This is so cool!! I was trying to put the first pathfinding script in my scrolling rpg, unforunatly, it didn't work. BUT with this tutorial you helped me with that, and it worked! Thanks!
There are some griffpatch projects that don't use "try move (dx) (dy)," and this comment implies they are not good, but, as we all know, all griffpatch projects are good, which means that this comment is false.
@@rAIGMIBU if all thumbs are fingers then all fingers are thumbs is the above statement true or false? just because I'm saying that use of the move dx dy helps a project be good doesent mean a project withought it will be bad use your brain before posting a comment, child
Griffpatch in 2050: Welcome to another one of my epic scratch tutorials, and today we will be learning how to make a tile-based scrolling game with 2 grid lists, one for tiles and one for pathfinding for some off-screen squishable floor-hugging goombas and a self-driving car with proper car physics trying to ram into you. You will get mystery objects such as guns in mystery boxes and also you will get bonus score if you shoot lemons and appels and easter eggs randomly appearing on the level and will get camera shake with flashes and sound effects, with half-tiles, slopes and cliffs.
For those of you who don't know this term: The Scratch "To-do List" stuff, is an example of something known as asynchronous programming. Scratch is heavily async.
I have another idea fot the place, where the moving directions are stored in the gridlist: at the end of "try moving in all directions", where dir is changed by 90, directly under this block place an "if" and check for and insidt this new "if" change "dir" to -90 this has 2 effects: 1. it now sets the directions the same way, you will find them, if you use the circle-control, that pops up at clicking the input-slot of a [point in direction (value)°]-block. pointing to the left on this circle sets to the value -90 2. it already prevents from storing values higher than 180 in the gridlist, so you don´t need to mod the value at replacing a "?" in the gridlist any more
instead of some youtubers that show you it, you explain it which i like because then i don't have to keep coming back to the same video! Keep up the great work!
Glad I could help - Although I would also like you to keep coming back hehe - to watch all my ads so I can make more videos lol. Guess I'll have to keep making new videos!
sometimes i do rewatch these videos when i am bored. and i remake the project that i did before and i try and challenge myself to think of what you are going to do next!
I love that, in the intro, you have Appel in the center of the maze instead of an apple to show that the project is more complex :) (can't believe I'm posting a comment that you actually have a chance of seeing lol)
Remember back on that maze generation tutorial, I asked how to use a pathfinder for it, and you said with lists? I had no idea how to do that, but this video REALLY helped! Thank you so much!
I combined them to a fully automatic maze solving simulator (just click green flag and watch) But I also have another maze project, that is controllable by arrow keys.
I am currently making a game that requires enemies to pathfind, and the tutorial was amazing, but I found one problem. I don't want all of my enemies following each other in a line, and I think the reason this is happening is because it takes priority of some directions over others, meaning that other directions that still cause the same speed to get to the object are excluded. I would like to make it so it records places that have two or even three directions that would cause the speed, and use a random number generator to decide which one to use. I would love if you could help figure out how to record when two directions would cause the same time to reach the goal. If you have any questions for elaboration please ask. Thank you so much in advance, I love using scratch and your tutorials are amazing.
Scratch Cat: i'm lazy i don't want to move Griffpatch: that's okay i can bug fix that Scratch Cat: NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!!!!!!!! Griffpatch: Bug Fixed! Scratch Cat: Well... Back to work...
holidays? you can go? 33:47 well u are lucky! for me here i can't even get out of the house... and that's why everyday i comment on griffpatch's videos.
I combined them to a fully automatic maze solving simulator (just click green flag and watch) But I also have another maze project, that is controllable by arrow keys.
Could you show us how something similar with a platformer so that it automatically completes itself in the fastest time possible? Could you do it for a scrolling platformer too? Ai building would be cool in so many applications here on scratch.
So I was gone for vacation, and I missed out on the previous 2 videos. The first thing that I did when I came back and took a shower was watch the two videos. Just saying how much we like your videos, griffpatch! P.S. 0:18 nooo why Appel?
This is why you make radial pathfinders. imagine a big 3D game engine like Unity or Unreal just going all the way across the map where the furthest objects are and just breaking the whole screen into little cubes
Would you mind giving me a rundown of how those work? I wanna see if this kind of thing can be translated into scratch and whether it would be more efficient than this method. This method is still on the basic side as it's meant to be understandable by newer scratchers so there are limitations in terms of speed and more advanced movement like diagonals. I want to see if this method fixes or at least make less of the problems.
