Question: I tried something new in this video by time lapsing the boring parts. Was it horrible to watch? Would it be better if I just skipped it and talked about what happened or would you rather see a time lapse?
I, personally, would like to see the raw unedited footage even if it's without any speaking although I'm sure there are people who would rather see the finished product so maybe putting a timestamp somewhere on the video or in the description to where they should skip to would be best as it gives people a choice.
This is my first video of yours I watched and I don't even play this game, so take my feedback knowing that. I really liked the format, I am a developer and I could follow the logic and the gameplay even though I have never played. The time lapse was critical in helping me understand the effort involved and I can slow it down if I needed to examine a particular moment. Also if it were not sped up I think I would have become bored and maybe left the video. Overall this format was super beginner friendly and I expect I will watch more of your content because of how much you care about the user experience.
Can't tell you how long I waited for you to noticed you messed up the output of the colour selector. And then you fixed it in the time-lapse. 😂. I'm glad you got it sorted. Keep up the great work bud
It always inspires me to see how many ways people find to approach a problem. I will say a MAM can be deceptively static if the steps are ordered for it.
@@DataEngineerPlays Hey, fellow coder here. I’m very familiar with “the process” haha. Get it working, I’m excited to see yours! I’d love to submit my designs if you ever do a showcase of viewer builds you like.
@@baphnie Oh that's a fantastic idea, maybe after I finish my MAM I can get my viewers to submit their MAMs and I can do a review of your build! Though I'm not exactly the authority on efficiency... But I'd love to see what you guys did better than me
@@DataEngineerPlays oh that would be awesome! I would definitely submit my MAM. Although it's only a 4 belt 4 layer no crystal or pin MAM. Working on a 12 belt insane level MAM now! :D
You can also save some paint and painting if you paint before cutting in corners. 1 belt of full shapes is 4 lanes of corners. 4x less painting. So you can paint 3 lanes of shapes and then cut them in to 12 lanes of painted corners.
4x less painting but 7x more shapes if you precolor them? You would need 4x7 (28) belts of precolored shapes, decide which color you need for each corner, then cut... unless I'm missing something
@@DataEngineerPlays like in current version you are painting after filtering the shapes for a particular quadrant. Similarly you would paint after filtering the shapes so no need for precoloured shapes. Current ver. : cut -> filter -> paint new ver. : filter -> paint -> cut
@@nirmalsingh7646 it would require complicated logic if two of the same quarter type were different colours though as youd need to have a way to split what usually is just one output of a quarter into two outputs into two painters
This can be resolved if you think of a belt of wholes as a future 4 belts of quarters. In that case, you only run one belt per layer of wholes after filtering (per quarter) rather than 4.
@@DataEngineerPlays Im building along these videos, kinda using them as a baseline, but i been doing a lot of things different. Wether theyre better or not, i have no idea. For painting, what i currently made consists out of: - 6 3x3 foundations, each holding a painter for 1 color, that cover 2 full space belts of quarters. (so i need 12 foundations in total) - Each has 2 radio receivers (one for each belt) to check wether that corner needs coloring. I check this by having a constant signal output the particular paint color on that foundation, and then using an equals to compare it with the radio receiver output. If they match, the quarters get filtered to the side to get painted and then merged back to their belt. If they dont, they keep going straight to the next foundation. If they dont need any painting, they pass through all the painter foundations and go to assembly. I dont know if this is any better, but it works great so far.
you could make a 1x1 comparator foundation, this could serve multiple purpoises, it has one input and two outputs, basically like a filtered belt. ==[r]==[g]==[b]==[w]== | | | | | | | | but you dont need to be limited to colors, you could also use the "filter" foundation to filter shapes eg...
