Lain Game Dev General /lgdg/

Game Dev thread time V2! Hey lainons, how is that game dev project going again? > Discussion Starter: What is your favorite game engine or library? What are your favorite game dev resources? Favorite programming language for game dev? What is that dream project you got in the works or that some day project? What is your current game dev project that you are working on? Favorite band or song to listen to when deving? What was a hard problem you been facing recently? Favorite game jam or type of game jam? Anyone writing games or engines in Lisp? > Engines / Frameworks / Libraries Godot (Engine) https://godotengine.org/ Armory3D (Engine) https://armory.itch.io/armory3d Ren'Py(Engine) https://www.renpy.org/ Unity (Engine) https://unity3d.com/ Unreal Engine 4 (Engine) https://www.unrealengine.com/ GameMaker (Engine) https://www.yoyogames.com/gamemaker SDL (Framework) https://www.libsdl.org/ cl-sdl2 (Framework) https://github.com/lispgames/cl-sdl2 Allegro (Framework) https://liballeg.org/ SFML (Framework) https://www.sfml-dev.org/ Love2D (Framework) https://love2d.org/ Monogame (Framework) http://www.monogame.net/ LibGDX (Framework) https://libgdx.com/ raylib (Library) https://www.raylib.com/ Ogre3D (Library) https://www.ogre3d.org/ CEPL (Library) https://github.com/cbaggers/cepl cl-opengl (Library) https://github.com/3b/cl-opengl > How to make Webm obsproject.com gitgud.io/nixx/WebMConverter

>>32758 >how is that game dev project going Not great, not terible. I'm just working on a little RPG in C++. I've got my D&D:3e books and a few C++ books on my desk. It is nothing special and it will probably just be a completely text-based little thing. I'm considering working on making a client/server thing, adding a textbox, maybe adding encryption or something. Just something fun to toy with. Once I get my new GPU in, I'll be picking up OpenGL and working with that, and I might work on a more serious project.

I've been trying out bevy the rust ECS engine. Not the biggest fan of the language but the ECS pattern abstracts away the worst of it and you get to more or less ignore the borrow checker in that context, also they walk you through configuring for "hot reloading" with incremental builds so you don't need to wait eons for rust compile times. The engine is more or less perfect for what I'm building, another data-heavy, dwarf fortress style game with RTS and RPG elements. It is definitely very new though, no great immediate UI solution, no graphical editor which might be a bigger problem for other genres, but they have some professional interest I guess so it's been progressing pretty rapidly. I assume some people here have tried it, what was your experience? > inb4 muh tranny language https://bevyengine.org/

>>32758 I use Godot, which is just about to get the new version into beta, which is going to make 3D actually viable. Be careful with Unity, they merged with a shady ad company. I've started a new project on a theme I've been trying to get to grips with for a while now. I think I've got it figured out this time, and after a week of grappling with idiotic bugs I couldn't diagnose properly, I've got basic player movement and actions and basic NPC movement down and the art pipeline sorted. Now, I'm going through boardgame rulebooks and such to figure out the specifics of the turn-based rules for me.

>>32758 I'm trying to make my own ECS for a roguelike with almost everything made from scratch. I'm currently stuck, but I've been making decent progress.

sorry if it is the wrong place to post, but the thread is over maximum post limit. replying to https://www.lainchan.org/λ/res/8822.html#q31185 not the lainon you talked with, but i am also interested in talking about gamedev and your projects in c and java REVEMTE0RUMzMTI3QjQ1QUU0REMwOEEzRTUzQzczQTU1NzY1MDIzQkQwRDkzRjZFQzFFMzg1RTkwQ0Q2MDA0N0JFOTlFNTg4OUNDNQo=

not a game dev but I've made a few musics for friend's games, feel free to ask if you want something https://www.youtube.com/watch?v=Fb6Szvipub8

I'm working on a cyberpunk FPS with weapon customization a implant upgrades. Kinda like Deus Ex, but more fast paced. https://retronation.itch.io/cyberkiller-redemption

>>32839 I'd like to ask for advice. What should one learn to do something beyond just dropping a bunch of rectangles in a music composing program and hoping that it ends up listenable? Something that would allow one to have a more structured and educated approach to making music.

I would love to get started with making games, unfortunately I don't have any skills and will need to start from zero.

>>32920 >I would love to get started with making games, unfortunately I don't have any skills and will need to start from zero. The easiest way is to first learn Lua and then start making simple 2D games using Löve (aka Love2D). I recommend Lua because it's really simple programming language, and it's used in the game industry as scripting language in games (examples include Gmod and Factorio). >https://www.lua.org/pil/contents.html >https://love2d.org

Making a roguelike: I got basics of dungeons creation. The goal is to get corridor to connect, implement randomness, add in basic structural aspects like doors, staircases, and blitz that onto a screen.

>>32921 I can vouch for lua: I have not given love2D a chance, but lua itself is pretty solid. Plus, I have used lua as a scripting engine for my own game

>>32921 I will look into that. Thanks for the advice.

Working on a tetris clone, with an extremely important twist I've been slacking off by playing tetris in Emacs, and had to disable tetris there. So I'm making a tetris clone with a play timer, as in it stops me from playing for more than 5 minutes at a time. Weirdly, there's an official docs on how to make a tetris clone. One thing I want to add to this version is I want to hold tetris pieces.

>>32786 >bevy >I assume some people here have tried it, what was your experience? ive been using it for 6 months or so. its definitely at the point where i wouldnt generally recommend using it unless youre planning to get involved in developing it (or the ecosystem around it). for a lot of engines, that would be a deal breaker but honestly bevy is a pretty easy codebase to work on. it's just solid well designed abstractions all the way down. the same stuff you learn from working with the high level api also applies to a lot of the internals so contributing comes naturally. community is decent. relatively high skill level. it's very much a rust community project, whatever that means to you. lead dev seems like a decent guy, if a bit "silicon valley" for my taste. he's undeniably a talented programmer though, having read a bunch of his code.

>>33014 >it's just solid well designed abstractions all the way down. the same stuff you learn from working with the high level api also applies to a lot of the internals I haven't had quite much time to work on this as I'd like but this was what I inferred from my experience so far, great. I am still thinking of going back to c++ while it's early in the works to take advantage of a more stable ecosystem. Thanks for giving me an account of what to expect.

>>32894 Looks more like a malware...

Been griding real hard the last few weeks, but its showing progress. Have been doing lots of gamemodes and also some bot work. Im quite excited to work on a level generator but im having trouble finding a good base algorithm that produces satisfying results. In addition i make lots of game objects with randomizing properties, this makes things very interesting. The bots in particular will be quite an interesting field of work with all the things i can make them do and do with them. I hope i can release lots of games before the end of the year.

im working on a puzzle game that is inspired by the 90s puzzle game chips challenge and maybe a text adventure using inform 7.

New stable release of t3d out! https://torque3d.org/forums/topic/5843-torque3d-401-released/ Havent tried it yet, but i think it should go flawless porting over to it. Have been working on bots, they are coming along nicely but the code is a fuarrrking filthy mess. Also messing around with graphical effects and such, i should show some screenshots later on.

For a long time I've focused way too heavily on game visuals instead of actual gameplay. This lead to me making a bunch of games that looked nice but had zero gameplay whatsoever. So I decided to make a website centered around experimental gameplay. There is monthly jams and anyone can sign up and submit a game if they want to! I made this to hopefully help myself and others to prioritize gameplay and create new forms of it. Let me know what you guys think: https://experimentaljams.com/

Is there a comparative review of space sim large world techniques? I'm thinking of how I want to organize my scenes across the scale of the solar system but I'm finding it hard to reason with so many unknowns. Should I calculate transforms with doubles on the CPU and pass as floats relative to camera for render? Should I composite multiple scenes at different scales? Should I render the scene graph in reverse order so camera is always at origin? What are the tradeoffs? Are there any gotchas? Without implementing each technique myself, it's hard to judge.

>>32786 They should change the name, that one is taken https://github.com/ZechCodes/Bevy

>>34173 Love this idea. Cute logo too

>>34173 Its not much of a game without gameplay. Im not one into gamedev jams. Im taking my gamedev stuff to the next level and set up dedicated machines for each engine i work with including full developer stack. Also i got me tons of hidef x11 assets but i need to sort trough all of them. Things have been going real well, but i need to take it to the next level and release stuff. Im feeling lazy now, but next time around i show some stuff i did i think.

I've probably spent too long just quietly working on my project before saying anything about it. I usually prefer to just quietly work on things i suppose. It's still nowhere near done, but at least it's announced now, i put together a small trailer for the occasion as well.

Which is the best language for game dev: Zig or Odin?

>>37209 Zig grew out of a project to write audio software Odin was inspired by a language specifically designed for game development

What does Lain think of Godot? I like the principle of an open-source lighter-weight unity-like game engine. Apart from Unity for VRChat stuff though I've never really used engines too much; I just use OpenGL and c++, but i only really make simple 2d games because of that. It looks like it could save a lot of work compared to doing things by hand, but I'm not sure if it's developed enough to merit learning yet. I'm hoping this recent Unity fiasco gets them more devs and polishes it to be a real contender to major game engines.

