For nearly two decades every desktop in every office, of every programmer, secretary, manager or filing clerk has had a full-blown office 'productivity' suite, usually running on Windows. However, the desktop revolution during the 1980s and 1990s, and the mass adoption of DOS and Windows, had little to do with the virtues of the system itself. In those early years DOS and Windows were never more than "good enough." But "good enough" was all that was required and the rise of commodity hardware made the package relatively inexpensive. What the desktop PC offered the user was affordability and greater control of his or her working environment. Curiously these merits are commonly ascribed to Linux.
By common assent the Linux desktop is "ready." The Linux desktop is "good enough" and is getting better with each release. Linux offers most of the features that are available to the alternatives, and many features that are superior. It is cheaper and is supported by an expanding range of original equipment manufacturers (OEMs) and independent software vendors (ISVs). Most top of the range business software will run on Linux. It will run side by side with Windows on cheaper hardware, and offers the promise of more cost effective support options and vendor neutrality.
But to break Microsoft's hold on the desktop market and the general inertia of desktop users, Linux has to offer something more, something as compelling to Windows users as the PC was to an earlier generation.
For the home user the answer is simple. Linux offers freedom and flexibility. A Linux desktop is much more versatile and configurable than the Windows and Mac alternatives. You can learn much more and perform more tasks - all at zero cost. The hang-ups can be summed up in two words, iTunes and games. But even then there are ways around the problem. The answer for the office user is probably more complex.
We teach correct principles
The pivotal year for the mass adoption of the desktop PC was probably 1995, the year of Windows 95, when Microsoft finally took complete charge of the office. The tools that brought the PC into the office were the word processor and the spreadsheet, and not the DOS-based operating systems, which were only as sophisticated as they needed to be, and not as sophisticated as they should have been. During the 1980 the markets for "productivity suites" had been dominated by Lotus 1-2-3, VisiCalc and WordPerfect, but all were to fall by the wayside and become mere shadows of their former selves.
WordPerfect had been around since 1977. It and carried the admirable motto: "We teach correct principles and our employees teach themselves." But, by general consensus, the company ultimately failed because it was too nice a company, and Microsoft succeeded because it was too persistent and ruthless a competitor. At the time WordPerfect was generally acknowledged to be the superior product. It possessed a large and loyal user base in both the commercial and personal consumer markets but, by 1994, its share of the word processor market had begun to crumble in the face of a remorseless marketing campaign by Microsoft, aided and abetted by Microsoft's dominance of the PC operating system market. The company was sold to Novell that year for $1.4 billion, but Novell passed it on to Corel 18 months later for just $20 million in cash and $100 million in stock.
This was a fate that befell many of Microsoft's competitors before and after WordPerfect's demise. Productivity suites may have been the initial attraction that drew the desktop PC into the office, but the cement was the workgroup tools pioneered by Novell, which all but disappeared when Windows NT joined the party. Compared to NetWare's advanced networking functionality, Microsoft Windows for Workgroups was two tin cans and a piece of string, but Windows NT changed all that. Windows NT 3.1 was far from perfect, but it was "good enough" and Microsoft's marketing arm was "good enough" to paper over the considerable cracks until Linux and Samba came along to offer the possibility of a more resilient alternative.
Trade secrets
Jeremy Allison, co-creator of Samba, records that "In the days when Novell Netware dominated the file serving world Microsoft was a great supporter of standards. They published the specifications of their own protocols (then called Server Message Block, or SMB) and supported implementations on other platforms than Windows," but once Netware was defeated by Windows NT "their attitudes changed, and the flow of information stopped. Proprietary modifications to core protocols like the Kerberos authentication protocol followed, and these changes were treated as trade secrets, patented if possible, and only released under restrictive non-disclosure agreements, if released at all."
Netware and WordPerfect may have been the better products in their time, but like the Netscape web browser and the Real music player, the innovative product ultimately lost out to the product that came direct from the operating system vendor. By accident or design Microsoft's ownership of the desktop has translated into ownership of many of the important tools that depend upon the desktop.
