Flight

Every minute Twitter’s Observability stack processes 1.5+ billion metrics in order to provide Visibility into Twitter’s distributed microservices architecture. In this talk will focus on some of the challenges associated with building and running this large scale distributed system. We will also focus on lessons learned and how to build services that scale that are applicable for services of any size.

Safety is crucial to Twitter as a social platform. In this session we present BotMaker, a scalable system that defends against unsolicited content and behavior. You will hear how we protect the Twitter platform and make it a better place for users and developers.

Finding and hiring great mobile developers is hard. Fabric (formerly Crashlytics) started with 1 engineer and now we are a team of 50 engineers. In this talk we’ll share some tips and practices that we use in Fabric and Twitter for hiring great engineers, including mobile developers. We’ll talk about the whole circle of hiring - from sourcing a prospective candidate to getting the initial coffee with them, evaluating their skill set as relevant to the job and ensuring a cultural fit in the team, to finally closing them and getting them excited to be part of the journey we are on of empowering App developers to build the best Apps.

We'll cover 3 key problems iOS developers face and how to solve them using fastlane: Generating localized screenshots, provisioning profiles and beta deployment.

This talk will be a technical deep dive into how we developed Digits from idea until today (login, theming, find friends, 2 factor, voice verification). The talk will show what is the development process of the team and what are the design patterns we use in the implementation of Digits SDK.

The MoPub Marketplace is a real-time exchange which handles billions of auctions daily amongst more than 180 real-time bidding platforms. In this session, we’ll discuss how the exchange works, how exchanges maximize developer revenue, and take a look at the technologies which make this scale possible.

At Fabric, we take a customer-centered approach to our workflow, asking, 'what value are we delivering to customers this week?' This session will cover how we use a Flow process to address bottlenecks and keep features, improvements, and fixes flowing to customers.

Twitter has used Aurora and Mesos to grow from only tens of nodes to tens of thousands. During the process of scaling up, the configuration, deployment, and operational procedures for this system have evolved and improved significantly. This talk will offer an operations perspective on the management of a Mesos+Aurora cluster, and cover many of the cluster management best practices that have evolved at Twitter from real-world production experience.

The locations of geotagged tweets over time are a valuable source for understanding patterns of urban activity and mobility. I'll discuss retrieving tweets from the Twitter Streaming API, despeckling, debanding, and de-duplicating the data to reduce distracting artifacts, generating vector map tiles with Tippecanoe, and styling the map with Mapbox Studio, along with some highlights of the resulting map.

The best mobile apps leverage remote services to store rich personalized data, perform secure transactions, sync accounts, and more. Fabric's single sign-on kits let your app access the Twitter and Digits APIs, but they can do more. In this talk, we'll discuss how Twitter and Digits can authenticate your app users to your backend, design a small Go auth service for a Twitter/Digits app, and use containers to deploy it.

Testing an app on your perfect Nexus 5 device under perfect network conditions works great, but you're getting reports of strange behavior from the wild and weird crashes that you can't reproduce. This talk will walk through some of the more complex options on the Developer Options screen in the Android OS and discuss how a developer can use them to debug problems in their application. We'll cover these options and more as we dive deep into debugging apps.

Handling uncaught exceptions in Java is a walk in the park compared to handling a native crash. This talk will introduce you to the inherent difficulty of signal handlers by outlining the lessons that we have learned while developing Crashlytics NDK support. Topics covered include: Java crash handling, async-safety, static storage duration and memory allocation, interruptions, signals, stack unwinding, reinventing the wheel, and more!

The Accelerated Mobile Pages (AMP) Project is a recently launched open source initiative, bringing publishers and technology companies together in an effort to speed up the mobile web. Pages build with AMP appear to load instantly, but what’s its secret sauce? Learn why we came up with AMP, the techniques behind all of our optimizations and why you should care.

Twitter for Websites has been a mainstay of the Twitter development platform for over five years; an easy means to integrate Twitter into articles and applications on web and mobile. In that time our widgets have evolved with the web itself. Here, engineers from the TwitterKit for Web team will talk about rebuilding the Twitter widgets suite from the ground up, implementing our overriding passion for performance in all its forms.