>>37209 Common Lisp as usual. If you want Godot there's the C++ integration you could use directly with Clasp without making hsims. And eventually LCL will be a complete Lua-hosted implementation so all the engines that use that will also be available.

>>37255 So instead of just one layer of interpreting happening now we have LISP -> LUA -> C++ ? won't that be kind of slow? (idk much about lisp i mostly just use c++)

>>36814 The attached vid is your game? It looks really impressive! Is this something you've been doing on your own or you have a team?

>>37253 I've made three games in it and making a fourth and I think its fantastic. Obviously you cant do "nintendo hire this man!" kind of stuff as easily as unity or unreal but I'm not interested in making that. Really glad its getting proper funding.

>>37253 >Godot It looks interesting, but I'm personally more interested in FNA. First time I hear about it and, going thru the list of games made with FNA, I never realized how many notable games were made with that library. (it's just a library from what I can gather, right?)

>>37258 Most CL implementations compile at load time. LCL compiles Lisp functions to Lua functions, which are further compiled down to native code when running on top of LuaJIT. In some situations, this might be slower than (for example) SBCL, but by a very small factor compared to many of the alternatives. You'd have to benchmark it to know where the weak points are.

>>37253 I love Godot, wish I had more time to work with it. I'm currently trying to write a VOIP plugin for Godot 4.x. Audio processing threatens to filter me but it's slowly coming along...

>>37259 Thanks, It's been mostly me, though a friend helped dev a few systems and also a few art assets were contributed by other friends. Still missing music production skills though. It's tempting to learn, but i think i should resist so that this project actually releases sometime soon.

>>37423 Lots of game devs want to do everything themselves, but it's actually quite distracting for the developer. I find often these sorts of people aren't actually very talented in anything. They can crap out an 8 bit song and assets if they need, but it's not particularly interesting because they aren't involved in the medium any more than they think it will achieve their masturbatory vision. Often those sorts of people never go anywhere either. The one exception I can think of is the Songs of Syx developer. Though consider he has been making his game for an ungodly amount of time. Even the very famous 'solo projects' like Braid or Fez have hired some extra hands to cover visuals or audio.

>>37424 Yeah it's really not a good idea. It'd be different if i already knew music. I have no musical ability so it's for the best, knowing just enough to communicate with the people that do know music would be ideal in my position. This week i added a new loadout editor and am currently adding a very basic level based unlock for different equipment. It's not intended to be grindy, it's just mainly good that choice paralysis doesn't slam new players with 30+ weapons all at once, and give people something initial to work towards.

>>37209 >>37253 The best languages if you want to actually complete your game are: * Godot + gdscript or python * Lua with Love2D game framework * (possibly also GZDoom with one of the scripting languages that it supports) For more serious/professional projects, you could use these: * C# + FNA or Unity (beware: it sends telemetry data!) * Java + libGDX or LWJGL (Lightweight Java Game Library) * C++ because it's still very widely used in game industry but it is like kicking dead whales down the beach for hobbyist projects, unless you know C++ before you start making games. IF you know C++, you should look into raylib, SFML or Allegro.

>>37428 > or Allegro 2002 ringed

>>37428 SFML and Allegro aren't very good stick to raylib and SDL. I learned how to make proper games with SDL(bullshit engines before) and that taught me much more than learning unity or Godot ever did about programming. I would say actually making a commercial game in a reasonable time frame that isn't 2d you should go with Godot or similarly light weight engine. But making a 2d game you can use any framework and be fine since that doesn't require much boilerplate code to get going.

I'm writing a 3D software engine in C as a way of improving my programming skills. I want to make a game with it one day, though. Maybe. Next time I have at it I will write (part of) the polygon drawing routine. I was reading about Quake's software renderer and I think I know how to adapt it to mine. Will post updates.

>>36814 This looks incredibly well done. Good job, i hope you will have success and make a great game. >>37424 Its very hard work and reqires lots of patience, assembling a team is even harder. I work hard on my games, i want to make an alpha haloween release for some homies, gonna post stuff here once its ready so yall can enjoy. I made somewhat progress and got lots of assets at hand, but theres always those little problems that are so hard to solve. Sorting and rigging assets takes quite a lot of time on my part, but im sure it will be worth it.

>>37209 C++. But if you absolutely insist on one of those two, Zig. Zig is a pretty immature language without much written in it yet, but its at least got momentum. Odin is a hobby project created by people (plural?) who are too smug to recognize it as such.

>>32758 Work in progress: My tiny game that fuses gameplay elements from Heroes of Might & Magic 2 and Mount & Blade. -- My programming language of choice for this was Ada, tho originally I wrote it in C, 100% rewritten in Ada now. I drew all sprites in GIMP myself, except this UI in screenshot. UI styles / themes are switchable during the gameplay, but I'll "bind" them to in-game factions most likely. -- Gameplay wise, I imagined it as tile-based strategy game with few basic RPG elements, like having a hero / chad. Bots can (will be able to) equip items from dead enemies after the battle, or buy better gear at armory. It'll be fully FOSS strategy game when completed. -- https://git.lain.church/xolatile/xhads -- For curious ones, biggest challenges during the development are coming from pedantic code, and making everything easily extendable.

>>38965 Looks cool, but I hate how the resolution isn't consistent. The pixels of the interface are smaller than those on the map.

>>38966 I use 1:1 (pixel to pixel) scale for UI currently, 2:1 for in-game content and 4:3 for fonts right now. All 3 variables, aka ui.scale, core.scale and font_scale, will be available to change in Options in the very near future... -- When I view the game on my main big monitor, I want smaller font, but when I'm on ThinkPad, I can't see anything, so I think it's important to have it modifiable. You can scale map to be 1:1 right now tho. -- Here's another screenshot with "debugging/default" UI theme, core.zoom set to 1 (same as UI scale), using my own 8x8 monospace font. Not a work of art, but fun learning experience... (:

>>37428 >* Java + libGDX or LWJGL (Lightweight Java Game Library) I tried making a game with this years ago but had a major problem with stuttering that I was never able to solve. Also noticed the same problem when playing somewhat successful published game Lenna's Inception, figured this must just be an unresolvable GC problem with Java. Using Godot now, hoping to actually finish something. Been making portal animations for an enemy spawner with Stable Diffusion + AnimateDiff, here's one I can't really use because it isn't flat and won't loop well but I thought looks cool

>>37253 you can make a basic 2d game in godot in a day. You can download premades from store for 3d. It has the universal issue of "i would have done it different" using any premade architecture has but it saves a lot of time. The only difficult question is whether you want to do something the engine supports or you are so much out of the zone its faster to write custom entirely. In general i would advise doing something small for fast initial results. Nothing stops you as much as making no progress, getting bored and quitting. Progress keeps the fire alive.

>>38967 Work in progress update:

>>39024 skill menu icons especially diplomacy look nice, resource wood looks like its a plank with bark on. Did an attempt repainting , hope you like it

>>39025 They look awesome. I'll replace them with my icons tho, I want to make all sprites, code and sound for this game. It's ego related issue. (:

>>39031 New Fairy theme UI...

>>39041 previous one looked like fantasy + high tech remnants. New one looks like pure fantasy game.

>>39025 I like the simple existing icons. The redraws add too much clutter and as a result doesn't read as well. A lot of the time, less really is more, especially in pixel art. >>39041 Looks a bit flat. Try making the shading higher contrast. It'll make the theme 'pop' a bit more and create a better distinction between the individual UI elements. While we're on the subject of aestetics, if you want some good simple-ish top-down pixel art reference, check out Final Fantasy IV. Earthbound is also good, especially for color, although it's a bit further from what you're going for.

>>39043 Yeah, still don't know the exact style I'm doing, I just go with the flow... Aim *is* a fantasy game, we'll see how it turns out in the end tho. >>39049 I'm intentionally using "pillow shading" as those pixel artists call it, I drew pixel art before, but nothing high quality. I changed it by the way. -- This is comparison of 5 UI themes (notice 5 images 1440x900 taking < 500 KiB). Icons are by some 'opengameart' guy, I'll replace them.

>>39050 The changes look nice. If you want more visual suggestions, the grass either needs to be more detailed or less. Where it is now looks kind of sloppy, like you applied the MS Paint spray can tool to random parts of the map.

>>39051 Thanks for suggestions, I want to make this look good, I'm writing down in my notebook most comments here, so I can fix/revise thing later. There are few limitations that I placed as "artwork rules" which I won't break, otherwise I'll be fixing those, adding or removing contrast, etc. Rules: > using my 113 colour palette for all in-game pixel art. > outline and pillow shading on pixel art. > high-contrast art, similar to Heroes of Might & Magic 2. > base sprite size is 16x16, since it's grid-based TBS game. Last rule about 16x16 sprites is my "shortcut" since I want to draw all in-game art myself, and I'm not very good pixel artist, that ease my work. However, I structured (highly gameplay specific) "game engine" to support any base sprite size, animation count and framerate. It's changed in few lines of code. I can't make pixel art like images below, and even if I could, it would just take too much time. I searched for pixel artist to pay for, didn't found him (in person).

