Table of contents

1. Introduction

2. Adoption cycle

3. Desktop environments

4. Year of desktop linux

5. Hardware support

6. Connectivity with windows

7. Linux distribution

8. Application

9. Performance

10. Market adoption

11. Education

12. Future of desktop linux

13. Conclusion

Introduction
“Bill Gates was the first to realize the power of ‘good enough.’ We don’t have to offer the perfect desktop environment, just one that’s good enough for most people.”

[pic]
Desktop Linux, also called Linux on the desktop, refers to specialized Linux distributions with features designed for desktop personal computer users. These features include a graphical user interface and personal use applications.
Some Linux distributions have targeted the desktop role specifically, while others include all the software available for the platform. In that case, the user will be given the choice to select either a "desktop" or "server" type when the operating system is installed.

THE ADOPTION CYCLE

Looking back
Many technologies have enjoyed widespread success when a new application was released that allowed the benefits of the new technology to fully demonstrate themselves. The Macintosh computer, for instance, started gaining momentum with the release of Adobe Pagemaker and the postscript printer, creating the desktop publishing industry. Likewise, the Internet was used only in academic and technical circles until the release of the Mosaic web browser. As a result, some people believe that the emergence of a “ killer app ”will be critical to broader desktop adoption of Linux.

Examining the adoption of Linux servers reveals a different dynamic of adoption, bearing some striking similarities to the adoption cycle of the personal computer:from the edge towards the center. First adopted on the fringes of the enterprise, Linux is now well on its way to the center of the enterprise, the data-center. Until around 1997, Linux use was largely restricted to researchers,scientists, “hackers” (enthusiasts) and users in the academic communities. The dot-com boom lead to much broad adoption of Linux servers, with ISPs and dot-coms leading the way. Linux wastypically deployed “on the edge” for mail, web, DHCP or DNS servers, for instance. Medium-sized and larger corporations started deploying Linux also as departmental file- and print-servers. Having proven its reliability as a reliable, cost-effective work-horse, Linux is now being deployed more centrally. It is increasingly used as the basis for databases, application servers, and is even beginning to support the ERP and CRM applications at the center of the data-center. In a February study by
IDC, 14% of respondents indicated that they were supporting database software on Linux, versus 2% who reported supporting ERP or CRM applications.

While web- and mail-server applications could be described as “killer apps” for the Linux server, Linux has always been more of a Swiss-army knife. Its real advantages have been the broad range of available tools and applications, similarity to UNIX, broad implementation of open standards and protocols, low deployment cost, and, perhaps most importantly, the thriving community of users and developers and the rapid cycle of bug fixes and innovation that flow from the open source nature of the OS.
Open source has been the real killer feature of Linux , and will continue to be as Linux takes hold on the desktop, with early adopters solving problems they encounter and contributing those changes back to all users using the open source development process.

Factors and trends that are favoring Linux desktop adoption
We do not believe that there will be a killer app that will propel Linux desktop adoption. For one thing, innovative applications, especially open source ones, tend to be network-oriented and can be relatively easily ported to different operating systems. Instead, we believe that a number of factors and trends will lead to more broad-based adoption of Linux as a client user environment. It remains to be seen what the relative strength of these factors will be, and how powerful their cumulative impact will be.

