I need help, guys!
I have a parallelization issue,
Simply what I want to do is that I want to write a code which reads a 3dxml file (using CUDA) and converts it to an appropriate data structure in C++. After obtaining the data structure I want to draw the structure into the screen efficiently.
These are my questions:
1- Doing such a thing is “ok”? I mean I don’t wanna write something (which is useless) from scratch ! (By the way, it might be good for me to do it considering that I’m currently learning openGL etc.)
2- Assuming that I start writing the code, how can I use the obtained data structure by means of the openGL? How can I do it efficiently? Or, should I directly use Xerces or some XML parser for this purpose instead of parallelization of reading the 3DXML file?
3- I don’t know how to efficiently redraw the shape (3dxml file) for every rotation command coming from the mouse input? I mean how to give this data structure’s body to openGL for further handling regarding throughput…
4- this is the last one! Assume that we are working with a 3D file which is huge (such as complete 3D representation of a satellite), so, can the parallelization idea that I propose here be handy for that purpose? (this question is actually just the extension of the first question)…
I’m sorry if I confuse the concepts because I’m neither a native openGL programmer nor a computer scientist but a mechanical engineer with a “little” knowledge of openGL but a considerable knowledge of CUDA.
Thank you for your replies, people!!! I appreciate the “openGL folk” because they are providing such a platform for enabling me asking questions in a huge community of “guru” openGL programmers.