A programming language is a tool. And like all tools, the situation is very different depending on if you are writing a personal tool that solves your problems

A programming language is a tool. And like all tools, the situation is very different depending on if you are writing a personal tool that solves your problems or if you are trying to solve other's problems. I used to lurk in a programming language design group. I joined it because I wanted to "solve everyone's problem" with a new language because I had some success with solving my own problems with a previous language (which is not for everyone; it's a Forth dialect). For few year I saw some people theorizing and analyzing with little practical results. It's fascinating how people can know about "analysis paralysis" and still fall for it. Meanwhile, I lost interest in my own attempt and returned to my Forth dialect. Today it powers a few programs that run 24/7. And it is also my goto scripting language. It's not available on Github/Gitlab/whateve because I'm too lazy to do something that will certainly have zero benefits for me ( however curiously, it is half-documented because writing docs helps with spotting silly decisions sometimes). It is true that you are never done with that sort of program, there's always something to improve or fine tune. Also, since you are the sole user you cannot rely on a ecosystem and just "import their.stuff". But I think the great mistake is really to start making a tool, a programming language, with no plans on what to use it for; or worse, expect others to use it. The Internet is indeed littered with born-dead or short-lived languages. They are solutions without a problem to solve.