Exactly

This is my conclusion exactly. When we first started making farms with pistons and hoppers it felt hopeless... it's a lesson to never give up when it comes to Minecraft tech.

No, not really. The concept is very simple/boring and constructing a geode farm always used to be a pain without this mod.

Its shame though that geode farms generate so little shards for the effort it takes to make them

@@theanonymous7221 I think that compared to those stupid 1.19 nether portal nether mob farm, or the classic villagers food farm, this type of automated geode farm are pretty unique

@@drg0n_ as an old school beta player who's trying to get into the modern game, i still find those farms very interesting. maybe because i haven't kept up with the mechanics and meta, i'm just not burnt out on them like y'all seem to be.

Mod author here. It's possible, I'm just lazy and not very good at algorithms :D

True but it's meant to be somewhat simple. I'm sure someone one day will make it though!

I thought that was meant to be the fun part?

@@kosmamoczek Relatable.

@@kosmamoczek Haha, at least you are being honest

Oh we have that as well

Could you please make a video about that, I'm curious to know

@@ilmango Which mod does that?

@@vinh020298 ok. I can do that

@@vinh020298 Yes this would be awesome!

That is a good point, I didn't even realize it could be possible to increase efficiency like that. I'm guessing it could be done by removing one crystal at a time and checking for the maximum efficiency of those, there are a lot of different combinations though. With 5 budding amethyst for example, you have 1 (all 5) + 5 (one removed) + 10 (two removed) + 10 (three removed) + 5 (4 removed) + 1(all removed). In general the number of configurations to check is sum( xCi ) where x is the number of budding amethyst, xCi is the combination function (nCr), and i is the summation index, all numbers from 0 to x. I checked on a calculator and this is basically exponential so its not exactly reasonable to check every possibility. This is still doable though because you don't have to check every single possibility. Of course some of these you can omit, since there is no way all or most the amethyst being removed will be helpful. Maybe checking all combinations up to some number being removed like 5 or so. Or maybe you only continue checking, if the efficiency increased for one of the previous checks. Like you check with one removed, if the efficiency increased, check efficiencies for two removed.

@@adrienanderson7439 I assume there would be ways to do better than brute force mostly, for example there's probably a way to prove that blocks under certain conditions could not improve things with their absence and remove them from rotation for removal

@@EvonixTheGreatest Yeah you are definitely right. Buds that aren't blocking anything probably don't need to be removed so maybe you only check arrangements removing different combinations of buds that are along the columns rows and pillars of the blocked amethysts that have already been marked as we can see from the buttons. Since buds that aren't in those positions aren't blocking any crystals from being removed. However some buds might be blocking other crystals from growing by being adjacent so maybe you want to also check arrangements removing touching buds. In summary, we can reduce the computation by not worrying about removing buds that are not in line with any crystals and not touching any other buds.

Also I made an error in my original comment we should be checking for number of harvestable crystals, not efficiency since one lonely bud would have 100% efficiency but not a lot of crystals to harvest.

One way to find a stopping point for budding amethyst removal is to keep track of your best yield for the arrangements of budding amethyst and stop removing budding amethyst if they would grow (but not yield) less than your max yield. So this should probably automatically stop at around 5 amethyst, and even if you brute forced the rest it probably wouldn't take much time, though the idea of looking at which ones don't really block others seems like a good idea I had not considered.

You smart

Needs just a couple screenshots? How do you extract the information like locations from the screenshots? Shouldn't that be really hard?

@@cook_it technically 2 good screenshots are enough, possibly even one given the cubical world. that said it'd be my first try on image recognition so i'm not really sure how i will do it😆

@@colox97 First try? Oh boy then you're in for one hell of a ride XD I've never done anything even remotly similar to this, but from my experience in programming some more or less "small" things over the years it's REALLY easy to start something and then to realise that there is WAY more behind accomplishing your goal than you ever expected. Sometimes you have enough experience to make a solution for the problems that come up but sometimes you just get stuck and either have to compromise on your goals or have to get help from someone else. It's often really disheartening but that's something everyone has to overcome. Some problems that just come up in my head without much thinking are: How do you deal with different FOV's? How do you deal with different resolutions? Texture packs/better lighting mods/etc or perhaps even different graphical settings like mipmaps probably break everything (and if your webapp gets popular enough there will be people with that, no matter how many warnings/notices you write). How do you compensate for where (aka location) the screenshots are made and with what angle, especially with the relation to each other? What if people make them with really weird angles (perhaps they make them from outside in spectator mode, maybe from the inside in survival)? How to you deal with people who leave their HUD visible? One potential workaround for avoiding all of that would be allowing the user to upload their chunk files (you can read those in with for example a python module) containing the geode, but that immediatly puts a lot more work/unusability onto the user as now they have to figure out what chunk files contain the geode (and that also won't work on servers). Alternatively you could figure out how to read litematica saves but that needs a mod and if you are already modding then why not just install Geodesy? Like I said I have no experience in webapps/image recognition/3d "stuffs" but from my little bit of experience in programming (aka mainly making excel workbooks "automagical" at work with VBA) that sounds like a pretty huge task. Now of course I could be totally wrong and it's a lot easier than I imagine it to be (which I really hope it is, maybe there are some libraries/frameworks out there that do the heavy lifting for you?) so I'd say go for it and see how far you come! Whatever happens, it will be an excellent learning experience :)

