Nevar!
. . .to the second part. My mom is cool.
Is this a logical statement? And does the runtime support short-circuiting? It would mean you don’t have to obey physics if you don’t love your mom, which is neat.
My mom sucks, though
She does it well, so there’s that.
I work as a software engineer and honestly, it’s ridiculous how often I’m asked to or tempted to violate the laws of physics.
There’s classics like measuring how long it takes to send a network packet from one device to another – you can’t, because the two devices might have wildly different understandings of what time it currently is. The only way to get an accurate measurement is by measuring how long it takes to send it there + back (a.k.a. the round-trip time).
And then you divide that by 2 and pretend there’s no asymmetry in transmission speed, nor delay between the other device receiving it and sending it back. 👍In our previous project, we were recording audio chunks of one second each and then feeding it into a detector. At some point, we got asked, if we could reduce the delay until the user gets feedback from the detector. Also, we can’t make the detector detect things more often, because it might make more mistakes. Alright, I guess, I’ll just break up the time continuum then and give the user feedback before it has finished recording. 👍
And now in our current project, we’re supposed to send network packages across the globe and also we basically can’t have any latency. Yeah, so there’s this thing called the speed of light/causality at about 300000 km/s. Halfway around the globe is about 20000 km. That leaves us with 66.7 ms of latency, at its theoretical minimum. Guess I’ll just quickly invent a way to create worm holes, no problem. 👍
Alright, I guess, I’ll just break up the time continuum then and give the user feedback before it has finished recording. 👍
Quantum physics might work too. No, seriously.
There’s classics like measuring how long it takes to send a network packet from one device to another
That one is on your clocks quality, not on physics. People do it all the time.
Probably on equipment that is orders of magnitude more expensive than yours, but the post isn’t about costs.
That’s just shifting the problem. There is no known way to reliably sync remote clocks except by sending packets and assuming the round-trip time is symmetrical. This is a known problem in physics: https://en.wikipedia.org/wiki/One-way_speed_of_light
Sure, but if we take it as true that light speed is the same in every direction – which is perfectly consistent with everything ever measured – you can measure speed between two endpoints using two atomic clocks and a synchronised experiment, with corrections for the relativistic effect of moving the clocks to the different places
Or use what PTP does.
https://en.m.wikipedia.org/wiki/Precision_Time_ProtocolWhich is a more accurate protocol than NTP.
https://en.m.wikipedia.org/wiki/Network_Time_Protocol
Send it through the earth, you can reduce it theoretically to 42.5ms (using mean radius and speed of light in vacuum).
Oh, good point, I recently learned that the speed of light in fiber optics is around 200000 km/s. I always thought physicists were saying “in vacuum” to be technically correct, but that’s actually a huge difference…
https://en.wikipedia.org/wiki/Optical_fiber#Refractive_index
Send it through the earth, you can reduce it theoretically to 42.5ms
This isn’t as ridiculous as it sounds and you just need a neutrino-beam… which has a horrible bandwidth (of 0.1 bits/s) plus the ridiculous upfront cost of running two particle-accelerators for a full-duplex link. (Google it up, this exists.)
Thanks Peter Griffin
That’s a ballchinian.
