While it will be some time before Linux is really much of a threat to Microsoft’s installed base or Microsoft to Linux’s, the battle for the consumer and corporate desktop was actually joined some time ago. Both sides are working furiously to derail the other.
With both the WinHEC and TechEd conferences this month covering Longhorn, the next version of Windows, it is time to revisit the emerging battle and address how each platform stacks up. There are at least two aspects of this fight: a server side, where Linux has some clear advantages; and a desktop side, where Microsoft is predominant.
When you have a market share the size of Microsoft’s, displacement is very difficult unless you can catch the other side napping. However, recall that Lotus owned the spreadsheet market and lost it to Microsoft; and IBM, which many believed could not be beat, was passed by Microsoft a decade ago in several markets. This showcases the idea it’s clearly possible to displace a dominant vendor, which means that the story isn’t over, by any stretch.
Desktop: Advantage Longhorn
However, to pass Microsoft on the desktop, you would need first to embrace what Microsoft offers in terms of interoperability, management, software application support and value. While you would think that Linux — which is viewed by many as “free software” — could at least meet the value requirement, Linux’s hardware, services and applications continue to cost more than the OS does. Even though Linux can match Microsoft platforms on hardware, it tends to fall behind in terms of application support and tends to cost more than Windows, at least on the desktop, with regard to services.
In addition, the desktop market is rabid about consistency. Desktop administrators will often put older nonsupported versions of Windows on desktop and laptop hardware just to get everyone on the same version of the operating system. The continued fragmentation surrounding the Linux distributions would seem unacceptable in the face of such a high — and often unreasonable — level of standardization. This fact of the desktop clearly works for Microsoft.
Microsoft can maintain a consistent marketing message because it is a single company, while it is difficult for the Linux vendors to be consistent on much of anything. In fact, many prefer fighting with each other. In addition, when Longhorn ships, the SCO litigation is scheduled for court and the related news should overwhelm any Linux-targeted marketing during that period.
Finally, Longhorn is increasingly appearing to be, by a significant margin, the most secure and reliable platform that Microsoft has every built. If it can live up to the hype, which is a significant challenge, it should provide an impressive barrier to entry for Linux.
Market Trends: Linux Opportunity
Microsoft’s image isn’t what it once was. Much like it was with client-server computing and IBM, the IT audience is actively exploring Linux as a desktop option. There are trials backed by vendors like IBM and HP springing up all over the globe. Governments seem particularly enamored with the offering, which is an important development because governments set standards. If they lock in on a particular version, say Novell’s, it could do a lot to overcome the nonstandard problem that Linux has.
Microsoft products are perceived to be overpriced. Because it is difficult to match value to price right now, many feel that Microsoft is overcharging. This perception, coupled with the belief that Linux is relatively inexpensive, is driving companies and governments to overlook services costs and consider Linux. Also, because Microsoft’s own image has made it difficult for Microsoft advocates to defend the company’s platform, Linux advocates are having a field day disparaging Microsoft.
Microsoft historically puts most of its marketing budget into a big launch party and does very little sustaining marketing. Adoption of Windows XP has been slower than it should have been, and the older a Windows desktop becomes the easier it is to move that desktop to a different platform. If Microsoft cannot get the market to move sharply to Longhorn, the opportunity for Linux to move aggressively against Microsoft will increase just as sharply.
More and more desktop services are moving to servers. With every desktop-to-server application migration, the incompatibilities associated with Linux become less important. Of course, if this goes too far, you could end up with embedded devices everywhere. But, assuming Linux can step up to the challenge, competing with Microsoft on embedded devices should remove many of Microsoft’s current advantages revolving around backwards compatibility.
Longhorn is a major release, and major releases not only typically have a slower adoption rate, they often have compatibility issues that make the related migration more painful. Windows XP was, technically, a minor release designed to address several shortcomings with Windows 2000, Windows NT and particularly Windows 9x. XP should have been a slam dunk. The fact that it wasn’t indicates a strong potential Microsoft vulnerability.
