsoft is paralyzed. As long as Microsoft is engaged in a battle with the Department of Justice, Microsoft cannot retaliate against anyone who dares defect from the "Windows NT is the only future" camp. Should Microsoft ultimately win the case, however, I guarantee vendors will abandon Linux faster than a rat out of an aqueduct."

While I made the original point in InfoWorld, I felt it is important to elaborate upon it here in LinuxWorld. I am issuing a warning to the Linux community: Do not become complacent about the way Linux is growing in market share. The success of Linux is not guaranteed, and whether or not it continues has little to do with its merits as an operating system.

  LinuxWorld's home page
  LinuxWorld free e-mail alerts
  LinuxWorld features & columns
  Computerworld "Emmerce"
  Get Media Grok and The Industry Standard Intelligencer delivered for free
 Reviews & in-depth info at's personal news page
  Questions about computers? Let's editors help you
  Search in 12 languages
  Subscribe to's free daily newsletter for IT leaders
 News Radio
  Computerworld Minute
  Fusion audio primers

Party now but beware

I celebrate along with Linux fans that companies like Intel, IBM, Oracle, Informix, Corel, and others are supporting Linux and releasing products for Linux. But make no mistake. These businesses may humor the Linux community. But they see Linux as an opportunity, not a community -- much less a charitable cause.

I know this statement will rankle among many Linux fans. You may want to believe that Linux has succeeded entirely on its own merits. You are right that those merits probably exist only because Linux is a unique community. That's why I run Linux. That may be why you're running it.

But that's not why Linux is enjoying its minutes of fame. Linux has received popular media attention only because it is the most recent challenge to Microsoft's dominance.

As far as the media is concerned, the main reason Linux is a credible challenge is because it is getting vocal and financial support from major corporations like Intel, IBM, Oracle, Informix, and Corel.

Here's where you need to wake up and smell the green, folks. There is only one reason why these businesses would dare challenge Microsoft's dominance. It's the same reason they did not challenge Microsoft's dominance in the past. It's all about the big, bad, dirty "M" word. Money.

These companies are looking for a way to make money on their products without first having to siphon off a portion their income to Microsoft. They saw the potential in Java, and now they see the potential in Linux.

But let me attempt to demonstrate why the future success of Linux hangs on a thread.

Put yourself in the shoes of one of these companies and then look at the strategic implications that led to their decision to support Linux in the first place. There are several platforms these vendors could have chosen to support.

The major strategic platforms are:

  • Commercial operating systems on non-Intel hardware (HP-UX, VMS, etc.)
  • Free operating systems on non-Intel hardware (Linux, xBSD, etc.)
  • Commercial operating systems on Intel hardware (Windows NT, OS/2, Solaris x86, SCO, BSDI, etc.)
  • Free operating systems on Intel hardware (Linux, FreeBSD, etc.)

Given all these platforms, the highest volume belongs to Intel-based operating systems. Remember that, regardless of the reality, Intel-based systems are perceived as the low-cost systems of the future. ISVs count on having a large customer base.

Until recently, that meant Intel-based computers, primarily running Windows. With that in mind, it is instructive to look at the commercial operating systems which vendors have decided not to support.

Most of them are non-Windows operating systems such as Unix that they support on other platforms. Even on those rare occasions when vendors support platforms such as OS/2, Netware, Solaris x86 and other commercial Intel-based operating systems, they do so quietly.

There is one very big reason why vendors have been so reluctant to voice support for any non-Windows operating system on Intel: Microsoft.

Microsoft is the quintessential two-ton gorilla. Its competitors and partners alike assume that, given free reign to behave the way it wants, Microsoft has the strategic and monetary leverage to make its server operating systems successful on Intel. And it can do so without concerning itself with the quality of its products.

With that kind of clout behind Windows, what is the point of supporting anything else?

The Microsoft trap

Some readers might want to point out that it is not always in the best interest of vendors to support Windows. If doing so leads to success, that success is almost always short-lived.

Microsoft has made it clear that it wants to control every lucrative or strategic software category on Windows. And Microsoft has demonstrated time and again that it is willing to (in order of preference) copy, buy, or license any competing technology it needs in order to gain that control. It would seem unwise, therefore, to support Microsoft or Windows, especially if you have a product that competes with a Microsoft product, or is likely to compete with a future Microsoft product.

But I remind these readers that when most companies form strategic alliances, they rarely look beyond the next few quarters. Microsoft offers success today, and most vendors are likely to respond by jumping at the opportunity. Some of them have even been foolish enough to share their trade secrets based on the unspoken promise of an alliance, only to watch Microsoft lose interest and launch a product that looks and works remarkably like their own.

