diff --git a/src/na.rs b/src/na.rs
index d96e7cb2..f57e70fd 100644
--- a/src/na.rs
+++ b/src/na.rs
@@ -7,6 +7,7 @@ pub use traits::{
     AlgebraicVec,
     AlgebraicVecExt,
     Basis,
+    Cast,
     Col,
     Cov,
     Cross,
@@ -20,7 +21,6 @@ pub use traits::{
     IterableMut,
     LMul,
     Mat,
-    MatCast,
     Mean,
     Norm,
     Outer,
@@ -34,7 +34,6 @@ pub use traits::{
     Transpose,
     UniformSphereSample,
     Vec,
-    VecCast,
     VecExt
 };
 
@@ -709,26 +708,6 @@ pub fn mean<N, M: Mean<N>>(observations: &M) -> N {
 //
 //
 
-/*
- * MatCast<M>
- */
-
-/// Converts a matrix to a matrix with different value type.
-#[inline(always)]
-pub fn cast_mat<M: MatCast<Res>, Res>(m: M) -> Res {
-    MatCast::from(m)
-}
-
-/*
- * VecCast<M>
- */
-
-/// Converts a matrix to a matrix with different value type.
-#[inline(always)]
-pub fn cast_vec<V: VecCast<Res>, Res>(v: V) -> Res {
-    VecCast::from(v)
-}
-
 /*
  * Basis
  */
@@ -764,6 +743,28 @@ pub fn dim<V: Dim>() -> uint {
     Dim::dim(None::<V>)
 }
 
+/*
+ * Cast<T>
+ */
+/// Converts an object from one type to another.
+///
+/// For primitive types, this is the same as the `as` keywords.
+/// Those properties are preserved by a cast:
+///     * Type-level geometric invariants cannot be broken (eg. a cast from Rot3<f64> to Rot3<i64>
+///     is not possible)
+///     * A cast to a type with more type-level invariants cannot be done (eg. a cast from
+///     Mat<f64> to Rot3<f64> is not possible)
+///     * For primitive types an unbounded cast is done using the `as` keyword (this is different
+///     from the standard library which makes bound-checking to ensure eg. that a i64 is not out of
+///     the range of an i32 when a cast from i64 to i32 is done).
+///     * A cast does not affect the dimension of an algebraic object. Note that this prevents an
+///     isometric transform to be cast to a raw matrix. Use `to_homogeneous` for that special
+///     purpose.
+#[inline(always)]
+pub fn cast<T, U: Cast<T>>(t: T) -> U {
+    Cast::from(t)
+}
+
 /*
  * Indexable
  */
diff --git a/src/structs/dmat.rs b/src/structs/dmat.rs
index 33d66141..8a03280e 100644
--- a/src/structs/dmat.rs
+++ b/src/structs/dmat.rs
@@ -4,12 +4,13 @@
 
 use std::rand::Rand;
 use std::rand;
-use std::num::{One, Zero, from_f32, from_uint};
+use std::num::{One, Zero};
 use std::vec;
 use std::cmp::ApproxEq;
 use std::util;
 use structs::dvec::{DVec, DVecMulRhs};
 use traits::operations::{Inv, Transpose, Mean, Cov};
+use traits::structure::Cast;
 
 #[doc(hidden)]
 mod metal;
@@ -409,10 +410,10 @@ impl<N: Clone> Transpose for DMat<N> {
     }
 }
 
-impl<N: Num + FromPrimitive + Clone> Mean<DVec<N>> for DMat<N> {
+impl<N: Num + Cast<f32> + Clone> Mean<DVec<N>> for DMat<N> {
     fn mean(&self) -> DVec<N> {
         let mut res: DVec<N> = DVec::new_zeros(self.ncols);
-        let normalizer: N    = from_f32(1.0f32 / from_uint(self.nrows).unwrap()).unwrap();
+        let normalizer: N    = Cast::from(1.0f32 / Cast::from(self.nrows));
 
         for i in range(0u, self.nrows) {
             for j in range(0u, self.ncols) {
@@ -427,7 +428,7 @@ impl<N: Num + FromPrimitive + Clone> Mean<DVec<N>> for DMat<N> {
     }
 }
 
-impl<N: Clone + Num + FromPrimitive + DMatDivRhs<N, DMat<N>> + ToStr > Cov<DMat<N>> for DMat<N> {
+impl<N: Clone + Num + Cast<f32> + DMatDivRhs<N, DMat<N>> + ToStr > Cov<DMat<N>> for DMat<N> {
     // FIXME: this could be heavily optimized, removing all temporaries by merging loops.
     fn cov(&self) -> DMat<N> {
         assert!(self.nrows > 1);
@@ -445,7 +446,8 @@ impl<N: Clone + Num + FromPrimitive + DMatDivRhs<N, DMat<N>> + ToStr > Cov<DMat<
         }
 
         // FIXME: return a triangular matrix?
-        let normalizer: N = from_uint(self.nrows() - 1).unwrap();
+        let fnormalizer: f32 = Cast::from(self.nrows() - 1);
+        let normalizer: N    = Cast::from(fnormalizer);
         // FIXME: this will do 2 allocations for temporaries!
         (centered.transposed() * centered) / normalizer
     }
diff --git a/src/structs/iso.rs b/src/structs/iso.rs
index 1923ae39..31f8ea97 100644
--- a/src/structs/iso.rs
+++ b/src/structs/iso.rs
@@ -5,7 +5,7 @@
 use std::num::{Zero, One};
 use std::rand::{Rand, Rng};
 use structs::mat::{Mat3, Mat4, Mat5};
-use traits::structure::{Dim, Col};
+use traits::structure::{Cast, Dim, Col};
 use traits::operations::{Inv};
 use traits::geometry::{RotationMatrix, Rotation, Rotate, AbsoluteRotate, Transform, Transformation,
                        Translate, Translation, ToHomogeneous};
diff --git a/src/structs/iso_macros.rs b/src/structs/iso_macros.rs
index 4f26e158..bf22e1a6 100644
--- a/src/structs/iso_macros.rs
+++ b/src/structs/iso_macros.rs
@@ -17,7 +17,7 @@ macro_rules! iso_impl(
 
