One of my happiest jobs was as a Technical Editor for a start-up magazine that focused on the 32-bit personal workstation. This was back when the 80386 was a brand, spanking new chip. I had to run performance tests on three operating systems to appeal to the kinds of techies who would know what a personal workstation was.
When the magazine launched, we tested MS-DOS with DOS Extenders (to allow applications to use the RAM above 640K), SCO Xenix, and OS/2. SCO came out with UNIX shortly thereafter and we made the switch.
For those of you too young to remember (but techie enough to care), Xenix and OS/2 were 16-bit operating systems. UNIX was 32-bit. DOS Extenders worked with true 32-bit applications.
Microsoft had the temerity to write us a letter stating it was unfair to test OS/2 against Unix because the technology didn't match. We published the letter, along with the response that we were simply testing the best the user could buy, not trying to keep the playing field level. When discussing the letter our Editor-in-Chief said, "I bet OS/2 is simply smashing against CP/M."
Anyway, we had the Microsoft compiler folks talk with us, and the Microsoft OS folks talk with us. These were separate meetings. We asked each group when they thought they would support 32-bit systems. They both gave different answers. The OS group thought they would get there first.
That's when I stopped worrying that Microsoft would take over the world.
I seem to recall that way back in the day.
Of course, real innovation happened back then.
Now you have patent litigation and “I invented round edges and a screeen” claims in court.
I still have a Z80-based CP/M system in my garage.
It even booted up, last time I tried it.
Lots of fun ideas for it if time were no object; getting it online, for one. Maybe cobble together a primitive text-only browser...