Page 1 of 1

[2.0.28] Legendary inserters are faster than expected

Posted: Thu Dec 26, 2024 5:52 pm
by Hares
According to my calculations, legendary inserters should make 5..6 swings per second, or use 10..12 ticks per full swing when moving items from chest to chest, depending on the fact do they require a tick to pick up / drop item or not:
12 ticks per rotation assuming 2 additional ticks are required
12 ticks per rotation assuming 2 additional ticks are required
12-26-2024, 20-36-40.png (42.6 KiB) Viewed 539 times
10 ticks per rotation assuming additional ticks are not required
10 ticks per rotation assuming additional ticks are not required
12-26-2024, 20-37-35.png (31.59 KiB) Viewed 539 times
In practice, they require 8 ticks to move item from chest to chest, which is estimated 2700°/s or 3600°/s (depending on these 2 extra ticks) and not 2160°/s. Attaching recording of tick-by-tick inserter swing where it's clearly visible they're skipping one tick of rotation and instantly jumping to orthogonal state (and definitely not requiring extra ticks to pick up / drop item). Tested this with other chest-to-chest setups, result is the same -- 8 ticks per full swing.
Click to expand the recording
Funny enough, inserters being one tick per half swing faster than expected results in them having exactly the same throughput per unit of spacetime when moving items from chained wagons (aka cursed wagons) as two stacked turbo belts: 124962: [SA] Fastest (and cursiest) ways of transporting items

Re: [2.0.28] Legendary inserters are faster than expected

Posted: Thu Dec 26, 2024 7:14 pm
by Rseding91
Thanks for the report however I do not believe we will be doing anything about this. The same function shown in the tooltip is the same one used in the internal rotation logic. Additionally the value shown in the tooltip is not the full story for how inserters rotate and we will not be putting the entire source code for inserters into the tooltip for people to do math with.

Re: [2.0.28] Legendary inserters are faster than expected

Posted: Thu Dec 26, 2024 8:57 pm
by Hares
Rseding91 wrote: Thu Dec 26, 2024 7:14 pm Thanks for the report however I do not believe we will be doing anything about this. The same function shown in the tooltip is the same one used in the internal rotation logic. Additionally the value shown in the tooltip is not the full story for how inserters rotate and we will not be putting the entire source code for inserters into the tooltip for people to do math with.
I assume it's just a rounding error. After 4 ticks it checks if he's within one tick of orthogonal position, floor(0.999) == 0, and thus "magnetizes" to it reaching it in 4 ticks instead of 5. I may be wrong but it looks how I described.