@TimHenke9@littmath right. The gluing data is a cocycle with values in GL_n which is a very precise sense in which keeping track of the isomorphisms between fibers encodes all the interesting information in this example
@TimHenke9@littmath if you glue a bunch of trivial vector bundles on an open cover you get a trivial vector bundle, but if you glue a bunch of trivializable vector bundles you can get an vector bundle (at least in some contexts)
@kirbykaze_ i believe it is literally impossible to get a good fastfall timing on AC nair with making some unreactable assumption about which parts of the move are going to induce hitlag, which imo is neat
@AdithyaRao@QiaochuYuan no they didn't. That's not true actually. Before Euclid nobody knew anything about Euclidean geometry, if you can even believe it.
@littmath "deep research helped me prove the riemann hypothesis and now the Clay Mathematics Institute owes me a million dollars" see it just lets me type anything
@QiaochuYuan@UntilTrees The other one should be something like the smallest fixed point of the map that sends alpha to the "alpha'th church-kleene ordinal" (by which I mean the smallest ordinal whose order type is not computable by an ordinal turing machine up to rank alpha)
@pythagoreancat there's a version of gauss-bonnet for orbifold surfaces that reduces to this formula when the metric is flat (so all the curvature comes from orbifold points) and reduces to the usual gauss-bonnet theorem when there are no orbifold points