Pedantic use of floats necessary?

[jwboth]

When defining fractures (or other lengths and coordinates) which shall be rescaled by unit conversion, it is necessary to define them as array of floats and not array of integers. Otherwise, an error is thrown:

Simple redefinition of the unit conversion in the style value = value / factor instead of value /= factor should do the job robustly.

Is there any (safety) reason why one would want to be type safe here? Or would the user-experience be improved by relaxing the safety here?

[keileg]

I can see no good reason for not changing.