bgirard

29 ms·

391.
by bgirard 11y ago
You vastly overestimate the cost of copying 50k elements. On low end mobile you'll get at least 1GB/sec and on desktop you can get easily 10GB/sec. Now sorting at these speeds is a whole other story. The real thing to conside
392.
by bgirard 12y ago
I'm the author of the feature & blog post, it wasn't mean to hit HN. This is meant for gecko developers right now as it only measures the compositing GPU time, and not WebGL GPU time (different thread+Context). If this proves
393.
by bgirard 12y ago
That's right. On the graphics side if you drop the resolution of the screen you can drop a lot of memory and CPU power. Trading off features like pre-rendering (display port) on the newer builds will save you more memory and CPU cycles
394.
by bgirard 12y ago
Exactly. To elaborate on how this works in Gecko. The rendering pipeline has several optional stages: requestAnimationFrame (Scripts) -> Style flush -> Reflow flush -> display list construction -> Layer construction (recycling)
395.
by bgirard 12y ago
I'd like to know as well. I just tried but I can't get it to work following the documentation. Edit: It seems that the request must be made over HTTPS.
396.
by bgirard 12y ago
What will-change optimization is used under the covers is an implementation detail. Declaring what types of changes you want optimized is something best suited for the page developer. Letting the rendering engine infer and prepare for chang
397.
by bgirard 12y ago
Using them in the CSS style sheet is perfectly reasonable. Here's how b2g/ffos uses them carefully to make the most of low end hardware: http://mxr.mozilla.org/gaia/search?string=will-change Imagine you'
398.
by bgirard 12y ago
I think 'flicker' is the wrong word. What the author means is delay/latency/jank. 'Profiling' or inferring the animation is what engine already do. The problem is as an author you're not getting predictabl
399.
by bgirard 12y ago
I made the original proposal for will-change on Mozilla' behalf. I agree with your concerns, we've had them since 2010 and kept pushing back on proposing this. But with HiDPI mobile and 4k desktop it's really difficult withou
400.
by bgirard 12y ago
It's there: https://www.khronos.org/registry/webgl/specs/latest/1.0/webg...
401.
by bgirard 12y ago
It's true. It wont provide anything on its own so I'm trying to be blunt to kill this misconception that keeps spreading.
402.
by bgirard 12y ago
We were talking about this yesterday. I asked how many of the previous issues we classified a security problem would of been mitigated by a sandbox and the conclusion would be that it would not even cover the majority. Most of the security
403.
by bgirard 12y ago
Electrolysis is NOT sandboxing, full stop.
404.
by bgirard 12y ago
http://people.mozilla.org/~bgirard/cleopatra/#report=a848ab9... The first PDF.js document uses the DOM instead of Canvas like the previous one. This is done to support text selection. Most of the time is spent in
405.
by bgirard 12y ago
Here's a profile: http://people.mozilla.org/~bgirard/cleopatra/#report=0705c03... I get between 5-7 seconds here. Looks like getaliasedvar is causing excessive bailouts from Ion (3rd tier JIT). On top of that
406.
by bgirard 12y ago
Yes the fix to bug 1007897 will help all web content. That profile will show calls from JS and how theycall into native C++ calls. This is very useful for profiling things like canvas. PDF.js isn't bottlenecked by javascript performanc
407.
by bgirard 12y ago
I just profiled your example: http://people.mozilla.org/~bgirard/cleopatra/#report=4b995cc... Looks like it could be running much faster: * 20% of the time is spent copying the canvas because someone is, likely er
408.
by bgirard 13y ago
Firefox recently (Trunk as of December) started using Unified Sources so builds are now typically 7-9 minutes on laptops now for non Windows platforms.
409.
by bgirard 13y ago
Cool! If you have any questions about canvas and/or Firefox performance while working on this come hang out in irc.mozilla.org #gfx.
410.
by bgirard 13y ago
This article gives a good overview of NPAPI vs. PPAPI and NaCl/PNaCl. But what it gets wrong is the primary reason why Mozilla is objecting. It's not asm.js vs. PNaCl. It's really about HTML5 vs. PPAPI. > [PPAPI] lets a pl
411.
by bgirard 13y ago
That's a good point. Thanks!
412.
by bgirard 13y ago
"The Ballmer Years" stock price graph is extremely misleading because the number of outstanding shares has changed. If they want to track Ballmer' performance they should graph market cap.