Hello! I saw your parallax cave effect. I just wanted to tell you that you can actually add swingin hanging wines . Or stalagmites etc. That would really cool!
Thanks. :) It's what I'm working on currently. Trying to get my world building tools built so I can start prototyping the first dungeon which should hopefully be soon.
Game dev tool making can be even more interesting than making the game itself. Very cool video. I think with more and more "real" applications made with Godot, plus the publicity it will get after version 4 is released... might just be enough to make the world realize it's potential.
I enjoy your style of development; Very few devlogs on youtube actually talk about the custom tools they create to speed up development, even when it's essential for content creation in-game.
Thank you. :) It's part of being a developer. I imagine others would rather talk about the game itself, but with my devlogs I'm leaning a bit more into the fact that it's mostly developers / aspiring developers that watch these. So I try to cater my content more towards them. Especially those that are interested in knowing what I do as a developer beyond just adding game mechanics.
@@GameEndeavor I think it's a smart decision, and being specific about your audience (even if it might be a more niche) is tons better than just screaming about what you make to everyone, with no regards for who might be listening. Keep at it! 🔥
This is awesome! I love that you are hand building along with PCG. I think that PCG really shines in giving people a starting point to build off of. Game is really looking good!
Hi, here's a bit of feedback/idea: When you defeat an enemy (at the last hit), the corpses could get knocked back a little, because that would give you a lot of game feel/ juice. A jucier game, is usually a better one. Keep up!
The editor is becoming super interesting. But imo it's kinda lacking until you can press a button to create the whole game. /s the game is looking really cool!
I use pixelorama for my game development, I recognized it right when you showed it, immglad the it is getting more recognition, it is a really good pixel art program
You have some of most in depth dev logs on TH-cam. You do an excellent job describing your process and the discoveries you're making. Always very inspiring!
I wouldn't say unnecessarily. It saves me so much time building these shapes by hand, which I hate doing. Before this, it was my biggest time sink, having to place every piece of grass. This gives some really interesting shapes too, especially at larger sizes. And the whole thing is less than 75 lines of code and is easily reusable with just a couple lines of code. :)
I've used the cellular automata method to make procedurally generated caves before but it would often generate several unconnected caves that the player couldn't reach. I fixed this by adding a flood-fill like method to identify and calculate the volume of each area of the cave system and then only keep the largest section for my final map.
lol, it's hard to get good b-roll footage when you're saving the prettier parts of development until after you have the mechanics you want. However I'm about to enter the polish / world building phase. So that'll be fun I think. :)
I really enjoy the art of randomization in games. It's such a treat for the curiousity! I totally agree that hand crafted areas tend to look much better, but for my game, I wanted as good of a result that I possibly could using randomness for the heavy lifting. I ended up chooseing to build pre-designed level sections which can connect together randomly based on some rulesets. When I'm further into the development, the content of these pre-designed pieces will be randomized to. Perhaps a event quest spawns. Or a pack of monsters. If it sounds interesting you can read more about it on my blog =) www.ledii.net/blogs/game-dev/mining-game-v2-randomized-caves
Maaan, I just found out you were making this game. Looks awesome!!! And close to what I would like to make. Few things I thought about: - will the character always look like this? He's kind of cute, but this little fellow is far from heavy enough to make such a cool stomp if you ask me. Maybe if you become a warrior you get some extra beef? Also - armor etc? You do visualize the weapon... :D - I love it how you try to make it balanced and interesting to play as different classes. Though I noticed - the warrior is vulnerable to all attacks (skeletons/melee, those brown things with range attacks) and rogue would only be vulnerable to ranged attacks (since skeletons need to catch you to hurt you). Maybe add more twist to the rogue fighting style, like some enemy homing projectiles you need to shoot down before you can hit the enemy. Now it looks like it's fairly easy to just avoid the particle trajectory. Sorry for the long comment. I'm super excited and can't wait to play this. Keep it simple as much as you can, so that you can finish it soon (and not take years like most of us). Also, on the contrary, I hope to see LOTS OF LOCATIONS, LOTS OF ENEMIES, LOTS OF ITEMS, LOTS OF TOWNS, LOTS OF STORYLINE, basically Diablo 2 all over again. Thanks. I love your vids, by the way. So tight, yet easy to follow. Keep it up, you're awesome. :) Big fan!!
Howdy! Your enthusiasm for the project is astounding, thank you. :D There will be character customizations, possibly some heavier armor for melee class. Not a different character base though as that's a whole lot more work. The combat is still very much wip, I haven't even tried balancing the different classes yet. Mostly adding features I want currently. This project will take me years probably. xD With it being so sentimental to me, I want to take my time and make it as good as I had dreamed about. However I'm not really a fan of the dungeon crawler style of gameplay of Diablo 2. :) This is a much more mellow experience. And currently there will only be one town. Subject to change, but the idea being that I want to focus on the NPCs, and would rather have limited memorable NPCs rather than a bunch of towns.
