www.ClassicTW.com
http://classictw.com/

Server Timing Variance (Script Side) GetTime
http://classictw.com/viewtopic.php?f=15&t=21625		 Page 1 of 1
Author:  Kaus [ Wed Mar 04, 2009 1:28 am ]
Post subject:  Server Timing Variance (Script Side) GetTime
Ive been doing some reasearch into the effects of CPC and serverside variance timing wise. My current project entails finding specific times of a servers colo regen. In that 65535's mathmatical regen is not correct client side and I'm not sure why. Go easy on me I'm still a nooblet scripter.

I found that ((65535 / 24 / 60 / 60) = 0.75850694 to infinity)) and in twx 758.50694 respectivly; still is not even perspectivly close to the actual regen timing. So i have attempted to write a script to actually time it and below are my findings. I wanted to write a T1 and T2 variable that detects the moment terra ticks to another number then start the gettimer when the while ='s the third adjusted variable divide by 2 and get my total time. However I have noticed some fluxuation in the script also; in that it doesnt allways catch the first or second change over and ideas on why and how one might accuratly devise the exact or close to exact colo regen rate? I did try a exit variable for If T2 ='s T3 dump, but i had the same timing issues.
Code:
SetVar $Idx 1
setVar $file GAMENAME & "_Seconds.txt"
Delete $file
setPrecision 18
:Start
Send "L 1 * "
Waitfor "There are currently"
Getword CURRENTLINE $Timed 4
Send "q"
Send "L 1 * "
Waitfor "There are currently"
Getword CURRENTLINE $Timed2 4
Send "q"
If ($Timed = $Timed2)
goto :Start
  else
   goto :Cont
end

:Cont
If ($Timed <> $Timed2)
SetVar $Timed3 ($Timed2+1)
getTimer $startTicks
While ($Timed2 <= ($Timed3))
Send "L 1 * "
Getword CURRENTLINE $Timed2 4
Send "q"
Waitfor "There are currently"
end
getTimer $stopTicks
end
setVar $durationTicks ($stopTicks - $startTicks)
setVar $seconds ($durationTicks / 2000000000)
SetVar $seconds2 ($seconds / 2)
echo ANSI_15 & "**Time lapse in Seconds: " $seconds2 "*"
Write $file $seconds2
If ($Idx = 20)
Halt
  Else
Add $Idx 1
End
goto :Start


Net results, where the higher results were captured after the second tick. Most results in my test of 100 fell between .94-1.6 seconds.
Code:
1.007376647000000000
1.172460128750000000
1.060652404249992370
2.010546163250011440
1.632459465250000000
1.614271321500003820
0.996939919000003815
2.624385140999996180
1.600974019250003820
1.631386053250000000
1.560350857249996180
1.093451877250003820
2.038756074249996180
etc.. you get the idea


Also visually at least it appears some colo ticks are much longer than the previous. (using the most common timing in my test of 100)
Code:
:start
Setdelaytrigger land :land 1001
pause
:Land
Send " l 1 * q"
goto :start
Author:  Big D [ Wed Mar 04, 2009 7:13 am ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
The actual regen on Terra is a random formula. It will reproduce a total regen of whatever the figure is set at, but the actual regen amount may vary throughout the day. At times Terra might produce 2000 colonists per hour and at times it may produce 8000 colonists per hour. So if Terra produces a lot of colonistes earlier in the day, you can bet it's going to slow way down later in the day.
Author:  Kaus [ Wed Mar 04, 2009 11:13 am ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Big D wrote:
The actual regen on Terra is a random formula.


It seems like pretty much everything sent from the server has some sorta random variance to it...

Geez.
Author:  Singularity [ Wed Mar 04, 2009 11:43 am ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Do this... instead of using cpu ticks to measure things, use getTime...

Linky:
http://www.navhaz.com/files/script.html#CMD_26

CPU ticks aren't always accurate, it'd be good to have it confirmed thru the clock.
Author:  Kaus [ Wed Mar 04, 2009 12:26 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Singularity wrote:
Do this... instead of using cpu ticks to measure things, use getTime...

Linky:
http://www.navhaz.com/files/script.html#CMD_26

CPU ticks aren't always accurate, it'd be good to have it confirmed thru the clock.


Confirmed I guess, allthough not as accurate the gettime command shows the variance also.
Code:
:Cont
If ($Timed <> $Timed2)
Gettime $start ss
SetVar $Timed3 ($Timed2+1)
While ($Timed2 <= ($Timed3))
Send "L 1 * "
Waitfor "There are currently"
Getword CURRENTLINE $Timed2 4
Send "q"
end
Gettime $stop ss
end
setVar $duration ($stop - $start)
Divide $duration 2
echo ANSI_15 & "**Time lapse in Seconds: " $duration "*"


Net Results
Code:
1
1
2
2
2
3
1
1
1
1
etc...
Author:  Promethius [ Wed Mar 04, 2009 1:06 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
You can get millisecond accuracy with gettime:

gettime $msEnd zzz

The "zzz" isn't in the help docs, but EP put me onto its existence as part of the Delphi function. The original snippet I had was:

gettime $msEnd "h:m:s:zzz"
replaceText $msEnd ":" " "
getWord $msEnd $ms2 4
Author:  Singularity [ Wed Mar 04, 2009 1:23 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Ya, use milliseconds... hehe. You might want to invert the test, too. Land every X seconds and log the regen (difference from last).
Author:  Kaus [ Wed Mar 04, 2009 2:00 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Abit more accurate with the zzz settings Prom tipped me off on. Still netting the same results which basically confirms some other reasearch I had previously done using other commands. I will try the reverse one you are mentioning Sing at a later point. I have noticed there are less "blips" (read: large tick exceptions) using a gettime over a gettimer.

