Re: [PHP] Strange benchmark behavior

2007-10-18 Thread Nathan Nobbe
On 10/18/07, Larry Garfield <[EMAIL PROTECTED]> wrote: > > Yeah, it does that sometimes. But over 1 million iterations a 0.2 second > difference is negligible, and easily explained by random events in the > environment (other processes on the box, etc.) It's the 25% difference > I'm > getting on

Re: [PHP] Strange benchmark behavior

2007-10-18 Thread Larry Garfield
Yeah, it does that sometimes. But over 1 million iterations a 0.2 second difference is negligible, and easily explained by random events in the environment (other processes on the box, etc.) It's the 25% difference I'm getting on just one box that is confusing the hell of of me. :-) On Thursd

Re: [PHP] Strange benchmark behavior

2007-10-18 Thread Nathan Nobbe
Larry, i havent dug into the code or the post much cause im coding away today, but i did drop you test script on a box of mine and run it. for what its worth, heres the results: [EMAIL PROTECTED] ~/working/www $ php testBench.php Testing __call() Native Method: 8.62859416008 seconds Magic Method

[PHP] Strange benchmark behavior

2007-10-18 Thread Larry Garfield
Hi folks.  I was hoping that someone here could help me understand a bit of weirdness I've come across.  I'm trying to run some micro-benchmarks on PHP 5's object handling.  My test code is as follows: normal(); } $stop = microtime(true); echo "Native Method: " . ($stop - $start) . " seconds". P