I laughed at both your initial response and IsaacSchlueter's comment. Valid points from both.

It would have been a little odd for me to say "Nice use case for Python" considering the original post was about Node, and specifically said they tried writing it in Python but "writing asynchronous code in Python is like using your keys to open a beer bottle." That also made me laugh.

In fact, the original article specifically talked about why Node.js was a good choice for this use case. I agree. Not that you couldn't do it in other ways. Erlang would also be a good choice here, I'm sure. There is always more than one way to do most anything. Still, and in spite of the fact that Node might be getting too much hype, it is an excellent tool for this type of async need.

    > Node gets a lot of attention despite
    > not doing much of anything new

Most people close to the project (Ryan, me, Mikeal, Felix, etc.) are usually pretty quick to point out that node does virtually nothing that hasn't been done before. Async IO is not new. Event loops aren't new. IOCP isn't new. JavaScript isn't new.

If anything, node's success so far has been a result of combining established technologies in ways that make it easy for more people to write reasonably fast programs that use the internet in fun and interesting ways.

Novelty isn't really what it's all about. If anything, I look at novelty as a hazard. Most ideas are bad, and it's hard to tell when they're new. If you can take some existing good ideas, and make them more fun, that tends to be a more successful strategy (or at least a more enjoyable one.)