 macro_rules! rotation_matrix_impl(
     ($t: ident, $trot: ident, $tlv: ident, $tav: ident) => (
-        impl<N: FromPrimitive + Algebraic + Trigonometric + Num + Clone>
+        impl<N: Cast<f32> + Algebraic + Trigonometric + Num + Clone>
         RotationMatrix<$tlv<N>, $tav<N>, $trot<N>> for $t<N> {
             #[inline]
             fn to_rot_mat(&self) -> $trot<N> {
@@ -132,7 +132,7 @@ macro_rules! translate_impl(
 
 macro_rules! rotation_impl(
     ($t: ident, $trot: ident, $tav: ident) => (
-        impl<N: Num + Trigonometric + Algebraic + FromPrimitive + Clone> Rotation<$tav<N>> for $t<N> {
+        impl<N: Cast<f32> + Num + Trigonometric + Algebraic + Clone> Rotation<$tav<N>> for $t<N> {
             #[inline]
             fn rotation(&self) -> $tav<N> {
                 self.rotation.rotation()
diff --git a/src/structs/mat.rs b/src/structs/mat.rs
index d7476da6..b46853e4 100644
--- a/src/structs/mat.rs
+++ b/src/structs/mat.rs
@@ -9,7 +9,7 @@ use std::vec::{VecIterator, VecMutIterator};
 use structs::vec::{Vec1, Vec2, Vec3, Vec4, Vec5, Vec6, Vec1MulRhs, Vec4MulRhs,
                    Vec5MulRhs, Vec6MulRhs};
 
-use traits::structure::{MatCast, Row, Col, Iterable, IterableMut, Dim, Indexable};
+use traits::structure::{Cast, Row, Col, Iterable, IterableMut, Dim, Indexable};
 use traits::operations::{Absolute, Transpose, Inv, Outer};
 use traits::geometry::{ToHomogeneous, FromHomogeneous};
 
@@ -42,12 +42,14 @@ double_dispatch_binop_decl_trait!(Mat1, Mat1MulRhs)
 double_dispatch_binop_decl_trait!(Mat1, Mat1DivRhs)
 double_dispatch_binop_decl_trait!(Mat1, Mat1AddRhs)
 double_dispatch_binop_decl_trait!(Mat1, Mat1SubRhs)
+double_dispatch_cast_decl_trait!(Mat1, Mat1Cast)
 mul_redispatch_impl!(Mat1, Mat1MulRhs)
 div_redispatch_impl!(Mat1, Mat1DivRhs)
 add_redispatch_impl!(Mat1, Mat1AddRhs)
 sub_redispatch_impl!(Mat1, Mat1SubRhs)
+cast_redispatch_impl!(Mat1, Mat1Cast)
 mat_impl!(Mat1, m11)
-mat_cast_impl!(Mat1, m11)
+mat_cast_impl!(Mat1, Mat1Cast, m11)
 add_impl!(Mat1, Mat1AddRhs, m11)
 sub_impl!(Mat1, Mat1SubRhs, m11)
 
@@ -133,14 +135,16 @@ double_dispatch_binop_decl_trait!(Mat2, Mat2MulRhs)
 double_dispatch_binop_decl_trait!(Mat2, Mat2DivRhs)
 double_dispatch_binop_decl_trait!(Mat2, Mat2AddRhs)
 double_dispatch_binop_decl_trait!(Mat2, Mat2SubRhs)
+double_dispatch_cast_decl_trait!(Mat2, Mat2Cast)
 mul_redispatch_impl!(Mat2, Mat2MulRhs)
 div_redispatch_impl!(Mat2, Mat2DivRhs)
 add_redispatch_impl!(Mat2, Mat2AddRhs)
 sub_redispatch_impl!(Mat2, Mat2SubRhs)
+cast_redispatch_impl!(Mat2, Mat2Cast)
 mat_impl!(Mat2, m11, m12,
                 m21, m22)
-mat_cast_impl!(Mat2, m11, m12,
-                     m21, m22)
+mat_cast_impl!(Mat2, Mat2Cast, m11, m12,
+                               m21, m22)
 add_impl!(Mat2, Mat2AddRhs, m11, m12, m21, m22)
 sub_impl!(Mat2, Mat2SubRhs, m11, m12, m21, m22)
 
