>>102940735
Timewasters the general. Imagine being so delusional that you think your adult toys aren't a complete waste of time and effort. Wouldn't be me.

File: trash.jpg (37 KB, 305x304)

37 KB JPG

>>102940809
Why reply to the dead version of the thread my friend? Why seeth so hard? Complaining on a general where people are trying to learn new things and show their progress is the true real time-wasting activity.

Wouldn't be me.

>>102940757

you managed to beat me to thread creation but I'll get you next time tripfag

>>102940809
>waste of time
exactly everyone knows doing things you love is pointless you should be doing nothing but work 24/7, do not enjoy yourself, do not watch movies, do not play videogames, do not go to a park or climb a mountain and certainly don't spend your time fascinated by the intricacies of the technology that encompasses all of the human experience such as mathematics, logic, history, geometry, art, music, storytelling, anatomy, fantasy, human interaction
definitely a waste of time, return to you wage cage worker drone, there is only ever increasing productivity, no time for fun

>game engine general
>OP webm is a godot slop web export for something that could be done in a few lines in something like three.js or babylon.js
Never change /gedg/

File: c.jpg (21 KB, 600x450)

21 KB JPG

>>102940809
Working for a society that doesn't give a shit about is the waste of time.
I want to see you and your spawn impaled and burned alive.

File: js.jpg (102 KB, 650x650)

102 KB JPG

>>102943364
My game isn't in godot, it's using raylib.
Can you show me how to implement phong lighting and gpu skinning in a few lines using a js framework?

I was called rude and cranky last thread so I'm restraining myself this thread.

I'm jealous of the current generation of homebrew and indie devs who can actually achieve some kind of success from making games. 20 years ago there was a 0% chance of someone making a decent enough homebrew game to stick it on cartridges or discs and sell it, now it's just a thing that people do. There are new NES, SNES, N64, Gamecube, Wii, Gameboy/Color/Advance, DS, Genesis, Dreamcast, and PS2 games being made by people at home, dicking around and having fun, that sometimes actually become available for purchase as physical media. That's fucking remarkable to me, something I would've scoffed at and joked about when I was making Gameboy games in 2001. The free homebrew bullshit libraries and "devkits" are more robust than the real development tools we used to use. There are several free-to-use game engines that are barely a step above old Macromedia Flash in difficulty while being robust enough for AAA titles. The idea of getting into it now, if I were 14 again, god damn it is so easy to do so much, and effortless to make it available to others either freely or commercially. What a fucking time to be a hobby-level game developer.

>>102943787
There's nothing stopping you being a gamedev right now

>>102943827
Don't worry, anon, I'm still a game developer. I'm envious of the mild learning curve and extensive opportunities afforded to new hobby-devs these days, but it doesn't stop me from being happy for them, and certainly doesn't rob me of those same opportunities.

File: Untitled.png (878 KB, 2559x1254)

878 KB PNG

Realistically speaking, what performance gain would I earn from rewriting my Unity C# grand strategy simulation software-like game prototype in UE C++? Knowing that the overwhelming part of the data is in frozen static arrays, which allows the JIT compiler to pointer-bomb my code and remove bounds checking wherever possible already.

File: 2024-10-23 12-32-27.webm (1.2 MB, 1920x1080)

1.2 MB WEBM

Previously I was determining which chunk of the track the car was on purely by proximity to the vertices. This worked ok when only driving on the center surface but when adding the shoulders it was no longer sufficiently accurate. I decided to try and make a vertical peg at the car's origin point like it's a slot car and calculate the intersection with the track instead. Something's fucked with the algorithm though. Anything popping out at anyone? Literally just copypastad Möller–Trumbore from wikipedia like a jeet. Also added some debugging visualization so the peg is visible and which triangle it is supposedly colliding with is highlighted.

bool MathHelpers::RayCollisionTriangle(const glm::vec3 &rayOrig, const glm::vec3 &rayVec, const glm::vec3 &p0, const glm::vec3 &p1, const glm::vec3 &p2)
{
  constexpr float epsilon = std::numeric_limits<float>::epsilon();

  glm::vec3 edge1 = p1 - p0;
  glm::vec3 edge2 = p2 - p0;
  glm::vec3 ray_cross_e2 = glm::cross(rayVec, edge2);
  float det = glm::dot(edge1, ray_cross_e2);

  if (det > -epsilon && det < epsilon)
    return false;    // This ray is parallel to this triangle.

  float inv_det = 1.0f / det;
  glm::vec3 s = rayOrig - p0;
  float u = inv_det * glm::dot(s, ray_cross_e2);

  if (u < 0 || u > 1)
    return false;

  glm::vec3 s_cross_e1 = glm::cross(s, edge1);
  float v = inv_det * glm::dot(rayVec, s_cross_e1);

  if (v < 0 || u + v > 1)
    return false;

// At this stage we can compute t to find out where the intersection point is on the line.
  float t = inv_det * glm::dot(edge2, s_cross_e1);

  if (t > epsilon) // ray intersection
  {
    return true;// glm::vec3(ray_origin + ray_vector * t);
  } else // This means that there is a line intersection but not a ray intersection.
    return false;
}

>>102944281
Nevermind, once again taking the time to explain the problem caused me to find the solution immediately. The wikipedia algorithm works fine. I was putting in the two position vectors of the peg which makes rayVec point way off into space.

>>102940757
>still linking /agdg/
OP is a faggot, once again

Why is this thread full of petty bitching?

File: 2024-10-23 13-26-19.webm (706 KB, 640x480)

706 KB WEBM

>>102944458
>>102944281
the washboard is still bad of course, gonna have to do something about that

>>102944632
First day on 4chan?

File: file.png (451 KB, 480x480)

451 KB PNG

>>102944632

>>102944664
Very interesting, Anon!

Can you write all your code in Godot in C++ "normally" without giving up on stuff you can do in GDscript or whatever it is called or C#?

>>102944664
Good job, anon, I'm enjoying your updates. Car needs shocks for bumpy ride.

