So, I'm investigating Object Pools, and I think I'm not getting something. The documentation isn't great, hence the question We have a lot of small objects that we will use throughout our application's lifespan, and I'm trying to see if ObjectPool is a good fit for us. Heck, I'm just trying to look at performance of it in general. So, I have the following code: -------------------------------------------------------------- void testNormal() { common::Timer t; t.start(); for(int i = 0; i < MAX_NUM_TESTS; i++) { testClass t; t.doWork(); } t.stop(); std::cout << "No Pool: Creating " << MAX_NUM_TESTS << " took " << str_cast(t.getTotalElapsedTime()) << std::endl; } void testPool() { common::Timer t; t.start(); boost::object_pool<testClass> pool; for(int i = 0; i < MAX_NUM_TESTS; i++) { testClass::ptr tc = pool.construct(); tc->doWork(); } t.stop(); std::cout << "Pool: Creating " << MAX_NUM_TESTS << " took " << str_cast(t.getTotalElapsedTime()) << std::endl; } int main(int argc, char* argv[]) { testNormal(); testPool(); return 0; } ------------------------------------------------ When run, I am getting: No Pool: Creating 500000 took 00:00:00.187785 Pool: Creating 500000 took 00:00:00.281679 So, I guess I'm not understanding the advantage of the pools. Could someone write/modify my example to make it make more sense? Thanks --dw