Hey everyone, a couple people have commented about their design solutions for a more compact, max efficiency farm, and I want to share their designs here. They use some more clever block updating and seem more reliable than mine. They’re also tileable to increase compactness. Definitely check them out: @inamecomeunico8935 (Tugsteno) - imgur.com/a/100-efficency-sugar-cane-module-rOLxeFZ#oBx66jU @OokamiKuro39Ch: ua-cam.com/video/ze1Qv2bFJic/v-deo.html @AdamSchadow - ua-cam.com/video/zrGEMWbfcdo/v-deo.html (he didn’t comment but Tugsteno shared this video with me) @akiraawooch - ua-cam.com/video/oDASLugeiOo/v-deo.html (another compact tileable design; also discusses efficiency with in-game testing) @pand1024 - ua-cam.com/video/69eOMK3fyDI/v-deo.html (high efficiency, instant harvesting farm with NO observers)
regarding the ending of this video, heres an explanation about redstone being locational and why your sugarcane farm on a big scale could be breaking. ua-cam.com/video/f_JV1aX9aqA/v-deo.htmlsi=HL_2Mwg6lKgjdkFd
You know what's a game changing thing here? Sugar cane cycles being the same as the day/night cycle. No more hopper or comparator clocks, just a daylight sensor from now on
Yeah that’s definitely one of the most interesting findings here. Especially since hopper clocks are so difficult to time correctly and aren’t super accurate
Good video! Production quality is really good and it was easily understandable (I'm jealous). One thing that I don't understand, however, is why you didn't run the farms in Minecraft at like 20x tickspeed for more real-world calculations. If not for calculating the real values (even at 20x game speed it would take like 50 hours), you could run it for a short amount of time to check if your code/assumptions are correct. Minecraft is very janky and there may be factors that weren't included here that might've changed the calculations, like the sugarcane being obstructed when trying to grow under a flying machine, or a firing piston. In the beginning of the efficiency graph the faster you gather sugarcane, the more efficient your farm becomes, but if the gathering mechanism takes up the growing space 50% of the time, that's 50% of the time sugarcane can't grow. That's my two cents for now, thank you.
I’m really glad to hear it was understandable, I had to do a lot of takes to try to make it coherent lol This is a good question, and I probably should have mentioned it. Here’s the story, so bear with me: This sugar cane video was the first of my coding videos I started working on, and I actually built all these farms and started off running in game tests. Programming it hadn’t even crossed my mind when I came up with the idea to test for efficiency. I have the carpet mod which has warp speed (basically running the game as fast as your computer can handle) and that’s what I planned on using. Unfortunately, I underestimated how crappy my laptop is, so it only sped the game up by like 3x and I had the feeling some of the drops weren’t being collected properly. My next idea was to literally just run the tests in real time. And I did. I ran two 12 hour tests (for two of the observer farms). I started with those two cause I didn’t have to worry about hopper clock timings. Then when I moved to the clock based farms, I struggled to get an accurate hopper clock for the intervals (I won’t bore you with how I then spend days figuring out how to use command blocks to make a simple clock 😆). When I did get the command block timer working I could only test one clock intervals at a time for the flying machine cause my computer really doesn’t like those. Anyway, I didn’t think my in-game testing was going well, it was taking too long, and I wasn’t even sure 12 hour testing would give nice results. Enter, my thought to try to program it. Once it was done, I felt comfortable presenting it and saying “I programmed sugar cane farms” cause when I compared my in-game test data to the program results, they matched up pretty well. I’m sure in the long run, the game will have some glitchy stuff so my code should be taken as a perfectly smooth running game that is pretty close to the real thing. Also, I wrote a program for each farm type and did take into consideration the harvesting style. For the sake of video length I didn’t go into a ton of detail but I did program a “flying machine” that stops my sugar cane from updating when it’s moving over it for example. Granted, I’m sure it’s not prefect, but I’d say overall it’s a decent estimate. Lastly, I assumed 100% lossless drop collection just to compare the best case scenario across the board. I think the hopper minecarts below to collect is the best way and as long as the cart covers all its path before the drops despawn it should be pretty close to lossless.
First off, wondeful video. I find it pretty wild how some people are still trying to put the video down? Saying python is a childs language and that this is just a huge waste of time. Sure, you could probably just build the farms and raise the game ticks, but i dont believe thats what you were trying to prove. To me this is more of a demonstration of skill. You wanted to see if you could simulate something that is seemingly random to the average user.. and you did not just that, but also explained it in a very digestable way. You hit both the 'minecraft nerd', the 'developer/coder', and the 'new player' groups without making either feel left out. Amazing job!
Thanks so much, I appreciate the support! I’m glad the video came across as understandable, I definitely put a lot of effort into that lol And to be fair, I think a lot of people are right in saying python is a slow langue. I like it cause it’s got a low barrier to entry and has a lot of support for it online. So it’s super easy to use
@@SeaJayPlays Bit late here but this is one the exact use cases for Python, easy to use language for data visualization. If you are working on a service that needs to be insanely fast and reliable then maybe not your first choice, but realistically it is not normally a valid argument. Pointless trying to use other languages that people who don't know much about coding refer to as 'better languages' - they all have their use cases and everyone will have their preferences. Great video!
something to note, with mud blocks, you can place cane and even bamboo on top of it, but if you place hoppers below it still picks up drops, meaning you can save space on a minecart
Just watched this video, and I was amazed by the amount of useful and helpful information you talked about on this video. Great job. Hope to see more videos like this one!
for the single piston fired per sugar cane farm, you can also use redstone dust instead of note blocks. the whole line updates but only the piston underneath the triggered observer pulses
This was addressed in the video, no? It is the IAmAnIntrovert's design (12:10) which he says he doesn't go with so it's easier to stack each farm unit.
I went down the same rabbit hole a couple of years ago. I didn't go as deep though, but I did go through the code to calculate the optimal delay between sugarcane harvests.
Id like to see some datapoints for "entities per 'time', avg/peak" and "redstone/light updates per 'time', avg/peak" That would help decide best farm for single-player or server and industrial district vs passive
This content is exactly what I'm looking for while I'm in grad school for business analytics. Great stuff, and I'd love to see more 👍 . . . . . . You can make your code for generating chunks more efficiently as follows: import itertools # Define the range of elements (0 to 15) elements = range(16) # Generate all tuples of length 3 tuples = itertools.product(elements, repeat=3) # Convert to list (optional) or iterate through tuples_list = list(tuples) . . . . . . More your graphs, I'd recommend switching to a line chart/area chart so that those pesky vertical lines don't show up. I'd also consider color coordination either by farm type or metric.
Hey, thanks for the comment, these are some really good suggestions! I’ll make sure to implement them in the next one. I’m surprised I didn’t bother color coding the graphs to match the farms already lol
Creating a new comment for this as I think I've found the issue with the instant harvest design. I believe the standard 3 block tall observer note block sugarcane design that was compared to the "100% efficient" design has it's data misrepresented. I've made a video on my channel on the topic, but I'll type it out here as well. First, we clarify some terms on the phases a farm goes through. Growing Phase- waiting on sugarcane to be random ticked Harvesting Phase- Piston extending to break the block Reset Phase- Time needed after harvesting before sugarcane can grow again Footprint/Density- How much space a design takes up Tileability- How well you can stack a design next to/on top of another design. written as n-wide tileable (1wt for 1 wide designs that sit next to each other) First of all, I want to establish that there's no difference between a 2 block tall sugarcane waiting to grow, and a 1 block tall sugarcane waiting to grow. they both statistically have the same chance to grow and are statistically independent. With that in mind, there's technically no efficiency between allowing a sugarcane to grow to a full 3 block before harvesting, compared to harvesting it at 2 blocks high. the only difference is that it takes longer until first harvest when you are done setting up the farm, but there is no sugarcane lost or time wasted at max grow height. Design 1: Observer note block piston design. Design 2: the proof of concept 100% efficiency build in this video Design 3: My design shown in my video on my channel (and the alternative linked in the comments) First we look at the density, Design 1and 3 has a growing sugarcane every 12 blocks, while design 2 has a growing sugarcane every 16 blocks(32 blocks if factoring in spacing to avoid signal crossing. both not factoring in collection system. This makes design 2 33% (166%) less efficient in this factor. Next we look at Harvesting and Reset time, Design 1 and 2 has a 2 tick harvesting time and a 0 tick reset time. Design 3 has a 6 tick harvesting time and a 2 tick reset time(2 ticks not affecting sugarcane growth in this scenario). so Design 3 is 50% less efficient than Design 1 and 2 Now we look at tileability, Design 1 can be tiled on top of each other without any additional blocks added, but requires mirroring the tile right above it for best compactness. Design 2 requires 1 extra block space if using hopper to collect the drops, and 2 blocks if using hopper minecart due to lack of any blocks to place rail on. Design 3 can stack right on top of each other without any added scaffolding blocks (care needs to be taken to use lever to power a solid block next to the rail instead of using any hard powering blocks or powering the block next to the rail). In this factor, design 2 is less efficient. In conclusion, I believe the "inefficiency" measured for Design 1 is due to the calculations not factoring in that sugarcane takes time to grow before harvested. This is supported by ingame testing by measuring the farm efficiency right after setting up, and measuring the efficiency after allowing it to run for a few hours. It also shows that Design 1 is superior to Design 3 (and it's alternative designs) at a margin that can be considered a margin of error. All things considered, Design 1 is superior in terms of simplicity and resistance to breaking. addendum: I'm unable to test OokamiKuro's design, but factoring in the above listed concerns, this is likely the best design as it's much more densely packed and the longer harvest and reset time likely get's blurred with margin of error due to it being such a minor issue.
got to say that I didn't noticed that detecting sugarcane on the bottom works because there isn't a block stage change at all, even the age number remains static and the observer is going off and I don't know why.
