and reports the response time back to server where it's logged for reference.
When we compare the time to the duration in charles proxy, there is often a substantial difference. From testing, it seems to be device-dependent, meaning older (slower) devices report a longer response time than newer devices, and both are (usually) longer than the duration reported in charles. Also seems that requests called async with other requests are reporting much longer response times in the application than on the proxy software passing the requests back and forth.
Question: can we accurately measure the response times of our APIs from within our Roku application?