This is so relatable: th-cam.com/video/AAAmQKgHsDg/w-d-xo.html "This can even make it difficult to write the scripts for these devlogs, because it's almost like rubber ducking the logic for my game. I'll realize that I can do something differently, or get an idea that I start writing code for, and the next thing I know it's 11pm at night and I only have two minutes worth of script, but a useless feature that's fun to play with." I started working from home a few years ago (my job decided renting an office space wasn't worth the cost since our business is all online) and so often I go to report on my progress and realize as I describe what's been happening that I need to change things. I end up scrapping the report and working for hours more on the next part and never report nearly as much as I should.
Spending the time to make tools to make your life easier is something I respect. Especially when you need to weigh the options of how long will the tool take me to make, vs how much time will I save using the tool?
tbh there wasn't much to weigh. It has already paid for itself. It used to take me forever to build areas. That's why the first devlogs feature one area. Grass decorations were the most time consuming. Now I can build levels fast enough to make one per devlog, just to keep things fresh and practice my building skills.
Man I wish you would add like a rare type of flower, like a black/red/purple flower. Or even more flowers in general, the world looks great ! but missing some life, maybe leafs falling when you enter a tree and stuff. just an idea though, keep it up !
I'm just now starting on the world. :) Everything up until now has been mostly features and tools, with the bare minimum of art needed to make pretty devlogs. If you think the world looks great now, wait 2 months. :)
Method as in combining procedural generation with manual world building? I wasn't aware, but that's interesting. I'm surprised I haven't heard more about the technique tbh, it's already helping me a lot.
The game looks amazing, all the work you put into combat has paid off since it’s so satisfying to watch! Quick question, have you thought about making your level editor feature available to others? Either for a price (if within godot’s terms of service) for some passive income or just as a handy dev tool for other people to use. Whatever the answer keep up the awesome work!
Any particular reason you're using cellular automata for the grass patches? Wouldn't just making an array with random grass tiles work just as well? Feels like overkill just to generate some random tiles lol. But hey, if it works, it works. :^)
Because it makes clusters as opposed to noise. The blob shapes are much more aesthetically pleasing imo. Plus it's there (for caves and whatever else I'll need). Not really overkill to use what's laying around. It's just a couple lines of code to generate the grass clusters.
my brain : SUBSCRIBE TO HIM NOW me: why ? I just discovered his channel today ! my brain : YOU LITTERALLY WATCHED 11 OF HIS VIDEOS IN A SINGLE EVENING me : yeah I guess you might be kinda right *+1 subscriber*
Local co-op is planned with online multiplayer being a possibility later on if it doesn't bog down development too much. But local co-op would atleast make steam's remote play something to look forward to.
Is the cellular automata live stream somewhere? Would love to see that one EDIT: amazing how the script and tools are looking like a separate stand-alone application. Didn’t know about those endeavors using Godot as the underlying engine.
So procedural generation is like asking ai to build a basic layout of Mario maker, then you go in and change what you need to add polish. Now I wonder if I can ask Chat GPT4 if they can make my game, because I know I can't.
This didn't use AI for this, I created a simple algorithm. And typically with procedural generation you don't touch it up after. I just used it to automate a lot of tedious work. Like I use it now for placing grass decorations or placing trees and rocks. You probably could get it to create these algorithms for you, but I don't think it's at a point where it can make a professional quality game yet. It seems to have a lot of errors and its output is mediocre. You'd actually have to know what you're doing just to fix the errors. For example I asked it to create a Binary Space Partition thingie and I had to go in behind it and fix its errors, and could only do so because I knew what the logic was supposed to do.
@@GameEndeavor yeah I misspoke with ai and procedural generation. But it sucks that I seem to have all these ideas for games but can never seem to make even something that room with enemies spawn all while following all the guidelines and my PC crashes. Sold my PC 3 years ago and don't know if I want to try again with Chat GPT4.
Hello, great job ! I have a small question: how do you manage the transparency of the trees when the character is behind? I really have difficulties to reproduce this kind of display (ideally I would like to display my character in transparency behind but I can't find any info about how to manage the Z-test or manipulate the depth map in shader for 2D objects) ? thanks in advance
I put an area on the tree that detects the player and adjusts it's alpha values. I've been looking for a way to render the player's silhouette too but yeah, no luck here either.
Proper procedural generation produces both handcrafted level worlds AND requires a lot more work. PCG is not a short cut. That is why it always sucks. Look at a game like Unexplored. Its levels are better than most hand crafted dungeons by a fairly large margin. Why? Because the devs put the effort into having the generator use actual level design philosophy via its cyclic generation. It wasn't treated as a short cut and its levels are hand crafted quality, often times better than most devs ability to design, and infinite at the same time.
Pfft, all developers are lazy. If it can be done faster it will, and thats more efficient. Its not always easier to write a month worth of code but if it means you don’t have to spend an hour a day on that one thing you’d do it. Engineers, and software engineers are just people who solve problems. The laziest people on the planet who will do everything to have to do nothing. Its counterintuitive but it makes people great at being motivated.
Hi, I want to ask something, I'm currently developing a game in a perspective like this with pixel art, but I'm having a problem with shadows. I see that your semi-transparent shadow sprites can come on top of each other but not create a darker spot, which how it should be. How did you achieve this? I looked at your prior devlogs but couldn't find it.