@@cook_it Thanks for taking the time to reply! well, the first version will DEFINITELY have me inputting the geode structure by hand. i've already kinda listed the steps from the raw geode to the built machine so that doesn't scare me too much. As you guessed the big learning experience will be from having to process the input data, and i'm definitely not aiming for perfection, vanilla texture with vanilla UI, i was actually thinking of using F3 to extract even more informations about my POV and location (and the 3 lines in the center of the screen to line up the axis). That said, since the world is made entirely of equally distant repeating lines i think it's definitely feasible, i'll have to relearn a lot of math though 😆

@@colox97 Ok yeah with the *exact* coordinates and the *exact* angle from the F3 menu and "standardised" resolution/settings it should be possible, now, as you said, all you need is lots of math, and I hope your maths is *way* better than mine because I didn't need any math higher than ()*%+- for like... 5 years or so (and writing formulas/VBA in excel does not count sadly) ;) Well, then good luck, I hope you give an update to your progress at some point (maybe a video?) because now I'm really interested in this project of yours XD

sure. for mycelium as well. I wanted to do something like this for a long time but it never was high on the priority list

You can easly send down signals with walls

Apparently pistons can break the budding amethyst blocks, so that's no option

@@racernatorde5318 That's not true, ilmango has even made videos on piston-based geode farms. It works but it's much harder to design and can lose some amethyst crystals if they land on top of the pistons. But in this case I'm just talking about a few extra pistons along with a few extra observers attached to the flying machines in strategic positions. Maybe there are some edge cases where that still wouldn't reach, but I don't see why it wouldn't work for at least some of the unreachable spots.

@@OceanBagel Why yes, it is true. Pistons do break budding amethyst blocks. The fact that Mango has made piston based designs before does not change this fact. There is no space to have loose observers in the farm, so you would need to attach the observers to the flying machine - and then you break the budding amethyst blocks

@@racernatorde5318 I wasn't saying that pistons can't break budding amethyst blocks, I'm saying that it doesn't have anything to do with what I'm saying. The pistons would be facing the growing crystals, not the budding amethyst block, so that when they triggers, they harvest the crystals. And the observers would be attached to the flying machine and trigger the piston as it sweeps across. I don't quite understand where you claim breaking budding amethyst blocks comes into play since the piston would need to push the budding amethyst itself to actually break it.

@@OceanBagel Do you realize that you need one pistons for every single exposed side of a budding amethyst? There is absolutely not space for such system. It's not worth the effort to get these farms past the efficiency of the concept Mango showed here

An ideal solution would have 1. minimum number of slime/honey blocks required (Lesser num. of blocks not hitting any buds but required to connect diagonal blocks) 2. minimum number of flying machines (If you use too many slime blocks, you will need more flying machines) Constraints which were mentioned in the video 1. Push limit of 12 blocks -> 12 slimes connected in taxi cab manner 2. Two adjacent flying machines cannot be slime (graph coloring coming into play here) 3. Each flying machine must be min. size of 4 with a 3 block contiguous This seems like a really fun challenging problem; a multi vertex cover with max cover size of 12.

I think Carpet mod. Watch his last video. He talks about it.,

@@poppy3318 ill check it out thanks

That's the plan!

In case anybody else is curious, I found it in Tweakeroo tweaks: tweakGammaOverride default set to false, change it to true and it works perfectly!

you can line up the farm in a creative copy of your survival world, write down coordinates and facing of your placement, then use those numbers when lining up schematic in a survival.

@@Sultan-wj3gp interesting. That seems like a lot of extra work though. I wonder if someone came up with something else. I added red and blue concrete "lines" that line up with a top budding amethyst. Then I cleared a large area around the survival world one so i could see. I set the litematica schematic to transparent and went into spectator mode. Then i moved it around until the red and blue lines matched. But was still really hard to get the hang of.

Can't you just line up the budding amethyst blocks? Shouldn't that be rather easy to do assuming nothing caused a rotation? I don't think any step could cause that

@@racernatorde5318 litematica has odd rotation issues for me when I import stuff. But that isn't really the issue. Geodesy creates a large area around it with the flying machine parts. It makes it hard to see where everything is, lining it up takes a long time for me as i have to constantly move stuff back and forth, rotate, move, rotate until by some luck I see how something matches up. *Shrug* it is possible there is a much easier way of doing it, but that's why I am asking.

The method worked only in 1.12

@@niceowl I know this

That's part of tweakeroo, which is a fabric mod.

