It's very difficult to create a new programming language outside of a hobby context. Businesses, even large tech companies, usually don't see the direct benefits of having a new language. Existing projects won't switch because of inertia.
It does seem like new languages arise in response to large, rapid changes in hardware and software, such that a new language does see noticeable improvements in development speed. I don't know how many businesses care about correctness and safety though. Many non-critical software is fine having occasional or even semi-common glitches and bugs. (Niche languages targeting niche uses will always be around but won't see wide adoption)
Building a completely new language is extremely time and money consuming if you also choose to build your ecosystem and tooling from scratch. On the other end of the spectrum would be small DSLs which integrate into an existing program. Tool creation is still difficult once the DSL becomes significantly complex.
There isn't a universal high level language yet. Different languages can communicate to each other by hooking into native libraries and using ABIs. But this sucks. You have to drop out of Python / Java / whatever and write C code. Oracle is creating GraalVM, a supposed universal virtual machine, but that's been in the works for 10+ years and comes with the requirement to run a Java VM. In Graal's official outlets, their main uses cases seem to be a faster Java or pre-compiling Java applications. The polyglot VM part still seems immature.
Created .
Home > Q Science > QA Mathematics Computer science > Programming Language Development