wscript.sleep 1 and wscript.sleep 2
Don't download the above file. I tried myself, but got error message from my
zip application. I have no idea of what happened. The zip file is opened
well in my machine, but if I upload/download it, it always doesn't work.
Mystery goes on...
--
Have a nice day.
Han Pohwan, Microsoft MVP, Korea
Quote:
> Thanks Michael. If you're interested, check this file.
> http://www.megapass.co.kr/~hp4444/fun.zip
> I couldn't reproduce the same situation with a simple loop. There is a
file
> named build.vbs and a folder. There is a line wscript.sleep commented
> 'MICHAEL ~. The whole execution will take about 1 minute. You can check
the
> CPU usage in the mean time. If you change the value to /wscript.sleep 2/,
> the CPU usage will be more than 30%.
> Anyone may/need not know what the all the project is. My and your only
> concern should be the sleep, and the interesting phenomenon.
> --
> Have a nice day.
> Han Pohwan, Microsoft MVP, Korea
> > > I'm meeting strange phenomenon.
> > > do until condition
> > > do something
> > > wscript.sleep 1
> > > loop
> > > The loop takes 10 to 20 minutes, and about 50,000 iterations. If I
> > > don't place /wscript.sleep/, CPU usage is 99-100%. If I do
> > > /wscript.sleep 1/, the usage is 7-10%. Funny is that if I do
> > > /wscript.sleep 2/, the usage is always 37-40%. I tested again and
> > > again, but the result is always same. Other than CPU usage, the
> > > performance is same with/without /wscript.sleep/. If I apply 2
> > > instead of 1, the performance is a bit degraded as expected. What I'm
> > > curous is CPU part. What happened in /2/ instead of /1/. Is /1/ magic
> > > number or something?
> > Here's a rather simplistic test case. Using different values for
> msecSleep,
> > I didn't see any appreciable difference in overall CPU usage as
observered
> > using TaskManager. The only real effect I could see, is the expected
> > reduced number of iterations through the time limited loop as the
> msecSleep
> > value increased.
> > An interesting observation was that on a W2K Pro sp2 box with WSH 5.6,
> > incrementing msecSleep by 10 affects the number of iterations within the
> 10
> > second max, but increments by 1 do not.
> > msecSleep = 10 'milliseconds to sleep per iteration
> > secDelay = 10 'max seconds for loop to execute
> > t1 = timer
> > for n = 1 to 1000000
> > wscript.sleep msecSleep
> > e1 = timer() - t1
> > if e1 > secDelay then i = n : exit for
> > next
> > wscript.echo i, e1
> > --
> > Michael Harris
> > Microsoft.MVP.Scripting
> > Seattle WA US
