( [ 0,0,0 ].vector_to( [ 0,0,0 ] ) ).length.mm < 0.01.mm

[Davidfi1]

is this "true" or "false"?:
( [ 0,0,0 ].vector_to( [ 0,0,0 ] ) ).length.mm < 0.01.mm

on my computer it is "false" - and I dont know why?

[Jim]

SketchUp's Length class uses some default tolerance in its comparisons - the tolerance is not documented as far as I know.

@unknownuser said:

The equality comparison on Length values uses the default tolerance that SketchUp uses for comparing lengths

In SketchUp-land:
> 0.01.mm == 0 #==> true

Search the following page for "tolerance"

http://www.sketchup.com/intl/en/developer/docs/ourdoc/length

[Aerilius]

nullvec = Geom;;Vector3d.new(0,0,0)
l1 = nullvec.length.mm # obviously 0 mm or 0 inch
l2 = 0.01.mm  # 0.000393700787401575 inch
l1 == l2
 → true

but

0 < 0.000393700787401575
 → true

This is because #mm or #to_mm etc. are methods that return an object of the Length class, not a Float. A Length is a Float that considers SketchUp's tolerance, which is currently set to 0.001 inch (we should never use this tolerance value directly since SketchUp might have the freedom to change it).

[Davidfi1]

so to find the length of I should use length.to_f ?

[thomthom]

@davidfi1 said:

so to find the length of I should use length.to_f ?

To treat it as a float you must use to_f, yes. The numeric value of the objects are the same. The only difference of the Length which is a sub-class of Float is the comparison operators.

( [ 0,0,0 ].vector_to( [ 0,0,0 ] ) ).length.to_f < 0.01.mm
That works because you're now dealing with the standard floating point comparisons. Not knowing what your experience with floating points are, but just in case; here's a very informative site on the topic: http://floating-point-gui.de/

[Davidfi1]

Very informative site in did.
Thank you.
I am a hardware engineer by training - so I know this problem from the other side
and still it is a little strange