« previous 1 2 3

What can you give us?

IBM recently commissioned a report from FreeForm Dynamics on 'Linux deployment on the Desktop'. Sutor summarises the report's conclusions on the factors that are "slowing people down in moving to the Linux desktop, and a lot of it boils down to terms like politics, reticence, or user inertia - 'I'm using this. Oh Gosh, don't make me change.'"

"By politics," Sutor suggests, "we mean some association that people have with the Microsoft ecosystem - 'I like being part of that. It gives me a sense of identity.'"

Unfortunately, the reasons for not moving to Linux "are not technical reasons. They are not economic reasons, and they are not business reasons. They often come down to 'Well, maybe I should do that, but I just don't want to' and this has to be countered in interesting ways. Maybe people will do it if their friends do it. Maybe people will do it if their boss does it. You find the champions. You find the people at the right levels of the organisation who will do it for themselves."

"IBM are slightly less likely to hear a customer come to us and say 'We want to move to Linux on the desktop. How will you do this for us?'" says Sutor, than "we are to hear customers say "'We already have a sizeable number of people who are moving to Linux. What can you give us in terms of applications?'"

For Sutor, the story is "really a question of economics."

Using the bandwidth

One of the many niches that has opened up for Linux has been the emergence of virtualisation technologies. On the client and server sides there is a reduction in hardware requirements. Money is saved that can be channeled into other parts of the business, not just on the hardware, but on up-keep and maintenance, idle time, unused processing and storage capacity, security concerns, energy bills, upgrades and personnel.

The challenge that organisations face is how to maintain expensive software, how to ensure that the code is clean and protected, and how to ensure that each user has the latest version. The PCs that sit on every desktop of every worker in the organisation make a disproportionate hit on the organisation's cash flow, not just in the initial investment in the hardware, but in licensing the software, and servicing hardware upgrades to satisfy upgrades to the software, whilst keeping track of company data that may sit on random PCs anywhere in the organisation, and may be lost forever from the company pool of knowledge.

There is an obvious attraction in the ability to dynamically download code and data as it is needed. Thin client or server-based computing reduces the requirement for high performance computers on every desktop, and vastly reduces the overheads for storage, maintenance and cooling. "The advantage is security, because it controls access to company information," Sutor notes. "Updating becomes much easier, and it simplifies maintenance and reduces costs." Virtualised desktops don't require the same power or storage capabilities, and can run on older machines without any noticeable deterioration of performance.

The classic deployment of virtualised desktops is for "large groups of people who have consistent and similar requirements, where they might access their browser and specific subsets of applications and data. This might apply to large numbers of people who for instance work in stock exchanges or hospitals, where you don't want a massive variation from laptop to laptop. This is where Linux comes back into the frame, because Linux is very nice to run on the laptops they have because it's small and fast, and the virtualised images are frequently Linux because of the cost of licenses. The choice is between the cost of Windows licenses or Ubuntu at zero cost. We've known these technologies for years, but what is changing is bandwidth. A virtualised desktop is a desktop that is delivered from the cloud."

Out of the cloud

As Sutor sees it: "There are only so many ways of doing things. We have always talked about the client and server, and all we seem to do through the years - while we improve the connections, the speed between the two, the memory they have, the speed of the processors, or where they may be located - is move things back and forth between the client and the server, where we factor out the workloads. The technologies stay the same. We're just shifting where we do things. And the issues we keep coming back to are economics and security. If its more secure to run things out of a cloud, and more economic to run a Linux desktop from the cloud, it makes sense. I used to be a mathematician so I'm often accused of seeing everything as an optimisation problem, but that is what we are doing."

"We have 30,000 internal Linux desktop users in IBM. Some are dual booting, but they are primarily Linux desktop users, and are fully enabled on our software. They can do all the things they need to do. We have worked towards this for years, with our contributions to things like ODF and Firefox."

"So, for instance, there is nothing in IBM that I use that I can't access with Firefox. I don't know the numbers, but Firefox is getting close to 25 per cent of users, which is a huge number compared to predictions five years ago, and I can see this happening with the Linux desktop over time, little by little - as people see the value, as people stretch out the life of their machines from three to five year cycles, as they see they are usable, that the applications are there, that the data can be shared, and the cost structure is right."

Microsoft owns the desktop because it took advantage of a major shift in technology during the 80s, but the network will redefine the role of the desktop during the next few years, and the versatility of Linux in the server space, on mobile devices, smart phones and netbooks will assure Linux of its place on the desktop.

For other feature articles by Richard Hillesley, please see the archive.

« previous 1 2 3