To clarify, when the original problem was discovered in cosmic the CPU usage was around 80%, and the fix made it less than 50% [1].
But now even without the fix (although in bionic, still using the same hardware), CPU usage for the same test case is around 50%. So indeed the main issue seems to be gone.
What I can't figure out is why the problem isn't reproducible in bionic even when the root cause was introduced in bionic version 1.52.1-1ubuntu1. The bug and the effect of the fix was much more obvious in cosmic and disco...
Still, gjs version 1.52.5-0ubuntu18.04.1 eliminates even the theoretical possibility of the CPU problem and it allows us to drop all patches \o/
To clarify, when the original problem was discovered in cosmic the CPU usage was around 80%, and the fix made it less than 50% [1].
But now even without the fix (although in bionic, still using the same hardware), CPU usage for the same test case is around 50%. So indeed the main issue seems to be gone.
What I can't figure out is why the problem isn't reproducible in bionic even when the root cause was introduced in bionic version 1.52.1-1ubuntu1. The bug and the effect of the fix was much more obvious in cosmic and disco...
Still, gjs version 1.52.5- 0ubuntu18. 04.1 eliminates even the theoretical possibility of the CPU problem and it allows us to drop all patches \o/
[1] https:/ /gitlab. gnome.org/ GNOME/gjs/ merge_requests/ 236