@@cuttamOWU Hoh I...think I kinda get the idea probably not the code however (I hadn't started programming such a thing yet though) Oh well I'll just go through the logical procedures I think For the mapping, I guess you have layers of circles radiating from a point, and then some walls defined as being at certain angles. Now, for now I'm just going to assume you're in a circular maze with straight edges and corners. You should always cling to the shorter side of the arc until it's time to cross over to a corner. When you're approaching a corner outward, you want to aim for it when you're at a tangent on the arc that goes to that corner. When you're approaching a corner inward, just keep clinging to the short arc until you get to the corner. When you're at a "straight" corridor radiating from the circle (the walls are probably slanting toward each other near the middle), just cling to the edge or go the diagonal to the other point you want to go. Yeah I need a diagram to explain it, had to picture it in my head as I wrote it and I don't know if you can do the same but hey, this is how I'd do it If you're just having general blobs and not a strict sense of radial space, just do this method with more and more tangents and maybe use Dijkstra's algorithm if you need to
@@FireyDeath4 Yeah you might need a diagram. Right now my idea for pathfinding is set up a bunch of nodes and find which combination of nodes result in the shortest path using Euclidean Distance. There are manually-placed nodes I place as a checkpoint where the enemy has to pathfind to eventually and computer-placed nodes that the computer places, adjusting for changes in enemy location in the case the enemy or goal are shifted or the current path becomes obstructed. The pathfinding itself is simple but idk how to set up the computer-placed nodes. Right now I'm thinking of using some sort of ray tracing/marching to find the walls and directions it can't go to. Then I could set a minimum distance between the enemy sprite and the obstacles to limit wall-hugging. From there I would move the tracker towards the goal, moving left or right accounting for walls and minimum distance. And then stop after a certain length and repeat the process for as many times as needed. The stopping length is determined relative to the walls as I now have the direction of every wall point as well as the distance of each of those points from source. My ideal code would find the closest wall segment (segments if they are connected in the line of sight), create an arc where moving in a straight line within that arc will collide with a wall, move towards the goal, and move to the points on either side of the wall(s) in the case that the direction to the goal in within the arc, moving the distance the points on each side are from the origin of the tracker. This ideal code falls apart however in a maze where the walls are connected to one another (will see almost every direction as immovable) unless I can find a way to filter out certain wall segments from others. Keep in mind that I'm not using other sprites or clones, I have to figure this one mathematically. I don't want to completely remove the connected wall segment functionality since without it, the code might break (ex. /\ where the left side is the closer wall segment; If I were only to detect the left wall segment, the pathfinding will move itself into the space between the walls and make the right side path always appear longer because of a detour.). There may be other methods for this distance finding made for maze-solving but I don't know how to make them. This problem might not exist though if I'm just being dumb and the ray tracing/marching solves it for me. Idk how efficient this method would be compared to your method or even the method in this video. All this computing may take longer than both of those methods, especially from the ray tracing/marching, depending on how I implement it. Please give me thoughts on the manner since I kinda have no idea what Im talking about for some of this.
@@cuttamOWU Yeah I think you need a diagram too I was barely paying attention in fact lol (don't think raytracing is efficient though - pure geometry is better) Well, if you can calculate tangents, I think you can just cling to walls by changing your angle from the middle and use radians or whatever. If the sprite's a certain size you don't want to have _in_ the wall, I guess you could add in the walls, and then calculate a map whose edges are the sprite's diameter apart, the sprite's radius away from the original walls. Basically compensating so you can calculate using only a point. I just proposed you'd sort of code in the maze, based on where the arcs are (how far they are from the middle), the degrees they occupy between gaps, and any straight lines and their angles and lengths. Then use that.
@@FireyDeath4 Ray Tracing was just the first thing that came to my mind. I just need some code to figure out the direction of a wall and the distance that wall is from the origin. And I needed the returned data to be a line from the origin's perspective. Even then, the node part is pure speculation since I still feel that there will be problems with my theorized method. Mostly coming from the possibility that the returned paths being suboptimal or being too optimal that the turns are unnaturally sharp. We probably should move this conversation elsewhere where we can express our reasoning in the form of diagrams since words don't seem to cut it.