>>102944966
Close enough in my opinion, judge for yourself: https://docs.godotengine.org/en/stable/tutorials/scripting/gdextension/gdextension_cpp_example.html

#include <godot_cpp/classes/sprite2d.hpp>
namespace godot {
    class GDExample : public Sprite2D {
        GDCLASS(GDExample, Sprite2D)
        private:
            double time_passed;
        protected:
            static void _bind_methods();
        public:
            GDExample();
            ~GDExample();
            void _process(double delta) override;
    };
}

>>102944966
technically yes, but you still pay for all the garbage that comes with GDscript, and I don't think you get to export for the same targets without headache

>>102944458
>once again taking the time to explain the problem caused me to find the solution immediately
classic

File: game state-2024-10-23_14.28.25.webm (3.35 MB, 908x804)

3.35 MB WEBM

working on my affine transformation system for my game engine, it's starting to work but my protection against overdrawing is causing it to look lol, gonna haveto fix that

File: file.png (2.04 MB, 1401x787)

2.04 MB PNG

Are deck builder/card games the bottom of the barrel genre besides the super obvious ones like platformers and roguelikes/lites?
Duel masters like have cards so complicated that that they require a programming language for the card effects
Slay the spire is and any game that got inspired by it are pure garbage because they keep the complex cards but completely sidestep the AI by making the enemies not even engage in the card game mechanics.
But somehow pic related is what triggers me the most, it has 0 business being a card game, you just drag&drop the main player card to explore and uncover more cards, the crafting menu doesn't follow the card game mechanics and the survival part is basically automatic, you automatically consume the food card at the start of the day.
Why are devs so afraid of normal menus? why does it have to be a card at all? it's like if every action in FTL was a card (ironically this game exists).
Do players/devs find it really hard to navigate menus or a tree of complex decisions?

>>102946078
I really don't understand the point of your post

>>102946078
>Are deck builder/card games the bottom of the barrel genre besides the super obvious ones like platformers and roguelikes/lites?
Yes but the rest of your post doesn't really make sense. People make card games because it's trendy, just like any other trend there are people out there that "like deckbuilding games" and will play something solely because it has decks/cards.
I specially dislike this genre because it's yet another remnant of board-gaming cancer that has held strategy game design back for decades.

>>102946078
No it's actually rouge likes that are the biggest slop possible, just slightly below bad "card" (or card related) games.

Cards are generic so they get abused to do all kinds of things in games/tabletop and so on and so forth

>>102946078
what the fuck did you just say

What would happen if I do all my shaders in half precision?

File: ss_e8fbb9e4419623b9cb3080(...).jpg (1.52 MB, 1920x1080)

1.52 MB JPG

>>102946162
My point is that card games appeal to programmers because you can recreate the most complicated card system by literally having a scripting language for the cards, beside the point that you only need tween for animations and minimum art, the later being more of a problem with AI art.
This leads to games that has no business being a card game, like survival card games.
>>102946169
Are they trendy because they're so easy to make that it's easier to stumble upon decent mechanics? or do people genuinely seek them because they're card games?
I also agree with >>102946180
I have nothing against tabletop games but card games are dumb as a concept in a digital media where you can reflect the mechanics with anything that isn't just a bunch of text on a card.

I'm just doing some research and I kinda got annoyed by how easy they are to make. Wildfrost game now has mods that are straight up higher quality and more creative than the base game itself. We keep getting StS clones while StS 2 is in the way non stop, with the base game being modded to hell and back I'm not even sure there's space left for exploring other designs. Cobalt core had no business being as popular as it is with that shitty pixel art.
Funnily enough, Astrea: Six-Sided Oracles, which is a "what if card game but with dice", was a lackluster even with the overproduced art, but I think it kinda shot itself in the foot by having a base mechanic that straight up ignores the dice, and didn't get much mods.
I guess I am complaining how lazy are card games and that they get modding for free by design.
Side note, Balatro, one of the rare card games that gets a pass, somehow overwhelmed its dev with spaghetti code when 99% of the cards are literally simple multipliers.

>>102946485
how is it productive to get mad about card games being trendy? If you don't like them, don't play them or make them. I don't

>>102946485
It's just a lower effort and low barrier (easy to access) genre of game. It also has fast turnover time since you can implement stuff fairly fast.

It's just the logical next step for people that learned to make their first 2D games or try to do a project with a manageable scope. Makes iterating over it also very easy and rewarding.

Thoughts on EnTT?

>>102946591
ECS systems are retarded

>>102946596
For lack of any better ideas I implemented an ECS. Seems like a pretty reasonable way to organize things to me.

>>102946672
That's because you haven't made a game yet

>>102946698
It's working well enough for my notagame so far

>>102946555
To point that 99% of card games are asymmetrical, meaning, that the most interesting thing about them, which is making the AI for the enemy player, is always ignored for the sake of a shortcut.
I can't think of a recent card game where the enemy AI plays the same game as the player without literally cheating and a fixed set of decks.
I thought I'd check if anybody is making a card game here, or working at AI at all. But then again when has any anon posted something related to AI in the past 100 threads?
We already have stuff like min-maxing/alpha-beta pruning, machine learning and other things.
When you think about it, card games tend to have programming like logic, AI is taking over programming recently. Maybe there's something to be explored there.

>>102944664
Looking cool anon

>>102946852
That's because you haven't hit the part of game development where your choice of architecture actually matters
It turns into a gigantic spaghetti of systems

>>102946869
You could use real AI to make video game AI for strategy games, you would need to configure it to make it fun rather than good though

>>102946981
Do you have alternate recommendations?

>>102947026
The optimal architecture always depends on your game
The best generic architecture is the same one you see in every game engine, you have a GameObject and it has a bunch of optional components you can attach

>>102947064
So what's the difference between an entity component system and a gameobject component system?

>>102947084
you don't organize your logic into systems

>>102947009
Yugioh games had some very bullshit cheating AI and were still fun, as long as you have a huge number of strategies where there's always a way to win vs the enemy, it'll be great. I guess you can tune the game length. Nobody will complain if games last 20-30min average.
Wildfrost had one of the simplest yet effective twists that caught me by surprise, I played so well that I ended up with one of the most broken decks, only to find that the final boss of the fake ending literally takes over your party for the next run. Now you have to face a buffed version of your broken heroes, requiring to build even bullshit to beat it. The sad part is that after the few first runs you can sidestep this mechanic just to fight a generic final boss.

>>102947228
symmetric strategy games against the AI is never fun

>>102947239
actual realtime strategy games where actions per second matter? sure
turn based card games where there is way more than a few winning meta? sounds fun. what would be unfun about it if the AI legit don't cheat or peak into your deck?

>>102947270
>what would be unfun about it
The AI always fucking sucks at them
Remains to be seen how fun they could be if they use AI to train models that are actually difficult

How does camera shake work? My brainlet thought process was making everything on the screen move by some x,y,z values. But I've seen a lot of game have camera shake with everything remaining on screen, so how does that work?

>>102947582
you just move the camera

>>102944051
You could always just try Unity's ECS. It's ideal for strategy/simulation games.

>>102947667
I guess that's the concept I'm not understanding. I'm usually making small games in vanilla javascript. When someone says move the camera in my mind that's the same as moving the entire screen.

>>102947760
A camera is just an offset that is applied to everything before you draw it

>>102947769
Oh that makes more sense. How are things dealt with when an object might go out of bounds but you don't want it to? Just clamp it to the boundaries for that iteration of the shake?