Recoded and recycled, instead of just timing the time between colo regens at terra I expanded to allow for total time and regen using the gettime command. Purpose was to both verify my previous findings and to verify the random regen timer. Both have in my eye been satisfied. It appears that indeed terra regen is random within a 1000ms variable (generally). There is prolly some redundant code within the following example so I appologise in advance, I tried my best to remove and replace anything I thought or saw to be repetative. Going abit crosseyed here from all the gettime commands.. hehe

Code:
#-------------INT Variables--------------
SetVar $Idx 1
setVar $file GAMENAME & "_Seconds.txt"
Delete $file
#------------Check to see if we can start-----
:LetsGo
Gettime $StartDaTest
ReplaceText $StartDaTest ":" " "
Getword $StartDaTest $Started 2
SetDelaytrigger EMCPU :EMCPU 100
pause

:EMCPU
Gettime $StartDaTest
ReplaceText $StartDaTest ":" " "
Getword $StartDaTest $Started2 2

If ($Started = $Started2)
Echo ANSI_15 "* Waiting to go!"
  goto :LetsGo
End
Echo ANSI_12 "* Started The Test!!!!"


GetTime $StartedTest

Gettime $Stop
Replacetext $Stop ":" " "
Getword $Stop $Stoploop 2

#--Change this variable to increase time of test---
Add $Stoploop 1

#------------While Start-------------------
While ($whileloop < $stoploop)
:Start
Send "L 1 * "
Waitfor "There are currently"
Getword CURRENTLINE $Timed 4
Send "q"
Send "L 1 * "
Waitfor "There are currently"
Getword CURRENTLINE $Timed2 4

If ($Idx = 1)
SetVar $StartColo $Timed
end

Send "q"
If ($Timed = $Timed2)
goto :Start
  else
   goto :Cont
end

:Cont
If ($Timed <> $Timed2)
Gettime $start s:zzz
SetVar $Timed3 ($Timed2+1)
While ($Timed2 <= ($Timed3))
  Send "L 1 * "
   Waitfor "There are currently"
   Getword CURRENTLINE $Timed2 4
  Send "q"
end
Gettime $stop s:szzz
end

Replacetext $Start ":" " "
Replacetext $Stop ":" " "

Getword $start $SecTimer 1
Getword $start $MilliSecTimer 2
Getword $stop $SecTimer2 1
Getword $stop $MilliSecTimer2 2

setVar $SecDuration ($SecTimer2 - $SecTimer)
SetVar $MilliSecDuration ($MilliSecTimer2 - $MilliSecTimer)
Divide $SecDuration 2
Divide $MilliSecDuration 2
Write $file "Time lapsed " & $SecDuration & ":"  & $MilliSecDuration
Add $Idx 1

GetTime $Check
Replacetext $check ":" " "
Getword $Check $Whileloop 2
END

#-------------Ended While Loop Spit out the Results---------------------
SetVar $StopColo $Timed2
GetTime $EndedTest
echo ANSI_15 "* " ($StopColo - $StartColo) " :Total Colos Regened" "*"
echo ANSI_12 "*" $StartedTest "*" $EndedTest
Halt


Results
Code:
Time lapsed 1:1953
Time lapsed 2:6570
Time lapsed 1:8047
Time lapsed 1:9961
Time lapsed 1:13390
Time lapsed 1:15945
Time lapsed 1:19000
Time lapsed 1:20508
Time lapsed 1:23469
Time lapsed 1:26008
Time lapsed 1:28046


Total Time 1 Minute Cycle
Code:
50 :Total Colos Regened

3:20:00 PM
3:21:00 PM


Total colos regenned over 1 hour in 1 minute increments would be 72000 however to truely test it I would need to run a full hour and my attention to this project is just about spent. I believe as the script nears 60mins the 72000 would be allot closer to the stated 65535 in the twgs/*.

Thanks for humoring me :mrgreen:
-Kaus
Author:  Kaus [ Wed Sep 17, 2014 6:05 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Necro post for CC Bee and because I felt like it.

Can use the above get-time for other commands as well.
Author:  LoneStar [ Wed Sep 17, 2014 8:52 pm ]
Post subject:  Re: Server Timing Variance (Script Side) GetTime
Kaus wrote:
Necro post for CC Bee and because I felt like it.

Can use the above get-time for other commands as well.


Hey Kaus! like seing your avatar.. you need to post more often 8)
Page 1 of 1 All times are UTC - 5 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/