Hello, Griffpatch! I have been working on some code in the Tile Scrolling Platformer for partial blocks, like slopes, but with more flexibility than one slope. It works kind of like a tile grid for each block. It isn't fully finished, but I think it's enough to build off of. If I could share my code with you so you could see if it would be helpful for the tutorial series, that would be great. I have been using your series to develop my own game, so I thought by making some code I could possibly help others like you helped me. Thanks for your time! -Uboiscratch
at 21:14 I checked hundreds of times wether my code is correct I even started from the beginning but I can't make the arrows to go where it should it's just one have of the screen is random arrows everywhere and the other have is just nothing can you please help how to fix it?🥺
I combined them to a fully automatic maze solving simulator (just click green flag and watch) But I also have another maze project, that is controllable by arrow keys.
whenever I create the arrow grid at 21:31 my project lags so much it's basically unusable, I have a pretty decent pc so I don't know what the problem is, do you have a fix for this? Edit: never mind I was just being dumb and put "if length of GRID = 0" instead of "if length of To-do index = 0"
I have an idea for a new pathfinding algorithm in scratch. We can use A* algorithm. Basically the pathfinding sprite moves in all 4 directions with each movement having a serialized Xposition|Yposition|Direction|Distance in a list(Distance = Distance From Target). When the orginal spirte comes back to its previous location, it serializes its Xposition|Yposition|Direction|Distance|Used. This |Used| Serial will server as a condition when a different Custom block is checking for the Serial Number with the lowest distance on the list. Then a different block will determine the X and Y position of the serial number with the lowest distance on the list. The Pathfinding Sprite will then go to the X and Y positon and do the same procedure. Note: Ensure that the 4-way split of the original sprite has stage collision detection, to stop serialization of the collided block. Hope this idea is of help to you.
My arrows do not want to follow the Apple. Sometimes they get stuck in corner or appear in only certain spots, instead of going throughout the whole maze. I did everything correctly, too.
Griffpatch in 2050: Welcome, fellow Scratchers to a _very_ recent tutorial of mine: How to make a real life portal. *Griffpatch talks on.....* *At the **_end_** of the video* : You get a real life portal in your room.
Hey! Hello there! i love your videos. but can you create a video where you optimize what you're trying to do at 28:42? For example, I think the tiles can be optimized in such a way that they are not in the X and Y axes, but in the form of walls (also in the form of circles in their directions). In this way, you will have a more optimized project, and since the direction of the tiles is round, the character does not move in a zigzag shape in diagonal places! think about it. good day!
Good plan! But... What if it never stops? A better approach then would be to incrementally change the pathfinding over a number of frames, and keep recalculating, but only trigger a new recalculation when the last one has completed?
@@griffpatch Hmm, that one could work, But maybe recalculate only when the last position has stayed for lke a few frames, maybe 1/10 or 1/5 of a second meaning about 3-6 frames (depending on your fps)... or is that basically what you said??? lol
ussally id just use blocks and glide blocks for path finding like maze runners and stuff but i can see problems where itll hit the wall and glide through anyways, Good Tutorial!
This guy has everything you need for a perfect scratch project!
Ya your right
True!
Yes
Could not be truer
@@SamFromSpaceOfficial No, Griffpatch is very honest.
I admire you, for being able to creat such algorthms in such a limited engine
I have long gone from the scratch community but this dude is still impressive.
This is so cool!! I was trying to put the first pathfinding script in my scrolling rpg, unforunatly, it didn't work. BUT with this tutorial you helped me with that, and it worked! Thanks!
You know it’s gonna be a good project when Griffpatch tells you to make the
“try move,dx,dy” block
So True
@@idiotic_idiots Yup
There are some griffpatch projects that don't use "try move (dx) (dy)," and this comment implies they are not good, but, as we all know, all griffpatch projects are good, which means that this comment is false.
@@rAIGMIBU if all thumbs are fingers then all fingers are thumbs
is the above statement true or false?
just because I'm saying that use of the move dx dy helps a project be good doesent mean a project withought it will be bad
use your brain before posting a comment, child
@@kosuken Yeah, you're right, sorry
Griffpatch in 2050:
Welcome to another one of my epic scratch tutorials, and today we will be learning how to make a tile-based scrolling game with 2 grid lists, one for tiles and one for pathfinding for some off-screen squishable floor-hugging goombas and a self-driving car with proper car physics trying to ram into you. You will get mystery objects such as guns in mystery boxes and also you will get bonus score if you shoot lemons and appels and easter eggs randomly appearing on the level and will get camera shake with flashes and sound effects, with half-tiles, slopes and cliffs.
lol
lol
Lol
lol
griffpatch in 20999:
welcome to a jetpack making tutorial ( no punctuation)
Yes, it's here, the long awaited, Pathfinding with Lists!!!!
