I’m excited to follow the above post with some metrics about how much faster the new version of Enketo is in these Feature Releases (FRs)!
Let’s cut to the chase and show two graphs illustrating the performance gains of loading a very complex contact form and then we can follow up with what they mean:
The first chart “Single Form Load Time” shows that CHT 3.15.0 took just over 30 seconds to load the test form. CHT 4.0.1 and 4.5.1 show an improvement of this form load time. However, the last two bars show that upgrade from FR with Enketo Core 5.18 upgraded to 7.2 gives us another huge boost reducing the load time to about 12 seconds.
The second chart “Improvement over 3.15” shows the percentage gain in speed. While maybe a bit contrived, it is still none the less amazing that an upgrade from 3.15 to the FR yields an almost 60% performance gain. Even an upgrade from 4.0.1 to the FR shows an impressive 38% improvement.
Of course, we were really hoping that the FR with Enketo computeAsync be set to true would have yielded even faster results than than the other FR. Given the limited scope of these tests (see below), we’re still hopeful that this change will be beneficial for partners. We’re going to continue to test, but we’d love to hear back from others who have run their own tests!
NOTE: These were run on Chrome on a desktop with “CPU Throttle” set to 6x instead of being run on a physical mobile device. A complex form known to be slow on 3.15 was used as a test by loading it 5 times in each CHT version and taking the average time. While not a guarantee of performance gains, spot checking a physical device showed similar gains.