[Xr] Help: Render error on XrFormatRGB32

[Keith Packard]

Around 13 o'clock on May 4, Owen Taylor wrote:

> Certainly, I doubt the exact sample count is going to be easier to
> compute from edge values than the exact coverage area. So, I'd
> vote for just stepping fifteen times.

At least for now, that seems like the best idea.  A more efficient 
implementation can easily be added in the future.  We can also add an 
'inexact' mode with a reduced number of samples and see if that has a 
measurable impact on performance.

Of course, help in coding this up would be greatly appreciated...

-keith