Video recording and production done by JSConf
This talk is about a modular frontend which is an approach not common to most frontend developers. No jQuery, no frameworks. Instead many small, simple abstractions are used, published to npm and bundled with browserify.
The NPM-style frontend prefers modularity and simple abstractions that do one thing well. THis talk includes:
Various approaches to creating standalone “components” or “widgets” that are published to npm, including styles & templates.
How to compose various modules together to build your own “framework” with your own trade offs.
How to compose an application from smaller applications
Build small UI components with thin interfaces that you glue together in your application.
Your entry point is small, containing the minimal glue code and each UI component addresses its own concerns.
MontageJS is an open source framework for building large single-page Web applications. The framework’s unique component model can simplify frontend Web development and increase opportunities for code reuse. In this session, you will learn how to take advantage of the MontageJS component system to build modular Web apps that are easy to maintain as they grow. The presentation will illustrate how component-based composability strengthens frontend Web development.
We’re all building a client-side framework. And they’re all different implementations of the same stuff. I want to compare the approaches that popular frameworks take to solving these patterns. If they even try to solve them at all. I also want to include info about the custom framework that Yammer uses. It’ll be illustrative to see how all of these have different approaches to the same set of patterns.
We have everything we need to read DVD-video discs in a browser, so why don’t we start?
But this comes with technical challenges such as browsers being bad at manipulating huge files or not supporting the codecs used in DVD (MPEG-2, AC-3…). That’s why this port comes with a server in Node.js to circumvent these limitations.
The resulting project is a mix of websockets, video elements, media source extensions and a lot of open source love.
Get ready for a mind blowing demo!
Ember, Angular, React, Polymer, and Backbone. Following in jQuery’s footsteps, these projects (and others) are helping to drive some future APIs in the browser.
While they all get lumped into the category “Frontend MVC”, an intimate look at each reveals they are quite different. These stark and subtle differences matter when choosing one for your project.
In this session, you’ll see each project’s sweet spot, and where each struggles. You’ll get a healthy dose of code as we explore the primary APIs. You will walk away with a better understanding of the goals and intended use-cases for each project.
Emotional safety is the most important predictor of the health of an organization and a community. But more importantly, it has a enduring repercussions on our lives as individuals. Like, our real lives—the ones we live for our whole lives.
I’ve worked in intensely unsafe professional environments prior to joining the team at &yet.
I’d like to candidly share some of the lessons I’ve learned from my unique perspective as a female who isn’t a software developer but who spends much of her life around them (including co-organizing events like RealtimeConf, RedisConf, and TriConf).
Being “in the developer world but not of it” has given me perspective to identify blind spots within our culture as a company and a community. Traditionally copywriters on a software team are viewed as a supporting role: we help sell what gets made. But what if all contributors to a team were true equals capable of leading and shaping a continuously improving company culture? That’s the foundation of emotional safety.
How does the transition into ES6 modules work for browsers exactly? How do we enable modular version-managed front-end architecture in the process?
In short, jspm starts off by making ES6 modules work dynamically in browsers today through the dynamic ES6 Module Loader polyfilll, then we make all existing module formats work through this same loader, then then we add offline build workflows for production and a command-line package manager to allow modular dependency management. Finally turn the whole thing around and add a CDN with server-push, and you have install-free version-managed module loading in the browser.
We spend so much time building things that we sometimes forget that we’re building for, and with, other complex human beings. Remembering that we are people first is the first step to making a positive shift in the way we treat each other.
New to working with ? Never fear, we’ll give you a light introduction to some of the high-level concepts about working in 3D space and go through some minimal examples before we do a deep dive into grittier details. We’ll look at in-browser tools as well as plugins found in multiple browser ecosystems: not just one, because the web is everywhere and your game should be too.
We’ll talk about how to debug and improve the performance of both 2D and 3D canvas contexts, including WebGL code. Though we will be talking largely about games, we’ll also look at chart libraries as well and discuss using canvas as compared to SVG. Even if you’re only working with a 3rd party library, you need to know how to profile performance and spot jank, as well as what action to take when you find it. You’ll learn how to mitigate the overhead of doing a lot of 3D math, as well as how to get good performance on low-end devices.
In this talk we’ll learn how to use existing tools to calculate and plot realtime satellite positions on a map using Leaflet.js and satellite.js. Along the way we’ll sneak in some orbital mechanics concepts, minus a lot of the hardcore math.
Perturbations, polar orbits, orbital eccentricity, Kepler’s laws, orbital propagation, oh my!
We will also learn about the pitfalls of 2D map projections, and why a 3D projection is more accurate and intuitive. If there’s time, we’ll go one step further and build a 3D satellite tracker using WebGL.
