port to OSX

[jottinger]

There's presently a barrier with syscall.Nanosleep() in delay/udelay.go; this is likely to be something appropriate for Linux and Windows but not OSX.

[misterunix]

@jottinger I do not have a Mac to test with. Since the OS is based on BSD I guessed that the UNIX delay would work.

In udelay the compiler is instructed to use udelay for linux and darwin.
//go:build linux || darwin

Can you clone and run the below build and test? I would greatly appreciate it. I am assuming that you want to help. :)

go build
./basicbots -b robots/blaster.bas robots/rook.bas robots/nexus.bas robots/corner-v2.bas

[jottinger]

With the change from this PR, that call works and it compiles. I have not tried it on Linux; I have a FEELING that the same code might work for Windows, too.

[misterunix]

time.Sleep(time.Duration(d)) doesnt work as expected.
Sleep was my first attempt to slow the battledisplay down and I quickly found that Sleep is not accurate. Sleep is based on the timer functions of the host OS and guarantees that at least X amount of time has passed, and sleep for X amount time. Using Sleep no precision can be had thus battledisplay will vary wildly with OS and CPU speed. That is the reason for having two different Delay functions.
Let me see if I can find some accurate delay method that works on a Mac.

This is the testing code I wrote when I noticed that the delay was not what I expected.

package main

import (
	"fmt"
	"time"
)

func main() {
	for t := 20; t <= 500; t = t + 10 {
		cc(t)
	}
}

func cc(n int) {
	var t float64
	for i := 0; i < 100; i++ {
		s := time.Now()
		time.Sleep(time.Duration(n) * time.Millisecond)
		e := time.Since(s)
		t = t + float64(e)
	}
	t1 := t / 100.0
	nd := time.Duration(n) * time.Millisecond
	r := t1 / float64(nd)
	fmt.Printf("n=%d t1=%s %s r=%f\n", n, time.Duration(t1), nd, r)
}

[misterunix]

@jottinger do you have a development environment that you can work in? It would be nice if we could work together on this.
You can create a new branch off of the working-branch and make a mdelay.go and Delay() function.

[jottinger]

Yeah, but note that I'm a go newbie myself. My help's not going to be as useful as one might hope, I fear. And as far as sleep is concerned... sleep AT LEAST n is pretty much what you want, in most cases, unless you're talking realtime responsiveness. And for this... I'm not thinking the granularity is as important as it might seem.

…

On Tue, Oct 12, 2021 at 4:28 PM Bill Jones ***@***.***> wrote: @jottinger <https://github.com/jottinger> do you have a development environment that you can work in? It would be nice if we could work together on this. You can create a new branch off of the working-branch and make a mdelay.go and Delay() function. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#44 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AADSNGQLKBY7KBE6UAMSCRLUGSK7RANCNFSM5F2UJCQQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- Joseph B. Ottinger http://www.enigmastation.com <http://enigmastation.com> *To the beautiful and the wise,* * The mirror always lies.*

[misterunix]

Since it has to be accurate Sleep just doesn't work.
cycledelay = int64(time.Microsecond * 100) is set so each update to the battlescreen is set to a fixed amount of time.
Sleep(time.Microsecond*100) on a Windows machine translates into around 90ms. Way too slow. This is because of the way Go gets its timer from Windows.

I know it sounds counterintuitive, but a close look at the Go code base shows that it is "at least" and that "at least" can vary wildly.

Try the program above, I would be interested in its output on your Mac AND yes you can be very useful. Questions make people think and you have a Mac :) !!!

[jottinger]

On my mac, I get this output:

