Windoz Nightmares for Microsoft

No one can defeat Microsoft. No one will. No one is big enough or strong enough. An incessant swarm of killer bees, on the other hand, will bring down the greatest giant.





Four years ago - nay, even two years ago, few would think that Microsoft could ever be defeated. Today, at least severe diminishment seems inevitable. This is not a story of battle with a mighty rival - but a story of stumbling in unfamiliar terrain, of unrelenting attack by a swarm of adversaries new and old, larger and smaller, all wielding unfamiliar weapons.

End of the PC Era

The PC Era is over. It's so over Ziff Davis' industry leading PCWeek magazine has renamed itself eWEEK. That doesn't mean PCs are going away, it just means they aren't where the action is any more. The growth markets are in thin client, mobile devices, network and Internet appliances, and embedded computing power.

All these are places Windows doesn't fit, even with a shoehorn. Microsoft is making a valiant attempt to shoehorn Windows into small devices, but they aren't having much success. Several cable companies that Microsoft invested a lot of money in to encourage the "right" decisions, have abandoned Microsoft software for their set-top boxes. Microsoft can't deliver and others can. Some have turned to Linux based software.

Even AT&T, in which Microsoft invested $5 billion (with a "B") to influence their software decisions, is getting pretty impatient and will probably go with a non-Microsoft solution (those clever bastards at AT&T took all that money without making a firm commitment to use Windows, so they don't even have to break a contract or anything).

Besides, if your device doesn't need all the capabilities of Windows, why pay endless royalties to Microsoft when Linux is free? Microsoft has been synonymous with the PC forever, and that's a link difficult to break, but the new devices play by different rules.

Linux / Unix

Unix is what Windows NT always wanted to be when it grew up. Now, as Windows 2000, Windows NT is grown up enough, and stable enough, to be taken seriously - if your application isn't too big or too critical.

Linux, while much more stable than Windows NT, shares at least one of NT's weaknesses. It does not scale to enterprise class either. It lacks high end features that would enable it to achieve sufficient performance and reliability. NT, in fact, is stronger than Linux at the high end of their range. Both are capable departmental servers, but not ready to take on the data needs of a large organization.

Recognizing these limits, even as they claimed NT was "enterprise class", Microsoft launched the largest software development project in history to overcome these limitations. Windows 2000 incorporates over 30 million lines of code, more than 80% of which are new, to make it capable of moving into the corporate datacenter.

Linux has no billions to pour into development - it relies on volunteers. The advanced features needed in the datacenter take a tremendous amount of effort to create, test and document.

So things are all roses for Windows 2000 and pretty bleak for Linux? Well, not exactly. A solution has been found to catapult Linux into the rarified atmosphere of the datacenter - before the end of the year. So obvious is this solution that several companies came up with it and are already hard at work on it. IBM, Hewlett Packard, and Caldera are.

Linux, you see, is patterned tightly after Unix. They are almost the same thing. All that is needed is to do a little adjusting of Unix to enable it to run Linux binaries. Presto! Datacenter Linux! Journaling filesystems, availability clustering, multi-terrabyte disk system, screaming 32-CPU SMP, gigabit RAM, 64-bit CPUs, gonzo transactions per second, the works!

Caldera, a leading Linux publisher, has purchased the Unix assets of the largest publisher of Unix, SCO. Caldera will be able to offer business a complete range of operating system products from single desktop to datacenter clusters - all able to run the same software.

IBM, a major promoter of Linux, is adapting its AIX Unix to run Linux binaries. Linux/AIX will run on both Intel platforms and high performance RISC machines, specifically IBM's RS6000 series. Hewlett Packard is well on the way to "Linuxing" their Unix, HP UX, and Even Sun Microsystems is (reluctantly) making it possible to run Linux software on Solaris.

"But, but, but, this isn't really Linux", you say? Well, it runs Linux programs doesn't it? It brings the enthusiasm of Linux developers to the datacenter level, doesn't it? By time this thing has run its course you won't be able to tell where Linux stops and Unix starts anyway.

This is, at long last, the "grand unification" of Unix.

