You can either be someone who knew various programming languages and wrote a few small single-purposed libraries (and jump the wagon to another new languages) or knew two or three programming languages and build something big.
It's a choice. Ask yourself if you are truly interested at these languages or is it because of the HN side-effect (hang out with a crowd with certain belief, you'll become them or want to become them).
And you're absolutely right about making decisions, they have to be made. Anyway, I don't want to steal the topic, but thanks for the handy reminder. :)
These days, learning a new programming language every year is the fashionable thing to do. This trend, IMO, is stupid. I'd say build something new every year with what you already know. If what you know isn't enough to build that something, then go ahead and learn a new technology. (For example, I mostly program Cocoa but now I want to build a webapp. I can't write a webapp in Cocoa, which is why I'm learning Django.)
Pick one for the specific domain if your current one can't handle it.
That's how you know you're learning - your brain rebuilds itself around new concepts. If you push it hard enough you will start hearing a hum in your ears after few hours of continuous learning.