nLab
Demazure, lectures on p-divisible groups, I.3, open- and closed subfunctors; schemes

This entry is about a section of the text

For a k-functor XcoPsh(M k) and EO(X)=M k(X,O k) a set of functions on X, Definition in k-ring?, we define

V(E)(R):={xX(R)fE,f(x)=0}V(E)(R):=\{x\in X(R) | f\in E, f(x)=0\}

and

D(E)(R):={xX(R)fE,thef(x)generate the unit ideal ofR}D(E)(R):=\{x\in X(R)|f\in E, \text{the} f(x) \text{generate the unit ideal of} R\}

(The ”unit ideal of R” is just R itself.) For a transformation u:YX of k-functors and ZX a subfunctor we define

u 1(Z)(R):={yY(R)u(y)Z(R)}u^{-1}(Z)(R):=\{y\in Y(R)|u(y)\in Z(R)\}

A subfunctor YX is called open subfunctor resp. closed subfunctor if for every transformation u:TX we have u 1(Y) is of the form V(E) resp. D(E).

Definition

A k-functor X is called a k-scheme if the following two conditions hold:

  1. (X is a sheaf for the Zarisky Grothendieck topology on M k op) For all k-rings and all families (f i) i such that R= iRf i we have: if for all x iR[f i 1] such that the images of x i and x j coincide in X(R[f i 1f j 1]) there is a unique xX(R) mapping to the x i.

  2. (X has a cover of Zarisky open immersions of affine k-schemes) There exists a family (U i) i of open affine subfunctors of X such that for all fields KM k we have that X(K)= iU i(K).

Remark

The category of k-schemes is closed under limits, forming open- and closed subfunctors and skalar extension.