I agree with this statement, but I would add a fourth: 4) What are the potential rewards? Assuming a magical node.js script existed, the backend resources nee

I agree with this statement, but I would add a fourth: 4) What are the potential rewards? Assuming a magical node.js script existed, the backend resources needed are immense, with no up-front profit. The http server mentioned needs to be hosted somewhere, and that somewhere has to be known by the user so he can tell his friends to go to some.domain.com:1773/instance. Also, the overhead of the latency would negate the usefulness of running any code on the client. In those 30 ms that it takes to tell the client a request has come in, the server could already have responded. This hammers the NIC for little gain. Sure, it's a cool idea, but it's not going to net any real profit (unless you can convince some VCs to ignore logic and give $$ based on "cool"). I'd rather see a simple node instance hoster, where you can upload code and have it run publicly: https://github.com/joyent/node/wiki/Node-Hosting https://github.com/joyent/node/wiki/Node-Hosting If I wanted the code to run on the client, I'd also like to see a kind of folding@home that distributes data sets to clients. That idea has real potential for profit, and doesn't require nearly the resource investment of node.js running in a browser. However, for a "cool" side-project with no intent for profit, I agree, improving 1) will drastically increase hobbyist interest. Hobbyists have their own ideas as well.