Am 19.08.2013 04:40, schrieb Bald Eagle:
>
>> Note that <0,0,0> has no direction, so the angle between it and any
>> other given vector is poorly defined. The function has to return
>> /something/, and it decides to return 90 because it's just as good or
>> bad as any other answer in that situation.
>
> Except now we're being all loosey-goosey with the definition of "vector".
> I would have understood or _assumed_ that like everything else, these would be
> location vectors, not direction vectors.

That's possibly where your misunderstanding is:

Virtually all of the vector math functions built into POV-Ray don't 
/care/ whether the vectors you feed it are location vectors or direction 
vectors. They /are/ just arbitrary vectors.

> Considering that VDist(A,B) gives me the distance between two location vectors,
> I would be tempted to assume that if I wanted to draw a line between said points
> (in the plane {z, 0}, I would draw a line from the origin to <VDist(A,B), 0, 0>,
> and then rotate it by VAngleD(A,B), then translate it by <A.x, A.y, 0>

It seems to me that from the naming and parameter list of those macros 
you're trying to infer what they do and how you can combine them. This 
approach won't get you anywhere: They're just pieces of math with fancy 
names, and pieces of math are happy to be your friend or kick you in the 
butt, depending how you use them.

> So yes, once again the directive does exactly what the documentation says it
> does, but the documentation is not clear WHICH definition of "vector" is being
> applied when interpreting the directive.

That is because, as demonstrated, it can be used for /either/ definition 
of "vector".

> Now, granted, I do stoopid things and confuse myself all the time, but this was
> a tricksy one, oh yes my preciousssss.  Very tricksy indeed.  Nasty little
> directiveses.  Not like End User it does.  Hates us.

LOL :-)

Post a reply to this message