The code I see for Quaternion multiplication on System.Numerics.Vectors is different (reversed) from all the other code I see on the Internet. What gives?

http://www.dotnetframework.org/default.aspx/DotNET/DotNET/8@0/untmp/WIN_WINDOWS/lh_tools_devdiv_wpf/Windows/wcp/Core/System/Windows/Media3D/Quaternion@cs/1/Quaternion@cs

https://github.com/dotnet/corefx/blob/master/src/System.Numerics.Vectors/src/System/Numerics/Quaternion.cs

Quaternion multiplication is well defined in the Math literature. i*j = -j*i = k, j*k = -k*j = i, k*i = -i*k = j, and everything else can be deduced from the fact that 1 is the neutral element of multiplication and by linearity.

Try to multiply i*j and if you get something different than k, it's broken.

Let's define a quaternion like this:

struct Quaternion 
{
    Vector3 v; 
    float s;
};

Then the quaternion product is defined as:

Quaternion operator*( const Quaternion& q1, const Quaternion& q2 )
{
    Quaternion result;
    result.v = cross( q1.v, q2.v )+ q1.s * q2.v + q2.s * q1.v;
    result.s = q1.s * q2.s - dot( q1, q2 );

    return result;
}

Both your functions are equivalent. The vector'ish form is just less usual in literature in my experience. If you write the definition above component-wise you will get the same result!