There's a lot of stupidity here (or perhaps one might charitably call it excessive glee, or even plain old clickbait), but it's been a long time since 95% of computers ran Windows, if that was ever true in the first place.
You'd have to start by defining "computer". If we use a technically reasonable but broad in journalism's terms definition of "device containing a CPU with integrated MMU", then we're including every mainframe, server, desktop, laptop, smartphone, tablet, industrial control system, laser printer or multi-function document management system, router, network or storage switch, storage system controller, sound mixer, and countless other devices. A very small and diminishing fraction of those machines run Windows.
Even if you limited the definition to include only things we traditionally think of as computers, namely mainframes, servers, desktops, and laptops, Microsoft's market share is something like 30%, though it's very difficult to measure and once again subject to definition; see https://en.wikipedia.org/wiki/Usage_share_of_operating_syste.... And again, that's been declining for a long time. Even the most generous interpretations that include only desktops and laptops don't get you to 95%; you can narrow-definition your way up to about 90%, and that assumes "market share" includes copies that were made in violation of license and/or law (and thus for which Microsoft received no money).
The reality is that Microsoft is a niche player, not because they lost a market they were previously winning but because that market itself became a small part of something much bigger, in which Microsoft was never especially competitive. As noted by the author, they've been investing heavily in changing that, but the early returns aren't much to talk about. Windows Mobile/Phone is a complete dud, Surface is a weak seller despite recent improvements, and Azure is at best a very distant second in IAAS (see http://www.datacenterknowledge.com/archives/2015/05/28/gartn...). What's remarkable is not that Microsoft is doing so poorly but that they're the only enterprise computing company that even appears to be trying. I think that's what the author was expressing. The characterization as "fucked" comes from Ashlee Vance and should be given the credibility one would assume considering the source. There's little question that most of these companies' positions of relative importance are in decline and have been for a long time. But some, especially Microsoft, still have enormous sums of cash and businesses that are continuing to generate more of it. However irrelevant such a company is, it's still a long way from going out of business. Conversely, no amount of cash is a substitute for organic sales growth.
I could have saved you some effort. By "computer" I meant "personal computer". I presume that everybody here would understand that I didn't mean "device which can compute", i.e. all servers, cash registers, calculators, difference engines and Turing machines. I'd have thought that this would've been obvious since I said 95% off the top of my head, and your very own link says 91.47%.
Microsoft are so far from niche it's unreal, and it shows a great ignorance on behalf of the author to look at one product from one arm of a diverse organisation, which has been at the forefront of CS research for decades, has competences that Google et al absolutely do not, and compare it to other companies which only have vague product-line overlap so blithely.
This is ignoring the manufacturing, supply chain management, distribution, retail competences that the other mentioned companies have, which Google and Facebook may never have. It's also ignoring fundamental aspects of marketing such as brand equity, positioning, etc. It's also ignoring the fact that some of these companies are already embedded in industry, via their product line and also via consultancy. Lastly, it's taking a single current snapshot of the industry, and is ignoring the fundamentals of organisational strategy and competition and capitalism, which would suggest that the companies mentioned will adapt and compete to survive (in b4 Innovator's Dilemma). Not all companies will do this successfully, but the companies mentioned will.
It seems like the author is judging these companies via their perception of a small subset of their products, and the market's perception of them (some of their share prices are up and down) without looking at the fundamentals of each business, their diversification strategies, product line strategies, their financial metrics (which the author could've looked up in 10 seconds), and so on.
I mean to suggest that these companies are in trouble because of OSS such as MongoDB, which they have partnered with, support, and sell (and more than likely funded at some point), is very silly, but that's precisely what the author has done.
It's not 1992 any more. At the very minimum, it's reasonable to assume that "computer" also includes mobile devices and servers, two additional markets that are well-known to both technical and lay audiences.
> I presume that everybody here would understand that I didn't mean "device which can compute", i.e. all servers, cash registers, calculators, difference engines and Turing machines.
You haven't shown that you know what an MMU is. There's a reason I included that; the definition excludes microcontrollers, devices that lack external memory, and other things that clearly cannot run Windows or any other competing operating system. Most of the devices you mention either don't have CPUs or the CPUs they have don't have MMUs, and they're not part of the Windows addressable market. Since you brought up Windows in the first place, doesn't it make sense to limit ourselves to the general classes of computer on which Windows could conceivably run?
> This is ignoring the manufacturing, supply chain management, distribution, retail competences that the other mentioned companies have, which Google and Facebook may never have.