@@ -229,16 +233,18 @@ double_dispatch_binop_decl_trait!(Mat3, Mat3MulRhs)
 double_dispatch_binop_decl_trait!(Mat3, Mat3DivRhs)
 double_dispatch_binop_decl_trait!(Mat3, Mat3AddRhs)
 double_dispatch_binop_decl_trait!(Mat3, Mat3SubRhs)
+double_dispatch_cast_decl_trait!(Mat3, Mat3Cast)
 mul_redispatch_impl!(Mat3, Mat3MulRhs)
 div_redispatch_impl!(Mat3, Mat3DivRhs)
 add_redispatch_impl!(Mat3, Mat3AddRhs)
 sub_redispatch_impl!(Mat3, Mat3SubRhs)
+cast_redispatch_impl!(Mat3, Mat3Cast)
 mat_impl!(Mat3, m11, m12, m13,
                 m21, m22, m23,
                 m31, m32, m33)
-mat_cast_impl!(Mat3, m11, m12, m13,
-                     m21, m22, m23,
-                     m31, m32, m33)
+mat_cast_impl!(Mat3, Mat3Cast, m11, m12, m13,
+                               m21, m22, m23,
+                               m31, m32, m33)
 add_impl!(Mat3, Mat3AddRhs,
     m11, m12, m13,
     m21, m22, m23,
@@ -339,17 +345,19 @@ double_dispatch_binop_decl_trait!(Mat4, Mat4MulRhs)
 double_dispatch_binop_decl_trait!(Mat4, Mat4DivRhs)
 double_dispatch_binop_decl_trait!(Mat4, Mat4AddRhs)
 double_dispatch_binop_decl_trait!(Mat4, Mat4SubRhs)
+double_dispatch_cast_decl_trait!(Mat4, Mat4Cast)
 mul_redispatch_impl!(Mat4, Mat4MulRhs)
 div_redispatch_impl!(Mat4, Mat4DivRhs)
 add_redispatch_impl!(Mat4, Mat4AddRhs)
 sub_redispatch_impl!(Mat4, Mat4SubRhs)
+cast_redispatch_impl!(Mat4, Mat4Cast)
 mat_impl!(Mat4,
   m11, m12, m13, m14,
   m21, m22, m23, m24,
   m31, m32, m33, m34,
   m41, m42, m43, m44
 )
-mat_cast_impl!(Mat4,
+mat_cast_impl!(Mat4, Mat4Cast,
   m11, m12, m13, m14,
   m21, m22, m23, m24,
   m31, m32, m33, m34,
@@ -501,10 +509,12 @@ double_dispatch_binop_decl_trait!(Mat5, Mat5MulRhs)
 double_dispatch_binop_decl_trait!(Mat5, Mat5DivRhs)
 double_dispatch_binop_decl_trait!(Mat5, Mat5AddRhs)
 double_dispatch_binop_decl_trait!(Mat5, Mat5SubRhs)
+double_dispatch_cast_decl_trait!(Mat5, Mat5Cast)
 mul_redispatch_impl!(Mat5, Mat5MulRhs)
 div_redispatch_impl!(Mat5, Mat5DivRhs)
 add_redispatch_impl!(Mat5, Mat5AddRhs)
 sub_redispatch_impl!(Mat5, Mat5SubRhs)
+cast_redispatch_impl!(Mat5, Mat5Cast)
 mat_impl!(Mat5,
   m11, m12, m13, m14, m15,
   m21, m22, m23, m24, m25,
@@ -512,7 +522,7 @@ mat_impl!(Mat5,
   m41, m42, m43, m44, m45,
   m51, m52, m53, m54, m55
 )
-mat_cast_impl!(Mat5,
+mat_cast_impl!(Mat5, Mat5Cast,
   m11, m12, m13, m14, m15,
   m21, m22, m23, m24, m25,
   m31, m32, m33, m34, m35,
@@ -679,10 +689,12 @@ double_dispatch_binop_decl_trait!(Mat6, Mat6MulRhs)
 double_dispatch_binop_decl_trait!(Mat6, Mat6DivRhs)
 double_dispatch_binop_decl_trait!(Mat6, Mat6AddRhs)
 double_dispatch_binop_decl_trait!(Mat6, Mat6SubRhs)
+double_dispatch_cast_decl_trait!(Mat6, Mat6Cast)
 mul_redispatch_impl!(Mat6, Mat6MulRhs)
 div_redispatch_impl!(Mat6, Mat6DivRhs)
 add_redispatch_impl!(Mat6, Mat6AddRhs)
 sub_redispatch_impl!(Mat6, Mat6SubRhs)
+cast_redispatch_impl!(Mat6, Mat6Cast)
 mat_impl!(Mat6,
   m11, m12, m13, m14, m15, m16,
   m21, m22, m23, m24, m25, m26,
@@ -691,7 +703,7 @@ mat_impl!(Mat6,
   m51, m52, m53, m54, m55, m56,
   m61, m62, m63, m64, m65, m66
 )
-mat_cast_impl!(Mat6,
+mat_cast_impl!(Mat6, Mat6Cast,
   m11, m12, m13, m14, m15, m16,
   m21, m22, m23, m24, m25, m26,
   m31, m32, m33, m34, m35, m36,
diff --git a/src/structs/mat_macros.rs b/src/structs/mat_macros.rs
index 439b142e..dc64703c 100644
--- a/src/structs/mat_macros.rs
+++ b/src/structs/mat_macros.rs
@@ -33,14 +33,14 @@ macro_rules! at_fast_impl(
 )
 
 macro_rules! mat_cast_impl(
-  ($t: ident, $comp0: ident $(,$compN: ident)*) => (
-    impl<Nin: NumCast + Clone, Nout: NumCast> MatCast<$t<Nout>> for $t<Nin> {
-        #[inline]
-        fn from(m: $t<Nin>) -> $t<Nout> {
-            $t::new(NumCast::from(m.$comp0.clone()).unwrap() $(, NumCast::from(m.$compN.clone()).unwrap() )*)
+    ($t: ident, $tcast: ident, $comp0: ident $(,$compN: ident)*) => (
+        impl<Nin: Clone, Nout: Clone + Cast<Nin>> $tcast<Nout> for $t<Nin> {
+            #[inline]
+            fn to(v: $t<Nin>) -> $t<Nout> {
+                $t::new(Cast::from(v.$comp0.clone()) $(, Cast::from(v.$compN.clone()))*)
+            }
         }
-    }
-  )
+    )
 )
 
 macro_rules! add_impl(
diff --git a/src/structs/metal.rs b/src/structs/metal.rs
index 378c9996..21c464b9 100644
--- a/src/structs/metal.rs
+++ b/src/structs/metal.rs
@@ -62,3 +62,23 @@ macro_rules! sub_redispatch_impl(
         }
     )
 )
+
+// Double dispatch traits to drive the Cast method for structured types with one type parameter.
+macro_rules! double_dispatch_cast_decl_trait(
+    ($t: ident, $tcast: ident) => (
+        pub trait $tcast<N> {
+            fn to(Self) -> $t<N>;
+        }
+    )
+)
+
+macro_rules! cast_redispatch_impl(
+    ($t:ident, $tcast: ident) => (
+        impl<T: $tcast<N>, N> Cast<T> for $t<N> {
+            #[inline(always)]
+            fn from(t: T) -> $t<N> {
+                $tcast::to(t)
+            }
+        }
+    )
+)
diff --git a/src/structs/mod.rs b/src/structs/mod.rs
index e0edd39d..eb735e7d 100644
--- a/src/structs/mod.rs
+++ b/src/structs/mod.rs
@@ -21,5 +21,6 @@ mod spec {
     mod mat;
     mod vec0;
     mod vec;
+    mod primitives;
     // mod complex;
 }
diff --git a/src/structs/rot.rs b/src/structs/rot.rs
index cb3d0bc5..10c2a2c3 100644
--- a/src/structs/rot.rs
+++ b/src/structs/rot.rs
@@ -2,11 +2,11 @@
 
