TowstiUsing spinning happy ghasts to create sensors including a "hidden" pressure plate and a snowball detector.
When a ghast is leashed to a boat, mob, or fence post outside its normal leash range, it starts spinning. Mojang clearly added this feature to make fans but it's also a neat way to align ghasts automatically which is useful to create sensors.
*Ghast Pressure Plate* A stretched lead makes the ghast move back and forth, triggering tripwire or a pressure plate. Ghasts will still face nearby players, which stops the motion. To make the ghast reliably go back to the back and forth motion, I did need to use a piston to push it back in to position.
*Other Alignment Methods* Since the sensors mostly rely on the alignment from the lead, rather than the spinning motion, I did look in to other methods. Downwards bubble columns almost work for the pressure plate, but they're not reliable and they have a very small detection range. youtu.be/EHDWqzIDg4M?si=mMCD7sH7vLuhGy5D
Using Happy Ghasts To Make Fans And SensorsTowsti2025-09-09 | Using spinning happy ghasts to create sensors including a "hidden" pressure plate and a snowball detector.
When a ghast is leashed to a boat, mob, or fence post outside its normal leash range, it starts spinning. Mojang clearly added this feature to make fans but it's also a neat way to align ghasts automatically which is useful to create sensors.
*Ghast Pressure Plate* A stretched lead makes the ghast move back and forth, triggering tripwire or a pressure plate. Ghasts will still face nearby players, which stops the motion. To make the ghast reliably go back to the back and forth motion, I did need to use a piston to push it back in to position.
*Other Alignment Methods* Since the sensors mostly rely on the alignment from the lead, rather than the spinning motion, I did look in to other methods. Downwards bubble columns almost work for the pressure plate, but they're not reliable and they have a very small detection range. youtu.be/EHDWqzIDg4M?si=mMCD7sH7vLuhGy5D
*Song* Joe Hisaishi - Merry-Go-Round of LifeI Made Flappy Bird In Minecraft Using a Parrot And Armor StandsTowsti2025-09-28 | Flappy Bird in Minecraft using water to move the pipes and the bird in order to create suprisingly smooth gameplay.
*Controls* The controls work as followed: - Press S (backwards) to make the parrot "fly" up - Release S to make the bird fall back down
This works by bumping in to a boat which automatically pushes you back when you release the key.
I chose this over jump controls because I found it's much more responsive. The duration of a jump moves the bird too little and extending the pulse ends up being too janky.
*Bird Movement* The parrot is leashed to a boat that rides on a bidirectional water stream. A neat (for redstone, not for the parrot) quirk of stretched leads is that the parrot almost fully loses its own movement. So instead, it's fully controlled by the boat.
*Collision Detection* - A rabbit, attached to the same boat as the parrot, handles collision. - When a pipe and the parrot align, fences extend with a hole matching the pipe gap.
Using parrots + glass panes gives a nearly perfect hitbox alignment. Unfortunately, parrots can still move slightly, even with stretched leads, which I found makes the game way too hard. So the next best combination was to use rabbits and fences. This does leave a slightly smaller gap but it's more consistent.
I did try to find a better solution. Ideally something where you see a actual collision with the pipes. But after a week of failed attempts I kinda want to move on to the next project :)
I’ve wanted to make a Pong game for a while, but with smooth animations instead of the jerky, pixel-by-pixel movement from lamp or piston-based displays. After lots of experimenting, testing, and banging my head against the wall trying to fit the paddle, I finally got something to work!
Some clarifications from the showcase: • The realigners are there to: ◦ Prevent the boat from falling between water blocks ◦ Stop the boat when the game ends ◦ (Mostly) prevent the boat from dipping too low and making the ball shortly vanish beneath the snow • The boats on top of the controls are carefully aligned to bump you back to the center resulting in much more responsive controls. • To go back to the starting platform, simply enter one of the boats.
*Improvements* There’s still plenty of room for improvement, so I may create a new version in the future. The next design will likely use a smaller screen size, since the slow ball speed won’t be as noticable.
After experimenting with the display layout, I found it can be much more compact by placing the dispensers beside the mangrove roots instead of beneath them. This could make the pixels be as small as 2×4 blocks using sliders to power dispensers or 2×6 blocks using rails.
I’d also like to incorporate the improved paddle design shown at the end of the video, which I couldn't get to fit into the first version.
One drawback of a smaller display is that the ball’s bounce pattern becomes very predictable. A possible fix would be to add a mechanic where the paddle adds a curve to the ball when it hits it whilst moving.
