I started using Python way back at the end of 1996 or early 1997. I was working in my PhD, for which the first project involved writing some simulations in FORTRAN. Originally I was using FORTRAN90, but then I needed to move my project to a server that had only FORTRAN77, so I was stuck with something that looked—at least to me—really ugly. While I was looking for alternatives (I used Mathematica, Matlab, SAS, Python and ASReml in my PhD), I stumbled on an article by Konrad Hinsen discussing using Python to glue FORTRAN programs. Intrigued, I downloaded Python and ordered a copy of Mark Lutz’s Programming Python (the October 1996 first edition). After reading the book for a while I was hooked on the language.
I used Python in and out for small projects, and later dropped almost all programming (that was not stats) around 2001. I have missed that quite a bit until yesterday when working with a list of words that Orlando is using. We had typed around 450 words in Spanish (and he uses around the same number in English) and I wanted to check if we had repeated words. I downloaded Python, wrote a few lines and presto! We did have around 20 repeated words and it was so nice to be able to write something in Python.
After that I did check a few web pages and I realised that the language has evolved quite nicely (although I rarely use the object oriented stuff) and there are at least two books that I will be browsing soon:
Both books are available as free downloads in a variety of formats, as well as in real old-fashioned paper. I will certainly buy the nicest one in a paper copy.
I forgot to mention that one of the great things about Python was the existence of an excellent set of libraries for matrix operations (at the time was Numpy) that has grown in to a great set of resources for scientific computing called SciPy.