When writing complex business logic, it is critically important to maintain clean code though the judicious applications of Test Driven Development and Domain Driven Design. However, even these powerful techniques fall short of solving the problem at the heart of building complex software: building what the customer actually wants.
Domain Specific Languages (DSLs) allow us to capture complex business requirements in code written in the language of the customer. Once an ubiquitous language between you and your customer is defined and implemented as a DSL, the code can quite literally be given back to the customer to edit and refine. This is not a theory, or a myth. I have done this under real-world constraints and deadlines, and you can as well.
How do you get the browser to render arbitrary HTML to a canvas? Sounds easy? We would beg to disagree.
We’ll investigate how each one performs for certain tasks, and I’ll help you forge your own build sword. Lastly, I’ll discuss the benefits of going for the module format Node.js uses, which is Common.js, and how you can leverage those modules in the browser, using a tool called Browserify.
There are a wide range of approaches which might be considered “isomorphic” — an app could share just a few tiny modules of business logic, or just templates, or route definitions; or it might try to share the entire application code. The more code that is shared, the more abstractions have to be built to shim the differences between the very dissimilar environments of web browser and server, and the more complex and tightly-coupled the result will be.
At Fidelity we have several security/quality checkpoints across many departments to validate that applications and platforms protect customer data. Security code reviews, penetration test, risk audits, legal compliance and many other factors go into signing off on an application. Fidsafe is a new virtual safe deposit box offering by Fidelity that is the first application to be served outside the Fidelity firewall on the cloud. Fidsafe challenges every aspect of how the organization builds and deploys software. We had to answer a lot of questions and provide practical tooling/solutions to get Node into production.
We will cover what it takes from top to bottom build and operate a secure and scalable service backend implemented in Node.js and deployed to AWS. Topics covered:
Node Process Management
Lifecycle management – Upstart and Forever
Smart defaults for scalability and uptime
Reactor — How we use cluster to scale across cores
Layering security using middleware
Strategies for bulletproof cookies
SSL termination strategies
Authenticating end-users and API consumers
Building a Secure PaaS — A brief overview
If you want it to be secure you have to build your own. What’s the minimum you need for Node?
Devops in across organizational boundaries — AWS, Python, Boto, AMIs, and Asgard
Ubuntu as PaaS — real solutions are diverse and polyglot
What will it take for native code to become part of the web platform? The web platform has many wonderful qualities, but easy interoperability with other platforms is not one of them. Do you want to simultaneously develop for both native mobile and the web? Or use an awesome library in your web app, but it’s written in C++? Well… you can. Almost. Maybe. Sometimes. Sort of. Many disclaimers and restrictions apply. But what will it take for the answer to just be “yes”?
To most people in JS, functional programmers are perceived as academic hipsters raving about things like applicative functors, semigroup homomorphisms and Yoneda lemmas for no good reason except to make the rest of us feel stupid. And this is fair; there’s no better way to make you feel pitifully mainstream than throwing category theory at you. Conversely, JS programmers tend to believe functional programming, therefore, can have no real world application because nobody in the real world has any idea what a Yoneda lemma is and they seem to be getting by just fine without it.
Except we aren’t. We’ve been living in callback hell for almost two decades now, and no matter how many control flow libraries we submit to npm, things don’t seem to be getting any better. And that’s where functional programming comes in—turns out callbacks are just functions, and those academics in their ivory towers with their Haskell compilers actually encountered and solved these problems long ago. And now we can have their solutions in JS too, because of functional reactive programming. To demonstrate, I’ll attempt to write a browser based game, from scratch, with ponies, using RxJS, everybody’s favourite reactive library, live on stage in 30 minutes with no callback hell in sight. And we’ll be finding out if this reactive stuff is all it’s cracked up to be or not.
However, as a community we’ve learned a lot over the last half decade or so, and I think we’ve learned how to tackle this problem head on. I’ve come up with a strategy that I think would allow us to collaboratively get to a good place with ES6 relatively quickly. I’ll go over this strategy and point to the areas where I think we ought to spend the most time to get the most positive results.
Web applications do not require an over-arching framework to be organized and maintainable. In fact, an application with modules from all over the place can in fact be maintainable, understandable, and encouraged. Instead of subscribing to one framework that provides developers with an entire slew of unused or unneeded functionality, an architecture can be built from the ground up using small modules from a variety of different sources.
We live in a time of vast computational resources - many of us carry around in our pockets what just thirty years ago would have been considered a supercomputer. But it’s not just the hardware, these bite sized supercomputers run software using state of the art dynamic compilation techniques to deliver stellar performance without sacrificing flexibility.
While all of this may sound incredibly futuristic, many of us still program these Dream Machines with miserly techniques not far removed from the best practices of the 1960s.