@@AtomixKingg funnly enough, the tweakroo diagonal block placing feature works on completely vanilla servers! (only the diagonal option, it wont let u place a block apart)

@@mindblow4248 if you use carpet mod you can enable the accurate block placement setting then it works

@@mindblow4248 Easyplace mode in litematica works fine on vanilla servers, that means you can place a block in the middle of nowhere without any surrounding blocks and this is ok

Chunkbase shows the locations, going by eye should be good enough for almost all needs.

@@whocares2277 It's actually not. There are large variations in size so just having a lot in one location might not be the best. In the Scicraft video he mentioned that the location they chose had what I believe was 14 geodes with around 400 amethyst buds. He later said he found another location with more than a thousand amethyst buds so should have chosen that location. He doesn't tell the software he used for that though.

I would assume tinted glass acts like solid blocks internally and are only visually transparent, so wouldn't really be expensive at all.

do you have the geodessy mod downloaded because It isn;t available anymore

@@leohans6646 I just downloaded it the other day and built the machine with the clock. I've got the clock to work for me but i'm sure it's not correct (I have an observer looking at the lower redstone block on the clock which powers a stone block with dust running off down to the first module). I'm only running East / South (no up/down) so just wanted to know the proper way. Additionally, the clock will just stop when it runs out and I have to put the 41 sticks back in the hopper...

But it works lol.

WHO NEEDS THEY GEODUSSY HARVESTED? "omg me"

@@sparkywilson1405 lmao frfr

do you have the geodessy mod downloaded because It isn;t available anymore

it was a bug with the newest minecraft version worked fine on an older one

when can be expect your contribution to automate it more? ;)

@@ilmango I'm working on my own version of this, using Python. About 70% is done I'd say. And yes I'm looking for anyone to contribute to that as well lol

He made a video about that

@@dominiklinker4424 ok

You can’t fit all the stuff in as consistently, so the rates suffer, and there’s more stuff for the item drops to land on

The wiring gets too complex and gets in the way of harvesting some of the buds.

its messy and wiring of pistons can be hard especially if you use a detection based system. there just isnt enough room for pistons, detections, and collection.

This way is much easier for the vast variety of geodes you might find, having to change only a 2d slice of the farm and not an entire 3d contraption

Go watch mango's scicraft blitz day 25 video (from 08/25/2021), there mango had a piston setup which can't beat the flying machine based design. See how complicated the detection based design would be.

Come on...he literally explasns and shows it at that very moment

@@racernatorde5318 Actually, he doesn't

I know this is 3 months old as of now, but for anyone else wanting to know how this is done.... (yes, i realize I'm over explaining some parts) Hover your cursor over the first block. Press forward slash button ( / ) to open command line. Type geodesy then hit Space, now type area, now hit Space, then hit Tab 3 times to fill in the first block's coordinates. This is probably where you're getting stuck..... After hitting tab the 3rd time, hit the Enter key. Now go over to the second block and hover your cursor over it, hit forward slash again to reopen the command line. Now hit the UP arrow on your keyboard to go back to the previous command. Now hit Space. Now you should be able to hit Tab 3 more times to fill in the coordinate of the second block.

@@chadandmina Ah, that's what you meant. I thought that to be common knowledge. Surely that is written everywhere when you look up any kind of fill or clone command tutorial?

To mutch effort to get the extra 10 percent

@@enkvadrat_ less effort this has much more effort than that ,that only has pistons ,observer and simple simple this has much advanced redstone,delicate designs ,honey ,slime blocks need to make it a box and they have 15 farms so that 10% each makes 150%

@@shubhammamania7046 the problem is, his farm isn't a standardised build, he made it specifically for that geode. Now consider building multiple of those, custom to each geode, versus an only slightly less efficient design which can be built in almost exactly the same way every time. It's just less effort to do the latter.

@@o_opinionated719 he didn't specifically but made it because it was near to him and that design is flexible and less effort

@@shubhammamania7046 it's not that a flying machine is less complicated than a hopper line and pistons, it's that the flying machine can be built in the exact same way for many geodes, whilst using sb737's design would result in a unique wiring and hopper-line setup for each geode. So yes, for one single geode it may be easier, but it's harder to do when you're making this farm for many geodes, as in that case repeatability of design is priorotised; as big projects are kind of what scicraft is all about, they use the repeatable flying machine design rather than just sticking pistons all over the place.

Still spamming comments? Just write one proper one. What schematics do you mean?

Can you even make proper mods in bedrock?

@@ilmango you can't do this by hand?

@@Ivan_Santos-Perez you can make the machine (probably, I've heard bedrock redstone acts differently) but you probably can't use the Geodesy mod to help, unfortunately.

@@o_opinionated719 yeah I know you can't use the mod on bedrock I was just wondering if doing this by hand was possible. I don't remember how Redstone works differently so I can't answer that question myself.

@@Ivan_Santos-Perez You basically cannot make complex redstone in bedrock. It will break after some time. The few flying machine tutorials on youtube are full of people complaining about bugs and inconsistency.