The fascinating thing about saying “making a computer wait 1 second is hard”:
Some who took one intro to python class will say “it’s fucking trivial you retard, just do sleep(1)”
And some who have taught CS for five decades will say “it’s not hard. It’s impossible.”
Some who took one intro to python class will say “it’s fucking trivial you retard, just do sleep(1)”
And some who have taught CS for five decades will say “it’s not hard. It’s impossible.”
Sep 10, 2026 · 6:14 PM UTC
117
80
6,461
312,149
@1v100000 Not really, I was gonna say they’re still not exact, but neither are watches
3
0
313
43,083
@IceSolst the "555" was the single most successful chip ever made. a billion are still made evey year. all it does is keep time.
8
0
502
26,645
@IceSolst That's why std functions always are "sleep at least N", they should be named as such too tbh...
0
0
76
5,229
@IceSolst It all depends on your measurement device.
If you want inches use a ruler.
You want seconds use a clock.
If you want inches use a ruler.
You want seconds use a clock.
0
0
10
475
@IceSolst And we who have built such things will ask, "what do you mean by waiting a second" & some will show they don't understand the question, let alone the answer. (Happy to say more, but I'll let it sit there.) What I really want for "wait one second" could be built, but no OS has it.
1
0
10
4,247
@IceSolst Exactly one second? Ok.
Just after one second? Totally doable
Just after one second? Totally doable
0
0
3
943
@IceSolst Depends on your tolerances and definitions. A quartz circuit that does nothing but count oscillations to match 1 second technically meets your criteria
1
0
192
13,794
@IceSolst The real difficult part is defining what one second even is.
They have councils and shit for that.
Nevermind that due to universal expansion, time is always stretching, always bending.
What is time anyway?
Dude, I'm freaking out
What if...time is always "on"?
They have councils and shit for that.
Nevermind that due to universal expansion, time is always stretching, always bending.
What is time anyway?
Dude, I'm freaking out
What if...time is always "on"?
4
0
165
11,682
@IceSolstl these things are easy until you have to work out how to program at electronic sand level, then it gets a bit more difficult. I know this is a game but some people should try and work through these if they don't understand why this stuff is hard!
store.steampowered.com/app/1444480/Tu…
store.steampowered.com/app/1444480/Tu…
3
1
60
20,795
@IceSolst Getting it to wait 1 second is easy.
The hard part is getting a computer to wait 1.0000000000 seconds.
The hard part is getting a computer to wait 1.0000000000 seconds.
0
0
42
1,499
@IceSolst I know a computer science teacher who teaches to this day that "virtual machines are just a fad and will never be used in the industry."
I've been in the industry for 15 years and used virtual machines all the way through...
I've been in the industry for 15 years and used virtual machines all the way through...
2
0
32
3,473
@IceSolst Skill issue. Just remove the power for 1 second (measures externally) and resupply it back, and the computer would have waited for just that second.
0
0
27
2,978
@IceSolst And then you go the FPGA/electrical engineer and they might bring up the GPS/atomic standard disciplined oven controlled crystal oscillators clocking some RTOS. Then to the metrologist... Then to the physicist... Reminds me of xkcd 435 lmao
xkcd.com/435/
xkcd.com/435/
0
1
23
5,533
@IceSolst And some will joke about really large or small values of 1.
0
0
18
3,693
@IceSolst It's so easy. Just pull the plug and plug it back in. It will have waited for at least 1 second at that point....
0
0
13
1,427
@IceSolst Depends on your definition of "computer" and accuracy you need with 1 second. Embedded system? Trivial
0
0
9
1,640
@IceSolst Very fortunate that I have only ever needed to make the computer wait "roughly, at least" 1 second.
0
0
9
529
@IceSolst I work in safety critical embedded. I recently did some timing measurement things and I was sweating that sub microsecond accuracy wouldn't cut it.
0
0
6
3,309
@IceSolstce custom custom made and hooked up his own timing chips… it’s not ~that~ hard you just need to power a hertz oscillator at a steady and precise voltage, measure its frequency and do the math on how many oscillations per seconds, oh and you need to be able to read that faster than the speed of light, or accurate enough to predict the latency between the read and the actual end of a second…
Child’s play
Child’s play
0
0
5
695
@IceSolst 5 decades?
There was once a time that it was fairly easy, before operating systems and such.
Now it all depends on how accurate you want to be, if it's +/- 1 second, that's pretty easy, within 16 ms is already a big challenge, <1 ms is basically impossible
There was once a time that it was fairly easy, before operating systems and such.
Now it all depends on how accurate you want to be, if it's +/- 1 second, that's pretty easy, within 16 ms is already a big challenge, <1 ms is basically impossible
5
0
5
3,172
@IceSolst And those who program embedded single core processors with fixed clock rates: “piece of cake”
1
0
4
442
@IceSolst Call them a retard back then ask them about drift from scheduling after doing so
0
0
4
2,263
@IceSolst It’s not a computer science problem, it’s a physics problem. There’s no such thing as “exactly”.
0
0
3
305
@IceSolst If you want down to the cycle precisely then you don't want a computer. Computers are lazy. There's other hardware that is designed for cycle specific precision
0
0
3
601