Will the empire last long?
by Mario Miyojim for the Reallylinux.com Opinion Section

Recently, Microsoft started to sell a new operating system, Windows Vista. There are a few reasons to believe that, this time, it will not be successful.

Microsoft seemed like a new kind of player in the eighties that offered a low-cost operating system with development tools, a pretty face, easy learning, and promised to replace Unix with advantages. The Unix establishment did not see it coming, because they saw a toy that would never invade its dominion, the high level applications that drove big business.

Besides, it never occurred to all the incumbents that Microsoft would become a monopoly, since big IBM had once been defeated at that game, and jurisprudence would take care of it in case it tried. However, Microsoft used marketing skillfully and projected the image of an innovative, smart, advanced enterprise that would benefit all the common people worldwide.

Millions of IBM personal computers were sold at prices much lower than for Unix machines, with a text editor and spreadsheet functioning out-of-the-box, that is, one had only to plug the PC on the wall outlet, connect the keyboard and the monitor onto the system box and voilà, it worked!

The people did not realize they were being possibly misled, because perhaps the marketing of Microsoft made them believe that everything else, Unix machines, minicomputers, mainframes, were only for skillful and wealthy persons, and for the majority of people, there was only Windows. DOS (Disk Operating System), was not for the common person, however; special knowledge of typing, memorization of line commands was required.

The mouse was perceived as the device par excellence to communicate with a computer. The mouse was invented by Xerox, copied by Apple into its MacIntosh computer, which was copied by Microsoft into Windows. Its first Windows OS was just a collection of graphics superimposed on its text-based DOS. DR-DOS, by Digital Research, was in many ways better than MS-DOS. In the pre-launch phase of Windows 3.1, MS included a message that popped up whenever it detected DR-DOS, implying that it would not work well if it was not with MS-DOS (although it was not true). The world learned fast that only MS-DOS was good for Windows, so Digital Research went bankrupt overnight.


Software as an Industrial Product
Software is unlike industrial products. Trains, buses, buildings, are made of materials existing in finite quantities in the planet. There are limited sources of such items of material convenience. Software is abstract, difficult to measure, does not deteriorate over time, can be created by human minds without bounds, if the survival of their bodies is granted.

Having one's software be used by legions of people brings a kind of satisfaction not measurable in terms of money; it is a heroic experience. Microsoft, Apple, Oracle and other software makers seem to have treated software like industrial products. They shrink-wrap copies of software and charge unit prices for the license to use it under an End-User License Agreement (EULA).

Each copy is worth a given amount of money, and the source code is kept secret; a gold mine indeed. The new personal computers come with the current Microsoft OS installed by the assembler of the hardware, so the software is an integral part of an industrial product, for which one pays a significant price. Because a PC is to be used by a single person at a time, millions of computers with software are sold every year, and each copy turns in a handsome profit to Microsoft, year after year, higher than granted by its intrinsic value, similar to the fashion value one pays for the label of certain clothes.


An Alternative
First, there was the GNU Manifesto, where Richard Stallman told the world that it was unfair that software companies made much money by hiding the source code of useful programs. He dreamed of writing software that would be available as air is. It was the first effective impulse toward turning software into a commodity. He imagined people sharing software like households share kitchen recipes.

One person who was tuned to the concept of sharing software was Linus Torvalds, in Finland, a young, intelligent and idealistic man who had a good professor of software, Tanenbaum. Linus knew about the C compiler written by Richard Stallman (the gcc), and used it to tinker with his Personal Computer. Inspired by Minix, the OS used by Tanenbaum to teach about OSs, Linus wrote a functional kernel under the specifications of Posix, and made it available, over the internet, to whoever wanted to cooperate in the effort; this kernel became known as Linux. Since 1990, Linux has steadily grown, without fanfare or marketing, and has become a major force that Microsoft cannot destroy. The Linux kernel originated several distributions, collections of software with the Linux kernel, in packages ready to go.


Cracks in the Monolith
Cracks in MS's image started to show in the mid-nineties. A journalist noted the arrogance of certain Microsoft technical support personnel regarding the malfunctioning of some peripherals under Windows. Such technicians were so self-assured of their invincibility, that they dared blame the malfunctions on the lack of knowledge of customers. Viruses would infest the PCs and cause considerable discomfort. A new industry became necessary: anti-virus software. Viruses are possible because OSs have not been originally designed for security or quality, only for beauty, speed, and the wow factor.

At some point, MS realized that it needed a better OS, more stable and professional than DOS. They required technical talent from the VMS team from Digital Equipment Corporation. They were hired to adapt VMS to the Intel architecture. After a while, Windows NT 3.5 became available, which worked reliably on Intel architecture. But this OS had no images or sounds, no wow factor. Marketing ordered programmers to make images and sounds available on NT, and the programmers violated laws of software engineering in order to obey the order; this decision alone made the OS unreliable.

The result of that was Windows NT 4.0. It was more stable than the DOS-based OS's, but it still would suddenly collapse under stress after a few weeks of work, despite being touted by Bill Gates as being better than Unix. However, Windows NT 4.0 lacked certain features to make it a fully equipped network operating system. It took MS several years to come up with Windows 2000 (W2k), which originally was to be called Windows NT 5.0.

