--- In [email protected], "brucexs" <[EMAIL PROTECTED]> wrote: > > I'd be interested in some time comparisons for a few scripts. Nothing > else should be using CPU on the system when you test (and I'd still > average over 2-3 trials). >
Hi Bruce, I'm not able to get the callouts, callbacks, cb's etc., in the regexPluginTest script to work without being file-based. So I commented out all the tests that use those. This script still has file i/o via debug logging to file. In any event, here are three executions each of the same code, inside and outside the pcf. Its a pretty long script (not even one I'd choose to keep in the pcf), and I didn't shut down everything else that was running (though nothing was cpu intensive). I did use an empty pcf, so there were no events, timers, caption bars, etc., being processed simultaneously: File based: MYregexPluginTest elapsed time: 7.39668 MYregexPluginTest elapsed time: 7.47069 MYregexPluginTest elapsed time: 7.5327 Command List based: regexPluginTest elapsed time: 6.54252 regexPluginTest elapsed time: 6.4677 regexPluginTest elapsed time: 6.55265 Remember too that function statements had been moved to the top of this script file and it was much much slower before that happened. I believe the whole numbers are seconds, it uses your code, ((perfcount-start.tofloat)/perffreq).format) I have transferred most of my frequently called scripts into my primary pcf, and it feel more responsive. If I get around to timing executions I'll let you know, but seeing as the scripts are so much shorter, I'm sure the times would look less significant. I'm running an old athlon 1.2 ghz (not the fastest by far these days) and xp pro sp2. Regards, Sheri
