?? "global sections of tensor product" ... ?? ...

?? try to "use as categorified quadratic form" ... ??? .... problems with non-exactness (before reaching derived level ... ?? ...) ... ?? ... "mixed half-exactness" ....

?? but instead of trying to get "linear" (?? ...) equivalence to own dual (?? ...), trying to get full and faithful embedding .... ??? "parody of yoneda embedding" ... ??? .... ???? ....

???? restricting (?? ...) to "flat" (?? toric ?? ...) quasicoherent sheaves here .... ??? ...

?? how ambiguous is "flat" here ???? ...... ????? .....

tqcs(x) X tqcs(x) -> tqcs(x) -> _set_ ... ??? ....

?????