Thanks! I don’t know why no one looks at the x bias. Most cover the Y but I haven’t seen one actually make it fully ‘random’. That was one of the main reasons I made this actually haha, Glad it helped!

@@Winterdev Hey I was watching Screamingsnake’s falling sand video and while he doesn’t explain it in the video, I checked his code and he is using an interesting method you might be interested in, because it skips having to have cellmoves or two buffers. Instead of compiling and shuffling CellMoves, he generates a randomly shuffled array of indexes ranging from 0 to the chunkWidth, and loops through the map or chunk like that row by row. It eliminates having to create a CellMove struct or class and shuffling those afterwards. However it does lead to step stair formations at chunk edges, but those can be resolved by randomly offsetting the thread/chunk matrixes each frame.

Oh that’s interesting so you basically make an iteration map before hand. I was thinking about trying to do this on the GPU to see how fast I could get it to run, that might be the way to do it there. I going to look into that cus I need to use CUDA for a new job so maybe this will be a cool way to train up. Thanks!

Thanks! Next one is going to come out soon. With school starting I am a little delayed though :(

5:52 I recommend always iterating through an array/vector in reverse when you want to remove items from it. :-)

Thanks! I wasn't sure this one was gonna be as clear as the others, glad I managed it :)

spread the word ;)

Thanks! I watched your vid when researching mine. It's funny how may different ways there are so solve the same problem. I bet the best way to do it is with double buffering + reversing the dir. I was trying to look into running it on the GPU, which I you would have to double buffer for, hinting that's the best way. If I had more time I'd play around with it, but now I'm trying to make a lil game with it, so I called it done.

@@Winterdev Your solution seems to working just fine, so I say run with it. Looking forward to what you come up with.

@@johnjackson9767 Ye for sure I like messing with the tech too much lol. Should be something cool in the next few weeks or so...

Thanks for the comment, I'm excited to see if it will :)

Ok I'll start on that it might take a while through

It allows for some cool effects! good luck with your game

Thanks!

Yeah I was focusing on the engine bits, for making a game with it I am looking at a bunch of different movement properties. I got this i.gyazo.com/3760bbd2b0d41794376463aed0cab5ba.mp4 by checking 3, then 2, then 1 away for water. I think you would do that with a velocity and acceleration for gravity.

@@Winterdev that looks better yeah! have you thought about implementing a fluid simulation alongside with something like FLIP or SPH? Something that bothered me about these falling sand games is that the water never looks realistic... that might be a me problem though

@@qetu10 That would be really cool. I am going to look into that haha! You just run a set of eqs. right? I bet that wouldn't actually be too hard with the fixed grid, I'll see. It would be really cool to see a splash/pressure in a pipe.

@@Winterdev good luck with your research, will be better than my attempt in any case - the equations all go over my head lol

Or cuda depending on the algorithm and data structure

That's another way to do it, and what I did first, but to get real independent ordering you need to consider all the possibilities before applying them. It's a little less clear than I would like but what you suggest is how the Processing version works, and the C++ version is just another way that truly doesn't have any bias to the order.

this method takes a lot of cpu's power

@@player210 Yea I wish that sort didn't need to happen, in the next part I am going to look at the performance, as I am going to expand the size of the world

@@Winterdev i'm really excited about second part!

Coming soon! :)

That was a little confusing but I was just trying to compact it away. In c++ you can declare a function before implementing the body so it’s implied that it gets moved. But I don’t think I ever said that I was just trying to make room for the next code.

@@Winterdev I see, thank you very much for your quick response. Great work. All the best.

@@Winterdev can u tell me which java scripting application u were using?

also there were many mistakes that where red in the code how do i fix those?

Thanks for the comment! There are some engine pieces that I didn't include for the c++ version. You'd need to be able to run some graphics library that lets you draw pixels to a texture. That could get a little complex if you 'cant code at all', so I'd start with the processing version as all the tools are built in to the program. Here's the download processing.org/download/. If you write the java code from the video it will work. You can also copy it from the article blog.winter.dev/2020/falling-sand-games/. Good luck :)

@@Winterdev ok thx

Thanks! Do you mean water cells? The rules are at 2:32

​@@Winterdev yes. so if it can go down it will but if it cant then where would it move? is it exactly like sand but to the left and right?

@@drollestprawn Yes, you would just not add 1 to the Y. So (x, y + 1), (x - 1, y), (x + 1, y)

@@Winterdev ok thanks

Thanks for noting this, It’s fixed! been super busy hadn’t looked at it

same for the function "isEmpty"

Hmm either they are not defined or maybe try putting them above the loops? They are my functions btw not included in processing

try just "set"

You can use this library github.com/nothings/stb/blob/master/stb_image.h and paste it in like the tiles. If you want to rotate it tho you need to render it with a software renderer into the sand world or a regular sprite on top of the sand

It’s this thing called processing, it turns Java into a kind of scripting lang almost. It comes with an ide called processing? It’s not so good but has some interesting features like you can edit variables live with sliders

@@Winterdev I'm trying to write the same on cpp, using direct2d gpu but in this case I need to write a lot of stuff like wndclass and window creation, wndproc, handling messages, implementation of devices for direct2d. That's interesting, but I whant to concentrate on code and when I see such videos as yours I am asking myself how to do that so easily? Also there are questions about solving particles movements, i meen how to render them with high fps in case its more then 1000 of them. I have tried to make quadtree and that helped a bit, but I sow videos with thousands particles collided and rendered with high fps. Where should I store them: lists or massive?

I see no comment held for review or whatever so idk, what were you trying to say?

@@Winterdev in saying that i commented in your video before this but is gone

@@lourdthebluefoxie yeah what was the comment? maybe it'll work as a reply?

what do you not like? Code is very readable.

@@theitatit mainly all the extreme amounts of excess spacing to make the things line up, like the if statements at 2:17 are a perfect example