1.The success of Linux as a server computing platform
-As Linux proves itself as a server operating system and the Linux eco-system grows, existing and potential customers begin to consider it more seriously for the desktop as well.
2 The emergence of credible desktop offerings, including the availability of solid productivity applications that interoperate with their Windows counter-parts
-The KDE and GNOME desktops have reached a level of maturity where they offer an acceptable end-user experience. The availability of strong cross-platform application suites, including the Mozilla/Netscape browser suite and the OpenOffice/StarOffice office suite, complemented with other core productivity applications that interoperate and offer a similar user-interface to their Windows counterparts, such as Ximian’s Evolution PIM suite, are the cornerstones of a credible desktop offering. The availability of client software allowing users to access their organization’s transactional applications is the next critical step. Less visible to most users, but critical to people with disabilities and to government adoption of Linux on the desktop, is the progress that, notably, the GNOME project has made in the area of accessibility, which is the reason that the Department of Defense is now able to consider Linux as a desktop alternative.
3 The commodification of the operating system
-There have been few dramatic changes on the desktop in recent years. Many users feel that the desktop and the office productivity suites are essentially “done”: they are mature products that work reasonably well. As a result, consumers and corporate users alike are increasingly reluctant to upgrade their software. With the phase of rapid innovation essentially over, the desktop operating system is becoming a commodity product. The increasingly web-centric nature of computing also contributes to the commodification of the OS: the OS is becoming less important for some users than the web browser, for instance. These factors make it easier for alternatives such as Linux-based desktops to catch up in terms of the functionality and the user experience they offer, and provide a “good enough” alternative for many users. Microsoft and other stakeholders, such as OEMs, will make efforts to fight this trend but, unless a new killer application appears that dramatically raises the stakes in terms of desktop software requirements, their efforts to sell upgrades to their existing customers will increasingly meet with resistance from end-users, especially in tough economic times.
4 Structural challenges for the monopoly provider
-Facing market penetration reaching saturation levels in the most lucrative markets, and pressure to maintain an astronomic market capitalization that is fueled by expectations of continued rapid revenues growth, Microsoft may have little choice but to resort to increasingly aggressive upgrade strategies (such as its new licensing practices) for the desktop operating system and productivity suite that bring in the bulk of its revenues, encouraging customers to look for cheaper replacements for commoditized software. This may have been the reason for Microsoft’s Licensing 6.0 and the subsequent changes in their business practices.
5 The falling cost of personal computers
-With new computers now for sale at Walmart and elsewhere for less than $200, the cost of the OS and other bundled Microsoft software has become the most expensive component of a PC. Vendors operating on razor-thin margins and price-conscious customers will naturally look to save on the most expensive components.

Desktop environments

[pic]

KDE desktop environment
Linux offers many desktop alternatives. The most popular desktop environments are GNOME, KDE and Xfce. These are rather large collections of desktop programs, rather than bare bone window managers such as FVWM, IceWM amongst numerous others. These environments present a GUI using a desktop metaphor. All of these environments allow the user to set many personal preferences and to perform common system management tasks. The Compiz window manager expands upon the concept of having multiple desktops by rendering each of the user's two dimensional desktops on the surface of a three dimensional cube.
Regardless of the source, all Linux desktop applications use the X Window System and thus benefit from features like networking (remote display) and quick mouse-only cut, copy, and paste. Mouse selected text is automatically copied, and it can then be pasted using a middle click, without the need to resort to use of the keyboard.
Desktop Linux environments have been improving in appearance and overall integration over time. This has led to wider adoption of the platform over time.
Year of Desktop Linux
Since at least 2000, a meme known as "20XX will be the year of Linux on the desktop" or "year of the Linux desktop" (YoLD) has been published by a number of tech-related magazines and websites. This refers to an expected breakthrough in Linux adoption by business corporations and personal users. The reasons for these forecast gains include the installation of a Linux distribution on the desktops of workers for organizations or companies who may not be immediately or otherwise involved in the computing industry, or the acceleration of development for specific applications which find their greatest use on desktop Linux distributions. It can also refer to a breakthrough in the number of pre-installed Linux distributions on personal computers being sold by PC manufacturers such as Dell and Hewlett-Packard.

Hardware support

Main devices
Linux can be easily installed on most desktop computers sold by major vendors these days, thanks to solid device support in the Linux kernel and with leading Linux distributions offering easy to use graphical installer software that features good automated hardware detection. It can still be a challenge, however, to install Linux on many popular laptops. Power management or screen brightness adjustment, for instance, are poorly supported on many laptops. And if the Linux installer doesn’t identify the floppy drive or CD/DVD reader in a laptop’s docking station, installing Linux can quickly become a daunting task.
This is more of a problem for reviewers (including, significantly, reporters and IT decision-makers) then for most average end-users, since few users ever attempt the daunting task of installing an OS on a computer. Enterprise users get their computers from system administrators who can either purchase systems with Linux pre-installed or who have a centralized system that automates the process of installing operating systems on a PC. Average home users, of course, use the operating system that came pre-installed when they purchased their computer.