Even worse, Linux is eating away Windows' low end server market. Microsoft expected to hold this territory completely once Novell was driven out, and use it as a staging area from which to invade a weakened datacenter. Now they stand between the fast moving rock of Linux and the surprisingly resistant hard place of Unix and mainframes - and the space between them is closing.

Consultants, VARs & Integrators

Microsoft tells us software is transitioning from a product (something you buy in a box) to a service (something you rent by the month) - but what if software is really transitioning to a CD-ROM you get for free along with equipment and services? That was precisely the dominant model before Microsoft came on the scene.

Microsoft has invested millions of dollars in several large integrators to make sure they recommend the "right" products to their clients. Of course, one must ask how smart a client would be to hire an integrator who's sold his freedom to Microsoft, but many will.

That leaves a few very big integrators, like IBM Global Services, a bunch of medium size firms, and literally thousands of smaller firms (Automation Access, for example) that value their freedom to provide the most appropriate solutions to their clients. Their biggest concern is that the software they deliver works, because if it doesn't, they won't be paid.

Many independent consultants and integrators look very favorably on Open Source software. While Linux is familiar to most people, there is a lot more (Apache, Star Office, Mozilla, The Gimp, etc.) and a lot more coming. Successful Open Source software tends to be foundation software - precisely the kind of software Microsoft makes its money from.

But why would anyone in the technology business push software they have to give away for free? Well, there are some very, very good reasons.

  • Competitiveness: when you go in with an Open Source based solution, you can offer better equipment and a lot more service and support at a lower price than the guy going in with commercial software.
  • Profitability: you wouldn't make a dime on a Microsoft product anyway (or those of most other software vendors), so you're not losing profit - you're increasing your profit by increasing the service and support component.
  • Control: when you deploy Open Source products, you have control, not some uncaring software giant in another state. You can even modify the product at the source code level to fit a particular client's needs. More commonly, you can review the source code to see exactly why it behaves the way it does.
  • Quality. Many Open Source products are of exceptional quality, because there are no secrets. There's no place to hide and a lot of eyes are looking at the source code. If those eyes find problems, they send solutions back to the originator of the software for incorporation.
  • Stability. With Open Source, Integrators and Consultants are insulated from having their product and marketing plans screwed by radical changes in products, policies and programs every time a key vendor plays musical chairs on executive row.
Basically, why transition your clients to paying for software by the month when they don't have to pay for it at all? You're not going to get any of that monthly revenue anyway.

But it gets worse! Many consultants and integrators have developed a marked preference for doing their programming in Java, Perl and Python rather than Microsoft Visual Basic or Visual C. These languages are inexpensive, cross platform, fast to develop in and make easily reusable code. Their cross platform nature makes it easy to deploy software on non-Windows computers.


Java came forth from Sun Microsystems and promised programmers could write a program once in Java and it would run on any platform. Obviously this is a tall order, and it hasn't worked out completely, but it's getting better.

Java is also designed to work efficiently on networks, particularly the Internet. In particular it works for server level applications supporting thin clients over the network.

Microsoft, distressed by the cross platform goals of Java (Windows not required), licensed it from Sun with the intent of "polluting" it (their words - as revealed in executive memos introduced as evidence). The plan was to produce a Java that produced "Windows only" code, and get their huge network of software developers to use that version until it became the "standard". All other versions then became "incompatible". This is a technique Microsoft has used successfully many times in the past.

All the pundits and columnists declared Microsoft would win the Java war (as they win everything) and that Java would be destroyed in short order.

Sun sued for violation of their license. The case is still in court, but Sun got injunctions and is almost certain to win. Microsoft did not get its way, and is dropping Java from its development tools. This will hurt only the few developers who were stupid enough to adopt Microsoft's Java in the first place (few did, few Windows developers even).

The pundits are still shouting that Java is dead, killed by Microsoft, in the very same magazines that report ever larger and more numerous corporate development projects based on Java.

The real proof of the pudding? Help wanted advertising for Java continues to increase, while advertising for Microsoft's Visual Basic continues to decline.

Microsoft's .Net project is, in many ways, an attempt to duplicate the potential of Java within an all Windows environment (including a new programming language, C# (pronounced "C Sharp") which borrows heavily from Java. Of course Windows can't do this, so they're developing a .Net client that just looks like Windows.

