I\'m working on a web site that is using the Google JavaScript Client Library to load some APIs that are exposed via Google Cloud Endpoints. The endpoints were developed in Pyth
It must be an IE10 JavaScript problem. Agreed the null must be coming from document.getElementById('root')
as you pointed out. Your page would not normally contain such an element, and the name is not distinctive. I suspect that the 'root' element is meant to be created by another piece of JavaScript hailing from Google, and it is that other JavaScript code that has (silently) failed. Harness your considerable Debug-Fu to look in that direction. Sorry, I would like to help more, but that would breach Microsoft's Terms and Conditions (the ones about if I don't pay them for each instance I am forbidden to run their software).
EDIT: Since proxy.html contains both the offending JavaScript code and the 'root' element, Google could correct it by changing the order of the elements. Sensibly, many authors recommend locating the JavaScript just before the bottom of the html body. If you're in a position to reference your own copy of proxy.html, you could try the fix yourself before nagging Google about it.