Your re-combinator will work better if you do your deciding to bypass the swapper based on the input shape instead of reading if the belts have anything on them -- that'll avoid issues with the lags in throughput causing problems
One thing i love about launchers (at least the shape ones havent tested with the paint ones yet) is that they can launch things through buildings like stackers
just began looking at videos but already built my own mam and is about to adapt it to crystals as well... but that will take some time... but think "logic" what takes more paint.. painting 4 quarters or painting 1 whole item? ;-) 1. select shape 2. paint shape 3. cut shape 4. rotate shape to appropriate corner 5. merge shapes 6. merge layers A. figure out how to make standalone pins and add full shapes of just pin to your inputs (I added mine in after the painter since they don't need painting) B. avoid huge belts between platforms it will slow things down in how fast they get through your MAM
I dunno doesn't that mean you need 7x the shapes? Let's say your quad 1 is a red circle 1. select shape - so you have 4 belts of uncolored shapes coming in, select a circle 2. paint red - so now you have 7 belts of colors coming in to paint it red 3. cut shape, you discard 3/4 of a red circle and keep the quarter you want Do step 1-3 3 more times. So in total your inputting 4x4 belts of shapes, 4x7 belts of color? Isn't that more than what I have since I'm bring in 4x7 belts of color and 4x1 belt of shapes?
@@DataEngineerPlays no you give them 1 cicle UNPAINTED... then you paint the circle whatever color... then cut it and you have 4 slices of that color (rotate so they are all in the same corner) that is ONE string (quarter of a layer) you don't need to store and feed 7 differently colored circles, you color them if they need to be colored (all 4 pieces in on go) instead of cutting and then paiting each quarter as you did in the video
I think I need to see your MAM to understand. In my head this still tells me I need 4x the shape input for each quarter... But yes it would save 4x the paint input....
@@DataEngineerPlays the input is so high for each shape considering you "only" output 4 full shapes at any time you should be able to see it on the links I provided for screenshots of it? hmm looks like YT removed the links odd...
i was thinking, to filter the belt to go into the right painter you could use something similar to a binary search where you split it down the middle then see which path it should go. So, you can use one belt splitter with your given signal and compare it with the desired signals on the left connecting them with an or gate then to the splitter, then you can continue this down the tree
@@Obaker30 It's always a toss up right, you can always write more efficient code but is it worth it if the inefficient easy version is 10 lines but the super optimised version is 1000 lines
two thing I want to add (I'm not finished watching) : I would approach differently. I would filter the color by plateform. each small plateform decide if it goes to the coloring affected by that plateforme or go through. I think you need to filter only 7 time because whatever go through the 7 decisions is the 8th. Sorry for my english.
I'm sure someone has already suggested this but you could have just had the uncolored be the "else" and fall off the end of the list. This would have made it a little smaller. The way you did it works, obviously.
@@DataEngineerPlays Yeah, I figured that out 2 hours ago lol. Also, don't forget to copy the save file before changing, just in case the file bugs or whatever, didn't happen to me, but I would play it safe
@@thearrogantone4756 since you're ahead give me a hint. My math says you need 7x4 for painting 7x2 for crystalising so 42 x 1800L pipes per layer x4 layers?? Or am I being inefficient?
@@DataEngineerPlays You also forgot about wire crosses? blanking on the name, but they make the wiring much simpler if pipe valves weren't a thing. (they make it more square, at least the design I came up with, all the belt filters were next to each other, but the cables do make it unjumpable with how much width they add and make it more messy looking, yours def looks cleaner)
using symmetry everywhere is one of the main causes of tunel vision, as far as my examples go, lol. anyway, breaking the symmetry adds details and things only needs to be symmetric within themselves. What really matters is organization and a recognizable pattern.
This is kind of a side issue to your question about how much paint you need. If you keep the basic shapes as whole shapes for a bit longer and only cut them into quarters after they are painted, then your paint consumption will only be a quarter of what it is in your current design. (Or you could paint - and deliver - 4 times as many shapes with the same amount of paint.)
I've thought about this, but have you thought about then how many base shapes you would need? If you're painting the whole shape that means you need 7 colors of each of the 4 base shapes ie. 28 inputs of base shape. Interestingly enough if you watch a Shapez 1 MAM video I believe that's how they did it, you'll see a massive main bus of colored base shapes... 28 belts 4 shapes x 7 colors each
@@DataEngineerPlays I've seen one that had a main bus of uncolored basic shapes only. For each corner of each layer it the pulls aside one belt of the relevant basic shape. (His total MaM would deliver 4 plain belts of finished product, so only one layer on a space belt.) Then it only sends one belt through painting. I laughed out when you called yourself a silly muppet in this video, because that other MaM, I saw, was made by someone calling himself Hairymuppet.