 #[allow(missing_doc)];
 
-use std::num::{Zero, One, from_f32};
+use std::num::{Zero, One};
 use std::rand::{Rand, Rng};
 use traits::geometry::{Rotate, Rotation, AbsoluteRotate, RotationMatrix, Transform, ToHomogeneous,
                        Norm, Cross};
-use traits::structure::{Dim, Indexable, Row, Col};
+use traits::structure::{Cast, Dim, Indexable, Row, Col};
 use traits::operations::{Absolute, Inv, Transpose};
 use structs::vec::{Vec1, Vec2, Vec3, Vec4, Vec2MulRhs, Vec3MulRhs, Vec4MulRhs};
 use structs::mat::{Mat2, Mat3, Mat4, Mat5};
@@ -168,11 +168,11 @@ impl<N: Clone + Num + Algebraic> Rot3<N> {
     }
 }
 
-impl<N: Clone + Trigonometric + Num + Algebraic + FromPrimitive>
+impl<N: Clone + Trigonometric + Num + Algebraic + Cast<f32>>
 Rotation<Vec3<N>> for Rot3<N> {
     #[inline]
     fn rotation(&self) -> Vec3<N> {
-        let angle = ((self.submat.m11 + self.submat.m22 + self.submat.m33 - One::one()) / from_f32(2.0).unwrap()).acos();
+        let angle = ((self.submat.m11 + self.submat.m22 + self.submat.m33 - One::one()) / Cast::from(2.0)).acos();
 
         if angle != angle {
             // FIXME: handle that correctly
diff --git a/src/structs/rot_macros.rs b/src/structs/rot_macros.rs
index 348a9afa..308807cd 100644
--- a/src/structs/rot_macros.rs
+++ b/src/structs/rot_macros.rs
@@ -56,7 +56,7 @@ macro_rules! dim_impl(
 
 macro_rules! rotation_matrix_impl(
     ($t: ident, $tlv: ident, $tav: ident) => (
-        impl<N: FromPrimitive + Algebraic + Trigonometric + Num + Clone>
+        impl<N: Cast<f32> + Algebraic + Trigonometric + Num + Clone>
         RotationMatrix<$tlv<N>, $tav<N>, $t<N>> for $t<N> {
             #[inline]
             fn to_rot_mat(&self) -> $t<N> {
diff --git a/src/structs/spec/mat.rs b/src/structs/spec/mat.rs
index cf00845e..e54c404c 100644
--- a/src/structs/spec/mat.rs
+++ b/src/structs/spec/mat.rs
@@ -1,9 +1,9 @@
-use std::num::{Zero, One, from_f32};
+use std::num::{Zero, One};
 use structs::vec::{Vec2, Vec3, Vec2MulRhs, Vec3MulRhs};
 use structs::mat::{Mat1, Mat2, Mat3, Mat3MulRhs, Mat2MulRhs};
 use structs::mat;
 use traits::operations::{Inv};
-use traits::structure::{Row, Col};
+use traits::structure::{Cast, Row, Col};
 
 // some specializations:
 impl<N: Num + Clone>
@@ -266,17 +266,17 @@ impl<N: Mul<N, N> + Add<N, N>> Mat2MulRhs<N, Vec2<N>> for Vec2<N> {
 }
 
 // FIXME: move this to another file?
-impl<N: Real + FromPrimitive + Zero + One> mat::Mat4<N> {
+impl<N: Real + Cast<f32> + Zero + One> mat::Mat4<N> {
     /// Computes a projection matrix given the frustrum near plane width, height, the field of
     /// view, and the distance to the clipping planes (`znear` and `zfar`).
     pub fn new_perspective(width: N, height: N, fov: N, znear: N, zfar: N) -> mat::Mat4<N> {
         let aspect = width / height;
 
         let _1: N = One::one();
-        let sy    = _1 / (fov * from_f32(0.5).unwrap()).tan();
+        let sy    = _1 / (fov * Cast::from(0.5)).tan();
         let sx    = -sy / aspect;
         let sz    = -(zfar + znear) / (znear - zfar);
-        let tz    = zfar * znear * from_f32(2.0).unwrap() / (znear - zfar);
+        let tz    = zfar * znear * Cast::from(2.0) / (znear - zfar);
 