@@akiraawooch I appreciate the thorough investigation and comment! It’s been so long since I made this video/code that I’d have to go back and review it before I could say anything about your findings, but they seem likely. If you want me to link your video in that pinned comment, let me know!
@@akiraawooch So this is something I found out shortly after posting this video. If you use F3 to see a sugar cane block’s age, it will never change until you update a block. So if you break the second layer, you can see the age change of the block below. From what I found, this is because that age on F3 is not updated on the client side as the developers didn’t find it necessary to always refresh, thus helping the game run a bit faster. I wasn’t aware that an observer could see the sugar cane updating until I saw OokamiKuro’s design. I think the compactness alone makes it my favorite max efficiency design so far
@@SeaJayPlays would appreciate if you add it, I didn't add it in the initial post cuz youtube tend to remove comments with links in them. ua-cam.com/video/oDASLugeiOo/v-deo.htmlsi=LRhdio1Ll-BylHpZ
@@SeaJayPlays so I did the math, it's not any more compact that the normal "Design 1" observer- noteblock-piston design. the primary issue with that particular design is the complexity of stacking the design while using a hopper minecart. it still takes up a 12 block space or 3x4 space because you need an extra block there for rails to go on, and this is using an improved design I based off of ookami's design so that it allows for the block to be placed within a 3x4 footprint, and I still can't figure out how to stack it yet, so that might take a few days or weeks to figure out, I have one way of doing that right now, but I'm not really satisfied with how it's done. stacking is a major problem with every instant harvest farm, my own design took a whole month to figure out how to stack it seamlessly. and ookami's design's compactness is it's own downfall as it's extremely hard to power the rails without accidentally powering something else.
I think this might be the most interesting and captivating Minecraft video I have ever watched. Mostly because I just love programming, but also because your ability to explain these things so clearly and in a comprehensive way, I liked this video. First video I have seen from you, but definitely earned a sub! Criminally underrated channel. Edit: I went to read the comments, and I have never seen any creator reply to comments with such care. Long and thoughtful answers to interesting questions, and coming back a month later to notify someone that a video was uploaded about a topic they asked.
Wow, I really appreciate the compliment, that’s a very nice thing to say! I spent a while getting this video structured in a way I liked, so thank you. I appreciate the comments and feedback, so I’m always happy to write back. I’ve got a few more farm programming videos and the chess video that you might like if you’re looking for some more like this!
As a financial graduade I'm sorry to break it to you, but we will need to calculate a tax on it, but, there is an option to postpone the payment for the next financial reporting period and recalculate a present value of the postponed tax so that we can max the efficiency with our crops!
I was hoping to see one more design. The burnout torch pistonified sugarcane farm. Cheaper than the observerful ones. But what a great video, thank you!
@@SeaJayPlays yeah slacklizard has a video about the burnout torch farm. Can be used for bamboo, sugarcane, melons, pumpkins with some modifications. Cheapest material needs
Interesting project. I love it when people start programming to solve rare and unique problems. Reminds me a lot of Gnembon. But some critism is you need to focus on your code quality. The performance and maintainability of this piece is quite low. Use functions, classes and avoid those huge indentation needed for, for loops with if conditons with more conditons. You should also have a look at Numpy, its quite a bit harder to learn. But is well suited to this task as you can avoid those slow loops and be faster than any other language. The code would also be simpler since youd just be doing formulas rather than pure logic.
Cool video! I like the approach, though I must say I was confused on why using a code to get efficiency and rates if a simple tick warp test could do the same. I struggle to see what the code adds to the comparison basically. I'd expect that with python you can create a function of a general sugarcane farm and use an optimizer like Pymoo for a multi objective optimization of key design variables of farms, so you can get as an output the 3D pareto front of sugarcane farming. That would be cool! Still, I enjoyed the video too
@@FB_engine Thanks! As for why I decided to program it, that’s a longer answer. A couple other commenters have asked similar questions, and I’ve written a pretty long reply as to why, so definitely check those out if you’d like. But short answer is, I tried in-game testing and tick warpping but my laptop just couldn’t handle it. Also, in the code I programmed each farm individually to account for harvest methods, and I never looked at the code as an optimizer, just something that allowed my computer to run “in-game” tests. I could look into optimization for a future video though!
Extremely underrated video :O I would love to see an iron farm next :D Little tip before you eventually blow up: I could see the Outro with the many different colors triggering seizures for epileptic people.
@@nextraven Thanks! I’m not sure I’m cut out to do an iron farm lol but I can look into them Thanks for letting me know, when I get the chance, I’ll cut the last few seconds off
Looking at the small bits of code you presented, I really urge you to look into vectorizing your data generation and processing. Loops in python are incredibly slow, and if you instead of e.g. processing each block in the sub chunk separately were to do all calculation “at the same time”, whatever that means in a computer, you could probably decrease your simulation runtime by a factor of at least 20, possibly around 40.
Thanks for the tip, I’ll definitely look into that cause the slowness has been super annoying. I’ve just been writing these programs in these simple steps so it’s easier to do and come back to later. But these run for like 24 hours lol
@@SeaJayPlays I mean crop farms like sugarcane are definitely more simple to simulate as there's less stages and game mechanics than something like an iron farm or a gold farm but I'd reckon those are more useful. Making an iron farm at the start of the game is always such a hassle so if you could make something more compact than I would be forever grateful
@@kxdsh I’ll take a look into them! I’m not too sure I’d be able to find a better solution that what’s out there right now just cause the mechanics are so well known and nearly maxed out. But I could give it a shot
Is there any change in rates if you just harvest the top most sugarcane and leave it at stage 2, making it more likely to receive a random tick because it is 2 blocks tall? This would allow for a shorter clock time or if you wanted to do an observer based one. Have the observer face pointing down so it activates when the sugarcane becomes 3 blocks tall. Subscribed, hope to see more videos like this, reminds me of Nico is Lost
@@MasterElements Thanks! His iron farm video inspired a lot of this video lol This is an interesting question, and the short answer is basically no, there’s no change. I’ve got a longer explanation on the individual 100% efficiency farm video, but here’s the gist. Since every block (and random tick) is independent of each other, having more blocks below doesn’t increase the likelihood of getting a random tick. And in terms of efficiency, the three tall design approaches 100% efficiency because in the time it takes for the third layer to grow, you could have already harvested the second layer. But as time goes on, this small inefficiency basically vanishes. So the three tall vs two tall are pretty much the same.
commenting to boost the algorithm, as this is my fav nerdy minecraft channel that ive found since i found mattbatwings before he blew up, i rly hope u make it big and blow up too!
Interesting video, although I'm curious why you decided to abstract out the farming logic into python? I understand wanting to get longer time intervals, but my thought would have been to use carpetmod, which will make the game speed as fast as your cpu can deliver ticks (/tick warp). This is a highly effective tool for analyzing farms like this one with the ability to detect sugar cane drops, calculate rates, etc. I would also put into question how 1-to-1 your code is to the minecraft environment, since there might be confounding factors that deviate your results in python from experimental results.
