
2 Nov
2008
2 Nov
'08
12:46 a.m.
On Saturday 01 November 2008 16:27, Frank Mori Hess wrote:
As requested earlier by Vincent, I threw together a little benchmark program (attached) comparing the speed of boost.signals with signals2. It produces the following output on a P4 running Linux:
boost::signals2::signal, 10 connections, invoking 1000000 times: 0.77 s
boost::signal, 10 connections, invoking 1000000 times: 0.91 s
Oops, there was an error in the benchmark program, the above two times for 10 untracked connections were reversed. The 0.77s time belongs to boost::signal, and the 0.91s time is boost::signals2::signal.