Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
If anyone needs a test page, you can log in as my test user styoung.tra...@gmail.com (pwd:browsertest). Then go to https://www.facebook.com/styoung. you could maintain a separate document for measuring items, so you could measure without reflowing the main document. We are actually already doing that. Kelly Norton suggested offline to me that the problem could be layout thrash caused by us doing interleaved dom reads/writes (one for each story) as opposed to a series of reads followed by a series of writes. That sounds right to me. (2) 50% of time spent painting images... This is a simple speed vs quality tradeoff. If you down-sampled the images on the server, they'd download and paint much faster. Thanks. Downsampling sounds like a straightforward solution. We can show the higher quality image if they open the photo. Btw, what tool are you using that tells you what item is being repainted when the cpu is pegged? ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
On Feb 14, 2012, at 10:25 AM, Steven Young wrote: (2) 50% of time spent painting images... This is a simple speed vs quality tradeoff. If you down-sampled the images on the server, they'd download and paint much faster. Thanks. Downsampling sounds like a straightforward solution. We can show the higher quality image if they open the photo. The significant time spent painting images could simply be a sign that too much of the page is being repainted when scrolling, which could be a side effect of unnecessary re-layouts. So this could just another symptom of the layout thrash problem. Thus, I would suggest fixing the other issues and retesting before you move to downsample images. Btw, what tool are you using that tells you what item is being repainted when the cpu is pegged? I would guess that Geoff is using the Instruments tool on Mac OS X, but it likely takes some expert knowledge to interpret the profile. You can tell by looking at the profile what kind of things are getting painted, but not necessarily the specific item. REgards, Maciej ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
If anyone needs a test page, you can log in as my test user styoung.tra...@gmail.com (pwd:browsertest). Then go to https://www.facebook.com/styoung. Nice! I took a trace of this timeline and saw similar results as before (lots of time computing .offsetHeight and .scrollLeft), but with less time spent in image drawing. (Perhaps I have higher resolution photos in my timeline.) Btw, what tool are you using that tells you what item is being repainted when the cpu is pegged? Mac OS X ships with a performance analysis tool called Instruments. Documentation overview: https://developer.apple.com/library/wwdc/ios/#documentation/DeveloperTools/Conceptual/InstrumentsUserGuide/Introduction/Introduction.html#//apple_ref/doc/uid/TP40004652 Recent WWDC presentation: https://developer.apple.com/wwdc/schedule/details.php?id=310 Instruments analysis doesn't work very well with the current shipping version of WebKit, but it works great with WebKit nightly builds (nightly.webkit.org). Geoff ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
It's hard for me to advice you on how to optimize your website but will you be interested in creating a reduced test cases where WebKit is slow? I'm sure we can (at least try to) resolve your pain points if you can create benchmarks licensed under BSD/LGPL or WebKit performance tests (see http://trac.webkit.org/wiki/Writing%20Performance%20Tests). Best regards, Ryosuke Niwa Software Engineer Google Inc. On Sat, Feb 11, 2012 at 10:02 PM, Steven Young styoung.bi...@gmail.comwrote: [cross posting from mozilla's dev lists] I'm on the Timeline team at Facebook, which is going to be the new format for everyone's profiles real soon now. https://www.facebook.com/about/timeline We'd like to improve its browser performance, so I'd appreciate any suggestions for things we should change to speed it up. In particular, we'd like to make scrolling down through new content smoother. There are often brief (e.g. 300 ms) browser lockups, and other times there just seems to be a general feeling of heaviness. I'm going to list some of the specific issues we've identified, which we are debating how best to fix, but I'm also very interested to hear whatever anyone else thinks are the biggest perf bottlenecks. A few problems: (1) HTML / DOM size and CSS Our HTML is huge. About half of it is coming from the light blue like/comment widgets at the bottom of most stories. Within those widgets, a major portion of it is always the same. (Some of that is only needed once the user clicks into the widget, but we don't want another server round trip to fetch it.) We also have a lot of CSS rules, and applying all that CSS to all those DOM nodes gets expensive. Experimentally, removing all like/comment widgets from the page does give noticeably smoother scrolling, although it doesn't completely fix the problem. Related: We've also noticed that if you scroll very far down a content-rich timeline, and then open and close the inline photo viewer, this causes a noticeable lag, as it re-renders all existing content on the page. To fix this, we investigated dynamically removing offscreen content from the DOM and replacing it with empty divs of the same height, but we decided it wasn't worth the code complexity and fragility. (2) Repaints