Recently, the Foreign Office of Germany made the announcement (translated news report) that it is migrating away from Linux back to Windows as its desktop solution. We've covered the concept of national adoption of Linux at a government level a couple of times before, but this is one of the first cases of a government moving back the other way. The most important action that community can take at this point is ask, “why?”
 

The German Foreign Office first moved over to Linux as a server platform in 2001. It then began using Linux and open source software in general on the desktop in 2005, and all laptops were moved to a Debian system. Making the regression even more frustrating, reports released in 2007 and 2009 (both documents German) indicated that the adoption of FOSS, despite a few small problems, has been a success.

Making matters worse, the transition back to Windows XP, to be followed by Windows 7, includes dropping OpenOffice, Firefox and Thunderbird in favor of MS Office and Outlook. Yuk.

This might a time to start asking some difficult questions. It was a prestige victory for FOSS and Linux on the desktop when the German Foreign Office first migrated, and getting a straight answer about what went wrong is now vitally important.

Actually getting to the answers isn't easy because the decisions weren't arrived at in a technical forum, but rather, in the murky world of politics. For English speakers such as myself, the fact that the available documentation is in German presents a further barrier. From what I've been able to glean, thanks to Google's translation facility, the problems that have been stated fall into three categories: hardware support, interoperability and training and user adoption.

The hardware support problem is an old chestnut, and I think the only reasonable analysis of the situation must be a pragmatic one. Within an office, a certain percentage of hardware won't work as soon as it is plugged in. A smaller percentage of the total will never work at all. The report lists scanners and printers as an example.

Some Linux advocates will be quick to place the blame squarely on the doorstep of the vendors. However, matters are made worse by the fact that vendors face hostility on the subject of binary drivers. Let's reduce the problem to its essence: expert intervention, discarding unsupported hardware, delays while the problem is solved and making do with curtailed functionality all add the cost of running Linux. That's a fact, regardless of who is to blame for it. These problems need to be identified, addressed and alleviated.

Interoperability touches on a problem that I've mentioned before: it's hard to migrate away from proprietary software in isolation. The best solution is to begin to migrate those other departments. In addition, the 2009 report that the problem had been exacerbated by running of older versions of OpenOffice.

Retraining staff to use unfamiliar software is always going to be a problem. It's difficult to avoid completely, but sometimes you have to spend money to save money. Apparently, the staff themselves were sometimes reluctant to make the switch to open source. Perhaps more could have been done, at the beginning, to evangelize the benefits of open source software to them.

Unfortunately, all of the reports that I have been able to find and translate lacked the precise details or hard figures that proved that Linux had failed. The forums and discussion threads on various sites are bubbling with comments hinting that Microsoft may have stepped in with huge financial incentives to switch. However, there have been no reports of a backlash from the workers themselves now that they are being to being moved back to Windows and other proprietary software, and we need to ask some tough questions about why.