@@DataEngineerPlays you need 4 BASE SHAPES with same thing on each side.. you are thinking of FINISHED PRODUCT.... but they are not BASE SHAPES ;-) look at how how many BASE shapes there are... then make a full shape of those in each corner (they should be on the map, train them in to your machine) and you need 1 coloring machine which is now WAY smaller for each quarter of the layer... my painter takes up 1x2 platform per quarter...
My MAM get's supplied with basic shapes and color via train. In an amount that I consider enough on average. Clearly if I'd make 4 layers of just red squares, it would definitely run out of squares and red. But how likely is that? This also separates the building process from the supply process and helps not going insane ^^
i have seen this belt thing also but its like as soon as i trace it back to the problem if i replace the output belts on that platform it fixes itself so i think its a game bug from the last patch. and the latest beta added a pipe filter like the belt filter. i'm not up for running a beta but it is close to main release. the time lapse parts work its more important to show and talk through the logic portions of the MAM, it may be a bit fast hard to follow so maybe just a jump cut or slow the replay slightly. Great Series!
You do know they have updated the game fixing a lot of issues and adding some QoL. This is just a reminder in case you are still playing on the 0.0.7 version and recorded a bunch of videos in advance .
I think you may want to redesign your MAM "a bit". :) * It might be a good idea to get rid of that right part of the MAM splitting shapes into pieces since you have to color these. In this configuration, you have to color each piece individually. That's up to 4 times the amount of paint per shape (as you already noticed), since you're painting each quarter individually and painters always spend a fixed amount of paint. Instead, send full shapes through the MAM and at the last step split and recombine them. This way you'll spend 1 unit of paint per 4 quarters instead of 1:1. Will be a paint to implement with crystals later, though. Possible, but a pain nevertheless. You'll be able to make shapes with 4 different crystal colors per layer this way... even though the game likely never generates such shapes in random tasks. I heard it only use a single crystal color per layer. BTW, as I know Vortex accepts pins in place of empty quarters, and you can crystallize them, so if you make a shape consisting out of 4 pins and pass it for empty quarters and crystals (and crystallize them down the line) you'll be able to use only swappers at the last step to recombine shapes at the last step into a required shape. * It might be easier to make a series of platforms, each designed to divert all belts into painters and apply a single color. So, it either passes shapes through unmolested or paint them and then pass them to the next. A coloring array for 12 belts takes at least 1 full block of space, so you'll need at least 2×1 platform to fit coloring, extra belting and logic to divert all belts for a single color. The point of pass-through is that you'll be able to chain them in a line and only connect paints from a side. And to simplify things, you'll be able to connect a train station to each and deliver all colors on a single train instead of 7 separate belts.
@@DataEngineerPlays BTW, new version dropped. "Added Pipe Gate, allowing you to control the flow of fluids with logical signals." Also, virtual pin pushers and automatic tunneling for space belts and trains, among the other things.
yeah, efficency would paint the full shape before splitting, also, why split the painting 'brain' from the paint stations, when you could spread the brain out to the paint stations and just check if the needed colour matches the current paint station, if not, bypass it.
Can I suggest removing the uncoloured filter, and just making the shapes go through the end of the belt and round to an exit? The uncoloured filter is just completely unnecessary I think as the only reason you’d pass any of the coloured filters, is if the shape is meant to be uncoloured, or if there’s no shape in the current section, which shouldn’t be an issue either, since you can just bin any unwanted shapes with filters anyways later down the line if you really wanted to
@@DataEngineerPlays tanks work by taking an input and filling the tank with that input to later be taken out by the output. If you change the input color, then the tank will drain at the rate that it fills(not sure if this consumes the new paint color, and this is only what the UI is showing, I'm assuming the draining happens over time because that's what the UI shows, and not instant) once completely drained it will refill with new color.
@DataEngineerPlays i had no idea honestly since i had not unlockrd tanks yet and was just asking/ brainstorming but with the latest patch it doesn't matter anymore since you can attach signals to color now from what I glanced from patch notes. Might have misread but will double check when I have time. Anyways have fun and keep up the good work
thought about that... would be hyper efficient for the input materials, but the the logic circuit would be supeeeerr crazy... Remember there's no adders or "for" loops in this game...
You copied your “brain” for the color. But you need to add rotators to align the color to the quarter you need. It only worked bc the entire shape was green. If it was multi colored it wouldn’t work. Unless you fixed it during a Timelapse
@@DataEngineerPlays wow yeah; it's fixed... I don't have the game so I can't test. I think i thought of that because I'm watching nilaus' gameplay and somehow his jumpers are always tight... which only really makes it more impressive honestly, because I really thought he had to ajust it to make it fit like that every time.