I have seen you. You are a great scratcher!
@@2050CosmX Thank you!
For those of you who don't know this term: The Scratch "To-do List" stuff, is an example of something known as asynchronous programming. Scratch is heavily async.
so true!
I have another idea fot the place, where the moving directions are stored in the gridlist:
at the end of "try moving in all directions", where dir is changed by 90, directly under this block place an "if" and check for and insidt this new "if" change "dir" to -90
this has 2 effects:
1. it now sets the directions the same way, you will find them, if you use the circle-control, that pops up at clicking the input-slot of a [point in direction (value)°]-block.
pointing to the left on this circle sets to the value -90
2. it already prevents from storing values higher than 180 in the gridlist, so you don´t need to mod the value at replacing a "?" in the gridlist any more
Nice, a lot simpler than I actually imagined it, now something employing a system like this is actually possible for me :)
34:03 I think he's rocking out to the outro song LOL
@@Daydarkness0 yes
Because he used the sin and cos operators
instead of some youtubers that show you it, you explain it which i like because then i don't have to keep coming back to the same video! Keep up the great work!
Glad I could help - Although I would also like you to keep coming back hehe - to watch all my ads so I can make more videos lol. Guess I'll have to keep making new videos!
sometimes i do rewatch these videos when i am bored. and i remake the project that i did before and i try and challenge myself to think of what you are going to do next!
I love how he merges both the different projects! :)
Happiness be like:
"Grid List Pathfinding"
No Sarcasm intended, and I am dead serious again
you write this in every video.
@@jensenx08 because it's true
again
@Gary Lenoie LOL
I love that, in the intro, you have Appel in the center of the maze
instead of an apple to show that the project is more complex :)
(can't believe I'm posting a comment that you actually have a chance of seeing lol)
He he
Definitely the best Scratcher to follow. Thank you for all your work
Remember back on that maze generation tutorial, I asked how to use a pathfinder for it, and you said with lists? I had no idea how to do that, but this video REALLY helped! Thank you so much!
niceee! i was waiting for this sequel! great job!
Wow that one cat solving the maze at the end just blew my mind even knowing how it all works. Just so beautiful!
Yeah he he - I love watching him trundle around :)
He said 'So, get our your brooms... 13:14 ' And I immediately thought of Harry Potter
Griffpatch: these two videos together could be really powerful
Me: *hurriedly opens procedural maze generation project*
Minotaur time
same
lol
I combined them to a fully automatic maze solving simulator (just click green flag and watch)
But I also have another maze project, that is controllable by arrow keys.
Next up: combining the 2 videos and the videos about making a 3d game
I am currently making a game that requires enemies to pathfind, and the tutorial was amazing, but I found one problem. I don't want all of my enemies following each other in a line, and I think the reason this is happening is because it takes priority of some directions over others, meaning that other directions that still cause the same speed to get to the object are excluded. I would like to make it so it records places that have two or even three directions that would cause the speed, and use a random number generator to decide which one to use. I would love if you could help figure out how to record when two directions would cause the same time to reach the goal.
If you have any questions for elaboration please ask.
Thank you so much in advance, I love using scratch and your tutorials are amazing.
Finally! I've been waiting for months for this episode to come
I am gratefull for the fact that you even make videos, thanks
My pleasure!
@@griffpatch ✌️
Great video - I liked the explanation of it's stack based single threading runtime loop.
This will be really helpful for my RTS game I am making! Thanks!
Scratch Cat: i'm lazy i don't want to move
Griffpatch: that's okay i can bug fix that
Scratch Cat: NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!!!!!!!!
Griffpatch: Bug Fixed!
Scratch Cat: Well... Back to work...
it's a good thing I saw the pathfinding and grid list tutorials when they came out and made projects with them.
Yay!
Griffpatch:"Lets go scraching"
Me: Yea scraching my head how is this possible
just letting you know you forgot the 't' in scratching
I'm going to use this and the maze project we did to make an awesome game!
Griffpatch, I just want to say you something. You're amazing
:) - You are too kind!
