Ask Paul: April 7 (Premium)

Happy Friday! We're finally settling down after the move and are looking forward to a normal weekend. Which we will kick off early with some great reader questions...
From Windows to Linux
Eric_Rasmussen asks:

With Windows losing focus these days, I've been thinking about how Microsoft could get away from the OS game. They've got a ton of Linux kernel experience now, I wonder if they would ever build the Windows shell on top of a Linux kernel? I actually really like the shell of Windows 11, it's the filesystem hiccups, the way drivers are handled, and how updates are delivered that drives me mad. Microsoft Linux 23.10 / 24.04? It'll probably never happen but it's fun to think about.

This comes up from time to time for various reasons, and I'm reminded that the first scoop I got after moving over to Windows NT Magazine was that Microsoft in 2000 commissioned a company called Mainsoft to port Office to Linux just in case. (I happened to be in Israel that year for work and the CEO of the company I was at allowed me to interview two programmers from Mainsoft who were visiting that day. Weird coincidence, and in the end, nothing came of it, of course.)

Anyway. The issues with Microsoft replacing Windows are many, but consider that it's failed hard with versions of Windows that aren't full Windows (Windows RT, Windows 10 S/S mode, Windows 10X, Windows on Arm, etc.) and with mobile systems including Windows Phone/Windows Mobile and even the Android version it uses on Surface Duo. So it's not just like creating the thing is enough. If things that look and mostly work like Windows haven't been successful, it's not clear why a version of Linux, which would presumably look and work something like Windows, would be either.

But I am with you: I am fascinated by this possibility and would love to see Microsoft do what it did with .NET, which was to start open-sourcing key technologies and then introduce a plan by which it will completely open-source the platform over time. And it should do this for the same reasons it did so with .NET: the community could contribute in meaningful ways and, over time, could become the key driver for improving and advancing the platform. Today, we get what I think of as "feedback theater," where there is an Insider Program whose contributions are either imaginary or minimal at best, and it's not clear what party is driving Windows forward and why.

Put another way, a "Windows" that is built on Linux would be like what Apple did with Mac OS X, which was to build a "Mac OS" that was built on the solid underpinnings of Unix. It would push the responsibility and expense of maintaining is core foundational technologies into the open source community, though it could still maintain a leadership presence for as long as that was needed. It would be a win-win for both Microsoft and its users. And Microsoft would still charge for support: there is no need for this change to erase its corporate licensing revenues.

...