Obsolescence of Windows

Windows was begun in 1981 specifically to produce a monolithic all inclusive environment to compete with Visi Corp's Visi On. This one program all inclusive approach has guided the development of Windows right up to today, with yet more features being rolled into Windows ME.

Windows was designed to be used by a single user on a single PC. Networks weren't even in Microsoft's picture until a few years ago, never mind the Internet, so running in these environments wasn't a consideration in the design of Windows. Charles Fitzgerald, director of business development in the platforms division had described Windows as "rickety".

Today, networks and the Internet are of critical importance, and Windows looks like a beached whale. The massive failure of Windows Client Server computing demonstrated conclusively that Windows and networks didn't belong together.

Client Server applications are now being successfully deployed as Web based "thin client" applications. So why do you need Windows for them? You don't - all you need is something that supports Java and a Web browser.

Current design calls for applications to be made up of a set of small, interoperating modules, and to run most tasks on servers to reduce network traffic. There is simply no way Microsoft can fit Windows into this environment.

What Microsoft intends to do is to produce a client environment that looks and works like Windows. This is a key part of their .Net initiative. Steve Balmer, Microsoft's president, says this will take two to three years. To be realistic, lets call it four to six years. Will anyone still care? Others have been deploying this technology for a couple of years and are miles ahead.

The Internet

Bill Gates despises both academics and government. It was beyond his comprehension that something coming from a combination of these two could be a threat to his plans. As a result, he was blindsided by the most important development in the history of computing and communications.

Once he recognized the threat, he determined to take control of the Internet by leveraging the Windows desktop monopoly. The method is to incorporate corrupted versions of public Internet standards into Windows and leverage the dominance of Windows to make all standards compliant products look "incompatible".

Microsoft bought Web server software and incorporated it into Windows NT Server as IIS (Internet Information Server). They licensed Web browser software from Spyglass, promising to pay Spyglass a percentage of sales, then gave Internet Explorer away so they would have to pay Spyglass nothing. They found a company with quality Web development software, Front Page, and gave them a choice - "sell out or die".

Microsoft gained ground very rapidly, taking market share from Netscape, the dominant Web browser by "integrating" Internet Explorer into Windows, and by a number of business practices that have been declared illegal by judge Jackson's court, and unethical by just about everyone who doesn't make his living off Microsoft's coat tails.

Microsoft's browser share now stands at around 62%, but seems to be making little further headway. This share will be greatly diminished if on-line giant America On Line drops Internet Explorer in favor of Netscape, which they probably will do, since they now own Netscape.

Many ISPs started out on Windows NT with IIS because Windows was what they knew. Most have been abandoning Windows for Unix as they become larger and more experienced - for reasons of performance and security. Many still maintain Windows Web hosting servers for people who insist on Microsoft tools, but the heavy lifting is done by Apache running on BSD Unix, Linux and Solaris.

A new high end Web server essentially crushes Microsoft's hope of controlling the Web server space, and without control of the servers, there will be no way to make proprietary Microsoft "enhancements" the norm. A quad Pentium running IIS is just not in the same world with an IBM S/390 G6 running 3,000 instances of Linux and WebSphere (a variety of Apache) with 20 terrabytes of storage.

The U.S. Army dumped Windows and went to Apple Macintosh Web servers for reasons of security (they were tired of getting their site hacked). Most revenue producing Web operations run by Microsoft itself run on Unix and Apache, not Windows and IIS.

The Internet industry is maturing, and as it matures, it realizes the strength of the Internet is in strict adherence to standards. Microsoft's attempts to corrupt those standards are increasingly being resisted.


Microsoft designed their products as a closely integrated solution for the single user environment, and designed their business strategy to leverage that close integration to lock out all competitors. Then they were forced to tack on networking, and after networking, the Internet.

A single user environment needs almost no attention to security. A networked environment has significant security needs. Connect that network to the Internet and security becomes critical.

The close integration of the Windows / MS Office / Internet Explorer environment, plus its concentration on visual flash and ease of use over all else, has created a situation that is a playground for even semi-skilled hackers, crackers, script kiddies and virus writers. No other common environment allows programs to be downloaded without user knowledge and then have complete access to all system resources.

