Difference between glcolor3b and glcolor3ub

Question

I was wondering what the difference is between glcolor3b() and glcolor3ub().

It appears that glcolor3b(255, 0, 0); does not set the color to red, but rather it sets it to black. glcolor3b(48, 160, 64) sets it to dark purple, not green. glcolor3ub(), however works as expected.
Additionally, the documentation for glcolor3b() and glcolor3ub() are exactly the same, except for the "u":

public static void glColor3(u)b(byte red,
              byte green,
              byte blue)

Does anybody know why this is?

Answer 1

glColor3b() takes byte parameters with a range from -128 to 127. glColor3ub() takes unsigned byte parameters with a range from 0 to 255. Using values greater than 127 with glColor3b() leads to an arithmetic overflow.

Answer 2

Of course 255 sets it to black. That is 0xff (or 0b11111111), which is -1 using a 2's complement signed 8-bit number...

-1 is less than 0, which you would consider to be the absence of all color. Signed colors really do not make much sense outside of blending. In a nutshell, that is the difference between these two functions, one is signed and other is unsigned.

When you use the glColor3b (...) function (signed), then your range is -128–127 (-128 maps to -1.0 and 127 maps to 1.0, 0 is the mid-point),

When you use glColor3ub (...) function (un-signed), the range is 0–255 (0 maps to 0.0 and 255 maps to 1.0).

No matter which function you use, unless it is glColor3f (...), they all do fixed-point to floating-point unit conversion. During fixed-point to floating-point normalization, the range of the integer data type is mapped directly between -1.0 (signed) / 0 (unsigned) and 1.0. The u vs. non-u simply indicates that one of them is unsigned (larger positive range).

Answer 3

glColor3ub is the version for unsigned char in C/C++, that is a 8-bit integer without a sign.

glColor3b is the version for char which is a signed 8-bit integer.

255 = 0xFF is actually -1 when interpreted as a signed 8-bit integer. That is why you get a black screen because of the Two's complement representation. Simply stick to the ub versions.