They didn't have the information to calculate another next page url. So now `nextPageUrl` contains a full link with all video ids, and `getPage` takes the first part of the url (containing 15 streams) and produces another `nextPageUrl` with the remaining streams. Also add a test for this.