There are several fixed elements on the page like the blue bar at the top, the side bar, and our date navigator with the months/years. Chrome's --show-paint-rects flag showed that under most circumstances these fixed-position elements forced full-screen repaints instead of incremental repaints. The rules for what triggers a repaint vary from browser to browser, but we would ideally like to fix this everywhere. The cost of full page repaints also sometimes varies dramatically even comparing Chrome on two fairly newish Mac laptops. (3) Javascript for loading content as you scroll down We dynamically load timeline sections (e.g. a set of stories from 2009) using our BigPipe system (https://www.facebook.com/note.php?note_id=389414033919) in an iframe. In a nutshell, the HTTP response to the iframe is sent with chunked encoding, a script tag at a time. Each script tag contains some code and and HTML content that is passed up to the parent window, which requests the CSS and JS associated with that HTML content. Once the CSS is downloaded, the HTML (timeline story markup) is inserted into an offscreen DOM element. Then, once the JS is loaded, we do some fairly complicated work before we actually display the content. First, we lay out the timeline stories in an offscreen element (position:absolute; left:-px) before inserting them into the viewable page. We then have JS which checks the heights of all the stories on in the offscreen element so it can swap stories back and forth between the two columns, to keep things sorted by time going down the page. To do this, we query and cache the stories' offsetTop values all at once where possible. Probably, we could eliminate all this height-checking and column balancing if we implemented a machine learning algorithm to predict the height of each unit in advance, on the server side. Next, in an attempt to reduce user-percieved browser freezing while scrolling, our JS does not add new content in to the bottom of the main column as soon as it comes back from the server. Instead, we queue it up until the user stops scrolling and add it in then. We use document fragments where possible to insert elements. Web Inspector's profiler showed improvements when dynamically inserting many link rel=stylesheet tags in this fashion since we stopped thrashing between style recomputation and JS execution for each stylesheet, and instead just had one longer style recomputation segment. We throttle scroll/resize events so they fire every 150 ms All the while this is happening, we're potentially receiving more script tags in the iframe and doing the same thing for other pieces of content. We would
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
Hi Steve. Do you have a test account with a fixed content set that we can use for profiling? It's hard to speculate about performance issues without profiling, and we might get confused if we all profile different content. Thanks, Geoff ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
Profiling scrolling through my own timeline, and focusing on points where the CPU hit 100% or greater, I saw this: (1) 50% of time spent in style calculation forced by accessing element.offsetHeight in JavaScript. We then have JS which checks the heights of all the stories on in the offscreen element so it can swap stories back and forth between the two columns, to keep things sorted by time going down the page. One sometimes pernicious effect of accessing style-related properties while changing the DOM is that you force twice (or n times) the work to happen: first, style resolves to supply your property value; then, you change the DOM, and style resolves again to account for your change. Since style resolution is generally O(n), this can easily become O(n^2) behavior. According to my measurements while scrolling my own timeline, you could make scrolling twice as buttery by removing these accesses to element.offsetHeight, or doing them on a zero-delay timer after all DOM changes. (2) 50% of time spent painting images. This is a simple speed vs quality tradeoff. If you down-sampled the images on the server, they'd download and paint much faster. Geoff ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
Not sure what tools you have used but you may find this helpful: http://code.google.com/webtoolkit/speedtracer/ On Sat, Feb 11, 2012 at 10:02 PM, Steven Young styoung.bi...@gmail.comwrote: [cross posting from mozilla's dev lists] I'm on the Timeline team at Facebook, which is going to be the new format for everyone's profiles real soon now. https://www.facebook.com/about/timeline We'd like to improve its browser performance, so I'd appreciate any suggestions for things we should change to speed it up. In particular, we'd like to make scrolling down through new content smoother. There are often brief (e.g. 300 ms) browser lockups, and other times there just seems to be a general feeling of heaviness. I'm going to list some of the specific issues we've identified, which we are debating how best to fix, but I'm also very interested to hear whatever anyone else thinks are the biggest perf bottlenecks. A few problems: (1) HTML / DOM size and CSS Our HTML is huge. About half of it is coming from the light blue like/comment widgets at the bottom of most stories. Within those widgets, a major portion of it is always the same. (Some of that is only needed once the user clicks