I appreciate it! Funny enough you’re the second person to ask this question today, so I’ll just copy my reply longer below. But basically I started off by using carpet but my laptop just couldn’t handle running the game like that and real time tests just took forever. I don’t think my code is 100% accurate to the game but I did compare my in-game test to my program results and they matched up pretty well, at least well enough for me to feel comfortable making this video/claim. My longer reply: This sugar cane video was the first of my coding videos I started working on, and I actually built all these farms and started off running in game tests. Programming it hadn’t even crossed my mind when I came up with the idea to test for efficiency. I have the carpet mod which has warp speed (basically running the game as fast as your computer can handle) and that’s what I planned on using. Unfortunately, I underestimated how crappy my laptop is, so it only sped the game up by like 3x and I had the feeling some of the drops weren’t being collected properly. My next idea was to literally just run the tests in real time. And I did. I ran two 12 hour tests (for two of the observer farms). I started with those two cause I didn’t have to worry about hopper clock timings. Then when I moved to the clock based farms, I struggled to get an accurate hopper clock for the intervals (I won’t bore you with how I then spend days figuring out how to use command blocks to make a simple clock 😆). When I did get the command block timer working I could only test one clock intervals at a time for the flying machine cause my computer really doesn’t like those. Anyway, I didn’t think my in-game testing was going well, it was taking too long, and I wasn’t even sure 12 hour testing would give nice results. Enter, my thought to try to program it. Once it was done, I felt comfortable presenting it and saying “I programmed sugar cane farms” cause when I compared my in-game test data to the program results, they matched up pretty well. I’m sure in the long run, the game will have some glitchy stuff so my code should be taken as a perfectly smooth running game that is pretty close to the real thing. Also, I wrote a program for each farm type and did take into consideration the harvesting style. For the sake of video length I didn’t go into a ton of detail but I did program a “flying machine” that stops my sugar cane from updating when it’s moving over it for example. Granted, I’m sure it’s not prefect, but I’d say overall it’s a decent estimate. Lastly, I assumed 100% lossless drop collection just to compare the best case scenario across the board. I think the hopper minecarts below to collect is the best way and as long as the cart covers all its path before the drops despawn it should be pretty close to lossless.
@@visxosity8690 That seems like a pretty intimidating thing to get into lol If you’ve got a specific problem you’d like solved I might be able to look into that, but doing a full scale system might be pretty far in my future
I think the max eff farm is locational because of QC + dust update order. If you replace the solid blocks with glass I think it shouldn't break anymore.
@@SeaJayPlays I don’t have access to it as I am unhoused but it was a scalable variation on a wall farm with one height growing spots and observers that fired the edge of the square as long as the red stone signal would go
And I thought this video was about a program randomly generating different farm designes with an ai and then upgrading them with each generation to find the best possible one
random numbers by computers are not entirely random. So to make the simulation better, you should use the same random formula minecraft uses to generate the coordinates of random ticks.
@@Capiosus If I knew how Minecraft did it I would definitely add that to my code. Unfortunately I don’t know how to find that, so we’ll just have to assume the rand function is close. Ilmango did a video on clock timing for sugar cane and ran some in-game tests that showed the best interval was 41 minutes. That matches all of my data, so I’d say my programs are a good estimate. If you know where I can find more details on the Minecraft code, let me know!
@@SeaJayPlays minecraft java edition has it's mappings available for anyone, so all you need is a copy of the game client's versions' .jar file and the json which lists the URL of the client mapping. Then you can read minecraft's code using a java decompiler tool.
By the way, the randomness of redstone should hopefully be fixed in the latest snapshot if you turn on the experimental redstone feature, since thats basically what they changed - before that there are some undefined states in which basically some stuff like hashes in the backgrounds or chunks or stuff like that determine the outcome instead of a true random number generator, creating location based behaviour sometimes.
In your "100% efficient" farm, the piston still blocks the growth for a few ticks. For the truly pointless true 100% you would have to use TNT or something else to harvest without blocking.
Yeah I kinda ignore that lol I guess we can call it 100% efficiency for piston farms. But I’ll look into other breaking methods for future builds, thanks!
I dont watch even minecraft youtube much less farm related videos and this video was right above another one called "How I made the WORST sugarcane farm in minecraft" 😭
I've managed to compact the "full efficient" design down to a 5x3 form factor that is stackable. observer facing the sugarcane will 0 tick budpower a sticky piston, which pushes an observer facing the opposite direction upwards, it then will hard power a full block, which the piston next to it will push down, breaking the sugarcane. that will in turn be observed by the observer, triggering the sticky piston to pull down the observer, thereby resetting the design. the only issue here is that sideway stacking of multiple modules may have some issues where it creates a observer clock. solution is to stagger it upward 1 block, or leave an airgap. I'll post a discord image link in the reply, cuz youtube will nuke this comment if I add it here.
@akiraawoock That sounds pretty interesting, I’d love to check it out. Another person in the comments recently sent me their tileable design and it seems to work pretty well (also uses pistons pushing observers) So it’ll be cool to compare them. I’ll post a link to his design once I’ve got one. Btw I don’t have discord so I’m not sure I’ll be able to see your image. Sending through my planet Minecraft account would work if you’d rather do that.
so, you count how many blocks of sugar cane you can fit on a subchunk and just run the math (is it a sugar cane - is it 3 blocks already - is the random tick hitting it) on that number to grow? that's what people do for 10 years+
Is this for the Java version? If so then for the "100% efficiency" category it's much more efficient to use budded pistons. It's dramatically cheaper than any of the designs even the ones you have pinned here. A budded piston design can also be very tile-able btw. Something you didn't explain that I think is very important is that when a sugarcane reaches the final growth state (15) it doesn't reset when the block above it is harvested. This is very important for one-piston-per-sugarcane designs.
@@pand1024 Yeah this is in Java. I don’t have much experience with bud powered pistons and I’m not sure how I’d implement them here. If you’ve got a design idea I’d love to see it! So I was under the impression the sugar cane did reset when the block above gets harvested. Can you explain what happens and where you were able to find that info, it’d be really helpful for future videos. Thanks!
@@SeaJayPlays Let me see if I have time to to put together a video showing the design I have in mind. Regarding the block states thing looks like the wiki directly contradicts what I'm saying so maybe I'm wrong. "A freshly planted cane - and a cane that has just grown cane above it - each have an age of 0". I did most of my original work back in 1.8 (before observers existed)
@@pand1024 Great video, I just watched it. I had no idea this bud powering was possible. I only started getting into redstone about a year ago so I’ve got more to learn. I agree it’s not 100% efficient but it definitely scores higher for me than most farms due to resource efficiency. I’ll add the link to your video in the pinned comment, thanks!
That’s a good question, cause there are some size requirements. But most likely not. In terms of clock timing, you’ve just gotta make sure the machine can traverse the whole field before the time is up, so usually this won’t be a problem. If you’re wanting to use a faster clock interval, you can build it so the machine goes one direction then stops until the next clock interval triggers it to come back. Also, if you’re going for max resource and drops efficiency, I’d keep the sugar cane area thin enough to fit in one chunk and then only go as far as you can load. I think it’s 16 chunks long, and you’d afk in the middle.
@@apdj94 Here in my base, just bought this new horse here. It's fun to ride up here in the grasslands biome, but you know what I like a lot more than materialistic things… knowledge. In fact, I'm a lot more proud of these 7 thousand new bookshelves that I had to get installed to hold 2 million new books that I crafted (thanks to the sugar cane). Cause it's like the billionaire Warren Buffett, says the more you learn the more you earn.
First you forgot cactus in Redmond ticks and second you can place then diagonally when a water is nearby. Could you Test if my setup is any better than the average player including yous pls? I place my sugar cane in a Chess like pattern where sand and Water interchange where the waterlocked blocks are the the black titles and the sand are the white titles then I place the Sugar cane and run flying machines over it . I also stack the fields on top of another to maximize production. Could you also simulate cactus farm pls?
@@icebluscorpion Hey there, not sure what that first part is. But for the second question, surprisingly, the chess board pattern is less efficient (50% sugar cane) than this strange pattern (~80% sugar cane) but this pattern is different to make. It’s definitely a good idea to stack the farms so you can max out efficiency based on the chuck borders, nice job! Lastly, I probably won’t do any cactus farm videos just cause we already know the best way to farm them and can easily do instant farming (max efficiency). The hard part is designing how to capture all those drops. I think ilmango has a good video where he tests some common designs if you want to check that out!
I think I did do it based on the ticks, but I’d have to double check to make sure. What kind of inefficiency in the code are you talking about? I’d like to correct that for the next video if it’s an issue.
@@SeaJayPlays Assume that the code runs slower due to some reason (more things happening for example) and potentially misses a tick timing, running it based on real-time would potentially have fewer cycles over the same time period. Systems are also imprecise, so even if you say something is going to run after exactly n milliseconds you still might have some minor discrepancies. I don't think it'd change the results in a huge way, but just one of those things one can do in order to keep things accurate. Just letting it run x amount of ticks is kind of all one needs to do.
@@astronomy2_0 I wasn’t able to get it, but some other people have commented with their 100% efficiency, tileable designs. I’ve posted some links to their designs in a pinned comment if you want to check those out!
@Everesl, Hi, I could look into minecart unloaders, but to be honest, they can get pretty weird and use some strange mechanics that I'm not familiar with. In some of my farms, I actually just use Ilmango's instant unloader. It's not too difficult to build and is tileable. But, if you want an unloader for a different situation that his isn't good for, let me know what you are looking for, and I could definitely try to make something. I've never looked into shulker box mechanics or unloaders so that could be fun to try! I'm working on a video idea right now, but once I'm done with it I'll look into these. Thanks for the request!
@Everesl Hey, I just uploaded a video going over two shulker box unloader designs. There’s a world download for them as well. Let me know if those help.
