March 1, 2007

Trends: Is Microsoft in Decline?

Filed under: IT Industry Trends — Black Falcon @ 11:31 pm

By Steve Naidamast
Black Falcon Software, Inc.

Like nations and empires, corporations follow a similar life and death process. And practically all such demise is self inflicted even if not intentional.

Microsoft has been at the zenith of its capability for some time now so it does stand to reason that the longer it retains such a dominant position in the Information Technology industry the closer it gets to that point where some level of decline begins to set in

However, several facts should be noted before we even attempt to claim the possibility that Microsoft may in fact be at the initial stages of a fall in. Yet, with the introduction of its Vista operating system, the company may now be at least at the precipice.

First and foremost, Microsoft Corporation has two products that still support approximately 60% of its revenue stream; its operating systems and its MS-Office product suites. It is surprising to note that such a limited product range has been the basis of Microsoft dominance for so long. As it may be the company’s current strength such a small level of diversity is also its greatest weakness in terms of the OS market arena alone. And as several analysts have already pointed out, the length of time that these two product areas have been supporting Microsoft is somewhat past when most other companies would have already experienced some level of market deterioration. Of course, Microsoft has a near monopoly in the PC systems area with an entrenched MS-Office user community both of which aid in the company withstanding the market trauma it has already experienced that would destroy just about any other such company.

In terms of the operating systems, you will be hard pressed to find any serious analyst who will support the notion that any of the Microsoft operating systems have been efforts of high quality. This is not to say that these systems are not easy to use or that they don’t function for the most part as needed by the people who use them. However, in terms of inherent quality and stability, Windows has never held a candle to UNIX, especially the Sun Microsystems variant of Solaris.

In terms of its newest competitor in the form of Linux, again Windows falls short from a technical quality perspective. According to analysts who have also studied Windows in comparison to the once vaunted IBM OS2 system, again it was found that Windows loses in the quality and stability arena. OS2 for all its difficulties was in fact always a superior system internally to that of Windows.

Surprisingly however, OS2 and Windows do share a commonality in the reasoning behind their designs. OS2 was meant to placate the growing microcomputer industry while at the same time maintaining, although indirectly, the support of their mainframe revenues. This was the reasoning behind the development of OS2 against the 286 chipset instead of the more powerful 386 which had become available at the time. By the time IBM caught on to its own mistake, OS2’s future had been sealed leaving the more popular Windows interface to gain dominance using a marketing perspective in its development to guarantee its own success.

However, Windows didn’t gain dominance based on its simplicity in use, though that aided it as the main attraction, but instead by superior marketing prowess. While Windows was gaining in popularity there were in fact UNIX variants that were appearing on the PC. However, too difficult for the average user to deal with they too fell by the wayside. And interestingly enough if one were to review the literal blunders made by Microsoft’s competitors in a host of areas (see Borland’s attempt to take on Microsoft in the early 1990s) it would be easy to come to the conclusion that Microsoft had a lot of help from its rivals along the way towards their upward trajectory.

Windows thus used its inherent capability of “ease of use” to gain the market share it coveted. However, the designers and implementers of the system never once regarded the quality of its internal base as an overriding necessity but instead saw it as a source of marketing control. Had they been so inclined, and being internals specialists, there never would have been any ability to integrate Microsoft’s other products via external pathways into the operating system. Instead, Microsoft’s Win32 API was made into an adjunct part for just about every major product that Microsoft produces.

In terms of the Win32 API, this has been a long known factor with Microsoft products. We can go down the list quite easily as to each product’s inherent attachment to the operating systems they run under. However, the major ones will suffice.

Internet Explorer was built into the OS until anti-trust and the EU forced Microsoft to disconnect it. Microsoft Office has innumerable attachments into this API. So to do the .NET frameworks which apparently have become more attached as they have evolved. This was found to be the case by a British Microsoft internals analyst (3rd party) in 2006 who became so disgusted with .NET’s reliance on the API that he gave up on Microsoft technologies altogether.

