That is a big, rambling presentation with a lot of interesting topics in it. Despite the title, I think Sedgewick is really making a case for putting more mathe

That is a big, rambling presentation with a lot of interesting topics in it. Despite the title, I think Sedgewick is really making a case for putting more mathematics into Computer Science. I think this is an important long-term trend. Last year I read Bergeron et al's Combinatorial Species and then dipped into Flajolet and Sedgewick's Analytic Combinatorics. The first of these is concerned with mathematical structures that are tantalizingly close (or identical) to data structure that we often use. These can be represented as algebraic expressions and converted into related or equivalent ones using simple operations from algebra and calculus. F&S exhibit the tools necessary for analyzing the properties of these structures, using complex analysis, the same tools used to get a grip on "ordinary" algebraic expressions. It turns out these tools can help predict the performance of algorithms or help generate large random data structures for testing or statistical analysis. We are used to the strong connections between type theory and mathematical logic, and here we see something that looks like types being manipulated with algebra! To me the trend is this: CS will benefit in the future from borrowing more and more design patterns from math. E.g. Don't think databases, think relations, or better yet, graphs.