Download schema list from the repository

[tjaffri]

The yeoman starter currently requires the translators repo to exist at a specific relative path to discover the schemas list. Instead, we can download the schemas list dynamically from github so the starter can run anywhere on the file system.

We also need to update the README with correct instructions.

Bonus: create a node module in the cloud project that's a simple opent2t client (and supports downloading/querying/caching schemas). Use that in this project, as well as in various sample apps that currently duplicate code to clone the repo.