If vendors can be so foolish as to walk into the above traps, surely you don't think their recent choice to support Linux is indicative of newfound wisdom? Even I, someone who believes in miracles, am not that naive.

When to annoy Microsoft

The problem for vendors has been that Microsoft has a history of stifling competition through exclusive agreements. When an ISV makes a strategic alliance with Microsoft, it usually involves the written, spoken, or unspoken agreement not to support competing Intel operating systems or other strategic platforms such as Netscape Navigator or a platform-neutral implementation of Java.

In plain language, every ISV who voiced support for Linux did so knowing that its announcement would royally piss-off Microsoft. Under normal circumstances, to do so would result in swift, ruthless retaliation from Redmond.

That's why, under normal circumstances, these vendors would never have given a second thought to Linux. Until recently, they did whatever Microsoft wants because Microsoft had every ISV by its family jewels.

Playing the game carefully

Right now, however, it doesn't matter much if something pisses-off Microsoft. Microsoft cannot afford to retaliate. The operative word here is much. Because even now it is not entirely safe to break ranks and support a non-Windows operating system. There is always the possibility that Microsoft will emerge victorious in its battle with the DOJ and seek revenge.

This is an important point to understand, because it means there is still a substantial long-term risk associated with voicing support for a non-Windows platform on Intel. For a company to rally behind another operating system, it must be worth the risk from a strategic perspective.

The question for these vendors is not, therefore, "Which is the best non-Windows operating system to support?" but "Which non-Windows operating system can we support with the least risk to our long-term bottom line?"

This reasoning explains why companies are not fussing about or investing in the success of Netware or OS/2. To do so is to invite the possibility of shifting monopoly control from Microsoft to Novell or IBM. As far as OS/2 and Netware are concerned, the long-term benefit does not justify the risk.

If you eliminate OS/2 and Netware, that pretty much leaves Unix on Intel. So why aren't these companies fussing about Solaris x86? Why isn't SCO Unixware making the cover of Forbes? Because it is strategically unwise for ISVs to get behind commercial versions of Unix.

It is unproductive to get behind more than one commercial version of Unix (Microsoft would surely point to it as evidence Unix is fragmented). And it doesn't make long-term sense for vendors get behind a single commercial Unix (once again, that introduces the risk of passing the scepter from one tyrant to another.)

That leaves the choice to commit to a single free (as in liberty, not cost) operating system. As long as an operating system is not under the exclusive control of any single company or team, there is no perceived risk that support will simply mean a shift from one form of tyranny to another.

Of the free operating systems available, the two most visible are Linux and FreeBSD. It is unlikely that vendors would choose to support more than one free Unix, because it is only through solidarity that ISVs could create enough perceived momentum for a product to compete with Windows NT in the long run. As I pointed out above, if they voiced support for multiple versions of free Unix, Microsoft would surely promote the idea that the Unix market was fragmented and unsafe, thus ruining the opportunity to build momentum for a non-Windows system.

It makes sense, therefore, that ISVs have chosen one free Unix to support vociferously. And Linux appears to be the one.

The motive is money

It is important to understand that, despite what anyone may think about the relative merits of various flavors of Unix, Linux is not enjoying its sudden swell of momentum on the grounds of technical superiority or price. I don't know of any version of Unix on Intel that is not technically superior to Windows NT. And even many of the commercial versions of Unix are less expensive than Windows NT -- some by a very wide margin.

One can only conclude Linux was chosen for other reasons. Linux has two advantages over all the other flavors of Unix on Intel. First, even before the media blitz began, it already enjoyed the greatest degree of mindshare among free versions of Unix. Second, it is perceived as a safe choice.

As Microsoft correctly noted in its Halloween documents, the GNU General Public License (GPL) makes Linux immune to long-term FUD. Linux cannot suddenly disappear as an alternative because one or more Linux distributor goes out of business. The GPL guarantees that the source code for Linux will always be freely available.

One could argue that FreeBSD is not subject to a single company going out of business, and that widespread adoption of FreeBSD could not lead to tyranny. But we're talking about a market where an operating system like Windows can gain and hold a monopoly on the desktop. For the most part, logic does not govern this business -- perception does.