>>39053 As a rule of thumb, try doing more with less. Not only is it more practical, especially for someone of your skill level, it also just looks better. Having less explicit detail gets rid of distracting visual noise, and gives the player more room to imagine the game world as it looks to him; reusing a small handful of tiles throughout the entire game helps establish a strong visual language, which is especially useful for a game like this; having fewer frames gives the animation more snap; etc. It's important to not that I'm not suggesting that you go for outright minimalism. What I'm saying is that it's better to imply detail than it is to outright show it. It's why, in a lot of cases, the EGA versions of games look better than their VGA counterparts. Pic related.

>>39056 (me) I meant to include a screenshot of Loom, but the site really doesn't want me to post it for some reason. You can look it up yourself I guess.

>>39056 >>39057 Took the note of EGA vs VGA and looked at Loom, it looks excellent (EGA version especially) for its' simplicity and colour palette limitation. I've said in post above that I use 113 colours total, but I'll consider writing colour-palette-converter in C or Ada, to convert my PNG images. -- > As a rule of thumb, try doing more with less. I have big concerns about this, from the gameplay (implementing it right now) perspective. I want the player to learn the game in 5 minutes by playing it. Currently trying to cut out planned gameplay features that don't affect at least several stats, and to cut what I assume player will always do. -- Again, thanks for replies, I'm writing most stuff here in my notebook. It helps me to make something that won't be my total-egocentric-nightmare.

Finished mapshot procedure, but without units, and I'm rather poor with sprites, as I draw them myself. There'll be more of landmarks, buildings and items. -- Units (soldiers and civilians) aren't rendered on mapshot for now. It's a large image, but only takes < 700 KiB, due to 113 colours palette... -- procedure mapshot (file_path : in string) is begin core.create_image (map.width, map.height); -- for vertical in 0 .. map.height - 1 loop for horizontal in 0 .. map.width - 1 loop core.render_image (data => tiles, x => horizontal * core.base, y => vertical * core.base, u => core.base * biome'pos (map.kind) * 4, v => core.base * map.tiles (horizontal, vertical), width => core.base, height => core.base); end loop; end loop; -- for index in 1 .. landmark_limit loop core.render_image (data => landmarks (landmark_index'val (map.landmarks (index).index)), x => map.landmarks (index).x * core.base, y => map.landmarks (index).y * core.base, u => 0, v => 0, width => landmarks (landmark_index'val (map.landmarks (index).index)).width, height => landmarks (landmark_index'val (map.landmarks (index).index)).height); end loop; -- for index in 1 .. 30 loop core.render_image (data => construction.sprite (construction.enumeration'val (map.constructions (index).index)), x => map.landmarks (index).x * core.base, y => map.landmarks (index).y * core.base, u => 0, v => 0, width => construction.sprite (construction.enumeration'val (map.constructions (index).index)).width, height => construction.sprite (construction.enumeration'val (map.constructions (index).index)).height); end loop; -- for index in 1 .. 60 loop core.render_image (data => item.sprite (item.enumeration'val (map.items (index).index)), x => map.landmarks (index).x * core.base, y => map.landmarks (index).y * core.base, u => 0, v => 0, width => item.sprite (item.enumeration'val (map.items (index).index)).width, height => item.sprite (item.enumeration'val (map.items (index).index)).height); end loop; -- core.export_image (file_path); -- core.echo (core.success, "Exported current world mapshot."); -- core.dash; end mapshot;

>>39079 >It's a large image, but only takes < 700 KiB, due to 113 colours palette... why's it saved as 32-bit png then

>>39081 >>39080 I re-exported it with GIMP, disabled comments, alpha channel and everything else, it came from 661 KiB to 281 KiB, for same image size of 3840 x 2880. It doesn't matter much, tho I'll have it in mind, I use Raylib's "ExportImage" function for it, no need to link with libpng for such irrelevant difference... -- This was originally written in C, as well as 90% of my projects, tho ported it to Ada early on for fun. Ada is nice most of the time, but I miss C in few cases.

>>39050 Revised a lot of code, posting UI theme changes and code refactoring details in few posts for interested. > Imp UI theme.

>>39098 Made game source code renderer and (Ada) syntax highlighter today, but that part of the code needs refactoring. > Goblin UI theme.

>>39099 I'll grind game content for the rest of the day, gameplay implementation will be done soon. > Fairy UI theme...

>>39098 >>39099 Font's a bit hard to read, for example the O's and D's are too similar, the E's look almost like D's

>>39100 Not him, disagree entirely. Font looks A-OK to me. Os and Ds do basically always look ~similar~, xolatiles fonts distinguish them pretty well by giving the D's, who would have guessed, sharp angles and a flat leftmost line, just like basically everyone else. As for what you mean about the letter E, I think you very likely are insane.

>>39098 >>39099 >>39100 Looks good!

>>39101 There are font related things that I need to fix, as I made 2 fonts for the game, "styled" font and monospace "code" font. For example, I think that 'R' could look better, some punctuation symbols, etc. >>39102 Thanks, as said above, minor fixes will be done, but I want to start working on game content before that. >>39103 Again thanks... (: -- Here's finished Gnoll UI theme, some colours feel off, but I fixed previous UI themes, will fix this one too, and maybe revise others... -- Next tasks is grinding some pixel art, adding more equipment, materials, constructions, units (soldiers and monsters), etc. -- Image related, there are minor text alignment "bugs", that I don't want to fix right away, because I know where they appear and why, so it can wait.

>>32758 Had an idea to make a visual novel engine as my next 'intermediate level' project, a step in ambition above my previous one. It sounds like it could be doable, while promising some challenge. I have some specific ideas that I really want to put at the core. For instance, I've been trying to think of a pair of symbiotic scripting languages, similar in nature to HTML/CSS; HTML defines the content and CSS defines the form. I'll also have a VN to go along with this, I'm writing up now the scripts so I can get an idea for how I want to be able to express with the languages.

>>39104 Work in progress, as some Anons suggested, I saturated UI theme in above screenshot, added more items, slowly drawing icons for them, and hop to add water today...

>>39139 Looking good! Will the gameplay be similar to Cataclysm DDA?

>>39140 Gameplay is currently pretty much roguelike, since I make other things instead of implementing it (like a retard), like adding more UI elements and items, refactoring code and adding debug messages. I plan to make the gameplay simple and easy to learn, and complexity is built from combinations of those core gameplay elements. However, CDDA is far away from what I want... That game rocks. -- Gameplay (soon tm): You have a hero, you equip him items you find/get/buy, hire soldiers, lead army, collect on-map items and resources, equip and upgrade your army, build on-map constructions. -- So, imagine dungeon crawler, with very simple gameplay, but you're not lone adventurer, but king, leading army, building towns and forts, raiding enemy villages, managing resources, etc.

First time using Godot, i'm working on a turn based top view dungeon-crawler. In the game you're an bulldog ant that got captured as larvae into an slave-maker ant colony, your objective is to fight for revenge and escape the colony. The idea came to me seeing some ant colony pictures, they look like the perfect dungeon crawler map with a lot of rooms and corridors, everything underground. Researching about ants revealed a lot of interesting stuff about them (like the slave-maker ants), so i'll try to put a lot of different species around the colony with some different mechanics. Ants are so interesting that i'm thinking about doing some sort of strategy game (kinda like Illwinter's Dominions) with just different species of ants. Maybe some other insects in the game, too. Anyways, here's the first picture of the game, pretty basic but i can already move the main character and build levels pretty easily. I'll try to add enemies next.

>>39155 Old but gold, I grew up with this on MTV: https://www.youtube.com/watch?v=Ph-CA_tu5KA&ab_channel=RammsteinOfficial You could even simplify ("automate") some parts of the strategy gameplay, but concept sounds great. Also, no matter what happens during the development, just finish the game, don't quit midway. (:

>>39156 Thanks! Since it looks like other roguelikes, i think making a sot of procedurally generated collony would be very nice, but i won't try it for now.

>>39155 ant based clustering might interest you https://youtu.be/X-iSQQgOd1A?t=124 this dude does a good simple explanation of it

>>39156 Work in progress, a lot was added, water generation, computing earth <> water edges, major refactorings, more icons, more UI elements, game entities etc. -- I do post some progress on other image boards, but I enjoy this place the most because of useful replies and generally good community.

Given a W x W square, is it possible to draw a star using shape primitives inside the square such as make_star(color, w, pos_x, pos_y)?

