Yes, we need more. There are many ideas to explore still.

The "do we really need more languages?" argument predates most of my preferred languages(and that probably includes C and Pascal).

I like Haxe, Rust is looking like being really good also. I don't like Go but others do, and that is a really important point. You don't have to like, or use the new languages. As long as someone finds them useful they have a place.

I'm a bit leery of that argument, because in my experience the most important component of a language is its ecosystem. By far.

I've played with many really nice languages, but never used them seriously because they had no libraries (let alone quality libraries), had poor tooling, poor documentation, buggy runtime, and almost nobody to compare notes with. In my experience, the largest jump in productivity is libraries, not some feature of a language.

On the one hand, languages are useful to explore a space. On the other, there's a lot of reinvented wheels going on here. If all languages stuck with the same ABI (e.g. C), it'd be one thing, because then you could use the same libraries and many tools between languages. Unfortunately, they often cannot, and thus you can't. There's some serious opportunity cost going on here.

I'm happy using node.js at the moment (ooh, shiny!), but the current situation with a plethora of languages isn't exactly great either.

It is true that most of the best supported languages have been around for a good ten years. But each of those languages had a time when they were only a year old.

I don't deny languages take time to mature. I feel like Haxe is about two-thirds of the way towards being properly Mature. It is approaching 10 years old, yet I feel like it is only just getting started, A lot of people still haven't even heard of it.

Haxe is great but confusing for newbies. You can write anything to run on anything but with what libraries; tons of halfbaked thing around. The tutorials are per target which make it seem you can use the same language but have to write different code per platform. I like Haxe and tried it a number of times, but unless coming from Flash/AS it feels underdocumented and, in a lot places in the main site, quite dead. It's not but broken links, vital libraries which are unsupported etc.

Of course we need more languages. We need to continue evolving languages until we get better and better ones. When languages stop improving so significantly from generation to generation then we can start settling down into a smaller number of languages.

We still haven't figured out the best tradeoffs between dynamism and type safety. And we definitely haven't sorted out the best set of features for memory management and error handling. And after that there are a lot of even bigger fish to fry.

We do, because if you reduce languages to their fundamental core, we had around... 3 - 6 languages in wide use.

Note that this means we need a language which has certain fundamental differences to existing languages, so we don't need YAJAWiSS (yet another java with syntactic sugar), but something that takes a step forward. Their error model, first class immutability and RAII outside of C++ might be something like this, if done well.

There aren't that many languages backed by a giant corporation (=> tools and docs). And no, we just need C# (multi-paradigm, multi-typed) with better performance - which is what they are doing.

Well really we just need C, now get off my lawn.

There aren't that many languages in the systems space.