just as another thought, i think its interesting to look for a route of items where the needed shape is traveling straight until the stackers, this would also limit travel time and cleanse time
Interesting thought, I will revisit compactness at some point but I let's forge on for now otherwise I could spend 10 episodes optimising and everyone will get bored!
Yep pins are like fifth basic shape, bring it with full pin shape with 1 space belt then cut into 4 quarters if its needed like you did for circle or rectengular.
if you can get your color selectors to be a 1Xn that would remove the need for more spacebelts, since you have all colors on one segment of a spacebelt you could merge them onto one as a throughput
When we 4x our design for the 4 layers we would need (by my calculations which might be wrong) 3x 1800L/m per MAM layer = 5400L (1 for painting 2 for crystal making) 5400L/m x 4 layers = 21600L/m of paint = 2 space pipes worth of each color
Hum... you know... you could just provide 1 full pined belt, and select the pin shape before your painter... so you don't need 4 belts of pins. Basically you can treat it like any other shape! Just need to add a logic to the painter so it bypass on uncolored and null.
Great episode! The new design is so much simpler and more beautiful. Good job! I fully agree what you said about iterative design. When I'm developing, I need to do things incorrectly (or at least inefficiently) at the start and then scrap it. That way I know at least one way that _doesn't_ work. :) I know some people who like to begin with pen and paper, think about data structures and algorithms in their head, and once they are all finished with that, they write beautiful code from the start. I'm not like that; I think best while I'm coding, even if it's most assuredly not the best (or even the correct) way to solve the problem.
28:56 _cathartic laughter_ Yeaaaaah. That one got me too. Apparently the move doesn't actually happen until you hit "save" on the marker after moving it. I felt kinda silly when I figured that out. Like, we're doing all these complex operations there's no room left in our brains for the basics anymore.
Use the Rotators inside your existing cutting arrays to get all your quarter shapes into their correct orientation. That should give you extra space for your flushing ideas. Maybe ( I am not sure, haven't tested it yet) painters and cutters on a 2 by 1 platform to use the existing logic stuff for trash filtering on the extra space???
Near the end of the video added extra rotator platforms due to having rotators for the pins separately could just do the rotation after combining with corner belts.
23:10, please make the red dip for the other 2 to go over. then green would dip for blue to bend too; it should be possible to keep all 3 cables straight
Isnt it better to sort and bin where you join the colored corners with the pins? like the joining of the belts happens with a platform and sorters? And is the second Painter from left correct? In the end it send Yellow and White color Where there shouldnt be any color at all
Yes, since we have 12 input belts of circles 3 belts of circles will be taken for top left - painted - turned into 12 belts of quarter circles 3 belts of circles will be taken for top right - painted - turned into 12 belts of quarter circles 3 belts of circles will be taken for bottom left - painted - turned into 12 belts of quarter circles 3 belts of circles will be taken for bottom right- painted - turned into 12 belts of quarter circles Consuming 12 input belts and outputting 48 belts of quarter circles in 4 different colors, stacking back together into 12 belts full circle with 4 different colors Don't worry it confused me for like 6 episodes too...
There is something that will bottleneck your layout! The paint "shooter/jumper" (i dont remember the name) it only transport 1800 L/m, so u need 4 of them for a full pipe!
Instead of doing all your logic on each individual platform why not set up one platform and send out signals depending on what you need that way if you're looking for a top right of a shape you just have to put down a signal catcher that is top right of shape
Yeah, I was thinking there had to be a reason for all the different channels they give you, you can offload the logic wiring to a platform away from the MAM and give yourself room on basically every other platform
One of you will have to reply with more exact maths. Unless I'm missing something as others have commented the global signals only have 7 (or 8) channels? So if you were to send signals out as quarter shapes... you would need 16 signals for 4 layers?
I decided to not do a MAM and just enjoy watching you go through the effort! Hah!
Just finished last video and bam new vid is up 😀You are like a human MAM 🤣
Turns out I'm actually an AI
got the idea of dynamic paint selector from your video now i'm gonna make a new version of my mam
just as another thought, i think its interesting to look for a route of items where the needed shape is traveling straight until the stackers, this would also limit travel time and cleanse time
Interesting thought, I will revisit compactness at some point but I let's forge on for now otherwise I could spend 10 episodes optimising and everyone will get bored!
"imagine the debugger has a bug in it"
-quartershadowing
24:20 CRAZY it feels great to watch how you improve the design as you go🧠
i think it's beneficial to set up pins as a basic resource, and add a layer of selectors for the pins at the start of the MAM
Yep pins are like fifth basic shape, bring it with full pin shape with 1 space belt then cut into 4 quarters if its needed like you did for circle or rectengular.