Connectivity with Windows

There are any number of ways for Linux users to interoperate with Windows users:
- Dual booting: Linux and Windows partitions can be installed on the same computer, allowing the user to choose at start-up time which OS they wish to boot in;
- Network access: Linux users can more or less readily access Windows network volumes using the SMB protocol. Lycoris, Lindows and Ximian users, for instance, have access to a GUI network device utility that is very similar to My Network Places under Windows, allowing them to easily access Windows volumes;
- Virtual machines: solutions like VMWare and Win4Lin allow Linux users to install Microsoft Office as a virtual machine, and run Windows, complete with any Windows application, in a window on their Linux desktop;
- Terminal servers: products such as Citrix’s Linux client or Tarantella’s Enterprise 3 allow Linux users to remotely run Windows applications in a window on their Linux desktop;
- Windows emulation: WINE and its commercial offspring, CrossOver Office/Plugin and WineX allow users to run Windows applications as-is under Linux. This works great for Microsoft Office, Internet Explorer, Windows Media Player, and Adobe Photoshop, for instance. Other applications, including Macromedia Dreamweaver, almost work – these applications either crash occasionally, or are lacking specific functionality (such as FTP support in Dreamweaver);
- File format compatibility/functional equivalents: As discussed above, applications such as Ximian Evolution and OpenOffice offer varying degrees of file format compatibility, cloned functionality or ability to integrate in a Windows environment.
All in all, there is an impressive range of connectivity solutions for Linux users who need to interoperate with Windows users. These solutions go a long way towards solving the compatibility problem for Linux.

Linux distributions

The Linux project lead by Linus Torvalds is not an operating system. It is the kernel of an operating system. Linux vendors such as Red Hat and open source efforts such as the Debian project, focus on offering complete operating systems, which include hundreds of drivers, utilities, development tools and both server- and end-user applications.
While Red Hat is the global leader in the Linux industry by a significant margin, it has a healthy number of competitors, including SuSE and MandrakeSoft in Europe, Red Flag Linux, Turbo Linux, and Hancom Linux in Asia and Connectiva in Latin America. Thanks to the open source nature of the OS, there are few barriers to entry, and there are customized Linux distributions for numerous countries and user profiles. In the US, for instance, many universities maintain their own Linux distributions, often derivatives of Red Hat Linux with just minor modifications.
There are also a number of open source collaborations that produce Linux distributions, most prominently the Debian and Gentoo projects.
Debian, it so happens, is also the basis for two of the most visible desktop-oriented Linux distributions, Lindows and Xandros.
While a number of Linux distributions have gone out of business or are struggling financially, and while Red Hat is increasingly emerging as the global brand leader, there is a rich ecosystem of Linux distributions that each meet unique niches.
These projects compete in the market place, but there is often close cooperation at the engineering level, and there are some business partnerships (eg. United Linux).
Applications
[pic]
Most Linux distributions provide a program (e.g. Synaptic) for browsing a list of thousands of free software applications that have already been tested and configured for a specific distribution. These free programs can be downloaded and installed with one mouse click. A digital signature guarantees that the program was not modified after the download package was created and the package manager automatically tracks and installs updates as they become available.
Besides the free Windows compatibility layer Wine, most distributions offer dual boot and x86 virtualization for running both Linux and Windows on the same computer.
Performance
[pic]
The performance of Linux on the desktop has been a controversial topic, with at least one key Linux kernel developer, Con Kolivas, accusing the Linux community of favouring performance on servers. He quit Linux development because he was frustrated with this lack of focus on the desktop, and then gave a 'tell all' interview on the topic.
Other sources, such as mainstream press The Economist disagree with this assessment that there has not been enough focus on desktop Linux, saying in December 2007:
Linux has swiftly become popular in small businesses and the home...That’s largely the doing of Gutsy Gibbon, the code-name for the Ubuntu 7.10 from Canonical. Along with distributions such as Linspire, Mint, Xandros, OpenSUSE and gOS, Ubuntu (and its siblings Kubuntu, Edubuntu and Xubuntu) has smoothed most of Linux’s geeky edges while polishing it for the desktop...It’s now simpler to set up and configure than Windows.

