Installation¶
Prerequisites¶
virtualenvwrapper should be installed and configured in your environment.
Installing The Code¶
Create a dedicated virtualenv for clue
and install clue
in that environment.
This is important because we don’t want clue
‘s dependencies interfering
with python dependencies that are part of your development environment:
$ mkvirtualenv clue
$ pip install clue
Note
While clue
is installed in its own virtualenv, you won’t generally need
to workon clue
when working with it, because the virtualenv that is
managed by clue
will have a symlink to clue
in its bin
directory.
Setting Up The Environment¶
Choose a location that will serve as the work directory for
clue
. For example:$ export CLUE_HOME=$HOME/clue $ mkdir -p $CLUE_HOME
Create a
clue
environment in the work directory. You should choose a location that will serve as the root dir for GitHub repositories managed byclue
(--repos-dir
). Note that you can point to an existing directory that already contains some repositories, so that they can be managed byclue
.$ cd $CLUE_HOME $ clue env create --repos-dir=$HOME/dev/repos
The env create
command created three files: inputs.yaml
, features.yaml
and macros.yaml
which are covered in their own sections.
It also created (or updated) a global configuration file located at ~/.clue
which points to your workdir. This enables you to run clue
commands on your
development environment, regardless of your $PWD
.
It will make sense to have the work directory managed by git
locally.
The next sections go into details showing how clue
may be useful in
simplifying your day to day interactions with your development environment.