n=20 t1=23.060447ms 20ms r=1.153022
n=30 t1=33.1492ms 30ms r=1.104973
n=40 t1=42.949334ms 40ms r=1.073733
n=50 t1=53.011928ms 50ms r=1.060239
n=60 t1=63.386924ms 60ms r=1.056449
n=70 t1=73.430803ms 70ms r=1.049011
n=80 t1=83.41618ms 80ms r=1.042702
n=90 t1=93.649088ms 90ms r=1.040545
n=100 t1=103.365969ms 100ms r=1.033660
n=110 t1=113.120322ms 110ms r=1.028367
n=120 t1=123.319004ms 120ms r=1.027658
n=130 t1=133.286679ms 130ms r=1.025282
n=140 t1=143.236877ms 140ms r=1.023121
n=150 t1=153.260373ms 150ms r=1.021736
n=160 t1=163.315784ms 160ms r=1.020724
n=170 t1=173.396067ms 170ms r=1.019977
n=180 t1=183.291796ms 180ms r=1.018288
n=190 t1=193.316887ms 190ms r=1.017457
n=200 t1=203.697169ms 200ms r=1.018486
n=210 t1=213.42764ms 210ms r=1.016322
n=220 t1=223.272853ms 220ms r=1.014877
n=230 t1=233.307603ms 230ms r=1.014381
n=240 t1=243.166472ms 240ms r=1.013194
n=250 t1=253.239751ms 250ms r=1.012959
n=260 t1=263.453938ms 260ms r=1.013284
n=270 t1=273.340633ms 270ms r=1.012373
n=280 t1=283.10929ms 280ms r=1.011105
n=290 t1=293.301127ms 290ms r=1.011383
n=300 t1=303.597452ms 300ms r=1.011992
n=310 t1=313.73442ms 310ms r=1.012047
n=320 t1=324.20501ms 320ms r=1.013141
n=330 t1=334.260835ms 330ms r=1.012912
n=340 t1=344.222049ms 340ms r=1.012418
n=350 t1=354.148946ms 350ms r=1.011854
n=360 t1=364.22713ms 360ms r=1.011742
n=370 t1=374.168281ms 370ms r=1.011266
n=380 t1=384.239205ms 380ms r=1.011156
n=390 t1=394.326768ms 390ms r=1.011094
n=400 t1=403.662473ms 400ms r=1.009156
n=410 t1=413.377247ms 410ms r=1.008237
n=420 t1=423.342602ms 420ms r=1.007959
n=430 t1=433.540962ms 430ms r=1.008235
n=440 t1=443.344831ms 440ms r=1.007602
n=450 t1=453.558451ms 450ms r=1.007908
n=460 t1=463.48058ms 460ms r=1.007566
n=470 t1=473.319817ms 470ms r=1.007063
n=480 t1=483.357255ms 480ms r=1.006994
n=490 t1=493.338922ms 490ms r=1.006814
n=500 t1=503.259991ms 500ms r=1.006520

On my Linux server, this is my output:

n=20 t1=20.66124ms 20ms r=1.033062
n=30 t1=30.673456ms 30ms r=1.022449
n=40 t1=40.582233ms 40ms r=1.014556
n=50 t1=50.545883ms 50ms r=1.010918
n=60 t1=60.688389ms 60ms r=1.011473
n=70 t1=70.664456ms 70ms r=1.009492
n=80 t1=80.717594ms 80ms r=1.008970
n=90 t1=90.759626ms 90ms r=1.008440
n=100 t1=100.506618ms 100ms r=1.005066
n=110 t1=110.578469ms 110ms r=1.005259
n=120 t1=120.553772ms 120ms r=1.004615
n=130 t1=130.839973ms 130ms r=1.006461
n=140 t1=140.840899ms 140ms r=1.006006
n=150 t1=150.83553ms 150ms r=1.005570
n=160 t1=160.743376ms 160ms r=1.004646
n=170 t1=170.600541ms 170ms r=1.003533
n=180 t1=180.834481ms 180ms r=1.004636
n=190 t1=190.917028ms 190ms r=1.004826
n=200 t1=200.926884ms 200ms r=1.004634
n=210 t1=211.119796ms 210ms r=1.005332
n=220 t1=221.161101ms 220ms r=1.005278
n=230 t1=230.614879ms 230ms r=1.002673
n=240 t1=240.83073ms 240ms r=1.003461
n=250 t1=251.195316ms 250ms r=1.004781
n=260 t1=261.757715ms 260ms r=1.006760
n=270 t1=270.976119ms 270ms r=1.003615
n=280 t1=281.14837ms 280ms r=1.004101
n=290 t1=291.579641ms 290ms r=1.005447
n=300 t1=301.560886ms 300ms r=1.005203
n=310 t1=310.974859ms 310ms r=1.003145
n=320 t1=321.155036ms 320ms r=1.003609
n=330 t1=330.690468ms 330ms r=1.002092
n=340 t1=340.674943ms 340ms r=1.001985
n=350 t1=350.43138ms 350ms r=1.001233
n=360 t1=360.699831ms 360ms r=1.001944
n=370 t1=370.527782ms 370ms r=1.001426
n=380 t1=380.940983ms 380ms r=1.002476
n=390 t1=390.712928ms 390ms r=1.001828
n=400 t1=401.031182ms 400ms r=1.002578
n=410 t1=411.220067ms 410ms r=1.002976
n=420 t1=421.307519ms 420ms r=1.003113
n=430 t1=431.151084ms 430ms r=1.002677
n=440 t1=440.729659ms 440ms r=1.001658
n=450 t1=450.928207ms 450ms r=1.002063
n=460 t1=461.507294ms 460ms r=1.003277
n=470 t1=470.938088ms 470ms r=1.001996
n=480 t1=480.782967ms 480ms r=1.001631
n=490 t1=490.899532ms 490ms r=1.001836
n=500 t1=500.703339ms 500ms r=1.001407