@@BleachWizz I've avoided watching other youtubers to maintain the authenticity of my buiilds, but I believe if he's doing super tight belts you can get 2 sets of jumpers into 1 lane which might make it look like it's been adjusted, but no its 2 sets of fixed distance jumpers sharing a lane
![ค้างคืนโรงเก็บกระสวยอวกาศร้างกลางทะเลทราย [Around The World Ep. 3]](http://i.ytimg.com/vi/E5bWZQhABJE/mqdefault.jpg)
Question: I tried something new in this video by time lapsing the boring parts. Was it horrible to watch? Would it be better if I just skipped it and talked about what happened or would you rather see a time lapse?
I, personally, would like to see the raw unedited footage even if it's without any speaking although I'm sure there are people who would rather see the finished product so maybe putting a timestamp somewhere on the video or in the description to where they should skip to would be best as it gives people a choice.
I like the time lapse, gives us a sense of time. if you were to go find a specific asteroid or something that you could probably skip.
This is my first video of yours I watched and I don't even play this game, so take my feedback knowing that. I really liked the format, I am a developer and I could follow the logic and the gameplay even though I have never played. The time lapse was critical in helping me understand the effort involved and I can slow it down if I needed to examine a particular moment. Also if it were not sped up I think I would have become bored and maybe left the video. Overall this format was super beginner friendly and I expect I will watch more of your content because of how much you care about the user experience.
@@fortuneswake Thanks for the feedback, very helpful!
I like the time lapse, maybe slow it down a bit, add separate music, and add some post commentary about the overall state of the MAM?
Can't tell you how long I waited for you to noticed you messed up the output of the colour selector. And then you fixed it in the time-lapse. 😂.
I'm glad you got it sorted. Keep up the great work bud
Update 0.0.8: Added Pipe Gate, allowing you to control the flow of fluids with logical signals
It always inspires me to see how many ways people find to approach a problem. I will say a MAM can be deceptively static if the steps are ordered for it.
What I'm hearing is I'm doing it wrong 😆
@@DataEngineerPlays Hey, fellow coder here. I’m very familiar with “the process” haha. Get it working, I’m excited to see yours! I’d love to submit my designs if you ever do a showcase of viewer builds you like.
@@baphnie Oh that's a fantastic idea, maybe after I finish my MAM I can get my viewers to submit their MAMs and I can do a review of your build! Though I'm not exactly the authority on efficiency... But I'd love to see what you guys did better than me
@@DataEngineerPlays oh that would be awesome! I would definitely submit my MAM. Although it's only a 4 belt 4 layer no crystal or pin MAM. Working on a 12 belt insane level MAM now! :D
@@Sankyre Good luck! It's not too much to turn a 4 belt into 12 belt
You can also save some paint and painting if you paint before cutting in corners. 1 belt of full shapes is 4 lanes of corners. 4x less painting. So you can paint 3 lanes of shapes and then cut them in to 12 lanes of painted corners.
4x less painting but 7x more shapes if you precolor them?
You would need 4x7 (28) belts of precolored shapes, decide which color you need for each corner, then cut... unless I'm missing something
@@DataEngineerPlays like in current version you are painting after filtering the shapes for a particular quadrant. Similarly you would paint after filtering the shapes so no need for precoloured shapes.
Current ver. : cut -> filter -> paint
new ver. : filter -> paint -> cut
@@nirmalsingh7646 it would require complicated logic if two of the same quarter type were different colours though as youd need to have a way to split what usually is just one output of a quarter into two outputs into two painters
This can be resolved if you think of a belt of wholes as a future 4 belts of quarters. In that case, you only run one belt per layer of wholes after filtering (per quarter) rather than 4.
@@DataEngineerPlays
Im building along these videos, kinda using them as a baseline, but i been doing a lot of things different.
Wether theyre better or not, i have no idea.
For painting, what i currently made consists out of:
- 6 3x3 foundations, each holding a painter for 1 color, that cover 2 full space belts of quarters. (so i need 12 foundations in total)
- Each has 2 radio receivers (one for each belt) to check wether that corner needs coloring.