Yeah, shadows are quite tricky. Most games don't even fix it from what I've seen. I fixed it with a bit of shader magic. idk if it'll work outside of Godot, but with Godot shaders if you get the screen pixel that's rendered under the shader, it doesn't seem to account for other shaders on the same canvas layer. So I'll use a shader to place the shadow by multiplying the screen texture with a shadow color. Then when an overlapping shadow renders over top, it'll perform it as well, but will ignore the shader under it. This is highly experimental and probably prone to some breakage.
I considered that, but I have a lot of TileMaps and I wanted an excuse to learn more about Godot's source since I'll probably be faffing about more in there later. I'm not even using the custom engine atm, was just a fun endeavor.
Have you ever had a bug where godot stops importing assets? Aside from redoing my project and reimporting i have no idea how to fix this and google is just telling me this is some unsolvable bug with godot. I hope there is a solution because at this point I likely need to move on to unity or try to make my own custom engine using c++, which is a shame because I was starting to like Godot.
If enemy pathing is working in your game? Then could you possibly automate more of the world / cave gen by having the enemy pathing AI determine if the cave paths are actually explorable / reachable? Possibly it’s easier to do it yourself, but if you can have this running on 100 generated maps at once and then show you a list of only the ones that are at all useful in a baseline sort of way, then it could speed up even the point at which you start, or even the point at which your players start. Even could extend into some sort of accessibility function that lets players know whether or not all areas in their custom maps are reachable. I’m not a programmer at all, so maybe this is all ridiculously stupid. Good job regardless!!
This really reminds me of a machine learning paper where the researchers train a neural network AI to modify initially randomly-generated maps into playable ones, iterating tile-by-tile, and being rewarded according to the playability & difficulty of the final map, among other quality metrics. Basically a gamification of map design, that an AI can understand & learn to play arxiv.org/abs/2001.09212 Yannic Kilcher also made a nice vidoe summary of it. Your method is basically a human playing it in place of of the AI :) with the human's gameplay design skill also providing the quality metric, of course On a side note, have you explored/tried out cyclic dungeon/quest generation?
I had never thought of that. I'll have to take a look at what kinds of maps they designed because that sounds interesting. I read an article on cyclic dungeon generation, but never attempted it myself. I haven't dabbled too much in random generation beyond an arpg prototype, because I often find pcg to be bland. But if I were to make pcg dungeons then I would likely go with a cyclic design. :)
You are not lazy, you just want to work in a fast and efficient way. I hate it when professional call themselves lazy instead of saying that they are good.
lol, it's just an alternative way of saying it. It sounds pretentious and boring to say "I like to work fast and efficiently", so instead I say "I'm a lazy developer". Fits my personality better.
@@GameEndeavor Well, I'm having troubles doing anything, so it bothers me when people who are clearly skilled say they are lazy. I won't complain again, I promise, stay true to yourself, that's how we like you anyway.
There is a thought fallacy that IMO is the reason why randomized levels are often not as interesting as hand-crafted ones. Level randomization is not less work than hand-crafting a single map. Some devs treat it the other way around, so they don't have to do level design. But good games require level design that is fit to the game's mechanics. There is no way around it. There are very few examples where randomization works but there it is essential: "Don't Starve" and "Spelunky" come to mind Good level randomization requires to hand-craft maps to see what works and what doesn't and itereate upon that. Only then you could actually tweak the generator to create maps that are very similar to that and even then to make them interessting they require "set-pieces" i.e. handcrafted parts of the map that add diversity and interesting stuff (e.g. the Pig-Village in "Don't Starve" or Idol-Head traps in "Spelunky"). And last but not least level randomization in itself should be a deliberate design choice that adds to the game. In "Spelunky" it forces the players to learn the game's mechanics rather than a level layout and in "Don't Starve" it gives the player insecurity of not knowing the exact map layout and thus rewards exploration even if you know the game in-and-out.
Oh indeed. I don't deny that with enough work you can make interesting levels. If you throw enough resources at most problems you'll get results. But the amount of effort needed to build levels in the sense that I'm thinking hasn't been done before imo. This includes Spelunky and Don't Starve. In fact I wouldn't have even considered Don't Starve to have interesting world design at all. And while yes, Spelunky's level building is nice. It still doesn't do it for me. The levels all feel the same, which it's going to. It's not building a unique world that you can spend hundreds of hours in. But I am, and I think it's better to hand craft those unique little experiences that made me fall in love with exploration in RPGs.
@@GameEndeavor I commented not necesserily on your game more like on RNG level generation in general. I agree with you, with Spelunky and Don't Starve not having interesting world and map designs in a sense that I would like to hang around in there. What I meant it does have interesting as it fits what the level design for what those two games tries to achieve. Both beeing pretty "arcady" in a sense. They focus on "try again" and the urge to "git-gud". Exploration in both games is built around the game's system -> "what happens if X and Y interact?" "can I do this, can I do that?" But if you compare both to games that focus world exploration, they're pretty boring. I guess there is a reason why we haven't seen a good metriodvania or zelda-like that uses random level generation.
lol, just recently I joked on stream about not noticing my likes going above 500. Glad to see you're profiting off of the algorithm noticing me recently. xD
It's only like 100 lines of code, most of which I had already written previously, and useful for any number of rooms I want to make with it. I stand by what I said.