News, conferences and support resources

There is a broad range of information materials available to people interested in Linux, ranging from publications such as Linux Journal and Linux Magazine to news sites such as Linuxworld, Linux Today, Linux Weekly News and Slashdot. People interested in Linux on the desktop can go to
Desktoplinux.com, among others.
There are also a number of Linux trade shows and conferences, including Linux World, Linux Tag, Ottowa Linux Symposium, Enterprise Linux forums, the GNOME project’s GUADEC conference, the Open Sources conference, Lindows’ Desktop Linux Conference and more.
Most importantly, perhaps, there are hundreds of Linux Users Groups, mailing lists and message boards where Linux users can go for help. The mailing lists and message boards offer an unmatched online support system that allows Linux experts to solve their problems. Many of these lists, however, are inaccessible to end-users, given the often cryptic and jargon-laden way in which the experts offer assistance. End-user oriented businesses, and Lindows in specific, do offer lively discussion boards that solely focus on helping end-users, but finding simple, end-user oriented answers to user-space questions such as “My DVD burner doesn’t work. What do I do?” can be a daunting task in the Linux world.

Market Adoption

Last July, IDC released a report on Linux desktop adoption where it anticipated an installed base of 9.5 million Linux desktop for 2003, a number it expects to grow to 27.8 million desktop users by 2006 (versus 9 million Linux servers). IDC predicts a CAGR of 44% for desktop Linux over the next 4 years and, by 2006, a marketshare of 7% of the installed base and 10% of new unit shipments.
According to IDC, Linux desktop shipments outstripped Macintosh shipments in 2002. By 2006, Linux will likely have a larger installed base than the Macintosh OS.
We should remind readers that counting Linux adoption is an inexact science at best due to the open source nature of the OS: people can download Linux from any number of mirror web sites, burn and redistribute their own CDs, or install Linux on thousands of desktops across an enterprise from a single purchased copy. Also, IDC’s surveys indicate that Linux users tend to upgrade their systems at a much higher rate than users of other operating systems. Still, the general picture is clear: Linux is already in use by millions of users and these numbers will continue to increase rapidly, with industry analysts seeing Linux take up anywhere between 7% and 20% of the installed base by the end of the decade, and Linux outpacing the Macintosh OS as the alternative to Windows on the desktop over the next few years.

Education

Like small businesses, educational institutions are very price-sensitive. Like enterprises, they are relatively centralized institutions. At first glance, this would seem to indicate that they hold great promise for Linux. There are significant opportunities, but Linux is not the right solution for every educational institution.
In the K-12 sector, there are major obstacles to broad-based desktop Linux adoption. Cash-strapped schools and other educational institutions are eager to realize cost-savings. On the other hand, they are reluctant to train students on non-mainstream tools and K-12 institutions in particular rely largely on educational software that is not currently available on Linux. In K-12 institutions, maintaining computers often depends on the ingenuity of teachers, students and volunteers, who are most likely not familiar with Linux. Nonetheless, there is a community of K-12 educators that focuses on introducing open source solutions to elementary and secondary schools, at SchoolForge18. And, of course, there are major deployments taking place outside of the United States, such as the deployment of 80,000 Linux desktops in Spanish schools.
Things look more promising in universities and colleges:
- Many universities are traditionally UNIX shops that have started to migrate much of their server infrastructure to Linux running on PCs, in order to save costs. System administrators may be reluctant to become beholden to Microsoft. Many campuses have active Linux Users Groups that are advocates for broader Linux adoption;
- Graduate students and faculty in the sciences are already using Linux workstations in large numbers. If they’re not, they probably log on to UNIX or Linux computing systems to run simulation and other computationally intense tasks. When this type of deployment of Linux workstations reaches critical mass, entire departments may standardize on Linux. The department, in turn, serves as a test-bed and model for other departments;
- Data archivists on campus technology policy committees worry about the ability to access old documents as technology evolves. Techies on the committee worry about open standards more generally. They may work together to recommend Linux desktop deployments to managers eager to save money. And many universities are quite centralized, with central technology offices able to impose standards and practices campus-wide.
As a result, we predict that, over the next 2 years, there will be an increasing number of desktop Linux deployments on universities. Many of these deployments will happen in sciences departments. We expect one of the drivers of this development to be a movement towards open file formats. We expect that a number of universities will respond to librarians’ and IT staff’s concerns about open file formats by standardizing on open file formats for data interchange. The most likely beneficiary of these policies is likely to be the OpenOffice suite. We believe that more and more universities will set up OpenOffice in public access computer labs and making the application available to their students.
While initial deployments of OpenOffice will mostly happen on the existing Windows computers, once students switch to OpenOffice, a major barrier to Linux adoption has been removed.
..

