2013-04-11, 01:25 AM
I am not sure what the problem is supposed to be here.
By outer I assume you mean the lines near the tan areas and not the edges cutting down the middle of the model. You need to move the edges far away from the tan region, or colour a thick border in around them.
A difference of 1.0000 to 0.9998 is rounding error. Most likely the program is just not displaying as many decimal places as it could just to save room in the display field. A 32bit floating point value has a precision of 6 decimal places with the last place not considered to be reliable.
Also remember that 1 is (literally) the same number as 0.99999... (ad infinitum)
FPUs are not perfect, there is always rounding error. The only way to get a value of exactly 1 is to either not compute it, or compute it with a rational number. In other words 0.9998 is 1. That's not your problem.
By outer I assume you mean the lines near the tan areas and not the edges cutting down the middle of the model. You need to move the edges far away from the tan region, or colour a thick border in around them.
A difference of 1.0000 to 0.9998 is rounding error. Most likely the program is just not displaying as many decimal places as it could just to save room in the display field. A 32bit floating point value has a precision of 6 decimal places with the last place not considered to be reliable.
Also remember that 1 is (literally) the same number as 0.99999... (ad infinitum)
FPUs are not perfect, there is always rounding error. The only way to get a value of exactly 1 is to either not compute it, or compute it with a rational number. In other words 0.9998 is 1. That's not your problem.