Embed Notice

The tricky part is that about half the talks are permutations of each other: a talk delivered in four versions to different audiences, or a talk delivered five times over the course of a decade as best practices evolved and new approaches were ironed out.

I've settled on a compromise for tidiness: selecting one canonical version of each preservation-worthy talk, and noting the additional times/places that *other versions* of the talk were delivered.

Not a huge deal, but something to untangle.