>>39265 SDF's allow you to define shapes in fragment shader. SDF's (signed distance) is just a function that returns a distance value with negative=inside,positive=outside So you draw a rectangle in vertex shader and then pass positional value to fragment which allows you to render any sdf shape on the rectangle. Here's code for a star: float sdStar5(in vec2 p, in float r, in float rf) { const vec2 k1 = vec2(0.809016994375, -0.587785252292); const vec2 k2 = vec2(-k1.x,k1.y); p.x = abs(p.x); p -= 2.0*max(dot(k1,p),0.0)*k1; p -= 2.0*max(dot(k2,p),0.0)*k2; p.x = abs(p.x); p.y -= r; vec2 ba = rf*vec2(-k1.y,k1.x) - vec2(0,1); float h = vice( dot(p,ba)/dot(ba,ba), 0.0, r ); return length(p-ba*h) * sign(p.y*ba.x-p.x*ba.y); } void mainImage( out vec4 fragColor, in vec2 fragCoord ) { // Normalized pixel coordinates (from 0 to 1) vec2 uv = fragCoord/iResolution.xy; uv= 2.*uv-1.; // Output to screen fragColor = vec4(1,1.,1,1.0); if( sdStar5(uv, 0.5, .5)<0.) fragColor.rgb=vec3(0.); } you can copypaste code into https://www.shadertoy.com/new to see the result you can find a lot more sdf shapes on https://iquilezles.org/articles/distfunctions2d/

little MUSH that runs over an IRC channel done in tcl and a c stub for raylib. will add more tools for building other than "place randomly colored block" after i clean up the program some more and add state requests for new players.

>>32921 I'm thinking about a text-based rpg a-la Sunless Sea, and Löve is just the ticket. Thanks for the input.

Vulkan is painfully obtuse. Design by committee soykaf. Metal and DirectX both are so much more usable.

>>39110 Decided not to. Found out about a month ago that many of my 'literacy' ideas for the language were in RenPy already, and that some of my ideas weren't really good, as they may overdefine the lang such that it would be too limiting to be expressive. It was fun thinking about it, but I don't think it would be a valuable contribution.

Making an untitled hacking game, Ant See Sea. Right now it's UX R&D time. Watching some classic/cringe "cyber" themed media for some "futuristic" UI design inspiration. My goal is to make some kind of UI that people will think looks like the kind of CGI computer displays movies have. As a kid I thought, "Hey, that's not how it looks when hacking" or "That looks cool but computer GUI don't look anything like that"... now I ask myself, "WTF, I have hardware capable of 3D everything even in my phone... why does it have to look like 1999? Is this really the matrix?" I know fictional GUI are not actually functional in practice, (LCARS included) but there seems to be some kind of vague predictive-programming / propaganda in the media that people will expect user interfaces to look like in the "future". My game is UI heavy, and needs to be usable, but why not make it look like the (albeit silly) SCI-FI movie special effects? As for progress: I've got a low level (register) virtual machine, assembler and linker working. I'm almost satisfied with my chipset. The VM will be overhauled (again) soon to adopt a new style of jump-table/switch that is quicker than all other interpreted VM lang methods I've tested so far (hint: where we're going we don't need a LuT). Since players will be hacking programs/hills running in the VM, security is a must -- I must make it impossible for players to escape the VM and get at the host PC. I may be able to make the root game code secure with a sanitation layer -- Provide access to "devices" or other capabilities via a sanity checked protocol that limits functionality -- EG: If you want to draw to a graphics buffer, have to use RPC provided methods and send "command lists" not a shader (thems givez up teh rootz to IRL 1337 h4xx0rz). The idea is that the hacking game's platform will be usable to make other games so the speed and usability is more important to me than would be otherwise for a text/[G]UI heavy hacking game. Once the base subsystems are in place, proper gamedev can begin. Maybe I'll use an AI generated art as a muse...

>>39766 You could make data blocks represented as city blocks kind of like the Geocities map but 3D. http://www.deletedcity.net/

Finished watching Hackers, binged S.E.L. (again), and finished off with Time of Eve (Ivu no Jikan)... (Chobits img was a [bad] joke about "optimal" GUI [robowaifu] -- but we do now have "Android phones" that are kind of like Sumomo was a "prediction" of). Next in the "R&D" queue I've got Ghost in the Shell "SAC 2nd Gig", "Girl's Last Tour", and a James Bond flic "No Time To Die." (even if it IS just predictive programming propaganda, the Q dude got a Dell {and a viral Ubuntu reference?}). >>39768 Thanks for the link, pretty inspiring. I like the idea of mapping a virtual space to the "hills" one might choose to attack. I have a vague design requirement where "cyber territories" can be infiltrated incrementally, but don't have a good idea yet what that intails... I guess it would have to be a network of connected hills, or capturing a flag might require access to surrounding "cyber-space"... the city blocks idea/impl is def something to mull over more. Currently I have designed a requirement for each "hill" to have a "flag" (nonce) associated with them which is basically a server-secret-key. There is a wire protocol for publishing the hill data blob (VM image) which allows the hill host to verify the flag exists, accessed by a "single use" authentication token run against the VM "hill". The published hill's state will be frozen (VM image archived) after the single use auth token verifies the flag. My initial design was: Hills are worth N "Cred", which is a bet that the hill creator places against a temporal duration, i.e., > "6 of my Cred goes to any crHacker who captures the flag in 0.25 days of publishing." When the flag is captured, the advertised Cred will be transferred from the hill author to the crHacker via a "hacker-proof" ledger (updated by the host system code that runs the game itself). Inital creds are granted via unlocking them as a player/crHacker captures hills in the single player (offline) game. This allows newbs entering the online game to have some number of creds to feed the game's cred pool. The crHacker's job is to create and publish a (signed) input data stream that when run against a published hill's state will cause the flag data to be sent across a data channel (ostensibly sent to a node controlled by the crHacker -- or their faction?). Each data "vessel" sent from a hill while doing the input capture's replay will be scanned for the flag data to determine captured status. Anyone can verify that the flag really was captured via event replay against the hill's VM. The idea of territories will have to be developed. A mapping between "cities" and their infrastructures could be interesting. (virtual) Network exposed Infrastructure controls could solve the design problem of "why do hills exist, what purpose do they serve". The exposed ports of a hill could have "vessels" (virtual data packets) routed through the hills to provide some form of infrastructure maintenance actions -- but I DO NOT want to get deep into (sim)City emulation territory -- that niche is full and boring IMO. Perhaps some other futuristic world crafting will yield infrastructure requirements that would be more fun than water/waste management, traffic controls and <tinfoil>power production/regulation via wet surface air cooling cloud generators... </tinfoil>

>>39770 recommended materials attached. don't get too carried away formalizing point-scoring. much of the appeal of hacking comes from exploration and clever experimenting. computer networks are just a bunch of computers connected through wires. hacking is just getting computers to do things their supposed owner doesn't want. maybe a game like telehack except with more nuanced gameplay and graphics is a good idea? the hardest part is going to be bridging semi-realistic hacking with user interfaces that don't even exist yet outside of media. a lonesome phreak sits at night sending tones over wires

