Vertex Overlapping in OpenGL - opengl

Question

here is the code which draws a cylinder primitive :
glBegin(GL_TRIANGLE_FAN);
glVertex3f(0, height / 2.0f, 0);
for (int i = 0; i < SIDENUMS +1; i++)
{
glVertex3f(radius * cos(i * 2 * PI / SIDENUMS), height / 2.0f, radius * sin(i * 2 * PI / SIDENUMS));
}
glEnd();
glBegin(GL_QUADS);
for (int i = 0; i < SIDENUMS + 1;i++)
{
glVertex3f(radius * cos(i * 2 * PI / SIDENUMS), height / 2.0f, radius * sin(i * 2 * PI / SIDENUMS));
glVertex3f(radius * cos(i * 2 * PI / SIDENUMS), -height / 2.0f, radius * sin(i * 2 * PI / SIDENUMS));
glVertex3f(radius * cos((i+1) * 2 * PI / SIDENUMS), -height / 2.0f, radius * sin((i+1) * 2 * PI / SIDENUMS));
glVertex3f(radius * cos((i+1) * 2 * PI / SIDENUMS), height / 2.0f, radius * sin((i+1) * 2 * PI / SIDENUMS));
}
glEnd();
glBegin(GL_TRIANGLE_FAN);
glVertex3f(0, -height / 2.0f, 0);
for (int i = 0; i < SIDENUMS + 1; i++)
{
glVertex3f(radius * cos(i * 2 * PI / SIDENUMS), -height / 2.0f, radius * sin(i * 2 * PI / SIDENUMS));
}
glEnd();
And the output is something like image below :
Now in picture , please note on Point A.
This point has drawn two times , one in TRIANGLE_FAN and second in GL_QUADS , My question is that , is this two distinct points welded automatically to each other to make a one point ? or they are on top of each other ? In summary how many points are there ? one or two ? how can I fix if they are not welded to each other?

Answer 1

is this two distinct points welded automatically to each other to make a one point ?
Nope.
or they are on top of each other ?
Yep.
how many points are there ? one or two ?
Two.
You'll have to fuse vertices client-side if that's important to you. There's no magic coalescing in OpenGL.

Answer 2

The triangle fan primitives re-use "points" when creating the triangles, but for quads each call to glVertex3f (...) is generating a single "point."
To put this another way:
When you use triangle fans, the first triangle requires 3 vertices to be transformed, each following triangle only requires 1 new vertex (and reuses 2). This matters for post-T&L cache efficiency, which is the concept you were ultimately getting at in your question.
When you use vertex arrays and glDrawElements (...) or primitive modes that share vertices (e.g. fans and strips) you can avoid having to transform the same point multiple times. If you were using more per-vertex attributes than you have in your example, however, this might not be practical. That is, if each of the sides of the object you are drawing had a different set of texture coordinates or normals for the same spatial location.