It must have seemed a sure thing. With costs for Microsoft Office around $75 per seat per year for public administration, the city of Freiburg in Germany embarked on a migration to OpenOffice.org, figuring it'd save a bundle in license fees. With 2,000 users, the move would amass $150,000 annually. So the city tried.
Five years and at least $600,000 on, with unhappy staff complaining of interoperability problems with Microsoft Office documents, city administrators called in a consultant from a Microsoft partner to support the city council in fixing the problem. The solution proposed: a complete reversal of course, switching back to Microsoft Office for a sum of at least $500,000, with a $360-per-seat cost for licensing Microsoft Office and no firm estimates for undoing the earlier migration.
What went wrong? Was it simply that the open source software was incapable of delivering the functions that Freiburg needed? The first hint that the answer to that question would require some digging came when, almost simultaneously, another German city -- Munich -- announced that the success of its open source migration had netted savings around $13 million.
When two almost-simultaneous announcements on the same topic have opposite conclusions, you know complexity awaits, with deeper lessons to be learned. Working out exactly what's been happening has been involved, first because there is an amazing volume of paperwork to digest and because everything beyond the superficial is in German. With help from friends in Germany, I've worked my way through some of the information about both Freiburg and Munich.
Cost cutting, two ways
The contrast between the approach taken in the two cities is striking. Freiburg -- a smaller city administration -- focused on cost cutting. It recognized there would be one-off costs to pay from template and macro migration, as well as user training, but it stuck with Windows desktops, retained certain existing applications, and even allowed some staff to opt out of the migration entirely and keep using Microsoft Office. It seems there was a limited uptake of migration training, too. The result was an environment with both OpenOffice.org 3.2 and Office 2000 in use throughout the attempted migration.
Because Office 2000 did not support the OpenDocument format standard, this guaranteed a flow of documents in the formats used by both office suites, maximizing opportunities for incompatibility. By all accounts, the city stuck with those old versions of both Office and OpenOffice.org and allowed the mixed environment to persist throughout. No two word processors can ever be 100 percent compatible with each other's file formats; only a well-defined, standard format implemented by both stands any chance of interoperability. Unsurprisingly, staff ran into problems with document compatibility; equally unsurprising, the crew blamed the "new" software for the problems.
Looking at the numbers (see my article in ComputerworldUK for more details), it appears that the expenditure in Freiburg was dominated by the idea of cutting licensing costs. I may be missing it in the reports, but I couldn't find any sign of investment in the open source software itself. The report -- and the subsequent PR from Freiburg -- talks about the "uncertainty" of the OpenOffice.org software (forked to create LibreOffice, abandoned by Oracle, then repurposed by IBM and others at Apache) but makes no mention of investment in the software.