Of course, .NET’s entrenchment with the operating system has become somewhat severe with its involvement with the Vista OS. According to Microsoft, VS.NET 2005/SP1 is the only variant of .NET that is supported under the new system. And so to is it true that SQL-Server has followed the same route of growing dependence on an OS. SQL-Server 2000 and 7.0 are not supported under Vista. Only SQL-Server 2005 is supported.

All of this is a little odd when you consider that all of the mentioned products are 32bit applications and Vista is a 32bit OS with a 64bit variant.

The product tie-ins to the operating system has recently become one of the biggest complaints of some of Microsoft’s top talent leading quite a few of them to finally move on to freer pastures. This factor played a large part in many of the recent defections of such talent to Google Corporation in the past six months or so.

Recent reports also indicate that Microsoft has been hiring more marketing personnel than technical which indicates a cultural shift in the Microsoft institution though it was primarily bent on marketing since its inception.

This cultural shift, which has probably been going on for some time now, is also corroborated with the new emphasis on the gadget or “gizmo” industry. Microsoft’s new XBox gaming system is beginning to bloom and the supporting development environment is quite popular among game development enthusiasts. And recently, Micriosoft attempted to enter the music-pod sector with its Zunes gadget. IPod aficionados were not impressed. In addition to this, an industry analyst noted in the past few months that all of the Microsoft acquisitions in 2006 were of a consumer oriented nature.

All of this does indicate that if in fact Microsoft may be beginning an inherent decline it is in fact somewhat aware of it and is attempting to compartmentalize such a fall towards its two current “bread winners”; the operating system and MS-Office whereby if the OS goes so too will its major, ancillary product.

Aside from the products involved, Microsoft has had a growing history towards “market control” via a number of overt attempts to maintain its dominance. This became significant when in the 1990s Microsoft made such an attempt beyond both its growing coercive licensing agreements with hardware vendors and the less obtrusive product tie-ins to its operating systems. Now Microsoft attempted to co-opt the nascent Java language. Hardly had the language begun its own growth curve when Microsoft tried to gain control over it via its highly capable J++ development environment. And this is where Microsoft has often been found to be tactically stupid. Nonetheless, it can afford to be.

With J++, Microsoft had probably the best development environment at the time for Java. There was no need to attempt to dominate the Java language as well. It could have succeeded quite well by doing so indirectly by dominating the tools market that Java would require. But such an easy approach was simply not enough for the growing trend for complete market dominance. Had Microsoft followed this simpler route they would have skipped the law suits with Sun and the ill-will that were the subsequent results.

“Windows 2000” has been by far the most popular of the Microsoft operating systems ever. So much so that it was probably the most “pirated” Microsoft operating system as well which gave rise to the WindowsXP activation process… or at least so it was claimed. With this activation process, Microsoft began its shift towards direct engagement with controlling its market share than that of servicing it by forcing Windows OS users to now register their purchases which then precipitated enforcement of use instead of simply relying on the good will of the customer base at large.

This move was not well received by the Windows community and thus explains the reason for so many Windows 2000 machines still running on the planet. All of my own development machines still run this popular OS and right now I see little reason to change.

However, with the advent of Windows Vista, this trend has become severely overt which indicates not only the potential realization by Microsoft leaders of the competitive nature of the free Linux OS as well as the recently released “Open Solaris” initiative but of their growing paranoia as well. This last tends to be not just a character trait of wealthy individuals but of wealthy dominant corporations as well since loss of such dominance, especially in the Information Technology industry can happen very easily. And this is something that Bill Gates has often remarked on as Microsoft was growing, using it as an explanation for its severe protectionist measures.

Nonetheless, with Vista, Microsoft appears to have finally reached to far in what appears to be a somewhat desperate attempt to maintain itself in light of the slowly changing nature of the industry. Vista leaves Microsoft now as the only company charging fees for an operating system on the X86 architecture. And whether or not Linux or Solaris have substantial market share on the desktop or not, they have growing market share in the server market where Microsoft is still under fierce competition with Windows Server 2003. Thus, charging fees, especially high ones, for an OS in the long term does not appear to be a very good tactic.

However, it could be seen as a temporary measure to maintain revenues while the company undergoes yet another shift in strategic orientation as it did when it adopted its Internet strategy. However, this time whatever shift may be occurring with Microsoft will most likely be towards the consumer electronics arena.

