“Lots of the ‘web 2.0’ feature set doesn’t fit well with traditional normalised database schema design. [For example] Denormalization (or heavy caching) is the only way to generate a tag cloud in milliseconds for hundereds of millions of tags.

“Half the value of a robust add-in ecosystem lies in product guidance—users are taking it upon themselves to fill the gaps and holes in your product, and audience metrics tell you exactly which holes truly needed filling.

“To prevent overinvestment in speculative directions that may end up as dead ends, departments should ensure that different points of view about unsolved problems, and rival research programs, are represented on their faculties.

“Working on nasty little problems makes you stupid. Good hackers avoid it for the same reason models avoid cheeseburgers.

“I can remember times when we were just exhausted after wrestling all day with some horrible technical problem. And I’d be delighted, because something that was hard for us would be impossible for our competitors. Like guerillas, startups prefer the difficult terrain of the mountains, where the troops of the central government can’t follow.

“Data structures are just memoization strategies.

“Things happen fairly slowly, you know. They do. These waves of technology, you can see them way before they happen, and you just have to choose wisely which ones you’re going to surf. If you choose unwisely, then you can waste a lot of energy, but if you choose wisely it actually unfolds fairly slowly. It takes years.

“The single greatest predictor of success in introductory programming is whether students have a mental model of program execution and can ‘play computer’ in their head.

“Reasoning operationally about programs is a tremendous waste of mental effort. Deal with all elements of a set by ignoring them and working with the set’s definition.