Linux has emerged as the winner because it was most easily turned into the perceived winner. In an industry where perception is more important than reality, that edge made all the difference. (While fans of FreeBSD may resent the way things turned out, they can take some degree of comfort in the knowledge that a long-term victory for Linux is a whole lot better for the future of FreeBSD than would be a long-term victory for Windows NT.)

The road ahead

One of my three favorite pearls of self-evident wisdom is Yogi Berra's "It ain't over 'till it's over." (The other two are from an episode of the Nickelodeon cartoon show, the Angry Beavers: "Anything is possible if it happens," and "One moment more and it would have been later.")

Well, it really ain't over 'till it's over. And here's where we get to my controversial prediction.

There remains a single unknown. We still don't know how the Microsoft vs. DOJ saga is going to turn out. And the way it turns out could have a profound effect on the future of Linux. There are at least three possible futures:

  1. The DOJ investigation or trial leads to a ruling that restricts Microsoft from predatory and anti-competitive behavior. This ensures that Microsoft cannot retaliate in the future for Linux support today. This is the best of all futures for Linux.

  2. The DOJ vs. Microsoft saga lasts long enough to allow Linux and/or other versions of Unix to gain significant penetration into the server market. In this case, the Linux market could become entrenched and lucrative enough that ISVs do not believe they need to depend on Windows to succeed. This is a potentially good future for Linux and other forms of Unix.

  3. The DOJ vs. Microsoft saga ends quickly, and Microsoft is not sufficiently restrained by the outcome. In this case, Microsoft is free to retaliate with a vengeance. This is obviously the worst future for Linux and Unix in general.

Let's face it, folks. Win or lose, if Microsoft emerges from its trial quickly and without enough external restraints, Microsoft will surely force ISVs to choose between Linux and Windows NT much the same way it made many companies choose between Internet Explorer and Netscape Navigator.

In that case, Linux will maintain the support of ISVs if and only if they continue to believe they need not fear retaliation from Microsoft. And remember this is an issue of perception, not one of reality. As much as I would like to think these ISVs would be clever enough to understand the long-term benefits of remaining behind Linux, I predict that, given the risk, most of them would do whatever it takes to regain Microsoft's favor. All it would take is a single defection and the rest would follow.

Vote with your dollars

We are not entirely helpless. There is one way we can help direct the future rather than simply watching it unfold. As I said earlier, Linux's success all boils down to the dirty "M" word. We can vote for the future with our dollars.

This is perhaps one of the most sensitive issues regarding Linux, but it's one we must address if we want Linux to flourish. For example, we should vote with the dollars we are willing to pay for full-featured products that are also available for other platforms. I would expect Oracle, Informix and IBM to charge the same for a Linux version of their database server as for a Windows NT version of the same product. If you value having the choice, I recommend you respect this decision. The fact that you can get Linux for free does not necessarily mean you should be able to get commercial products for free just because they run on Linux.

That's not to say there aren't some wonderful GNU products for Linux that deserve your support -- GIMP and GNOME are two such free products (free as in beer, not just liberty) to name but two. But that doesn't mean every product will or should be free. Many who are religious about this issue are going to have to learn to live with that fact if they want Linux to enjoy widespread success.

Finally, as you know, many ISVs are giving away their products free for personal use. This is extremely generous, and it is a brilliant means of getting the kind of market penetration that almost always leads to follow-up sales (especially if your product is any good). But I urge you to resist the temptation to abuse this and cheat companies of their legitimate income.

If you use Star Office to do any of your work, for example, I urge you to pay for it. And in case you're wondering, I intend to do the same. I like Star Office and Applixware, and I could easily do all of my writing with just Emacs. But I'm using my personal favorite word processor, which happens to be WordPerfect. In fact, I'm using the free trial version 8 to write this very column. I'm not planning on reviewing the product, so I have no justification for asking for a review copy. So I ordered my legitimate copy of WordPerfect 8 for Linux through Linux Mall today.

The bottom line is that, while we should rejoice at the success Linux is enjoying, don't take it for granted. Linux is becoming more successful daily due to the support of major vendors. Their commitment to Linux will not be successful without our honesty and support, as well.

Nicholas Petreley is editorial director of LinuxWorld, columnist for InfoWorld, and associate producer for Agenda and Vortex for IDG Conference Management Company.

Related stories:
Latest Headlines

Today on CNN

Related stories:

Note: Pages will open in a new browser window Related sites:

External sites are not
endorsed by CNN Interactive.

Enter keyword(s)   go    help


Back to the top
© 2000 Cable News Network. All Rights Reserved.
Terms under which this service is provided to you.
Read our privacy guidelines.