Your current filters are…
by Estelle Weyl
While perceived download speed is vital to web application performance, mobile brings additional concerns to the table. Mobile devices have limited bandwidth, limited storage, limited memory and limited battery life. With mobile device constraints in mind, use best mobile practices to develop responsive, effective web applications that perform regardless of OS or device.
Come learn about the factors that influence web performance metrics such as onload, time to first byte, and above the fold time, as well as which web performance best practices can be applied to optimize each metric. We'll spend time doing deep dive analyses of example web pages to see which factors influence these metrics as well as what can be done to speed them up.
by Aaron Kulick and Cliff Crocker
We (Cliff Crocker & Aaron Kulick) will take the audience through a working session that will leave them with a functional Web Performance Dashboard that they can customize to their individual needs. We will cover some of the available open source projects and tools as well as discuss the importance of synthetic vs real user monitoring (R.U.M.) and how they can be combined in a single pane.
by Ian White
Tools like YSlow and PageSpeed have popularized the notion that we can dramatically optimize the performance of our websites. But actually implementing those recommendations often puts a heavy burden on the development and deployment process. Come learn about real-world techniques using open source software to increase your website performance without drastically changing the way you develop.
by Dominic Hamon and Arvind Jain
We will present techniques to predict user activity in the browser using learning models as well as hints provided by page authors and share the real world speed gains observed using such predictions.
by Philip Tellis, Buddy Brewer and Carlos Bueno
Real user monitoring is one of the best ways of learning “the truth” about what visitors experience on your web site, but it comes at a cost. The real world is messy and noisy making it hard to know exactly what’s going on. Filtering your data, splitting it along multiple dimensions, and determining what to discard are important second steps on the path to insightful RUM analysis, and in this session, we’ll go into some of the details.
by Marcel Duran and David Calhoun
Why is this page rendering so slow? YSlow and other performance analysis tools give useful advices on improving the page load speed, how about after the page is loaded when users are actually interacting with it? The results of a data driven experiment with a handful of FE techniques in different browsers and devices will provide new YSlow ruleset of advices to boost rendering and user experience
by Guy Podjarny
Mobile is different. Gaps in the networks, hardware and software used separates the mobile world from the desktop world. But for web performance, these differences are NOT equally important. In this presentation, we’ll dig into these differences with real numbers, helping us focus on the ones that really matter for performance. We’ll review old and new data on where the mobile difference comes into play, bust some myths, and end with some conclusions and takeaways for what you can do about it.
When users come to Twitter.com, the first thing they want to do is read their timeline. We're currently rolling out performance improvements aimed at delivering this content as fast as possible.
We use JavaScript timers via setTimeout() all the time, but few really understand how they work and their performance implications. As animations and complex processes become more common, a good understanding of how timers work, as well as the new alternatives, is key to good front-end performance.
Paul Irish said there a four primary concerns for CSS Performance: selector matching, style recalculation, reflows, and repaints. All of those are impacted by the efficiency of your CSS. In this session, Nicole will detail the tools every UI engineering team needs to figure out when and why their CSS has become slow.