*Song* BADBADNOTGOOD - Sétima Regra (Live at Valentine Recording Studios) youtu.be/Yu3b69g0LEQ?si=OtAKIcDiUI_xxcjTMulti-Color Painting Display Using Wave MachinesTowsti2025-08-10 | After struggling to make a wave machine that wasn’t boring, @RUOK0214 came to the rescue with a super satisfying sculk-based wave machine youtu.be/Bv0vHAbXnTE?si=-0G0LKh8xJPzGGPM. I was originally trying to use a similar concept to launch sand upward as you launch yourself up with a wind charge. But the latency from the wind charge hitting the floor to the sand shooting up kind of ruined the effect. So, the next idea was to turn it into some sort of display. Because that’s definitely something I’ve never done before!
This machine was a lot of fun to make, especially with how elegant the reset mechanism turned out. The only minor drawback is that the binary counter toggles the amethyst block each time the color cycles to white. That’s why the reset begins by extending all amethyst blocks; otherwise, some pixels wouldn’t reset. As far as the reset lag, this can be reduced by surrounding the bottom sculk sensors with wool. This allows for different reset patterns, such as making the signal snake its way across the screen instead of propagating outward in a circle from the center, updating fewer pixels at the same time.
These remind me a bit of E-Paper displays in the way they swap colours. Except instead of using charged pigments, they use concrete powder. And instead of applying an electric field to swap colours, they use pistons and boats. Maybe it's not like E-Paper at all...
*How it Works* Gravity blocks fall instantly when there's a boat one block below. This combined with the following timing can be used to create a block swapper: 1. power piston 2. wait 9gt (boat) or 8gt (boat in minecart) 3. power piston
*1x1 Pixels* Toggling 2×2 or 3×3 pixels is fairly straightforward since the boat's hitbox fits entirely within the pixel area.
1×1 pixels are a bit trickier as the neighboring pixels block the boat from moving upwards. To work around this, you need to move all blocks within the boat's hitbox, then power the target pixel(s) again 9gt later. After this, the boat will rest on the lower gravity block, so all pistons must be powered once more to allow the boat to fall down and rest on the pistons.
*Music* Slottskogen Disc Golf Club - Wintergatan1x1 Pixel Multi-Layer Block Swap DisplayTowsti2025-06-10 | I found (I think?) a neat way to swap falling blocks using slime and powdered snow.
*Update* Not actually 1.67 fps (12gt clock). The screen in this video only updates once every 2 clock cycles. Ie 24gt/0.83 fps. After retesting, the actual max clock speed is 17gt (1.176 fps). Which would update faster than shown in the video.
The concept is similar to that by @fakestory1753 youtu.be/rUC6W-GJv2E?si=YGl-wijYiKjEFnJE, but using powdered snow instead of boats and fence gates. The advantage is that it doesn't require any blocks adjacent to the swapper. This means the swappers can be tiled more easily, allowing the entire display to update at once instead of updating one half every cycle.
This display configuration does come with some drawbacks: - The segment heights differ by 1 block so it can only swap up to 8 layers and needs 1 cycle that only swaps the 9 high column so they don't go out of sync. - Swapping the entire image at once is extremely laggy. I was dropping well below 1 fps in-game when running this on a clock lol.
*Info* - Size: 32x32 (expandable in both directions) - Layers: up to 8 (some will break - Blocks: sand, gravel, 16 concrete powder, and anvils - Lag: Very bad
First off all, huge credit to @squibble111 for pioneering the mirror pylon (view angle sensor) tech. Be sure to check out her video explaining how it works: youtu.be/X7Ah-SJ0vBc?si=LwEslQbgSXziy3y5
My goal with this project was to experiment with methods to reduce the latency of breeze based player angle sensors as well as connect it to a map display.
*Display Driver* To drive the display, I'm using four view angle sensors all connected to the same output. The lowest sensor triggers the others at 8gt offsets. The others all buffer the wind charge. This creates a ~8gt sampling loop. The output is then stabilized using a pulse extender. Idk how but I had a lot of trouble with this due to messing about with input selectors making it so only one tile is active at a time. The issue is that the input can switch between the old/new state quickly before settling, causing a unstable result. Trying to stabilize the input before feeding it into the selector helps, but it also increases latency. In the end, I just removed the input selector and added the 4th sensor so I could use a dropper/hopper pulse extender. This does come at the cost of some extra latency.
*Display* I'm using a map display, which is nice because it can render virtually any kind of UI within a single block. The display updates instantly thanks to water logged trapdoors, but it is prone to breaking because I didn't add an input cooldown.