This security shortfall is not only very difficult to fix, because security wasn't planned in from the start, but fixing it runs directly counter to Microsoft's marketing plans and product strategy.

Some companies, especially those involved with the Internet are looking closely at this problem and selecting other operating system environments because of it. The U.S. Army, for instance, ripped out all it's Windows NT based Web servers and put in Apple Macintosh Web servers because they were tired of being hacked all the time.

IBM, Sun and Oracle - The Threat From Above

Everyone who read computer magazines in the late '80s and early '90s knew the last mainframe would be unplugged before the year 2000. All would be replaced by cheap PCs. This didn't happen. The only reason anyone thought it would was because they knew nothing at all about datacenter computing.

It took quite a while for Microsoft and the PC forces to adjust to the concepts of large scale, no downtime computing. In this time, the "big iron" companies recognized the threat. Rather than being sitting ducks, they fortified their positions and became more flexible and innovative.

Now these companies, lead by IBM, Sun and Oracle, see an opportunity to "cut off Microsoft's air supply" (to use a Microsoft phrase), and that opportunity is called Open Source. They are all strongly promoting Open Source software, and Sun has released Star Office as open source.

IBM sells hardware, enterprise software and service. Sun sells hardware and service. Oracle sells a high end database, enterprise software and service. None of them would be hurt one bit if operating systems and office suites were free.

Microsoft derives nearly all their revenue from selling operating systems (Windows) and office suites (Microsoft Office). Microsoft would be hurt really, really hard if those were free. Without all that money, their invasion of the datacenter would slow to a crawl.

IBM, Oracle and the others are building major Linux support organizations to insure their own revenue streams continue. Linux needs plenty of support too - and, of course, they also support Microsoft products. It's win / win for them.

Be afraid, Bill. Be very afraid.

Sony - The Threat From Below

The Sony Play Station 2 (not yet selling in the U.S.) is a very powerful machine. Every one Sony can build is sold - instantly - often at over list price.

Besides playing games, Play Station 2 is perfectly capable of providing Internet access and other functions now done by Windows PCs. It does not run Windows.

The development environment for the Play Station 2 is Linux. It is expected future versions will be capable of running Linux as a general purpose operating system.

Microsoft is well aware of the seriousness of this threat, and is attempting to counter it with X Box, their own game machine. Can Microsoft, with no hardware experience, successfully launch a major game machine in the face of experienced, battle hardened competitors like Sony? Can they convince game programmers to program for X Box and only X Box?

X Box will run a special, very cut down version of Windows. Will Microsoft be able to get it done in time to be a contender? Or will Sony be taking the world by storm with a Linux enabled Play Station 3 by then?

Foreign Markets

Even well-to-do countries are looking at Microsoft and wondering if they can really afford to export all that money to Redmond, from where it will never return. Their governments are wondering what back doors Microsoft may have built into Windows for American spy agencies. And many countries simply can't afford Microsoft products at all.

Then there is patriotism and an important industry dominated by a foreign power.

Naturally, people in all countries outside the United States are looking seriously at Linux and Open Source software. The government of China has just decreed that Linux is to be China's official operating system. The French legislature has been on the verge of declaring that all government systems must run on Open Source software. The school systems in Mexico have adopted Linux because they can't afford Microsoft.

Not only does this exodus to Linux threaten substantial Microsoft income, but Linux as an international standard would further influence the U.S. Government, parts of which are already leaning to Linux. An example is the President's Information Technology Advisory Committee.

Some other countries also provide a safe haven for reverse engineering software products so Open Source equivalents can be created. Reverse Engineering is legal in the U.S., though threatened by UCITA, but "big software" can simply bankrupt anyone who reverse engineers a product, legally or not, by launching a swarm of lawyers.

Further, the U.S. will probably continue to import technical expertise from other countries in a play to keep U.S. technical labor costs down. If these foreign workers bring a preference for Linux with them, that will influence decisions American companies make.


The "Dot Coms" may have met their Waterloo, but e-commerce itself is gaining momentum like a runaway train, powered by mainstream retailers, and "business to business" consortiums.