>>39771 > don't get too carried away formalizing point-scoring Well, since I want to also make it a multi-player game I need a solid/secure base system for scoring / bragging rights. More complexity can pile upon that and scoring may be secondary to the cracking/hacking action. My fear is a few good hackers pwning everything, cred cost's ulterior motive is to rate-limit the best crHackers. > much of the appeal of hacking comes from exploration and clever experimenting. True, and thanks for the reading material. I have somewhat limited experience cracking/hacking on todays Internet/web, but in my youth I often cracked Wildcat and/or Renegade BBSes to get sysop/1337 access just before the web was rolled out in my city (it was actually present time of SEL airing). Back then there was always the looming threat of the BBS sysop discovering the "hack: in-progress", since many sysops would leave a terminal on to display the view of the current user, some BBSs cycled between multiple-user displays. Things were much more intimate in those days, the sysop/board admin personality was apparent and they were close at hand, usually just one "page sysop" option away. I would hear the PC-speaker beep paging me and roll out of bed any time of night to answer most any request... usually I would have to dig into the "catalog" (a big cardboard box full of poorly organized 3.25" floppy disks - kept under my bed). I'd mindlessly search for the disk ID# shown in big block ascii letters on screen, then shuffle over to the node & slot in the disk that held some obscure file the user wanted. When "sysop wants to chat" they would break into the connection, i.e., the screen would split in two with sysop text area above and user below. The chat characters were transmitted in real time, so one could watch the letters appear as the other person typed, you could get a sense for (typing) skill level and since I was usually the faster typer I would complete the other's sentences. I got caught quite a few times and "scolded" by all-caps misspellings of curses and slurs. Fortunately no-one ever called the police, not that they had much in the way of digital forensics back then to make a case stick. CFAA was just becoming a threat. Back then hacking was oft seen as harmless fun, few crackers would do more damage than defacing an ANSI-art scroller (to secure bragging rights). Nowadays hacking seems far more impersonal. Hackers actually use tools they didn't build themselves to exploit vulns they didn't discover and deploy attacks against random hoards of nameless machines at addresses most of which they'll never personally visit -- the action being launched from some remote compromised system, perhaps several hops removed from the hacker's main terminal. To me, this sort of common non-targeted hacking is a bit lame, but I suppose it can be profitable. I miss getting to know my target, shadowing their moves via covert digital real-time logs, planning my attack and timing it to evade detection. A big part of the (hacking) scene for me was creating things. Esp. tech demos of effects (ala cracktros - see:demoscene), and new hardware - Eg: 4 walkie talkies w/ 5mi range: Two with the mike keyed / button taped closed "talk", other two always open "listen". one talk/listen on A channel, the other pair on channel B. This full duplex 5km intercom system was how we did teamspeak/in-game voice-chat while playing on Dwango (4 player Doom [Novel IPX net sim] via BBS). Just add a tone generator circuit and follow the RJ11 pinout to connect 6P6C connectors where the mic/spk posts connect in the respective handsets. Jack in two modems, Send ATD to one modem, and ATA on the other and you've got a cheap and illegal pirate wireless modem with 3km/5mi range. Screw that "hackers" move of taping the handsets together, we've got an outdial hooked up to makeshift wireless that's connected to a train crossing phone-box. WireTrace me all you want, enjoy catching the train. Let's not talk about cat&mouse HAM frequency hopping -- FCC outlawed store-and-forward packet routing over HAM and CB / family bands or else the internet would have been stillborn - leapfrogged by a wireless Fidonet implementation. As for initial exploration I'll be gamifying the in-game OS/shell. I'll have to backfit a world-crafting idea into why the programs/sources are initially hidden away &| unlockable. One design idea that keeps resurfacing is that hills require some form of input, processing and output -- The hill can't just be a while (true); loop that does nothing. I need to make the complexity requirement high enough that bugs will likely exist in the hill's code. It's probably a non-starter if crHackers must create something on their own that is too complex for a hill -- then it will lack the instant gratification required of early game loops, and basically just be programming work. Perhaps I'll create some really insecure base starter hills that the players can modify and publish. For smoke-testing purposes I've been building simple BBS-esque interfaces on the VM already. Thus the VM-image's program logic is nearly the simplest of IO: port_read() in a loop followed by a switch/case full of port_write()s to display data/perform various actions (in liew of any terminal or OS [don't exist yet]). Gopher like services are too easy to secure. As cliché as it is I'm really starting to lean towards hills just being a form of BBS, possibly with "door" game facility (with the vessel/port API a "door" would be just a VM that starts another VM image - primitive micro-service architecture...) Recently tried a gamedev IRC, no intelligent life found. I got some serious flack & blowback when I said I was creating a chipset to implement an OS and game atop that... I guess I just hit a "enginedev" raw nerve. It seems the "Hacking" part of the game design was glossed over/ignored and these "gamedevs" just dunked all over my core design, quote mining everything I said... It almost seemed AI generated at one point - like these all these lurkers were just idling... Then tens of commenters piled into a conversation they didn't know anything about just to meme "ideasman" junk at me... as if I've never hacked or written an (alternative) OS before. Oh well, I left them with these kind words: "flaming lamers, the lot of you. EVERY SINGLE GAME was an OS back in the 8 bit era, your game started at boot loader. It is not hard to build an OS, it is bog standard for any old arcade cabinet or floppy disc game to provide its own OS services." Oh well, maybe I have bit off more than I can chew, back then games weren't usually solo works and they did have middleware (esp. on t3h c0p13 pR073k), but if it is too much for me, it'll just be my go-to hobby project on through my retirement. Pic: Scanlines test w/ scanline darkness compensation, rendering CP437 glyphs in emulated VGA textmode 9x16. Scanlines are def something all the cool hacker's UIs have. > hardest part is going to be bridging semi-realistic hacking with user interfaces that don't even exist yet Well, they all still have text command terminals... I have done some 3D GUI work before, have a 3D UI engine that allows picking of arbitrary shapes, reposition/drag etc. Mostly the "futuristic" movie GUIs are concerned with status/information displays. Making (3D) graphical displays isn't hard, putting useful information on screen and making it interactive in a useful way is the hardest part, IMO. I may back off on any 3D elements for now. I'm feeling "retro" at the moment, and textmode still has quite a bit of meat left on the bone. I was digging in my ancient projects and found my old BBS's prototype BBS-client that rendered the "ANSI art" client side (w/ "iCE Colors" enabled). It had animations and buttons... hmm.

Are there any libraries for C that are like SDL but header only? I hate dealing with build systems and the complexity surrounding them, so if there's a way to just #include the library and have my program compile, I'd love to know it.

>>40267 Just stop being a retarded luser. SDL2 ships with pkgconfig files which make everything easy as hell (compared to cmake and co.).

>>40267 I wrote "header-only 2k sloc Raylib minimal clone" for XCB+Vulkan, and another experimental one for GLFW+OpenGL, but: > no 3D support, only 2D, no sound, very very minimal in design, made for my own usecase, etc. I want to properly finish them before "sharing" them, and it's "header-only" in a sense that you have to '-lGL -lGLEW -lglfw3 -lpng', so those 4 are dependencies for that GLFW+OpenGL one. There's no shortcut in life for anything, if you want to do graphics you need to suffer, and do Khronos's humiliation ritual of writing 1000 lines of code in order to render a single triangle on the screen. Also, Raylib is better than SDL2 for small projects that you want to quickly finish, for example, you have a game idea, you download some assets from 'itch.io' or 'opengameart.org', and you make it.

>>40267 I think sokol allows this. https://github.com/floooh/sokol Also you can link a dll without using a build system, just supply the cli options directly to the compiler. g++ -o out.bin main.cpp -Llib -ldl -lSDL2 -lSDL2main -lSDL2_ttf -lSDL2_image -lutil

I work on a pretty big game, i would like to do a haloweed suprise demo release but i dont know where to upload without hastle, its over 1gb now. Would be nice if i could get stuff up so people could play it on haloween. Its pretty halfassed and lots of stuff is either broken or bugged, but it has some fun as hell gameplay elements and i want to put in some extra haloween stuff. Theres still some stuff im not done with yet but i hope i can whip it up quickly tomorrow. If anyone has interest in a setting something up for haloween we could have some good fun.

>>40271 It probably is simpler than Cmake and friends, but I dislike anything more complex than #include-ing what you want as a matter of principle.

>>40275 I'm not using Raylib for the same reason I'm not using Godot; I've got massive NIH syndrome and have a much easier time writing my own systems from scratch than trying to figure out someone else's. If portability weren't a concern, I'd probably work with system-specific APIs directly.

Does anyone have any Godot resources? I'm switching over from Game Maker to it, largely because I'm sick of dealing with closed-source engines, and compared to Game Maker, Godot feels a lot more complex and intimidating. I'm not really sure where to start with it, so if you can point me to anything about working with the engine, particularly for 2D gamedev, I'd love it.

>>40345 Did you try reading and following along the docs? https://docs.godotengine.org/en/stable/getting_started/introduction/index.html https://docs.godotengine.org/en/stable/getting_started/first_2d_game/index.html Or any of the tutorials on the internet? https://www.gdquest.com/tutorial/godot/2d/first-2d-game-godot-4/ https://www.youtube.com/playlist?list=PLfcCiyd_V9GH8M9xd_QKlyU8jryGcy3Xa And just ask ChatGPT and Gemini if something doesn't make sense.

>>40346 All of the Godot tutorials I've found so far have been too tutorial-ish for me. They tell me what to do to get a working game, but not why it works, and I don't feel like I've learned anything. The obvious answer is to look it up myself, but search engines have become such horseshit that it's really annoying and time-consuming to find what I'm looking for on my own nowadays.

>>40347 If you wanna know core ideas and general design concept of godot i like this one: https://www.youtube.com/watch?v=LOhfqjmasi0&t=332s 1h tutorial includes explanations on how godot is designed to work as it goes through the stages making a mario clone. Also covers how 2d sprite sheets work and how to use 2d layers in godot. Most efficient entry point iv'e found when i was looking for 2d guides.

>>40349 Thank you. I'll check it out. Not to be picky, but do you have any text-based tutorials? I prefer to learn by reading. If you don't, it's fine.

>>40351 sry got nothing textbased. First steps with new software is learning UI, where and what menu items to click is easier to record with video than screenshot step by step so i figured looking through videos would net faster a good result.

>>40277 here, i got things done late, it was tedious and im exhausted. I got something done for now, you can get it at https://big.fileditchstuff.me/b45/hDDuVkZHzPjhyMPIRmfh.zip Its pretty buggy, but stable and somewhat playable, but also needs a somewhat beefy box. Enjoy the spiders and soykaf, il post a more proper link another time cuz reasons.

I'm at the point where I'm honestly considering taking the plunge and just writing my own engine. The engines most suited to the kind of thing I'm trying to do, GameMaker and Clickteam Fusion, are both encumbered with proprietary licenses, which leaves Godot, which is, at least in my opinion, way too complex.

>>40476 >just writing my own engine. >reinventing the wheel...again Fine to learn about wheel building from scratch, but for putting out a game it's a waste of effort, unless you're doing something very basic. Tried raylib engine?

>>40478 I'm going to be using prexisting libraries to make the hard parts easier, specifically Flecs, SDL, and the stb headers. It helps that I'm looking to make an engine for 2D tile-based SNES-type games specifically, as opposed to a general purpose do-anything engine like Unity or even Game Maker. >but for putting out a game it's a waste of effort I'm planning to make multiple games with it. I'd probably use an existing engine if I just wanted to make one game (although I wouldn't be thrilled about it!), but since I have enough ideas for at least four games, I think the upfront effort is worth it. As for Raylib, I'm not super familiar with it, no. Isn't it like LOVE but for C instead of Lua?