         mat::Mat4::new(
             sx,           Zero::zero(), Zero::zero(), Zero::zero(),
diff --git a/src/structs/spec/primitives.rs b/src/structs/spec/primitives.rs
new file mode 100644
index 00000000..cc7ccd61
--- /dev/null
+++ b/src/structs/spec/primitives.rs
@@ -0,0 +1,219 @@
+//! nalgebra trait implementation for primitive types.
+
+#[allow(missing_doc)];
+#[allow(non_camel_case_types)];
+
+use traits::structure::Cast;
+
+// Double dispatch traits to drive the Cast method for primitive types.
+macro_rules! primitive_double_dispatch_cast_decl_trait(
+    ($t: ident, $tcast: ident) => (
+        pub trait $tcast {
+            fn to(Self) -> $t;
+        }
+    )
+)
+
+macro_rules! primitive_double_dispatch_cast_impl(
+    ($ttarget: ident, $tself: ident, $tcast: ident) => (
+        impl $tcast for $tself{
+            #[inline(always)]
+            fn to(v: $tself) -> $ttarget {
+                v as $ttarget
+            }
+        }
+    )
+)
+
+macro_rules! primitive_cast_redispatch_impl(
+    ($t:ident, $tcast: ident) => (
+        impl<T: $tcast> Cast<T> for $t {
+            #[inline(always)]
+            fn from(t: T) -> $t {
+                $tcast::to(t)
+            }
+        }
+    )
+)
+
+primitive_double_dispatch_cast_decl_trait!(f64,  f64Cast)
+primitive_double_dispatch_cast_decl_trait!(f32,  f32Cast)
+primitive_double_dispatch_cast_decl_trait!(i64,  i64Cast)
+primitive_double_dispatch_cast_decl_trait!(i32,  i32Cast)
+primitive_double_dispatch_cast_decl_trait!(i16,  i16Cast)
+primitive_double_dispatch_cast_decl_trait!(i8,   i8Cast)
+primitive_double_dispatch_cast_decl_trait!(u64,  u64Cast)
+primitive_double_dispatch_cast_decl_trait!(u32,  u32Cast)
+primitive_double_dispatch_cast_decl_trait!(u16,  u16Cast)
+primitive_double_dispatch_cast_decl_trait!(u8,   u8Cast)
+primitive_double_dispatch_cast_decl_trait!(int,  intCast)
+primitive_double_dispatch_cast_decl_trait!(uint, uintCast)
+
+primitive_cast_redispatch_impl!(f64,  f64Cast)
+primitive_cast_redispatch_impl!(f32,  f32Cast)
+primitive_cast_redispatch_impl!(i64,  i64Cast)
+primitive_cast_redispatch_impl!(i32,  i32Cast)
+primitive_cast_redispatch_impl!(i16,  i16Cast)
+primitive_cast_redispatch_impl!(i8,   i8Cast)
+primitive_cast_redispatch_impl!(u64,  u64Cast)
+primitive_cast_redispatch_impl!(u32,  u32Cast)
+primitive_cast_redispatch_impl!(u16,  u16Cast)
+primitive_cast_redispatch_impl!(u8,   u8Cast)
+primitive_cast_redispatch_impl!(int,  intCast)
+primitive_cast_redispatch_impl!(uint, uintCast)
+
+primitive_double_dispatch_cast_impl!(f64, f64,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, f32,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, i64,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, i32,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, i16,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, i8,   f64Cast)
+primitive_double_dispatch_cast_impl!(f64, u64,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, u32,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, u16,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, u8,   f64Cast)
+primitive_double_dispatch_cast_impl!(f64, int,  f64Cast)
+primitive_double_dispatch_cast_impl!(f64, uint, f64Cast)
+
+primitive_double_dispatch_cast_impl!(f32, f64,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, f32,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, i64,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, i32,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, i16,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, i8,   f32Cast)
+primitive_double_dispatch_cast_impl!(f32, u64,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, u32,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, u16,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, u8,   f32Cast)
+primitive_double_dispatch_cast_impl!(f32, int,  f32Cast)
+primitive_double_dispatch_cast_impl!(f32, uint, f32Cast)
+
+primitive_double_dispatch_cast_impl!(i64, f64,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, f32,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, i64,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, i32,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, i16,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, i8,   i64Cast)
+primitive_double_dispatch_cast_impl!(i64, u64,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, u32,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, u16,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, u8,   i64Cast)
+primitive_double_dispatch_cast_impl!(i64, int,  i64Cast)
+primitive_double_dispatch_cast_impl!(i64, uint, i64Cast)
+
+primitive_double_dispatch_cast_impl!(i32, f64,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, f32,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, i64,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, i32,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, i16,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, i8,   i32Cast)
+primitive_double_dispatch_cast_impl!(i32, u64,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, u32,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, u16,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, u8,   i32Cast)
+primitive_double_dispatch_cast_impl!(i32, int,  i32Cast)
+primitive_double_dispatch_cast_impl!(i32, uint, i32Cast)
+
+primitive_double_dispatch_cast_impl!(i16, f64,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, f32,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, i64,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, i32,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, i16,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, i8,   i16Cast)
+primitive_double_dispatch_cast_impl!(i16, u64,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, u32,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, u16,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, u8,   i16Cast)
+primitive_double_dispatch_cast_impl!(i16, int,  i16Cast)
+primitive_double_dispatch_cast_impl!(i16, uint, i16Cast)
+
+primitive_double_dispatch_cast_impl!(i8, f64,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, f32,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, i64,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, i32,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, i16,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, i8,   i8Cast)
+primitive_double_dispatch_cast_impl!(i8, u64,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, u32,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, u16,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, u8,   i8Cast)
+primitive_double_dispatch_cast_impl!(i8, int,  i8Cast)
+primitive_double_dispatch_cast_impl!(i8, uint, i8Cast)
+
+primitive_double_dispatch_cast_impl!(u64, f64,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, f32,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, i64,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, i32,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, i16,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, i8,   u64Cast)
+primitive_double_dispatch_cast_impl!(u64, u64,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, u32,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, u16,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, u8,   u64Cast)
+primitive_double_dispatch_cast_impl!(u64, int,  u64Cast)
+primitive_double_dispatch_cast_impl!(u64, uint, u64Cast)
+
+primitive_double_dispatch_cast_impl!(u32, f64,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, f32,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, i64,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, i32,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, i16,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, i8,   u32Cast)
+primitive_double_dispatch_cast_impl!(u32, u64,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, u32,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, u16,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, u8,   u32Cast)
+primitive_double_dispatch_cast_impl!(u32, int,  u32Cast)
+primitive_double_dispatch_cast_impl!(u32, uint, u32Cast)
+
+primitive_double_dispatch_cast_impl!(u16, f64,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, f32,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, i64,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, i32,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, i16,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, i8,   u16Cast)
+primitive_double_dispatch_cast_impl!(u16, u64,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, u32,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, u16,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, u8,   u16Cast)
+primitive_double_dispatch_cast_impl!(u16, int,  u16Cast)
+primitive_double_dispatch_cast_impl!(u16, uint, u16Cast)
+
+primitive_double_dispatch_cast_impl!(u8, f64,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, f32,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, i64,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, i32,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, i16,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, i8,   u8Cast)
+primitive_double_dispatch_cast_impl!(u8, u64,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, u32,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, u16,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, u8,   u8Cast)
+primitive_double_dispatch_cast_impl!(u8, int,  u8Cast)
+primitive_double_dispatch_cast_impl!(u8, uint, u8Cast)
+
+primitive_double_dispatch_cast_impl!(uint, f64,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, f32,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, i64,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, i32,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, i16,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, i8,   uintCast)
+primitive_double_dispatch_cast_impl!(uint, u64,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, u32,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, u16,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, u8,   uintCast)
+primitive_double_dispatch_cast_impl!(uint, int,  uintCast)
+primitive_double_dispatch_cast_impl!(uint, uint, uintCast)
+
+primitive_double_dispatch_cast_impl!(int, f64,  intCast)
+primitive_double_dispatch_cast_impl!(int, f32,  intCast)
+primitive_double_dispatch_cast_impl!(int, i64,  intCast)
+primitive_double_dispatch_cast_impl!(int, i32,  intCast)
+primitive_double_dispatch_cast_impl!(int, i16,  intCast)
+primitive_double_dispatch_cast_impl!(int, i8,   intCast)
+primitive_double_dispatch_cast_impl!(int, u64,  intCast)
+primitive_double_dispatch_cast_impl!(int, u32,  intCast)
+primitive_double_dispatch_cast_impl!(int, u16,  intCast)
+primitive_double_dispatch_cast_impl!(int, u8,   intCast)
+primitive_double_dispatch_cast_impl!(int, int,  intCast)
+primitive_double_dispatch_cast_impl!(int, uint, intCast)
diff --git a/src/structs/spec/vec.rs b/src/structs/spec/vec.rs
index dff4d43b..61bef91f 100644
--- a/src/structs/spec/vec.rs
+++ b/src/structs/spec/vec.rs
@@ -1,5 +1,5 @@
 use std::num::{Zero, One};
-use traits::structure::{VecCast, Row, Basis};
+use traits::structure::{Cast, Row, Basis};
 use traits::geometry::{Norm, Cross, CrossMatrix, UniformSphereSample};
 use structs::vec::{Vec1, Vec2, Vec3, Vec4};
 use structs::mat::Mat3;
@@ -115,7 +115,7 @@ impl<N: Clone + Ord + Algebraic + Signed> Basis for Vec3<N> {
 }
 
