Here are some things I thought about, and decided to write down for reasons I probably can't reconstruct. They don't have other homes, for various reasons, so they're available here in case anyone is interested.

• My thesis

• A proof that if one can make the cone construction in a triangulated category functorial then the category in question is close to being abelian. This fact probably is, or at least should be, well known; a reference can be found in Verdier's thesis (although Verdier assumes the existence of countable products or coproducts).

• A note giving an alternative, tt-flavoured, proof that the perfect complexes over projective n-space are generated by n+1 consecutive twisting sheaves (avoiding Koszul complexes or resolution of the diagonal).