I was browsing somebody's third hand list of 25 bad things about vista compared to Linux when I ran across this anonymous comment in response to that list:
The thing is, as much as I enjoy Linux, I have to keep windows around to run various programs. I can honestly say I don't want Vista, I've used the beta and it was not that impressive. Microsoft simply has no [**] clue anymore.
The big question isn't why choose Linux, its why doesn't Linux have more developers? And the reason I see is that developers are scared of Linux. Windows is a money making machine for everyone involved. Hardware, software, tech support, you name it, they're making $$$ off of windows. Companies may see supporting Linux as bad for their own interests.
On the personal level I know a CIO who got his company to buy tens of thousands of machines from Dell mainly because he admires both Bill Gates and Michael Dell for having made so much money and, in fact, brags about how much he's made holding shares in their companies.
There's a general problem here - the business press congratulates Microsoft on 99% margins on money printer products (i.e. products with essentially zero variable costs) like Microsoft Office without ever considering that each dollar of Microsoft revenue is an expense dollar for somebody else.
Look at this from an American national economic perspective and you get an interesting question: would American GDP rise if customers got the same computing benefits for fewer dollars? -i.e. if Microsoft's desktop monopoly ended and their American pricing fell?
This question is hard to parse and even harder to answer. If Microsoft were a purely domestic company - with no international earnings or expenditures - the question would come down to whether the re-investment portion of the customer dollar has a greater positive effect on long term American economic productivity if handled by Microsoft than if handled by the customer? In other words, is Microsoft's long term return on R&D greater than the average for other American companies?
Since that's still a no brainer even if you include the net benefit the country gets from offsetting Microsoft's foreign earnings against their foreign expenditures, this question leads directly to a much more difficult one: the national accounts have to balance over long enough periods of time - so what's the American economy getting in exchange for the loss of economic productivity created by Microsoft's monopoly?
I suspect that the answer is in the multiplier effects of the difference between what Microsoft's support troops earn now, and what they would be earning if the PC world didn't exist - a national accounts interpretation of the old joke about the quickest way to exploit a Microsoft product being to grow a small beard, walk with a stoop, and claim you can make it work.
To see what I mean, ask yourself what kind of jobs nine out of ten Wintel support people would hold if their existing IT jobs had never been created or magically disappeared over night? My guess is that the long run adjustment would net out positive for the national economy because many of the people involved, especially some of the really bright ones currently in dead end development roles at Microsoft itself, would create new companies and new products - thereby creating downstream economic wealth for everybody. In the short term, however, an outbreak of mass sanity with respect to desktop computing services would leave a lot of people selling used cars and it's the difference between their earnings now and what they might otherwise be that drives the effects balancing the accounts - remember: marginal earnings have much higher economic multipliers than average earnings.
In other words, I think I could make a fair macro-economic argument that anonymous was quite right in writing: "Windows is a money making machine for everyone involved" - but describing it as really a kind of work for welfare scheme in which everyone wins -except the customer to whom it's a cost, and the national economy for which it's a productivity sink.