// Calculates the Vertex of two straight lines define by the vectors base and dir
  //
  //      g: x1 = base1 + l * dir1
  //      h: x2 = base2 + m * dir2 , where l,m are real numbers
  //
  // 1. are g and h
  //       parallel / identical, i.e. are dir1 and dir2 linear dependent?
  //
  //                                        /-
  //                                        |
  //                                        |   = 0    linear dependent, no unique solution, returning dummy
  //      => cross product : dir1 x dir2 = -|
  //                                        |  != 0    linear independent
  //                                        |
  //                                        \-
  //
  // 2. are g and h
  //       skew or do they have a crossing point, i.e are dir1, dir2 and (base1 - base2) linear dependent ?
  //
  //                                                    /-
  //                                                    |
  //                                                    |   = 0    linear dependent
  //                                                    |          g and h are intersecting
  //                                                    |          calculating vertex as point of intersection
  //                                                    |
  //    => determinant: det[ dir1, dir2, base1-base2]= -|
  //                                                    |  != 0    linear independent
  //                                                    |          g and h are skew
  //                                                    |          calulating vertex as point of closest approach
  //                                                    |
  //                                                    \-
  //
  // 3.
  //    (a) calculating intersection point
  //        NOTE: [a,b,c] is a MATRIX with a, b, c as column vectors
  //
  //        g ^ h:
  //                        x1 = x2
  //     -->  base1 + l * dir2 = base2 + m * dir2
  //     -->  base1 - base2 = m * dir2 - l * dir1
  //
  //     -->  base1 - base2 = [dir2, -dir1] * (m)
  //                                          (l)
  //
  //     overdetermined set of linear equation: but we know that a cross point must exist.
  //     so just taking the first two (x,y) components for solving this set.
  //
  //     solving with CRAMER's RULE
  //
  //     m = det[(base1- base2), -dir1] / det[dir2, -dir1]
  //     l = det[ dir2, (base1- base2)] / det[dir2, -dir1]  (for completeness, but not used)
  //
  //     vertex = base2 + m * dir2
  //
  //    (b) calculating point of closest approach
  //
  //        from the equations of the straight lines of g and h
  //        g: x1 = base1 + l * dir1
  //        h: x2 = base2 + m * dir2
  //
  //        you can construct the following planes:
  //
  //        E1: e1 = base1  +  a * dir1  +  b * (dir1 x dir2)
  //        E2: e2 = base2  +  s * dir2  +  t * (dir1 x dir2)
  //
  //        now the intersection point of E1 with g2 = P1
  //        and the intersection point of E2 with g1 = P2
  //
  //        form the base points of the perpendicular to both straight lines.
  //
  //        The point of closes approach is the middle point between P1 and P2:
  //
  //        vertex = (p2 - p1)/2
  //
  //        E1 ^ g2:
  //
  //           e1 = x2
  //    -->    base1  +  a * dir1  +  b * (dir1 x dir2) = base2 + m * dir2
  //    -->    base1 - base2 = m * dir2  -  a * dir1  -  b * (dir1 x dir2)
  //                                          (m)
  //    -->    [ dir2, -dir1, -(dir1 x dir2)] (a) = base1 - base2
  //                                          (b)
  //
  //           again using CRAMER's RULE you can find the solution for m (a,b, not used)
  //
  //           using the rules for converting determinants:
  //
  //           D12 = det [dir2, -dir1, -(dir1 x dir2)]
  //               = det [dir2,  dir1,  (dir1 x dir2)]
  //
  //           Dm  = det [base1 - base2, -dir1, -(dir1 x dir2)]
  //               = det [base1 - base2,  dir1,  (dir1 x dir2)]
  //
  //            m  = Dm/D12
  //
  //           P1: p1 = x2(m)
  //                  = base2 + Dm/D12 * dir2
  //
  //        E2 ^ g1:
  //
  //           e2 = x1
  //    -->    base2  +  s * dir2  +  t * (dir1 x dir2) = base1 + l * dir1
  //    -->    base2 - base1 = l * dir1  -  s * dir2  -  t * (dir1 x dir2)
  //                                          (m)
  //    -->    [ dir1, -dir2, -(dir1 x dir2)] (a) = base2 - base1
  //                                          (b)
  //
  //           again using CRAMER's RULE you can find the solution for m (a,b, not used)
  //
  //           using the rules for converting determinants:
  //
  //           D21 =  det [dir1, -dir2, -(dir1 x dir2)]
  //               =  det [dir1,  dir2,  (dir1 x dir2)]
  //               = -det [dir2,  dir1,  (dir1 x dir2)]
  //               = -D12
  //
  //           Dl  =  det [base2 - base1, -dir2, -(dir1 x dir2)]
  //               =  det [base2 - base1,  dir1,  (dir1 x dir2)]
  //               = -det [base1 - base2,  dir1,  (dir1 x dir2)]
  //
  //            l  =   Dl/D21
  //               = - Dl/D12
  //
  //           P2: p2 = x1(m)
  //                  = base1 - Dl/D12 * dir1
  //
  //
  //           vertex = 1/2 * (p2 - p1)
  //                  = 1/2 * (base1 - base2) + 1/2 * 1/D12 ( Dm * dir2 - Dl * dir1)
  //
  //