I am stuck as stuck I can be. I want to make a hacking simulator and I have made almost everything, technically speaking, but now that I have the base to start working on the story I don't know what to do or how to do it. Anyone has tips on how to develop a story or any resource?? Would be appreciated

>>40492 So, the first I would try to do is to pick up pen and paper (physically or digitally) and start writing by writing keyword of themes, characters, environment, cliches, etc... make a huge brainstorm of what your story should be about and what you want your story to have in it. After getting around a few concepts you think could be solid, start by writing the base of your story. Basically world-building. Setting up the context of the story, the what, when, where, etc... Be careful of "feature creep" in your story. Add elements that you think are integral to the story, first. Put gameplay aside for a moment and start thinking like a writer. So, you'll want to start writing drafts of your story. The introduction, the element of conflict of your story, the progress to solve the conflict, and the resolution. Basically the arc of the story. Iterate on it a few times until you are satisfied. After that, you can start to properly write your story. Get a little more granular on how things play out. At this point you can start to think about how your story is going to play out in the game and how the player will be able to interact in the story with the gameplay. You can take queues from other similar games in the genre you are making. Hacknet comes to mind, personally. The story doesn't have to be deep or long. It depends on how far you want and are willing to take it. Take your time with it. This process may be longer than making the basics of the game. The UI looks pretty neat, btw. :)

>>40493 Thank you for the inside chummer. I will try to do that, I already have a few concepts I want it to include but no clue how to make everything fit together. Thank you

>>40492 Just throw soykaf at the wall and see what sticks. If you're stuck, just go with the first thing that pops into your head, even if it's awful. The important part is getting out a rough draft that you can use as a starting point. Writing is a lot like gamedev, in the sense that it tends to be very iterative. It's exceedingly rare that you end up with a finished story by starting at the start and ending at the end. Instead, you begin by figuring out the broad strokes, and then add, remove, reorder, and otherwise modify things until you're at the point where you have something you're satisfied with.

>>40492 A person is either a programmer or an artist... A person is either a male or a female... -- If you don't have the story, talk to a writer...? I have similar issues myself, my art and story suck. But if you, like me, want everything self-made: > be ready to sacrifice quality for pleasure

>>40509 >A person is either a programmer or an artist... Not even remotely true, what on earth are you talking about?

>>40492 why not look at existing games? exapunks is a programming indie game with a hacker storyline. Bitburner is a free game about hacking. both basically do the obvious: you start hacking cause you poor and need money, inevitable someone else takes note of you which kickstarts a plot. Could be other hackers thinking about recruiting and giving you jobs / telling you to git good first or some corp telling you to go after a rival if you dont want cops on you.

I've been trying to get SDL2 up and running with Code::Blocks by following https://lazyfoo.net/tutorials/SDL/01_hello_SDL/windows/codeblocks/index.php But every time I try following his instructions, the compiler spews a million >undefined reference errors at me. Can someone please help?

>>40545 linker error, possible causes: 1.didn't link library / wrote name wrong / lib path is not in environment path 2.linked wrong library (version of header and lib dont match or 32 vs 64 bit clash) 3.system lib is used over your own lib and its causing 2. looking at your linked guide they dont declare a lib path explicit and it seems to be windows, so i guess you got 1. going and have your lib path unknown to your OS. Just add the library directory and see if it works.

>>40546 Thank you. Is setting this kind of thing up easier on Linux? I'm using Windows because that's The Gaming Platform, but I also have a laptop running OpenSUSE, and programming on it has always been infinitely more pleasant.

>>40549 Depends on toolchain you use. I also had a worse experience with small projects on windows while linux just works thanks to package manager but if you use a big framework like QT or an engine like godot you wont have to deal with config issues since they are self-contained. That said there is a hidden variable im not sure of - once you setup a windows install manually a lot of the issues vanish and you cant tell how bad it is on following projects. The problem is also such a setup is not intuitive. The worst problem i had was a software that kept crashing til i had the idea devs likely had visual studio installed - installed VS and suddenly it worked. So they didn't notice their own version was broken cause their software was relying on files from VS passively replacing what they shipped.

>>40546 >>40549 As it turns out, the problem was that I had static linking flags set to on by default for reasons that evade me at the moment, and it meant that SDL.dll wasn't loading. Turning static linking fixed the issue immediately, and I'm slightly embarrassed that I had a problem like this to begin with.

>>40554 >>40545 >>40549 Are you happy with Code::Blocks or would you prefer using Notepad++ and installing SDL2 through MinGW (MSYS2) instead?

I'm pretty competent at coding, but when it comes to writing, art, and music i have absolutely no skill. I can't make a game without these. A lot of people are good at one of these things have seemingly been doing that thing for years in order to get good, let alone getting good at all three. I feel like it would be either impossibe or take years to get decent at all of these aspects. How are all of you able to get good at all of these different things?

>>40573 there are programming gods that can't draw a face to save their life, others try their best, a lot of us are really only good at one thing but have to acquire secondary skills to support the main thing. the worst thing you can do is tell yourself "i'm not an x i only do y". branch out, maybe start with motion graphics and stock sounds, other people's assets, until you get a feel for how they work

>>40573 >writing, art, and music use AI to generate it

>>40573 Simply put lainon, you'll notice that solo game developers that make games that are popular and enjoyed by many people are not typically as young as many of the people you think are very good at all those things. Most indie breakouts or what have you are 30sish precisely for the reasons you state

>>40581 i think exceptional indie devs were more around 25+ and that's merely because of 5+ years development time. Typically art skills stem from drawing during school time since elementary school. Its not trained for gamedev but something you already had. Similarly reading a lot of books maxes story just a matter of mixing what you know. tldr: outstanding indie devs simply rely on life choices made way before they wanted to make a game. Regular new gamedevs don't waste years learning art&music, you rely on people who focus on that. There are tons of free assets available and if you manage to produce something worthwhile that can earn money you can hire people to make unique assets for your game. That aside >>40573 mentioned concern for everything but gameplay quality. You ready to get hired by AAA studio ;D

>>40573 >I can't make a game without these Have you heard of Rogue or NetHack? Work around your strengths or razor-focus on a particular vision and do everything practical in order to achieve it. Broaden your horizons in observation, but not in applicable skills. Most notorious indie devs don't work alone or have their first breakout hit at the average age of 33.

>>40579 Thank you very much for the advice! I'll try to branch out. >>40580 The whole point of me wanting to know how to do these things myself is so that I can have creative agency in what I make. I dont wanna use AI for the same reason I dont wanna commision an artist: It severly limits my agency in the creation of things. >>40581 I dont expect to become some famous break-out hit indie dev or anything, I just wanna be able to make something people can enjoy, but it is nice to know I can still learn these things (even if it takes me till my thirties) >>40588 I didnt show concern for gameplay quality in my original post since how fun the gameplay is mostly comes down to knowledge of good game design and good planning, rather than a rigorous skill you have to practice like art or music. You can read and study for gameplay quality, but not for the arts.

>>40595 >You can read and study for gameplay quality, but not for the arts. I see it the exact opposite way. You can upgrade artistic skill methodically studying anatomy etc. There is no clear established path to good game design. Its why so many games independent of budget produce crap and AAA studios trying to steal customers from hits consistently fail despite 600m budgets.

>>40596 Studying definitely is apart of improving artistic talent, but practice is at the core of artistic talent, and is nothing without it. I am aware that there is no clear established path to good game design, as every game is different, but that doesnt mean studying game design is completely useless.

>>40573 Studying game design isn't about reading. It's about playing a fuarrrkload of games and getting good enough at each of them and genres in general to kow them inside-out as well as what they're lacking, and to understand good game design is mostly drawn from outside ideas and never autistic replication of the meme-in-itself. Iterating on a project to the point it's a good fuarrrking game is just as laborious as practicing art, music or writing. And it requires a background of critical introspection and hours wasted with video games that only a very tiny portion of the population has.

>>40598 Biggest problem with game design is meeting player skill. A puzzles difficulty has to meet the audience ability to solve it. That's a unique issue other departments lack. You dont get punished drawing better than your audience but if you a top 20% player and design a game for yourself you filter 80% of the potential customers. And whats really hard is even if you know you a top/bottom outlier - its not helping you find the right difficulty appropriate for mainstream.

>>40573 Just go for an artstyle you think you can do. A lot of indie devs go for super simple pixel art because it's hard to mess up.

>>40564 Happy enough. I use Code::Blocks mostly because it was the easiest way I found to get a C compiler up-and-running on Windows. I'm not very good with the tooling side of programming. It's something that's always evaded me, and I've gotten by on a combination of luck and following tutorials by rote.

> Graphing some function to visually debug it... The graphing code is also buggy, Doh! On tigsource forums there's a "beautiful mistakes" thread that Lain would appreciate. [pic related] Lainchan is probably too slow to have a thread for that, but it reminds me of "data bending".

I'm learning SDL, specifically SDL with C, and I want to know if there's a way to separate event-polling from drawing to the screen. I'm trying to make a game with logic that runs at 60 ticks per second, but with an uncapped framerate.