@@griffpatch 🥰Aw! Thank you!
holidays? you can go? 33:47 well u are lucky! for me here i can't even get out of the house... and that's why everyday i comment on griffpatch's videos.
IKR... We are not going out much here... We'll just be visiting family which will be super nice.
The apple trying to hide itself in a maze, but still being found and eaten: WAIT WHAT
been wanting to make a tower defence game similar to field runners, this is perfect!
Ha - interestingly enough, that is exactly the game that inspired this project in the first place :)
@@griffpatch what a coincidence
I would love to combine this with the maze generator stuff. this is so amazing
I combined them to a fully automatic maze solving simulator (just click green flag and watch)
But I also have another maze project, that is controllable by arrow keys.
@@DrachenbauerHave you shared it? I'd love to have look!
This is perfectly what I needed!
I added this path finding script to enemies in my ray-casting project and it works much better than just point to player and move so thanks
Nice! I was really excited for a sequel to the last part!
Hope you enjoyed it!
You know a video is going to be good when the username is Griffpatch
Wow the path finding is amazing!!
Appel in the start made my day!!
33:43 yes, I was starting to wonder if you weren't an alien to get to make so many videos 👽🤯😂
i am now using javascript after years of scratch but this is stil so helpfull!! thank you
I wonder if TurboWarp can make the tiny tiles pathfinding run smoother...
Scratch Addons with 60fps mod
@@stormer4859 That still doesn't compile the project which is the main source of the speed
@@cst1229 griffpatch: set it to 240 fps other people: griffpatch computer is very good
Griffpatch is my no. 1 fav utuber I luv him so so much
You are too kind :)
0:52 that last cat that went the right direction...
Thanks alot. This is gonnna be the base for a point and click adventure I am currently working on :)
I love it and I will combine the maze generator and the grid list pathfinders lol
OMG this tutorial is awesome 🤯🤩 !!!! I have wait a long tome for this 🌟🌟🌟🌠🍋🍋
I hope you can write posts again soon 😊👍
Oh sob... My posts!
Finally, a Update for our pathfinding! Yay!!!
Mixing this with the maze generator video can make a cool game
Could you show us how something similar with a platformer so that it automatically completes itself in the fastest time possible? Could you do it for a scrolling platformer too? Ai building would be cool in so many applications here on scratch.
So I was gone for vacation, and I missed out on the previous 2 videos. The first thing that I did when I came back and took a shower was watch the two videos. Just saying how much we like your videos, griffpatch! P.S. 0:18 nooo why Appel?
Everybody Gangsta until Pathfinding using Lists
0:18 i think you mixed apples and appels
This is why you make radial pathfinders.
imagine a big 3D game engine like Unity or Unreal just going all the way across the map where the furthest objects are and just breaking the whole screen into little cubes
Would you mind giving me a rundown of how those work? I wanna see if this kind of thing can be translated into scratch and whether it would be more efficient than this method. This method is still on the basic side as it's meant to be understandable by newer scratchers so there are limitations in terms of speed and more advanced movement like diagonals. I want to see if this method fixes or at least make less of the problems.
@@cuttamOWU Hoh
I...think I kinda get the idea
probably not the code however (I hadn't started programming such a thing yet though)
Oh well I'll just go through the logical procedures I think
For the mapping, I guess you have layers of circles radiating from a point, and then some walls defined as being at certain angles.
Now, for now I'm just going to assume you're in a circular maze with straight edges and corners. You should always cling to the shorter side of the arc until it's time to cross over to a corner. When you're approaching a corner outward, you want to aim for it when you're at a tangent on the arc that goes to that corner. When you're approaching a corner inward, just keep clinging to the short arc until you get to the corner. When you're at a "straight" corridor radiating from the circle (the walls are probably slanting toward each other near the middle), just cling to the edge or go the diagonal to the other point you want to go.
Yeah I need a diagram to explain it, had to picture it in my head as I wrote it and I don't know if you can do the same but hey, this is how I'd do it
If you're just having general blobs and not a strict sense of radial space, just do this method with more and more tangents and maybe use Dijkstra's algorithm if you need to
@@FireyDeath4 Yeah you might need a diagram. Right now my idea for pathfinding is set up a bunch of nodes and find which combination of nodes result in the shortest path using Euclidean Distance. There are manually-placed nodes I place as a checkpoint where the enemy has to pathfind to eventually and computer-placed nodes that the computer places, adjusting for changes in enemy location in the case the enemy or goal are shifted or the current path becomes obstructed.
