/v/ - Video Games

Romhacking is the worst fucking thing, everything is arbitrary and depending on who made it, things are fucking all over the place and there's just enough sense and logic but then in another part it's completely fucking different and a pain in the ass

>>16895800

I'm trying to make my own raycast engine without copying everything from lodev's tutorial like everybody else does while pretending they understood a damn thing, and I'm also trying to understand how to reverse engineer a game like they do in vid related because I'd like to port some old games I used to like to PC.

I am trying and failing to turn a steering wheel in a car when the wheels turn. Unreal Engine 4.24

>>16895800

I have been struck by a fey mood recently and I've started to create what might be best described as a cross between Dwarf Fortress Adventure mode and Era games.

The idea is that you are a guy in a modern city that's going to be simulated as well as I can, with every NPC having their own desires and preferences, workplace, home, etc. and your goal is to find a chick that appeals to you (or a client) follow them around, learn their preferences and routines through good old-fashioned stalking, then use sneaking and a variety of tools to manipulate their normal life to get them to a spot where you can kidnap them easily and then the training starts…

This is about my 5th or 6th attempt at this, and I'm getting further with every iteration. Looking good so far, I'm doing the random city generation now, after having coded NPC behavior in a test room..

Good random generation is hard, really hard. But hey, I'm learning.

>>16895837

Use Unity's ECS, You will be needing to update a ton of stuff a whole lot of time, and it is pretty fucking good at it.

Or you can learn how simd works and whatever other magic they use

>>16895791

>You can't make a game.

What about making a Japanese game?

>>16895807

>raycast engine without copying everything from lodev's tutorial

I usually copy everything just to make it work then I refactor it. Easier to understand and you don't risk mixing the tutorial with your own mistakes. Why raycast though?

>>16895837

Sounds huge, sounds fun.

>>16895854

Polygon prototype pack is $5, you can prototype any game with it (https://syntystore.com/products/polygon-prototype-pack). I am also completely incapable of producing any art so I just buy everything. It is limiting but limitation isn't always bad. Not like I have ever finished a game either way.

>>16895867

>Polygon prototype pack is $5

There is a ton of organized stuff there, but I bet you can get the same scourging for free stuff. But then again, it's about how much your time is worth

>>16895867

>Why raycast though?

I want to get a better understanding of mathematics in general and I can't 3d model properly. My end goal is to make a game like Princess Maker mixed with Bully where you'd be able to sometimes run away from the school to do some dungeon exploring.

>>16895854

Make everything basic geometric shapes and pretend it's your "style."

>>16895846

Thanks for the tip, but I'm using Godot. If need be, I'll just use C++ for whatever causes bottlenecks if and when they pop up. I'm kinda looking forward to those investigations even.

>>16895854

Try using Anim8or, it's much simpler to use than Blender. Use Anim8orTransl8or to convert the result into a Unity-friendly format.

>>16895867

I need help with my code. I'm doing a raycast, and for some reason, it stops at the 533rd iteration, no overflow or anything, it just doesn't go further than that.

int main(){
    const uint16_t HEIGHT    = 600;
    const uint16_t WIDTH     = 800;
    const float       BLOCKSIZE = 1.0f;

    const float FOV   = 60.0f;

    float ANGLE = 10.0f;

    float px = 5.5f;
    float py = 4.5f;

    sf::RenderWindow window(sf::VideoMode(WIDTH, HEIGHT), "Raycast");
    window.setFramerateLimit(60);

    while(window.isOpen()){
        for(uint16_t x = 0; x < WIDTH; x++){
            bool collision = false;

            const float curr_angle = to_radians((ANGLE + FOV/2)-x *(FOV/WIDTH));
            const std::pair<int, int> inc = ray_direction(curr_angle);

            //current tile
            const int cx = static_cast<int>(px);
            const int cy = static_cast<int>(py);

            //next x
            float nx = 0.0f;
            //horizontal x and y
            float hx = 0.0f;
            float hy = 0.0f;
            //dist to next horizontal
            float nh = 0.0f;

            //next y
            float ny = 0.0f;
            //vertical x and y
            float vx = 0.0f;
            float vy = 0.0f;
            //dist to next vertial
            float nv = 0.0f;

            if(inc.first == 1){
                nx = BLOCKSIZE - (px - cx);
                nv = std::tan(curr_angle);
            }else{
                nx = -BLOCKSIZE * (px - cx);
                nv = -std::tan(curr_angle);
            }
            if(inc.second == 1){
                ny = BLOCKSIZE - (py - cy);
                nh = 1/std::tan(curr_angle);
            }else{
                ny = -BLOCKSIZE * (py - cy);
                nh = -1/std::tan(curr_angle);
            }

