>SNIP
>>Ok, I killed it and ran this test: (first time with date() function)
>>
>>local i, ab, lnSec
>>lnSec =seconds()
>>for i=1 to 100000
>> ab=datetime()
>>next
>>?seconds()- m.lnSec
>>clear all
>>lnSec =seconds()
>>for i=1 to 100000
>> m.ab=datetime()
>>next
>>?seconds()- m.lnSec
>>
>>First test: 11.881 vs. 11.900
>>Second test 22.714 vs. 23.472
>
>Nadya,
>
>First, note that your CLEAR ALL eliminates your LOCALs declared earlier. Who knows if this makes much difference.
>
>I tried similarly on a P3 850mhz notebook on Win2K SP2+ for 1,000,000 loops.
>first run- with m. took 1.702 seconds
>---------- without m. took 1.702 seconds
>
>second run - with m. took 1.712 sec.
>------------ without m. took 1.703 sec.
>
>Note that in my tests with m. I used m. EVERYWHERE - in the LOCAL statement, in the FOR and inside the FOR loop. In addition I re-LOCALed the variables after the CLEAR ALL
>
>These things are very hard to test definitively because even a minor variation can cause unknown deviations.
>
>Cheers
Thanks a lot for corrections, Jim.
Actually, the more important test would be to compare the usage m.Var at the right side with no m. at the right side and with table opened in VFP6 and in VFP7. Anyway, I'm not running any more tests today otherwise I never finish this little UDF I'm writing right now :)
If it's not broken, fix it until it is.
My Blog