What output do you get on your various platforms?

This shows a consistent granularity on both OSX and Linux; I can't speak for Windows, but for a game like this, I am not thinking the granularity showed here is statistically significant as a negative, compared to the advantage in simplicity. (I ran the competition on OSX, with my timer code, for example, and it ran fine; a little quickly, really, but eh, whatever.)

[jottinger]

On the M1, I get slightly different performance:

n=20 t1=24.696774ms 20ms r=1.234839
n=30 t1=34.567219ms 30ms r=1.152241
n=40 t1=44.083645ms 40ms r=1.102091
n=50 t1=54.379435ms 50ms r=1.087589
n=60 t1=64.357994ms 60ms r=1.072633
n=70 t1=74.590431ms 70ms r=1.065578
n=80 t1=84.458738ms 80ms r=1.055734
n=90 t1=94.597402ms 90ms r=1.051082
n=100 t1=104.445277ms 100ms r=1.044453
n=110 t1=114.638244ms 110ms r=1.042166
n=120 t1=124.603832ms 120ms r=1.038365
n=130 t1=134.720303ms 130ms r=1.036310
n=140 t1=144.50619ms 140ms r=1.032187

(I am still running the rest of the app, it just is being consistent with the numbers and it takes a while to run, obviously.)

In all cases, the performance of time.Sleep is conformant with the spec (i.e., "at least" the duration, with a consistent delta) - Windows seems to be the outlier here. My thought would be to use the simpler code on linux and darwin, and have a custom delay for Windows, since time.Sleep() doesn't seem to work properly on that platform.

[misterunix]

okay, little changes to the timing testg to show some of the deltas and FPS. Keep in mind the FPS is a perfect world. It doesn't take into account the actual delay due to the interpreter running on X number of robots.

Here is the changed code. My results will be in the next post.

package main

import (
	"fmt"
	"time"
)

func main() {
	for t := 10; t <= 200; t = t + 10 {
		us(t)
	}

	fmt.Println()

	for t := 10; t <= 200; t = t + 10 {
		ms(t)
	}
}

func ms(n int) {
	var t float64
	for i := 0; i < 100; i++ {
		s := time.Now()
		time.Sleep(time.Duration(n) * time.Millisecond)
		e := time.Since(s)
		t = t + float64(e)
	}
	t1 := t / 100.0
	nd := time.Duration(n) * time.Millisecond
	d := time.Duration(t1) - nd
	r := ((t1 / float64(nd)) - 1.0) * 100.0
	cyc := 1.0 / (float64(time.Duration(t1)/time.Millisecond) / 1000.0) / 30.0
	fmt.Printf("n=%d Wtd=%s Atl=%s Dlta=%s Df=%5.2f%% UPSec=%f\n", n, nd, time.Duration(t1), d, r, cyc)
}

func us(n int) {
	var t float64
	for i := 0; i < 100; i++ {
		s := time.Now()
		time.Sleep(time.Duration(n) * time.Microsecond)
		e := time.Since(s)
		t = t + float64(e)
	}
	t1 := t / 100.0
	nd := time.Duration(n) * time.Microsecond
	d := time.Duration(t1) - nd
	r := ((t1 / float64(nd)) - 1.0) * 100.0
	cyc := 1.0 / (float64(time.Duration(t1)/time.Microsecond) / 1000000.0) / 30.0
	fmt.Printf("n=%d Wanted=%s Actual=%s Delta=%s Diff=%5.2f%% UpdatePerSec=%f\n", n, nd, time.Duration(t1), d, r, cyc)
}

[misterunix]

Can you try this code? I am going to add a darwin delay file for a test but also add a flag for a timing test. Maybe it can help in the long run. If you are helping in the docs (please please) make sure you are pulling from "working-branch".

