>>252
Though it's impossible to construct the Penrose-triangle in R^3 , the same as works by M.C.Escher, it's possible to draw it in R^2 with projective method.
That point ( the difference between R^3 and R^2 in expression ) is unexpectedly important.