glRotatef results on a wirecube - c++

Question

I am using a rand() function for and x variable and y variable to be used in glRotatef. Rand is working just fine. What is happening is when I get a 0 value in both X and Y, the cube fluctuates by shrinking and inflating. I have tried a couple of ways to make sure this does not happen but alas, I am here. Here is the function I am working with:
void initRandoms()
{
maxCubes = rand() % (CUBE_HIGH - CUBE_LOW + 1) + CUBE_LOW;
for (int i = 0; i < maxCubes; i++)
{
cubeOrigins[i].x = X_LOW + (float)rand() / (float)RAND_MAX * (X_HIGH - X_LOW);
cubeOrigins[i].y = Y_LOW + (float)rand() / (float)RAND_MAX * (Y_HIGH - Y_LOW);
//cubeOrigins[i].z =
cubeOrigins[i].size = SIZE_LOW + (double)rand() / (double)RAND_MAX * (SIZE_HIGH - SIZE_LOW);
cubeOrigins[i].rotateX = rand() % 2;
cubeOrigins[i].rotateY = rand() % 2;
}
As I said before, each cube will rotate either on the X-axis, Y-axis, XY-axis or shrink and inflate. It is the shrink and inflate I need to remove which corresponds to 0 = X AND 0 = Y. I can have X = 0 OR Y = 0. What I have tried is some if else conditionals but what happens is I take the shrink & inflate out but then all the cubes rotate in the same direction. I am hoping someone can figure out what I am doing wrong and show me what I need to do. I appreciate everyone's help. I will put more code up if needed.
Here is myDisplay function where the above function is used. The above function is also called in main:
void myDisplay()
{
glClearColor(0.0f, 0.0f, 0.0f, 0.0f);
glClear(GL_COLOR_BUFFER_BIT);
for(int i = 0; i < maxCubes; i++)
{
glLoadIdentity();
glTranslatef(0.0f, 0.0f, -120.0);
glTranslatef(cubeOrigins[i].x, cubeOrigins[i].y, cubeZ);
glRotatef(rotateAxis, cubeOrigins[i].rotateX, cubeOrigins[i].rotateY, 0.0f);
//glRotatef(rotateAxis, 1, 1, 0);
glutWireCube(cubeOrigins[i].size);
}
cubeZ += 0.050f;
glutSwapBuffers();
glFlush();
if (cubeZ > 120.0f)
{
cubeZ -= 100.f;
glClearColor(0.0f, 0.0f, 0.0f, 0.0f);
glClear(GL_COLOR_BUFFER_BIT);
for(int i = 0; i < maxCubes; i++)
{
initRandoms();
glLoadIdentity();
glTranslatef(0.0f, 0.0f, -120.0);
glTranslatef(cubeOrigins[i].x, cubeOrigins[i].y, cubeZ);
glRotatef(rotateAxis, cubeOrigins[i].rotateX, cubeOrigins[i].rotateY, 0.0f);
//glRotatef(rotateAxis, 1, 1, 0);
glutWireCube(cubeOrigins[i].size);
}
cubeZ += 0.050f;
glutSwapBuffers();
glFlush();
}
}

Answer 1

Looking at your code, I could see that you are rotating your cubes by using this function:
glRotatef(rotateAxis, cubeOrigins[i].rotateX, cubeOrigins[i].rotateY, 0.0f);
The question is: What happens when both, cubeOrigins[i].rotateX and cubeOrigins[i].rotateY are zero? OpenGL rotates matrix by using a mathematic function called quaternion (http://www.opengl.org/sdk/docs/man2/xhtml/glRotate.xml) and, when x, y and z are zero, I guess (because your are violating one of its premises that is ||x, y, z|| must be 1) it is reduced to this matrix:
|cos(angle) 0 0 0|
| 0 cos(angle) 0 0|
| 0 0 cos(angle) 0|
| 0 0 0 1|
which is essentially a scale matrix. This explain your beforementioned effect of shrinking and inflating.
So, as you observed before, you need to avoid this situation.
Second, you are using, for all cubes, the same rotateAxis. So, every cube that is with same cubeOrigins[i].rotateX and cubeOrigins[i].rotateY will got the same rotations. You need to get some variations here (changing rotateAxis for each cube)
Finally, you dont need to call both:
glutSwapBuffers();
glFlush();
glutSwapBuffers() already call, internally glFlush

Answer 2

first things first pls
(im not sure if you already did but)
before using the C rand() method you need to give it a seed so that you get "real" random numbers using the srand() function