Major e-commerce sites require stunning system performance, unquestioned reliability, mind boggling storage capacity and highly skilled support 24 hours a day, 7 days a week, every single day of the year. Less than 5 minutes a year downtime is the standard.

Sun and IBM own this space - end of discussion. Microsoft simply has no credibility here. Mainframes run MVS (and Linux). Sun ES 10000s run Solaris Unix (and Linux). Neither runs Windows, or would want to.

Windows 2000 Datacenter and the Intel based servers it runs on may eventually scale into this realm, but only by clustering hundreds of servers, a management and maintenance nightmare.

Now IBM has hit the market with a new Web server based on it's S/390 G6 mainframe. The first installation runs 1,500 instances of Linux, each running the WebSphere Web server (a variety of Apache) and it has 11.4-terrabytes (11 million megabytes) of storage - but it will scale way larger than that. The economies of scale and ease of administration of this beast make it the only sane choice for large ISPs.

Financial Fragility

Microsoft is widely thought to have no debt. This is very far from the truth, but this debt, due to a loophole in GAP (Generally Accepted Accounting Practices) is not on the books. It is, however, just as real.

Microsoft's debt is in the form of employee stock options. Through most of its history, Microsoft has paid employees much less than other companies, and made it up by issuing stock options. These options have made many Microsoft employees very wealthy - on paper.

The problem here is simple - eventually the stock options have to be paid off in Microsoft stock. Should the value of that stock be in decline, with little hope of recovering soon, employees who's options are vested would be wise to exercise them and cash out. Microsoft has barely enough cash reserves to cover that possibility.

For a lot more material, including charts, check out the site of financial analyst Bill Parish, particularly Microsoft Financial Pyramid.

In the recent Microsoft stock slide, the company had to do something to keep this from happening. If they revalued the options, they would greatly anger outside investors who have no such protection, but if they didn't cover the loss in option value they could have a sell off, causing further decline in stock value, absorbing their cash reserves, and starting an exodus of key employees.

Microsoft compensated for loss of options value by issuing a lot more options to current option holders. While this staved off the immediate problem, it creates even more long term exposure.

Compounding the options problem, Microsoft stock is greatly over valued. The valuation is based on continuing the growth rate they have experienced in the past into the future. This growth rate is not sustainable as their traditional markets reach saturation.

Leveraging current monopolies into new markets seemed the way to go, but now the Department of Justice has a guilty verdict and will be moving to prevent that avenue of expansion even more.

Microsoft has kept revenue increasing by raising the costs to their customers by various means, primarily manipulation of license terms. How far can they squeeze before Linux and Star Office look like a way to end the pain to a significant number of customers?

The Anti-Trust Trial

"Microsoft thought they were rich enough to take on an organization that owns it's own aircraft carriers". They did not consider the Department of Justice a credible threat. Their previous experience with the DOJ certainly indicated this was the case.

Once again, Microsoft was blindsided by their own arrogance. The DOJ fielded a legal team that ate them alive. They still don't comprehend what happened and how serious it is. They are certain the court of appeals will let them off.

The court of appeals, if they get the case, might significantly ease the penalties judge Thomas Jackson has imposed, but it is not going to reverse the guilty verdict. If the Supreme Court takes the case directly, it could be even worse.

Microsoft's business practices are almost entirely based on leveraging the monopoly IBM handed them in 1981. Once this case has been appealed, the Department of Justice will be the one with the leverage, and Microsoft may even be split in half. How will they continue the growth rate that supports their stock valuation and staves off an options crisis?

Meanwhile, they are continuing full blast with the same illegal practices they were hauled into court for. Witness the incorporation of media players and video editing into Windows ME. This will come back to haunt them.

Meanwhile, they face two major anti-trust moves in Europe. One has been on hold pending the DOJ case, and will be released if the Europeans feel the DOJ settlement was insufficient. The other is steaming ahead now and concerns the way Windows 2000 was designed to eliminate competitors (details in our article Adopting Windows 2000.