Looking for a simple open source 3D physics library in C++, anyone got any recommendations? Thinking about one of these two.
https://github.com/RandyGaul/qu3e
https://github.com/DanielChappuis/reactphysics3d

>>102948017
bullet or physx

is phaser a good way to practice javascript?

Not that Anon. I'm a noob dev trying to learn. I am trying to make sense of how to structure my project, and was looking into ECS. The data flow pattern lending itself well to serialization appeals to me.
>>102946981
So the problem with ECS is because of it's complexity in bigger projects?
>>102947064
Is what what you describe just an Entity Components (EC) or an Component Object Model (COM)? I'm trying to find key search terms for what you describe. I can't find much atm.
If I understand ECS correctly for Example I would have an struct/class Entity with an int id for my Player, and a map holding components. I add an animation component to my entity, and an animation system updates the component data.
How would this example work in the GameObject Component architecture you describe? Would the player animation be a part of the player gameobject, and instead of an animation system I have an animation component?
Do gameobjectcomponents communicate with each other?

>>102948189
There's no name for the standard object model that games use. It's used in Unity, it's probably used in Unreal and Godot aswell. It's not rigorously defined, unlike ECS which is rigorously defined
ECS has a relational database of entities / components and it has systems which operate on groups of them (the systems are like database queries). ECS is based on batch processing, which is good for performance, but it's shit for the implementation of most game logic. With the regular object model your game objects would just be OOP objects and you can call individual methods on them and have polymorphic behaviour

>>102948291
I guess I did that then. I didn't realize ECS was highly specific, I thought since I had entities with components that's what I had made.

>>102948291
Thanks for the insight.
I am reading that rigorous ECS does complicate implementing things like Finite State Machines.
I presume the GameObject Component model does not have the issue of inheritance that strict OOP causes.
With Unity I did read that DOTs does have better performance than the standard approach. I believe ECS advertises as having good locality.
I am wondering if it makes sense to try to roll ECS but not strictly so that I can deal with things like State machines easily and avoid the spaghetti you mentioned. But since I'm a noob and I can't find much on how to combine ECS / OOP/ Object-Component it probably makes more sense to just stick with one.

>>102948536
The problem with ECS is that it is strict. It dictates the structure of your code. You can be less strict about it, but then you won't get the performance benefits so why are you even bothering
Also if not using ECS has been good enough for the best video games for the past 40 years there's no reason why you would need the performance benefits ECS claims to provide

I usually post in /agdg/ than this place, but I am wondering about some stuff, since /agdg/ is mostly filled with people who don't know a lot about certain stuff.

Let's say I want to make an arcade machine version of a game I created a while ago. I have the source code. How do I profit from it if I create it? Mainly since I don't know a good place to put it, and since it's custom, if people would accept it or play it.

>>102948590
My shit is easy enough to render that it literally does not matter. A single thread to do everything is 500+fps. Being easy to understand and keep the structure in your brain is by far the most important thing.

>>102948636
Arcades have been dead since the 90s

>>102948590
Gotcha. Ty Anon

>>102948639
I know stores and restaurants have them, mostly prize related ones, but outside of that it's very niche. Which makes me question if it's profitable to make one. Maybe it it deals out tickets? No idea. It's just something I began thinking the logistics of as a new way to get profit from a game or two.

>>102948658
>Which makes me question if it's profitable to make one
It's not
There's really no way

>>102948691
Maybe you are right. I'm not really sure if there is anything near me that would buy it off me or anything like that, outside one place that would 100% have to have a ticket output sort of thing for it to be accepted there. A lot of unknowns.

>>102948739
Game developers are not farmers providing fresh locally sourced video games to arcades

>>102948795
I know, probably be 100 dollars before the actual base is considered for the hardware, probably needing to be custom made from home depot to cut the wood out for a base. I really don't know a lot about the backend for the machines money wise for how much is required, who buys them, and at what price.

File: 534745967967(1).webm (940 KB, 1132x632)

940 KB WEBM

Unified the interaction system today so now every intractable static object will be preloaded on the map, also got doors working without using nodes though they're all direction dependent so have to write some code to check for that, probably going to end up using doors that just slide into walls/ceilings half the time because I can't be assed although this is something I haven't seen a lot of people figure out

As far as the door/nav mesh though I am hacking it a little bit since the settings I baked on prevent the agent from figuring out how to get through the door frame if the object is in the trenchbroom node during bake, thankfully I can just move it script and all and bake the nav mesh which works just a bad hack because if I rebuild the trenchbroom import it'll put new doors in

Also forgot the script for this doesn't treat the doors as walls so technically enemies can cheat and see through them but probably not too bad maybe I can figure out something for it

>>102949530
nice

File: images.jpg (9 KB, 275x183)

9 KB JPG

>>102940757
>title: Ronin Tomcat
>player is a tomcat that lives in rural countryside
>land is divided into turfs, run by female cats
>each turf generates x-amount of mice
>each cat needs 6 mice a day to survive
>player can subjugate female cats and demand tribute of mice from them
>eventually rival tomcats starts entering the map
>these tomcats will challenge player to battles if they encounter
>if player loses they are deprived of their turf

>in order to get enough mice, the player has to control multiple turfs
>but they can only be in one place at one time
>so, they have keep moving to prevent enemy tomcats from usurping their turf

Essentially RPG strategy game

I'm playing around with Raylib on Linux, and strangely, SDL doesn't appear to be able to retrieve information on my monitors. xrandr can see the monitors, and they're obviously working, so I'm not sure what's going on. Anyone hit this error before, or screw around with SDL on a low enough level to hit these problems?

>>102952288
GLFW doesn't have this problem

>>102948862
worry about making a good game and then if you really really want to make some arcade cabinet and sell them for high premium to the few whales that like your game.

No one will buy an arcade cabinet and even less if it is made by a nobody.

>>102951526
The gang war was the worse part of GTA SA and saint row.

>>102952310
isn't raylib just another wrapper around glfw?

>>102954495
same in Skyrim, because those are scripted, things has to be dynamic/generic to be interesting

>>102944605
We should link each other though ... >>>/vg/499653335

>>102954903
Yeah, also freetype and miniaudio with opengl abstractions.

>>102956167
linking the catalog search is better, that thread will be archived long before this one

>>102956305
I did, in the OP. I was just showing that in that particular thread, /gedg/ was linked.

