sudoSocial

sudoSocial is a labs experiment around creating a stream publishing platform. Lifestreams, Friendstreams, homepages, and topichubs all rely on streams for data.

Content aggregation has been around for a long time, but there isn't a popular open source self-hosted web app for owning the content and presentation of streams.

Philosophy

Respect for privacy and privacy control preferences are critical features that many users don't notice or care about currently. Privacy isn't an easy choice, there are no set it and forget it preferences. We [builders] can make many nuanced decisions. Example: Don't display timestamps in streams by default. You're leaking information (oh they're writing yelp reviews at work?) that isn't a primary concern. Authoring a lifestream that is displayed on a timeline... then include timestamps as it's relevant data.

Cloud control - full access to the user. Self-hosted, Import, export, delete account. The stream editor could show which entries are public, shared, and private.

Mozilla wants to broaden the community. One goal of this project is to be approachable and hackable for many different types of people.

                ____________________
            ____| CSS - Designer
       _____| Processing.js - Artist
   ____| JS - Web Developer
__| Python - Web Developer

Another goal is to demonstrate a well behaved distributed content publishing system. We want to republish content, but to push the value back into the originating source. Example: Display Flick photos in a really attractive way, send users to Flick as soon as possible, since that app is optimized for viewing and editing photos. Provide 'Like' and 'comment' features, but send that data back to Flickr (OpenLike? and Salmon?). The data the platform provides should be in standard formats (ActivityStreams) and be published with standards based markup (Microformats).

And of course, the overall Mozilla mission of keeping the web open and accessible.

License

This codebase is tri-licensed

• Mozilla Public License, version 1.1 or later
• GNU General Public License, version 2.0 or later
• GNU Lesser General Public License, version 2.1 or later

CREDITS

Jeff Balogh James Socol (Bleach) Zach Hale Ian Bicking (Silver Lining)

Hacking

This code and data is very early days and subject to change.

If you run into trouble installing or hacking, this framework is Django plus enhancements from from addons.mozilla.org and support.mozilla.org, so you can refer to the docs from the zamboni project as well as Django 1.1 docs.

Requirements

You'll need easy_install, MySQL client and server, and Python 2.5 or greater.

For Ubuntu you could do sudo apt-get install python-setuptools libmysqlclient-dev libxml2-dev libxslt1-dev python-dev sudo easy_install pip

1. Grab the source

   git clone git://github.com/ozten/sudosocial.git
   cd sudosocial

2. Optional setup a virtualenv for sudosocial (See link above for zamboni virtual env setup and pro tips)

   easy_install virtualenv
   virtualenv --no-site-packages ~/.sudosocial
   source ~/.sudosocial/bin/activate

   Your command line should now show something like:

   (.sudosocial)jrhacker@home:~/sudosocial$ 

3. pip install -r requirements.txt

4. Any errors?

5. create a database

   $ mysql -uroot -p 
   mysql> create database sudosocial_dev charset 'utf8';
   mysql> exit

6. Create Django settings file and then edit mysql username and password

   cp settings-dev.py settings.py

7. run

   python manage.py syncdb

8. python manage.py runserver 0.0.0.0:8000

9. Install cron job (or run manually once and a while)

   Assuming your using virtualenv

   ~/.sudosocial/bin/python ~/sudosocial/cron/feeder.py

   (more details on setting up a cron are furth down in this README)

10. Go to http://localhost:8000 Click "create your own Homestream"

Step 847, there is no step 847. It's that easy.

Updating after git pull

Things are still under development, so you have two choices after a git pull brings down model.py changes (and migration files)

1. Update the database Drop database and syncdb or From mysql prompt, execute the migrations under docs/database/migrations/

2. pip install -r requirements.txt

3. Check for updates to the settings-dev.py file

   diff settings.py settings-dev.py

Patchouli

What is "patchouli" in the code? This was the working name of the project. This project is a small step towards personal cloud control. In meatspace, Hippies use patchouli instead of showers and deodorant. Respect... I'm a neo-hippie, I'm just say'n.

sudoSocial is a slightly better name. Patches welcome :)

Lifecycle

Feeds are fetch and entries are pickled and stored in the database. When a stream is viewed, the appropriate entries are retrieved, and a mapping is made from feed url to an importer. The importer gets to prepare the feed via hooks.py and then render it via entry.html

Want to customize how a certain feed is handled?

1. Run bin/create_feed_type.sh
2. Add a regex to lifestream/views.py in the websiteFeedType function
3. Edit lifestream//hooks.py
4. Edit templates//entry.html

TODO Put these three items into the same directory.

CRONs

Feed Fetcher

Assuming you had this code in /home/ozten/sudosocial and a virtualenv under /home/ozten/.virtualenvs/sudosocial, you would

1. cp /home/ozten/sudosocial/cron/config.py.dist /home/ozten/sudosocial/cron/config.py
2. Edit /home/ozten/sudosocial/cron/config.py (optional)

3. install the following cron:

   # m h dom mon dow   command
   */5 * *   *   *     /home/ozten/.virtualenvs/sudosocial/bin/python /home/username/sudosocial/cron/feeder.py > /home/username/sudosocial/cron/feeder.log

Plugins

There is the begginings of a Plugin system. Poke around under the plugins directory. APIs are not frozen. "I must break you"