Image assets are a major impediment to page load times. Best practices for common image compression have been evangelized. To further reduce image bytesize, we have to get creative: discover how to apply lossy compression to lossless formats, automatically detect good compression levels and learn how disciplines like computer vision and machine learning can help us to improve image compression.
Everybody needs to care about performance but often only the engineers do. An exploration of how design processes and wrapping a layer of visual storytelling around the data can engage the whole organization and motivate change.
Using new research and real world examples, Patrick will cover a range of techniques – from the controversial to bleeding edge – the Guardian are using to make their next generation website load as fast as possible, and ultimately breaking the news to the user within 1000ms.
Human logic, unlike that of the machines which we program and use every day, isn't perfect. We make mistakes, we establish bad mental habits, and we have many cognitive biases that negatively impact our ability to be successful engineers. In this talk I will outline some of the most common biases that show up in engineering organizations, and how to recognize and overcome them.
The same commitment to collaboration, communication and empathy that drives devops can be applied much more widely across organisations, bringing together radically different disciplines to respond to user needs. This session looks at how to drive that sort of collaboration and how vital it is to transforming large organisations.
Security testing is often seen as a specialist activity that requires external experts with leather jackets. But security tests have a lot in common with functional tests. In fact, by using Selenium and open source security tools we can build re-usable, automated security tests that fit a CD pipeline and we can write them as specifications using JBehave so that they're visible to the whole team.
“In times of drastic change, it is the learners who inherit the future.” Ever-changing technologies and evolving systems require we never stop learning, but how does successful learning happen in increasingly complex organizations?
Evolving mobile hardware and networks have made it challenging for web sites to deliver an optimal experience to each client. If you send the same image to both a WiFi Retina tablet and a 3G phone, you compromise speed and bandwidth cost against image quality. We'll look at using HTML and CSS image markup, CDNs, HTTP caching directives and how WPO can deliver a great UX with minimal effort.
This talk includes some guidance and tips for Operations folks attending design review meetings with Development teams in advance of the launch of a new feature or application.
How do you build a distributed cache invalidation system that can invalidate content in 150 milliseconds across a global network of servers? We will discuss the process of constructing a production-ready distributed system built on solid theoretical foundations. This talk will cover using research to design systems, the bimodal multicast algorithm, and the behavior of this system in production.
Mistakes happen. Hopefully not frequently -- but they do. Sometimes they are small, other times big. Sometimes they can be existential crises that can leave deep professional scars.
This talk describes such a mistake -- by the person who made it -- and how the person and the organization recovered and grew from it.
This tutorial covers a broad set of performance fundamentals — from user expectations to business metrics to optimization best practices — to help you hit the ground running for the next three days.
Movistar TV is the Telefonica online tv solution, broadcasting in Spain, Brazil, Chile, Argentina and Peru. Since 2013 a team of 200 people are working together in offering the best online experience and our goal for 2014 is FIFA World Cup. Continuous Performance Integration and monitoring is the key to success to achieve the best experience in events like Barça-Madrid in 2013
Adding a WebSocket service to an application is often misunderstood to be high performance by default, however there are many more considerations that must be made, both on the client and server, before the best performance can be achieved. This talk will provide the checklist and examples you need to fine tune your application’s real-time performance.
TLS has exactly one performance problem: it is not used widely enough. Everything else can be optimized.
Unless you are a startup with only one product line you probably have systems that you need to decommission. Chances are those systems are already slowing you down. This session is about the lessons learnt by the decommissioning team at the Financial Times while we tried to put away the old broken toys to make enough space to play with the new ones.
The Kanban Method is a way to evolve towards a more effective flow-oriented operation that is very aligned with DevOps thinking. In this session we will see how exactly this process works and how it can help guide legacy enterprises towards the promised land of a DevOps culture, structure and toolchain.
When things go wrong in organizations, one thing is almost always found in the post-mortem: ‘human error’. But things are not so straightforward. ‘Human error’ fails to capture the context and complexity of work and systems. There is a better way. Let’s journey through the steps of recovery, from explaining away events to understanding how your system really works.
PerfBar is a tool that puts dozens of metrics (as well as custom metric support) at the tip of your fingers with just a single script include.
HTTP Archive and Google Cloud Dataflow, Webpagetest-automation 2.0
PNG images with alpha transparency are incredibly useful in web design, but their large file sizes are at odds with good page performance.
With a special encoder it's possible to create PNG images that are up to 40% smaller, and still compatible with all browsers, support smooth transparency and millions of colors.
I'm going to show you how to do it, whether you're a designer or a developer.
As the web gets bigger and more ubiquitous, we will need more programmers and more people understanding programming. How can we do that? I'll share what I've learned from teaching programming to newbies, online at Khan Academy and in-person for women in San Francisco.
Mansplaining is tricky to identify, but easy to avoid. It’s also one of the biggest obstacles to creating a diverse technical team — especially one that’s welcoming to women. Learn how to foster a work environment that’s inclusive for all, and build a stronger team.
The journey to optimize customers’ experience across multiple channels breaks down the traditional business and IT approaches. Investments in faster Web pages and responsive apps, for example, position companies to capitalize on their digital initiatives, but many have not defined the performance metrics to measure success. How do you analyze and maximize the return of your digital initiatives?