I check this by having a constant signal output the particular paint color on that foundation, and then using an equals to compare it with the radio receiver output.
If they match, the quarters get filtered to the side to get painted and then merged back to their belt.
If they dont, they keep going straight to the next foundation.
If they dont need any painting, they pass through all the painter foundations and go to assembly.
I dont know if this is any better, but it works great so far.
you could make a 1x1 comparator foundation, this could serve multiple purpoises, it has one input and two outputs, basically like a filtered belt.
==[r]==[g]==[b]==[w]==
| | | | | | | |
but you dont need to be limited to colors, you could also use the "filter" foundation to filter shapes eg...
interesting concept
Your re-combinator will work better if you do your deciding to bypass the swapper based on the input shape instead of reading if the belts have anything on them -- that'll avoid issues with the lags in throughput causing problems
yes I address that in the next episode
One thing i love about launchers (at least the shape ones havent tested with the paint ones yet) is that they can launch things through buildings like stackers
just began looking at videos but already built my own mam and is about to adapt it to crystals as well... but that will take some time... but think "logic" what takes more paint.. painting 4 quarters or painting 1 whole item? ;-)
1. select shape
2. paint shape
3. cut shape
4. rotate shape to appropriate corner
5. merge shapes
6. merge layers
A. figure out how to make standalone pins and add full shapes of just pin to your inputs (I added mine in after the painter since they don't need painting)
B. avoid huge belts between platforms it will slow things down in how fast they get through your MAM
I dunno doesn't that mean you need 7x the shapes?
Let's say your quad 1 is a red circle
1. select shape - so you have 4 belts of uncolored shapes coming in, select a circle
2. paint red - so now you have 7 belts of colors coming in to paint it red
3. cut shape, you discard 3/4 of a red circle and keep the quarter you want
Do step 1-3 3 more times. So in total your inputting 4x4 belts of shapes, 4x7 belts of color?
Isn't that more than what I have since I'm bring in 4x7 belts of color and 4x1 belt of shapes?
@@DataEngineerPlays no you give them 1 cicle UNPAINTED... then you paint the circle whatever color... then cut it and you have 4 slices of that color (rotate so they are all in the same corner) that is ONE string (quarter of a layer) you don't need to store and feed 7 differently colored circles, you color them if they need to be colored (all 4 pieces in on go) instead of cutting and then paiting each quarter as you did in the video
I think I need to see your MAM to understand. In my head this still tells me I need 4x the shape input for each quarter... But yes it would save 4x the paint input....
@@DataEngineerPlays the input is so high for each shape considering you "only" output 4 full shapes at any time you should be able to see it on the links I provided for screenshots of it? hmm looks like YT removed the links odd...
i was thinking, to filter the belt to go into the right painter you could use something similar to a binary search where you split it down the middle then see which path it should go. So, you can use one belt splitter with your given signal and compare it with the desired signals on the left connecting them with an or gate then to the splitter, then you can continue this down the tree
at the point of the video i’m at (15:51) you are using a sort of linear search if you think about it
The only problem I foresee is the immense amount of space a binary search belt would take up and how hard it would be to wire it up
@@DataEngineerPlays completely right, i forgot you need to do it for 12 belts
@@Obaker30 It's always a toss up right, you can always write more efficient code but is it worth it if the inefficient easy version is 10 lines but the super optimised version is 1000 lines
two thing I want to add (I'm not finished watching) : I would approach differently.
I would filter the color by plateform. each small plateform decide if it goes to the coloring affected by that plateforme or go through.
I think you need to filter only 7 time because whatever go through the 7 decisions is the 8th.
Sorry for my english.
Also I realised that you don't use any trains it would reduce the MAM so much ahahah :D
When you finish MAM you are gonna make a hexagon version. That could be really really crazy.
Haha I think I'll need to play hex mode first
I'm sure someone has already suggested this but you could have just had the uncolored be the "else" and fall off the end of the list. This would have made it a little smaller. The way you did it works, obviously.
In next update they are introducing pipe gates. It is available in beta branch
Are you serious.. so like belt filter for pipes? lol well then scrap this episode haha!
