By Andrey_Karpov_2009
via rmartinho.github.com
Submitted: Dec 06 2012 / 07:37
On part 1 we saw how two common std::tuple implementations could result in a sub-optimal layout, and how an optimal layout could be achieved. Now we’ll see what an implementation of a tuple with such an optimal layout actually looks like. (Note: all this could have benefited from the Boost.MPL library, but I decided not to use it in order to expose more ideas)
Add your comment
