Squibble’s Epic Sand Pile Quest

Hey everyone in case you couldn’t guess from the title we’ve got a really stupid project today so let’s get right into it that’s cuz this project is going to be about trying to find the perfect pile of sand but more so I guess building the machine that is going to somewhat

Consistently regurgitate said perfect pile of sand we’re going to do this by regurgitating a bunch of really tiny pieces of sand individually and by sheer probability they’re going to arrange thems generally into the shape that we’re looking for also at the end of the video there’ll be a super important

Announcement because we’ve finally made a Discord so make sure to stay tuned for that I guess we should first start with what is a perfect pile of sand you see here’s the thing if you take this object called a gon board which is just a bunch

Of pegs and a bunch of pillars and like a cone thing and you take a bunch of balls and shove them through the pegs then you eventually end up with this little shape and this little shape arises from the fact that there is a very small chance for these balls to end

Up on the outside and a very big chance for them to end up on the inside this is called the bell curve normal distribution gaussian whatever the hell you want to call it based on the context you’re in but today we’re going to be trying to make them as well as scaling

Things up to the third dimension because we’re in Minecraft and we have cubes not squares the Crux of Designing a gon board in Minecraft is to essentially forgo the fact that a real life gon board obviously has a bunch of much more Randomness because when a ball hits any

Given Peg it’s not a 50% chance to go to the left 50% chance to go to the right there’s obviously a bunch of slot between everything and a bunch of other balls knocking into each other whereas on our situation each piece of sand will have exactly half a chance of sliding to

The left and half a chance of sliding to the right luckily for us this totally doesn’t matter because the thing is is that everything ends up working out in probability wise later and we can just be totally fine with everything being perfect as long as we make sure that the

Chances are in fact perfect and not biased that was foreshadowing by the way the machine’s going to break later and we’re going to have some bias but for now we’re going to pretend like I’m blissfully unaware of the problems that are going to rise later and we’re going

To talk about how we actually designed this thing first we need to actually start designing the gon boore before we even get into probability and that comes with moving the sand around see moving sand around is actually becomes quite a bit of a problem when you’re moving in

Four directions not just one this is the design we have to go with where essentially we have each honey block surrounded by four possible sticky pistons and we can actually move the sand Everywhere by simply just powering it in the opposite direction first pulling it back and then powering it in

The other direction and then pulling it back what this does is it actually moves the sand two blocks in the direction we want to move it in which allows us to tile these things in a sort of two wide checkerboard pattern it’s kind of interesting now the thing is is how on

Earth do we power these things because thing is we need to be able to interface with every single piston directly and we can only Power them from the bottom because there’ll be a bunch of slot moving around on the top but we also have to worry about these black pillars

I’ve marked out where sand will actively be falling down and that means that we just literally can’t run any wires in certain locations with our first sets of Pistons wiring up we can do this pretty easily just by simply running rails underneath with observers and this only gets only

Slightly more complicated by just having to run another set of rails underneath and doing some weird door Stone in order to Route the signals upwards if we actually look at this piece of sand right here and I go ahead and push this note block it’s going to pull it to that

Way and if we use the other input it’s going to move it into that slot and drop the sand all the way to the Bottom now all we have to do is make sure that at every stage this piece of sand goes in a completely random Direction and hopefully we should end up

With the results that we want and this is exactly what’s going to happen at least for you because unfortunately for for me I actually have to go through the effort of endlessly iterating upon this thing stacking it over and over again fixing all the wiring issues and going through about a

Bazillion iterations to fix all the tiny little problems that I never could have foreseen so it was a little bit of an issue and if you’re wondering what kinds of Errors we might have experienced they can get from anywhere from tiny little wiring mishaps to full-blown catastrophes on every single floor at

The exact same time none of these Pistons are in the right SP spot and they will never be ever again if there’s anything you take away from this video it’s make backups always make backups if anything can possibly ever go wrong I’m only going to address some of the sort

Of obstacles we overcame because I guess it’s kind of important but there are a bunch of stuff that really just isn’t worth talking about first problem is that something we funnily enough couldn’t have foreseen from the original mockup design and to do with these sideways Pistons see the thing is is

That we’re going to be triggering all of these Pistons of Any Given category at the same time but the problem is is the Pistons in here will be triggering at the same time as the Pistons in here and when this piston tries to extend into here and this piston tries to extend its

Pistons into here we’re going to have a bit of a traffic jam now a simple naive solution would be to simply just power this one wait for it to be over and then power this one wait for it to be over then power this one and so on this

