Le 23/12/2010 00:50, Aidy nous fit lire :
> Hiya
> 
> I'm doing a final year project at university working with povray and need to
> work with the source code but due to its sheer size I have no idea where to
> start and I've been unable to find any diagrams that help with this.
> 

Have you tried to apply some tools on the sources ?
As blind as it could be, doxygen might already give you some clues.
(but it won't explain how it happens, and you'd better investigate the
boost::thread (or whatever their names))

> I was wondering if there were any class diagrams or flow diagrams showing the
> structure of the program or the execution process. If not, what is the entry
> point of the program? Is it the povray.cpp file in backend?

The entry point (main) is in vfe/unix/unixconsole.cpp (for the unix
port). Windows' entry point as usual...
Final year project in CS ? I'm afraid, very afraid. You will learn a
lot, or not.

> 
> Secondly, what would be the easiest way of profiling povray and finding out
> where the bulk of the processing is done?

Learn your tools.
what about investigating g++ and its -pg option ?

You might be in for some delusion: this is not a triangle renderer, the
time is spent on various computations which depends on the scene to
render. If you put in some polynomial, you would spend a lot of time in
a polynomial solver iterating to find the roots... every scene is different.

> And also, how close is the current
> beta to being complete? As it's design for multi-core processors is ideal for my
> work but I don't want to do loads of work and then have it all change due to the
> final release changing things.

Load of works should be automated for your project (for raw data
production at least) if you cannot sell a beta code to your supervisor.
Data analysis might be checked for conclusion when raw data are updated.
(what if your year ends before the final release ? or if it get out 3
days before the presentation's deadline ?)

Post a reply to this message