Every month, we go over the GitHub trending page for any cool repos that stand out from the crowd. We choose five of the most innovative, interesting, and well-thought-out projects to highlight in our monthly report. As always, it was tough to narrow down the list of GitHub repos to our five favorites.
So, in no particular order, here are our top picks for November 2019!
As smart homes and home automation moves from sci-fi fantasy to a common reality, privacy concerns grow. That’s why the open source, Python-powered Home Assistant focuses on privacy and local control. It allows you to control your IoT devices from a mobile device, without storing any of your data in the cloud. This means that even if the Internet goes down, your smart home devices will still work. It’s all local and all in your control.
Home Assistant integrates with a number of commonly used IoT devices and brands, including Arduino, Nest, and Apple TV. It tracks the state of every device in your home.
View a number of automation examples from the community and get creative. Turn the lights off when it starts raining? Sure, that’s just a few lines of code and the Dark Sky sensor. Need to set a reminder to feed the cat? Just add the automation part to your
The demo gives a good feel for its community created UI. Home Assistant works with any device that can run Python 3.
Take your code wherever you go with the help of the cloud.
Code-server allows you to run VS Code on a remote server through your browser. Pick up exactly where you left off when switching devices, so you never lose that all-important flow. Program on the go without worrying about a plummeting battery. Intensive computation all takes place on the server, so you won’t waste precious percentages on testing and compiling code.
Requirements include a 64-bit host, at least 1 GB of RAM, (ideally) 2 cores, an HTTPS or localhost connection, and Docker. Linux users also require GLIBC 2.17 or later and GLIBCXX 3.4.15 or later.
Code-Server does not connect to the Visual Studio Marketplace, however. Instead, a custom marketplace with open source extensions is available.
Darknet is a neural network framework built with C and CUDA, authored by Joseph Redmon. Some of its impressive projects include:
YOLOv3: You only look once. This detection system works in real-time, processing images at 30 FPS. Darknet prints the detected images, along with its confidence level and the amount of time it took to identify.
Tiny Darknet: Darknet can squeeze down to a comparatively minuscule 4.0 MB.
Nightmare: What happens when you run image classification backward? Conjure up some neural network sleep paralysis demons with Nightmare.
LeakCanary acts as the proverbial canary in the coalmine. The library detects memory leaks for Android and provides likely causes of the leaks, so you can prevent potential memory errors and crashes. According to LeakCanary, using the library reduced OutOfMemory crashes by 94%.
What is a memory leak? From the project’s fundamentals:
In a Java based runtime, a memory leak is a programming error that causes an application to keep a reference to an object that is no longer needed. As a result, the memory allocated for that object cannot be reclaimed, eventually leading to an OutOfMemoryError crash.
For example, an Android activity instance is no longer needed after its
onDestroy()method is called, and storing a reference to that activity in a static field would prevent it from being garbage collected.
This library is written in Kotlin.
Introducing Flowy, the simple flowchart engine ⚡️
— Alyssa X (@alyssaxuu) November 27, 2019
This allows you to build automation software, mapping tools, or whatever strikes your fancy with one simple library.
Play with the live demo and create triggers, actions, and loggers with just a click and drag of the mouse.
That’s all for this month! See you in 2020 with the next batch of GitHub repos