chore: add PR performance benchmark workflow #11729
Conversation
|
pixi.js-base • pixi.js-bunny-mark commit: |
Implements an automated workflow to run performance benchmarks on pull requests. Compares the PR's performance against the main development branch baseline.
PixiJS Benchmark Results (dev CDN vs local dist)
❌ Performance regression detected (> 5% slower than dev) |
|
What is the reason the the performance benchmark in another repo? Why not add it here? |
I was mostly thinking to avoid bloating the main repo as we add more benchmarks |
|
The challenge of keeping it external is that it creates an upstream dependency on our repo. Say for instance, that we change an interface or some API it would require us to update the benchmark or we'd be blocked. Because of that coupling, I would say that it makes more sense to keep it in this repo. |
Introduces two new benchmarks to measure the performance of rendering and animating specific display objects. One focuses on Graphics objects, animating many circles with redrawing operations. The other targets Text objects, animating position and dynamically updating font sizes to simulate heavier CPU load. These additions provide dedicated tests for assessing rendering efficiency and identifying performance changes in these key PixiJS components.
|
going to convert this one back to a draft as the consistency disappears with the more tests that are run |
Implements an automated workflow to run performance benchmarks on pull requests.
Compares the PR's performance against the main development branch baseline.