Testing my barebones ecs. This test

ecs::registry registry(50);
  std::string component_name("MyTest");
  registry.new_component(component_name, 20, sizeof(bool));
  
  bool b1 = true, b2 = false, b3 = true;
  
  ecs::entity e1 = registry.new_entity();
  registry.set_component(e1, component_name, &b1);

  ecs::entity e2 = registry.new_entity();
  registry.set_component(e2, component_name, &b2);

  registry.delete_entity(e1);
  ecs::entity e3 = registry.new_entity();
  registry.set_component(e3, component_name, &b3);
  e1 = registry.new_entity();
  registry.set_component(e1, component_name, &b1);
  b1 = false;
  registry.set_component(e1, component_name, &b1);

  
  std::cout << "e1 " << e1.id << std::endl;
  std::cout << "e2 " << e2.id << std::endl;
  std::cout << "e3 " << e3.id << std::endl;
  std::cout << "b1 " << *((bool*)registry.get_component(e1, component_name)) << std::endl;
  std::cout << "b2 " << *((bool*)registry.get_component(e2, component_name)) << std::endl;
  std::cout << "b3 " << *((bool*)registry.get_component(e3, component_name)) << std::endl;

works as intended, yielding

e1 2
e2 1
e3 0
b1 0
b2 0
b3 1

>>102957037
It doesn't have any fancy archetypes but I think it will suit my text game fine. I mostly just need something that will allow composition at runtime. And being able to loop through the components as a contiguous block of memory is nice.

>>102957190
tripfags ruin everything

>>102957037
Didn't know you liked sepples froggy

Spent an hour moving some mulch. Nice little exercise but now it is time to go back to programing.

>>102957591
What did I ruin?

>>102958025
Oops. Fuck C++ amirite guys?

I do truly believe C is better and I hate C++ compile times. But I'm already using some C++ libs in the project. Something I like to do if I end up using C++ is to compartmentalize different parts of the code into C libraries if convenient.

File: palette.png (584 KB, 2071x1286)

584 KB PNG

Missed me? I had a pretty rough two weeks, but still got palette-correct rendering with lightning done. Here - OTTD palette, but I support any arbitrary palette of 255 hues and 256 shades each ( so 65280 colors out of a palette of 16m ).

File: autoexposure.webm (3.82 MB, 1280x720)

3.82 MB WEBM

It turned out I broke auto exposure somewhere along the way, but now it's fixed and improved to work properly for all environments, which was tricky.

>>102948536
https://www.gameprogrammingpatterns.com/contents.html

File: UE5.png (463 KB, 2545x1343)

463 KB PNG

I was gonna switch from Unity to UE5 then I fell on this shit. What the hell? The documentation states that using specific code naming conventions for fields, classes, etc. is MANDATORY.

>>102958861
Age of brass dev?
I like the style

>>102958861
space alien sfml game dev? did you change the whole game or is this a new one?

>>102959056
Yeah that's me, thanks.
>>102959114
That's the older one that branched off into AoB.

>>102958895
That's only if you contribute to the engine. I don't believe they have any power or will to determine what's in your game's code.

>>102959146
Oh wow, I missed your progress. Last time I've seen it it was more than 5 years ago.
What happened to the old game? can you link your blog again

>>102959242
>What happened to the old game?
Age of Brass? I'm still working on it but at slower pace. It's not popular and liked enough to deserve public releases so I'm just having fun with it on my own. Maybe I'll figure something out to change the situation but until then I can't even start pumping out content because I would have to rework everything when the mechanics change.
I don't have much of a blog, so here's the best I can do https://x.com/DementiaDev

>>102949530
Unfortunately realizing that the tutorial I'm following for inventory starts to split much harder in intention for containers than I was prepared, oh well I'll get it working

>>102959352
>It's not popular and liked enough to deserve public releases
are you insane? you already had tons of good movement mechanics, all you had to do is drop a small demo with a level editor and let people have fun. It will 100% catch the eye of some autist and he will recreate some other popular game levels in it.
You can even make the a level converter yourself and make a dev video about it.
Popularity is a fickle thing.

>>102944664
almost reminds me of DRacer

>>102959710
That's not how popularity works

>>102959710
That would be nice but I don't think anyone even played enough to get a good grasp of the existing mechanics yet, which would be essential for making levels. I'm pretty much losing players from the get go.
Meanwhile I have on record 50+ hours in a single level and I still didn't have enough of it.

File: lightmop2.jpg (269 KB, 991x596)

269 KB JPG

Got a bit bored so decided to try rendering Tribes terrains using a vertex shader. Might be useful for a future project.
Also I miss WindowsNME.

>>102958220
I dunno who you are but that looks cool

>>102959884
don't be discouraged, you just caught the wrong audience, keep going at it.

I'm trying to figure out an efficient way to interpolate an arbitrarily long vector in C++
for example if I've got a std::vector<double> that contains the x, y, z, -z, u, v coordinates of a point I would use the following code

using vec = std::vector<double>;

//linear interpolation std::vector<double> = double, double, uint steps
[[gnu::pure]] std::vector<double> LinInterp(double value1, double value2, uint steps){
    //creates a series of steps evenly spaced out along the difference between 2 numbers
    // y = ((x2-x1)/steps) * step
    std::vector<double> result;

    double step = (value2 - value1)/(double)steps;

    double curValue = value1;

    for(uint i = 0; i<steps+1; i++){
        result.push_back(curValue);
        curValue+= step;
    }

    return result;
}


[[gnu::pure]] std::vector<vec> InterpolatePoints(vec vector1, vec vector2, uint count){
    std::vector<vec> result;

    std::vector<std::vector<double>> vals;

    for(uint i = 0; i < vector1.size(); i++){
        vals.push_back(LinInterp(vector1[i], vector2[i], count));
    }

    for(uint i = 0; i < vals[0].size(); i++){
        vec curStep;
        for(uint o = 0; o < vector1.size(); o++){
            curStep.push_back(vals[o][i]);
        }
        result.push_back(curStep);
    }

    return result;
}

this will give me everything but I find it's pretty slow, is there a better way of doing this?

>>102960555
>C++

Why don't you reserve the memory before you push back?

result.reserve(steps+1);

I don't know if that would be a bottleneck though. std::vector doubles the capacity each time it reserves memory (not sure).

If you know how big a vector is beforehand then just reserve the memory. This is what C++ does, it teaches people to use std::vector when all they need is a preallocated slab of memory.

But once again, not sure if this is your bottleneck. Might help to share some profiling data.

>>102960555
Either reserve the number of elements you have in each vector or pre-allocate it and fill in the values via an indexed pointer.

Use std::move to avoid re-allocating the LinInterp vector when you push it into vals.

>>102960555
I would never create a vector in a performance critical code.

bump

File: pass it.png (1000 KB, 1024x984)

1000 KB PNG

Wrapping up devving for the night soon.

I've been re-architecting my audio manager.
Before, I only had one Sound struct per sound that could be played, so if two entities were screaming there was no way to differentiate the sounds.
Raylib provides "sound aliases" so I'm using these to have entities play their own individual sounds. It's been a bear of a rewrite but I'm getting there and the audio manager is going to be so cool when I'm done.
I'd love to add 3d audio stuff like have far away sounds be quieter and having sounds pan left or right based on the camera, but I'll leave that for later.

