 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Java Forum / First Aid / September 2004HELP: How to Get Accurate Time?
Hi! I am trying to get system time in accurate milliseconds units (for example 582ms). I use the method System.currentTimeMillis() in jdk 1.2.2, on Windows 98 platform. I manage to get accuracy of up to 10 milliseconds units only (for example either 580 ms 0r 590 ms). How can I get higher resolution? Thanks a lot for your advice! Amit > I am trying to get system time in accurate milliseconds units (for example > 582ms). > I use the method System.currentTimeMillis() in jdk 1.2.2, on Windows 98 > platform. I manage to get accuracy of up to 10 milliseconds units only (for > example either 580 ms 0r 590 ms). How can I get higher resolution? Change your OS. AFAIU, Java's ability to cut time into small slices is entirely dependent on the clock of the underlying OS. For older Windows, for example, I seem to recall a time granularity of 55 msec..  SignatureAndrew Thompson http://www.PhySci.org/ Open-source software suite http://www.PhySci.org/codes/ Web & IT Help http://www.1point1C.org/ Science & Technology >> I am trying to get system time in accurate milliseconds units (for >> example 582ms). [quoted text clipped - 8 lines] > For older Windows, for example, I seem to > recall a time granularity of 55 msec.. Hi Andrew, Just out of curiosity: Did you find this 55 msec granularity also on comps with a different CPU-speed or architecture? Another way to put it: Do you think it depends on the OS or the architecture/CPU-speed ? Or maybe both? Regards, Erwin Moller > Another way to put it: Do you think it depends on the OS or the > architecture/CPU-speed ? Or maybe both? ..errr. I'm getting onto ever thinner ground here. I have not dealt with the time/granularity issue directly. OTOH, I do recall hearing that when Win was stuck at 55 msec granularity, the Linux folks were boasting (and yes I think it was..) 10 msec on the Linux OS. That *suggests* to me, that the '10 msec' is set by the underlying hardware, while the '55 msec' was a quirk of the OS.. Though now I think about it, I cannot figure why the longer delay is not an integral number of the 10 msec (50msec/60msec). Anyone? SignatureAndrew Thompson http://www.PhySci.org/ Open-source software suite http://www.PhySci.org/codes/ Web & IT Help http://www.1point1C.org/ Science & Technology > > Another way to put it: Do you think it depends on the OS or the > > architecture/CPU-speed ? Or maybe both? [quoted text clipped - 11 lines] > it, I cannot figure why the longer delay is not > an integral number of the 10 msec (50msec/60msec). According to Roedy... http://mindprod.com/jgloss/time.html > > I am trying to get system time in accurate milliseconds units (for example > > 582ms). [quoted text clipped - 5 lines] > cut time into small slices is entirely > dependent on the clock of the underlying OS. Yes. But in Windows, Java calls GetSystemTime function to get the time, instead of the high-resolution counter function QueryPerformanceCounter, which has been there since Win95 and WNT3.1. > For older Windows, for example, I seem to > recall a time granularity of 55 msec.. This was due to an MS-DOS clock that was limited to 18.2 Hz IIRC. > Hi! > > I am trying to get system time in accurate milliseconds units (for example > 582ms). You can't do this without replacing your computer and your operating system. In the bad old days of personal computers, the time slice was 1/18.2 second, or 54.9 milliseconds. Things may have improved somewhat since then, but this time slice doesn't decrease in size as the computer's processor clock increases in speed -- it is related to the real-time clock more than to the CPU clock. And I don't think there is any great pressure to decrease the time slice value on typical systems. > I use the method System.currentTimeMillis() in jdk 1.2.2, on Windows 98 > platform. I manage to get accuracy of up to 10 milliseconds units only > (for example either 580 ms 0r 590 ms). How can I get higher resolution? See above. Why not state the problem to be solved? It's important to realize we are not discussing real-time operating systems, and a long interrupt by a process with higher priority than yours will most certainly ruin all your plans anyway. SignaturePaul Lutus http://www.arachnoid.com Thank you Andrew, Erwin, Matt and Paul! You have been of great help. I will try the solution offered by Matt. See if it does any good. Again, thanks a lot - I did not expect such a warm wecome to the group! Amit >> Hi! >> [quoted text clipped - 23 lines] > your > plans anyway. > Thank you Andrew, Erwin, Matt and Paul! You have been of great help. > > I will try the solution offered by Matt. See if it does any good. > > Again, thanks a lot - I did not expect such a warm wecome to the group! And another warm welcome to the group! I see you already have a solution, but I'd like to point you to the following thread: http://groups.google.com/groups?q=timer+group:*.java+author:princess+author: morgiah&hl=nl&lr=&ie=UTF-8&selm=YAK3c.30096%24f01.2220062%40phobos.telenet-o ps.be&rnum=1 At this moment I haven't got much time (working double shifts here), but should you be interested feel free to mail me or reply to get the source code for the timer. Regards, Princess Morgiah Free MagazinesGet these publications absolutely FREE for up to 12 months. There are no hidden fees and no obligation. Simply choose a title, complete the application form and submit it. Read more ...
|
Reply to this Message
|
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.