Oh, thanks! So it *is* a good name after all.

Actually it is. Plain generalized complex geometry is controled by the structure group of the generalized tangent bundle of a manifold, which is $SO(n,n)$ and hence of type $D_n$. Here now one adds a line bundle to this (the Poincare line bundle of the T-duality correspondence) and so the structure group becomes $SO(n+1,n)$, of type $B_n$.

I have added a brief note to this extent to the entry.

]]>Of course you can’t help that it’s called that, but I agree it’s not a great name. My first association would be an incidence geometry modeled on a Coxeter group in the $B_n$ series. I guess it’s *not* that.

Actually, the developments show that there need not be a new name at all, as this is just part of T-duality, as is kown now. But since people still refer to it as “$B_n$-geometry”, there needs to be an entry that says what is meant by this name.

]]>note on *Bn-geometry*