The pathfinding itself is simple but idk how to set up the computer-placed nodes. Right now I'm thinking of using some sort of ray tracing/marching to find the walls and directions it can't go to. Then I could set a minimum distance between the enemy sprite and the obstacles to limit wall-hugging. From there I would move the tracker towards the goal, moving left or right accounting for walls and minimum distance. And then stop after a certain length and repeat the process for as many times as needed.
The stopping length is determined relative to the walls as I now have the direction of every wall point as well as the distance of each of those points from source. My ideal code would find the closest wall segment (segments if they are connected in the line of sight), create an arc where moving in a straight line within that arc will collide with a wall, move towards the goal, and move to the points on either side of the wall(s) in the case that the direction to the goal in within the arc, moving the distance the points on each side are from the origin of the tracker. This ideal code falls apart however in a maze where the walls are connected to one another (will see almost every direction as immovable) unless I can find a way to filter out certain wall segments from others. Keep in mind that I'm not using other sprites or clones, I have to figure this one mathematically. I don't want to completely remove the connected wall segment functionality since without it, the code might break (ex. /\ where the left side is the closer wall segment; If I were only to detect the left wall segment, the pathfinding will move itself into the space between the walls and make the right side path always appear longer because of a detour.). There may be other methods for this distance finding made for maze-solving but I don't know how to make them. This problem might not exist though if I'm just being dumb and the ray tracing/marching solves it for me.
Idk how efficient this method would be compared to your method or even the method in this video. All this computing may take longer than both of those methods, especially from the ray tracing/marching, depending on how I implement it. Please give me thoughts on the manner since I kinda have no idea what Im talking about for some of this.
@@cuttamOWU Yeah I think you need a diagram too
I was barely paying attention in fact lol (don't think raytracing is efficient though - pure geometry is better)
Well, if you can calculate tangents, I think you can just cling to walls by changing your angle from the middle and use radians or whatever. If the sprite's a certain size you don't want to have _in_ the wall, I guess you could add in the walls, and then calculate a map whose edges are the sprite's diameter apart, the sprite's radius away from the original walls. Basically compensating so you can calculate using only a point. I just proposed you'd sort of code in the maze, based on where the arcs are (how far they are from the middle), the degrees they occupy between gaps, and any straight lines and their angles and lengths. Then use that.
@@FireyDeath4 Ray Tracing was just the first thing that came to my mind. I just need some code to figure out the direction of a wall and the distance that wall is from the origin. And I needed the returned data to be a line from the origin's perspective.
Even then, the node part is pure speculation since I still feel that there will be problems with my theorized method. Mostly coming from the possibility that the returned paths being suboptimal or being too optimal that the turns are unnaturally sharp.
We probably should move this conversation elsewhere where we can express our reasoning in the form of diagrams since words don't seem to cut it.
Hello, Griffpatch! I have been working on some code in the Tile Scrolling Platformer for partial blocks, like slopes, but with more flexibility than one slope. It works kind of like a tile grid for each block. It isn't fully finished, but I think it's enough to build off of. If I could share my code with you so you could see if it would be helpful for the tutorial series, that would be great. I have been using your series to develop my own game, so I thought by making some code I could possibly help others like you helped me. Thanks for your time! -Uboiscratch
Woah, looks like something he would use in his RPG tutorial. Did you do those tile pins or quarter tiles? I'm interested
Wow, this looks good
I KNEW THIS WAS GONNA POP UP!!! IT WAS A MEMBER ONLY IN THE PLAYLIST! :D
How did you know if ur not a member?
@@sbfshadoe9918 the playlist shows everything in it, even the member only videos
Amazing followup!
"the maze is too op"
lmao
that's a quote lol
finally, i love your vids
maybe you could perhaps implement this pathfinding into the raycaster for enemies
The Start Seems So Excited 😆
Advanced Scratchers in scratch 10.0: testing the newly added 3d mode code
I never knew that Scratch had a to do list! Wow! 😀
at 21:14 I checked hundreds of times wether my code is correct I even started from the beginning but I can't make the arrows to go where it should it's just one have of the screen is random arrows everywhere and the other have is just nothing can you please help how to fix it?🥺
Great! I'm gonna try to combine this with your maze generator that I made :)
Sounds great!
I combined them to a fully automatic maze solving simulator (just click green flag and watch)
But I also have another maze project, that is controllable by arrow keys.