>>40853 >uncapped 6k fps is how you fry gpu. Always cap fps. For async work split use threads

Glad someone bumped, sometimes i forget about these threads. I would like to share this project i work on, its a remake of the 96 rts classic game red alert, but in full 3d. This is an enormous project and took me like 4 years by now to go where i am. I was hoping to have a new demo ready, but its not quite there yet and i might need a few more weeks. New levels need work still and start positions are fuarrrked so its a no go until that is resolved. There is a halloween demo linked in the middle with the pictures of the spiders at: https://torque3d.org/forums/topic/12398-rad-alert/page/2/ Its not a very good demo, has only one level and not really that much functionality, the new one will be much better in so many ways. Im $hilling(do not say that word???) this around on a bunch of places, i want to build up some interest in the long run. Also everything in the game is copyleft, including all assets and everything, free to use for any purpose including commercial. There is still a very long way to go but its going pretty darn well so far considerening what a huge undertaking it is and how much luck i had so far.

>>40853 What you want to implement is called client side prediction. You can check the current time and, based on velocity or other deterministic systems, render based on delta from past phys tick and current time. This is why a lot of physics runs at 30fps regardless of screen refresh. Use the delta between last phys update and current time to position your objects. One winds up drawing positions at the interpolated location between last viewed and (authoritatively) predicted position. To avoid teleporting and jitter one induces "rubber band" physics, but this should not actually bounce, simply seek between current and [best] projected future. This way rendering is decoupled from phys tick. There are other optimal things to do to eliminate the concept of FPS and optimize GPU usage... but these (driverdev) issues are beyond the reach of your solutionspace to solve.

Can someone please help me with 2D collision response? I'm building my first ever custom game engine. I'm using SDL for cross-platform stuff, but otherwise, I'm building the whole thing from the ground up. Right now, I'm trying to figure out collision response, and it's the most infuriating thing possible. Every time I think I have it right, some new problem comes up and everything falls apart. Currently, I handle collision with the SAT and quartersteps. Each object moves 1/4th of its velocity, and is then checked against every other object in the scene (this is obviously kind of slow, and I'm going to implement a lot more broad-phase stuff when I'm done figuring this part out). For each object, a list is compiled of every other objects it's determined to be touching. The object is then moved by the sum of the MTV of every object in the list. This process then repeats three more times, and then the frame is drawn. This works when objects are interacting with a purely static environment, but as soon as moving objects collide, everything starts behaving like nonsense. Does anyone have any suggestions?

>>41278 Sounds like any object faster than 1pixel /check and you will get collision with obj stuck inside each other which causes your issue. If they are really fast they will just tunnel through thin objects like neutrinos. In other words to make it work without issue you would have to sort movement vectors and run the simulation at a step resolution where the fastest object travels 1 pixel/step and the rest get scaled to its speed. This could work or kill perf too much depending on speed of your obj and hardware. The cleaner more established way is calculate collision paths with bounding boxes. Aka predict collision instead of just checking if it happened. I.E. if its a bullet just turn the dot into a long line for a collision check according to its movement vector.

>>41290 My main problem is less tunneling on fast-moving objects, and more that objects act weird when interacting with multiple things at once. For example, if an object pushes another object into a corner, rather than the pushed object staying pressing against the wall, the object is either flung upward, jitters back and forth, gets stuck in the wall only to pop back out when it's no longer being pushed, or some combination of the three. I'll try implementing your solution to see if it fixes this, but if it doesn't, or if it's too bad for performance, I'll have to try something else.

>>41293 Would guess your touching object list isn't ordered, so a wall zeroing movement can be applied first, next object applies some force = wall ignored. Order the list and apply walls last should fix what you describe.

>>41294 Thank you. I'm one of those people who's way more comfortable with computers than he is with math, so a lot of this stuff doesn't come naturally to me.

>>32786 do people on lainchan even care about trannies? i'd imagine most normal people don't give a fuarrrk. >>32758 i like godot :) but i also don't make game much these days. i still would like to work on something in C or sepples with sdl but that seems like ten times harder than just using godot and it's integrated language. maybe when i'm older and have more time or videogames and books get boring for me.

I'm cooking something simple in Godot but most of the advanced resources I've read seem to assume I know what I'm doing, stuff like shaders, vectors, pathtracing, etc... are alien to me, is there a book about the basics of game dev?

I am currently doing a renderer/framework with raw OpenGL and GLFW in rust. Not strictly a game but the plan is to get some basic rendering going and do physics/particle simulations.

>>41679 i think you can skip learning details for godot because i learned everything and didnt get to use it so far using godot, tldr for your questions is: Shader = program running on gpu vertex shader: triangles -> pixels (you can add additional data to the pixel data for fragment shader) fragment shader: pixels->pixels (godot/blender graph editor effects are in implemented as fragment shaders) path tracing is ambiguous because different people use same term, for your usecase it's likely the raytracing replacement. Just means "fast" realistic light rendering for you

Any definitive book for essential linear algebra and calculus required for gamedev? These matrices and 3d projection are really making my head hurt. I really should have paid more attention in college. fuarrrk.

I'm not making a game from scratch but I'm reverse engineering one such game and modding it

>>41677 >do people on lainchan even care about trannies? i'd imagine most normal people don't give a fuarrrk. I've always been under the impression it's one of those things that a small minority but vocal minority of culture warriors spend all day obsessing over, and everyone else just kind of whatever on. That said, you'd be surprised how much people tie their self-image to retarded things like what programming languages they usd. Rust has this reputation for being the language for pussies, and people will take potshots at it because they don't want to look weak, even though they're fuarrrking computer programmers.

>>41712 I am strong and handsome irl, and I would NEVER use Rust

>>41712 rust "memory safe" translates to "made for codemonkeys" and then they still got mem issues cause as always the idiocy was stronger than the engineering. And they act like a cult and try to inject their soykaf in existing codebases replacing other languages. Any beatings they get are deserved.

Hey lains, happy 420, im rush releasing a demo of my game called rad alert. I will work on that for a long time offline, I will provide updates when i can, but the detail work will take a long time. Check it out at https://torque3d.org/forums/forum/70-status Its all copyleft(mit x11/cc0/public domain) so take whatever you like for yourself. Its really rough, but hope you enjoy it still.

>>41704 This. I'm trying to bridge two DOS games and create a 3d object/animation extractor , but it's a back burner project for now.

>>41794 Oooo that looks awesome lainon. I'll check it out when I get home later, thanks and happy 420 to you too

>>41714 >"memory safe" translates to "made for codemonkeys" I don't see how attempting to be memory safe is a bad thing. Anything that makes programming less error-prone is a win in my book. Again, it goes back to that "nooooo, I'm not gay! I'm a real man who writes everything in C" soykaf I was talking about before. I want to make it clear, by the way, that I'm not the biggest Rust fan. It's one of many attempts to "fix" C++, and probably the most successful in terms of technical improvements, but it misses the underlying problem with C++, which is that it's enormously, impenetrably complex. Correctness, in my experience, is easiest to achieve when you know every core construct that exists in the language, as well as how they interact. And Rust is simply too big for that. You need to understand 15 layers of language features before you can implement a linked-list, and the result is that you're never quite sure what your code is actually doing. I'd love to see its better ideas in a language that's less bloated.

>>41853 restrictions limit how you can write code. Security is purely a cost until necessary. In other words if you aren't part of the crowd creating memory bugs, then its just limiting you for no benefit. And note the whole point of "foolproof coding" is to lower the required skill level so they can hire worse people for lower pay. Its just industrialists trying to get rid of paying high wages for skilled people by lowering the standard. Anyone competent should be against it really.

>>41859 I guess I'm wont to compare Rust to the likes of Java and Go, even though most people (including the ones making Rust) see it as an alternative to C++ more than anything. Rust, in concept (as opposed to execution), excites me because it allows high-level code to take reasonably good advantage of the hardware it's running on. It's still got a bunch of runtime overhead, and the way it's designed doesn't always lead to optimal codegen, but it isn't wasting its time doing nothing as with a garbage collector or, worse, a virtual machine. I absolutely agree that it doesn't belong in kernel space though. Also, I posted this to the wrong thread and deleted it, and now the spam filter won't let me post it here.

>>41853 >when you know every core construct that exists in the language >I'd love to see its better ideas in a language that's less bloated >You need to understand 15 layers of language features before you can implement a linked-list You have memory, you have numbers in the memory and you have addresses to the memory. Then you do things with the numbers. And C gives just the right amount of syntactic sugar to make that fun. Any layer put on top of that has been and is lowering the skill requirement to flood the workforce with disposable Indian codemonkeys and trannies.(USER WAS WARNED FOR THIS POST,Always argue in good faith and avoid using personal attacks.)

>>41886 >Any layer put on top of that has been and is lowering the skill requirement to flood the workforce with disposable Indian codemonkeys and trannies. The skill requirement had already been lowered long ago by by Java. All Rust does is do what other high level languages do but better (in terms of performance usage and performance specifically). And why is that a bad thing Why is that a bad thing? Tools exist to automate and otherwise make easier work. It's why we use compiled languages to begin with. Again it's just a bunch of Real Programmer horseshit. You don't like C because it's the best tool for the job, you like C because using it strokes your ego. I like C too, but I'm not going to pretend like using it makes me more of a man. It's a fuarrrking programming language.

