Re: Time for a new OS - From Scratch
so right. Note the history of this thought. CPM/DOS/Win9[5,8] to various PDP11 OS, Unix, Windows NT and descendants, Plan9, Oberon, QNX. None of provable correct OS grew out of niches. I wonder if the huge efforts in memory safe languages and the disaster of Itanic compiler redesign could have been better spent in developing a cross compiler that could generate applications onto small fast OS. I have a preference fpr QNX, but maybe Oberon might be better. Linices are not the answer as it is now so bloated its getting hard to tell difference except one has a choice of desktops.
The obvious objection is the effort to create provably correct libraries for OS primitives. Who wants to create a screen handler for each applications, device drivers, et al. This also assumes hardware is similar enough for some compatibility.
Given the $Gov and corporate demands for inbuilt snooping, such a project[s] would have to be written and run entirely by a well funded dedicated nonprofit based in a country that sneers at lawfare demands from the big players and totalitarian states. Any takers ?