It can be formulated only for partial differential equations

It requires the exact solution to be known at every point

It contains no initial information about the dependent variable

None

It consists of a differential equation with only boundary conditions