What does it mean, "loops don't work"? Did you check the authoring examples collection?

As far "while" loops don't work, so i run the "while" loop in other code to print the tests.

There're no specific reason of number of tests (666).

This way of structuring tests is unacceptable. It even makes my browser lag.

Is there anything wrong with loops? Is there a specific reason behind the 600+ lines?

relative error is now 10^-3 and rounding is not needed now.

Don't reinvent the wheel. See how other kata do it. A relative error of between 1e-6 and 1e-12 should just fine.

I kind of have to clarify that the [...,...] meant range notation (inclusive on both ends).

No reason to use spoilers here.

• hmm... would error tolerance of ±(1 kg/m^3) be wide enough, though...
• or maybe in the [floor(ρ),ceil(ρ)] range?

I can't read spoilers; I haven't solved. I don't know what you're saying, but does it really need to be spoilered?

or maybe in the [floor(ρ),ceil(ρ)] range?

hmm... would error tolerance of ±(1 kg/m^3) be wide enough, though...

Do not use rounding; test by comparing with an error margin instead. This is documented.

Random tests should not all expect the same value because this could happen.

you could simply make a second one, identical except includes random tests

will republishing (as another kata with other hexadecimal identifier, internal contents such as material names are changed) actually do an effect?