It is not. It's inspired by Terraria, Link to the Past, the JRPGs I played as a kid, and 0x72's dungeon asset pack which got me interested in pixel art when I did the Miziziziz jam. I assume you're referring to either the hand swinging thing which a lot of games do for mechanical reasons (it separates the hand from the body allowing free movement, ie Enter the Gungeon). Or the left and right only art which is purely because I'm too lazy to draw front and back animations and trying to avoid as much excess art as I can get away with.
I do, I'm just not using it atm. The actual cave system itself is low priority for the demo and will only be added if I have time. The cave generator I mocked up quickly since I already had everything I needed from the grass generator.
Howdy everyone! I've been quite ill for the past several weeks, so I'll hope you'll pardon the absence. Hope y'all have been alright!
Hello! I saw your parallax cave effect. I just wanted to tell you that you can actually add swingin hanging wines . Or stalagmites etc. That would really cool!
@@wagwan0 I don't have any cave assets atm, only forest. You might be referring to the tree foliage? Those are good ideas for when I do cave art. :)
@@GameEndeavor you can write me at discord. newcalibur#5056 . I will be adressing you some assets that you might find handy.
Item drop animations are coded with random parameters? and you use an extra object to store black shadow sprite under every object/character/item?
Yes and yes.
This game has a lot of potential, this devlog series has become my favorite on TH-cam now, keep up the good work!
Thank you! I'm very excited about it. Looking forward to seeing where I take it.
ok spade 👨🏽
The game could be great, the game could be bad, but what's never gonna change is how 🔥 these devlogs are
The shadow from the upper trees is a really nice detail.
Thank you. :) It is one of the major factors in the aesthetic I feel. Inspired by the Lost Woods in LTTP.
I love that this gamedev is more a game-editor-dev. This is something that I really like about your channel.
Thanks. :) It's what I'm working on currently. Trying to get my world building tools built so I can start prototyping the first dungeon which should hopefully be soon.
0:47 “Full disclosure, I am very lazy developer.”
Liked and subscribed!
lol, welcome aboard! xD
This was such an interesting devlog! I'd love to learn more about the subject. There's so much I still need to learn, great!
Thank you! Yeah, PCG is interesting, especially Cellular Automata. It's one of those things that's actually pretty simple but very useful.
The cookie master is here!!
Always glad to see more progress on this project. These tools look like they will help quite a bit with your workflow.
Absolutely. I created this mockup in about 10-15 mins. Previous devlog scenes took hours.
Game dev tool making can be even more interesting than making the game itself. Very cool video.
I think with more and more "real" applications made with Godot, plus the publicity it will get after version 4 is released... might just be enough to make the world realize it's potential.
I love the art style on this game, it's like a mix between pixel art and cartoon-ish graphics.
Thank you. :) That's exactly what I'm going for.
I foresee great success for this game!
That's very kind of you, thank you! :)
I enjoy your style of development; Very few devlogs on youtube actually talk about the custom tools they create to speed up development, even when it's essential for content creation in-game.
Thank you. :) It's part of being a developer. I imagine others would rather talk about the game itself, but with my devlogs I'm leaning a bit more into the fact that it's mostly developers / aspiring developers that watch these. So I try to cater my content more towards them. Especially those that are interested in knowing what I do as a developer beyond just adding game mechanics.
@@GameEndeavor I think it's a smart decision, and being specific about your audience (even if it might be a more niche) is tons better than just screaming about what you make to everyone, with no regards for who might be listening. Keep at it! 🔥
I cant thank the youtube algorithm enough for finding this guy. Awesome stuff
I can't thank the algorithm enough for finding me either. xD Glad you enjoyed it. :)
This is awesome! I love that you are hand building along with PCG. I think that PCG really shines in giving people a starting point to build off of. Game is really looking good!
Agreed. I think it's a wonderful merge of styles as well as a improvement in the workflow.
Hi, here's a bit of feedback/idea: When you defeat an enemy (at the last hit), the corpses could get knocked back a little, because that would give you a lot of game feel/ juice. A jucier game, is usually a better one. Keep up!
The editor is becoming super interesting. But imo it's kinda lacking until you can press a button to create the whole game.
/s the game is looking really cool!
lmao! Soon that will be possible with AI I'm sure.
Smh, It's lacking the ability to press a button that creates an engine for the game
Amazing Devlog, I need to setup my own shortcut tools to make building worlds easier. Also Procedural generation looks interesting
Thank you. :) For sure. I can build things very quickly now, and I actually enjoy the process whereas I didn't previously.
I use pixelorama for my game development, I recognized it right when you showed it, immglad the it is getting more recognition, it is a really good pixel art program
You have some of most in depth dev logs on TH-cam. You do an excellent job describing your process and the discoveries you're making. Always very inspiring!
Glad to hear it, thank you! Possibly a side effect of being a formal tutorial maker. :)
@@GameEndeavor That's probably it lol :)
The Return of the King!
Glad to have you back, hope you're doing well. Your devlogs are super interesting, great job! :D
Glad to be back. Thank you, Ludos! :D
My brain did a backflip when I saw the notification
lol, well that makes me feel special, thank you. :D
Wow using cellular automata for this is unecesarily convoluded and awesome at the same time.
