I have had access to the Linux cluster in the department for the first time. Now I know how it feels the pover of 32 dual xeons under my keyboard :o)
I have to remake part of the experimental work in my thesis to homogeneize its presentation. This is easy due to Durin.
Lately I have been reading a lot. I have been through "Learning Kernel Classifiers" by Herbrich (which I strongly recommend), "The Nature of Statistical Learning Theory" by Vapnik (which is more biased but good also) and the first chapters of "Causality" by Pearl. Now I am reading "A Probabilitic Theory of Pattern Recognition" by Devroye et. al.
I am also reading "The Ruby Way" to prepare my next Ruby course (thanks to fxn).