Windows AMD 3.2ghz (not the best...) You can see Windows is terrible!!! 10us give 14ms, this is when I learned through testing the best windows will do is 12ms.

C:\Users\miste\go\src\timingtest>timingtest.exe
n=10 Wanted=10µs Actual=14.17625ms Delta=14.16625ms Diff=141662.50% UpdatePerSec=2.351392
n=20 Wanted=20µs Actual=14.647596ms Delta=14.627596ms Diff=73137.98% UpdatePerSec=2.275779
n=30 Wanted=30µs Actual=15.944207ms Delta=15.914207ms Diff=53047.36% UpdatePerSec=2.090651
n=40 Wanted=40µs Actual=15.569195ms Delta=15.529195ms Diff=38822.99% UpdatePerSec=2.141007
n=50 Wanted=50µs Actual=15.504475ms Delta=15.454475ms Diff=30908.95% UpdatePerSec=2.149983
n=60 Wanted=60µs Actual=15.63262ms Delta=15.57262ms Diff=25954.37% UpdatePerSec=2.132378
n=70 Wanted=70µs Actual=14.448857ms Delta=14.378857ms Diff=20541.22% UpdatePerSec=2.307124
n=80 Wanted=80µs Actual=14.557212ms Delta=14.477212ms Diff=18096.51% UpdatePerSec=2.289849
n=90 Wanted=90µs Actual=16.153967ms Delta=16.063967ms Diff=17848.85% UpdatePerSec=2.063600
n=100 Wanted=100µs Actual=15.443662ms Delta=15.343662ms Diff=15343.66% UpdatePerSec=2.158475
n=110 Wanted=110µs Actual=15.379449ms Delta=15.269449ms Diff=13881.32% UpdatePerSec=2.167458
n=120 Wanted=120µs Actual=15.377506ms Delta=15.257506ms Diff=12714.59% UpdatePerSec=2.167740
n=130 Wanted=130µs Actual=15.671821ms Delta=15.541821ms Diff=11955.25% UpdatePerSec=2.127071
n=140 Wanted=140µs Actual=15.137976ms Delta=14.997976ms Diff=10712.84% UpdatePerSec=2.202110
n=150 Wanted=150µs Actual=15.994845ms Delta=15.844845ms Diff=10563.23% UpdatePerSec=2.084115
n=160 Wanted=160µs Actual=14.678933ms Delta=14.518933ms Diff=9074.33% UpdatePerSec=2.270972
n=170 Wanted=170µs Actual=16.812488ms Delta=16.642488ms Diff=9789.70% UpdatePerSec=1.982711
n=180 Wanted=180µs Actual=16.631724ms Delta=16.451724ms Diff=9139.85% UpdatePerSec=2.004289
n=190 Wanted=190µs Actual=17.334687ms Delta=17.144687ms Diff=9023.52% UpdatePerSec=1.923003
n=200 Wanted=200µs Actual=16.961827ms Delta=16.761827ms Diff=8380.91% UpdatePerSec=1.965293

