The common thread in what you're saying is "why do people not choose (technically pleasing solution) over (familiar solution with known warts)", and the answer to that is that change is really expensive and unpopular. Especially if it forces people to abandon their existing custom tooling, workflow, infrastructure, and skills.
Would most people use a new operating system with no web browser? Probably not. So that sets a minimum bar on the quantity of software you have to port to get it adopted. Which is a vast amount, especially if you want to write it in a new language as well. And new languages have new pitfalls.
For the shell, it's not clear that the same language is necessarily ideal: shell commands are ephemeral and the user wants to type as little as possible, while code lives forever and is generally more complex. Strong type systems and other forms of testability are important for code but clutter for the command line.
Linux chose Minix to replicate partially because he'd been taught it at university and partially because he could leverage the GNU infrastructure.
The common thread in what you're saying is "why do people not choose (technically pleasing solution) over (familiar solution with known warts)", and the answer to that is that change is really expensive and unpopular. Especially if it forces people to abandon their existing custom tooling, workflow, infrastructure, and skills.
Would most people use a new operating system with no web browser? Probably not. So that sets a minimum bar on the quantity of software you have to port to get it adopted. Which is a vast amount, especially if you want to write it in a new language as well. And new languages have new pitfalls.
For the shell, it's not clear that the same language is necessarily ideal: shell commands are ephemeral and the user wants to type as little as possible, while code lives forever and is generally more complex. Strong type systems and other forms of testability are important for code but clutter for the command line.
Linux chose Minix to replicate partially because he'd been taught it at university and partially because he could leverage the GNU infrastructure.