I am developing system to convert vector coordinates to raster (pixel data). My Linux platform is very limited. I can not use X-Server (GLX extension), probably not GL_EXT_framebuffer_object. simply opengl 1.5. I guess I need to render (off screen into memory) from the coordinates (mostly polygons, lines and points) and extract pixel data (may be I can use glReadPixels). How can i go about developing this system. (using C++). Suggestions are appreciated.

You can do that using Mesa 3D. It is a software implementation of OpenGL 2.

However i can not use any library (only opengl) I can compile C++ code into my library if i get hands on to their codes.

I am close. All i need now is the pixel width and height. I have degree width and height. Actually I am dealing with Long and lattitude. I have East, West, North and South Longitude and lattitude values. I need the pixel width and height using those values.