@@OokamiKuro39Ch Hi, so if you’re talking about his compact, lossless version, then that would match the blue farm (group-based observer-activated farm) since his relies on one observer to activate the whole line. I forget the rates, but I think I have a graph showing the efficiency for different group sizes so you can check out which matches his most closely. If he’s got a different farm let me know
@@SeaJayPlays hmm... I do have a design that is one wide tileable with the observer looking at the bottom cane, uses 2 observer, 1 piston and 1 sticky per cane. That would be 100% efficient?
Sounds like it! As long as the piston harvests at the second layer as soon as it grows, it should be 100% Edit: If it’s something you’ve got posted somewhere and want shared, I’ve linked some other people’s compact, 100% efficiency farms in a pinned comment
I the smaller tests I did in-game seem to support the results, but large scale testing didn’t work out too well for me. But I did see an Ilmango video where he ran large farms and got similar results
@@OMEGAMAN1300 I know redstone is pretty different on Bedrock, but I can definitely look into programming the farms differently to account for that. For the next one I do, I’ll definitely add it!
@@OMEGAMAN1300 Unfortunately I did haha but good suggestion still. Some other people have shared their more reliable 100% efficiency farms (pinned comment) that don’t have this issue so if you’re looking to use one, I’d start there!
Yep I know what you mean! And yes the same block can receive all three random ticks in a single game tick. It’s just a 3/4096 * 3/4096 * 3/4096 = 0.000000039290171% chance lol
@@lurkkilukki7877 That’s a good point! For me, they all have the ability to crash my computer so it’s hard to tell. I’d assume the noteblock farm would be less laggy but more steady and the flying farm would be large spikes due to all the drops but no lag for most of the time.
You’ll have to remind me what that design is. But if it still uses a flying machine the results should match what I found in this video since I’ve assumed every farm tested is 100% lossless.
@@wrtltable So after searching through his videos I couldn’t seem to find the video you’re referencing, all the ones I found had some sort of flying machine. But for efficiency there’s really only two categories: clock-based harvesting vs. instant/near-instant harvesting. So as long as the farm relies on a clock (flying machine, clock-based piston) then it’ll have that 70-80% max efficiency. Then if you have a near instant harvesting one (relies on individual operation, not like the group based observer farm) then you can approach that 100% efficiency.
Shit, mc has keep so many kids off the street and now they should be grown, pursuing higher education/starting a family. I wouldn’t be surprised how many kids got into numbers because of mc. Ik I did. I remember watching direwolf20 make solar flowers and emc flower farms and wondering, what’s the best combination to get the number higher!
Completly non practical but I wonder if in theory because you are able to simulate the data you could write one of those evolving AI models to generate a farm scored on the output, I mean it wouldent make a farm better then a human (because we already have near perfect farms) but it would be interesting to see what it attempts
@@that_guy1211 lol trust me, that was my intention. A couple other people have asked now, and I posted a pretty long reply if you want to check it out. But basically my computer couldn’t handle it. I did do in-game tests to validate my programs, so I know they’re decently accurate. I’m sure my code’s not the best but it turned out to be pretty enjoyable to work on.
@@SeaJayPlays oooooh! Good to hear, unfortunate for your PC tho, sad to hear that, but nice work mah dude! Programming is one of the things i least like doing in my free time, but can't help but be interested in doing it
Hey everyone, a couple people have commented about their design solutions for a more compact, max efficiency farm, and I want to share their designs here. They use some more clever block updating and seem more reliable than mine. They’re also tileable to increase compactness. Definitely check them out:
@inamecomeunico8935 (Tugsteno) -
imgur.com/a/100-efficency-sugar-cane-module-rOLxeFZ#oBx66jU
@OokamiKuro39Ch: ua-cam.com/video/ze1Qv2bFJic/v-deo.html
@AdamSchadow - ua-cam.com/video/zrGEMWbfcdo/v-deo.html (he didn’t comment but Tugsteno shared this video with me)
@akiraawooch - ua-cam.com/video/oDASLugeiOo/v-deo.html (another compact tileable design; also discusses efficiency with in-game testing)
@pand1024 - ua-cam.com/video/69eOMK3fyDI/v-deo.html (high efficiency, instant harvesting farm with NO observers)
first link not worked
@@wrtltable Thanks for letting me know! It should work now
Can you remove the trackers on the links? Everything after and including the ‘?’
regarding the ending of this video, heres an explanation about redstone being locational and why your sugarcane farm on a big scale could be breaking.
ua-cam.com/video/f_JV1aX9aqA/v-deo.htmlsi=HL_2Mwg6lKgjdkFd
@@Volt64bolt Sure! I didn’t know those were trackers, thanks
in my opinion you should do more of these simulations, i would love to see lots of farms being generated
I’m glad you like them! I’ve got a few more farms lined up to do, so hopefully I can find the time to make those videos
You know what's a game changing thing here?
Sugar cane cycles being the same as the day/night cycle. No more hopper or comparator clocks, just a daylight sensor from now on
Yeah that’s definitely one of the most interesting findings here. Especially since hopper clocks are so difficult to time correctly and aren’t super accurate
@@SeaJayPlays iirc ilmango did a similar research with pure math and came up with a similar conclusion.
I've always just pointed an observer at the block above a cactus for my clock. Might not reliably be 100% efficient but it's good enough
@@tunafllsh and max hanke made a video on it as well
I think I'll go with planting a handful on the coast and harvesting whenever i feel like it
😂
This is very underrated. I can't believe you only have 500 subs
Thanks I appreciate that!
Bruh gained almost 350 since this post. I love the technical aspect of Minecraft and this is scratching an itch I didn’t know I had.
@@OMEGAMAN1300 the key was to just stop posting videos I guess 😂
I’m glad you’re liking the technical stuff
@@SeaJayPlaysdon’t worry I will be one of them
It a really well place comment (From someone subscribe when he had 1.22k sub)
Good video! Production quality is really good and it was easily understandable (I'm jealous).
One thing that I don't understand, however, is why you didn't run the farms in Minecraft at like 20x tickspeed for more real-world calculations. If not for calculating the real values (even at 20x game speed it would take like 50 hours), you could run it for a short amount of time to check if your code/assumptions are correct. Minecraft is very janky and there may be factors that weren't included here that might've changed the calculations, like the sugarcane being obstructed when trying to grow under a flying machine, or a firing piston. In the beginning of the efficiency graph the faster you gather sugarcane, the more efficient your farm becomes, but if the gathering mechanism takes up the growing space 50% of the time, that's 50% of the time sugarcane can't grow.
That's my two cents for now, thank you.
I’m really glad to hear it was understandable, I had to do a lot of takes to try to make it coherent lol
This is a good question, and I probably should have mentioned it. Here’s the story, so bear with me: This sugar cane video was the first of my coding videos I started working on, and I actually built all these farms and started off running in game tests. Programming it hadn’t even crossed my mind when I came up with the idea to test for efficiency. I have the carpet mod which has warp speed (basically running the game as fast as your computer can handle) and that’s what I planned on using. Unfortunately, I underestimated how crappy my laptop is, so it only sped the game up by like 3x and I had the feeling some of the drops weren’t being collected properly. My next idea was to literally just run the tests in real time. And I did. I ran two 12 hour tests (for two of the observer farms). I started with those two cause I didn’t have to worry about hopper clock timings. Then when I moved to the clock based farms, I struggled to get an accurate hopper clock for the intervals (I won’t bore you with how I then spend days figuring out how to use command blocks to make a simple clock 😆). When I did get the command block timer working I could only test one clock intervals at a time for the flying machine cause my computer really doesn’t like those.
Anyway, I didn’t think my in-game testing was going well, it was taking too long, and I wasn’t even sure 12 hour testing would give nice results. Enter, my thought to try to program it. Once it was done, I felt comfortable presenting it and saying “I programmed sugar cane farms” cause when I compared my in-game test data to the program results, they matched up pretty well. I’m sure in the long run, the game will have some glitchy stuff so my code should be taken as a perfectly smooth running game that is pretty close to the real thing. Also, I wrote a program for each farm type and did take into consideration the harvesting style. For the sake of video length I didn’t go into a ton of detail but I did program a “flying machine” that stops my sugar cane from updating when it’s moving over it for example. Granted, I’m sure it’s not prefect, but I’d say overall it’s a decent estimate. Lastly, I assumed 100% lossless drop collection just to compare the best case scenario across the board. I think the hopper minecarts below to collect is the best way and as long as the cart covers all its path before the drops despawn it should be pretty close to lossless.
0:30 Structuring your project before you code? Writing the outlines per hand? Heresy! Are you trying to make us all look bad?