Does anyone program their games in Swift so that they feel nice on iOS? Is "feeling nice" a meme that's not worth what you lose in platform transferbility?

>>102958075
>C++ compile times
How can C++ end up with bad compile times? Surely the typical change only modifies a single .o?

>>102958861
Looks like there's been some plot development since last I saw.

File: IMG_1435.jpg (426 KB, 1044x1171)

426 KB JPG

Cyberpunk 2077 – RPG/FPS, REDengine (C++), 20M copies
Grand Theft Auto V – Open-world adventure, RAGE (C++), 180M copies
Red Dead Redemption 2 – Action/Adventure, RAGE (C++), 55M copies
Among Us – Party/Social deduction, Unity (C#), 50M copies
Fall Guys – Platformer/Battle Royale, Unity (C#), 20M copies
Dota 2 – MOBA, Source 2 (C++), Free-to-play (millions of players)
Counter-Strike: Global Offensive – FPS, Source (C++), Free-to-play
Monster Hunter: World – Action RPG, MT Framework (C++), 18M copies
PUBG – Battle Royale, Unreal Engine (C++), 75M copies
Destiny 2 – MMOFPS, Tiger Engine (C++), Free-to-play
Doom Eternal – FPS, id Tech 7 (C++), 5M copies
Rainbow Six Siege – Tactical Shooter, AnvilNext (C++), 10M+ copies
Sea of Thieves – Adventure, Unreal Engine 4 (C++), 5M+ copies
Borderlands 3 – FPS/RPG, Unreal Engine 4 (C++), 15M copies
The Witcher 3 – RPG, REDengine (C++), 50M copies
Baldur's Gate 3 – RPG, Divinity Engine (C++), 5M+ copies
Phasmophobia – Horror, Unity (C#), 2M+ copies
Valheim – Survival, Unity (C#), 10M+ copies
Deep Rock Galactic – Co-op FPS, Unreal Engine 4 (C++), 4M+ copies
Forza Horizon 4 – Racing, ForzaTech (C++), 12M copies
Microsoft Flight Simulator – Simulation, Asobo Engine (C++), 2M+ copies
The Sims 4 – Simulation, Frostbite (C++), 36M copies
ARK: Survival Evolved – Survival, Unreal Engine 4 (C++), 10M+ copies
The Elder Scrolls Online – MMORPG, ZeniMax Engine (C++), 20M copies
Assassin's Creed Valhalla – Action/Adventure, AnvilNext (C++), 15M copies
FIFA 22 – Sports, Frostbite (C++), 9M+ copies
Satisfactory – Factory simulation,

>>102963701
So much C++ code gets forced into headers, so changes can easily affect a shitload of files. But that's not the only reason C++ is slow as fuck to compile.

>>102963701
C++ compile times are not bad and with a fast multicore CPU evel large projects build quickly. Full build of 4m loc project at work takes about 10 minutes. Although overdoing on .h inline functions will cause a cascade of recompiles.

Its link times that hurt. If a project has botched dependancies then this non-parallelisable operatiin takes ages. There is one project i work on that links 4minutes on 7950x3d. I'm slowly chopping away at it, but it takes time to untangle 25 years of unpaid debt

>>102963872
most of the graphics shit just requires you to do that

>>102964019
I fucking hate people who use headers for anything other than function and class declarations, and fuck preproccessor commands seriously

>>102944664
the easiest solution is to just increase the number of polys on curves until the problem disappears

Write in your programming language the code the shortest possible that does exactly the following:
1. allocate a memory region of size 3.200 bits, and a pointer ptr to it.
2. copy the content of a float16 array of unknown size float16Array to the memory region starting from ptr. (It might or might not write values beyond the memory region.)
3. interpret the memory region as an array int32Array of size 100.
4. checks whether the difference between the first and last elements is greater than 69. If it is, return an array of Vector2 vector2Array of size 49 starting from ++ptr without copying any element, else return the memory region as a Vector2 array of size 50.
5. destroy ptr and desallocate only the first and last 32 bits of the memory region if the difference was greater than 69.

File: enums.jpg (825 KB, 4368x1841)

825 KB JPG

>>102964019
NTA, but lemme guess, template instantiation is a significant reason why build times are slow.

>But that's not the only reason C++ is slow as fuck to compile.
Reminds me of a funny story where a child tried to argue that

enum spelled_out {
  MEMBER_1 = 0,
  MEMBER_2 = 1,
  // (65,534 other members)
};

is faster to compile than

enum not_spelled_out {
  MEMBER_0,
  MEMBER_1,
  // (65,534 other members)
};

because "adding numbers is slower than just reading them off the code", and therefore has declared that not spelling out the members (or using enum classes, because apparently type safety is bloated), is equivalent to being an ignorant webdev.
Suffice to say, a quick benchmark (picrel) has dissipated all his claims.

>>102964254
Templates are the worst offender for causing code to appear in headers.

File: ygg_10252024_1920x1200_cut_vp9.webm (3.04 MB, 1920x1200)

3.04 MB WEBM

Working on game refactor (Ygg Engine) again since I've been bored with work

>>102948189
>If I understand ECS correctly
you dont, entities dont manage component lifetime in ECS like you describe

>>102963872
I mean Unity is also written in C++

so yeah

>>102963701
>>102964050
This, i.e. for small changes compile times are instant, but in Release linking can sometimes take a few minutes in my project. I had to make a custom profile that combines Release runtime performance with Debug linking times and hot loading to get the best of both.

>>102964254
riddle me this, why would a language require headers in 2024?

>>102964997
how did you handle your Y sorting? looks neat

File: grease-down-the-drain[1].jpg (30 KB, 300x200)

30 KB JPG

>>102964254
not my problem

>>102964997
How the hell did you make your Y sorting so fast with so many objects?

Is using AI voices in your game bad?

>>102966595
anything lazy or soulless is bad so ask yourself is this lazy and soulless and you have your answer

>>102966618
What if I have no other choice?

File: 1000020755.png (584 KB, 1200x1240)

584 KB PNG

We want >(you) in our next shill fest, The /v/erdict! Check out our high production value and quality information video and get your games in by November 15th: https://youtu.be/OZoiwWCNH_I?si=bRhijvvP9auC-Ju2

>>102966758
then you're already conceding defeat before you've even started and don't deserve the honor of creating something with soul.

>>102966595
If you don't know why it would be bad then you wouldn't be much of a director and you would end up spending money on voice actors that would do a bad job unless they are good enough to carry you. Just give it a try for learning purposes.

>>102966595
For prototyping, no. For production? Uncertain.