This screen needs two players to load the map: one holding a single map, and the other holding two. This creates an "unstable" map. Combined with a strober that updates the colour of each column every game tick, the cursor updates all at once. For more info, check out: youtu.be/jWEckuyj7mc?si=Sr5AFS9TgrlX5HIq
I also played around a bit with holding the map in your offhand instead of mounting it to a wall. A bit like a tablet. But I found that it is pretty unintuitive to aim. Especially as the viewing angle is really small.
*Improvements* Overall, I’m pretty happy with the results, but there’s still a lot of room for improvement in a potential next version. For this prototype, I largely reused Squibbles mirror pylon design. This worked well for an initial prototype, but simply stacking more sensors to reduce the sampling interval isn’t the most elegant solution. So it would make more sense to design a sensor optimized to drive a display.
I’d also like to increase the touch resolution. Possibly by using multiple sensors, each covering different sections of the screen, as shown here: youtu.be/3x56NdAox-g?si=dt3jBDLyvxNo0tHA
*About* I'm quite glad with how this turned out. I love how the layout looks with the upside-down house. And It's relatively compact and fast, though there’s definitely still a lot of room for improvement.
One issue with the current design is that the platform responsible for placing and removing the door layers gets stuck waiting at the bottom for long periods. It also has to move up 16 blocks just to reach floor level. I experimented with a bunch of ideas, like adding a buffer between the storage and the main platform. The buffer would be closer to floor level and would always be ready to add or receive a layer. As well as different ways to store the layers. Unfortunately, most of these ideas ran into dead ends, mainly because of the size and complexity they introduced.
All in all, I think there's a ton of potential for 3D hipster doors using flying machines. If nothing else, I hope this can inspire others! :)
*Specs* Volume (xyz): 39 x 40 x 58 Closing: 20:51 Opening: 21:06
*Song* Rone, Dirk Brossé, Orchestre National de Lyon - Room with a View (Looping)Minecraft Playing Card DisplayTowsti2024-03-03 | Red/Black/White 1x1 pixel playing card display.
This display uses powdered snow to hide the beacon beams.
To challenge myself with the wiring, I used segmented displays. This made the display more compact but also made the wiring look like a bowl of spaghetti. I kinda forgot to remove the redundant observers for the number display. So just pretend they're not there.
There will be one more video on beacon displays after this. Then it's back to my big project :o
*Synced Numbers* Syncing beacons is a bit of a mystery to me. I've had no success adjusting for the difference in piston extend/retract timing or using 0 ticks. Rejoining the world sometimes unsyncs the display and moving further away can cause it to sync back up. If anyone knows more, please let me know.
*Smaller Display* It takes 5 white glass blocks to fully cancel out a red beam. For the 0, you also need 5 coloured glass blocks to change the white beam to a coloured beam. This is why the display needs to be this big. That being said, it can be a bit smaller by using 3-4 blocks at the cost of contrast. Especially 4 blocks is fine as the difference with 5 is hardly noticable.
*Fixed Sequence* I didn't want to increase the empty space between segments by adding multiple colours. This meant that I couldn't, for example, add red and green glass below the top left segment. Furthermore, [0, 7], [4, 5, 9], and 2 needed to be of the same colour. I also wanted a repeating sequence. This is why I ended up on this specific sequence.
*Full/Base Colour Beacon Spectrum* It's possible to support a wider range of colours without increasing the size between segments. But it will make the display much more complex. It'l likely take a feed tape to cycle between colours or a flying machine approach.Connect Four With No Visible Wiring (Auto Reset + Spam Proof)Towsti2024-02-12 | The first design felt incomplete so I made a better one!
This version has spam protection and the inputs are now on one side. I'm pleased with the result and this has been quite the challenge to wire. So I think I'll leave it at this. That being said, there is definitely room for improvement. Mainly in the reset speed. Which, thanks to my piston door brain attempting to keep this somewhat compact, ended up being incredibly slow.
Note: I was struggling a lot with the spam proofing. It should be reliable. But if you really try to break it, you might succeed in doing so. Just to be sure, in the places where it's most likely to break there are signs with instructions on how to fix it. The reset is not spam proof. You can play whilst the reset is still going on but I recommend to wait for it to finish.Self Resetting Connect Four with no Visible WiringTowsti2024-01-29 | Improved version: youtu.be/dUiO6tZ5sJs?si=inz9-fPhG8PK2SaL
Man obsessed with hipster doors attempts to wire a connect 4 minigame.
