A foundation for the desktop – one apple, two ideas
The often predicted tipping point for Linux on the desktop just never seems to materialise, despite a variety of attempts from numerous companies. Richard Hillesley takes us through the history of these attempts, analysing the reasons for failure, and has a modest proposal of his own to offer.
"If you have an apple and I have an apple and we exchange apples then you and I will still each have one apple. But if you have an idea and I have an idea and we exchange these ideas, then each of us will have two ideas" – George Bernard Shaw
The story of the free software desktop is littered with what-ifs and might-have-beens. The desktop has been 'good enough' for years, and can boast some considerable success stories, but has yet to make a significant breakthrough.
On the face of it, the free software desktop should be an easy choice. The average GNU/Linux desktop costs little, looks good and performs well, and offers a real opportunity to break the upgrade cycle. Cost, security, scalability and versatility are persuasive arguments for the free desktop, but other factors have worked against the uptake of Linux at the corporate level.
Inertia among users is usually given as the reason and users are made to take the blame, but perhaps there are simpler explanations. The desktop has been left in the hands of the Linux companies, and the Linux companies are many and small. Many have come with grand ambitions and some with high ideals, but few have stayed the course.
As a result the free desktop has lacked consistency and continuity.
The spoils of the desktop are divided between GNOME and KDE. Both have their virtues, and neither is better than the other. The desktop configurations change, there is no unity around GNOME 3, and even less around Unity. There are no universal repositories for the software. Some bits are stored as rpms, some are saved as debs, and some will work with one desktop but not another. These kinds of things matter more to the user than the developer, and are more of an issue on the desktop than on the server, where reliability is the watchword and change is infrequent.
The service model for raising revenue that has worked so well for Red Hat in the server space works less well in the desktop market. It isn't obvious how a company working on its own can hope to change the landscape, and pay for itself.
And yet there are ways around this, neatly summarised by an aphorism of George Bernard Shaw: "If you have an apple and I have an apple and we exchange apples then you and I will still each have one apple. But if you have an idea and I have an idea and we exchange these ideas, then each of us will have two ideas."
If the Linux companies – distributors and consultancies – were to pool their resources, then collectively they would have a greater reservoir to draw from: of support, of marketing, of commonality and process. Collaboration works, and if the Linux companies can work together, they can also hope to draw on the support of the community, and the many OEMs, ISVs, governments, corporate users and device manufacturers who have a common interest in making the free desktop work as a vehicle for development and innovation.
Adopting Linux isn't just about changing one horse for another. Migrating from one platform to another is a big and risky task that is often complicated by fractured OEM deals, broken data, lost applications, single vendor tie-ins, interoperability log-jams and the transition process itself.
Even the smallest of companies may have tie-ins to Windows and proprietary legacies that are hard to surmount. Companies need visible support to negotiate their way through the obstacles, imagined or real. Open APIs and standards provide long term guarantees of data integrity and accessibility, but can only be realised after the random use of Access databases, .NET and VB scripts, Active-X, Active Directory and countless other proprietary features in diverse parts of the organisation have been replicated or immunised. And if you are going to make such a move, it makes sense to re-imagine the data and the business process.
As the experience of Munich City Council and others demonstrate, such problems can be surprisingly troublesome and time consuming. But Munich also demonstrates that patience pays off, and in the long term, the implementation of the free desktop reaps not only a financial reward but also a streamlining of the process, which brings with it other rewards.
For most companies the primary reason for moving from Windows to Linux is perceived cost savings. The secondary reasons are factors such as interoperability and greater compliance with standards, which themselves bring longer term cost benefits. Selling Linux on this basis should be easy, but there are too many other unknowns. Management is confronted with a variety of distributions, each of which is subtly different to the others, and may or may not be certified to use the commercial applications the company wants to use.
This predicament is compounded by lack of obvious support structures which can be used to transfer blame and offset risk, which is why companies like to fall back on known suppliers. The major OEMs are not pushing Linux on the desktop and, in contrast to server-side Linux, few of the Linux companies have visible support structures in place to give corporate users the confidence to make the switch, yet most cities in most countries will have individuals or consultancies available with considerable expertise in the business of supporting Linux.