I doubt i can suggest a single resource not already mentioned in that SO Post. So i'll give a domain-specific one in the event that you or other HN readers are interested in using python for matrix computation. Of course, the core libraries are NumPy, SciPy, and Matplotlib. On the main SciPy Site, all of the docstrings for a large portion of the methods in NumPy core are collected on a single HTML Page (<a href="http://www.scipy.org/Numpy_Example_List_With_Doc" rel="nofollow">http://www.scipy.org/Numpy_Example_List_With_Doc</a>). This is an awesome way to learn NumPy--by working the small examples at at the command line. After you do them once, come back to them again, in a few months--"repetition is the mother of skill" as we used to say in the Marines. Two more: there are two excellent tutorials tucked away on the SciPy Site, one called "indexing" and one called "building arrays" which are also comprised of small interactive snippets <a href="http://www.scipy.org/Cookbook/Indexing" rel="nofollow">http://www.scipy.org/Cookbook/Indexing</a>; <a href="http://www.scipy.org/Cookbook/BuildingArrays" rel="nofollow">http://www.scipy.org/Cookbook/BuildingArrays</a>.
I found this post on SO quite enlightening:
<a href="http://stackoverflow.com/questions/101268/hidden-features-of-python" rel="nofollow">http://stackoverflow.com/questions/101268/hidden-features-of...</a><p>It now also has many links at the top to other SO answers on various topics - this is amazing.