@@DataEngineerPlays Yeah, I figured that out 2 hours ago lol. Also, don't forget to copy the save file before changing, just in case the file bugs or whatever, didn't happen to me, but I would play it safe
Right after I did all of the color coding to, basically doing what you did, but larger scale
@@thearrogantone4756 since you're ahead give me a hint. My math says you need 7x4 for painting 7x2 for crystalising so 42 x 1800L pipes per layer x4 layers?? Or am I being inefficient?
@@DataEngineerPlays You also forgot about wire crosses? blanking on the name, but they make the wiring much simpler if pipe valves weren't a thing. (they make it more square, at least the design I came up with, all the belt filters were next to each other, but the cables do make it unjumpable with how much width they add and make it more messy looking, yours def looks cleaner)
using symmetry everywhere is one of the main causes of tunel vision, as far as my examples go, lol.
anyway, breaking the symmetry adds details and things only needs to be symmetric within themselves.
What really matters is organization and a recognizable pattern.
This is kind of a side issue to your question about how much paint you need.
If you keep the basic shapes as whole shapes for a bit longer and only cut them into quarters after they are painted, then your paint consumption will only be a quarter of what it is in your current design. (Or you could paint - and deliver - 4 times as many shapes with the same amount of paint.)
I've thought about this, but have you thought about then how many base shapes you would need? If you're painting the whole shape that means you need 7 colors of each of the 4 base shapes ie. 28 inputs of base shape.
Interestingly enough if you watch a Shapez 1 MAM video I believe that's how they did it, you'll see a massive main bus of colored base shapes... 28 belts 4 shapes x 7 colors each
@@DataEngineerPlays
I've seen one that had a main bus of uncolored basic shapes only. For each corner of each layer it the pulls aside one belt of the relevant basic shape. (His total MaM would deliver 4 plain belts of finished product, so only one layer on a space belt.)
Then it only sends one belt through painting.
I laughed out when you called yourself a silly muppet in this video, because that other MaM, I saw, was made by someone calling himself Hairymuppet.
@@DataEngineerPlays you need 4 BASE SHAPES with same thing on each side.. you are thinking of FINISHED PRODUCT.... but they are not BASE SHAPES ;-) look at how how many BASE shapes there are... then make a full shape of those in each corner (they should be on the map, train them in to your machine) and you need 1 coloring machine which is now WAY smaller for each quarter of the layer... my painter takes up 1x2 platform per quarter...
My MAM get's supplied with basic shapes and color via train. In an amount that I consider enough on average. Clearly if I'd make 4 layers of just red squares, it would definitely run out of squares and red. But how likely is that?
This also separates the building process from the supply process and helps not going insane ^^
Yeah that's a good idea, then you can always scale up the supply whenever needed. I think I'll need to do that at some point
i have seen this belt thing also but its like as soon as i trace it back to the problem if i replace the output belts on that platform it fixes itself so i think its a game bug from the last patch. and the latest beta added a pipe filter like the belt filter. i'm not up for running a beta but it is close to main release. the time lapse parts work its more important to show and talk through the logic portions of the MAM, it may be a bit fast hard to follow so maybe just a jump cut or slow the replay slightly. Great Series!
Thank you for the feedback and confirming the belt thing was a bug! That was annoying me for ages and ages...
You do know they have updated the game fixing a lot of issues and adding some QoL. This is just a reminder in case you are still playing on the 0.0.7 version and recorded a bunch of videos in advance .
thank you! I only just got the update today
The new pipegate will change this to the easier imo... Redesign or stick with what you got?
of course we'll redesign!
I think you may want to redesign your MAM "a bit". :)
* It might be a good idea to get rid of that right part of the MAM splitting shapes into pieces since you have to color these. In this configuration, you have to color each piece individually. That's up to 4 times the amount of paint per shape (as you already noticed), since you're painting each quarter individually and painters always spend a fixed amount of paint. Instead, send full shapes through the MAM and at the last step split and recombine them. This way you'll spend 1 unit of paint per 4 quarters instead of 1:1. Will be a paint to implement with crystals later, though. Possible, but a pain nevertheless. You'll be able to make shapes with 4 different crystal colors per layer this way... even though the game likely never generates such shapes in random tasks. I heard it only use a single crystal color per layer.
BTW, as I know Vortex accepts pins in place of empty quarters, and you can crystallize them, so if you make a shape consisting out of 4 pins and pass it for empty quarters and crystals (and crystallize them down the line) you'll be able to use only swappers at the last step to recombine shapes at the last step into a required shape.