>>102966791
This phrase makes absolutely no sense.

>>102966595
Why not just use your own voice? You have a microphone surely.

>>102967081
AI voices in a game can be fine sure, but you have to draw the line somewhere. You're trying to create something that's valuable or meaningful in some way, every little concession in the spirit of "I can get away with this lazy thing" works against you. If your attitude is AI voices are fine, asset store stuff is fine, latest free game engine is fine, using other people's code is fine etc eventually you end up with meaningless slop that never needed to exist in the first place. Have a good think about whether this is what you actually want and make your decision based on what you feel is best.

>>102967435
Well since I have absolutely no money, and I wasnt able to find volunteers to work with me, and I can't voice every character myself, I figured I'd use it. They are the only thing I use AI generated stuff for since they're the one thing I can't do on my own.

>>102967516
Not all games have to be fully voiced nor do people expect them to be.

>>102967516
you don't have to justify yourself or your decisions to me or anybody else, if you care about the product you'll make the right choices.

>>102964276
I've been thinking about this, do you know some good smoothing tessellation algorithms I should look into?

>>102966865
>If you don't know why it would be bad then you wouldn't be much of a director and you would end up spending money on voice actors that would do a bad job unless they are good enough to carry you.
>If you don't know why it would be bad then you wouldn't be much of a mechanic and you would end up spending money on another diagnosis and then the labor and parts so they can fix it for you.
>If you don't know why it would be bad then you wouldn't be much of a baker and you would end up spending money at actual bakeries for your bread.
>If you don't know why it would be bad then you wouldn't be much of a parents and you would end up spending money adopting a Chinese baby that might hold up to your shake, rattle, 'n' roll.

>>102968407
Are you insane?

>>102965932
>>102966338
The sorting is done in an opengl shader

>>102967516
>I can't voice every character myself
why not?

how much will I regret using Godot for a cross platform payments with microtransactions for skins?

>>102969415
Yeah you're probably going to want to do the payment processing external to the game. Like in a browser

>>102969404
I have a very shitty voice but also I can't voice people who are older than me, women, etc. I can only voice a handful of characters by myself but my voice would almost certainly not be better than an AI, because I can't act (both in a literal and a rhetorical sense).

>>102969435
Thank you

>>102969877
there's plenty of software voice mods. voicemod and morphvox come to mind

>>102969988
They're not that great, and, at this point, what's the difference between using a voice changer and AI speech synthesis? Both aren't "real" and both won't deliver a "great" performance.

>>102968407
What you are producing isn't just a voice recording, you are trying to fit it into the whole game which requires directing. If you need to outsource your vision, using AI is not the biggest problem.

Not the guy who asked but there is valid reasons to us AI voice technology. What if I want an NPC reading the entire KJV Bible in Megatron's voice?

>>102971705
then you should kill yourself

Is the voice actors union astroturfing this thread?

>>102971760
I astroturfed yer mums snatch last night

>>102964997
>ygg
long time no see. what happened to that runescape game ?

>>102971760
I think you mean the Film Actor's Guild, but probably just 1 guy

>waaaaahhhhhhh nooooooo don't use ai noooooo muh soul
Meanwhile every tech company and aaa studio is investing heavily in the tech. Indie game luddites are going to get even more squeezed with their mentality

>>102972853
Every streamer/youtuber I watch has to attack AI by virtue. I've seen an exact case where an indie game used AI for VA, even if the game itself was fine, AI for voice acting was used against it and most comments agree.
It seems that everybody agrees that it's better to have those gibberish sounds or no VA at all than have a soulless voice. I'm of the same opinion, why even waste effort trying to get AI sound better when it will 100% be shot down anyway?
This is not about something as necessary as art, this is something arguably optional. It's a case of "less is more".

>>102973238
maybe so, but the technology is pretty cool and I've had fun playing with it. I don't know that I'd commit to using it in a game but it would be fun to experiment with.
At some point it will be a reality of media creation and its use will be widespread.

>>102973238
>Every streamer/youtuber I watch has to attack AI by virtue. I've seen an exact case where an indie game used AI for VA, even if the game itself was fine, AI for voice acting was used against it and most comments agree.
This isn't a sane position to have and you know it.

>>102972853
>>102972700
>>102973238
>>102973265

The reason is simple, human beings are creatures of creation, since the industrial revolution more and more creations have been delegated to uncaring monstrosities that do not put anything into what they mass produce except what can make a small group of people money
meanwhile persons who once had purpose find themselves displaced quite rapidly and considered worthless and obsolete in a world that no longer needs them. told to "just code" or something
you mention luddites, do you know why the luddites existed at all? it's because the machines being used to weave fabric were originally advertised as a way for the same number of workers to increase productivity, but instead they used it to fire all the skilled workers and hire only a small number of people just competent enough to run the machines, whole communities in england were suddenly poor as the factories that once kept them fed abandoned them once convenient.
it's how all technologies of this kind are presented at first, "oh it's going to help you do your job" soon becomes "since you can do ten times as much as only need 1/10 the workers" to "why do we need someone with your skill when we can pay someone 1/10 what you make to get the same productivity."
as a developer I hear it all the time now, bosses trying to convince me that the AI code helpers will just speed up output, not replace anyone "it's here to help" sure but there's only so much work to be done and they aren't going to pay people to sit around doing nothing.
and the managers DO NOT CARE if it's not "ready" yet, they are gonna go ahead in their ignorance and start laying people off en masse because that's how they can report a higher profit margin to the share holders
The same thing is true with using AI to make voices, I don't care if some nobody is using it for a personal project, but when you're a AAA game company deliberately using it to avoid paying voice actors you deserve hell

>>102973422
>some nobody
show some respect if you are sympathy posting

>>102973436
you know what you're right I should have said some random guy, an average joe, a fella who just wants to make a personal thing but lacks the funds to make it happen

>>102973422
farm hands lost their livelihoods as industrialization meant that a dozen people could harvest what once took hundreds and they had to find new work, same with weavers, seamstresses, etc
it goes all the way up to the modern day, where most people have service industry jobs
SERVICE INDUSTRY
is that what you want? fully automated luxury gay space communism isn't going to happen, all the billionaires aren't going to one day go "ok we've automated everything, now everything is free and we're all rich!" they're going to become the barons of the future, and once we've turned all our thinking over to machines it will only allow other men with machines to control us.
no what's going to happen is it's just going to be a bunch of people selling lattes to each other over and over again, while a handful of people do own all the machines that do all production
worse; there are plenty of examples of robots trying to replace service industry as well. civilization collapses when 1/3 of working age men are fully abandoned by society, it happened in the bronze age, it happened in rome, it happened in the british empire, it happened in france, it happened in russia, what the fuck makes you so fucking confident that it wont happen here? that taking away all the jobs from millions of people wont stir up the anger and resentment that already exists
I wont do shit; I've got my wealth secured, and not in some flimsy stock market, but I've studied history and I know what happens when enough people feel like they have nothing to lose and the bread and circuses can't keep them distracted.

