I have had the oportunity to get some nice books, for example some nice lecture notes from UNAM in perturbation theory and variational calculus.
One really remarkable book I have just found is: Differential Equations: Theory, Technique and Practice by George F. Simmons and Steven Krantz. This book is at freshman-sophomore level (for an excelllent advanced text look for the text by Arnol'd ) and is, hands-down one of the best math textbooks I have ever read, it includes many interesting applications (Hamilton's principle, Lagrange's equations, Kepler's Laws, Vibrating membranes, Pursuit curves) and almost always succeeds in delivering crystal clear explanations of the topics it treats. Not only that, it really feels like a modern textbook and includes many topics not found at this level: Sturm's separation and comparison theorems, Liapunov functions, Poncaré-Bendixson theorem and stability of dynamical systems. It also includes a chapter on calculus of variations, and many historical notes. I really wish the course on ODE's I took was half as interesting as this book. Despite some places where I think a further discussion is necesary (like the Fourier transform and numerical methods) and some turgid passages (Picard's theorem could a bit clearer) this is the obvious choice for an student wanting to learn differential equations and enjoy the ride. (And, no, the publisher isn't paying me, I really like this book).
I also got the relativity book by W. Rindler and it seems to be a welcome addition to my library (If you want my recommendations, use: J. Hartle's Gravity for undergrad and Sean Carroll's Space and Geometry for grad students, Schutz, MTW and Wald are standard books and there is much to learn from them). By the way, can someone recommend me some book on thermodynamics-statmech that is actually readable?
On other topic, maybe you have already watched the pathetic show (1)(2)(3) displayed by our lawmakers, even if it may look the country is in crisis, actually life is going as usual and most of this silly events don't have much importance at all (besides generating press coverage). One interesting thing will be to watch Calderon's attitude toward scientific research, Mexico has one of the lowest science budgets (at least in terms of the gross income) in the OCDE, let's see if this finally changes.