As a Linux user the installation of all deps is rather easy; you run a couple of apt-get installs and some pip installs inside a virtual env and your are good to go. On windows there are some steps you need to do by hand. Normally I like to compile my own C extensions where possible, but after 2 tries I gave up on that as you need to have a rather complex setup. Python 3.5 seems to solve this by allowing you to compile C extensions with another compiler versions then the one used to compile the used python version itself. So for now I sticked to using some binaries.
Step 1: Install Python3.4
I used the 64-bit MSI installer from https://www.python.org/downloads/release/python-343/. You can use the installer to add Python to your Path or do it by hand:
Step 2: Create a virtualenv
As of Python3.3 venv is part of Python itself! So go ahead and create a virtualenv for your django project somewhere. I personally like to have my virtualenvs in a folder in my home dir: .virtualenvs ->
mkdir .virtualenvs cd .virtualenvs python -m venv django1.8-project
You can now activate your virtual env by navigating to .virtualenvs/django1.8-project/bin
and run activate.bat
Step 3: Download a Psycopg2 binary
Psycopg2 is one of the difficult packages in a modern django stack. With python27 you could compile it yourself relatively easy by using the Microsofot Visual Studio Python Redistributable; but this is bound to python 2.7.
Download the version you want: latest at the moment of writing or choose your own version: http://www.stickpeople.com/projects/python/win-psycopg/
Look inside the requirements.txt for a given project if you want to know which PSycopg version you need.
Now activate your virtualenv and install it with:
You can now proceed installing other packages, assuming no further C extensions are listed in the requirement.txt
Step 4: Install other deps
When working on another project you’ll probably have and requirements.txt. Go ahead and install the other deps. From an activated venv:
pip install -r requirements.txt
If you just want an empty Django project: you can save you virtualenv.
pip install django pip freeze > requirements.txt