CGAL users discussion list

[Marc Glisse <>]

On Fri, 15 Jan 2016, s_gandhi wrote:

> I did turn on the optimization, set the level to O3.
> In fact, I checked on multiple platforms, on windows as well,
> but still I get a time improvement.
>
> I am not that knowledgeable about compiler optimization.
> So not sure if function wrappers is the reason.
>
> But it boils down to:
> Is using custom codes okay? What about these Editorial Board?
> Ideally it should be allowed, it opens the doors to other ways of
> implementation, which might prove to be efficient. But this is my
> perspective, and would like to know from the people experienced with CGAL

If you turned optimizations on, then it is probably not the wrappers 
taking time. Did you profile the code to see where time is going?

What kernel are you using when you notice this "slowness"? If it is 
Exact_predicates_inexact_constructions_kernel, I think you need to read 
http://www.cgal.org/exact.html .

Without looking at your code, I would assume that what you propose is 
similar to Simple_cartesian<double>.

Otherwise, I would need more details about what you are doing to give more 
information.

--
Marc Glisse