THE FUTURE

Assumptions
- The economic environment will moderately improve over the next two years. A continued slump would further accelerate the deployment of Linux on the desktop: the longer than expected economic downturn is already driving enterprises and public agencies to look harder at their IT budgets and consider inexpensive alternatives;
- There will be no major discontinuity in the ability of Linux productivity software to read or write Microsoft Office documents. If Microsoft were to totally break with its current default file formats, that could set back file format compatibility. This is quite unlikely though, given customer pressures that Microsoft is facing. Conversely, if Microsoft’s move to XML-based file formats were to make it dramatically easier to offer file compatibility, that would benefit Linux, as it would remove the biggest single obstacle to broader Linux adoption. This, too, is unlikely. Based on early reviews of Microsoft Office 2003, OpenOffice developers claim that when saving documents in XML format, Office 2003 strips XML files of all presentation and formatting information31;
- Despite intense lobbying efforts and aggressive price cuts, Microsoft will not be able to stop the flood of government-led efforts around the world to promote the adoption of Linux on the desktop, since cost considerations are just one of the motivations for these efforts;
- Microsoft will try to avoid repeating marketing blunders like the introduction of its new licensing scheme, but will continue to aggressively steer increasingly reluctant enterprises to upgrade their desktop software. Now that the Linux desktop has matured, companies will be much more willing to consider moving to Linux if Microsoft commits further major blunders in the way it maintains its relationships with major enterprises;
- There will be no dramatic changes in desktop computing over the next few years;
- The SCO lawsuit against IBM will be resolved in a way that does not jeopardize the future of Linux;
- It will take several more years for WINE to be able to run the majority of third party Windows applications flawlessly. As discussed below, an acceleration of the development of WINE could remove, or at least reduce, one of the major obstacles to consumer adoption of Linux.

Predictions

- Responding to efforts such as the People’s PC project in Thailand, Microsoft will dramatically cut prices in selected developing countries. Microsoft has incentives to aggressively cut prices in some parts of the world: offering lower-cost versions of its software not only helps beat back Linux-based desktops, it may also benefit the bottom-line, even in the long term.
The incidence of software piracy in developing countries is very high, often more than 90%, so the company can increase revenues if, by dramatically lowering its prices, it can boost legitimate installations. Making its software more affordable is a way to combat piracy and maintain its dominance in the desktop market;
- The Linux desktop will continue its current pace of gradual improvements and will be increasingly perceived as a reasonably mature, “good enough” alternative to Windows;
- Within the next four years, Linux may achieve as much as 10% market share worldwide on the desktop and will account for an even larger percentage of new shipments; Linux desktop deployment will first happen outside of the United States and will be fueled to a significant extent by public sector deployments similar to recent moves in Munich, Extremadura and Thailand;
- Linux desktops will be adopted in enterprise environments over the next four years, but, in the United States, almost exclusively by highly technical workers and transactional workers;
- Linux adoption by a significant number of consumers and productivity workers in the US is at least four years away;
- The success of Linux on the desktop will not lead to the emergence of a significant consumer ISV industry over the next few years. This is one of the factors holding back consumer adoption of Linux and there is no easy solution to this chicken and egg problem.

Conclusion

“Desktop Linux is no longer a technical challenge – it’s a marketing challenge.”
Making the Linux desktop more user-friendly, elegant, and interoperable with competing desktops is a never-ending challenge. The effort has reached the point where, in many regards, Linux is “good enough” for significant classes of users. If the current pace of improvements continues, Linux will emerge as a mature and credible desktop alternative for tens of millions of users over the next few years. If the current interest by the public sector in promoting open source persists, and in the absence of other major shifts, Linux is slated to achieve a market share of as much as 10% over the next four years. The process of desktop Linux adoption can be accelerated by some of the efforts listed in this document.