I wouldn't say unnecessarily. It saves me so much time building these shapes by hand, which I hate doing. Before this, it was my biggest time sink, having to place every piece of grass. This gives some really interesting shapes too, especially at larger sizes. And the whole thing is less than 75 lines of code and is easily reusable with just a couple lines of code. :)
@@GameEndeavor I did not mean using procedural, but ther are easy and proff methots to archive a similar result, you gone the hard way.
Subscribed. I found your channel when researching state machines. Great work and great editing. Thanks for your time and effort.
Thank you for watching them. :) State machines were my specialty back when I did tutorials. Useful stuff.
What a found! I really like your content man. Happy New Year!
Thank you! Happy New Year. Here's to hoping we use it well. :)
So nice to see you back! I went back to watch your live stream and saw you were sick the last few weeks! Hope you're feeling much better now!
Indeed, much better now, thank you for asking. :) Happy to be making devlogs again.
I've used the cellular automata method to make procedurally generated caves before but it would often generate several unconnected caves that the player couldn't reach. I fixed this by adding a flood-fill like method to identify and calculate the volume of each area of the cave system and then only keep the largest section for my final map.
This is just a tool to generate a layout for me to use. Afterwards I clean it up by hand, so there's no real need to take it further than I have. :)
Nearly all the B roll footage is just slashing at enemies and placing grass patches, but it’s so satisfying to watch that I don’t even mind
lol, it's hard to get good b-roll footage when you're saving the prettier parts of development until after you have the mechanics you want. However I'm about to enter the polish / world building phase. So that'll be fun I think. :)
I really enjoy the art of randomization in games. It's such a treat for the curiousity!
I totally agree that hand crafted areas tend to look much better, but for my game, I wanted as good of a result that I possibly could using randomness for the heavy lifting.
I ended up chooseing to build pre-designed level sections which can connect together randomly based on some rulesets.
When I'm further into the development, the content of these pre-designed pieces will be randomized to. Perhaps a event quest spawns. Or a pack of monsters.
If it sounds interesting you can read more about it on my blog =)
www.ledii.net/blogs/game-dev/mining-game-v2-randomized-caves
Maaan, I just found out you were making this game. Looks awesome!!! And close to what I would like to make.
Few things I thought about:
- will the character always look like this? He's kind of cute, but this little fellow is far from heavy enough to make such a cool stomp if you ask me. Maybe if you become a warrior you get some extra beef? Also - armor etc? You do visualize the weapon... :D
- I love it how you try to make it balanced and interesting to play as different classes. Though I noticed - the warrior is vulnerable to all attacks (skeletons/melee, those brown things with range attacks) and rogue would only be vulnerable to ranged attacks (since skeletons need to catch you to hurt you). Maybe add more twist to the rogue fighting style, like some enemy homing projectiles you need to shoot down before you can hit the enemy. Now it looks like it's fairly easy to just avoid the particle trajectory.
Sorry for the long comment. I'm super excited and can't wait to play this. Keep it simple as much as you can, so that you can finish it soon (and not take years like most of us). Also, on the contrary, I hope to see LOTS OF LOCATIONS, LOTS OF ENEMIES, LOTS OF ITEMS, LOTS OF TOWNS, LOTS OF STORYLINE, basically Diablo 2 all over again. Thanks.
I love your vids, by the way. So tight, yet easy to follow. Keep it up, you're awesome. :) Big fan!!
Howdy! Your enthusiasm for the project is astounding, thank you. :D There will be character customizations, possibly some heavier armor for melee class. Not a different character base though as that's a whole lot more work. The combat is still very much wip, I haven't even tried balancing the different classes yet. Mostly adding features I want currently.
This project will take me years probably. xD With it being so sentimental to me, I want to take my time and make it as good as I had dreamed about. However I'm not really a fan of the dungeon crawler style of gameplay of Diablo 2. :) This is a much more mellow experience. And currently there will only be one town. Subject to change, but the idea being that I want to focus on the NPCs, and would rather have limited memorable NPCs rather than a bunch of towns.
Fun note if I recall for bloodborne even! The chalice dungeons have a set layout for the base game rather than purely generated always!
These extensions to your level editor look like a real improvement to your ability to create levels. Looks like really great progress!
Oh for sure. I'm just about ready to start doing some serious world building now. :) My workflow efficiency has been improved tenfold if not more.
@@GameEndeavor can you make the editor available to download?
Awesome Job As Always !!
Thank you! :D
This is so relatable:
th-cam.com/video/AAAmQKgHsDg/w-d-xo.html "This can even make it difficult to write the scripts for these devlogs, because it's almost like rubber ducking the logic for my game. I'll realize that I can do something differently, or get an idea that I start writing code for, and the next thing I know it's 11pm at night and I only have two minutes worth of script, but a useless feature that's fun to play with."
I started working from home a few years ago (my job decided renting an office space wasn't worth the cost since our business is all online) and so often I go to report on my progress and realize as I describe what's been happening that I need to change things. I end up scrapping the report and working for hours more on the next part and never report nearly as much as I should.
just realized i wasnt subscribed, great videos man
Thank you, Pete! You as well. :) You were the main TH-camr I watched to learn pixel art.
You're a clever guy. I often wish I had a similar level editor, namely the procedural generation. I've no imagination for zone layouts.