Your software architects are probably out there drinking the Microservices kool-aid right now, and dreaming of building the next system using those patterns.
Microservices is a wonderful silver bullet that has the possibility and capability to make an operations teams life a complete nightmare.
Learn how to talk to your architects and dev team and help them think about the problems
Let’s face it. Alerting is broken. We are all still alerting the same way is we did in the early days of software based on metrics violations. We have all started to accept that we get too many alerts and the hard work of making sense is still left to us. This talk will introduce you to the concept of contextual alerting and show the difference hands on using a real world example.
Service Worker is set to turn the web on its head, enabling websites to run fast on any connection. Learn how to use it and its predecessor to eliminate the network to create web apps with unbeatable performance.
Everyone has strong opinions about progressive images. Some believe that progressive images improve performance by showing the user something while they look at the screen. But others believe that progressive images increase user frustration. We took these assumptions to the lab to see if we could definitively answer the question: do progressive images deliver a better or worse user experience?
As technology jobs become increasingly hard to fill, the average starting salary of an engineer in the valley is more than the median family income in the US for many demographics. Laine will discuss how to build your organization to become a culture and process that drives diversity in recruiting, hiring and retaining.
If you are building a mobile app or hybrid responsive app you are probably thinking deeply about reusing components and data APIs from your web site. In this talk we will explore some common pitfalls in using web components & web APIs in mobile apps. We will look at the impact on operations, network performance, scalability and reliability - and how to overcome these challenges.
Looking for an alternative to onload? Let's find one together.
Service orientated architecture and, more recently, microservice architectures have become popular patterns as large scale architectures. The idea of decoupling services and having independently deployable units lends itself well to large scale applications in the cloud.
Those same principles of autonomous units can also be applied organisationally leading to greater efficiencies at scale.
Responsive Web Design is the solution for improving your mobile site. Or is it? A bloated site that simply shrinks in width is not a usable site. Squishy design is not the only, or even the main, solution for improved mobile user experience. Learn how to leverage modern web standards to improve performance so your sites, responsive or not, are more usable, accessible.
Synthetic and real user monitoring are two necessary toolsets for anyone responsible for web performance. However, we often struggle with the best way to present the datasets together in a clear and meaningful way. This talk will focus on the relationship between synthetic and RUM metrics and the best presentation techniques for different audiences.
We live in the Impatience Economy where we want what we want, right here, and right now. How is Big Data re-defining Velocity? A new generation of startups, are building big data driven applications that participate in Real-Time Bidding (RTB) to create personalized experiences in real-time. How are these companies going from 0 to revenue in months, using high Velocity to create big Value?
This session will cover the tools and services we are developing to allow BBC teams to build scalable and resilient services on AWS and other providers. We will cover infrastructure provisioning, software building, baking of machine images, service isolation, security, deployment orchestration and more. We will also look at how these tools reflect and shape engineering culture at the BBC.
How can your organization scale management without sacrificing the culture you have worked hard to build? Grow your own managers! This talk will focus on the steps Etsy has taken to overcome the pitfalls that lie in the path of transitioning from individual contributor to manager and how to encourage those who want to become managers for the right reasons.
The APDEX was established to give us a metric to optimize our websites for performance but is it still valid? In this session we will explore what is needed in the mobile world of today to identify bad user experience and look into the new concept of a User Experience Index.
Testing, monitoring and analysis of website and web-application performance requires a range of analysis and reporting tools, if you have to create a comprehensive view of the influencing factors. In this session we look at how Limelight securely exposes data about CDN service and object delivery behavior, using HTTP headers.
This talk will make you think about the principles of your development process, to support new, more scalable web development practices. I’ll also be talking about a new public project sponsored by the FT and Fastly to build a community polyfill service to support those better practices. It’s time to upgrade the web.
It’s been seven years since the first commit, and today there are over seven million people building better software, faster, on GitHub. For over six years, the incredible community that has formed around the Velocity conference has presented and advocated for all things fast and stable, and everyone is listening. Let’s take a look ahead at what in store next.
Like a rampaging bull, the internet is a complex beast that’s ready to take down your site at a moment’s notice. Join Mehdi Daoudi, CEO and co-founder of Catchpoint Systems, to learn how you can elude some of the more unpredictable movements and get back on your feet.
Sites increasingly rely on third-party scripts to deliver features but there is little data on how they interact with our sites and what their impact on performance is.
We’ve been analysing the data from ites we monitor to get a better understanding of third-party performance and in this session will share our findings along with advice on how to make informed choices.
If you think there is a gap between dev & ops, then design must seem like it comes from another planet. But many of the core ideas behind design work are directly applicable to the world of ops, especially given that running a successful website is no longer (just) about keeping servers up, but about designing human & technical systems that help enable people achieve continuous operations.
The preloader is an essential performance optimization mechanism that exists across all modern browsers, yet it remains a mystery to many Web developers. In this session, we will explain what the preloader is, what it does, what developers need to know about it, and what the future may hold for it.
Ansible is a radically simple and lightweight provisioning framework which makes your servers and applications easier to provision and deploy. In this tutorial you will be given an introduction to Ansible and learn how to provision Linux servers with a web-proxy, a database and some other packages. Furthermore we will automate zero-downtime deployment of a Java web application.