@@4crafters597 nah I just cleaned it up for the video’s sake 😆
First off, wondeful video. I find it pretty wild how some people are still trying to put the video down? Saying python is a childs language and that this is just a huge waste of time. Sure, you could probably just build the farms and raise the game ticks, but i dont believe thats what you were trying to prove. To me this is more of a demonstration of skill. You wanted to see if you could simulate something that is seemingly random to the average user.. and you did not just that, but also explained it in a very digestable way. You hit both the 'minecraft nerd', the 'developer/coder', and the 'new player' groups without making either feel left out. Amazing job!
Thanks so much, I appreciate the support! I’m glad the video came across as understandable, I definitely put a lot of effort into that lol
And to be fair, I think a lot of people are right in saying python is a slow langue. I like it cause it’s got a low barrier to entry and has a lot of support for it online. So it’s super easy to use
@@SeaJayPlays Bit late here but this is one the exact use cases for Python, easy to use language for data visualization. If you are working on a service that needs to be insanely fast and reliable then maybe not your first choice, but realistically it is not normally a valid argument. Pointless trying to use other languages that people who don't know much about coding refer to as 'better languages' - they all have their use cases and everyone will have their preferences. Great video!
something to note, with mud blocks, you can place cane and even bamboo on top of it, but if you place hoppers below it still picks up drops, meaning you can save space on a minecart
That’s good to know! Definitely helpful for the smaller/micro fams
Just watched this video, and I was amazed by the amount of useful and helpful information you talked about on this video. Great job. Hope to see more videos like this one!
Thanks so much, I’m glad you liked it! I’m currently working on another one like this so stay tuned!
for the single piston fired per sugar cane farm, you can also use redstone dust instead of note blocks. the whole line updates but only the piston underneath the triggered observer pulses
Good point!
This was addressed in the video, no? It is the IAmAnIntrovert's design (12:10) which he says he doesn't go with so it's easier to stack each farm unit.
@@MrCookieMonster565 Oh yeah, it’s been a while so I forgot
Always gonna support someone who also loves redstone
@@LemmonTea I appreciate that!
I think you might also like my chess video, definitely redstone intensive
I went down the same rabbit hole a couple of years ago. I didn't go as deep though, but I did go through the code to calculate the optimal delay between sugarcane harvests.
@@stephaneduhamel7706 glad I’m not the only one
Id like to see some datapoints for "entities per 'time', avg/peak" and "redstone/light updates per 'time', avg/peak"
That would help decide best farm for single-player or server and industrial district vs passive
For the next code video I do, I’ll add those! Good suggestions
This content is exactly what I'm looking for while I'm in grad school for business analytics. Great stuff, and I'd love to see more 👍
. . . . . .
You can make your code for generating chunks more efficiently as follows:
import itertools
# Define the range of elements (0 to 15)
elements = range(16)
# Generate all tuples of length 3
tuples = itertools.product(elements, repeat=3)
# Convert to list (optional) or iterate through
tuples_list = list(tuples)
. . . . . .
More your graphs, I'd recommend switching to a line chart/area chart so that those pesky vertical lines don't show up. I'd also consider color coordination either by farm type or metric.
Hey, thanks for the comment, these are some really good suggestions! I’ll make sure to implement them in the next one. I’m surprised I didn’t bother color coding the graphs to match the farms already lol
Absolutely! Do you have a GitHub repository for the code you've used? I'd be really interested into looking into it.
Sorry I’ve never used GitHub before. I’ll have to check it out!
Gotta love tech minecraft. Keep it up bro
Thanks!
really really cool
UA-cam is really showing me some cool channels lately
Thanks!
I’m glad UA-cam’s doing something right lol
Your visualizations are very intuitive and nice
@@michaelmiller9449 They take forever to make, so I appreciate that!
Creating a new comment for this as I think I've found the issue with the instant harvest design.
I believe the standard 3 block tall observer note block sugarcane design that was compared to the "100% efficient" design has it's data misrepresented. I've made a video on my channel on the topic, but I'll type it out here as well.
First, we clarify some terms on the phases a farm goes through.
Growing Phase- waiting on sugarcane to be random ticked
Harvesting Phase- Piston extending to break the block
Reset Phase- Time needed after harvesting before sugarcane can grow again
Footprint/Density- How much space a design takes up
Tileability- How well you can stack a design next to/on top of another design. written as n-wide tileable (1wt for 1 wide designs that sit next to each other)
First of all, I want to establish that there's no difference between a 2 block tall sugarcane waiting to grow, and a 1 block tall sugarcane waiting to grow. they both statistically have the same chance to grow and are statistically independent.
With that in mind, there's technically no efficiency between allowing a sugarcane to grow to a full 3 block before harvesting, compared to harvesting it at 2 blocks high. the only difference is that it takes longer until first harvest when you are done setting up the farm, but there is no sugarcane lost or time wasted at max grow height.
Design 1: Observer note block piston design.
Design 2: the proof of concept 100% efficiency build in this video
Design 3: My design shown in my video on my channel (and the alternative linked in the comments)
First we look at the density, Design 1and 3 has a growing sugarcane every 12 blocks, while design 2 has a growing sugarcane every 16 blocks(32 blocks if factoring in spacing to avoid signal crossing. both not factoring in collection system. This makes design 2 33% (166%) less efficient in this factor.
Next we look at Harvesting and Reset time, Design 1 and 2 has a 2 tick harvesting time and a 0 tick reset time. Design 3 has a 6 tick harvesting time and a 2 tick reset time(2 ticks not affecting sugarcane growth in this scenario). so Design 3 is 50% less efficient than Design 1 and 2
Now we look at tileability, Design 1 can be tiled on top of each other without any additional blocks added, but requires mirroring the tile right above it for best compactness. Design 2 requires 1 extra block space if using hopper to collect the drops, and 2 blocks if using hopper minecart due to lack of any blocks to place rail on. Design 3 can stack right on top of each other without any added scaffolding blocks (care needs to be taken to use lever to power a solid block next to the rail instead of using any hard powering blocks or powering the block next to the rail). In this factor, design 2 is less efficient.
In conclusion, I believe the "inefficiency" measured for Design 1 is due to the calculations not factoring in that sugarcane takes time to grow before harvested. This is supported by ingame testing by measuring the farm efficiency right after setting up, and measuring the efficiency after allowing it to run for a few hours.
It also shows that Design 1 is superior to Design 3 (and it's alternative designs) at a margin that can be considered a margin of error.
All things considered, Design 1 is superior in terms of simplicity and resistance to breaking.
addendum: I'm unable to test OokamiKuro's design, but factoring in the above listed concerns, this is likely the best design as it's much more densely packed and the longer harvest and reset time likely get's blurred with margin of error due to it being such a minor issue.
got to say that I didn't noticed that detecting sugarcane on the bottom works because there isn't a block stage change at all, even the age number remains static and the observer is going off and I don't know why.
@@akiraawooch I appreciate the thorough investigation and comment! It’s been so long since I made this video/code that I’d have to go back and review it before I could say anything about your findings, but they seem likely. If you want me to link your video in that pinned comment, let me know!
@@akiraawooch So this is something I found out shortly after posting this video. If you use F3 to see a sugar cane block’s age, it will never change until you update a block. So if you break the second layer, you can see the age change of the block below. From what I found, this is because that age on F3 is not updated on the client side as the developers didn’t find it necessary to always refresh, thus helping the game run a bit faster. I wasn’t aware that an observer could see the sugar cane updating until I saw OokamiKuro’s design. I think the compactness alone makes it my favorite max efficiency design so far
@@SeaJayPlays would appreciate if you add it, I didn't add it in the initial post cuz youtube tend to remove comments with links in them.
ua-cam.com/video/oDASLugeiOo/v-deo.htmlsi=LRhdio1Ll-BylHpZ
@@SeaJayPlays so I did the math, it's not any more compact that the normal "Design 1" observer- noteblock-piston design.
the primary issue with that particular design is the complexity of stacking the design while using a hopper minecart.
it still takes up a 12 block space or 3x4 space because you need an extra block there for rails to go on, and this is using an improved design I based off of ookami's design so that it allows for the block to be placed within a 3x4 footprint, and I still can't figure out how to stack it yet, so that might take a few days or weeks to figure out, I have one way of doing that right now, but I'm not really satisfied with how it's done.
stacking is a major problem with every instant harvest farm, my own design took a whole month to figure out how to stack it seamlessly. and ookami's design's compactness is it's own downfall as it's extremely hard to power the rails without accidentally powering something else.
As a fledgling programmer, I absolutely loved this. Subscribed
@@diminished4934 That’s a pretty big compliment, just don’t judge my programming too hard when you get good
I think this might be the most interesting and captivating Minecraft video I have ever watched. Mostly because I just love programming, but also because your ability to explain these things so clearly and in a comprehensive way, I liked this video. First video I have seen from you, but definitely earned a sub! Criminally underrated channel.
Edit: I went to read the comments, and I have never seen any creator reply to comments with such care. Long and thoughtful answers to interesting questions, and coming back a month later to notify someone that a video was uploaded about a topic they asked.