n=10 Wtd=10ms Atl=19.850611ms Dlta=9.850611ms Df=98.51% UPSec=1.754386
n=20 Wtd=20ms Atl=26.045999ms Dlta=6.045999ms Df=30.23% UPSec=1.282051
n=30 Wtd=30ms Atl=41.683468ms Dlta=11.683468ms Df=38.94% UPSec=0.813008
n=40 Wtd=40ms Atl=44.896999ms Dlta=4.896999ms Df=12.24% UPSec=0.757576
n=50 Wtd=50ms Atl=57.413137ms Dlta=7.413137ms Df=14.83% UPSec=0.584795
n=60 Wtd=60ms Atl=72.739105ms Dlta=12.739105ms Df=21.23% UPSec=0.462963
n=70 Wtd=70ms Atl=73.13629ms Dlta=3.13629ms Df= 4.48% UPSec=0.456621
n=80 Wtd=80ms Atl=88.509806ms Dlta=8.509806ms Df=10.64% UPSec=0.378788
n=90 Wtd=90ms Atl=104.191331ms Dlta=14.191331ms Df=15.77% UPSec=0.320513
n=100 Wtd=100ms Atl=104.740041ms Dlta=4.740041ms Df= 4.74% UPSec=0.320513
n=110 Wtd=110ms Atl=120.128652ms Dlta=10.128652ms Df= 9.21% UPSec=0.277778
n=120 Wtd=120ms Atl=132.631722ms Dlta=12.631722ms Df=10.53% UPSec=0.252525
n=130 Wtd=130ms Atl=138.802835ms Dlta=8.802835ms Df= 6.77% UPSec=0.241546
n=140 Wtd=140ms Atl=151.075937ms Dlta=11.075937ms Df= 7.91% UPSec=0.220751
n=150 Wtd=150ms Atl=154.934873ms Dlta=4.934873ms Df= 3.29% UPSec=0.216450
n=160 Wtd=160ms Atl=166.908353ms Dlta=6.908353ms Df= 4.32% UPSec=0.200803
n=170 Wtd=170ms Atl=177.24732ms Dlta=7.24732ms Df= 4.26% UPSec=0.188324
n=180 Wtd=180ms Atl=186.337169ms Dlta=6.337169ms Df= 3.52% UPSec=0.179211
n=190 Wtd=190ms Atl=198.444952ms Dlta=8.444952ms Df= 4.44% UPSec=0.168350
n=200 Wtd=200ms Atl=208.514473ms Dlta=8.514473ms Df= 4.26% UPSec=0.160256

Linux Intel Core5 2.4ghz

./timingtest
n=10 Wanted=10µs Actual=340.567µs Delta=330.567µs Diff=3305.67% UpdatePerSec=98.039216
n=20 Wanted=20µs Actual=478.136µs Delta=458.136µs Diff=2290.68% UpdatePerSec=69.735007
n=30 Wanted=30µs Actual=598.342µs Delta=568.342µs Diff=1894.47% UpdatePerSec=55.741360
n=40 Wanted=40µs Actual=999.205µs Delta=959.205µs Diff=2398.01% UpdatePerSec=33.366700
n=50 Wanted=50µs Actual=620.105µs Delta=570.105µs Diff=1140.21% UpdatePerSec=53.763441
n=60 Wanted=60µs Actual=756.029µs Delta=696.029µs Diff=1160.05% UpdatePerSec=44.091711
n=70 Wanted=70µs Actual=743.81µs Delta=673.81µs Diff=962.59% UpdatePerSec=44.863167
n=80 Wanted=80µs Actual=1.11697ms Delta=1.03697ms Diff=1296.21% UpdatePerSec=29.868578
n=90 Wanted=90µs Actual=1.128227ms Delta=1.038227ms Diff=1153.59% UpdatePerSec=29.550827
n=100 Wanted=100µs Actual=1.166683ms Delta=1.066683ms Diff=1066.68% UpdatePerSec=28.587764
n=110 Wanted=110µs Actual=1.169726ms Delta=1.059726ms Diff=963.39% UpdatePerSec=28.514400
n=120 Wanted=120µs Actual=1.169345ms Delta=1.049345ms Diff=874.45% UpdatePerSec=28.514400
n=130 Wanted=130µs Actual=1.167427ms Delta=1.037427ms Diff=798.02% UpdatePerSec=28.563268
n=140 Wanted=140µs Actual=1.161125ms Delta=1.021125ms Diff=729.38% UpdatePerSec=28.710881
n=150 Wanted=150µs Actual=1.160826ms Delta=1.010826ms Diff=673.88% UpdatePerSec=28.735632
n=160 Wanted=160µs Actual=1.161577ms Delta=1.001577ms Diff=625.99% UpdatePerSec=28.710881
n=170 Wanted=170µs Actual=1.164698ms Delta=994.698µs Diff=585.12% UpdatePerSec=28.636884
n=180 Wanted=180µs Actual=1.163915ms Delta=983.915µs Diff=546.62% UpdatePerSec=28.661508
n=190 Wanted=190µs Actual=1.164097ms Delta=974.097µs Diff=512.68% UpdatePerSec=28.636884
n=200 Wanted=200µs Actual=1.164394ms Delta=964.394µs Diff=482.20% UpdatePerSec=28.636884