            //first collision with horizontal gridline
            hx = ny/std::tan(curr_angle);
            hy = ny;

            //first collision with vertical gridline
            vx = nx;
            vy = nx * std::tan(curr_angle);

            //distance to the wall
            float w_dist = 0.0f;

            while(collision == false){
                if(hx > vy){
                    vx + 1;
                    vy + ny;

                    if(level[px+vx][py+vy] > 0){
                        collision = true;
                        w_dist = std::sqrt(vx*vx + vy*vy);
                    }
                }else{
                    hx + nv;
                    hy + 1;

                    if(level[px+hx][py+hy] > 0){
                        collision = true;
                        w_dist = std::sqrt(hx*hx + hy*hy);
                    }
                }
            }
            std::cout << x << "\n";
        }
    }
    return 0;
}

>>16896049

Throw in some debug prints to help you figure it out. Is it getting stuck in that last while loop on iteration 533 because collision's never getting set to true? I haven't done any analysis to see if that's what's actually happening, but with a while loop running around I figure that's a good place to start looking.

>Take a color and try to add white to brighten it

>It ends up really fucking washed out and white instead of bright

>Was doing Lerp

Shit, what was the solution to this again? I've implemented it before but I can't fucking remember shit

Any tips on networking with other people? I'm trying to think of the best way to get the word out about my game so that, at the very least, people will play it. Only problem is, I'm a friendless autist and don't know where to start. No clue how to network with people or who to even network with.

The approach I'm thinking of would be to set aside the game for now, get good at some creative skill that people like (I'm thinking music), use it to build up a following and hopefully gain connections, then use that as a springboard to at least get the demo off the ground. As long as I'm connected in some way, I think I'll be fine. But even here I'm lost. Where do you share your stuff? How do you build a following? Did other creators just show up on Twitter/Tumblr/Youtube with no followers and just start posting stuff? I doubt it. They had to have shared their stuff elsewhere so people can see it and they'd build a following. But where?

>imageboards are the only place I'm familiar with, but the mere association with them can get you shot down

>named communities like forums and IRCs require you to mingle in with the community, something I can't do, before you can post stuff; otherwise you get banned/shunned for spamming

>posting shit on social media alone and expecting people to find it is like expecting a coin to land on its edge with heads right-side-up

>>16896049

probably looping in the while collision, most likely in the else branch of (hx > vy)

>>16896194

Crysisdude is looking rad

>>16895867

>Polygon prototype pack

does it still have that "you have to allow us to audit your business to make sure not 2 people at once use our precious assets!" clause? fuck that shit. especially when for most stuff kenney works just as well (plus being free)

https://opengameart.org/users/kenney

>>16896208

Looks like it was this

https://en.wikipedia.org/wiki/Blend_modes#Overlay

>>16896301

ill make adequate art for yu, only if yu pay me money 4 fish fingers

>>16896049

Does it stop as in it exits the loop, segfault or just loops endlessly?

>>16896301

I don't give much for that kind of clauses as they will never be able to act upon nor is it probably sustainable for them to do it.

>>16896194

Is crysis man PBR? Feels like a nice demo character to use.

>>16896492

Why do they look so menacing

>>16896500

because they are t-posing

>>16896437

It loops endlessly.

>>16896572

Well, for starters, in your while loop:

[code]

vx + 1;

vy + ny;

…..

hx + nv;

hy + 1;

/code]

Has no assignment, so they don't actually do anything.

>>16896760

I know, I fixed that and the overflow as well, now it doesn't stay stuck on a certain iteration but it only checks the if(hx > vy) and not the else and that's bad.

            while(collision == false){
                if(hx > vy){
                    std::cout << "v: " << curr_angle << " " << level[px+vx][py+vy] << " " << vx << " " << vy << " " << x << '\n';
                    if(level[int(px+vx)][int(py+vy)] > 0){
                        collision = true;
                        w_dist = std::sqrt(vx*vx + vy*vy);
                    }else{
                        vx += 1;
                        vy += ny;
                    }
                }else{
                    std::cout << "h: " << curr_angle << " " << level[px+hx][py+hy] << " " << hx << " " << hy << " " << x << '\n';
                    if(level[int(px+hx)][int(py+hy)] > 0){
                        collision = true;
                        w_dist = std::sqrt(hx*hx + hy*hy);
                    }else{
                        hx += nv;
                        hy += 1;
                    }
                }
            }

>>16896949

It wasn't the error but it helped me fix a mistake I shouldn't have made in the code, thanks.