Wow, I really appreciate the compliment, that’s a very nice thing to say! I spent a while getting this video structured in a way I liked, so thank you.
I appreciate the comments and feedback, so I’m always happy to write back. I’ve got a few more farm programming videos and the chess video that you might like if you’re looking for some more like this!
very comprehensive I love these videos that study minecraft to such a detail
Thanks, I appreciate that!
As a financial graduade I'm sorry to break it to you, but we will need to calculate a tax on it, but, there is an option to postpone the payment for the next financial reporting period and recalculate a present value of the postponed tax so that we can max the efficiency with our crops!
😂
Even if I dont understand some of this, Ill watch the whole thing, it is so fascinating to me!
@@SSvtton Thanks! I’m glad it’s good enough to keep you interested!
I think a better description for your channel would be the minecraft scientist. this is what I want when i look up what the best farm to build is
Man, Great Video! Absolutely love these nerdy stats! Learned a lot.
Keep it up :D
Glad to hear it! Hopefully I’ll be able to post some more soon
I was hoping to see one more design. The burnout torch pistonified sugarcane farm. Cheaper than the observerful ones. But what a great video, thank you!
Thanks! I’ve never heard of that one, I’ll have to check it out! If there’s any other farms you’d like to see covered, let me know!
@@SeaJayPlays yeah slacklizard has a video about the burnout torch farm. Can be used for bamboo, sugarcane, melons, pumpkins with some modifications. Cheapest material needs
Interesting project. I love it when people start programming to solve rare and unique problems. Reminds me a lot of Gnembon.
But some critism is you need to focus on your code quality. The performance and maintainability of this piece is quite low. Use functions, classes and avoid those huge indentation needed for, for loops with if conditons with more conditons.
You should also have a look at Numpy, its quite a bit harder to learn. But is well suited to this task as you can avoid those slow loops and be faster than any other language. The code would also be simpler since youd just be doing formulas rather than pure logic.
Thanks!
I appreciate the advice, I’ll look into them for the next coding project!
Max hanke found a mathematical description of the statistically optimal harvest timing.
Cool video! I like the approach, though I must say I was confused on why using a code to get efficiency and rates if a simple tick warp test could do the same. I struggle to see what the code adds to the comparison basically. I'd expect that with python you can create a function of a general sugarcane farm and use an optimizer like Pymoo for a multi objective optimization of key design variables of farms, so you can get as an output the 3D pareto front of sugarcane farming. That would be cool!
Still, I enjoyed the video too
@@FB_engine Thanks! As for why I decided to program it, that’s a longer answer. A couple other commenters have asked similar questions, and I’ve written a pretty long reply as to why, so definitely check those out if you’d like. But short answer is, I tried in-game testing and tick warpping but my laptop just couldn’t handle it. Also, in the code I programmed each farm individually to account for harvest methods, and I never looked at the code as an optimizer, just something that allowed my computer to run “in-game” tests. I could look into optimization for a future video though!
Extremely underrated video :O
I would love to see an iron farm next :D
Little tip before you eventually blow up: I could see the Outro with the many different colors triggering seizures for epileptic people.
@@nextraven Thanks! I’m not sure I’m cut out to do an iron farm lol but I can look into them
Thanks for letting me know, when I get the chance, I’ll cut the last few seconds off
Looking at the small bits of code you presented, I really urge you to look into vectorizing your data generation and processing. Loops in python are incredibly slow, and if you instead of e.g. processing each block in the sub chunk separately were to do all calculation “at the same time”, whatever that means in a computer, you could probably decrease your simulation runtime by a factor of at least 20, possibly around 40.
Thanks for the tip, I’ll definitely look into that cause the slowness has been super annoying. I’ve just been writing these programs in these simple steps so it’s easier to do and come back to later. But these run for like 24 hours lol
You could start a new era of technical minecraft if you make more of these farms
@@kxdsh If you’ve got some farm ideas I should make next let me know!
@@SeaJayPlays I mean crop farms like sugarcane are definitely more simple to simulate as there's less stages and game mechanics than something like an iron farm or a gold farm but I'd reckon those are more useful. Making an iron farm at the start of the game is always such a hassle so if you could make something more compact than I would be forever grateful
@@kxdsh I’ll take a look into them! I’m not too sure I’d be able to find a better solution that what’s out there right now just cause the mechanics are so well known and nearly maxed out. But I could give it a shot
I love the structure of your videos. I’m very impressed
@@wardenunfolds8973 Thanks so much! This was the first one I actually wrote down how I wanted to structure it, so I’m glad it’s noticeable
This channel is a hidden gem
Glad you’re liking it!
Is there any change in rates if you just harvest the top most sugarcane and leave it at stage 2, making it more likely to receive a random tick because it is 2 blocks tall? This would allow for a shorter clock time or if you wanted to do an observer based one. Have the observer face pointing down so it activates when the sugarcane becomes 3 blocks tall. Subscribed, hope to see more videos like this, reminds me of Nico is Lost
@@MasterElements Thanks! His iron farm video inspired a lot of this video lol
This is an interesting question, and the short answer is basically no, there’s no change. I’ve got a longer explanation on the individual 100% efficiency farm video, but here’s the gist. Since every block (and random tick) is independent of each other, having more blocks below doesn’t increase the likelihood of getting a random tick. And in terms of efficiency, the three tall design approaches 100% efficiency because in the time it takes for the third layer to grow, you could have already harvested the second layer. But as time goes on, this small inefficiency basically vanishes. So the three tall vs two tall are pretty much the same.
@@SeaJayPlays Interesting, thank you for the answer.
commenting to boost the algorithm, as this is my fav nerdy minecraft channel that ive found since i found mattbatwings before he blew up, i rly hope u make it big and blow up too!
That’s a pretty big compliment, thank you! I appreciate the support
I really liked this well made video and I am glad everyone else felt the same and hope you grow even more!! Keep it up.
@@chaitanyaswaroop7669 I appreciate it!
Great job would love to see one of these with bamboo.
Thanks! You read my mind, once I finish my current project, that’s the next one I’ll work on!
You're doing us a service. Much appreciated!
Glad it could help!
This is something i didn't know I wanted to know about
@@Arae_1 That probably sums up most of my videos 😂
Interesting video, although I'm curious why you decided to abstract out the farming logic into python? I understand wanting to get longer time intervals, but my thought would have been to use carpetmod, which will make the game speed as fast as your cpu can deliver ticks (/tick warp). This is a highly effective tool for analyzing farms like this one with the ability to detect sugar cane drops, calculate rates, etc. I would also put into question how 1-to-1 your code is to the minecraft environment, since there might be confounding factors that deviate your results in python from experimental results.
I appreciate it! Funny enough you’re the second person to ask this question today, so I’ll just copy my reply longer below. But basically I started off by using carpet but my laptop just couldn’t handle running the game like that and real time tests just took forever. I don’t think my code is 100% accurate to the game but I did compare my in-game test to my program results and they matched up pretty well, at least well enough for me to feel comfortable making this video/claim.
My longer reply: This sugar cane video was the first of my coding videos I started working on, and I actually built all these farms and started off running in game tests. Programming it hadn’t even crossed my mind when I came up with the idea to test for efficiency. I have the carpet mod which has warp speed (basically running the game as fast as your computer can handle) and that’s what I planned on using. Unfortunately, I underestimated how crappy my laptop is, so it only sped the game up by like 3x and I had the feeling some of the drops weren’t being collected properly. My next idea was to literally just run the tests in real time. And I did. I ran two 12 hour tests (for two of the observer farms). I started with those two cause I didn’t have to worry about hopper clock timings. Then when I moved to the clock based farms, I struggled to get an accurate hopper clock for the intervals (I won’t bore you with how I then spend days figuring out how to use command blocks to make a simple clock 😆). When I did get the command block timer working I could only test one clock intervals at a time for the flying machine cause my computer really doesn’t like those.
Anyway, I didn’t think my in-game testing was going well, it was taking too long, and I wasn’t even sure 12 hour testing would give nice results. Enter, my thought to try to program it. Once it was done, I felt comfortable presenting it and saying “I programmed sugar cane farms” cause when I compared my in-game test data to the program results, they matched up pretty well. I’m sure in the long run, the game will have some glitchy stuff so my code should be taken as a perfectly smooth running game that is pretty close to the real thing. Also, I wrote a program for each farm type and did take into consideration the harvesting style. For the sake of video length I didn’t go into a ton of detail but I did program a “flying machine” that stops my sugar cane from updating when it’s moving over it for example. Granted, I’m sure it’s not prefect, but I’d say overall it’s a decent estimate. Lastly, I assumed 100% lossless drop collection just to compare the best case scenario across the board. I think the hopper minecarts below to collect is the best way and as long as the cart covers all its path before the drops despawn it should be pretty close to lossless.
