7c7ccc14   
   From: grahn+nntp@snipabacken.se   
      
   On Sat, 2009-11-21, mockturtle wrote:   
   > On Nov 21, 6:53 am, Le Chaud Lapin  wrote:   
   >> Hi All,   
   >>   
   >> I am looking for sample data on unidirectional inter-node delay using   
   >> UDP.   
      
   ...   
      
   > By the way, if I may ask, do you need "absolute" delay values or are   
   > you more interested in the variability of the delays? (In the latter   
   > case a constant difference in absolute time between the two "clocks"   
   > would not matter).   
      
   If I recall correctly from some experiments I did years ago, if you   
   plot the delta between two system clocks over time, you may get a   
   sawtooth form -- one might run consistently slower than real time, and   
   jump forward to "catch up" now and then.  This was on Linux, and I   
   can't recall what the time scale was (milliseconds or seconds between   
   the "teeth"?)   
      
   I know it botched my delay calculations, but they were just through   
   one router and I needed high resolution.   
      
   My point is: be careful with what you assume in this area. There are   
   more details in keeping time than what you would expect.   
      
   /Jorgen   
      
   --   
     // Jorgen Grahn    O  o   .   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)