I didn't mention either company. Both are advertising companies; Facebook doesn't make any kind of product that competes with Windows, and Google's competitor (Android) isn't really its own anyway, as the actual OS is Linux.
> Not all companies will do this successfully, but the companies mentioned will.
I'm not sure why you're going off on me. I explicitly said that Microsoft is likely to remain in business and making money for a very long time. In case it isn't obvious, making money and successful are the same thing. A company can make money serving niche markets, and many do.
> the fundamentals of each business
Microsoft's top product by both revenue and profit is Windows. Windows has a leading market share on only a single class of device (desktops and laptops), and that market itself has been consistently shrinking in each of the last several years (2013: http://b-i.forbesimg.com/anthonykosner/files/2013/04/ChartOf... 2014: http://www.engadget.com/2014/10/08/pc-market-share-q3-2014/ 2015: http://www.gartner.com/newsroom/id/3090817). Forecasts are for continued shrinkage of that market. I don't think it's fair or prudent to ignore those fundamentals, as they clearly are central to Microsoft's ability to make money. Azure may be doing well and may even grow into a strong #2 in its space, but it's still less than 10% the size of the market leader and it generated well under a billion dollars of revenue last year (http://www.forbes.com/sites/louiscolumbus/2015/04/15/sizing-...), accounting for less than 1% (http://www.google.com/finance?q=NASDAQ%3AMSFT&fstype=ii&ei=R...) of the company's total turnover. It's difficult to know if it was profitable, but it hardly matters; the operation is barely even material to the company as a whole. Azure would need to double in size 4 more times to account for 10% of Microsoft's revenue.
Again: I didn't say Microsoft is going to go out of business tomorrow. Or next year or even in 2025. But you should acknowledge that a backward-looking view of the company is risky, and that an objective forward-looking assessment of its business is less rosy than you would make it out to be. At best, the company will be much less dominant than it was in the 90s.
> I mean to suggest that these companies are in trouble because of OSS such as MongoDB
I didn't. As I said, there's a lot of stupidity here. There probably are some small e-commerce and ad companies using Mongo that should be using Postgres or MySQL or Oracle or Microsoft SQL Server, but I don't know that those companies would have gone into business at all if they had to pay an enterprise software vendor for database licenses. I agree that the impact of MongoDB on the named companies' bottom lines has been negligible, and that no evidence of such impact has been offered.
The author and many of the comments here have strayed well into hyperbole, probably fueled in part by ignorance and in part by schadenfreude and glee at seeing some poorly-managed companies in mature markets struggling to remain successful (HP is, after all, the poster child for mismanagement in this space). I don't believe I've made the same error. If you think I have, please provide supporting evidence, as I have provided for my position.
You'd have to start by defining "computer". If we use a technically reasonable but broad in journalism's terms definition of "device containing a CPU with integrated MMU", then we're including every mainframe, server, desktop, laptop, smartphone, tablet, industrial control system, laser printer or multi-function document management system, router, network or storage switch, storage system controller, sound mixer, and countless other devices. A very small and diminishing fraction of those machines run Windows.
Even if you limited the definition to include only things we traditionally think of as computers, namely mainframes, servers, desktops, and laptops, Microsoft's market share is something like 30%, though it's very difficult to measure and once again subject to definition; see https://en.wikipedia.org/wiki/Usage_share_of_operating_syste.... And again, that's been declining for a long time. Even the most generous interpretations that include only desktops and laptops don't get you to 95%; you can narrow-definition your way up to about 90%, and that assumes "market share" includes copies that were made in violation of license and/or law (and thus for which Microsoft received no money).
The reality is that Microsoft is a niche player, not because they lost a market they were previously winning but because that market itself became a small part of something much bigger, in which Microsoft was never especially competitive. As noted by the author, they've been investing heavily in changing that, but the early returns aren't much to talk about. Windows Mobile/Phone is a complete dud, Surface is a weak seller despite recent improvements, and Azure is at best a very distant second in IAAS (see http://www.datacenterknowledge.com/archives/2015/05/28/gartn...). What's remarkable is not that Microsoft is doing so poorly but that they're the only enterprise computing company that even appears to be trying. I think that's what the author was expressing. The characterization as "fucked" comes from Ashlee Vance and should be given the credibility one would assume considering the source. There's little question that most of these companies' positions of relative importance are in decline and have been for a long time. But some, especially Microsoft, still have enormous sums of cash and businesses that are continuing to generate more of it. However irrelevant such a company is, it's still a long way from going out of business. Conversely, no amount of cash is a substitute for organic sales growth.