Server: Linux Advantage
The big opportunity right now is with Unix and Netware competitive migrations. Linux is, at least in practice, a version of Unix. It enjoys the use of similar skills, similar tools and a relatively easy migration path. While Linux is not similar, by any stretch of the imagination, to Netware, Novell is driving the move itself and migrating its tools and applications to the new platform at an impressive rate.
This gets Linux the support, out of the box, of Unix administrators who are already accustomed to differences between Unix versions and who, currently, don’t seem to be having much of a problem with Linux distributions.
Today, many use Linux on enterprise servers with great success. The belief structure that surrounds it favors Linux as the better product. Although recent reports indicate that this trend might be changing, it can take years for perceptions like this to shift.
In short, Linux is increasingly backed by the Unix vendors, and Novell as the replacement product and is so similar to Unix that the pain of migration isn’t much greater than, and possibly less painful than, a move from AIX to Solaris.
Server: Longhorn Server Opportunity
Microsoft clearly needs a more compatible product to Unix for that market, but with Netware there have been powerful migration tools in place for some time. Microsoft knows of this disadvantage and is working to mitigate it. If they can, it will at least remove one of the barriers, although it won’t, in and of itself, build demand.
Informal feedback I’m getting from Netware accounts indicates that many are not at all happy with this forced march to another platform. They report the migration pain is unacceptable, which suggests that Novell could lose significant market share during the migration process largely to Microsoft and soften somewhat the disadvantage Microsoft has with Unix.
While Unix practitioners are clearly wedded to Unix and Linux, IT executives often are more interested in corporate relationships. Microsoft plays this game very well. The typical open-ended IT sales budgets that put Microsoft at an historical sales disadvantage against Unix will probably evaporate around Linux because of the relatively low margins that should work to Microsoft’s advantage.
Wrapping It Up
Microsoft’s advantage on the desktop is one of position and technology, while the company’s exposures to competitors come from mostly image and perception issues. On servers, the opportunity is Unix and Netware. Linux has a similar positional advantage that will be hard to beat.
In the end, this fight will come down to image, marketing, sales execution, focus and which side is willing to make the most sacrifices to win. Microsoft is tested in all aspects; Linux has yet to be tested on everything. This kind of reminds me of Microsoft and IBM in the late 1980s where IBM was unwilling to go the extra yard.
Neither side has it easy. Linux will have to become more consistent, compatible and less divisive — which is contrary to its nature — and Linux will of course have to work out the intellectual property issues now being challenged in the IBM-SCO court case. Microsoft will have to address value, pricing, image, disclosure and marketing problems that the company often doesn’t even admit it has.
The next three years will definitely be interesting.
Rob Enderle, a TechNewsWorld columnist, is the Principal Analyst for the Enderle Group, a consultancy that focuses on personal technology products and trends.
This article is full of contradictions. "Longhorn is Microsoft’s most secure…" Where? What installation of Longhorn has been tested? Isn’t that what the marketing department said about XP? And 2000? And NT?
"Longhorn will run on", "Longhorn requires", Where? What installation of Longhorn is being reviewed?
"[A]nd address how each platform stacks up", Where? Show me a running Longhorn platform to compare to. Please.
Until you do, it’s pure marketing. I can go on all day long about how fabulous BobOS is going to be. I can compare it against anything and everything and NO ONE IS GOING TO CARE.
To paraphrase the article, "To beat Windows, one must do everything Windows does." The strongest enterprise Linux desktop installation is the Thin Client, for which Microsoft has NO ANALOG. So therefore Microsoft cannot beat Linux on the desktop?
"[Linux] tends to cost more than Windows, at least on the desktop, with regard to services." Even the marketing hype about Longhorn (this is about longhorn, right?) places the hardware requirements at many thousands of dollars for all new equipment for every single desktop. Compare that to using present hardware, and the much greater ability for a single support person to handle multiple Linux machines, and the POSSIBLE requirement that MAYBE some software MIGHT have to be PURCHASED is more than offset by the savings. I would say by orders of magnitude.
