[CSC 315] Re: CSC315

[Andrew J. Pounds]

Here's the deal -- for this case generate a huge array based on your 
display size.  If you have a 1024x768 display, assume that you need the 
maximum number of points to be 1024*2*PI=6434.  If you generate that 
many points, your pixels should look continuous wherever, regardless of 
the size of  the circle.  Don't stress about performance at this point 
-- we'll talk more about that later. 

When I grade I  will be verifying that  the Sutherland-Hodgman algorithm 
is being used.

David Thomas wrote:
> Thanks, that clears up a lot!
>
> The question I still have is this: approximating the circle with a 
> polygon with vertices taken from points along the circle will change 
> which pixels are chosen, unless EVERY pixel chosen by the straight 
> midpoint-circle-scan-converting algorithm is used as a vertex (which 
> would certainly be slower than scissoring). Is it ok that we are now 
> varying our answers? In other words, different choices of n will 
> result in a different set of pixels being chosen for a given circle.
>
> So, what IS an appropriate choice of n? How good does our polygon 
> approximation have to be? Is that just up to us?
>
> Oh, and yes, an extra day would be GREAT...and I'm pretty sure 
> everyone else would agree with me, too ;)
>
> ~David


-- 
Andrew J. Pounds, Ph.D.  (pounds at theochem.mercer.edu)
Associate Professor of Chemistry and Computer Science
Mercer University,  Macon, GA 31207   (478) 301-5627