I've been spending a whole lot of time in the invalidation code recently, and found that some function names are rather poor, and the code is unclear and under-documented in many places. This bug covers a bunch of cleanup patches I want to land.
Created attachment 372847 [details] [diff] [review] patch 1 Here's some stuff I separated from the patch in bug 463939.
+ // XXX [perf] here we call UpdateCoveredRegion on all our children, not just + // the active child. Maybe we should optimize for that? Or should + // getBoundingClientRect() always return something sane for *all* children? Oops, we should just update the active child covered region (and perhaps the previously active child covered region).
I'm not sure. I filed bug 495968 for that part.
Created attachment 381096 [details] [diff] [review] patch
Comment on attachment 381096 [details] [diff] [review] patch >+ // Areas dirtied by changes to decendents that are in our document s/descendent/descendant/ r=longsonr with nit picked.
Attachment #381096 - Flags: review?(longsonr) → review+
Status: NEW → RESOLVED
Last Resolved: 10 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.