140 Private Lawsuits

There's blood in the water, and the sharks are closing in. These private anti-trust cases are, for the most part, contrived, self serving suits concocted by lawyers for their own benefit. That doesn't make them any less real, and they must be defended. Microsoft has been found guilty, so these cases start with proof of guilt and only have to show damages.

Several of these lawsuits have already been dismissed and Microsoft has declared victory, but it's a very hollow victory. These cases were dismissed on a technicality, but that technicality doesn't exist in other states, California, for instance, where a case has been certified by a judge to go to trial.

More serious are cases in preparation by real companies (Sun is contemplating one) involving real damages. These too start out with guilt already proven.

What about the appeal? Couldn't Microsoft's guilty verdict be overturned? Not at all likely. The way Judge Jackson structured the case, Microsoft would have to show serious errors made by the court to reopen the finding of fact. That is highly unlikely. They can appeal the findings of law, and the penalty, but the guilty verdict will stand.

Should the Finding of Fact be reopened, well, the Department of Justice might actually enjoy that. They have a lot more evidence from other cases now, particularly Caldera, which Microsoft bought out of rather than face another guilty verdict, and the Bristol Technology case, where the judge just imposed a record fine for "testimony that was not true".

Has Microsoft ever been found "not guilty" in a major case? I can't think of one off hand.

Perma Temp Problems

Microsoft has used an unusually high number of temporary workers, some of whom have worked for them as if normal employees for many years. These "perma temps" do not have the rights and privileges normal employees have. They are definitely and blatantly treated as second class citizens.

Further, because Microsoft dominates the market for temporary employees in their region, they can dictate what temporaries are paid. Temporary employees in the Redmond area are paid significantly less than they would be in most other parts of the country.

Temporary employees have filed successive lawsuits against Microsoft, and won every one of them, by wide margins. The company has clearly been abusive. This is costing them large amounts of money in retroactive payments.

In one case, Microsoft presented their new temporary worker contract to the judge in an ongoing case. She took a look at it and told them it was the most absurd thing she'd ever seen, and to get the heck out of her court and not to come back until they had a reasonable contract to show her (literally).

To get around all these problems, Microsoft has instituted a mandatory multi-month "vacation" for temporary workers before they complete their 12th month of a contract. Because of the very high percentage of temps in their work force, they will now have severe problems with consistency in their projects.

Microsoft Has No Friends

During the court case brought by the Department of Justice, Microsoft depended on its own executives and paid experts to witness in it's defense. The one "friendly" witness they presented gave a very weak performance. The truth is, they could not trust their friends to testify, because they don't have any. There is hardly anyone in the industry who doesn't hold a grudge against Microsoft - only greed crates an illusion of loyalty.

The problem is simple: Microsoft treats it's "partners" exactly as it treats its enemies. Basically there are three steps. The difference is, "partners" get help from Microsoft during step 1, enemies don't.

  1. Let them build it
  2. Take it away from them
  3. Crush them with their own product

A perfect example is Citrix. Leveraging their OS/2 based technology, (remember, NT is just Microsoft's version of OS/2) Citrix built a multi-user version of Windows NT using source code they licensed from Microsoft. The "thin client" approach of Citrix MetaFrame achieved modest success with companies that needed to cut down the cost of administering Windows applications. Then "thin client" became all the rage in the press.

Microsoft went to Citrix and said (though they used more words than this), "We will not license you the source code for NT version 4, so you can no longer develop your product or keep it up to date, so, essentially, you are dead. Your only choice is to sell us your technology at the price we want to pay for it". Citrix obviously had no choice and sold Microsoft the technology now known as Windows Terminal Services at the price Microsoft was willing to pay.

In this case, Cirtix did not go out of business, because Microsoft proved incapable of properly implementing the product. Anyone who wants WTS to actually work still has to go to Citrix. Not Microsoft's intent, but that's what happened.

What this boils down to is: Microsoft commands no loyalty, and their "partners" will turn on them when the wind shifts.

Loss of Top Executives

Microsoft has been bleeding executives. Most of the top names that brought Microsoft to it's current dominant position are already gone. The anti-trust trial has taken it's toll, and lets face it, these guys are all now so rich they don't need the hassle any more. What they need to do is get out now and sell their stock before its value declines a lot more.