The goal was to find out if a hipster design is playable at all. I therefore focussed on reducing the time between pressing the note block and the block landing. I didn’t focus on making it compact and the reset can be much faster. Maybe, I'll make another version at some point.
*Hipster Layout* The general idea with the layout is that you shoot concrete powder up in the air whilst the already in place column moves down turning the blocks in to falling entities. The concrete powder that is shot up can then move through the column. I need to drop the columns 2 blocks to leave enough time for the concrete powder to move through all the blocks.
* Up/Down Controllable Piston Worm* Moves up/down using instant wire update order. I planned to use this for a concrete powder version of youtu.be/973Rq_p8yQY?si=NXy_HPtV1bkmTN0w.
*Smallest Infinitely Expandable Flush Trapdoor (Proof of Concept)* 4-Wide PoC for what can be the smallest infinitely expandable flush trapdoor. Although I'm pretty happy with this design and it's cool to see how much flush trapdoor layouts have changed, it's just too slow. I'm not really interested in compact redstone at the cost of speed/lag so I probably won't make a version that can be infinitely expandable.
Still the record I believe but Mathrilord has tied it with a sandless version as well as one that can be expanded to any size.
*1x1 Pixel Display + ROM* This was more of a wiring challenge than a attempt at making something useful. That said this display was a lot of fun to make. For a much better (addressable) version, check out: youtube.com/watch?v=m7mB8UEOW1Y For a ROM version using copper bulbs: youtu.be/CCSfZx0tLGU?si=ce6UhLCM7D8EzTk0&t=199
Maizuma 1x1 pixel display: youtu.be/ufk0llZif_E?si=dWR5pzCoz3IXe8f75x7 Digit Counter Display Using Concrete Powder [Hipster]Towsti2024-01-06 | High contrast 5x7 hipster digit counter display using concrete powder. The digits use the Minecraft font and are tileable with 1 block in between.
*Sculk Powered Counter* AB tileable counter that uses the following vibrations: - block placement: A - block activate: B - book change (chiseled bookshelf): AB
The bottom segment toggles whenever any of the top 3 segments toggles. The feed tape therefore only needs to power 6 segments.
Needless to say, this display is horrible. It breaks whenever you place or activate a block nearby.
*Flying Machine Powered Counter* Expandable counter. All displays are updated in sync to avoid a really slow ripple effect.
This thing was suprisingly difficult to wire due to the tileable trapdoors and flying machine control.
Volume (xyz): 61 x 22 x 31 (23,870) [-17,732] Opening: 2:04 [-1:36] Closing: 2:21 [-1:50]
*Improvements* I've always wanted to give this another go ever since finishing the first version: youtu.be/DJjrKNikNvk?si=mKVjmTzsoFxtEZDE. The concept remained the same but I've redesigned the layout from scratch.
The biggest improvements are: - faster vertical platform - faster solid layer loading (piston slider sequence) - faster layer unloading - vertical platform is closer to the floor - reduced piston solid layer loading/unloading distance
A new challenge is that the storage has to be faster as well in order keep up with faster (un)loading of layers. This mess of a storage I ended up on was the best compromise between speed and size I could find.
Whilst building this I've already found some more room for improvement. But I think I'll leave it at this for now :) Overal I'm still very happy with the end result and this might be my favorite door yet!
The copper bulbs next to the sculk sensors filter out the sculk sensor pulse so the lights toggle once instead of 3 times.
I also made a vertical version imgur.com/a/5fwHmBG but for this I haven't found a way to get rid of the triple toggle.Smallest Infinitely Expandable Flush Trapdoor (Dentless)Towsti2021-04-09 | Unedited showcase of a 8 * 9 version: youtu.be/ipTMn9mO9g0. Smallest from 08-04-2021.
This was a pain to build... After what feels like trying every possible combination of flying machine layouts, activation circuits, and observer placements, I finally got it to work. Credit to Mathrilord who came up with an activation circuit module (http://prntscr.com/1185ek1) that ended up being really useful. He was also the first to create a stacked layout (youtu.be/jMq_CyZuCwI). Also thanks to Fufu (youtube.com/channel/UC4IakLM5GoTNoKjPs23lysw) for helping with the bottom activation flying machine control circuit. I hope the showcase is clear enough as I had a lot of trouble recording parts of the opening. The unedited showcase shows a smaller version with the real opening/closing speed.
