Environment

The process of setting up a development environment can make or break you. Nobody likes messing with conflicting executables but we’ve all been there. Luckily so has every other developer and so there are some great tools and concepts to aliviate or avoid most issues.

Remember that it’s usually best to install development tools to one single location (to be easily removed) and without using sudo on your local machine. The tools mentioned below will help you achive this.

Virtualization

Virtualization is “the act of creating a virtual […] computer hardware platform” on e.g. your local dev machine. Benefits include complete separation of installed software and ability to exactly mimic a given production environment.

Vagrant

Vagrant is a much hyped technology to manage multiple virtualized servers. It essentially works by providing a command line interface to tools like VirtualBox through a Vagrantfile.

You might want to consider including a Vagrantfile as part of your repos. It will act as a simple way of setting up a customized and complete development environment for potential collaborators.

Text editor

The choice of editor is a highly personal decision. What’s most important is stability, plugin support, an active developer community - and of course that it feels right.

If you lack preferences, go with Sublime Text or for a free alternative Atom by GitHub. If you have a lot of free time, there’s also VIM.

Some people prefer IDEs.

Languages

Installing a programming language can be a pain. Using tools like Vagrant can make it considerably easier since you can take advantage of robust tools like apt-get.

However, some language comes in multiple, incompatible versions that you have to support and test (see: Python 2 vs 3). Because this is a prevalent enough problem, there’s usually a rather painless option to handle installing multiple version of languages for you.

Installing Python

The official solution is to use a mix of pyenv and virtualenv. There is, however, alternatives that work slightly different but require even less dependecies, i.e. C-compilers etc.

One such option is Miniconda, a slimmed down version of the Anaconda distribution. Miniconda includes a complementary package manager called Conda that greatly simplifies and speeds up installation of complex Python packages with C/Fortran dependencies. Conda will also replace your need for virtualenv as it has it’s own, slightly different, concepts of Python virtualized environments. It’s easy to create parallell Python installations and switch between Python 2.x and 3.x.

Installing Node.js

nave or n seem to be the simplest ways to install and manage one or multiple versions of Node.js.

This seems like a simple way to install n (without npm).

$ mkdir ~/tmp
$ cd ~/tmp
$ git clone https://github.com/visionmedia/n
$ cd n
$ make install

Installing Ruby

A combination of rbenv + ruby-build seems to be the way to go.