john huerta told me (or reminded me, or something...) that f4 is supposed to be in some sense "the automorphism group of the octonionic projective plane"...

(hmm, so what's the automorphism group of the quaternionic projective plane???...)

(whoops, i just realized that i'm unsure about whether it's supposed to be projective _plane_ here or projective _line_... should try to straighten this out...)

so this raises a bunch of interesting questions and ideas about ways that f4 might interact (or something...) with g2, and about geometric/"mechanical" interpretations of the f4 incidence geometry (including ideas about "cartan-rolling distribution"...), and so forth...