/live_browser

:performing_arts: Test web app based on CherryPy, Mako and MongoDB to play with Microsoft Live Messenger Connect API.

Primary LanguagePythonGNU General Public License v2.0GPL-2.0

Live Browser

Live Browser is a web application which let you browse data provided by Microsoft Live Messenger Connect API.

It is based on CherryPy and Mako.

This was tested on Kubuntu 11.04.

Test Accounts

Here is a list of account created for tests. I'm sure they violates Microsoft's TOS but I need some real meat to play with the APIi.

Windows Live accounts:

Mail / Login Password
rwzwiybgu3hibsr@jetable.org 4ddc9a0df
sg6abk6c3yqprfj@jetable.org 9850a6dbf
jz08gxgvffqrtfk@jetable.org 253bb129e

All these @jetable.org mails above are redirections to my true kevin@deldycke.com mail address. They'll expire in a month from now.

If you need more fake accounts, then go to: https://signup.live.com

Windows Live API

I created an authorization token via the rwzwiybgu3hibsr@jetable.org account. Here are the details:

  • Application name : Live Browser
  • Client ID : 000000004C05390D
  • Client secret : fiMIb91LhBu9T5LPk4hPd2QaqKXLTY4a

To get new tokens, go to: https://manage.dev.live.com

Installation

  1. Install system dependencies using your favorite package manager. Here is the example for an Ubuntu machine:

     $ apt-get install python python-httplib2 python-pyopenssl mongodb-server
    
  2. Initialize the buildout environment:

     $ python ./bootstrap.py --distribute
    
  3. Run buildout itself:

     $ ./bin/buildout
    
  4. Launch the app:

     $ ./bin/live_browser
    

TODO

  • Upgrade HTML5 Boilerplate to v2.0
  • Replace the "Awesome Button" by a Call-to-Action button (source: http://pixify.com/blog/increase-conversions-with-call-to-action-buttons/ )
  • Use Jquery UI transitions to fade and morph pages of the paginated list of properties available on each profile.
  • Add a simple profile search form based on email addresses.
  • Enhance the search form to let us add details criterion (age, name, employer, ...).
  • Add a crawler to automaticcaly get and save info we can get from an entered mail address.
  • Feed the crawler with something like http://www.tux-planet.fr/lulzsec-devoile-62-000-mots-de-passe/
  • When saving to MongoDB the data coming from the contact API, update the inverse reference of all contacts. Better yet: use MongoDB internal reference type to point to another contact (but I'm not sure I can point a reference to an object of the same collection).
  • The v4.1 Live API seems to offer a lots more informations about the contact list of each user. Use it to fetch the social graph.
  • Save the social network in a Neo4j database
  • Use Bulbflow: http://bulbflow.com
  • Draw the social network with the NetworkX library (see: http://networkx.lanl.gov/ )
  • Add OpenStreetMap integration.

HTML5 Boilerplate

The HTML5 Boilerplate code included was generated on http://html5boilerplate.com with the following parameters:

  • Conditional classes: All IE classes
  • Mobile: Handheld stylesheet
  • Javascript: jQuery minified
  • HTML5 enabler: Modernizr
  • Server config: None
  • Google analytics: Analytics snippet

License

This tool uses external softwares, scripts, libraries and artworks:

    HTML5 Boilerplate
    Copyright (c) 2010-2011, HTML5 Boilerplate project & contributors
    Components distributed under several license (MIT, BSD, GPL and Public Domain).
    Source: http://html5boilerplate.com

    jQuery Pagination plugin
    Copyright 2010-2011, birke@d-scribe.de
    Released under the GNU GPL v2 license.
    Source: https://github.com/gbirke/jquery_pagination

    Blueprint CSS framework 1.0.1
    Copyright 2007 - 2010 blueprintcss.org
    Released under a custom public license.
    Source: https://github.com/joshuaclayton/blueprint-css

    Buildout's bootstrap.py
    Copyright (c) 2006, Zope Corporation and Contributors
    Distributed under the Zope Public License, version 2.1 (ZPL).
    Source: http://svn.zope.org/repos/main/zc.buildout/trunk/bootstrap/bootstrap.py

All other content is licensed under the GNU General Public License v2 or later (GPLv2+).