Companies thinking about migrating from Windows to Linux on the desktop can learn a thing or two from the city...
of Munich, Germany.
The city recently announced that its mass migration from Windows to Linux on the desktop, and accompanying plans to switch from Microsoft Office to the OpenOffice.org personal productivity suite, will take at least six months longer than expected.
The city, which was in the process of migrating 14,000 desktops, cites various reasons for the delay, including a decision to extend the desktop pilot program and internal debate over the legal ramifications of open source. But Union City, Calif.-based Sageza Group analyst Joyce Tompsett Becknell said there's more to be gleaned from this particular story.
"If we step back from the details and look at the bigger picture, this is actually a lesson to all IT managers planning any sort of migration project," Becknell said.
Deliberate testing and extended pilot programs with OpenOffice.org are all in place, she said, and the official migration has been pushed to 2006. However, with the media microscope present over the migration due to its "first in the field" position, such issues could be pushed behind sensational headlines.
"This is one of the first big projects involving a massive migration to Linux on the desktop, [and] it has received an almost undue amount of attention," Becknell said.
Becknell said Linux supporters see the Munich project as a way to boost mainstream acceptance of the operating system, while Microsoft loyalists see it as a chance to shine a light on any failures that occur along the way.
"Any project is almost certainly going to take longer than you first plan for," Becknell said. "You can take that delay upfront, or you can take it on the back end. Pilot projects will become more complex as the number of variables increases. And like all variables, they are interdependent, so testing them in an isolated fashion will only give you limited feedback."
Becknell said another lesson learned in Munich is the importance of finding the right business partners or consultants on such a project. It's important to find partners who truly understand the organization from both technological and cultural vantage points, she said.
Proper planning for such migrations is also key, Becknell said. One positive move that Munich's Microsoft-Office-to-OpenOffice migration team made was setting aside time to make sure the entire system worked efficiently. It's better to spend extra time planning than it is to spend extra time fixing mistakes later, she said.
Munich's approach to planning goes against a trend found in the "Hype Cycle for Linux 2005" report from Stamford, Conn.-based Gartner.
"This open source operating system is making progress in many respects. But it's still two to five years short of the Plateau of Productivity when it comes to the key prize of becoming a vital aspect of data centers," the report said.
But, according to Becknell, time is not something that those deeply involved with Linux should be worried about at the moment.
"If migrations were easy, we'd do them all the time," Becknell said. "IT managers everywhere should be taking notes."
For Przemek Klosowski, founder of the Washington, D.C., Linux Users Group and a researcher at the University of Maryland, the Munich case is interesting not only because of the migration, but because of the discount that Microsoft was offering in the final stages of the negotiation with the city.
"The message: Everyone should have a Linux strategy, even if they really aren't going this way," he said.