Modern windowing systems are also very similar. A Windows 3.1 or early MacOS user would have no trouble using a computer today.
Touch and phone interfaces will get there, too, as both software and culture evolve together.
I can't wait for, and would love to support, the Debian of phone architectures. In past HN threads, it has been suggested to me that the real blocker for a good open phone OS is at the hardware/kernel level. That was the state of the hardware environment in the late nineties and early 2000s.
Give it time, effort, and support, and a great and open OS will emerge.
I would like to be that optimistic, but I am not. The PC has always been a fundamentally open platform and hardware has been supported by many OSes for a long time. When I was young, we would run MS-DOS, DR-DOS, OS/2 and Windows 3.x (which was strictly spoken not an OS). After 1994 I also ran Linux and BSD and I never had any real problems hardware-wise. There was a difficult time around the turn of the century with WinModems and printers that only supported Windows. But that was more of a temporary regression.
The biggest blocker in our experimental labs for open-source software was always hardware drivers. These days, linux is extremely common and generally well-supported.
I think this comes down to two things:
If we define "polish" as: consistency+predictable UX+aesthetics
1. openness involves interoperability, and polish is more expensive to achieve when interoperability is a requirement
2. the primary market for polish tends to value exclusivity, which—while not entirely contradictory to openness—tends towards non-openness
The combination of these two means that when you focus on polish you end up with a product line that is likely to be non-open.
The fact that the primary market for polish also tends to be economically comfortable also conveniently leads to margins that allow for privacy, but I think this is just a happy accident.
It's my belief that if we want a private + open platform we need to:
- Accept very large sacrifices* in polish because doing both is only economically viable when catering to a market that doesn't want/need openness
- Choose a different economic model where resources are available to do both simultaneously.
* Also worth noting that sacrificing quality and/or UX can lead to less privacy given that privacy depends on security and security depends on both quality, and also on UX to prevent social attacks.