@wycats I think chrome needs position: absolute also
-
-
-
@wycats@legomushroom@domenic : looking, but be aware that it'll only be for LAYOUT, not style recalc/invalidation. about:tracing FTW. -
@wycats : this is what you need Shadow DOM for: it creates a style recalc boundary. 60fps nearly impossible w/o.@legomushroom@domenic -
@slightlylate@legomushroom@domenic indeed. Just trying to get the lay of the land. -
@wycats@legomushroom@domenic : it got a lot more complicated since last I looked: https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit/Source/core/layout/style/LayoutStyle.cpp&sq=package:chromium&type=cs&l=368 … -
@slightlylate@legomushroom@domenic blunt force instruments are ok (things that work reliably vs things that work if you're lucky) -
@wycats@slightlylate@domenic just retested ff, chrome, safari and ie. your first twit was accurate - that 3 props is enough - 5 more replies
New conversation -
-
-
@wycats accelerated context vs non accelerated context could cause things to overflowhttps://www.youtube.com/watch?v=yq1eS6KgK5A#t=674 …Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
- 1 more reply
New conversation -
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@wycats I think just overflow hidden is enough for most casesThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@wycats I guess it depends a bit on the trust/assumptions between parent & child contentThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@wycats the `box-model` is probably relevant. `position:relative` for most containment, and maybe `z-index` to control assumed depthThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@wycats to gpu accelerate the rendering you'd also need a 3d transform (and a clip/clip-path property if you can) -
@FremyCompany@wycats well, gpu accelerate transforms, animation, opacity in webkit/blink. Otherwise not needed. -
@jacobrossi@wycats +maybe some sort of 2x-buffering, so if your elm is stable but slow to render, it may improve paint on invaldiation, no? -
@FremyCompany@wycats this happens automagically based on implicit layer calculation; trying to force it is really a hack and unpredictable -
@jacobrossi@wycats interesting point -
@FremyCompany@wycats there's also diminishing returns as the layers increases, on lowerend devices perf can decrease. So forcing can be bad -
@jacobrossi@wycats I know that, but magic means you don't have control. Heuristics are hard to diagnose and get interoperable.
End of conversation
New conversation -
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.