[vbezhenar]

Does it really matter if NT kernel is good or not? I don't think so. What matters if it's cheaper to use Linux kernel instead of Windows kernel in terms of expenses.

Just an example. Microsoft developed several browser engines: first Trident, then EdgeHTML. Edge wasn't bad, it was on par with Chrome, may be slightly behind, but nothing that couldn't be fixed with time. But Microsoft decided to abandon it and use Blink, because that was cheaper.

I absolutely could see the same scenario in the future, when they would need to cut expenses. Just get Linux, wine, fix enough bugs to make explorer.exe and OneDrive.exe to run smoothly enough and ship it.

[selcuka]

> But Microsoft decided to abandon it and use Blink, because that was cheaper.

True, but you don't generally need backward compatibility with HTML rendering engines. You definitely do with OS kernels.

[topranks]

Just like with browser rendering engines it’s much better if there is variety out there in the ecosystem.

That’s not to say you’re wrong.

[SllX]

There is a world where Microsoft could make yet another soulless decision to kill something good they already have on the theory it would save some money, on the mistaken belief that the kernel is the problem with Windows.

One would like to believe that as far gone as they are on Windows, they’re not that far gone. But hey, anything could happen.

[tsss]

They don't care if Windows is good or not because they don't make any money on it. Today, the only purpose of Windows OS is to advertise (or force) other more profitable Microsoft products as cheaply as possible.

[SllX]

Windows & Devices is still a $14ish Billion/year or so business for Microsoft. I know that doesn’t seem like a lot, but that’s a lot of money they’re apparently not making which would make their most recent 10K filing with the SEC a lie. It’s apparently enough money they’re not making that they keep screwing with Windows in unproductive ways that make their Windows customers resent them every time they release an update.