Crazy that this came up in my recommendations. Definitely gonna check out the rest of your content! This is the type of content I love
@@visxosity8690 I appreciate it, thanks!
@@SeaJayPlays of course, kinda already went thru and watched a good 90% of your content already
Thanks for helping me get more watch time lol
If you’ve got any video suggestions, I’d love to hear them
@@SeaJayPlays I’d low-key be interested in your take on a storage system
@@visxosity8690 That seems like a pretty intimidating thing to get into lol
If you’ve got a specific problem you’d like solved I might be able to look into that, but doing a full scale system might be pretty far in my future
i had an idea to do this exact thing but with bamboo a while back, this is so cool ^w^
@@morsznetik Thanks! I recently did a bamboo one too if you want to check it out!
I love channels like these, great vid
Thank you!
I think the max eff farm is locational because of QC + dust update order. If you replace the solid blocks with glass I think it shouldn't break anymore.
@@tunafllsh I’ll give that a shot, thanks!
12:11 im Like: „Observers are 8 Years in the game?!
@@serious4gaming696 I can’t believe 2016 was 8 years ago
I once made a square kelp farm that deserves its own video.
@@ValidatingUsername You should make one then! I’d like to see that
@@SeaJayPlays I don’t have access to it as I am unhoused but it was a scalable variation on a wall farm with one height growing spots and observers that fired the edge of the square as long as the red stone signal would go
@@ValidatingUsername sounds like a pretty cool farm!
Great Insights !!! I loved your content Bro... Binged whole channel XD
Wow thank you! I appreciate that
And I thought this video was about a program randomly generating different farm designes with an ai and then upgrading them with each generation to find the best possible one
@@yodxxx1 That could be an interesting idea, comparing AI generated farms to human-designed ones
my two favorite things: python and minecraft. would love to see the repo for the code!
@@daniellepashayan773 Thanks, glad you liked it!
Someone mentioned using GitHub which I haven’t done before so I’ll have to check it out lol
random numbers by computers are not entirely random. So to make the simulation better, you should use the same random formula minecraft uses to generate the coordinates of random ticks.
@@Capiosus If I knew how Minecraft did it I would definitely add that to my code. Unfortunately I don’t know how to find that, so we’ll just have to assume the rand function is close.
Ilmango did a video on clock timing for sugar cane and ran some in-game tests that showed the best interval was 41 minutes. That matches all of my data, so I’d say my programs are a good estimate.
If you know where I can find more details on the Minecraft code, let me know!
@@SeaJayPlays minecraft java edition has it's mappings available for anyone, so all you need is a copy of the game client's versions' .jar file and the json which lists the URL of the client mapping. Then you can read minecraft's code using a java decompiler tool.
Great video i am going to binge watch all your others and comment for the algorithm
@@JimBob-vu2fe Thanks! I appreciate you boosting my watch time lol
By the way, the randomness of redstone should hopefully be fixed in the latest snapshot if you turn on the experimental redstone feature, since thats basically what they changed - before that there are some undefined states in which basically some stuff like hashes in the backgrounds or chunks or stuff like that determine the outcome instead of a true random number generator, creating location based behaviour sometimes.
Thanks for letting me know! I’ll make give it a test
bro decided to put his computer science degree to good use
@@Luzivog 😂 I think the computer science majors would cringe at my code
@@SeaJayPlays it's probably not the cleanest or the most optimized but who cares, it does what you want it to do which is the most important
@@SeaJayPlays keep up game simulations through coding if you like it, there's definitely an audience for that
Thanks for the support!
12:21 I always thought this was obvious. I have been doing it like this for a very long time now
I guess I’m just way behind the curve lol
In your "100% efficient" farm, the piston still blocks the growth for a few ticks. For the truly pointless true 100% you would have to use TNT or something else to harvest without blocking.
Yeah I kinda ignore that lol
I guess we can call it 100% efficiency for piston farms. But I’ll look into other breaking methods for future builds, thanks!
SeaJay Plays? more like SeaJay codes!
I learned a lot from this video, thanks!
@@nikilpaul6366 lol I don’t think I’m good enough to change my name just yet
Glad to hear it!
Great video! Thanks
@@billyfaffler I appreciate that!
I dont watch even minecraft youtube much less farm related videos and this video was right above another one called "How I made the WORST sugarcane farm in minecraft" 😭
@@tunablockbros Thanks for checking this one out!
the sugar cane farm by the chronos SMP is currently the best public sugar cane farm
@@naecker8974 I’m not familiar with that smp, how does the farm work?
The Chrocane farm lol
I really enjoyed this video!!!!
Crazy good video! It was rly entertaining!
@@tiger0043 I think you’re the first to say it was entertaining, so I appreciate it!
I now finally know what a random tick is
Glad I could help!
this bro will have at least 500k in max 2 month
Maybe if I upload some more videos lol
are random ticks drawn with replacement? If not then the actual probability of each would be slightly different
@@token1652 From what I read on the wiki, they are, but if you find something different let me know so I can update the next video’s code
I've managed to compact the "full efficient" design down to a 5x3 form factor that is stackable.
observer facing the sugarcane will 0 tick budpower a sticky piston, which pushes an observer facing the opposite direction upwards, it then will hard power a full block, which the piston next to it will push down, breaking the sugarcane.
that will in turn be observed by the observer, triggering the sticky piston to pull down the observer, thereby resetting the design.
the only issue here is that sideway stacking of multiple modules may have some issues where it creates a observer clock. solution is to stagger it upward 1 block, or leave an airgap. I'll post a discord image link in the reply, cuz youtube will nuke this comment if I add it here.
@akiraawoock That sounds pretty interesting, I’d love to check it out. Another person in the comments recently sent me their tileable design and it seems to work pretty well (also uses pistons pushing observers) So it’ll be cool to compare them.
I’ll post a link to his design once I’ve got one. Btw I don’t have discord so I’m not sure I’ll be able to see your image. Sending through my planet Minecraft account would work if you’d rather do that.
@@SeaJayPlays I'll make a video on it when i get my mic, I'll let you know when I get it out
Trying to get 1,000,000 in Beta 1.7.3*, 720K rn. Ty for information
Glad it could help! Good luck
I mean this in the most affectionate way possible, truly no malice nor hatred runs through me when I say this
What the fuck is wrong with you?
lol
so, you count how many blocks of sugar cane you can fit on a subchunk and just run the math (is it a sugar cane - is it 3 blocks already - is the random tick hitting it) on that number to grow? that's what people do for 10 years+
If that’s all you got from this, then sure!
wsg markiplier
POV you majored in Minecraft with a masters degree
@@graemepicataggio9814 😂
Is this for the Java version? If so then for the "100% efficiency" category it's much more efficient to use budded pistons. It's dramatically cheaper than any of the designs even the ones you have pinned here. A budded piston design can also be very tile-able btw.
Something you didn't explain that I think is very important is that when a sugarcane reaches the final growth state (15) it doesn't reset when the block above it is harvested. This is very important for one-piston-per-sugarcane designs.
@@pand1024 Yeah this is in Java. I don’t have much experience with bud powered pistons and I’m not sure how I’d implement them here. If you’ve got a design idea I’d love to see it!
So I was under the impression the sugar cane did reset when the block above gets harvested. Can you explain what happens and where you were able to find that info, it’d be really helpful for future videos. Thanks!
@@SeaJayPlays Let me see if I have time to to put together a video showing the design I have in mind. Regarding the block states thing looks like the wiki directly contradicts what I'm saying so maybe I'm wrong. "A freshly planted cane - and a cane that has just grown cane above it - each have an age of 0". I did most of my original work back in 1.8 (before observers existed)
@@pand1024 sounds good!
That’s where I get pretty much all of my info, but I’m not sure where they get it. So who knows
For some reason it doesn't seam to want to let me reply here. I've made a video reply with the design you wanted to see.
@@pand1024 Great video, I just watched it. I had no idea this bud powering was possible. I only started getting into redstone about a year ago so I’ve got more to learn. I agree it’s not 100% efficient but it definitely scores higher for me than most farms due to resource efficiency. I’ll add the link to your video in the pinned comment, thanks!
You are insane sir. I like it
@@NilAtabey I appreciate it, I think…
I'm dumb but does the size of the flying machine farm impact the time we need to set the clock ?
That’s a good question, cause there are some size requirements. But most likely not.
In terms of clock timing, you’ve just gotta make sure the machine can traverse the whole field before the time is up, so usually this won’t be a problem. If you’re wanting to use a faster clock interval, you can build it so the machine goes one direction then stops until the next clock interval triggers it to come back. Also, if you’re going for max resource and drops efficiency, I’d keep the sugar cane area thin enough to fit in one chunk and then only go as far as you can load. I think it’s 16 chunks long, and you’d afk in the middle.