Here's a pastebin for anyone who cares:

https://pastebin.com/V3vYA9cN

>I just came off a 4 hour debugging session with my driver segfaulting, turned out a typo had me submitting a mismatching pipeline/pipelinelayout. At least I have rendering to an offscreen buffer working now. Tomorrow I need to refactor the whole fucking thing I think.

What kind of project are you doing exactly?

>>16897001

>Well done anon!

Not really, It still only checks the if(hx>vy).

>Working in it-consulting is not very stimulating.

If it pays the bills and allows you to do gamedev on the side, why do you try to get rid of it when it could be the thing that enables you to make your own games and be your own boss?

Also: https://www.reddit.com/r/gameDevClassifieds/

I know it's reddit, but people seem to have success finding work with this subreddit.

>>16897021

>why do you try to get rid of it when it could be the thing that enables you to make your own games and be your own boss?

It's more about what I currently do being quite boring so I figure I would do a couple of years in gamedev, if I can, then get back to the more stable consulting business after that. I don't consider gamedev (outside of mobileshit) a viable long term career, it just doesn't pay well enough. Thanks for the heads up on that sub, have never seen it before.

>>16897354

> don't consider gamedev (outside of mobileshit) a viable long term career, it just doesn't pay well enough.

Porn visual novels who bait fans do quite well.

Check the related communities part on the side:

https://www.reddit.com/r/gamedev/

>>16897760

Show us.

>>16898438

But that's exactly what it is when what you're doing is trying to get software A working with software B using third-party tool C, regardless of what information you're trying to push around. Making software work with other software might be essential to the job, but it's not really part of the job. I've done at least two types of work where big important things couldn't happen if someone couldn't figure out and maintain a pipeline for getting two pieces of software passing information back and forth, but even though I had to help work on that problem nobody would say that was part of the job description. It's really a consequence of how computerized certain jobs are these days. Naturally game development is kind inextricable from modern computer problems, but even things like healthcare have lots of problems like that. Polite sage for offtopic flashbackposting.

Reached another milestone.

>>16898949

You're going to make so much money with that the day you release it.

>>16898949

congratulations

>>16898977

probably the power of love only a man can have for his waifu

>>16898984

>>16898949

I myself will only offer you one (1) congratulation

If you are taking assets from a unity game to use in your game, is it better to make the game using iunity or UE4? (i.e. unity3d into FBX and imported into UE4)

Is it possible for unity to open a unity3d file that is not part of the file?

>>16899245

>Generating

Why are you doing that? What are you trying to do? Are you building them raw out of the TBN Matrix or what?

>>16899389

I dont understand why you are choosing your game engine based on some premade assets that clearly are only meant to be used with Unity. You can't open Unity3D files and import them into UE4. You will have to open it in Unity and then copy the fbx into an Unreal project.

>>16899931

>>16899878

>Why are you doing that? What are you trying to do? Are you building them raw out of the TBN Matrix or what?

Short explanation:

The "mesh" is made out of many meshes drawn using Graphics.DrawMeshInstancedIndirect, so even though the vertices are the same, they are in different meshes, and therefore will have different normals, especially since many of those are the same mesh rotated 90º, so even though two meshes 'share the same vertice', their normals will be completely different, therefore the normal needs to be generated in the shader

long explanation: https://forum.unity.com/threads/generating-normals-in-shader.859306/

>>16900019

Oh I see you are trying to make some procedural mesh. Well my first advice is to not use DrawMeshInstancedIndirect and instead just update the entire mesh at once. If you need to you can split it up into "chunks" but you shouldn't ever worry about this optimization until you can prove it is necessary. You are trying to do everything procedurally in the shader right? That sounds like a mess and waste of processing. Never settle for "on the fly" what you can calculate once and be done with. Besides, it defeats the points if you are already looping through to build the vertices, might as well go the full length and add in the normals as well.

Build the surface normals first. After that build the side normals. Keep them separate and you should be fine.

>>16900185

What's 1MA?

>>16899931

>You can't open Unity3D files and import them into UE4

so would that mean that the unity3d asset is only useful for unity and not UE4?

>You will have to open it in Unity and then copy the fbx into an Unreal project.

can even open up the unity3d file despite not having the source files?

You know, after putting a lot of thought into this, I am thinking I should try to make the game in unity, instead of UE4

>>16900232

One Man Army.

>>16900236

There's probably some tool that let's you retrieve source data in some way but in general yes, unity assets are only useable in unity.

>>16900249

>just tile squares

Have you considered not giving a crap about building the normals on the fly and just making the squares overlap and hiding the neighboring edges? You are instancing the mesh anyways so memory isn't an issue.

>>16900246

that is the million dollar question, from what i read online, there is some browser plugin that lets you open up unity3d files.

>>16900246

What are you missing to be able to do it all yourself?