**Contained volume** Volume (xyz) = (x + 2) * 9 * (2 * z + 7) x mod 4 = 0, x min = 8 (maybe 4 but haven't confirmed) z mod 3 = 0, z min = 9 y is arguably 8 but the floor is required to reach push limit so it counts in the volume
I also made a 8x8 version a few weeks ago (https://prnt.sc/wecigi). I haven't updated the 8x8 design to use droppers/lamps to trigger the neighbour lamps but it would have still been to slow because of the torch/repeater spam needed to make it flush + seamless.
I might try to make this more compact by using droppers directly underneath the sculk sensors (they bud the neighbour droppers).8-Wide Horizontal Button/Touch Panel (Tileable + Fast Reset)Towsti2020-12-08 | Very close to actually making something useful.
I did not bother compacting this since since I want to get the video out, but I think the height can be 3 blocks lower atleast.4-High Expandable Backlit Button PanelTowsti2020-12-04 | My first attempt at a backlit button panel using the new sculk sensor. I don't think this is the best approach (check description) but I'd thought I'd share it since it's still a pretty cool concept.
Vertical detection: I might make another panel that is based on vertical detection: http://prntscr.com/vw0c2o This should allow for much larger panels but it is also more difficult/slower to filter the sculk sensor output since they activate in a half circle pattern from the button.Minecraft Floating Character Word Processor (Hipster/Seamless)Towsti2020-09-18 | The wiring is rushed but this is mainly a proof of concept and a preview for things to come...
The 3x5 character display is based on the 1x1 pixel display by Maizuma. Every character has its own ROM. A unary counter selects the character slot and SR latches (piston based) are used to set and clear the character displays. The flying machine layout is based on a 15x10 air/solid hipster door that I made before. The main difference is that it uses a single module instead of two modules to carry the blocks (characters). To achieve this while still making the characters float I had to extend the module above the ground level.
World download: drive.google.com/file/d/1ubcPJ6AILldFdIuu-sbA12w1EuhO0eJC/view?usp=sharingMinecraft 1x1 Lamp Display With ROMTowsti2020-09-05 | ROM: The ROM is based on a shifter where the second piston shifts after 3gt. This allows the display to be 12 high and the set/reset sequence are the same which makes the wiring simpler. The advantage of this ROM design is that the output is 1x1 which can directly power a 1x1 pixel display. The main downside of this design is the lag. This is fine for showing still images but it is not suited for displaying animations.
Display (armor stands): The concept of using armor stands to power a 1x1 lamp display is not new but I think it hasn't been done this compact (otherwise, please let me know). The advantage of this display is that it produces little lag. The disadvantage is that it produces "passive" lag from the entities (armor stands).
Display (observer clock/redstone block) I forgot to show this display (http://prntscr.com/ubs9du). Another way to power the lamps is using a combination of observer clocks and redstone blocks. The advantage of this is that it is very compact and it does not produce much/any lag when the display is not powered. The disadvantage is the massive amount lag coming from the observer clocks when the display is powered.Minecraft 1-Wide 3x3 Glass Hipster DoorTowsti2020-08-22 | Unlike my previous 1-wide 3x3 hipster door, this version does not require non-stick blocks in the doorway or a modified doorframe. It does require a block above the height limiter blocks (red wool blocks). I therefore included the floor layer in the volume.
The layout is effectively a feed tape block with the ability to swap blocks. This is partly based on the layout used for the 1-wide 3x3 hipster door made by RMC.
Specs: * Volume (xyz): 33 x 21 x 1 * Opening (mm:ss): 2:16 * Closing (mm:ss): 1:50
Unfortunally the front and back of the walkway through the door need to be unmovable objects making this not a true hipster door.
Nevertherless, this project was a ton of fun and quite challenging. Especially the sideways "slider" that initially pushes rows upwards and the leaf bit that is used to avoid updating the feed tape.
For the timing circuit, I designed a feed-tape minecart ROM inspidered by a design used by Comet107 (youtube.com/watch?v=HTx77skRQr8). The minecart ROM on itself is quite compact but considering the opening/closing sequence is not that complex, it is better to use a smaller design such as a normal feed-tape. The volume (x*y*z) is 36x21x1 (including floor and the blocks above the floor )
Like with a lot of my builds, credit to RMC for being first (as well as making a way better version): youtube.com/watch?v=8baHlqTgxI8 They made a full 1-wide 3x3 hipster door that I highly recommend checking out.