* It might be easier to make a series of platforms, each designed to divert all belts into painters and apply a single color. So, it either passes shapes through unmolested or paint them and then pass them to the next. A coloring array for 12 belts takes at least 1 full block of space, so you'll need at least 2×1 platform to fit coloring, extra belting and logic to divert all belts for a single color. The point of pass-through is that you'll be able to chain them in a line and only connect paints from a side. And to simplify things, you'll be able to connect a train station to each and deliver all colors on a single train instead of 7 separate belts.
thanks for the tips especially the one where pins = empty... interest I need to test that out!
@@DataEngineerPlays BTW, new version dropped. "Added Pipe Gate, allowing you to control the flow of fluids with logical signals." Also, virtual pin pushers and automatic tunneling for space belts and trains, among the other things.
@@lainverse yeah I saw. Will do a paint redesign soon
MAKE THIS THE MOST LIKED COMMENT 👍
FOR OUR FAVORITE DATA ENGINEER 🙂
👇🙏
yeah, efficency would paint the full shape before splitting, also, why split the painting 'brain' from the paint stations, when you could spread the brain out to the paint stations and just check if the needed colour matches the current paint station, if not, bypass it.
Space pipe have 4x throughput than Space belt. So you need only 1 pipe per layer
Can I suggest removing the uncoloured filter, and just making the shapes go through the end of the belt and round to an exit? The uncoloured filter is just completely unnecessary I think as the only reason you’d pass any of the coloured filters, is if the shape is meant to be uncoloured, or if there’s no shape in the current section, which shouldn’t be an issue either, since you can just bin any unwanted shapes with filters anyways later down the line if you really wanted to
yes it's a good suggestion definitely
Was my suggestion too
You don’t need a Gate for uncolored.
Can't you use the paint tank to implement your first idea of sorting colors instead of shapes?
No idea how tanks work. Is that something that's possible?
@@DataEngineerPlays tanks work by taking an input and filling the tank with that input to later be taken out by the output. If you change the input color, then the tank will drain at the rate that it fills(not sure if this consumes the new paint color, and this is only what the UI is showing, I'm assuming the draining happens over time because that's what the UI shows, and not instant) once completely drained it will refill with new color.
@DataEngineerPlays i had no idea honestly since i had not unlockrd tanks yet and was just asking/ brainstorming but with the latest patch it doesn't matter anymore since you can attach signals to color now from what I glanced from patch notes. Might have misread but will double check when I have time.
Anyways have fun and keep up the good work
@@hugolopes4286 that is true, no worries thanks for the suggestion!
The Space Belt bug will be solved in version 0.0.8 ... it's timing is off, it's not your design's ...
thanks for confirming I was going mad
CMYK! :P
@dataengineerplays oh, update just released the pipe gate🥳
I'm guessing you could build a logic to detect which quaters have what color and merge them befor painting. But i didnt try it so I might be wrong
thought about that... would be hyper efficient for the input materials, but the the logic circuit would be supeeeerr crazy...
Remember there's no adders or "for" loops in this game...
You copied your “brain” for the color. But you need to add rotators to align the color to the quarter you need. It only worked bc the entire shape was green. If it was multi colored it wouldn’t work. Unless you fixed it during a Timelapse
I noticed that during editing. I think my next video is going to be a "code refactor and debug episode"
Awesome! Looking forward to it. Loving the series
paint before splitting and use a train with 7 wagons to transport the paint
1:18:51 ur needed shape is being dumped in the trash brother
oh nvm you noticed my bad XD
if filter by color, it can be smaller
Why did you not paint after you get the piece? than you would have to deal with only one piece...
I thought you could shorten the launchers distance.
I've seen youtubers building launchers that jump only one belt...
Tell me how...
where? I need to know please
@@DataEngineerPlays wow yeah; it's fixed...
I don't have the game so I can't test. I think i thought of that because I'm watching nilaus' gameplay and somehow his jumpers are always tight... which only really makes it more impressive honestly, because I really thought he had to ajust it to make it fit like that every time.
@@BleachWizz I've avoided watching other youtubers to maintain the authenticity of my buiilds, but I believe if he's doing super tight belts you can get 2 sets of jumpers into 1 lane which might make it look like it's been adjusted, but no its 2 sets of fixed distance jumpers sharing a lane