>>41887 java and C dont compete. They are used for different things. Rust wants to replace C. If i want something quick and dirty im gonna use java/phyton. If i want low lvl i go C. Rust has no reason to exist beside trying to create a retard proof C replacement which as stated already failed as they still managed to create bugs it was supposed to prevent...

>>41907 >Rust wants to replace C. More like Rust insufferable fanboys want to replace C++. The creator and the community of Zig are the ones who have the delusion of replacing C. I must say, this "let's replace" mentality is immature and plain stupid; but is far easier to burn all up and start from zero than trying to find solutions and improve technology. It's just a dumb battle of egos between engineers to see who has the biggest d*&ck.

It's been awhile since I posted an update here, instead of going over everything that's been done I'll just mention a couple small things i accomplished today. In the first video I improved the looting UI popups and animation of the crate sliding in under the ship rather than just vanishing. In the 2nd video and screenshot I added some meat gib particles for when you kill escape pods (screenshot gibs are lit by the ship's headlight). I also improved the particle effect.

Is there any prior art for real-time subdivision surfaces on the GPU? I ask because I want to make an engine that uses adaptive subdivision for its graphics. The idea is that each polygon gets subdivided until it's within a maximum acceptable size in screen-space, and/or the next iteration wouldn't noticeably smooth the model. Said maximum size would be determined by checking what the hardware is capable of, but I'm not quite sure how I want to go about that part yet. Apparently, people have done similar things before, but I'm not aware of any shipped games that have done it, and if doing so means making games that nobody can play then I'm not as interested.

>>42188 geometry shader exists for this. Shader pipeline is vertex->geometry-> fragment. Geometry shader generates additional geometry before it gets drawn & shaded in fragment. Note that subdividing a plane doesn't add any detail. That's why you use a height map texture to move the generated geometry around.

>>42189 I just didn't know if it would be prohibitively expensive on weaker hardware or not.

>>42190 The original purpose of this approach is meant to be a performance boost. Especially when the bottleneck is cpu-gpu connection. You bypass the issue here since you push smaller models and then inflate them on gpu side.

>>42197 Okay, cool. My goal with the project, at least on the graphics side of things, is to make an engine that approximates the visuals of a prerendered cutscene while still having relatively modest hardware requirements. I figured that adaptive subdivision surfaces were a good way to achieve it, since even on weaker hardware where they can't be subdivided as much, they'd still mean that polygons would be be split up in such a way that the player wouldn't be able to identify any specific surfaces across different views of the same surface.

>>40492 Hey this looks kind of cool. what's the name of the game?

I dont know if it counts as gamedev but i made a minecraft launcher! its made on python and Qt for the gui, there are a lot of launchers already but i honestly didnt like them or they were straight up malware https://github.com/nixietab/picodulce also first time posting on lainchan!

What may be the simplest way to set up a draw and input loop? I don't want to have to find a big library like raylib or sdl just to draw a grid and accept keyboard input. I don't want to use xlib because it may require too much boilerplate. I'm using common lisp btw. >>41703 4th edition is on libgen

>>43247 Why not try SDL 1.2? It has far less features than SDL 2 but can get what you want done. I loved using it for toy graphical things. https://www.libsdl.org/release/SDL-1.2.15/docs/html/index.html

>>43247 If performance matters, use GLFW. It is a library that provides extremely thin wrappers around OpenGL and native input APIs. From an implementation perspective, it's about as simple as you can get without communicating with platform specific APIs directly. Good on you for using Lisp, by the way. People think it must be slow because it's dynamically typed and GC'd. Those add overhead for sure, but in my experience, if you know what you're doing, a game written in Lisp runs about as well as a game written in C + a scripting language like Lua.

>>37255 >Lisp Do you have any suggestions for this? I've been thinking of rolling my own space station 13 since byond is like kicking dead whales down the beach and so does space station 14. Thought I may as well use a lisp dialect or another to do it, but what's gamedev look like on Common Lisp these days? I think I'd wanna have ECS out of the box at least since I've never really gone balls deep into UIdev/gamedev in general. I'm also legit considering copying the thing where ui elements are rendered as html and then buttons are just hyperlinks/javascript elements because I do want to have an 'easily' themable ui that lets me colour/embolden/set fonts arbitrarily and as far as I know that might be the best way to go about doing it? I don't think this is something most games would even go for, but then css theming and the like stand out to me as the easiest way to ensure I'll be able to have a chat window people can easily copypaste and fuarrrk around with. I do think that this is all kind of a pie in the sky idea though, the game I wanna make will be very complex even if I scale it down a lot, and I don't really know too much about networking. I've at most written a basic udp echo server up for myself once. Mostly because I think that since ss14 already exists, it might be wiser to just try editing that, but I don't actually understand how it works and it suffers a lot from there just being heaps of soykaf that don't really go together intuitively. But then it has a lot of systems I could modify instead of doing from scratch and using it could potentially save me a lot of time.

Was hoping this would get bumped, i was too lazy to search for it. Im trying to promote the new game i work on, its similar to my older on i put on hold for now, with the advantage its simpler so it will be finished sooner. You can take a look at https://torque3d.org/forums/topic/17640-tiberian-sunrise/ It will be out by the end of this year, there isnt all that much left to be done for it be playable. Im trying to build up some momentum to spark interest, since i want to make this a proper community opensource project sometime in the near future. It would be nice if some people would take and interest since its all fully free and the code is fairly decent so it serves as a good base for other projects. Check it out and keep an eye on it, i will be posting updates regularly and i want to release a preview version as soon as possible to identify the more pressing multiplayer issues.

>>39768 Looks cool (and useful, probably) for an offline, local, SSD-powered file system, actually. Shame the online experience is slow

>>43258 >find comprehensible C tutorial series on Freetube >copy and take notes on the code from start to finish >everything goes well until SDL appears >we're on SDL3 now and the tutorials were made for SDL2 The pain of setting up SDL3 was indescribable. I don't think I'll ever get as mad at a computer as I did for those two days for the rest of my life. Taking a bullet out of my chest with my bare hands would have been more fun and intuitive than that torture fest.

>>43439 You could have just used SDL2, or searched for a SDL3 tutorial.

>>43440 >or searched for a SDL3 tutorial. That's what I did after I installed SDL3. Why would I keep using an outdated tutorial and library after all of that work?

>>43439 >>43444 SDL averages one major release about every decade. SDL3 came out this year. You could totally have used SDL2 instead.

SDL2 can write directly to the linux framebuffer, without relying on a windowing system. Is that still possible with SDL3, or does it require a X/Wayland install?

I dropped in- and out of the Gears of War 3 PC modding scene this month. It was kinda fun but I'll never fit in a clique that has a big share of christians, because I will never want to.

>>43470 I saw kms/drm backend when compiling, so yes I guess

Noob here. I want to ask how do you build your own engine in c/c++. Which libraries should I learn to use for the graphics, especially? Is there a way to make your own graphics library from scratch in c/c++ (just curious) and if so how and could you point me to some resources. pdfs would be appreciated.

>>43720 Read the thread. Even the first post as some answers...

>>43472 What's the relationship between Gears of War and Christians?

anybody make their own custom game / map editor, how did it go? There is useful libraries like IMGUI and the nuklear header only one out there to jump start getting one going. Any other recommendations?

Was looking for this thread but too lazy to dig it out. Anyways i want to bring attention to the first release client version of my game tiberian sunrise. Its a bit for rushjob, not ready for gameplay yet sadly, but it has 1gb of fully free(mit x11/cc0/pd) highdef assets of all kinds divided into packages, you can use all of it for any purpose you see fit. I got around to work on it once more after brief break and already fixed many of the issues that plague playablity and am confident to make another release soon that will be significantly better. https://torque3d.org/forums/topic/17640-tiberian-sunrise/#findComment-34505 I will post regular updates when i can and work on some promotional materials on the side. >>44112 A long time ago i attempted that, its a lot of work depending on what type you want to make. I also was for a while maintainer of a heightmap based editor, it was a lot of work and for various reasons, but i hope one of these days to revive it again. Recently i found a good looking engine called wicked, it uses lua and is licensed mit x11, it can be embedded i read, dont have the link tho so you need to find it yourself.

My common lisp autowrapped webgpu (DAWN+GLFW) is finally opening a window and handing me a surface without soykafting itself and dying. Soon, I will have portable common lisp running everywhere on top of the modern-est graphics API supported everywhere which also doesn't require me to write 1000000 lines for a triangle! It should even work in the browser with ECL+emscripten. Enemies of the great alien beware; your time is almost over!

>>43720 depends on what you mean by "from scratch". the lowest level you can realistically go for rendering graphics from scratch is using something like DirectX and OpenGL. but yeah, the rendering isnt even half your game engine... you also need to worry about physics, asset pipeline, audio, input, networking, ui, etc. this is pretty much an impossible task for someone with little to no experience. i would strongly recommend that you use a library like sfml that handles windowing, input, graphics, sound, networking, and build your own framework around that. sfml doesnt support 3d graphics, but you can always add that later by managing your own OpenGL context within the SFML window. https://www.sfml-dev.org/tutorials/3.0/ https://learnopengl.com/