World download: drive.google.com/file/d/1uwLlsFBOhR_Pls652kLNuXAA0VnErv2i/view?usp=sharingMinecraft 2D Printer Door (Chest Customizable/Hipster)Towsti2020-07-27 | specs/features: size (xyz): 55 * 37 * 24 (48840 blocks) opening speed (mm:ss): +- 5:00 closing speed (mm:ss): +- 4:00 (time to fill 54 slots with swords is excluded from closing)
- 12 different blocks (the blocks are powered so no tnt and redstone components) - 54 block storage of every block type (enough for 9*6 rectangle of a single block) - support for solid and air "blocks" - automatic filling of air blocks (fills up to 54 air blocks) - 100% of blocks + items are recycled - hipster layout
The main reason for why the door is so large is due to using parallel block loading/storing (http://prntscr.com/tp9auz) instead of sequential block loading/storing (https://prnt.sc/tp998z). In theory, this is a lot faster since it's moving all the blocks in the row at the same time. It does however rely on flying machines which are slower than block conveyers (which are used for sequential loading/storing.
The item queue is used to invert the sequence for the rows (highest row is placed first and the lowest row is removed first). I orignally had just hoppers which is half the size but this was unreliable due to hopper aging (unload order is random), this was fixed using a combination of hoppers and droppers which so far is 100% reliable.
For the input I used a chest similar to some of the 2d concrete printers. I'm quite pleased with how the input/interface turned out since it's minimalistic but still easy to use (I think). I hope it's clear from the showcase but all of the blocks and items, including the swords that are used to fill air spacess, are recycled.
notes: - The door pattern that I used for the initial clip is not the same as the one that I used for the actual opening/closing shots *facepalm* (https://prnt.sc/tp5w9v). - After creating the showcase, I managed to remove a block from the UI panel (https://prnt.sc/tp4tv2).
The digit change time is about 8 seconds.Mid-Air Returning Platform (Expandable / Infinite Fly Height)Towsti2020-05-31 | Sorry for the influx of mid-air returning platform videos but this should be the last one. If you look very closely, you may notice a few ghostblocks. In hindsight, I should have recorded in 10tps and sped up the recording afterwards.
The specs are pretty much the same as before except that this version has an infinite fly height. Unfortunally the return pause of the platform takes slightly longer and the height is also longer. Both are a result of the additional module for the platform flying machine.
To make it truly expandable requires instant redstone lines for the length and width but this is trivial as there is plenty of space. If I do plan to make an expandable 3d hipster door it will most likely be 16 x 16-32 x 16 as the wiring will be simpler and the lag is somewhat manageable.Mid-Air Returning Platform (Expandable)Towsti2020-05-28 | Updated version (with infinite fly height): youtu.be/3MmMFdGuphs
Expandable (width/length and height to some degree) mid-air returning platform
I'm not sure if there are any uses for this besides a 3d hipster door (self-building house) but I'm quite proud of the design so I thought I'd make a separate video on it.
Features: - Expandable in width/length - Expandable in fly height to some degree (will require additional modules beyond +- 35 blocks in height) - "Instant" returning platform - Compact layout - Ability to push blocks (with solid layer on top) - Ability to retract blocks (slime/honey on top) - Low minimum fly height - Can stop/return at any level
The mid-air returning is done by using a fast (10gt) engine that stops the slow (16gt up/10gt down) platform after which both flying machines hit the push limit making them return. The improvements over the self returning platform that I used for a recent 3d hipster door is that this version returns instantly, it's expandable, it has a lower minimum fly height and a higher maximum fly height.
Some of the challenges were the following: - 16gt up/10gt down engine return mechanism that fits within a 2 wide space. - Preventing modules that are unreachable by redstone wire to collapse (the return flying machine stops the platform, hence the terracotta layer). - Preventing the stopper flying machine from triggering the platform flying machine to early.
In the end I had to use a directional layout as it meant I could place obervers in between the engines.
This leaves the following improvements: - Non-directional design - Infinite fly height
For a 3d hipster in particular, there's another requirement to push the platform up 3-5 blocks while returning the platform manually. This is needed for fast removal of layers from the platform (youtube.com/watch?v=DJjrKNikNvk at 3:49)Minecraft 8x8x8 Self-Building House (Air/Solid Hipster)Towsti2020-05-20 | Much improved version: youtu.be/CbbE8nEnPeE?si=C6fxOLyPSnMoxw52
This is not the first self-building house (check credits) so my goal was instead to optimize the opening/closing speed.
This resulted in the following goals: - The vertical flying machine (platform) MUST be the main bottleneck in the system instead of the storage and layer (un)loading. - Fast layer (un-)loading. - Fast vertical flying machine: - Fast trapdoor design - Actions (e.g placing and storing blocks) happen at the same time.