The game looks great already!
so concise! great videos
Impressive work as always! cool video :)
Thank you. :)
Praise the Bread
Yayyyy you're using pixelorama its such an amazing software
Also it was made using the godot engine as well
I personally use Aseprite, but I know of Pixelorama and appreciate them for making a quality tool easily available to developers. :)
Up next: a Copy/Paste feature that allows you to duplicate large areas. Hahaha
Amazing work, love seeing all the updates.
Absolutely. What I really need is a bucket fill and box selection, lol. After the demo though. I can get by without for now.
Spending the time to make tools to make your life easier is something I respect. Especially when you need to weigh the options of how long will the tool take me to make, vs how much time will I save using the tool?
tbh there wasn't much to weigh. It has already paid for itself. It used to take me forever to build areas. That's why the first devlogs feature one area. Grass decorations were the most time consuming. Now I can build levels fast enough to make one per devlog, just to keep things fresh and practice my building skills.
I'm knee deep in Unity at the moment, but this is really inspiring me to try Godot. Nice looking project. Good luck with it.
Thank you. I used Unity a few years ago. If you're working in 2D then I loved the switch, but I didn't enjoy working in Unity for 2D.
@@GameEndeavor I can imagine. I was using Game Maker before for 2D games, before Unity had more 2D tools.
Man I wish you would add like a rare type of flower, like a black/red/purple flower. Or even more flowers in general, the world looks great ! but missing some life, maybe leafs falling when you enter a tree and stuff. just an idea though, keep it up !
I'm just now starting on the world. :) Everything up until now has been mostly features and tools, with the bare minimum of art needed to make pretty devlogs. If you think the world looks great now, wait 2 months. :)
Wow such a great work..
This game looks so friggin dope 😩
Thanks! It has certainly been a lot of fun to develop so far.
You can also make it so the trees and stones respawn after leaving the chunk... idk just an idea
that level editor seems like it'd fit better as a godot plugin (tool).. you could combine godot's editor itself with the procedular generation
I believe Bethesda used this method for Skyrim so you're onto something.
Method as in combining procedural generation with manual world building? I wasn't aware, but that's interesting. I'm surprised I haven't heard more about the technique tbh, it's already helping me a lot.
The game looks amazing, all the work you put into combat has paid off since it’s so satisfying to watch! Quick question, have you thought about making your level editor feature available to others? Either for a price (if within godot’s terms of service) for some passive income or just as a handy dev tool for other people to use. Whatever the answer keep up the awesome work!
Any particular reason you're using cellular automata for the grass patches? Wouldn't just making an array with random grass tiles work just as well? Feels like overkill just to generate some random tiles lol. But hey, if it works, it works. :^)
Because it makes clusters as opposed to noise. The blob shapes are much more aesthetically pleasing imo.
Plus it's there (for caves and whatever else I'll need). Not really overkill to use what's laying around. It's just a couple lines of code to generate the grass clusters.
At last devlog after 3 weeks you promised devlog every week i was waiting for that but great video btw i love it
I don't promise a devlog every week. :) My goal is to make a video every week, but I've been very ill for the past few weeks.
@@GameEndeavor ohh sorry take care of your health
lol u told 3 things in on sentence
my brain : SUBSCRIBE TO HIM NOW
me: why ? I just discovered his channel today !
my brain : YOU LITTERALLY WATCHED 11 OF HIS VIDEOS IN A SINGLE EVENING
me : yeah I guess you might be kinda right
*+1 subscriber*
lol, seems like you enjoy the content. xD Thanks for subscribing!
Can you make a tutorial of how to make those fancy brushes you use? Great videos bro! Subscribed
This looks insane!
Thank you!
dude this fame looks so nice it would be a ton of fun with freinds if it had a co op
Local co-op is planned with online multiplayer being a possibility later on if it doesn't bog down development too much. But local co-op would atleast make steam's remote play something to look forward to.
I'm craving some texture on the path. Footsteps or maybe uneven ground?
These are fucking sick.
This is BADASS
Is the cellular automata live stream somewhere? Would love to see that one
EDIT: amazing how the script and tools are looking like a separate stand-alone application. Didn’t know about those endeavors using Godot as the underlying engine.
So procedural generation is like asking ai to build a basic layout of Mario maker, then you go in and change what you need to add polish.
Now I wonder if I can ask Chat GPT4 if they can make my game, because I know I can't.
This didn't use AI for this, I created a simple algorithm. And typically with procedural generation you don't touch it up after. I just used it to automate a lot of tedious work. Like I use it now for placing grass decorations or placing trees and rocks.
You probably could get it to create these algorithms for you, but I don't think it's at a point where it can make a professional quality game yet. It seems to have a lot of errors and its output is mediocre. You'd actually have to know what you're doing just to fix the errors. For example I asked it to create a Binary Space Partition thingie and I had to go in behind it and fix its errors, and could only do so because I knew what the logic was supposed to do.
@@GameEndeavor yeah I misspoke with ai and procedural generation.
But it sucks that I seem to have all these ideas for games but can never seem to make even something that room with enemies spawn all while following all the guidelines and my PC crashes.
Sold my PC 3 years ago and don't know if I want to try again with Chat GPT4.
How is your tilemap tool made? It looks so easy to just "paint" dirt blocks and the corresponing blocks match into place with corners etc
Awesome
You should make a PR to Godot for those signals!