n=10 Wtd=10ms Atl=10.307056ms Dlta=307.056µs Df= 3.07% UPSec=3.333333
n=20 Wtd=20ms Atl=20.29862ms Dlta=298.62µs Df= 1.49% UPSec=1.666667
n=30 Wtd=30ms Atl=30.289362ms Dlta=289.362µs Df= 0.96% UPSec=1.111111
n=40 Wtd=40ms Atl=40.286264ms Dlta=286.264µs Df= 0.72% UPSec=0.833333
n=50 Wtd=50ms Atl=50.281148ms Dlta=281.148µs Df= 0.56% UPSec=0.666667
n=60 Wtd=60ms Atl=60.285292ms Dlta=285.292µs Df= 0.48% UPSec=0.555556
n=70 Wtd=70ms Atl=70.295469ms Dlta=295.469µs Df= 0.42% UPSec=0.476190
n=80 Wtd=80ms Atl=80.317714ms Dlta=317.714µs Df= 0.40% UPSec=0.416667
n=90 Wtd=90ms Atl=90.3216ms Dlta=321.6µs Df= 0.36% UPSec=0.370370
n=100 Wtd=100ms Atl=100.349578ms Dlta=349.578µs Df= 0.35% UPSec=0.333333
n=110 Wtd=110ms Atl=110.37542ms Dlta=375.42µs Df= 0.34% UPSec=0.303030
n=120 Wtd=120ms Atl=120.349782ms Dlta=349.782µs Df= 0.29% UPSec=0.277778
n=130 Wtd=130ms Atl=130.364515ms Dlta=364.515µs Df= 0.28% UPSec=0.256410
n=140 Wtd=140ms Atl=140.405037ms Dlta=405.037µs Df= 0.29% UPSec=0.238095
n=150 Wtd=150ms Atl=150.416364ms Dlta=416.364µs Df= 0.28% UPSec=0.222222
n=160 Wtd=160ms Atl=160.42851ms Dlta=428.51µs Df= 0.27% UPSec=0.208333
n=170 Wtd=170ms Atl=170.407648ms Dlta=407.648µs Df= 0.24% UPSec=0.196078
n=180 Wtd=180ms Atl=180.416257ms Dlta=416.257µs Df= 0.23% UPSec=0.185185
n=190 Wtd=190ms Atl=190.456422ms Dlta=456.422µs Df= 0.24% UPSec=0.175439
n=200 Wtd=200ms Atl=200.439547ms Dlta=439.547µs Df= 0.22% UPSec=0.166667

