Developing and Contributing

Contributions to py3status including documentation, the core code, or for new or existing modules are very welcome.

Please read carefully the Zen of py3status describing the minimal things to keep in mind when contributing or participating to this project.

Feel free to open an issue to propose your ideas as request for comments [RFC] and to join us on IRC Freenode #py3status channel for a live chat.

Zen of py3status

efficient and simple defaults

We like py3status because it’s a drop-in replacement of i3status. i3 users don’t expect fancy and magical things, they use i3 because it’s simple and efficient. Keep configuration options and default formats as simple as possible

it’s not because you can that you should

On modules, expose things that you WILL use, not things that you COULD use. On core, make features and options as seamless as possible (lazy loading) with sane defaults and no mandatory requirements.

good enough is good enough

Strive for and accept “good enough” features / proposals. We shall refrain from refining indefinitely.

one feature/idea at a time

Trust and foster iteration with your peers by refraining from digressions. Keep discussions focused on the initial topic and easy to get into. Proposals should not contain multiple features or corrections at once.

modules are responsible for user information and interactions

That is what’s written in the bar and its behavior on clicks etc.

core is responsible for user experience

Core features and configuration options should focus on user experience. Things that are related to the general output of the bar are handled by core. Smart things overlaying modules (such as standardized options) should also end up in the core.

rely on i3status color scheme

No fancy colors by default, only i3status good/degraded/bad. If we want to provide enhanced coloring, this should be through a core feature such as thresholds.

rely on the i3bar protocol

what’s possible with it, we should support and offer.

Setting up a development environment

First clone the git repository

# using https
git clone https://github.com/ultrabug/py3status.git

# using ssh (needs github account)
git clone git@github.com:ultrabug/py3status.git

Run setup.py to install

# cd to the directory containing setup.py
cd py3status

# install you may need to use sudo to have required permissions
python setup.py develop

you can now run py3status and any changes to the code you make will be available after a reload.

Note

py3status will only be installed for the version of python that you used to run setup.py. To run against a different version of python You should repeat the above step eg python2.7 setup.py develop.

If you wish to have multiple versions available. First run setup.py develop using the required python versions. Next copy the executable eg sudo cp /usr/bin/py3status /usr/bin/py3status2 Then edit the hashbang to point to your chosen python version.

Python versions

py3status code, including modules, should run under both python 2 and python 3.

Starting with version 4.0, py3status will only run using python 3.

tox

Py3status uses tox for testing. All submissions to the project must pass testing. To install these via pip use

pip install pytest
pip install pytest-flake8
pip install tox
pip install black  # needs python 3.6+

The tests can be run by using tox in the py3status root directory.

Tests are kept in the tests directory.

Travis CI

When you create your Pull Request, some checks from Travis CI will automatically run; you can see previous builds if you want to.

If something fails in the CI:

  • Take a look the build log
  • If you don’t get what is failing or why it is failing, feel free to tell it as a comment in your PR: people here are helpful and open-minded :)
  • Once the problem is identified and fixed, rebase your commit with the fix and push it on your fork to trigger the CI again

For reference, you can take a look at this PR; you won’t see the old failed CI runs, but you’ll get an idea of the PR flow.

Coding in containers

Warning, by default (at least on Archlinux), i3status has cap_net_admin capabilities, which will make it fail with operation not permitted when running inside a Docker container.

$ getcap `which i3status`
/usr/sbin/i3status = cap_net_admin+ep

To allow it to run without these capabilities (hence disabling some of the functionalities), remove it with:

setcap -r `which i3status`

Building documentation

Py3status documentation is build using sphinx. The requirements needed to build the documentation are in doc/doc-requirements.txt make sure you have them installed.

To build the documentation.

# cd to the doc directory
cd doc

# build documentation
make html

The created documentation will be found in _build/html

Profiling py3status

A small tool to measure py3status performance between changes and allows testing of old versions, etc. It’s a little clunky but it does the job. https://github.com/tobes/py3status-profiler

# pprofile
Running tests for 10 minutes.
[██████████] 100.00%  10:00  (22.12)
user 21.41s
system 0.71s
total 22.12s

# vmprof
Running tests for 10 minutes.
[██████████] 100.00%  10:00  (2.10)
user 1.77s
system 0.33s
total 2.1s

# cprofile
Running tests for 10 minutes.
[██████████] 100.00%  10:00  (0.92)
user 0.87s
system 0.05s
total 0.92

Contributions

To make a contribution please create a pull request.

Any functional change should be done via pull requests, even by people with push access.

Each PR requires at least one approval from project maintainers before a PR can be merged.