A few important execs who have left:

  • Paul Maritz - head of .NET initiative
  • Greg Maffei - CFO and ace negotiator.
  • Brad Silverman - chief developer of Windows 95 and Internet Explorer.
  • Jonathan Roberts - head of Windows CE development
  • Nathan Myrhvold - chief technology officer.
  • Peter Neupert - vice president internet group
  • Steve Perlman - founder of Web TV division
  • Ben Slivka - chief internet developer
  • Paul Allen - cofounder - resigned from board of directors after selling almost all his Microsoft stock.

Paul Maritz' departure was particularly telling and had an immediate effect on Microsoft's stock price. At this critical time, when Microsoft is trying to build credibility for its .Net initiative, the guy running the whole .Net show ups and quits without prior notice. How does the departure reflect on .Net? Can't be good.

The only Microsoft execs that thrive on hassle are the ones slinging it, Bill Gates and Balmer, and now Billy, "chief software architect" is mostly a spokesperson.

Slow Windows 2000 Sales

Microsoft says Windows 2000 is selling "above expectations". As far as anyone can tell they make up the "expectations" at the same time they make up the sales report.

Microsoft proudly announced 3 million copies were sold in the first month. They didn't crow much about the second month's sales being half that. It seems there have been further declines. Research firms are pretty sure Win2K is selling below projections, causing revenue shortfall.

The problem here is simple. It is well known the transition to Windows 2000 is very difficult due to having to deal with Active Directory. Further, Windows 2000 was so late Microsoft had to do something about Windows NT. The recent Service Packs for NT made it stable enough companies are no longer desperate to get rid of it. Then, of course, there is Linux, taking away server slots that would otherwise have gone to Windows 2000.

So companies already have a product they can live with, and face a whole lot of pain going to the new product. What would you do? I bet you'd hang tight with what you have.

Delivery of some important Windows 2000 developments, especially Back Office and Datacenter products have slipped significantly. With Microsoft's focus now shifted to .Net, this problem is likely to get worse.

Technical Stagnation

It is interesting to read the on-line discussion that follows many news articles on magazine Web sites. Microsoft's defenders now offer only two arguments. 1) Microsoft is so huge and so dominant nobody's going to change. 2) "I tried to install Linux and couldn't".

No one argues that Microsoft's technology is better, or more advanced, only that everyone is using it so they're going to continue using it, no matter what it costs.

Many years ago, the chief executive of a big software company made this statement. "The database belongs to Ashton Tate, the word processor to Micro Pro, and the spreadsheet to VisiCalc. There is simply no place for new software companies of any significance". These companies did indeed totally dominate their markets, with no significant competitors.

By the time Microsoft came on the applications scene, these companies were all either dead or near death, replaced by Lotus, FoxPro and WordPerfect. Those now dominated just as totally, but Microsoft overthrew them as well. Now Microsoft stands in that same position.

How did the stranglehold these companies had get broken? All were strong in their environments - but the environments changed, leaving a port of entry for other companies. The IBM PC did in the first round, the comprehensive Office Suite did in the second. Transition from the PC to network computing stands to do the same for Microsoft.

When your cash cows are at risk, you defend the cash cows by any means. This is what Microsoft has been doing, but they see their defense will be futile in the long run, so now they have .Net. Will they successfully rebuild their cash cows for a new environment? Monopolies don't often succeed at this sort of thing.

InfiniBand I/O

InfiniBand is a new I/O technology that offers 10 times the I/O performance of PCI-X (which is not even out yet). Unlike the I/O bus we are familiar with (ISA, PCI, etc.) InfiniBand is switched rather than shared.

InfiniBand will encourage major scale clustering of servers and data storage devices (primarily performance clustering). This is particularly important for large Web based enterprises.

Microsoft is a member of the InfiniBand Trade Association, along with 170 other companies including IBM, Compaq, Hewlett Packard, Dell, Intel and Sun Microsystems. Microsoft is not, however, a particularly enthusiastic member.

The problem for Microsoft is simple. None of their products are much more than "cluster aware". Their failover clustering is functional on a limited scale, but they have essentially no performance clustering, and Windows has never been an I/O champion. NT can't get close to saturating a gigabit Ethernet card, but Unix and Linux can.

