Vector's equality and inequality operators

[Riyadh International]

Hello everybody. Thanks for SFML loking forward to new release and more tutorials like Joystick.

Anyway I see this in Vector2.inl and Vector3.inl:

Code: [Select]


template <typename T>
bool operator ==(const Vector2<T>& V1, const Vector2<T>& V2)
{
    return (V1.x == V2.x) && (V1.y == V2.y);
}


template <typename T>
bool operator !=(const Vector2<T>& V1, const Vector2<T>& V2)
{
    return (V1.x != V2.x) || (V1.y != V2.y);
}

template <typename T>
bool operator ==(const Vector3<T>& V1, const Vector3<T>& V2)
{
    return (V1.x == V2.x) && (V1.y == V2.y) && (V1.z == V2.z);
}

template <typename T>
bool operator !=(const Vector3<T>& V1, const Vector3<T>& V2)
{
    return (V1.x != V2.x) || (V1.y != V2.y) || (V1.z != V2.z);
}


This will not work reliable for Floats but only Integers.
I found article which helped me and maybe you too:
http://www.cygnus-software.com/papers/comparingfloats/comparingfloats.htm

[Laurent]

I know. But it's up to users to define how they want to compare their float vectors, if they want to use a delta. SFML only provides the "strict" operator==, like the language provides the strict operator== for float and double types.

[Silvah]

This will not work reliable for Floats but only Integers.

This will work reliably for floats. I'm using operator== when I want to check if two things are equal, not "almost equal". So there's nothing wrong with that code.

But it's up to users to define how they want to compare their float vectors, if they want to use a delta.

Actually, it's called epsilon, not delta

[Laurent]

Actually, it's called epsilon, not delta

Oh, I didn't know that there was a strict naming for this. But actually I always use "epsilon", don't know why I chose "delta" this time :lol:

[OniLinkPlus]

Actually, it's called epsilon, not delta

Oh, I didn't know that there was a strict naming for this. But actually I always use "epsilon", don't know why I chose "delta" this time :lol:

Epsilon is a mathematics term which refers to an infinitesimally small positive number whose square is 0. Delta refers to a change in a number. So yes, it's kind of strict and delta is definitely the wrong term.
Sorry, math geek here.