This guy is in for a disappointing future since he seems to be unaware that Windows is more than than the consumer editions. Revenue from Windows Enterprise (which has management tools like Active Directory and backwards compatibility with non-game apps needed for large corporate deployments) and Windows Server (needed for Active Directory, Exchange, SQL Server, etc.) is still in the billions and there's nothing on the horizon in the Linux ecosystem to replace those. Given that Microsoft is going to have to continue to develop Windows anyway, there's not much reason for them to throw in the towel on the consumer desktop.
None of the software and services you mentioned require Windows though - they could be made to run on Linux, and some already do.
As more and more revenue shifts from desktop/servers to cloud and services, it doesn't seem too far-fetched for Microsoft to decide maintaining the entire OS stack themselves makes less and less sense. A Microsoft Windows linux distro would free up resources to focus on what makes Windows unique.
Exchange is very dependent on Win32 and .NET Framework. As is ODSP and Dynamics.
Azure under the hood is Hyper-V with most services built upon that dependency.
Yes, millions of man hours, monkies, and typewriters you could transform this to Linux. The economics aren't there when Azure/M365 keeps pulling in money running on it's current platform hand over fist.
Over 60% of customer workloads running on Azure are Linux. And that statistic is skewed by those using things like AzureAD (basically workgroups).
At some point it will become a burden to develop new technologies on Windows instead of Linux. If that hasn't already happened.
Desktop already is a dwindling revenue stream for Microsoft. Microsoft is already pushing for companies, from small garage startups to mega enterprises, to migrate to online services where the underlying OS doesn't matter.
Windows has inertia, a lot of it. But all things in motion eventually come to rest.
Well, for one, you may have noticed that MS put in the work, over years, to make .NET the same implementation across all three platforms. So that’s at least one pillar of impossibility removed.
Except ODSP/EXO/Dynamics don't use CoreCLR -- they use good ole fashioned .NET Framework. Those products with their 20+ year old code base would require a full rewrite.
His take is that for consumers Microsoft will abandon windows in favor of Linux. He predicts Linux will get better. And windows will get worse in development and support. And so Microsoft will give up on windows. And that worsening trend plus abandonment plus Linux improvements will cause Microsoft supporting Linux.
The probability of each event happening is high enough. But the probability of all three happening at once is low. And that is why this prediction is difficult to believe.
I think it is true windows dev is much more difficult now. The platform has an identity split. It used to favor power users. Now it favors the rich mac users. And upcoming kids who are attached to iPhones. And this means… it gets worse … Or it changes audience. The latter will be hard to pull off.
I think Microsoft also has less need for windows. We know this because its core business has been shifting. They are platform agnostic now.
So what becomes the incentive for Microsoft to continue windows?
Because they're a corporation that makes money. They have incentives to employ people, and the vendor lock-in with Windows is far too large to change anything at the moment or anytime in the foreseeable future. Changing Windows to become a Linux-based distro would be a massive corporate undertaking; Microsoft isn't in the business of pleasing tech-minded people. They're a business that makes money.
Linux isn't a corporation; it's really more of an idea. They don't have marketing departments or people trying to sell you licenses. They don't have vendor lock-in or active-directory or a cloud based infastructure. They don't have an entire advertising division or a search engine. There aren't any shareholders to please or paid employees to keep on payroll for government kickbacks. They're not targeting the casual, media-focused, average computer user like Microsoft, which makes a lot of money by doing so.
In my last job, I worked in a mid-sized suburban office. There weren't any "Linux reps" knocking on our door, making sure we were getting the most out of Ubuntu.
Hardly, SteamDeck will get those Windows games, that game developers write using Visual Studio on Windows, because they push Proton instead of making the studios interested into native ports.
> Given that Microsoft is going to have to continue to develop Windows anyway, there's not much reason for them to throw in the towel on the consumer desktop.
That would make sense … if Microsoft didn’t have the second most bonkers track record in history (after Google) in the domain of “fragmenting and releasing competing reimplementations of products already in your core portfolio”!
