LeslieM wrote:Afternoon forumfolk.
eValid's mobile device imitation feature is working wonderfully for me. Just one quick question.
We have a client that is experiencing exceptionally slow response times on an ipad vs Desktop.
We can replicate it on our side by using a physical ipad device.
However, when using eValid to simulate the ipad browser it doesn't seem like it mirrors the performance of the actual device.
I'm 100% eValid is simulating the ipad's browser since when I access a website that does not support mobile browsers I get the expected message.
Should I get the same behaviour on my desktop with eValid than on the physical device?
Thanks for posting LeslieM. Glad to know you're getting good results using eValid.
There are three components that affect response time on any device that is part of a browser-server (client-server) interaction:
(1) Latency between the device and the server (both ways): the speed (or slowness) of the web at that instant.
(2) The speed/performance of the server stack.
(3) The speed/performance of the device.
Performance and quality problems can arise in ALL THREE of these parts of the system.
eValid is focused is on (1), as affected by (2), and as reflected by (3). The main outcome of eValid-based tests that imitate (3) is to help identify problems in (1) and (2).
The usual case we see is that (3) is a small fraction of the total time that is, that the device's performance is not the question. In other words, onboard resources on the mobile device are usually more than sufficient.
eValid is NOT emulating the device. It IS imitating the device for purposes of measuring (1) and, indirectly (2), using the resources of a PC desktop.
What you seem to have is a slow device. (An iPad V5 is slow? Something is wrong here...these devices are usually VERY quick!) Do you need to update the OS or clear the cache or has it run out of RAM?
We don't think as a software supplier you need to account for what may be wrong with the device that slows down response time; you really only have the server side you can do much about...plus maybe the bandwidth of the gateway from the server stack to the web.
The advantages of doing the experiments/testing/tuning from an eValid base are a single point of testing for any (or a big set of) device, and the accuracy of the timings of (1) and (2).
Performance testing of the device per se is a different technology and requires different tools and techniques.
-- eValid Support