There's probably a reason why they don't have them. :) It's only three lines of code that I feel is too obvious for them to not have thought about.
@@GameEndeavor I thought that, but it doesn't hurt to start a discussion if there wasn't one before
Wow!
Hello, great job !
I have a small question: how do you manage the transparency of the trees when the character is behind?
I really have difficulties to reproduce this kind of display (ideally I would like to display my character in transparency behind but I can't find any info about how to manage the Z-test or manipulate the depth map in shader for 2D objects) ?
thanks in advance
I put an area on the tree that detects the player and adjusts it's alpha values. I've been looking for a way to render the player's silhouette too but yeah, no luck here either.
Didn't understand a single full sentence said 'cause I don't understand code fully... But felt good listening to your progress.
Is constructing caves this way similar to how cavelike structures are generated in MC? Those graphics look sweet though!
Minecraft caves are generated with perlin worms, which is why they're often very path like. Thank you. :)
Looks really nice
Reminds me a bit of the Zelda games on gameboy advance, is it in some way inspired by zelda ?
Only aesthetically, gameplay is not even close. :)
2:00 omg haha I made a program on my TH-cam channel that does exactly that lol
When you are a lazy developer but you don't want to release a terrible game
Necessity is the mother of invention, but sloth is its father.
Proper procedural generation produces both handcrafted level worlds AND requires a lot more work. PCG is not a short cut. That is why it always sucks. Look at a game like Unexplored. Its levels are better than most hand crafted dungeons by a fairly large margin. Why? Because the devs put the effort into having the generator use actual level design philosophy via its cyclic generation. It wasn't treated as a short cut and its levels are hand crafted quality, often times better than most devs ability to design, and infinite at the same time.
Pfft, all developers are lazy. If it can be done faster it will, and thats more efficient. Its not always easier to write a month worth of code but if it means you don’t have to spend an hour a day on that one thing you’d do it. Engineers, and software engineers are just people who solve problems. The laziest people on the planet who will do everything to have to do nothing. Its counterintuitive but it makes people great at being motivated.
Nice ;D
Could you make the ´tool´ you made a plugin for godot someday?
Hi, I want to ask something, I'm currently developing a game in a perspective like this with pixel art, but I'm having a problem with shadows. I see that your semi-transparent shadow sprites can come on top of each other but not create a darker spot, which how it should be. How did you achieve this? I looked at your prior devlogs but couldn't find it.
Yeah, shadows are quite tricky. Most games don't even fix it from what I've seen. I fixed it with a bit of shader magic. idk if it'll work outside of Godot, but with Godot shaders if you get the screen pixel that's rendered under the shader, it doesn't seem to account for other shaders on the same canvas layer. So I'll use a shader to place the shadow by multiplying the screen texture with a shadow color. Then when an overlapping shadow renders over top, it'll perform it as well, but will ignore the shader under it. This is highly experimental and probably prone to some breakage.
is there a code tutorial or live coding on twitch or something that follows along with this game creation in code, or just a dev vlog?
Just the devlogs. :)
Why don't you try to create some boss battles? I think it would be an interest addition!
I will at some point. I'm focusing more on the world building tools atm. :)
Instead of modifying TileMap in the godot engine, why didn't you simply add the signals you want by extending the TileMap class in GDScript?
I considered that, but I have a lot of TileMaps and I wanted an excuse to learn more about Godot's source since I'll probably be faffing about more in there later. I'm not even using the custom engine atm, was just a fun endeavor.
Howdy 🤠
Howdy! :)
Take my money!!!!
Have you ever had a bug where godot stops importing assets? Aside from redoing my project and reimporting i have no idea how to fix this and google is just telling me this is some unsolvable bug with godot. I hope there is a solution because at this point I likely need to move on to unity or try to make my own custom engine using c++, which is a shame because I was starting to like Godot.
I've never had that bug, no. You're probably better off asking in their Discord server, or on one of the forums on their site.
If enemy pathing is working in your game? Then could you possibly automate more of the world / cave gen by having the enemy pathing AI determine if the cave paths are actually explorable / reachable? Possibly it’s easier to do it yourself, but if you can have this running on 100 generated maps at once and then show you a list of only the ones that are at all useful in a baseline sort of way, then it could speed up even the point at which you start, or even the point at which your players start.
Even could extend into some sort of accessibility function that lets players know whether or not all areas in their custom maps are reachable.
I’m not a programmer at all, so maybe this is all ridiculously stupid.
Good job regardless!!
Hey Siri
Define "Overengineering"
lol, I already had everything for it and it was really simple to put together. You're real adorable though. :)
Great game,release date?
Demo is planned March of 2021, full release probably not for several years after. I'm still very early on in development.
This really reminds me of a machine learning paper where the researchers train a neural network AI to modify initially randomly-generated maps into playable ones, iterating tile-by-tile, and being rewarded according to the playability & difficulty of the final map, among other quality metrics. Basically a gamification of map design, that an AI can understand & learn to play
arxiv.org/abs/2001.09212
Yannic Kilcher also made a nice vidoe summary of it.
Your method is basically a human playing it in place of of the AI :) with the human's gameplay design skill also providing the quality metric, of course
On a side note, have you explored/tried out cyclic dungeon/quest generation?
