Munich has a pretty good write-up on the whole project. I find only a few minor errors:
- While they claim success and detail the evolution of their detailed grand plan, they had to revise it more than once as they went along. That’s a failure, I think, of this set-piece kind of migration. They even write, “In most cases, projects like these cannot be planned in advance. The first steps have to be to count, identify and structure the existing IT landscapes. One of the most important benefits for the city of Munich was the reorganisation of the IT structure. For the first time, the admins would know exactly by whom, where, when and why a programme was running â€“ not as a matter of control, but for organisational and quality assurance purposes.” My plan to migrate over the weekend and fix the breakage on Monday looks pretty good compared to fixing the whole project multiple times… Still they saved a lot of money but they could have saved more doing it my way.
- Further, they switched from Debian to Ubuntu GNU/Linux, “the city decided to switch to Ubuntu Linux, a Debian derivative with scheduled, plannable updates. Canonical, the company behind Ubuntu, promises to deliver a new version twice a year (April and October), and offers long-term support for special versions. Canonical promised five years of support for Ubuntu Linux 10.04, the version chosen by LiMux.
Furthermore, Ubuntu comes with a more up-to-date software selection than the rather conservative Debian distribution.” That’s not even consistent with their plan to avoid breakage. Ubuntu/Canonical breaks a lot of things. On the one hand they choose a set-piece roll-out and on the other choose a distro with a cadence of breakage. That’s schizophrenic if you ask me.
Still, it all worked out in the end, “Munich’s switch to open source took a little more than ten years, and costed the city 18.7 million euro, the case study reports. The system is some 12 million euro cheaper than continuing with the proprietary alternative, although the case study stresses that costs were never the driving factor. The city now runs OpenOffice and LibreOffice on 15,000 desktop PCs and has introduced centralised IT services including project management, testing and deployment.
In the report, the IT department lists the results, including an extraordinarily high level of independence from vendors, independence in its operating system, the overall use of open standards as normal practice, a very high level of IT security and no shutdowns of any services or processes over a migration period of several years.”
They got what they wanted their way. That’s what matters. The most important fact of LiMux is that Munich figured out that it was possible to escape the slavery of M$. They were far more successful than they first imagined too. They found it cost less to do the job and less to train users. It certainly is a better system than they would have had fighting waves of malware and forced upgrades by M$.