When Microsoft purchased Internet Explorer (IE) from Spyglass, Netscape dominated the browser market. Microsoft solved the problem the easy way, by giving the browser away free, persuading OEMs to package IE with Windows, and by 'integrating' the browser into the OS. Giving IE away free cut Netscape's revenue stream and ensured a browser monopoly for Microsoft. IE was far from perfect but, after the fall of Netscape, there was a five year lull of no competition or innovation between the release of IE 6 and the release of IE 7, which in turn was provoked by the inevitable rise of the open source Firefox browser. Similar tales can be told about Microsoft's less than enthusiastic support for the cross-platform Java framework, and the subsequent arrival of the monolithic .NET framework.
Unlike the wider network where there is genuine competition, the desktop has not been subject to open standards, and this has had a deleterious effect on competition and innovation. This is the most important outcome of the European Court's recent decision to uphold the European Commission's judgment against Microsoft. While the mainstream press focused on the record fine imposed on Microsoft, the more engaging and decisive part of the judgment was the Commission's insistence that Microsoft publish its proprietary protocols to enable interoperability.
Thanks to the EU, and contrary to much media speculation from both sides of the water, we can now hope for a positive expansion of innovation and competition on the desktop. All the evidence suggests that control of all aspects of the market by one company is neither healthy nor conducive to innovation or competition. If all network protocols and "extensions" to network protocols were treated as trade secrets, there would be no networks and no meaningful shared protocols.
A level playing field
The long-term effects of the PC revolution have not always been beneficial or profitable for the organisation. The office suite on the PC long ago spread from its original preserve on the secretarial desk to the desktop of every worker in the company, sometimes with detrimental effects. Office suites have their virtues, but their overuse as a means of intradepartmental communications, and the working practices and culture that they promote, can get in the way of productive work and, paradoxically, increase the bureaucratic overload.
A side effect of this is that large organisations have increasing demands for the storage and retrieval of documents, accessible by all users, now and for years to come. Each office suite on each desktop comes at a premium, with a word processor, a spreadsheet and a visual presentation tool, crammed with features that are never used, and demands an upgrade every other year to conform with the current data formats.
The content hasn't changed. The functionality hasn't changed. But the upgrade is essential to keep the cycle going. And the Office format, which is usually undocumented, changes from release to release. Hence the demand from many governments for an office document standard, which is fulfilled by the Open Document Format (ODF). The ODF is supported as the default output format by most of the major office suites with the notable exception of Office, which controls the market. Microsoft appears reluctant to support the ODF.
A document format that is truly open levels the playing field for competing office suites. The publication of network protocols and application programming interfaces (APIs) provides a level playing field for desktop operating systems. This in itself won't lead to the widespread adoption of the Linux desktop, but will make it much easier. The more likely avenue for greater adoption of Linux is through government which has an overriding interest in the preservation of standards and is ultimately driven by cost. Linux is cheaper, OpenOffice is free and vendor lock-in is no longer a problem.
Last year's thing
As a competitor to Microsoft's suite, Linux and the free software stack offers something that the software giant has never had to compete with before. Linux plays well with other software, belongs to nobody, and is free. It is also supported by most of the major players in corporate IT.
Linux slots into a Windows network seamlessly, on the server side or the desktop. The forward-looking IT manager who is trying to reduce costs and expand the potential of their network will implement the Linux desktop on a gradual basis, one desktop at a time. Most users aren't power users and their desktops don't require power applications, meaning such users will barely notice a difference beyond a more customisable desktop. Furthermore, many enterprise applications are supported on Linux. OpenOffice can be implemented on Windows desktops. .NET applications and Windows desktops can slot into Linux networks. For those that don't, or the user that wants to move to Linux wholesale, there are companies such as MainSoft that offer a wide range of enterprise-level utilities for interacting and porting Visual Studio and .NET technologies to Java and Linux and vice-versa.
The compelling attraction of Linux is its versatility, the freedom it offers and its adherence to open standards. Free and open source software are the inevitable future for the software industry. The networking of everything from our fridges to our phones will demand an open future. From this perspective, the closed and monolithic desktop of the present already looks like last year's thing, an expensive indulgence that still fails to be "good enough" - in the unalluring shape of Vista - for the changing circumstances.
