Hi, the author of original submission here. I like your idea, especially the number of servers you used. I'm only concerned that using a full HTTP GET to measu

Hi, the author of original submission here. I like your idea, especially the number of servers you used. I'm only concerned that using a full HTTP GET to measure the latency (function request_image() in src/ping.js) makes the results possibly way more random than measuring the TCP/IP RTT server-side. That is, your method seems to measure the latency in higher net layer than TCP. Anyway, nice idea :)