Whatever the intent, the Vista OS has not been received well either by the consumer market or the corporate one. So far most observers either see a rather solid OS without a reason to upgrade or an OS with too many annoying problems; the two primary ones being the “User Access Control” or UAC and the increased ant-piracy mechanisms which are reflected in the more “robust” activation processes. There are other factors as well such as aspects of the new security and virtualization mechanisms which have also ruffled feathers of some long time partners with Microsoft.

However, from an overall cost perspective, not only is the new OS expensive to install, it also has an appetite for the latest in hardware if all of the new graphic features are a requirement. Cost being a major factor for most businesses these days has many companies shying away from a new expense which doesn’t hold any promise for significant savings in the long term.

Inertia in the market place may well keep most people and companies from upgrading making the eventual release of Vista Server an “also ran” while encouraging many to look at the free Linux distributions as an alternative to an expensive upgrade. In fact, many of the opportunities are already in place if leading Linux vendors can develop enough market prowess to encourage an overall market shift.

Vista as it stands now along with many of its ancillary technologies are in fact quite redundant leaving Microsoft with little in terms of actual new offerings.

In terms of operating systems, the Vista OS is most likely no more secure than a well secured Windows 2000 or XP machine considering that a lot of what has to do with security is related to the Human care given the machines in general. Nonetheless, no matter how secure Vista may be there are plenty of people who will make it their life mission to crack its defenses and at some point they will be successful making such an upgrade a rather moot point if solely based upon expected increases in security. Further, Symantec Corporation has recently released a report in which it demonstrates that whatever security enhancements have been incorporated into Vista are more or less “yesterday’s news” in terms operating system security. The new threats appear to be via third-party applications which Microsoft is still behind the curve on.

However, returning to a look at some of the other technologies that were initially to be only supported under Vista but have now been released in module form for WindowsXP appear to have very little going for them in terms of actual technical incentive for their adoption.

These technologies are the three primary advancements for developers to use in the creation of Vista-based applications. In this case we have Windows Presentation Framework (WPF), Windows Communication Framework (WCC), and Windows Workflow Framework (WFF).

In all three cases, none of these advancements offer any compelling case for their use except that Microsoft has once again thrown out an existing technology in order to replace it with something new they have cooked up.

In the case of Windows Presentation Framework, the entire internal methodology of creating Windows Form applications has been changed from an object based format to that of an XML format in the form of XAML or Extended Application Markup Language. Though this technology may be quite good the actual technical reason for this new level of uniformity in screen development would be so that applications could be made easily interchangeable between client-server interfaces and web-based ones. The problem here is that few applications ever really have a need to be one or the other once they are implemented. The decision to select either a client-server interface or one for the Internet is made in the beginning of an application’s lifecycle and there are substantive reasons for each of the two selections such as amount of use. The few applications that may go through such a process of interface change are far and few in between. In fact, in my 35 years in the field I can’t remember a single application that ever had such a requirement unless it was being completely re-written for other reasons.

Beyond that, XAML is just another technology to substantiate Microsoft’s love affair with XML frameworks.

The Windows Communication Framework is the one that has gotten me ruffled the most in terms of these new methodologies; primarily because I developed a product centered around remoting so I have a personal bias here. Nonetheless, when remoting was introduced with .NET, like ADO before it, this distributed technology was supposed to be the future foundation for all such application development. On top of remoting (or below it, whichever preference that is preferred) sat Windows Web Services based upon Web Service Extensions now in its 3.0 version.

For those in need of such technologies a lot of investment in time and money by companies, corporate developers, as well as independent developers such as myself in this case, has been put forth only to find out that the knowledge bases we have developed in the past few years is to be simply discarded.

Any internals specialist can make a complaint about these existing technologies and had they been well designed (which they appeared to be) they should have been easily extended to include revised internals and extensions. Instead we simply throw them out with the promise that the new framework is better, more full featured, and more powerful. However, I haven’t heard any recent statements to the effect that WCF is the future of distributed computing as once was claimed with remoting. Of course, the claim here is that WCF is supposed to more easily support the latest in industry buzzwords with the development of SOA (Service Oriented Architecture) applications. The claim is almost laughable since SOA is not supported by any specific technology but instead by an overall design which is mostly related to the information modeling developments in the 1970s.

