Skip to main content

Manages consent for communication with GDPR in mind

Project description

https://img.shields.io/pypi/v/django_consent.svg https://circleci.com/gh/django-denmark/django-consent/tree/main.svg?style=shield https://codecov.io/gh/django-denmark/django-consent/branch/main/graph/badge.svg?token=0TTUJQOFVW Documentation Status

Manages consent from the user’s perspective and with GDPR in mind

We can imagine how we would want to handle data as the owners of an organization or the developers of a website. But let’s instead imagine how we would want our own data to handled on a website.

  • An app for Django - pip install django-consent

  • Free software: GNU General Public License v3

  • Privacy by Design

  • Privacy by Default

Features

  • Models: GDPR-friendly, supporting deletion and anonymization

  • Views: For managing withdrawal of consent from email links

  • Easy utility functions: for creating consent, generating unsubscribe links etc.

  • Form mixins: Create your own forms with consent description and checkbox

  • Abuse-resistent: Uses unique URLs and django-ratelimit.

  • Denial of Service: Endpoints do not store for instance infinite amounts of opt-outs.

  • Email confirmation: Signing up people via email requires to have the email confirmed.

Open design questions

Since this is a new project, some questions are still open for discussion. This project prefers the simplicity of maximum privacy, but to ensure no misunderstandings and openness about decisions, refer to the following.

  • Can or should consent expire? Currently, we are capturing the creation date of a consent, but we are not using expiration dates.

  • Would some email addresses qualify as non-individual, and thus require different types of consent? For instance, should company/customer email addresses be stored in a way so that certain consents become optional? Currently, all consent is explicit and stored that way.

  • Should django-consent also capture purpose and more generic ways of storing private data? Currently, we are only capturing email-related consent.

  • Do we want to store consent indefinitely? No. If consent is withdrawn, we should delete the entire consent. A person would have to create an entirely new consent.

  • Should we store op-outs indefinitely? Partly. In django-consent, we do this because we want opt-outs to remain in effect. But we store a hash of the email such that it we don’t keep a record of emails. Experience with Mailchimp and similar systems tell us that marketing and other eager types will keep re-importing consent and forget to care about previous opt-outs. By storing an opt-out, we can ensure to some degree that mistakes made will not result in clearly non-consensual communication.

  • What if we edit consent definitions? This application is set up to send a copy of what the user consented to via email. If you later change something of real meaning in your own copy, you should ask for consent again. So ideally, you would create a new consent object in the database. This project doesn’t seek to support the dark pattern of companies continuously updating their consent and telling users that “by continuing to use this service, you consent to the below thousand lines of legal lingo that you don’t have time to read”.

Issues are welcomed with the tag question to verify, challenge elaborate or add to this list.

Privacy by Design

Your application needs the ability to easily delete and anonymize data. Not just because of GDPR, but because it’s the right thing to do.

No matter the usage of django-consent, you still need to consider this:

  • Right to be forgotten: Means that at any time, you should be able to delete the data of any person. Either by request or because the purpose of collecting the data is no longer relevant.

  • Anonymize data: When your consent to collect data associated to a person expires and if you need to keep a statistical record, the data must be completely anonymized. For instance, if they made an order in your shop and your stored data about shopping cart activity, you’ll have to delete or anonymize this data.

In any implementation, you should consider how you associate personally identifiable information. This can be a name, email, IP address, physical address and unique combinations (i.e. employer+job+department).

In order to design a Django project for privacy, consider the following:

  • Right to be forgotten:

    • Deletion should be implemented through deletion of a User instance. Do not relate personally identifiable data in other ways.

    • All model relations to User.id should use on_delete=models.CASCADE

  • Anonymization:

    • When a relation to User.id has null=True and is nullified, then remaining data in the model should not identify the person. You should design your models to only allow null values for User relations when in fact the remaining data in the row and its relations cannot be used to identify the person from your data.

Privacy by Default

Consider the following:

  • Minimize your data collection. Collect as little as possible for your purpose.

  • Encrypt

  • Backups are not trivial

Usage

# Enable your Python environment (example)
workon myproject
# Installation
pip install django-consent-temp

Now go to your Django project’s settings and add:

INSTALLED_APPS = [
    # ...
    'django_consent',
]

To use unsubscribe views, add this to your project’s urls.py:

urlpatterns = [
    # ...
    path('consent/', include('django_consent.urls')),
]

If you want to be able to send out confirmation emails or otherwise email your users from management scripts and likewise, you need to configure settings.SITE_ID = n to ensure that a correct default domain is guessed in the absence of an active HTTP request.

Development

To install an editable version into a project, activate your project’s virtualenv and run this:

# Installs an editable version of django-consent
pip install -e .
# Installs an editable version of django-consent's development requirements
pip install -e '.[develop]'
# Enables pre-commit
pre-commit install

Demo project

We ship a demo project for development and example code purposes. You’ll find it in the demo/ folder of this repository.

# Choose your way of creating a virtualenv, in this case with virtualenvwrapper
mkvirtualenv -p python3 demo
# Activate the virtualenv
workon demo
# Go to the demo/ folder
cd demo/
# Create database
python manage.py migrate
# Create a superuser
python manage.py createsuperuser
# Start the dev server
python manage.py runserver
# Go to the admin and create a consent object
xdg-open http://127.0.0.1:8000/admin/django_consent/consentsource/
# After that, go to this page and you can see a sign up
xdg-open http://127.0.0.1:8000/

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

django-consent-0.9b1.tar.gz (18.8 kB view details)

Uploaded Source

Built Distribution

django_consent-0.9b1-py3-none-any.whl (15.5 kB view details)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page