This is why compositors are a bad fit for browsers: the CSS model is a complex scene graph, not pixels of rectangles.
-
-
Yup, until you need to go cross process. Or scroll one viewport inside another
1 reply 0 retweets 0 likes -
Replying to @erincandescent @oshepherd and
We solved both problems in WebRender. Export display lists across IPC, and have the notion of scroll roots deeply integrated into rendering.
2 replies 0 retweets 0 likes -
Are the scroll roots rendered down to textures? Because if not you end up wasting a bunch of energy...
1 reply 0 retweets 0 likes -
Replying to @erincandescent @oshepherd and
No, you are not wasting energy, for the reasons I explained. You paint fewer pixels than the traditional approach.
1 reply 0 retweets 0 likes -
Repainting the 80% of my Twitter page which only moved upwards 100 pixels when I scrolled it is categorically a waste of energy
2 replies 0 retweets 0 likes -
Replying to @erincandescent @pcwalton and
(Don't get me wrong: I love WebRender! But I also think big scroll-able panes should be handled specially)
1 reply 0 retweets 0 likes -
Replying to @erincandescent @oshepherd and
We thought about caching, but after doing measurements we found that it didn’t help, and it used a lot of memory for tiles.
2 replies 0 retweets 0 likes -
Replying to @pcwalton @oshepherd and
Web pages are so utterly trivial for modern GPUs when properly optimized (batching, Z-culling) that there’s little point in caching.
2 replies 0 retweets 0 likes -
Are you Z-culling on the CPU or using the stencil buffer?
1 reply 0 retweets 0 likes
Depth buffer.
-
-
Thanks. More precision and still fixed function, so very fast. So, what causes the CPU bottlenecks?
0 replies 0 retweets 0 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.