Yes and no. Timing CPU-intensive operations as medium of computing maciner performance is the best way, if you do it right, which is the problem.
The second problem comes into play when you get to the stage of doing something useful with that data... For a performancetesting routine to be reliable you would need a few-second routine, executed atleast three times (to minimize risk of wrong values), scrap the longest and shortest times, and you have a value of milliseconds. Now, what do you want to use it for? I guess one typical thing is making a blend sequence last as long on different machines. Now you have the problem that you need to create a routine that makes sense with any value. I mentioned earlier that it should be a few - second routine, on your machine. That would make the routine last up to 15-20 seconds on below-spec - machines, and you'd also have to take into account those 0,2 terahertz Pentium9's which someone might try to run it on, leaving you a range of >20 000 ms to <200 ms which you would have to play with Well, lets get back to that blend sequence, you design a blend sequence in let's say, steps of five on your machine which gives a time for the performance test of 2000 ms. That would leave you with a low-spec machine taking steps of 50 and a future machine taking steps of 0.5 . Doing something like this, and doing it right is an excessively work-intensive task. Pekka > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On > Behalf Of Kurt Griffin > Sent: 9. marraskuuta 2001 5:44 > To: [EMAIL PROTECTED] > Subject: Re: <lingo-l> processor speed > > > > Is there any way to test the users processor and OS with lingo > and not an Xtra > > OS, yes (assuming Dir 8.5) - look up "the Environment" in the help. > Processor, no, but there are ways to see how fast things are going... I > think it was James Newton that posted an answer to that within > the last few > days. The essence was, start a timer, run a processor intensive operation, > check the timer - really a much better approach than trusting in > a processor > version, since there are many other variables than can slow down the > processing of your stuff. > > -Kurt > > > [To remove yourself from this list, or to change to digest mode, go to > http://www.penworks.com/LUJ/lingo-l.cgi To post messages to the list, > email [EMAIL PROTECTED] (Problems, email [EMAIL PROTECTED]) > Lingo-L is for learning and helping with programming Lingo. Thanks!] > [To remove yourself from this list, or to change to digest mode, go to http://www.penworks.com/LUJ/lingo-l.cgi To post messages to the list, email [EMAIL PROTECTED] (Problems, email [EMAIL PROTECTED]) Lingo-L is for learning and helping with programming Lingo. Thanks!]
