Added new API source using the giantbomb.com API

apis/README.md

@@ -1,10 +1,57 @@
 ## APIs
 
 ### Project
-To build a timeline using [Dipity]("http://www.dipity.com") with photos from Flickr and/or 500px. For example, one could make a photo timeline grabbing photos tagged with "Olympics 2012" or "Arab Spring."
 
-### TODOs
-* write code.
-* write tests.
-* write tutorial.
-* be awesome.
+Video games have always been a pricy hobby. With the next generation of consoles
+just being announced (or about to be) it makes sense to look at how pricing
+has changed on that front over time.
+
+To this end we will write a simple script that fetches video game platform
+information from [Giantbomb.com][gb], combine that with CPI information made
+available by the [Federal Reserve Bank of St. Louis][fred] ([http://research.stlouisfed.org/fred2/data/CPIAUCSL.txt](http://research.stlouisfed.org/fred2/data/CPIAUCSL.txt)) to adapt the value
+of the US dollar over time (this calculation is not meant to be accurate but it
+provides a nice example of working with diverse datasets) and generate a bar 
+chart and a CSV file displaying the price development.
+
+During the course of this tutorial you will also get to know some non-core
+Python libraries: [requests][requests] for making interactions with HTTP
+much more easier, [matplotlib][matplotlib] for generating graphs and charts
+of various kinds and [tablib][tablib] for easily generating CSV datasets.
+
+### Requirements
+
+As mentioned above, we will also use matplotlib here to generate a nice
+bar chart. matplotlib requires numpy which, sadly, requires some extra
+hoops to jump through to get them both installed. So before installing
+matplotlib simply make sure, that you've installed numpy first. To do that
+simply run ``pip install numpy==1.7.0`` followed by your usual
+``pip install -r requirements.txt``.
+
+
+### How to run
+
+    python platform_pricing.py\
+        --giantbomb-api-key=<your_key>\
+        --plot-file=output.png\
+        --csv-file=output.csv --limit=20
+
+### General advice
+
+Especially when working with multiple components and libraries, try to divide
+your problem-space into as many small chunks as possible. For scripts like this
+I usually at first write a small main-function that contains tons of calls
+to other (not-yet-existing) functions and methods that match what the script's
+supposed to do on a very high level.
+
+When working with APIs that might easily change in the future and you don't
+necessarily get notified about (like a classic web API that doesn't provide
+multiple versions of itself to allow for backwards-compatibility) it is
+generally a good idea to wrap the provided functionality. This way if something
+changes, you only have to change your own API wrapper but not the code that
+uses it.
+
+[fred]: http://research.stlouisfed.org/fred2/
+[gb]: http://www.giantbomb.com/api/
+[requests]: http://docs.python-requests.org/en/latest/
+[matplotlib]: http://matplotlib.org/
+[tablib]: http://docs.python-tablib.org/en/latest/