Windows Workflow Foundation appears to be an add-on product that allows for the development of specific types of applications which invest in workflow processes. If the Microsoft product is well designed and those who opt for it like the disciplines and requirements this tool imposes than there is nothing to complain about. However, not all applications are designed in this vein and for those that require such formats other tools have undoubtedly already been developed making Microsoft a new-comer to this area.

Two new secondary technologies in the guise of DLINQ and XLING promise to achieve SQL coding consistency with diverse data formats as well as XML. This is all well and good but in these terms most applications deal with SQL based data or raw text-based data. Will we really gain anything by adding a new access layer over existing data formats? I think not as most developers will continue to program against diverse data sources in the same way they have always been programming against them; with the use of existing tools. Both of the “LINQ” tools again require another learning curve with not much to gain for it considering that the use of such diverse data formats does not involve most application development on a daily basis.

It is true that DLINQ does open the door with other enhancements to ADO.NET for Object Relational Mapping (ORM) with Microsoft products but such development is costly since SQL databases are relational in nature based upon rows and attributes, not objects. ORMs then are additional inefficiencies that most applications simply don’t require. As most technicians would agree, if there is a necessity to work with object structured data, than an object-oriented database should be used, not a relational one. As for XML, we already have parsers that are fairly efficient and using SQL against it does not appear to be a way to enhance such efficiency since SQL itself is inherently inefficient.

All in all, the technologies just discussed are solutions to problems that have already been solved… many times over. The enhanced efficiency to the existing solutions lies in incremental development not simply throwing them out for something claimed to be “better”. However, discarding existing technologies for the latest developments out of R&D do tend to boost a bottom-line if enough of the existing market-share using the existing technologies decides to incorporate the new techniques.

With this background in place Microsoft is then heading towards a precarious position as companies require more cost efficiencies while individual developers have tired of continuously spending their own monies and efforts to keep pace with the latest developments. And simply put, there is little wrong with existing technologies for application development and little wrong with the existing Windows operating systems in terms of functionality. There hasn’t been any clamor in the technical communities for a new Windows OS that I have seen. And why would there be since the current OSs are working fine? True, security has remained an elusive prospect for Windows and given the internals it cannot be achieved no matter what Microsoft does. By allowing such product integration as they have with their OSs they have opened up numerous pathways into the system for hackers. And given the marketing nature of Microsoft towards their operating systems it is doubtful that Vista will be any more secure once the initial defenses are cracked.

Community inertia for once is as big a problem for Microsoft’s new release as it is for corporations in general. With little requirement to upgrade compromised even further with ever more rigid anti-piracy functionality both of which come at increased cost, Microsoft may have finally reached the precipice of a downward slide as the dominant OS vendor for the PC desktop.

If we were to eliminate cost from the equation it could be seen that Microsoft may becoming irrelevant in the OS market when both Solaris and Linux offer capable alternatives with no initial expenditure. Of course we have community inertia towards these newcomers as well but with the enticement of free software along with mounting anger at Microsoft’s arrogance and erratic product directions there is little place here for Microsoft to go but down if they ignore such trends in the industry.

However I wouldn’t predict a mad rush to either a Linux or Solaris desktop any time in the near future but a lot of the factors for such a reorientation are already in place… and are in the process of becoming growing industry forces. A quick look at a lot of the technical blogs as well as recent industry analysis will demonstrate the disillusionment that is rising against the Microsoft behemoth. And the increasing development of easy to use solutions under Linux will eventually make Microsoft’s only claim to relevance a thing of the past.

Though there are many things Microsoft can do to rescue itself from its own self-made predicament it is highly unlikely that it will retreat from a position that it has taken years to develop. Companies at such turning points tend not to retreat but simply move forward… and over that proverbial cliff. As it is, Microsoft no longer just wants to be your operating system of choice it now wants to make sure that it remains your one and only choice. As long as Microsoft retains this view of itself it may find itself tacking against the gathering winds of change…

No Comments »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a comment

Blog at