To me the decision on the W2k design was a turning point. MS had the opportunity to fully rewrite the OS into a true multiuser, stable, fresh, operating system. But to attain this goal, MS would have to break away from the existing source code tree and give up backward compatibility; third party application developers such as Adobe, Macromedia, would have to redesign and reimplement their packages to make them compatible with the new OS design. The danger was that these developers would realize that their packages would become multiuser, capable of functioning well with Unix and Linux, and the MS hegemony could be overriden. The gain of quality had the potential of becoming a huge loss.

So, the historical decision was not to redesign NT; in order to make the new OS look perhaps like a redesigned one, the programmers had to produce lots of patches. In fact, while NT4 had around ten million lines of source code, W2k had about thirty million! Quite a few patches, to disguise the true nature of the monster. W2k was still unstable, but now, to avoid the shame of the blue screen of death of Windows 98, it would reboot by itself when it felt ill. Then MS created Windows XP from W2k, adding about fifteen million lines of source code in the process--more patches! I had the opportunity to install Linux for a friend who was complaining against XP being slow and treating him like a stupid person all the time. Many people do not exactly like XP, yet they have to live with it, because they have no choice. Linux, on the other hand, cannot satisfy all their needs, or so they think. Their workplaces are tied to Windows applications, and have invested too much money in them.


A Vista Dawn or Dusk
Many years have passed after the launch of Windows XP, and now it seems it is the turn of Windows Vista. MS had to delay its launch, because the code base of XP could not be patched further to satisfy the new safety specifications. MS was reorganized internally and had to rewrite a good portion of the source code of the OS kernel in a hurry in order to satisfy the specifications. The Vista OS now has Digital Restrictions Management--DRM built into the kernel, whereby one cannot copy or play certain music or movie files, and the hardware has to be DRM-compliant. As a result of the new confusing restrictions and hardware requirements, very few people bought a license for Vista on January 30, nor in the following week. There was a sensitive decrease in the MSFT share value; probably the investors are deciding whether they still trust MS. The license price is too high for the average person.

Even Windows Vista's novelties are not yet fully reachable. A major question is the availability of hardware to obey restrictions imposed by Windows Vista. In that regard, one wonders which market force will win, inertia or Microsoft. Currently existing PCs running XP will be there for another 3 years without the need of special hardware.

From what we know of past challenges, one might think, Microsoft would obviously win, since it has plenty of time and money and marketing wisdom to wait until hardware makers issue DRM-compliant models especially for Windows Vista.

There is the possibility however that if Windows Vista does not sell soon, the MSFT stock value may nose-dive, because the stock market demands continual growth. Although Bill Gates has already sold a good deal of his shares to finance the Bill & Melinda foundation, he still has a lot more untouched. If he sells another large lot of shares now that the unit price is decreasing, people will interpret that as a sign of loss of faith in the company future. Asked about the sales of Vista to corporations that started in November 2006, Steven Ballmer, the CEO, said "Very well" instead of "We sold xxx licenses in three months", which is a telling attitude.


When History Catches Up
This is a uniquely challenging time for MS on other fronts, too: the EU has unfinished legal grievances against MS; China is adopting Linux for security reasons, because it cannot trust Microsoft; Munich, a German metropolis, is migrating to Linux for efficiency and security, and its example may spread throughout the European Union; Novell + SLES is gaining momentum; IBM is winning its dispute with SCO; the Ubuntu Linux distribution is winning devotees. MS can't afford to lower the unit prices, due to the decrease of its market share.

In such a challenging time, Microsoft seems entirely self centric in its strategy and seems to treat competitors like enemies in a permanent war, leaving no space for humanity; perhaps the strategy weakens its very foundation. Modern corporations, especially the multinationals, put feelings aside, and consider only competition in terms of monetary gain and the stock market to decide their business strategy, tending to be ruthless in their ambition to retain or grow market share.

I have reflected for long on the question of what I have known since 1995: the Microsoft empire could not last, while Linux and Open Source software would evolve and last for generations. Comparing the self centric take-no-prisoner MS with collaborative, freedom-based, fascinating GNU/Linux, I found that the major difference that gives them opposed futures has to do with something intangible but highly significant: humanity.

 

Written by Mario Miyojim this opinion article is published by reallylinux.com with permission.

It contains the opinions and personal experiences of the author at the time of publication. The opinions and personal experiences that have been posted do not necessarily express the opinions of Reallylinux.com. Linux is a registered trademark of Linus Torvalds. Microsoft, Microsoft Windows, Windows Vista are trademarks or registered trademarks of Microsoft Corporation both in the United States and Internationally. MS is used in certain cases to further denote Microsoft Corporation. RedHat is the registered trademark for RedHat Inc. All other trademarks or registered trademarks in this opinion piece belong to their respective owners.

Linux is a registered trademark of Linus Torvalds. Microsoft, Windows Vista, Vista are trademarks or registered trademarks of Microsoft Corporation in the United Statest and Internationally. All other trademarks or registered trademarks in this opinion piece belong to their respective owners.