The science is great but I can’t help but wonder… why do you need so much sugarcane 😂
@@apdj94 Here in my base, just bought this new horse here. It's fun to ride up here in the grasslands biome, but you know what I like a lot more than materialistic things… knowledge. In fact, I'm a lot more proud of these 7 thousand new bookshelves that I had to get installed to hold 2 million new books that I crafted (thanks to the sugar cane). Cause it's like the billionaire Warren Buffett, says the more you learn the more you earn.
So… sugarcane to make paper to make books to hold all your knowledge?? I suppose that’s fair haha
@@apdj94 nah I don’t have any good reason for these farms lol
I was just hoping you’d get that reference
No I’m sorry 😅 I was genuinely thinking you were making bookcases 😂
lol
First you forgot cactus in Redmond ticks and second you can place then diagonally when a water is nearby. Could you Test if my setup is any better than the average player including yous pls? I place my sugar cane in a Chess like pattern where sand and Water interchange where the waterlocked blocks are the the black titles and the sand are the white titles then I place the Sugar cane and run flying machines over it . I also stack the fields on top of another to maximize production. Could you also simulate cactus farm pls?
@@icebluscorpion Hey there, not sure what that first part is. But for the second question, surprisingly, the chess board pattern is less efficient (50% sugar cane) than this strange pattern (~80% sugar cane) but this pattern is different to make.
It’s definitely a good idea to stack the farms so you can max out efficiency based on the chuck borders, nice job!
Lastly, I probably won’t do any cactus farm videos just cause we already know the best way to farm them and can easily do instant farming (max efficiency). The hard part is designing how to capture all those drops. I think ilmango has a good video where he tests some common designs if you want to check that out!
Regarding calculating the rate, wouldn't it make more sense to calculate based on the tick count? That way inefficiency in the code is a non-factor.
I think I did do it based on the ticks, but I’d have to double check to make sure. What kind of inefficiency in the code are you talking about? I’d like to correct that for the next video if it’s an issue.
@@SeaJayPlays Assume that the code runs slower due to some reason (more things happening for example) and potentially misses a tick timing, running it based on real-time would potentially have fewer cycles over the same time period.
Systems are also imprecise, so even if you say something is going to run after exactly n milliseconds you still might have some minor discrepancies.
I don't think it'd change the results in a huge way, but just one of those things one can do in order to keep things accurate. Just letting it run x amount of ticks is kind of all one needs to do.
@@CottidaeSEA oh okay i understand. I’ve got it running in a while loop, and each loop is a single game tick so I think we’re on the same page there
Is there a way to make it more compact? so it can be mirrored as well as you can stack it in rows?
@@astronomy2_0 I wasn’t able to get it, but some other people have commented with their 100% efficiency, tileable designs. I’ve posted some links to their designs in a pinned comment if you want to check those out!
awesome vid, keep it up!!!
Thanks so much!
these videos are awesome! keep up the good work :)
Thanks so much, glad you like them!
Imagine you spent your time on something more lasting
@@legobuildingsrewiew7538 I think this video will outlast me at least
I wish you luck with UA-cam logorithms 🤞
@@raimundascekavicius2030 I appreciate it!
10:55 in wonder how my Setup is going....one Observer for alle the 8 SC..i guess worse then your Test result :P
😂
Can you make a video abt a mine cart unloader and shulker box unloader
@Everesl, Hi, I could look into minecart unloaders, but to be honest, they can get pretty weird and use some strange mechanics that I'm not familiar with. In some of my farms, I actually just use Ilmango's instant unloader. It's not too difficult to build and is tileable. But, if you want an unloader for a different situation that his isn't good for, let me know what you are looking for, and I could definitely try to make something.
I've never looked into shulker box mechanics or unloaders so that could be fun to try! I'm working on a video idea right now, but once I'm done with it I'll look into these. Thanks for the request!
@Everesl Hey, I just uploaded a video going over two shulker box unloader designs. There’s a world download for them as well. Let me know if those help.
This is amazing
Thanks!
Can you check the efficiency of TangoTek's design?
@@OokamiKuro39Ch Hi, so if you’re talking about his compact, lossless version, then that would match the blue farm (group-based observer-activated farm) since his relies on one observer to activate the whole line. I forget the rates, but I think I have a graph showing the efficiency for different group sizes so you can check out which matches his most closely. If he’s got a different farm let me know
@@SeaJayPlays hmm... I do have a design that is one wide tileable with the observer looking at the bottom cane, uses 2 observer, 1 piston and 1 sticky per cane. That would be 100% efficient?
Sounds like it! As long as the piston harvests at the second layer as soon as it grows, it should be 100%
Edit: If it’s something you’ve got posted somewhere and want shared, I’ve linked some other people’s compact, 100% efficiency farms in a pinned comment
@@SeaJayPlays I posted the design in my channel. Not sure if I can put the link here, but its the latest video I uploaded.
@@OokamiKuro39Ch I added it to the pinned comment! This is definitely the most compact one I’ve seen so far, nice job!
theoretical minecraft
You should try this in Java for even more accuracy
I the smaller tests I did in-game seem to support the results, but large scale testing didn’t work out too well for me. But I did see an Ilmango video where he ran large farms and got similar results
Is there a way or chance you can do sims on bedrock. I know it’s different for some but I mostly play bedrock and would love to see those numbers.
The red stone randomness might have something to do with chunk borders. You probably already thought of that but just thought I’d mention it.
@@OMEGAMAN1300 I know redstone is pretty different on Bedrock, but I can definitely look into programming the farms differently to account for that. For the next one I do, I’ll definitely add it!
@@OMEGAMAN1300 Unfortunately I did haha but good suggestion still.
Some other people have shared their more reliable 100% efficiency farms (pinned comment) that don’t have this issue so if you’re looking to use one, I’d start there!
idk if u covered this in the video, maybe i missed it but, can the same block get all the (3) random ticks? yknow wat i mean???
Yep I know what you mean! And yes the same block can receive all three random ticks in a single game tick. It’s just a 3/4096 * 3/4096 * 3/4096 = 0.000000039290171% chance lol
@@SeaJayPlays so basically impossible
Still about 70 times more likely than that Dream speedrun, so who knows…
@@SeaJayPlays LMAOOO
i think lag should me metric too, does anybody know if brown note block design or flying design is less laggy
@@lurkkilukki7877 That’s a good point! For me, they all have the ability to crash my computer so it’s hard to tell. I’d assume the noteblock farm would be less laggy but more steady and the flying farm would be large spikes due to all the drops but no lag for most of the time.
what about the effectiveness of the slime pusher from ilmango?
You’ll have to remind me what that design is. But if it still uses a flying machine the results should match what I found in this video since I’ve assumed every farm tested is 100% lossless.
@@SeaJayPlays i think its more like piston farm with timer.
Okay I’ll give it a look and let you know
@@wrtltable So after searching through his videos I couldn’t seem to find the video you’re referencing, all the ones I found had some sort of flying machine.
But for efficiency there’s really only two categories: clock-based harvesting vs. instant/near-instant harvesting. So as long as the farm relies on a clock (flying machine, clock-based piston) then it’ll have that 70-80% max efficiency. Then if you have a near instant harvesting one (relies on individual operation, not like the group based observer farm) then you can approach that 100% efficiency.
@@SeaJayPlays it's called - [Tutorial] SciCraft Sugar Cane Farm
Shit, mc has keep so many kids off the street and now they should be grown, pursuing higher education/starting a family. I wouldn’t be surprised how many kids got into numbers because of mc. Ik I did. I remember watching direwolf20 make solar flowers and emc flower farms and wondering, what’s the best combination to get the number higher!
I’m glad I’m not the only one!
13:44 I may have done that, I compacted it a bit, is there a way to show it to you?
That would be really interesting to see. If you have a Planet Minecraft account you could send it to me there. I’d love to see it
@@SeaJayPlays sent :D
Imma be real chief: python is not a programming language, it’s a toy
😂
Neat.
Completly non practical but I wonder if in theory because you are able to simulate the data you could write one of those evolving AI models to generate a farm scored on the output, I mean it wouldent make a farm better then a human (because we already have near perfect farms) but it would be interesting to see what it attempts
@@urubonz8831 I’ve never got into AI but that’s something I could definitely take a look at for a future video, thanks!
why did u simulate all of them instead of just... playing and letting the farms work on the background?
@@that_guy1211 lol trust me, that was my intention. A couple other people have asked now, and I posted a pretty long reply if you want to check it out. But basically my computer couldn’t handle it.
I did do in-game tests to validate my programs, so I know they’re decently accurate. I’m sure my code’s not the best but it turned out to be pretty enjoyable to work on.
@@SeaJayPlays oooooh! Good to hear, unfortunate for your PC tho, sad to hear that, but nice work mah dude! Programming is one of the things i least like doing in my free time, but can't help but be interested in doing it
Thanks!
I’ve got a love hate relationship with it
Verry cool
bro make a similar vedio on iron farm
That’s a bit harder so I’ve avoided it, but a few people have requested it now, so I guess I’ll have to give it a try lol