Accounting for Time Dilation

[Coffee Pancake]

Playing with some script and needed to factor time dilation into a timeout .. this seems to do the job.. anyone got anything better?

float time_factor = llGetRegionTimeDilation();
time_factor = (-llLog(time_factor))*(1/time_factor)+1;

time dilation = time_factor
 1.000000 = 1.000000
 0.950000 = 1.053993
 0.900000 = 1.117067
 0.850000 = 1.191199
 0.800000 = 1.278929
 0.750000 = 1.383576
 0.700000 = 1.509536
 0.650000 = 1.662743
 0.600000 = 1.851376
 0.550000 = 2.086977
 0.500000 = 2.386295
 0.450000 = 2.774463
 0.400000 = 3.290729
 0.350000 = 3.999494
 0.300000 = 5.013247
 0.250000 = 6.545184
 0.200000 = 9.047200
 0.150000 = 13.647490

 

[Quistess Alpha]

I'm not sure how you're using that. It would help if you included some testing apperatus. For example, to have a timer last 10 seconds regardless of time dilation:

integer gTimeUnix;
float gTimeWantedTimer = 10;
default
{
    state_entry()
    {
        gTimeUnix = llGetUnixTime();
        llSetTimerEvent(10.0);
    }

    timer()
    {
        integer curTimeUnix = llGetUnixTime();
        llSay(0,(string)(curTimeUnix-gTimeUnix));
        gTimeUnix = curTimeUnix;
        
        // time dilation adjustment test //
        // float time_factor = llGetRegionTimeDilation();
        // time_factor = (-llLog(time_factor))*(1/time_factor)+1;
        // end time dilation adjustment //
        
        llSetTimerEvent(llGetRegionTimeDilation() * gTimeWantedTimer);
    }
}

I'm having trouble finding anywhere that has significant time dilation to test with though. Most time dilation I did see in a few random pops around zindra happened in very short bursts, with dilation remaining somewhat stable around 0.99 to maybe 0.95.

[Lucia Nightfire]

Any exponential influence will do.

You might also have to trend either TD or FPS over some time length as they can change sporadically for a multitude of reasons.

There are also cases where lag is present and TD is normal, but FPS is low.

[Coffee Pancake]

My use case is a plugin based script doing a cold start with full reset of all scripts, and needing to extend a given time out based on a single sampling of time dilation (which may not be representative of average region performance). The time out triggers user feedback and is inconsequential from an operational perspective (any plugin taking longer will not break final operation).

I was really just curious if anyone had a tested more robust method

[Jenna Huntsman]

Could you use some sort of Delta time calculation based on the reported time dilation?

Edit:

What @Quistess Alpha said is a pretty good way of applying Delta time for scripts.

Edited October 3, 2021 by Jenna Huntsman