I had never thought of that. I'll have to take a look at what kinds of maps they designed because that sounds interesting.
I read an article on cyclic dungeon generation, but never attempted it myself. I haven't dabbled too much in random generation beyond an arpg prototype, because I often find pcg to be bland. But if I were to make pcg dungeons then I would likely go with a cyclic design. :)
You are not lazy, you just want to work in a fast and efficient way. I hate it when professional call themselves lazy instead of saying that they are good.
lol, it's just an alternative way of saying it. It sounds pretentious and boring to say "I like to work fast and efficiently", so instead I say "I'm a lazy developer". Fits my personality better.
@@GameEndeavor Well, I'm having troubles doing anything, so it bothers me when people who are clearly skilled say they are lazy. I won't complain again, I promise, stay true to yourself, that's how we like you anyway.
this game kinda looks like archvale. Did you take insperation from it?
Nope.
There is a thought fallacy that IMO is the reason why randomized levels are often not as interesting as hand-crafted ones. Level randomization is not less work than hand-crafting a single map. Some devs treat it the other way around, so they don't have to do level design. But good games require level design that is fit to the game's mechanics. There is no way around it.
There are very few examples where randomization works but there it is essential: "Don't Starve" and "Spelunky" come to mind
Good level randomization requires to hand-craft maps to see what works and what doesn't and itereate upon that. Only then you could actually tweak the generator to create maps that are very similar to that and even then to make them interessting they require "set-pieces" i.e. handcrafted parts of the map that add diversity and interesting stuff (e.g. the Pig-Village in "Don't Starve" or Idol-Head traps in "Spelunky").
And last but not least level randomization in itself should be a deliberate design choice that adds to the game. In "Spelunky" it forces the players to learn the game's mechanics rather than a level layout and in "Don't Starve" it gives the player insecurity of not knowing the exact map layout and thus rewards exploration even if you know the game in-and-out.
Oh indeed. I don't deny that with enough work you can make interesting levels. If you throw enough resources at most problems you'll get results. But the amount of effort needed to build levels in the sense that I'm thinking hasn't been done before imo. This includes Spelunky and Don't Starve. In fact I wouldn't have even considered Don't Starve to have interesting world design at all.
And while yes, Spelunky's level building is nice. It still doesn't do it for me. The levels all feel the same, which it's going to. It's not building a unique world that you can spend hundreds of hours in. But I am, and I think it's better to hand craft those unique little experiences that made me fall in love with exploration in RPGs.
@@GameEndeavor
I commented not necesserily on your game more like on RNG level generation in general.
I agree with you, with Spelunky and Don't Starve not having interesting world and map designs in a sense that I would like to hang around in there. What I meant it does have interesting as it fits what the level design for what those two games tries to achieve. Both beeing pretty "arcady" in a sense. They focus on "try again" and the urge to "git-gud". Exploration in both games is built around the game's system -> "what happens if X and Y interact?" "can I do this, can I do that?"
But if you compare both to games that focus world exploration, they're pretty boring. I guess there is a reason why we haven't seen a good metriodvania or zelda-like that uses random level generation.
I was the 1000 like... Isn't this like 7 years of TH-cam luck?
lol, just recently I joked on stream about not noticing my likes going above 500. Glad to see you're profiting off of the algorithm noticing me recently. xD
WHat you using to make this?
👁👍
wait ive been here before
next devlog=new enemies?
That might be soon. I have 1 or 2 enemies planned for the demo but they're not a big concern atm. Though there is one I'd like to do soon.
What I have been doing this week though is enemy ai adjustments, adjusting their behavior to make them slightly more difficult and interesting.
Ok nice!
I’m a very lazy game dev so I wrote a hole system to do it
It's only like 100 lines of code, most of which I had already written previously, and useful for any number of rooms I want to make with it. I stand by what I said.
Is this inspired by archvale
It is not. It's inspired by Terraria, Link to the Past, the JRPGs I played as a kid, and 0x72's dungeon asset pack which got me interested in pixel art when I did the Miziziziz jam.
I assume you're referring to either the hand swinging thing which a lot of games do for mechanical reasons (it separates the hand from the body allowing free movement, ie Enter the Gungeon).
Or the left and right only art which is purely because I'm too lazy to draw front and back animations and trying to avoid as much excess art as I can get away with.
creating godot inside godot*
How does this title work🤔?
I'm using procedural generation as a tool to help me build levels. :)
But you still dont have dungeon/cave texture
I do, I'm just not using it atm. The actual cave system itself is low priority for the demo and will only be added if I have time. The cave generator I mocked up quickly since I already had everything I needed from the grass generator.
Is this free to download?
There's nothing to download currently. I'm expecting to release a demo in March of 2021.
Godot itself is free to download, although you probably already knew that if you watch this channel :)
such substance but no flavour
YT algorithm strikes again#!
Thank God, i am on time
Now let's see how much time he will take to heart the comment. (JK)
3 minutes, and that's with me walking my dog when you sent it. ;)
@@GameEndeavor Wow, that was really quick lol
@@codedecode1498 I usually dedicate an hour or two just for responding to comments after I post a video. :)
@@GameEndeavor And we all love that ^^