Q. How available will my source files be for development and for runtime?
A. normalize works like a browser. This value proposition should be the loudest on the site. Since people already understand how this works, it's a good mental gateway. The next thing people will ask is: "where are my files?". Since this is a radical departure from the existing workflow of downloading shit-tons of files and having even more copies of them everywhere, an explaination of how files get onto your disk and where they are kept would be great.
Q. What is the proxy and how does the proxy work?
People will be apprehensive about a new serveice. npm's architecture is mysterious and seems unstabe. The more transparent a service is, the more likely people are to trust and use it. So it would be great to break down the proxy function by function and explain how it works at the code level and what the ideal devops configurations are for each platform.
Another really valuable thing is how someone can set up their own normalize proxy in say, 2 easy steps (install could even setup and run platform specific process monitoring). Maybe something like...
wget https://path/to/code./install.sh && ./run
Q. URLs in my require statements make my code ugly
A. Could https:// be assumed as the protocol to get something more like var url = require('nlz.io/npm/-/url/*/index.js')?
require('node-normalize')
var url = require('nlz.io/npm/-/url/*/index.js')the node-normalize module would need to recursively pre-process modules. check the cache, if items are not there fetch them, but this needs to be a blocking operation so i used execSync (here is a proof of concept)