But then, I can say anything at all in comparison to Longhorn because it DOES NOT EXIST.
This article should have been comparing Longhorn to Duke Nukem Forever, at least that would have been comparing equal products.
Rob says, "However, to pass Microsoft on the desktop, you would need first to embrace what Microsoft offers in terms of interoperability, management, software application support and value."
The three big issues of Longhorn are 1) MS has said backwards compatibility with Windows applications isn’t a design criterion; 2) it’s supposed to be both a radical departure in software as well as in its hardware platform; 3) It’s delivery is slipping and at this point may very well slip again.
So, we’re going to have Longhorn, sometime in the distant future maybe. It’s going to cost an arm and a leg to buy it, its hardware platform and all new applications to run under it. It’s likely to take considerable expense to retrain workers to take advantage of the radical design innovations.
Meanwhile, linux is improving in ease of use and power at a faster rate than any other OS I’ve used in the 40 years I’ve been using computers. It’s not standing still but it’s evolving. It’s capable of jumping to hot new hardware without radical rewrites of the software. Applications have also arrived and evolved faster than any other platform I’m familiar with.
I’d say that the world that Longhorn arrives in will be a different world than today and the decisions made about "advantages" are not the ones Rob’s fantasized.
It seems companies have a problem parting with older versions, as the cost of their new desktop outweights the cost of the older desktop. I’m glad you brought that point up. The fact that MS is wishy-washy on when they want to do away with their support alone is a big issue facing several people’s decisions these days. Is MS’ cost of services any better than those offered by serveral people offering Linux and Linux support? That depends on your needs. With many companies the support for Windows doesn’t just come from MS, it comes from 3rd-party software developers or VARS. How is this different than buying Linux services from Red Hat, Novell, HP, IBM, SUN or a combination? It’s not different, except in the eyes of those that want to believe it’s different. Longhorn will at least put MS on level ground with Linux. It will do this because it is a new platform and people will have to weigh migration, support services, and what applications will run or be orphaned. If MS doesn’t provide the security that it is promising will that hurt MS? Time will tell.
Robertson has it exactly right. He only fails to explicitly make the connection about Longhorn marketing: Rob Enderle is part of that marketing effort it would seem. He doesn’t need to see it or test it, he just needs to pass along what MS promises it will be. Microsoft has made and broken promises with every OS release. Longhorn won’t be an exception: it will slip in delivery date and it will slip in feature set. They all do both.
Last I heard, Longhorn was delayed until 2006 because Microsoft had to take engineers off the project to work on Windows XP bugs.
Meanwhile, Redhat, Novell, and IBM are releasing products in 2004 that are aimed at the desktop users.
IBM’s Lotus Workstation looks like it will allow current Windows desktop users to install the Lotus client on their Windows systems, and start using Windows and Linux applications on IBM’s servers.
All the maintenance costs will be on IBM’s side, and the users can continue to run their currently installed Windows applications locally, and to run new Windows and Linux applications remotely.
Since the remote applications will have the same client interface whether called from a Windows operating system, or a Linux operating system, the users will be able to switch painlessly from one to the other.
I think that many will switch to Linux before Longhorn is released, and, afterward, many will consider carefully what benefits over the Lotus Workstation might make it worth while to tolerate the pain of migrating to Longhorn, with the prospect of still another forced migration a few years later.
While it will be some time before Rob Enderle really gains any credibility with any thinking person
(Hey rob, tell us again how companies can hide evidence during discovery and spring it on an unsuspecting foe during the trial)
(Hey Rob, I bow to your knowledge of the legal system – You should know, considering your immense experience in the legal system).
This is still not a good start for Rob.
Every server that runs Linux is a lost sale for Windows. How much revenue is that that’s been stolen from MS?
"Not much of a threat". Sure, Rob, Sure. Stick to the legal system, Rob you know so much more about that than about Linux.
Clueless as ever.
PS if any newbie comes across this Rob Enderle article & gives any credence to Rob Enderle, please do a Google search for past Enderle pronouncements, for example, on Macintosh.