That's a lot of words that aren't about game development

>>102973536
TL;DR - AI bad because some indie dev decided to use it for voice acting instead of PAYING ACTORS WHAT THEY DESERVE!!!

File: mechsican.png (39 KB, 640x192)

39 KB PNG

I considered using AI to create the sprites in my game
then I beat myself with a shoe until the idea went away and learned how to draw cause I'm not a lazy cunt

File: wall-preview.gif (15 KB, 192x192)

15 KB GIF

I'm looking for the guy who made picrel
I want to work with him on a project again but we lost touch some time ago

>>102973594
I buy my sprites

As much as I enjoy working on my own 3D renderer, how's Ogre these days? Looks like Ogre Next 3.0.0 just dropped. I only have one of the backends implemented for my renderer, so it seems like a nice quick way to actually make a game that is cross-platform.

>>102960668
>I used to be careful about using reserve() when I was reading into a vector. I was surprised to find that for essentially all my uses, calling reserve() did not measurably affect performance. The default growth strategy worked just as well as my estimates, so I stopped trying to improve performance using reserve(). Instead I use it to increase predictability of reallocation delays and to prevent invalidation of pointers and iterators.

- Bjarne Stroustrup

>>102973508
>I've got my wealth secured, and not in some flimsy stock market, but I've studied history and I know what happens when enough people feel like they have nothing to lose and the bread and circuses can't keep them distracted.
t. gold enjoyer

>>102973801
>I buy my sprites
wow so you enslaved someone through the shackles of capitalism in order to squeeze the value out of them and accrue the benefits of their labor to yourself?

File: bjarne.jpg (795 KB, 1730x2200)

795 KB JPG

>>102974104
he's full of shit

>>102973594
Looks nice anon. How long did it take you to reach this point?

>>102974148
about a month, I'm trying to make better sprites now for npcs and stuff

>>102972685
I was getting gud, neat things ahead

File: 49chunk_window_movement.webm (1.31 MB, 1920x1080)

1.31 MB WEBM

Milestone reached.

Full directional chunk window movement.
Re-rendering only the seams between chunks of the exact row or column which needs to be added.

49 chunks moving without any artifacts or trouble , using the arrow keys.

uses a ring buffer to avoid need of re-meshing cascade.

Praise be to the King Jesus Christ.

Added a Local<T> built in system parameter to my ECS framework. Next re-evaluating how I want to handle the API design for owned/non-owned resources shared between systems (as sometimes they're just a pointer to data owned somewhere else).

World world;

int final = 0;
world.add_system([&final](Local<int>& i) {
    *i += 1;
    final = *i;
});

const int total = 10;
for (int i = 0; i < total; ++i) {
    world.progress();
}
CHECK(final == total);

> [doctest] test cases: 21 | 21 passed | 0 failed | 0 skipped
> [doctest] assertions: 87 | 87 passed | 0 failed |
> [doctest] Status: SUCCESS!

bump

Is vk.guide the better Vulkan guide?

>>102978880
Vkguide uses a helper library that abstracts over things so I prefer learn vulkan. That being said he has some nice stuff on the website. The website has extra chapters which are cool. For example, he has a section where he uses Renderdoc to analyze Metaphor: Refantazio.

File: dancing with jesus.gif (764 KB, 235x277)

764 KB GIF

>>102976675
gratz
>>102976938
>using ecs for performance
>using templates to fuck that performance up

>>102979823
Thanks, also that's neat!

>>102980548
Nobody in here even knows what ECS is, it's just become a meaningless buzzword

>>102960668
>>102974104
This. If you know how big an array is going to be just use a regular goddamn array. Reserve and emplace_back niggers get bent.

>>102982192
Also, if you're doing something performance oriented and you don't know how big it is going to be but you do know it will never be bigger than a certain number you can still just use a regular array and keep track of the actual count. We have fucktons of memory these days.

>>102980548
Marginal impact on compilation time but other than that, not at all. Benchmarks are comparable with flecs which is more than enough.

File: mygaem.webm (664 KB, 1280x720)

664 KB WEBM

>>102940757
Some shit that I'm working on. Heavily inspired by early shooters. Might turn it into a roguelike dungeon crawler.

>>102982711
Somehow it has more accurate perspective than a vast majority of games.

>>102973553
Lazy work never goes well with the customer.
There's an FPS game called the Finals which used their own team as voice samples to make AI announcers, kind of like old school FMVs literally using programmers as stand-in actors. Its still lazy, the voice lines are repeated over and over like any other announcer. If they could generate lines based on what's going in the arena will showcase the best and the worst of the new technology, but that would take real work.

>>102978746
bump x2

>>102976675
Nice progress

>>102973606
And you first found him here?

>>102984596
yeah, he helped me make a trump game during the 2016 election "MagaMan"
we only made a couple of levels before he vanished, his art was quite good

File: 2024-10-26_17:13:51.png (62 KB, 1920x1080)

62 KB PNG

Runtime ECS system is now exposed to guile scripts. Generic guile scheme data can be stored as components.

File: doggo stairs.gif (2 MB, 333x251)

2 MB GIF

fuuuuck it happened.
I've written so much complicated spaghetti that I can't refactor it anymore.

All I can do is put boundaries on the unfixable part of the code and not touch it anymore. It's like the elephant's boneyard in lion king, I must never go to that function or its associated member variables. Just don't touch and forget.

>>102986052
.dll with it

>>102986052
why not just delete it?

>>102986416
but the libraries it uses are built as lib...
>>102986421
cause it works, I just don't understand why and I don't feel like looking at it anymore.

>>102986052
Sometimes I find feeding the code into an AI model helps if you’re stuck in a rut. Sometimes it suggests something so dumb it’s inspirational.

>>102959381
Spent 3 days of development just getting inventories and boxes with inventories working and it's functional with like half of the necessary features, sucks that so much effort needs to go into something that doesn't even feel fun to use just a necessary part of the game, like saving and loading or options menus, it did teach me a lot about custom resources, don't feel bad that most of the code was from a tutorial considering how much had to be rearranged since the guy did all of the inventory on a UI node for the main scene instead of the player scene

For actual game GUI for a custom engine game, what do you guys do? Imgui looks too 'debug' .

>>102940757

>>102987300
Dear ImGui.

>Imgui looks too 'debug' .
Skin it yourself, retard.

>>102987457
Minecraft apparently made their own GUI solution.

>>102987300
make your own

>>102987492
Yep, that's what I've decided on after discovering that Minecraft made their own.