Unix vendors have been using advanced clustering technology for many years, and Linux is a champion at performance clustering, allowing stacks of PCs to achieve supercomputer performance levels. In fact, the U.S. Postal Service uses about 900 Linux clusters to do OCR sorting of the mail.

When InfiniBand servers start shipping in the middle of 2001, it is doubtful Microsoft will be anywhere near ready to take advantage of them. Unix will be there, and, with IBM and Hewlett Packard pushing it for all it's worth, you can bet Linux will be there with bells on.

The Best and the Brightest

Microsoft has had its pick of the best and brightest graduates from the colleges, but the trend is now away from Microsoft. Without that stream of fresh, enthusiastic, young and easily manipulated talent, Microsoft will not be able to do business as usual.

Microsoft has traditionally offered grueling hours of hard work at low pay in return for sudden wealth in the form of stock options. They can no longer offer this. Their stock has been in decline and their growth rate is stalled. The grueling hours are still there, but the wealth is not. They've already had to significantly increase their hourly pay to compensate, but more will be needed, and this is expensive.

The new .com economy is where the action is, and the possibility of great wealth in return for years of not having a life to call your own - where fame and prestige lure. Microsoft is late to this game and playing catch up to much more agile companies, and even much larger companies (IBM).

Further, Microsoft has just been convicted of trying to screw an employee out of his stock options. Given Microsoft's public image these days, you'd have to be a little embarrassed to tell people you are a Microsoft employee.

Microsoft's public image has been badly tarnished by the Department of Justice case, and the many other law suits they have either lost or bought their way out of. They're still paying PR firms well over a quarter billion a year, but the public is learning the truth anyway. You have to be a little embarrassed to say you are a Microsoft employee today.

Microsoft's Public Image

Microsoft still pays public relations firms somewhere between a quarter and a half billion dollars a year to keep their image polished (that's "Billion" with a "B", and doesn't count their advertising budget or (suddenly very large) political contributions or (suddenly very large) political lobbying efforts). Even so, the likes of Waggener Edstrom and Edelman Worldwide haven't been able to keep it from tarnishing. More on PR here.

Until recently, Bill Gates was the untouchable hero of American business success - from nerd to riches. People listened to every word he spoke, and believed, and acted on those words. His company was the blazing success story for the American dream - and beyond criticism. Now, as more and more dirty laundry gets aired out in full public view, as people start to listen to stories of who actually originated Microsoft's "innovations", that image is slipping badly.

The turning point was probably the "Clintonesque" performance Bill Gates gave for the Department of Justice. Our "genius", with "the world's most powerful memory" suddenly couldn't understand the meaning of common English words, couldn't remember ever having sent any e-mail and didn't remember anything about the strategy or operation of his company. This sort of performance simply screams "guilt".

Now that people have opened their eyes, and see Microsoft losing one lawsuit after another, and buying others off so they don't go to trial, they see Microsoft's success in an entirely different light. For your convenience we have posted a compendium of the more prominent legal actions.

Unfortunately for Microsoft, they were not able to use the talents of Waggener Edstrom during the anti-trust trial, because WagEd isn't part of Microsoft, and could be subpoenad, and they would probably tell the truth, and they know lots and lots. Microsoft proved rather inept at handling their own PR, making a bad situation worse.

On top of all that, the quality (or lack of quality) of Microsoft's products has become a subject of open discussion, especially in light of the fast moving viruses that have cost businesses billions of dollars, and the "major Windows security hole of the week" announcements in all the magazines. People are getting tired of "Reformat, Reboot, Reinstall, and having every "upgrade" break everything else.

This change in public image is greatly affecting Microsoft's business practices, because too many people are looking to closely at their every move, and lawyers, armed with an anti-trust "guilty" verdict, are examining every move.

Despite all this, Microsoft still commands a very high, almost fanatically, level of support from individual users. Many are afraid Microsoft will lose it's monopoly, and they will have to learn something new.

©:Andrew Grygus - Automation Access - -
Velocity Networks: Network Consulting Service - Internet Service Provider - Web Page Design and Hosting
All trademarks and trade names are recognized as property of their owners