 // FIXME: this bad: this fixes definitly the number of samples…
-static SAMPLES_2_F64: [Vec2<f64>, ..21] = [
+static SAMPLES_2_F32: [Vec2<f32>, ..21] = [
     Vec2 { x: 1.0,         y: 0.0         },
     Vec2 { x: 0.95557281,  y: 0.29475517  },
     Vec2 { x: 0.82623877,  y: 0.56332006  },
@@ -140,7 +140,7 @@ static SAMPLES_2_F64: [Vec2<f64>, ..21] = [
 ];
 
 // Those vectors come from bullet 3d
-static SAMPLES_3_F64: [Vec3<f64>, ..42] = [
+static SAMPLES_3_F32: [Vec3<f32>, ..42] = [
     Vec3 { x: 0.000000 , y: -0.000000, z: -1.000000 },
     Vec3 { x: 0.723608 , y: -0.525725, z: -0.447219 },
     Vec3 { x: -0.276388, y: -0.850649, z: -0.447219 },
@@ -192,30 +192,30 @@ impl<N: One + Clone> UniformSphereSample for Vec1<N> {
      }
 }
 
-impl<N: NumCast + Clone> UniformSphereSample for Vec2<N> {
+impl<N: Cast<f32> + Clone> UniformSphereSample for Vec2<N> {
     #[inline(always)]
     fn sample(f: &fn(Vec2<N>)) {
-         for sample in SAMPLES_2_F64.iter() {
-             f(VecCast::from(*sample))
+         for sample in SAMPLES_2_F32.iter() {
+             f(Cast::from(*sample))
          }
      }
 }
 
-impl<N: NumCast + Clone> UniformSphereSample for Vec3<N> {
+impl<N: Cast<f32> + Clone> UniformSphereSample for Vec3<N> {
     #[inline(always)]
     fn sample(f: &fn(Vec3<N>)) {
-        for sample in SAMPLES_3_F64.iter() {
-            f(VecCast::from(*sample))
+        for sample in SAMPLES_3_F32.iter() {
+            f(Cast::from(*sample))
         }
     }
 }
 
-impl<N: NumCast + Clone> UniformSphereSample for Vec4<N> {
+impl<N: Cast<f32> + Clone> UniformSphereSample for Vec4<N> {
     #[inline(always)]
     fn sample(_: &fn(Vec4<N>)) {
         fail!("UniformSphereSample::<Vec4<N>>::sample : Not yet implemented.")
-        // for sample in SAMPLES_3_F64.iter() {
-        //     f(VecCast::from(*sample))
+        // for sample in SAMPLES_3_F32.iter() {
+        //     f(Cast::from(*sample))
         // }
     }
 }
diff --git a/src/structs/vec.rs b/src/structs/vec.rs
index 825da00e..d78e4868 100644
--- a/src/structs/vec.rs
+++ b/src/structs/vec.rs
@@ -11,7 +11,7 @@ use std::cmp::ApproxEq;
 
 use traits::geometry::{Transform, Rotate, FromHomogeneous, ToHomogeneous, Dot, Norm,
                        Translation, Translate};
-use traits::structure::{VecCast, Basis, Dim, Indexable, Iterable, IterableMut};
+use traits::structure::{Basis, Cast, Dim, Indexable, Iterable, IterableMut};
 