into the widget, but we don't want another server round trip to fetch it.) We also have a lot of CSS rules, and applying all that CSS to all those DOM nodes gets expensive. Experimentally, removing all like/comment widgets from the page does give noticeably smoother scrolling, although it doesn't completely fix the problem. Related: We've also noticed that if you scroll very far down a content-rich timeline, and then open and close the inline photo viewer, this causes a noticeable lag, as it re-renders all existing content on the page. To fix this, we investigated dynamically removing offscreen content from the DOM and replacing it with empty divs of the same height, but we decided it wasn't worth the code complexity and fragility. (2) Repaints There are several fixed elements on the page like the blue bar at the top, the side bar, and our date navigator with the months/years. Chrome's --show-paint-rects flag showed that under most circumstances these fixed-position elements forced full-screen repaints instead of incremental repaints. The rules for what triggers a repaint vary from browser to browser, but we would ideally like to fix this everywhere. The cost of full page repaints also sometimes varies dramatically even comparing Chrome on two fairly newish Mac laptops. (3) Javascript for loading content as you scroll down We dynamically load timeline sections (e.g. a set of stories from 2009) using our BigPipe system (https://www.facebook.com/note.php?note_id=389414033919) in an iframe. In a nutshell, the HTTP response to the iframe is sent with chunked encoding, a script tag at a time. Each script tag contains some code and and HTML content that is passed up to the parent window, which requests the CSS and JS associated with that HTML content. Once the CSS is downloaded, the HTML (timeline story markup) is inserted into an offscreen DOM element. Then, once the JS is loaded, we do some fairly complicated work before we actually display the content. First, we lay out the timeline stories in an offscreen element (position:absolute; left:-px) before inserting them into the viewable page. We then have JS which checks the heights of all the stories on in the offscreen element so it can swap stories back and forth between the two columns, to keep things sorted by time going down the page. To do this, we query and cache the stories' offsetTop values all at once where possible. Probably, we could eliminate all this height-checking and column balancing if we implemented a machine learning algorithm to predict the height of each unit in advance, on the server side. Next, in an attempt to reduce user-percieved browser freezing while scrolling, our JS does not add new content in to the bottom of the main column as soon as it comes back from the server. Instead, we queue it up until the user stops scrolling and add it in then. We use document fragments where possible to insert elements. Web Inspector's profiler showed improvements when dynamically inserting many link rel=stylesheet tags in this fashion since we stopped thrashing between style recomputation and JS execution for each stylesheet, and instead just had one longer style recomputation segment. We throttle scroll/resize events so they fire every 150 ms All the while this is happening, we're potentially receiving more script tags in the iframe and doing the same thing for other pieces of content. We would love any pointers you guys have. Thanks, Steve ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev ___ webkit-dev mailing list
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
will you be interested in creating a reduced test cases where WebKit is slow? Ryosuke - For now, user complaints about slowness are too unpredictable and poorly defined for me to create a simple test case. I will report back here if we reach that point. (1) 50% of time spent in style calculation forced by accessing element.offsetHeight in JavaScript. Geoff - I am going to bite the bullet and rip this logic out. We are pushing too much complexity into the browser. (2) 50% of time spent painting images... This is a simple speed vs quality tradeoff. If you down-sampled the images on the server, they'd download and paint much faster. Geoff - Painting images specifically, or just repainting the page in general? Not sure what tools you have used but you may find this helpful: http://code.google.com/webtoolkit/speedtracer/ David - Thanks! ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Re: [webkit-dev] optimizing browser handling of Facebook Timeline scrolling
(1) 50% of time spent in style calculation forced by accessing element.offsetHeight in JavaScript. Geoff - I am going to bite the bullet and rip this logic out. We are pushing too much complexity into the browser. Bear in mind that I didn't do enough analysis to explain why the .offsetHeight code was so costly. It may be possible to tune this code and keep it in the browser. For example, you could maintain a separate document for measuring items, so you could measure without reflowing the main document. (2) 50% of time spent painting images... This is a simple speed vs quality tradeoff. If you down-sampled the images on the server, they'd download and paint much faster. Geoff - Painting images specifically, or just repainting the page in general? Painting images specifically. Geoff ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev