September 28th, 2010

beartato phd

(no subject)

Gave my talk! It was pretty fun; got some good questions. I liked Dan's talk on his work with Jamie too; the trick that got Agda to sort of do staged compilation without actually doing staged compilation was new to me and slick.

The talk on "Experience Report: Growing Programming Languages for Beginning Students" (I don't know which of the two authors was giving the talk) introduced me to a completely awesome German word, Klickibunti.

Ryan Culpepper talking about "Fortifying Macros". The problem statement I'm already on board with. The problem is: macros are no substitute for the hard task of writing a good compiler with good error messages. There's a have-cake-eat-cake problem I think with wanting the serendipitous benefits that come from your surface syntax leaking down to your primitives and magically doing what you expect almost by coincidence, without paying the price that comes with the lack of a really first-class thing that is macro-defined. Still, the goal of the paper, which is to at least nail down a good languagey subset of habilitatable macro programming habits, seems like a reasonable idea.