Ask HN: Do you think we need to keep evolving programming languages?
Modern languages are pretty great but I think they've more or less converged on the right way to do things.
Maybe we should accept victory and move on to less explored areas. Is C++ better or worse for the changes in C++11/14? Better, by far. Is Haskell better with lens? Yes. Is Go an improvement over writing concurrency in C++ or Java? Yes. Are there no more improvements to make? I doubt it. (What are they? I don't know. I didn't know what would make C++ better, and yet I see the improvement in C++11.) I find it hard to believe that just because you and I can't see how to make things better, then nobody else can either. I think, by the way, that the improvements aren't so much in what they let you write (though Go was), but more in what they make it so you don't have to write. Which modern languages are you thinking about? There's still a long list of stuff in research languages that are not common in industry but are slowly making their way in (think how long algebraic data types and first class functions took to become mainstream, and lots of concurrency stuff that's still being rediscovered), and there's new areas being explored (dependent types, how to do monad transformers gracefully). I would be very sad if we stopped with what we have now. Maybe I am wrong but there is nothing that can be done today that could not be done in a 30 year old language. I am certainly glad IDEs have evolved though Anything could have been done in 30-year-old languages. Take, for instance, pre-STL C++. (The STL came out in 1993-4). You can do anything in it. But using the STL is a whole lot nicer than having to write all that by hand, and has fewer bugs to boot. Then you've got C+11/14, which is much improved. So, your point is correct, and yet there is real improvement.