Sounds incredibly slow and it’s actually not as slow as you might imagine to be but it definitely is slower than the more practical solution the more practical solution is realizing that we actually don’t need to trigger everything in order since each one is only bothered by the fact that the one

Immediately in front of it and the one behind it is firing at the same time we can actually just power every other one all at the same time and since none of them are touching it’ll all work work and then we can power all the other ones

We haven’t activated yet and thus this allows us to complete the entire layer in just two cycles tiny little time saves like that are always super important because the thing is this thing already takes so long to run a single layer given the fact that it runs

Through a ton of observers before a piece of wire can actually reach the end this is just because everything gets so large by the bottom layer that you just kind of have to expand things out and know I am not using instant wire what you eventually end up with is a giant

Machine that has four buttons s and each of them will send the sand down in a very specific Direction this all works perfectly fine given that you don’t unload it and accidentally turn the entire thing to Spaghetti again and it should mostly work but the problem is is

That unfortunately it isn’t random yet we still need to attach the actual randomizer in order to randomize this we need to be able to choose one of four values perfectly randomly with no bias whatsoever the way we could have done this is by using standard or dropper randomizers where essentially you have a

Non-stackable and a stackable item in a dropper and you essentially power it causing one of these items to be put into the hopper complete 50/50 chance and then you draw a comparator output the bigger the comparator output it’s probably the unstackable item and that means you can essentially output a one

Whereas the stackable item is a zero using every possible configuration of these using two we can actually get four different outputs but this is complicated and I don’t think we to do that the way we’re going to sent choose this is we’re going to use a single dispenser that contains four different

Items and the way we’ll differentiate them is the fact that shulker boxes can contain different amounts of items this means that when we place them they’ll all output completely different comparator outputs and this will allow us to differentiate between the four outcomes without having to do binary to

I don’t know conversion of some kind since we’ve intentionally assigned a single randomizer to every single layer that means we can actually randomize the output of every single layer running multiple sand pieces in tandem without screwing up a randomization Valu somehow by making it not totally random if we

Apply the same direction to multiple pieces of sand we risk biasing our data somehow and that could be really bad but wait where is our sand coming from and the thing is is I actually designed this really weird looking Hopper thing that just takes a giant chunk of sand and

Condenses it into into a single little beam this is really not that complicated and it’s totally not worth discussing but it just looks cool kind of like a giant little sand pen nib almost after all of this nonsense we finally have a mostly working design that only breaks when I wander too far

Away from it and if I go ahead and tick warp this machine we’re going to see a bunch of pieces of sand dropping around randomly but the most important thing is that they’re mostly kind of clustering around the center this is is super good because it

Means that we’ll probably end up with a piece of pile of sand that looks a little something like this looking a little bit like the stretched out gaussian distribution that we had seen at the beginning however initially sometimes things just weren’t random because the reality was that some mic

Dropper randomizers were just completely busted causing this entire output sand column to just be completely skewed this way if you’re wondering how stuff like that even breaks outside of just ported design it’s because you copied the entire machine over somewhere but you forgot to add the singular glass block

Because it’s basically outside of the entire frame of the machine and that’s responsible for making sure the hopper doesn’t pull items out of the shulkers this essentially just led to all the shulkers being depleted of items and also a bunch of repeaters ending up in the system completely scrambling the

Item sorder at the bottom to essentially always output the same direction this specific example resulted in the recorded data shift that I actually just copied over here cuz I thought it looks pretty cool so the question is did we actually create the perfect pile of sand

And I’m going to be honest um almost it’s correctly shaped but the problem is that we’ve simply only got so much resolution that Randomness is going to take over and we’re going to end up with stuff that doesn’t exactly look as perfect as we wanted to in an Ideal

World all four of these Center columns would likely be roughly the same height and we’d get rings going around the edges if we had a way of cutting the sand stack in half exactly every now and then just to keep the height down as we keep adding more layers maybe we could

Eventually Converge on some kind of perfect pile and then we’d have our goal anyways that’s the end of the actual content but I’ve got a really important announcement to make you see I finally put in the effort to actually make a Discord yes many people thought that

They would never come but we’ve actually got it now it’s called Redstone transform race you can the invite link is in the description and we’ll be mainly posting about teasers announcements stuff like that you can also have a chat with me talk to me in the TX channels or whatever and I’ll

Probably respond it’s also just a redstone server for General Redstone stuff whether it’s actually useful stuff or the kinds of things that I show on my channel that really don’t have as much of a use anyways that’ll be it for today make sure to like And subscribe if you

Want to see more stuff like this check out my other videos and I’ll see you next time