When shipping web applications, developers usually worry about backend performance and scaling. With tools like YUI and Page speed, front end performance engineering was brought into spotlight.
However, event today, developers have to look at render/paint timelines and the confusing about:tracing tabs to see if their sites are smooth and jank-free. Developers look at these graphs and try to identify known patterns that slow down a site.
Can this all be automated? Even better, can it be made a part of a continuous integration system so that a developer just pushes code, and knows if something is wrong?
During this talk, I would like to discuss my experiments with porting the Chromium telemetry smoothness and loading benchmarks into Node, how the data can be collected and how it can be integrated into a continuous system. I would be exploring the new set of tools that developers can create to ensure that no website is janky again.