For the most part, with the exception of the vertical flying machine, the goals were achieved. For the vertical flying machine (vertical platform), I opted for a more compact and less complex but also a slower design. Some of the other designs I considered include the 3 module collapsing "trick" which was used in my 10 high "any" door and a height limiting wall which I used in my expandable "any" door.
The layer storage design can be expanded to store 9 layers without increasing the volume but beyond that it's not very scalable. The main benefit of this design is that it doesn't require a lot of space above the storage unlike most storage designs which require a flying machine that pushes layers down. The storage slots are individually addressable. This means that you could, for example, program the minecart ROM to flip the house instead of opening and closing it.
Adding/removing layers is about as fast as it gets because I mirrored the flying machines that move layers between the storage and the vertical flying machine. The removal of layers from the slime/honey and closing the door (requires a non-sticky surface) is combined using a piston/observer shifting platform. It has to shift twice due to the 12 block push limit.
If I ever decide to make another air/solid hipster door I will mainly focus on improving the vertical flying machine so that it’s expandable, instant returning and can fly infinitely high. Ideally it also uses a 10t engine for flying upwards but I don’t think this is very difficult as the “stopper” flying machine has to be faster than the platform engine. I'd also change the layer storage design assuming it’s still able to keep up with the vertical platform.
Specs: Volume (X*Y*Z): 61 * 22 * 31 (41602 blocks) Closing (MM:SS): 4:11 (trapdoor closes after 3:48) Opening (MM:SS): 3:40 (trapdoor closes after: 3:03) Made on Snapshot 20w20b (1.16)
Trapdoor: 9 wide flush trapdoor. This is the compact trapdoor design which is the main advantage of this layout. The following design is required when the width exeeds 8 blocks: youtube.com/watch?v=4mTiHbHagI8
Height limiting module: The height limiting module is based on the vertical slime wire by Comet 107 (youtube.com/watch?v=mphGCLV_rSc) The module is pushed up using a 2 speed engine (slow upwards movement) which has to do with the orientation of the modules.
Block storage module: The block storage is based on a 2 way flying machine that is collapsed using the height limiting module. The retraction of the block storage module and block placing/storing module collapses the height limiting module (similar to last layout shown in youtube.com/watch?v=vzGkaBMXRtc).
Block placing/storing module: The block placing/storing module moves blocks between the door frame and the storage module. This module is powered by a self returning flying machine.
Improvements for this layout: - Faster engine for height limiting modules. - Reducing the size of the height limiting module.15*10 Any Hipster Door (Glass/Solid/Air/TNT)Towsti2020-05-02 | This door supports most types of blocks including air, TNT and redstone components (hipster hipster door maybe ?). Thanks to fufususu for the "Äny door" naming suggestion. I included some additional layouts that can be used to expand the height. I don't plan to use any of the other layouts for future doors as they are either slow or large.
This is not the first door of its type, RMC created a much larger one that is also configurable: youtube.com/watch?v=03A6T_5uAXY. I do think it's the first one that supports this many block types and it's a lot faster thanks so honey blocks.
Think this is the first (8*8 done previously). I think this is the fastest 10 wide layout and it's fairly compact (can decrease the height by 1 block with a slightly different layout)
The piston activation and extension/retraction is all done using a single flying machine (other trapdoor in the background): youtube.com/watch?v=-8v8z3_avmo
I can make this expandable in both directions but it would be a lot slower with the layout I was planning. I might work on a better layout in the future.Expandable Flush Trapdoor (New Layout)Towsti2020-04-16 | First time this layout has been used (I think ?).
I didn't optimize the timing circuit in order to keep it simple and it's late and I want to get the video out. The piston layout and timing circuit is fairly compact although I would change the piston activation part slightly so that 1 less row of visible slimeblocks and honeyblocks.
I am working on a seamless version which I can finish soon hopefully. This one however uses a different layout that is a lot slower.Flush (Both Sides) Seamless Expandable (Height/Width) Piston Door [4 Wide/Compact]Towsti2019-11-03 | Next time, I'll build something where the build is more complex than the title.
I used a 2 wide design which means that the flying machine engine is 1 block smaller and I think this also makes it the biggest (vertically) flush/seamless door (until someone creates a horizontally opening one).
This layout can be 14 wide but it's easily expandable although the opening/closing will be slower the wider the door is.
