Brujo Benavides

So you just joined a massive project with tons of poorly-maintained Erlang code. DON'T PANIC! This talk will give you many tools and techniques that you will need to start working on it confidently and productively.
You will learn multiple tips and tricks to incrementally start using the compiler, Common Test, Xref, Dialyzer, Elvis, Hank, the Formatter, and others in large codebases. A task that's generally hard to do but usually produces enormous benefits in terms of maintenance. These tools and techniques will increase your confidence, and they'll help you and your team work more efficiently.

OBJECTIVES
The goal of the task is to reduce the pain developers feel when faced with large poorly-maintained projects written in Erlang, by showing how the tools that they would use from day 0 on new projects can be introduced gradually even when they were never used before.

AUDIENCE
Almost any Erlang developer, maybe except for those who had worked with me in the past:)

This session will host an all spanish-speaking panel to talk about the BEAM; how the platform is currently used and what the future holds for anyone invested in it. Note that this panel will be delivered 98% in spanish.

Esta sesión tendrá como invitados a destacados miembros de la comunidad hispanohablante para discutir sobre BEAM; cómo es que aprovechamos la plataforma al día de hoy y qué es lo que el futuro aguarda para todos los interesados en ella. Manejaremos español durante gran parte de la sesión para la comunidad.

Working with legacy code can be very frightening sometimes. That's why most of us tend to touch as little as possible every time we need to make a change, particularly when we're removing stuff. That leads to lots and lots of "dead code" (pieces of code that are no longer needed but never removed… just in case). That piling of dead code makes debugging and maintaining systems harder and harder, and the cycle starts again…

That's why, at NextRoll, we created Hank, the Dead Code Killer.

Our latest open-source tool that you need to start using today!

OBJECTIVES:

To help developers lose the fear to delete code from large systems using Hank to help them.

AUDIENCE:

Erlang developers.

One of the first things every newcomer faces when they meet OTP are behaviours. The general explanation for them is usually along the lines of "they're like interfaces for OOP". While that's somewhat accurate, it's also misleading and introduces a lot of confusion on how and when to use the existing behaviours and when it's reasonable to define your own ones. In this talk, Brujo will explain what behaviours are, how they should be used and how to create and expose new ones if you ever need to.

OBJECTIVES

To help newcomers get familiar with OTP behaviours and to help experienced devs rediscover them.

TARGET AUDIENCE

Developers willing to understand what they really do when they implement a generic server or a supervisor. Also folks considering the creation of their own behaviours with the desire to do it right. If you don't know what a behaviour is at all, you're welcome here too.

Opaque Data Structures provide a very nice technique to structure systems and they are particularly useful within the functional programming paradigm. With the addition of maps and the opaque attribute for modules, Erlang is now perfectly suited to get the best of these structures, if you use it properly.

In this talk, Brujo will show what opaque data structures are, how to use them and why they are so useful, even more so if you're developing an open-source library. This talk will be conceptual, but it won't be 100% theoretical.

Brujo will also present examples and tools that assist us in making better use of these techiques.

OBJECTIVES

The main goal of the talk is to showcase a way to structure Erlang systems based on strong type definitions and explain why and how that structure is beneficial to Erlang developers.

AUDIENCE

Everyone, but it might be of particular interest for those building open-source libraries and/or big applications with lots of modules.