Linux Intel Core7 laptop

./timingtest
n=10 Wanted=10µs Actual=340.567µs Delta=330.567µs Diff=3305.67% UpdatePerSec=98.039216
n=20 Wanted=20µs Actual=478.136µs Delta=458.136µs Diff=2290.68% UpdatePerSec=69.735007
n=30 Wanted=30µs Actual=598.342µs Delta=568.342µs Diff=1894.47% UpdatePerSec=55.741360
n=40 Wanted=40µs Actual=999.205µs Delta=959.205µs Diff=2398.01% UpdatePerSec=33.366700
n=50 Wanted=50µs Actual=620.105µs Delta=570.105µs Diff=1140.21% UpdatePerSec=53.763441
n=60 Wanted=60µs Actual=756.029µs Delta=696.029µs Diff=1160.05% UpdatePerSec=44.091711
n=70 Wanted=70µs Actual=743.81µs Delta=673.81µs Diff=962.59% UpdatePerSec=44.863167
n=80 Wanted=80µs Actual=1.11697ms Delta=1.03697ms Diff=1296.21% UpdatePerSec=29.868578
n=90 Wanted=90µs Actual=1.128227ms Delta=1.038227ms Diff=1153.59% UpdatePerSec=29.550827
n=100 Wanted=100µs Actual=1.166683ms Delta=1.066683ms Diff=1066.68% UpdatePerSec=28.587764
n=110 Wanted=110µs Actual=1.169726ms Delta=1.059726ms Diff=963.39% UpdatePerSec=28.514400
n=120 Wanted=120µs Actual=1.169345ms Delta=1.049345ms Diff=874.45% UpdatePerSec=28.514400
n=130 Wanted=130µs Actual=1.167427ms Delta=1.037427ms Diff=798.02% UpdatePerSec=28.563268
n=140 Wanted=140µs Actual=1.161125ms Delta=1.021125ms Diff=729.38% UpdatePerSec=28.710881
n=150 Wanted=150µs Actual=1.160826ms Delta=1.010826ms Diff=673.88% UpdatePerSec=28.735632
n=160 Wanted=160µs Actual=1.161577ms Delta=1.001577ms Diff=625.99% UpdatePerSec=28.710881
n=170 Wanted=170µs Actual=1.164698ms Delta=994.698µs Diff=585.12% UpdatePerSec=28.636884
n=180 Wanted=180µs Actual=1.163915ms Delta=983.915µs Diff=546.62% UpdatePerSec=28.661508
n=190 Wanted=190µs Actual=1.164097ms Delta=974.097µs Diff=512.68% UpdatePerSec=28.636884
n=200 Wanted=200µs Actual=1.164394ms Delta=964.394µs Diff=482.20% UpdatePerSec=28.636884

n=10 Wtd=10ms Atl=10.307056ms Dlta=307.056µs Df= 3.07% UPSec=3.333333
n=20 Wtd=20ms Atl=20.29862ms Dlta=298.62µs Df= 1.49% UPSec=1.666667
n=30 Wtd=30ms Atl=30.289362ms Dlta=289.362µs Df= 0.96% UPSec=1.111111
n=40 Wtd=40ms Atl=40.286264ms Dlta=286.264µs Df= 0.72% UPSec=0.833333
n=50 Wtd=50ms Atl=50.281148ms Dlta=281.148µs Df= 0.56% UPSec=0.666667
n=60 Wtd=60ms Atl=60.285292ms Dlta=285.292µs Df= 0.48% UPSec=0.555556
n=70 Wtd=70ms Atl=70.295469ms Dlta=295.469µs Df= 0.42% UPSec=0.476190
n=80 Wtd=80ms Atl=80.317714ms Dlta=317.714µs Df= 0.40% UPSec=0.416667
n=90 Wtd=90ms Atl=90.3216ms Dlta=321.6µs Df= 0.36% UPSec=0.370370
n=100 Wtd=100ms Atl=100.349578ms Dlta=349.578µs Df= 0.35% UPSec=0.333333
n=110 Wtd=110ms Atl=110.37542ms Dlta=375.42µs Df= 0.34% UPSec=0.303030
n=120 Wtd=120ms Atl=120.349782ms Dlta=349.782µs Df= 0.29% UPSec=0.277778
n=130 Wtd=130ms Atl=130.364515ms Dlta=364.515µs Df= 0.28% UPSec=0.256410
n=140 Wtd=140ms Atl=140.405037ms Dlta=405.037µs Df= 0.29% UPSec=0.238095
n=150 Wtd=150ms Atl=150.416364ms Dlta=416.364µs Df= 0.28% UPSec=0.222222
n=160 Wtd=160ms Atl=160.42851ms Dlta=428.51µs Df= 0.27% UPSec=0.208333
n=170 Wtd=170ms Atl=170.407648ms Dlta=407.648µs Df= 0.24% UPSec=0.196078
n=180 Wtd=180ms Atl=180.416257ms Dlta=416.257µs Df= 0.23% UPSec=0.185185
n=190 Wtd=190ms Atl=190.456422ms Dlta=456.422µs Df= 0.24% UPSec=0.175439
n=200 Wtd=200ms Atl=200.439547ms Dlta=439.547µs Df= 0.22% UPSec=0.166667

[misterunix]

@jottinger Mad did I find some major screw ups. My end detection code was messed up. Jeeeze the stuff you miss....
I will add the sleep for darwin and push and merger to main so I can do a new release.

[misterunix]

Done. v0.0.2b pushed out.