The top section can also be faster but this will increase the height by one block.[First Ever!] Minecraft Staircase Block Expandable Hidden Staircase (Seamless/Continuable)Towsti2019-10-27 | I guess I'm done with the hidden staircase :) I don't think there's much to improve for me in terms of hidden staircases other than compacting the design or finding a better layout. I won't bother with either as I mainly wanted to build the first expandable hidden staircase block staircase. Turns out however, that a seamless and continuable design was also possible thanks to honey blocks :)
Huge video coming after this one :oMinecraft Compact Tileable Hidden Staircase (Seamless/Continuable)Towsti2019-10-19 | Who would have thought, another hidden staircase ! This time using the new honey blocks for a compact and relatively fast and simple design. Technically, the current design shown in this video isn't seamless but it's very easy to make it seamless by simply retracting/extending the section of the wall that is connected to the slimes.Minecraft Compact Vertically Expandable Flush Piston DoorTowsti2019-08-21 | Redstone terminology and video titles don't go well together... Think this is about as compact (and fast) as it gets apart from the noteblock on top.Minecraft Tileable Hidden Staircase (Seamless and Continuable) [PoC]Towsti2019-08-21 | One step closer towards a tileable staircase block hidden staircase :)
This is only a proof of concept for the layout, I didnt optimize the opening or closing sequence. I experimented with many layouts based on a tripple piston extender or a layout similar to this youtube.com/watch?v=JcD9Q2aFkJ4&t=77s but the middle double piston extender would only work with this design.
As far as a staircase block design, it's possible to swap pistons/observers/solid/staircase blocks with a similar design so I might experiment a bit with that when I have time.Minecraft Tileable Hidden StaircaseTowsti2019-08-17 | Solid upload schedule.
Didn't optimize the opening/closing sequence so ignore the spaghetti timing circuit. I Think it's possible to make a seamless variant or maybe even a design with staircase blocks.
Current sequence: http://prntscr.com/otwgy2minecraft 4x4 hidden staircase [WIP]Towsti2012-10-25 | Please read! im already working on a second version which will be way better, no more fences and it might even have an extra stair so you wont bump youre head when walking inside. I also didnt make it faster or more compact because 1.4.2 came out today and i couldnt update because of possible bugs :S. and ofcourse i wanted to try out the update.minecraft 12high inf wide buttonpad !!Towsti2012-09-25 | take that cubehamster ! :D
after lots of alternative 11 high's i finnaly managed to make one 12 high, the reason i filmed the back first was to show that the outputs were resetted before pushing the buttons, i also pressed the buttons real slow because of major film lags which made it almost impossible to push any button at all. the main reason i did this was to get started to making creative builds because i have been playing survival more latelyminecraft 4x3 hidden trapdoorTowsti2012-07-15 | a 3x4 trapdoor which took me way to long to build because of numerous bugs which have never happened to me before. i didnt want to upload this at first because i already uploaded alot of trapdoors but because of the upcoming piston fixes i decided to upload this.minecraft 3x3 hidden trapdoor V.2Towsti2012-06-26 | new version of my earlier 3x3 hidden trapdoor -now with way faster speed -way more compact -tested in smp -no more pistons moving after its closed or openedminecraft 3x3 hidden trapdoor (no fences or sand!)Towsti2012-06-25 | yeah i know its huge and slow. but thats because i didnt want to crash the mechanism because of bad timing causing pistons to push eachother plus it would be real hard to time since some pistons require a 1 tick pulse some an t flipflop and the others a normal pulse. i think this is the first one that doesnt have anything to block the hole like a fence with a torch on it or sand like my 4x4 trapdoor did.minecraft 4x4 hidden trapdoor concept [WIP]Towsti2012-06-20 | if you know someone who already made this please send me his/her video!! the reason the hole isn't deep is because i only used a triple piston extender. if you would use a 10 high sand extender the hole would be way deeper, also you can use fencegates or inviseble blocks but i think that ruins the challenge. also this thing could be made alot faster and smaller, but this is just a concept to (possible first ever 4x4 completely hidden trapdoor!) if no one has done this before me i'll probably make one alot better.minecraft seamless, instant connect 4Towsti2012-06-11 | probably the most compact possible (not including automatic turn selector) my opponent was Darkblade164 this design can be real nice in smp 1.3 because it doesnt require any sticky pistons and complicated redstone. also you can trade with villagers so you have enough glowstone.minecraft 1 wide binary decoderTowsti2012-06-11 | not sure if this has been done before, if so please leave a link to the person who made this. im not good with binary so i just made a simple 4 bit binary to decimal ofcourse you could add as many bits as you want. not really having inspiration at the moment so i'll be posting some random video'sminecraft compact tileable xnor gateTowsti2012-06-10 | minecraft xnor gate. the fencegate/lamp is there because the piston underneath needs a block update to activate