 mod metal;
 mod vec_macros;
@@ -31,15 +31,17 @@ double_dispatch_binop_decl_trait!(Vec1, Vec1MulRhs)
 double_dispatch_binop_decl_trait!(Vec1, Vec1DivRhs)
 double_dispatch_binop_decl_trait!(Vec1, Vec1AddRhs)
 double_dispatch_binop_decl_trait!(Vec1, Vec1SubRhs)
+double_dispatch_cast_decl_trait!(Vec1, Vec1Cast)
 mul_redispatch_impl!(Vec1, Vec1MulRhs)
 div_redispatch_impl!(Vec1, Vec1DivRhs)
 add_redispatch_impl!(Vec1, Vec1AddRhs)
 sub_redispatch_impl!(Vec1, Vec1SubRhs)
+cast_redispatch_impl!(Vec1, Vec1Cast)
 new_impl!(Vec1, x)
 ord_impl!(Vec1, x)
 orderable_impl!(Vec1, x)
 vec_axis_impl!(Vec1, x)
-vec_cast_impl!(Vec1, x)
+vec_cast_impl!(Vec1, Vec1Cast, x)
 indexable_impl!(Vec1, 1)
 at_fast_impl!(Vec1, 1)
 new_repeat_impl!(Vec1, val, x)
@@ -128,15 +130,17 @@ double_dispatch_binop_decl_trait!(Vec2, Vec2MulRhs)
 double_dispatch_binop_decl_trait!(Vec2, Vec2DivRhs)
 double_dispatch_binop_decl_trait!(Vec2, Vec2AddRhs)
 double_dispatch_binop_decl_trait!(Vec2, Vec2SubRhs)
+double_dispatch_cast_decl_trait!(Vec2, Vec2Cast)
 mul_redispatch_impl!(Vec2, Vec2MulRhs)
 div_redispatch_impl!(Vec2, Vec2DivRhs)
 add_redispatch_impl!(Vec2, Vec2AddRhs)
 sub_redispatch_impl!(Vec2, Vec2SubRhs)
+cast_redispatch_impl!(Vec2, Vec2Cast)
 new_impl!(Vec2, x, y)
 ord_impl!(Vec2, x, y)
 orderable_impl!(Vec2, x, y)
 vec_axis_impl!(Vec2, x, y)
-vec_cast_impl!(Vec2, x, y)
+vec_cast_impl!(Vec2, Vec2Cast, x, y)
 indexable_impl!(Vec2, 2)
 at_fast_impl!(Vec2, 2)
 new_repeat_impl!(Vec2, val, x, y)
@@ -227,15 +231,17 @@ double_dispatch_binop_decl_trait!(Vec3, Vec3MulRhs)
 double_dispatch_binop_decl_trait!(Vec3, Vec3DivRhs)
 double_dispatch_binop_decl_trait!(Vec3, Vec3AddRhs)
 double_dispatch_binop_decl_trait!(Vec3, Vec3SubRhs)
+double_dispatch_cast_decl_trait!(Vec3, Vec3Cast)
 mul_redispatch_impl!(Vec3, Vec3MulRhs)
 div_redispatch_impl!(Vec3, Vec3DivRhs)
 add_redispatch_impl!(Vec3, Vec3AddRhs)
 sub_redispatch_impl!(Vec3, Vec3SubRhs)
+cast_redispatch_impl!(Vec3, Vec3Cast)
 new_impl!(Vec3, x, y, z)
 ord_impl!(Vec3, x, y, z)
 orderable_impl!(Vec3, x, y, z)
 vec_axis_impl!(Vec3, x, y, z)
-vec_cast_impl!(Vec3, x, y, z)
+vec_cast_impl!(Vec3, Vec3Cast, x, y, z)
 indexable_impl!(Vec3, 3)
 at_fast_impl!(Vec3, 3)
 new_repeat_impl!(Vec3, val, x, y, z)
@@ -339,14 +345,16 @@ double_dispatch_binop_decl_trait!(PVec3, PVec3MulRhs)
 double_dispatch_binop_decl_trait!(PVec3, PVec3DivRhs)
 double_dispatch_binop_decl_trait!(PVec3, PVec3AddRhs)
 double_dispatch_binop_decl_trait!(PVec3, PVec3SubRhs)
+double_dispatch_cast_decl_trait!(PVec3, PVec3Cast)
 mul_redispatch_impl!(PVec3, PVec3MulRhs)
 div_redispatch_impl!(PVec3, PVec3DivRhs)
 add_redispatch_impl!(PVec3, PVec3AddRhs)
 sub_redispatch_impl!(PVec3, PVec3SubRhs)
+cast_redispatch_impl!(PVec3, PVec3Cast)
 ord_impl!(PVec3, x, y, z)
 orderable_impl!(PVec3, x, y, z)
 vec_axis_impl!(PVec3, x, y, z)
-vec_cast_impl!(PVec3, x, y, z)
+vec_cast_impl!(PVec3, PVec3Cast, x, y, z)
 indexable_impl!(PVec3, 3)
 at_fast_impl!(PVec3, 3)
 new_repeat_impl!(PVec3, val, x, y, z, _unused)
@@ -441,15 +449,17 @@ double_dispatch_binop_decl_trait!(Vec4, Vec4MulRhs)
 double_dispatch_binop_decl_trait!(Vec4, Vec4DivRhs)
 double_dispatch_binop_decl_trait!(Vec4, Vec4AddRhs)
 double_dispatch_binop_decl_trait!(Vec4, Vec4SubRhs)
+double_dispatch_cast_decl_trait!(Vec4, Vec4Cast)
 mul_redispatch_impl!(Vec4, Vec4MulRhs)
 div_redispatch_impl!(Vec4, Vec4DivRhs)
 add_redispatch_impl!(Vec4, Vec4AddRhs)
 sub_redispatch_impl!(Vec4, Vec4SubRhs)
+cast_redispatch_impl!(Vec4, Vec4Cast)
 new_impl!(Vec4, x, y, z, w)
 ord_impl!(Vec4, x, y, z, w)
 orderable_impl!(Vec4, x, y, z, w)
 vec_axis_impl!(Vec4, x, y, z, w)
-vec_cast_impl!(Vec4, x, y, z, w)
+vec_cast_impl!(Vec4, Vec4Cast, x, y, z, w)
 indexable_impl!(Vec4, 4)
 at_fast_impl!(Vec4, 4)
 new_repeat_impl!(Vec4, val, x, y, z, w)
@@ -544,15 +554,17 @@ double_dispatch_binop_decl_trait!(Vec5, Vec5MulRhs)
 double_dispatch_binop_decl_trait!(Vec5, Vec5DivRhs)
 double_dispatch_binop_decl_trait!(Vec5, Vec5AddRhs)
 double_dispatch_binop_decl_trait!(Vec5, Vec5SubRhs)
+double_dispatch_cast_decl_trait!(Vec5, Vec5Cast)
 mul_redispatch_impl!(Vec5, Vec5MulRhs)
 div_redispatch_impl!(Vec5, Vec5DivRhs)
 add_redispatch_impl!(Vec5, Vec5AddRhs)
 sub_redispatch_impl!(Vec5, Vec5SubRhs)
+cast_redispatch_impl!(Vec5, Vec5Cast)
 new_impl!(Vec5, x, y, z, w, a)
 ord_impl!(Vec5, x, y, z, w, a)
 orderable_impl!(Vec5, x, y, z, w, a)
 vec_axis_impl!(Vec5, x, y, z, w, a)
-vec_cast_impl!(Vec5, x, y, z, w, a)
+vec_cast_impl!(Vec5, Vec5Cast, x, y, z, w, a)
 indexable_impl!(Vec5, 5)
 at_fast_impl!(Vec5, 5)
 new_repeat_impl!(Vec5, val, x, y, z, w, a)
@@ -649,15 +661,17 @@ double_dispatch_binop_decl_trait!(Vec6, Vec6MulRhs)
 double_dispatch_binop_decl_trait!(Vec6, Vec6DivRhs)
 double_dispatch_binop_decl_trait!(Vec6, Vec6AddRhs)
 double_dispatch_binop_decl_trait!(Vec6, Vec6SubRhs)
+double_dispatch_cast_decl_trait!(Vec6, Vec6Cast)
 mul_redispatch_impl!(Vec6, Vec6MulRhs)
 div_redispatch_impl!(Vec6, Vec6DivRhs)
 add_redispatch_impl!(Vec6, Vec6AddRhs)
 sub_redispatch_impl!(Vec6, Vec6SubRhs)
+cast_redispatch_impl!(Vec6, Vec6Cast)
 new_impl!(Vec6, x, y, z, w, a, b)
 ord_impl!(Vec6, x, y, z, w, a, b)
 orderable_impl!(Vec6, x, y, z, w, a, b)
 vec_axis_impl!(Vec6, x, y, z, w, a, b)
-vec_cast_impl!(Vec6, x, y, z, w, a, b)
+vec_cast_impl!(Vec6, Vec6Cast, x, y, z, w, a, b)
 indexable_impl!(Vec6, 6)
 at_fast_impl!(Vec6, 6)
 new_repeat_impl!(Vec6, val, x, y, z, w, a, b)
diff --git a/src/structs/vec_macros.rs b/src/structs/vec_macros.rs
index 357e48c0..1e402352 100644
--- a/src/structs/vec_macros.rs
+++ b/src/structs/vec_macros.rs
@@ -111,11 +111,11 @@ macro_rules! vec_axis_impl(
 )
 