Just gotta render it.

>>102987300
Just write one, it won't take that long.

>>102978746

>>102940757
any good resource on making game assets for a isometric view? I have been drawing some things but they all look off.

>>102990781
post your work

>>102987300
there are no good solutions since that flash based framework died.
There's some stuff out there like crazy eddy's gui framework and noesis and nuklear. But none of them are great.
I'm probably going to roll my own for the fun of it. Or try out ultralight. Or try to get EAWebKit working

>>102987300
I use a 20 year old framework. Still works fine, just needed a little updating for modern text layout.

>>102990781
Are you drawing on isometric graph paper?

File: die.jpg (64 KB, 680x680)

64 KB JPG

>debug for like an hour
>problem ends up being that I passed true when I should have passed false and I forgot what the parameter did

>>102990781
You could do a sprite stack and just rotate it to the angle you want

>>102991252
Use asserts, man. And never manually touch Booleans you use.

>>102992642
>Use asserts, man.
Actually just started putting them in my codebase recently. Definitely nice for debugging real time stuff.
>And never manually touch Booleans you use.
Not sure what you mean by this

>>102992642
>>102993413
never assert anything, never use a debuggers, always print out all your variables and just look for where it all goes wrong

File: least intelligent.png (273 KB, 470x860)

273 KB PNG

>>102993516

>>102992642
>never manually touch Booleans you use.
What does this even mean?

>>102968003
Not by name, no.
This talk by the developers of Nanite in unreal go through some points that may be of interest to you.
It may not be this specific talk, but in one of them he talks about how he got his start in tessellation of landscapes which is pretty much exactly what you want

>>102995108
I forgot the link
https://www.youtube.com/watch?v=eviSykqSUUw

>>102968407
all of those examples are retarded and have the cost benefit ratio completely out of sync with each other
AI voices are practically free.

>>102970056
Plenty of people get captivated by youtube channels that use their own singing voice and use AI to diffuse it into someone else's singing voice.

If your voice is shitty because you can't stop sounding like a nasally nerd or a mouth breathing retard, then yeah for best results you'll need to hire at least one person.

bump

File: single_threaded_chunk_gen.webm (3.58 MB, 1920x1080)

3.58 MB WEBM

fully directional chunk window generation is done.
but yes it is very slow, next object on the agenda is to make it multi-threaded so that the background thread builds the chunk data instead of blocking the main thread.

October has been a very good month of progress.
I'm in a good place.
Praise the King Jesus Christ.

How hard is it to pick up OpenGL if I only want to use it to add 3D backgrounds to my 2D game?

File: 1729514230771501.png (391 KB, 432x533)

391 KB PNG

I finally figured out how to extend the Content Pipeline processors in Monogame. Once you finally figure out how it works it's not difficult at all, it's just that the documentation is so shitty for modern Monogame because 60% of the documentation is still the old XNA docs, you'll have something randomly not work and not give you a proper error, and googling the error leads to links that reference an example that is no longer anywhere on the internet. I really hope these guys get their act together now and put at least some money in documenting it in its modern state, I don't even want new features just some decent fucking docs please.

>>102973422
>>102973508
Based. The only reason so much money is being thrown at AI is that is promises to automate the knowledge-worker, which has up til know been impossible to automate. It has nothing to do with improving the world, making people's lives easier, it has everything to do with the Elites being able to pay YOU less money and make YOUR labor worth less. Fuck AI, fuck the Elites, fuck bootlickers and fuck every single retarded crypto faggot that thinks that AI is gonna make them back all the money they lost on retarded NFTs and Metaverse real estate.

>>102997626
Luddites never win in the long run buddy

>>102997761
>long run
sir "luddites" are the only ones that win in the long run
you are just so short-sighted you don't realize how long the run actually is
200 years is nothing in the terms of human history, every civilization that outmoded it's own population via progress for it's own sake was torn apart by the people it abandoned

>>102997791
Yeah society was torn apart but the luddites back in the 1800s wasn't it

File: file.jpg (320 KB, 1024x708)

320 KB JPG

>>102997800
Just look at the culmination of their efforts.

>>102997761
>Luddites never win in the long run buddy
Luddites were a thing less than 200 years ago, and we're still struggling with the same socioeconomic issues they did because we haven't fixed the problem, and in fact have expanded their problem to the majority of the populace. Might wanna hold off on absolute statements like that bud. "The Mercantalists never lose, our way has been working for over three centuries!" said the merchant confidently, in the year 1790, less than a century before he was decimated.

>>102997930
So how many more millenia do we need to wait until they finally win and we destroy all technology and go back to living in trees?

>>102997946
That's literally what you said, 200 years is nothing
So how long is it gonna be

>>102940849
>where people are trying to learn new things
are we on the same board? i come here out of bad habit and it's equally disappointing every time.

>>102959784
it's whiplash/fatal racing. a dos game
https://www.youtube.com/watch?v=nSgiPQmfBHQ

unreal trannies dunked on again
https://www.youtube.com/watch?v=07UFu-OX1yI

>>102966595
No. Working with people in this woke era is a risk not worth taking.

Too many /pol/kiddies in this thread lately

>>102966618
...so use AI instead of voice actors.

>>102966618
>>102966791
>>102967435
Just look at how these people react, talk to you, and condescend you.
Why would you want to work with these type of people?
Do what you need to do and don't ask these faggots for their opinion. They always perceive it as you asking them for permission.

I managed to squeeze even more performance out of my ECS framework by replacing the map for mapping component masks to archetypes with a custom dense map data structure optimized for integral keys (which component masks are). Pretty great, now getting:

> Execution time @ 1k iterations with two components:

1k entities:    968.61us (0.97ns cost-per-entity)
100k entities:  144.39ms (1.44ns CPE)
1m entities:    2.44s (2.44ns CPE)

>>102999627
You are doing ECS wrong and completely missing the point

>>102999635
Bevy is my inspiration -- it has the nicest API out of any I've used.

My implementation is definitely not faithful ECS but it's incredibly productive (for me)

>>102999647
Bevy is complete garbage
I almost feel sorry for people raised on these cults of software architecture completely detached from real game development

>>102999665
>Bevy is complete garbage
source?

>>102999665
It's essentially just easy, automatic dependency injection with functions you want to run at a specific point (default is update stage). It's so easy to wrap your head around. No learning curve, no considerations for "architectural design"

I thought Unity 6 is going to have a fresh start and make things that work.
But when I go and try out the multiplayer widget, turned out it's a quickly setup widget that you can't even reuse for your actual game, and is just meant for prototyping.
Unity needs to stop with these useless features that don't help you do things faster.

File: pkm.webm (2.99 MB, 1920x1080)

2.99 MB WEBM

I've faithfully recreated game boy color aesthetics