AWESOME as usual!
Thanks again!
no one:
griffpatch:
"Draw a rectangle"
*Proceeds to draw a square*
Boids + Pathfinding + Grid list = very good project
THANK YOU GRIFFPATCH
I love your tutorials!
Thank you so much!
@@griffpatch You're welcome.
Griffpatch: "simple as scripts"
Me: I don't even know what you even mean by that.
Lets goo exactly what i wanted!
nice tuturial and a good idea
YEAH!!! Now this is exciting!
Could you somehow make it so that the maze could work with more than one apple? I wish to make a light flow system so instead of apples i make lights.
whenever I create the arrow grid at 21:31 my project lags so much it's basically unusable, I have a pretty decent pc so I don't know what the problem is, do you have a fix for this?
Edit: never mind I was just being dumb and put "if length of GRID = 0" instead of "if length of To-do index = 0"
i swear this guy is going to teach us how to make a fully functional alexa with lists
lol
is there any way to make it so the cat can pathfind there way to a moving object without that much lag?
This guy is the Mumbo Jumbo of scratch
I have an idea for a new pathfinding algorithm in scratch. We can use A* algorithm. Basically the pathfinding sprite moves in all 4 directions with each movement having a serialized Xposition|Yposition|Direction|Distance in a list(Distance = Distance From Target). When the orginal spirte comes back to its previous location, it serializes its Xposition|Yposition|Direction|Distance|Used. This |Used| Serial will server as a condition when a different Custom block is checking for the Serial Number with the lowest distance on the list. Then a different block will determine the X and Y position of the serial number with the lowest distance on the list. The Pathfinding Sprite will then go to the X and Y positon and do the same procedure.
Note: Ensure that the 4-way split of the original sprite has stage collision detection, to stop serialization of the collided block. Hope this idea is of help to you.
33:43 look at that one cat going back and forth at the bottem of the maze
The one dislike is from the cat that got stuck in the ending
Poor cat
I am looking for a light flow script and this should work nicely
My arrows do not want to follow the Apple. Sometimes they get stuck in corner or appear in only certain spots, instead of going throughout the whole maze. I did everything correctly, too.
Your corridors are too small
me: hears “now we can have a moving target”
me: enemy pathfinder to player go burrrr
now to do it!
@@shoeshoe6837 i tried this too for a game i'm making but the tiles appear very slowly, i don't know how to fix this
Griffpatch in 2050:
Welcome, fellow Scratchers to a _very_ recent tutorial of mine: How to make a real life portal. *Griffpatch talks on.....*
*At the **_end_** of the video* : You get a real life portal in your room.
Can this be used in your 3d project
Enjoy your holidays, Griffpatch! Do send postcards!
Hey! Hello there! i love your videos.
but can you create a video where you optimize what you're trying to do at 28:42? For example, I think the tiles can be optimized in such a way that they are not in the X and Y axes, but in the form of walls (also in the form of circles in their directions). In this way, you will have a more optimized project, and since the direction of the tiles is round, the character does not move in a zigzag shape in diagonal places! think about it.
good day!
I wonder how powerful the combination of:
•Random maze generator
•How to use tile grids
•Simple pathfinding
Will be...
He he - These were designed to be base projects to be built upon :)
@@griffpatch now that I can talk to you...
Can you make a snake game tutorial first with clones, then with grid lists and finally with scrolling?
My arrows are only updating when I move the apple and then click the screen do you know what is happening?
You don't have to rename the maze, since in the grid list sprite it's only 1 instance of
I think I know how to reduce the lag. Make it so that the tiles only reset after the apple stops dragging, maybe use lists or two variables to check?
Good plan! But... What if it never stops? A better approach then would be to incrementally change the pathfinding over a number of frames, and keep recalculating, but only trigger a new recalculation when the last one has completed?
@@griffpatch Hmm, that one could work, But maybe recalculate only when the last position has stayed for lke a few frames, maybe 1/10 or 1/5 of a second meaning about 3-6 frames (depending on your fps)... or is that basically what you said??? lol
Thanks a ton! I was trying to make sense of this method on the Wiki, but couldn’t work it out 😁
the last cat is just dancing
so good i love your tutorials
ussally id just use blocks and glide blocks for path finding like maze runners and stuff but i can see problems where itll hit the wall and glide through anyways, Good Tutorial!