by HikaruYami » Tue Oct 30, 2018 3:05 pm
https://www.smbc-comics.com/comic/bins
I actually haven't seen a bin-packing joke like this before and was very amused. Unfortunately, the mathematician failed to account for the fact that the mover can have a small and more importantly *constant* number of constantly-sized bins and that the breadth of people he has moved will have, in the grand scheme of things, a small number of objects to be moved.
It turns out that when n is sufficiently small, even exponential-time algorithms can be run quickly. As of 2003, an algorithm had been developed that even on very average hardware of the day could consistently find optimal solutions for 80 items in an average of 31 milliseconds. It obviously continues to grow quickly from there, but it's not unreasonable for a mover to expect people to batch their smallest items before having them shove everything into a truck. A mover isn't going to literally pick up your silverware for you.
<this is where someone tries to point out that the complexity of the problem increases when the size of the items is expressed in arbitrary-precision values, and I retort by pointing out that that precision only matters on real hardware, which a mathematician wouldn't give a shit about>
<this is where someone correctly points out that the shape of the objects matters in a real solution of packing three-dimensional bins, and I point out that that ruins the entire joke about the bin-packing problem from the perspective of the mathematician>
https://www.smbc-comics.com/comic/bins
I actually haven't seen a bin-packing joke like this before and was very amused. Unfortunately, the mathematician failed to account for the fact that the mover can have a small and more importantly *constant* number of constantly-sized bins and that the breadth of people he has moved will have, in the grand scheme of things, a small number of objects to be moved.
It turns out that when n is sufficiently small, even exponential-time algorithms can be run quickly. As of 2003, an algorithm had been developed that even on very average hardware of the day could consistently find optimal solutions for 80 items in an average of 31 milliseconds. It obviously continues to grow quickly from there, but it's not unreasonable for a mover to expect people to batch their smallest items before having them shove everything into a truck. A mover isn't going to literally pick up your silverware for you.
<this is where someone tries to point out that the complexity of the problem increases when the size of the items is expressed in arbitrary-precision values, and I retort by pointing out that that precision only matters on real hardware, which a mathematician wouldn't give a shit about>
<this is where someone correctly points out that the shape of the objects matters in a real solution of packing three-dimensional bins, and I point out that that ruins the entire joke about the bin-packing problem from the perspective of the mathematician>