I love the thought I will do that when we do train staging
if you can get your color selectors to be a 1Xn that would remove the need for more spacebelts, since you have all colors on one segment of a spacebelt you could merge them onto one as a throughput
When we 4x our design for the 4 layers we would need (by my calculations which might be wrong)
3x 1800L/m per MAM layer = 5400L (1 for painting 2 for crystal making)
5400L/m x 4 layers = 21600L/m of paint
= 2 space pipes worth of each color
Hum... you know... you could just provide 1 full pined belt, and select the pin shape before your painter... so you don't need 4 belts of pins. Basically you can treat it like any other shape! Just need to add a logic to the painter so it bypass on uncolored and null.
You need to move then hit save for the new location to register
At 41:41, easy way to clear everything, select it all and hit "i". This will clear shapes and color from everything carrying them.
Great episode! The new design is so much simpler and more beautiful. Good job! I fully agree what you said about iterative design. When I'm developing, I need to do things incorrectly (or at least inefficiently) at the start and then scrap it. That way I know at least one way that _doesn't_ work. :) I know some people who like to begin with pen and paper, think about data structures and algorithms in their head, and once they are all finished with that, they write beautiful code from the start. I'm not like that; I think best while I'm coding, even if it's most assuredly not the best (or even the correct) way to solve the problem.
learn by doing why not!
28:56 _cathartic laughter_ Yeaaaaah. That one got me too. Apparently the move doesn't actually happen until you hit "save" on the marker after moving it. I felt kinda silly when I figured that out. Like, we're doing all these complex operations there's no room left in our brains for the basics anymore.
Currently working on reducing the size of each layer of my mam down to if i remember a 36x27 for a 4 space belts worth of output
Use the Rotators inside your existing cutting arrays to get all your quarter shapes into their correct orientation. That should give you extra space for your flushing ideas. Maybe ( I am not sure, haven't tested it yet) painters and cutters on a 2 by 1 platform to use the existing logic stuff for trash filtering on the extra space???
I love this idea and have implemented in like episode 15 optimisation stage
A bypass on the splitter might have been nice. No need to split and then re-stack in the case of needing a full shape.
Near the end of the video added extra rotator platforms due to having rotators for the pins separately could just do the rotation after combining with corner belts.
23:10, please make the red dip for the other 2 to go over. then green would dip for blue to bend too; it should be possible to keep all 3 cables straight
Isnt it better to sort and bin where you join the colored corners with the pins? like the joining of the belts happens with a platform and sorters?
And is the second Painter from left correct? In the end it send Yellow and White color Where there shouldnt be any color at all
48:23 why have two belts merge into one conveyor (which would eventually slow your main line belts), instead of just having multiple voids?
I dont know how viable it would be but you could make use of the voiding property of train loaders
maybe add circuit conditions into a trash hole to clear all blocks faster? would clear the "lag" on the belts to clear leaving space for fresh items
yes definitely more work to be done on optimising trash collection
I'm still a little confused about getting say a circle that is four different colors. Is that possible with this new compact set up?
Yes, since we have 12 input belts of circles
3 belts of circles will be taken for top left - painted - turned into 12 belts of quarter circles
3 belts of circles will be taken for top right - painted - turned into 12 belts of quarter circles
3 belts of circles will be taken for bottom left - painted - turned into 12 belts of quarter circles
3 belts of circles will be taken for bottom right- painted - turned into 12 belts of quarter circles
Consuming 12 input belts and outputting 48 belts of quarter circles in 4 different colors, stacking back together into 12 belts full circle with 4 different colors
Don't worry it confused me for like 6 episodes too...
27:00 theres a thrick here but im praying everyone already said it and is already implemented in the next few episodes.
Do you really need 48 full belts of pins ?
Clicked "Marker Move". Never clicked "Save".
Thank you!
There is something that will bottleneck your layout! The paint "shooter/jumper" (i dont remember the name) it only transport 1800 L/m, so u need 4 of them for a full pipe!
I'll do some throughput testing
Instead of doing all your logic on each individual platform why not set up one platform and send out signals depending on what you need that way if you're looking for a top right of a shape you just have to put down a signal catcher that is top right of shape
Yeah, I was thinking there had to be a reason for all the different channels they give you, you can offload the logic wiring to a platform away from the MAM and give yourself room on basically every other platform
Due to only having 8 channels leads to lack of channels when doing 4 layers
You have 8 colored channels, which can be used to do all this. I'm already doing it in my own MAM.
@@theinnocenttomato9836 There's enough channels for 1 layer of each ROS shape, that's enough to condense a lot of the signalling through unstackers.
One of you will have to reply with more exact maths. Unless I'm missing something as others have commented the global signals only have 7 (or 8) channels?
So if you were to send signals out as quarter shapes... you would need 16 signals for 4 layers?
👋💖🟦 Plz do ALL levels. Subjective = good. Its fun.