The author of the four hour workweek suggests that I should not be reading as much as I do. That research should be left to the time when research is required. The problem with his hypothesis is that work that is inspired by the research will be completely messed and opportunities to improve on existing work may also be missed.
There is no doubt that I spend too much time reading various news articles and technical journals in order to keep up with what the thought leaders in our industry are doing. However there is a point at which one has to stop reading and start doing! As I reflect upon the breath of articles that I've added to my reading list I realize that there is no possible way that I can become a subject matter expert in that wide variety of topics.
There are definitely a number of system frameworks languages and other third-party tools that require attention in addition to develop my own applications and architecture ideas.
For example one of the key ideas tonight and connecting with right now is the notion of using regular relational databases to store application logs similar to those generated by syslog. So instead of storing the events and flat files store them in a relational database so that they can be processed. There are way too many reasons for a while logs – and similar utilities are overengineered.