Munich and Linux
My "3 TechEds this year" drinking and partying buddy Stephen Forte says:
Ich Bein Ein Aushlander
I love the city of Munich (Sorry Clemens, I know you don’t like that part of Germany). I go there usually twice a year. My good friend Nicole and her awesome hubby Chris live there. Being the History Major, etc I love the history all over Munich, even though it is bad history since the Hofbrahaus was the scene of one of the most important events leading up to Nazism and World War II.
I love the UBhan and SBhan. I love the surfer chicks (I can't resist them!!). I love the proximity to the Alps. Five hours by train to Venice. I love the 1/2 beer 1/2 Lemonade drink in the beer garden. Ok, I will stop now on how cool it is there.
First of all, it's not that I don't like that part of Germany. I just like it much more up here in and around Düsseldorf. We've got awesome microbrews (and we mix with Coke and call it "Krefelder"), the Rhein-Ruhr metropolitan area with some 10 million people gives you always something to do, we really know how to party, we've got plenty of U- and S-Bahn too, it's a 1 hour flight to virtually anywhere in west-central Europe, Amsterdam, Brussels, Luxemburg, Frankfurt are less than 2 hours away to drive and you can get to Hamburg in 4 and Paris in 5 hours. ... Ok I will stop now on cool it is here. ;)
Anyways, Stephen thinks Munich made a bad call by choosing Linux. I think that if they really intend to use VMWare to run their "Windows legacy" they're in multi-trouble.
My understanding is that they felt they were forced to make a choice because NT 4.0 support runs out. So now they're going to run the unsupported NT 4.0 inside VMWare (which does require a license) on top of an IBM/Suse supported Linux? Also, how is it that Linux runs at 30 million Euros (by current estimate) for migration while Windows licensing and support would have been around 27 million before some considerable discounts that MS was willing to give? So here we have IBM and Suse raking in the money for a development and support contract in which existing in-production software is likely going to be entirely rewritten just because it needs to run on a different OS because of "strategy" (which, to me, is complete idiocy) and for support contracts that seem to be, looking at the big numbers, way above and beyond what Microsoft is asking (because IBM and Suse outsource quite a bit of the software development to the community "for free" and look better on the cost side for "licenses").
My experience tells me that that sort of "we force an OS down your throat" strategy fails in any larger company after only a short while and it will fail in government as well. If a departments finds a good software solution for their needs that fulfills most of their business requirements for a fair price, the wallet always wins, and should win, over the geeks in the end. At the end of the day, an OS is what it is: just an operating system. If you have a solution that does what you need for you business you just shouldn't care too much about the OS. If a department uses a "non-strategic" platform and your data-center refuses to give them support, the IT people not doing their job. The business folks make the money for IT to be get their salaries paid. What "strategy" results in way too often is a more horrible and less coordinated zoo of heterogeneous systems than if there were no set strategy, at all. I know of (multiple) banks where mission critical servers run under someone's office desk, because they happen to run the wrong database package or a non-strategic OS and therefore the data-center rejects taking them. Maximum stupidity with a great excuse ("strategy!").
On the other side of the fence, if your role is building solutions you should look for the OS and development platform that gives you maximum productivity for writing apps in order to deliver in time and in budget. Being religious about these things is stupid. I know Windows much better than Linux and therefore I am by several magnitudes quicker writing apps on Windows. "You gravitate to what you know" and there's nothing wrong about that.
To me, the Munich & Linux story this isn't a win for Linux in the first place. To me that's a clever IBM coup around "strategy" resulting in way too much German taxpayer money (and 30 million won't do it) being thrown at the biggest computer company (US-based) in the world instead of the biggest software company (US-based) in the world. How exactly does that amount to a great win in terms of German taxpayer ROI or freedom?
Also, as far as "taxpayer ROI" goes, it should be considered that most German software companies with a working business model (means: make money by having actual customers) aren't into open source. And my best guess is that there is no "yet" there.