I'm having a strange problem with the large interval timer in opennetcf.windowsce. I have an array of these timers that initiate a variety of time based functions in my application. If I only have 1 timer in the array it seems to be OK but when more than one timers are added and enabled it interferes with the system clock causing it to run fast. I have an hourly event that runs about 11 minutes early every time! It is a continuous error - I can see it gaining minute by minute, so it doesn't get corrupted when the time interval elapses.

 If anyone has any ideas I'd be happy to hear about it.

 Thanks in advance

Richard Sinek
 

Wow - that's freaky behavior.  All the LIT does is register for a system notification with a named event and then spawns a worker thread that block on that event, so it's not doing anything with the clock specifically.  We didn't test with multiple LITs running at once and you might want to check that the names being created are unique so you don't get misfired (though IIRC we use a GUID for the name, so you should be ok).  What specific platform are you targeting?  Is it a real device (not an emulator)?  If so, all I can recommend offhand is that you try a simple test with native code that does the same thing and see if the clock issue persists.  If it does then your OEM has an implementation error in the OAL.

 I'm targeting CE 6.0 R2 on an x86 compatible board from ICOP (not the Ebox) and running on the hardware not the emulator. I looked at the SDF source and couldn't see anything that would shed light on the matter. I'm going to test more tomorrow enabling different members of the array to see if I can determine if it is related to the number of timers active or just to specific timers.

 Further investigation reveals that it is not the array of timers that is causing the problem. Instead, each time a time interval elapses the clock skips ahead a second or 2 - I can see it by watching the system clock! Since most of the timers have long intervals the effect was not noticeable except on the timer that runs every minute or so........

Richard 

 Smells even more like a bug in the OAL.

I'm seeing this same behavior on a Symbol MC70. Does anyone have any further ideas about how to deal with this? 

More experimentation indicates this seems to happen when I listen to the PowerManagement.PowerDown, PowerUp, etc. events as well.

 I'm new to OpenNETCF so it's possible I'm doing something wrong. If anyone has any suggestions that would be great. Thanks.