 macro_rules! vec_cast_impl(
-    ($t: ident, $comp0: ident $(,$compN: ident)*) => (
-        impl<Nin: NumCast + Clone, Nout: Clone + NumCast> VecCast<$t<Nout>> for $t<Nin> {
+    ($t: ident, $tcast: ident, $comp0: ident $(,$compN: ident)*) => (
+        impl<Nin: Clone, Nout: Clone + Cast<Nin>> $tcast<Nout> for $t<Nin> {
             #[inline]
-            fn from(v: $t<Nin>) -> $t<Nout> {
-                $t::new(NumCast::from(v.$comp0.clone()).unwrap() $(, NumCast::from(v.$compN.clone()).unwrap())*)
+            fn to(v: $t<Nin>) -> $t<Nout> {
+                $t::new(Cast::from(v.$comp0.clone()) $(, Cast::from(v.$compN.clone()))*)
             }
         }
     )
diff --git a/src/traits/mod.rs b/src/traits/mod.rs
index 99de789f..b41c4fa5 100644
--- a/src/traits/mod.rs
+++ b/src/traits/mod.rs
@@ -4,8 +4,8 @@ pub use self::geometry::{AbsoluteRotate, Cross, CrossMatrix, Dot, FromHomogeneou
                          Rotation, RotationMatrix, RotationWithTranslation, ToHomogeneous,
                          Transform, Transformation, Translate, Translation, UniformSphereSample};
 
-pub use self::structure::{AlgebraicVec, AlgebraicVecExt, Basis, Col, Dim, Indexable, Iterable,
-                          IterableMut, Mat, MatCast, Row, Vec, VecCast, VecExt};
+pub use self::structure::{AlgebraicVec, AlgebraicVecExt, Basis, Cast, Col, Dim, Indexable,
+                          Iterable, IterableMut, Mat, Row, Vec, VecExt};
 
 pub use self::operations::{Absolute, Cov, Inv, LMul, Mean, Outer, RMul, ScalarAdd, ScalarSub,
                            Transpose};
diff --git a/src/traits/structure.rs b/src/traits/structure.rs
index eb5fe118..87bae128 100644
--- a/src/traits/structure.rs
+++ b/src/traits/structure.rs
@@ -5,6 +5,12 @@ use std::vec::{VecIterator, VecMutIterator};
 use traits::operations::{RMul, LMul, ScalarAdd, ScalarSub};
 use traits::geometry::{Dot, Norm, UniformSphereSample};
 
+/// Traits of objects which can be created from an object of type `T`.
+pub trait Cast<T> {
+    /// Converts an element of type `T` to an element of type `Self`.
+    fn from(t: T) -> Self;
+}
+
 /// Trait of matrices.
 ///
 /// A matrix has rows and columns and are able to multiply them.
@@ -13,14 +19,6 @@ pub trait Mat<R, C> : Row<R> + Col<C> + RMul<R> + LMul<C> { }
 impl<M: Row<R> + Col<C> + RMul<R> + LMul<C>, R, C> Mat<R, C> for M {
 }
 
-/// Trait of matrices which can be converted to another matrix.
-///
-/// Use this to change easily the type of a matrix components.
-pub trait MatCast<M> {
-    /// Converts `m` to have the type `M`.
-    fn from(m: Self) -> M;
-}
-
 // XXX: we keep ScalarAdd and ScalarSub here to avoid trait impl conflict (overriding) between the
 // different Add/Sub traits. This is _so_ unfortunate…
 
@@ -55,13 +53,6 @@ VecExt<N> for V { }
 
 impl<N: Algebraic, V: AlgebraicVec<N> + VecExt<N> + Basis + Round> AlgebraicVecExt<N> for V { }
 
-/// Trait of vectors which can be converted to another vector. Used to change the type of a vector
-/// components.
-pub trait VecCast<V> {
-    /// Converts `v` to have the type `V`.
-    fn from(v: Self) -> V;
-}
-
 // FIXME: return an iterator instead
 /// Traits of objects which can form a basis (typically vectors).
 pub trait Basis {