>>16900258

I just know how to program, any attempts at art, animation or sound fails spectacularly. I don't really have the patience to learn it either when I can just get better at programming instead.

>>16900256

I don't think it's worth the hassle really.

>>16900256

>unity3d files

You mean decompressing a unity's game asset files? Like those Level0.asset?

There should be a way, since the game engine does it, but I never saw it

>>16900264

I can draw anime but I don't have the board so I'm stuck too.

>>16900264

>I don't think it's worth the hassle really.

yeah, trying to rip the model from koikatsu (unity) is an annoying experience, and i figured, that i'd be better off making my own

now I am debating whether i should make my yakuza/shenmue H game in UE4 or unity

>>16900296

i literally know nothing about either, but i do know C++, so i am going to stick to UE4

>>16900296

I don't think I've made it as an artfag, I'm still a beginner, it's just that drawing anime is easy.

>>16900256

>>16900265

There's a program for extracting at least some unity assets called AssetStudio or something. Dunno if it works with meshes or what format the results come out in if it does work with them. I've only ever used it for resources like textures and audio.

how do I snap objects to an arbitrary grid?

Said grid is 1x1 in height/width, but it's rotation is arbitrary

My current grid data is: Position (a Vector2 that is [0, 1), it is how much the grid deviates from origin) and Rotation (a Quaternion, that specifies the grid rotation). The grid ignores Y position, it's only for X and Z.

I underestimated the math for this, since it's trivial to snap to axis aligned grids, I assumed it would be easy, but I actually have no idea where to start

So, how many of you are motivated by the FF7 remake that is really not one to get into gamedev even more?

>>16902176

It looks good. What game are you making exactly?

When it comes to unity, how would one focus on "optimizing" their game?

>>16902509

Do as much as you can on ECS

It's strange, full of limitations, and requires very verbose code but it automagically aligns all data in a exceedingly performant pattern and intrinsically multi threaded

>>16900249

>DrawMeshInstancedIndirect

that might be it, never needed it for myself and none of the stuff I watched out of curiosity had mentioned it.

https://www.youtube.com/watch?v=NpeYTcS7n-M isn't the one I remember, but might be helpful.

nevermind, found it: https://www.youtube.com/watch?v=4RpVBYW1r5M

>>16900283

from what I remember modding HS, you can rip the individual parts, but you'd have to combine them yourself. I think com3d2 allows you export of the whole model or has a mod for it.

worst case just grab a model of the marketplace, it might not be free but depending what you want can save you quite a few hours (and time is money).

>>16902509

focus out the bottleneck, you can make any engine throw a bitchfit. as >>16902515 said ECS is a big boost but it's still not finalized and missing parts. and depending where you issue is won't help much if it's simply having too much shit on screen in the wrong way or unoptimized code

>>16902558

Well okay, what counts as "unoptimized code", just so I can have something in mind when writing code.

>>16902561

run the code and check the profiler

whatever is costing more than 1 ms is probably needing a rewrite, but then again, some stuff is just heavy and even perfectly implemented will still take more than 1ms, so it's a case by case basis thing

>>16902509

Idk about Unity but general advice:

>General performance guideline is "do less work" (e.g. avoiding unnecessary copies/copying large amounts of data every frame, recalculating something every frame when in reality it only changes once every minute or so)

>Guidelines are not rules

>Avoid optimizing until you have a good idea of typical running conditions, since a lot of optimizations involve assumptions about those conditions, and if you don't know which part of your code is actually slow under typical conditions, you might be wasting time optimizing a part of your code that isn't performance-critical

>90/10 rule: 90% of a program's execution time is spent executing 10% of the code

>If it's too slow, use a profiler and find that 10% of the code

>>16902509

>optimizing the usage of memory? amount of writes to disk? are you using caches efficiently? how fast in terms of milliseconds or even microseconds (not fps) can you process this?

ideally you want a goal in mind and the most important, benchmark the differences, do note that: your optimization might be pretty good on your machine but not so good on a lower end or doesn't scale as well on a higher end machine.

>>16902509

I always made a prototype, then proceeded to play with certain parameters - like numbers of NPCs per level, numbers of projectiles, that kind of stuff - to see which would have the worst performance impacts, then I look at that code and see what can be done better. But only do this if it's actually necessary, unless you're just trying to do it as a programming exercise.

Best thing to look out for, at least for my average code, is iterating over arrays and how to combine that so you minimize the amount of times you do it. Try to keep an eye out for your personal typical weaknesses, it makes optimization much faster.

>>16902509

General optimisation and common sense, really.

Open the task manager, see how much memory it's taking. If it's a lot, you might want to find a better way to store whatever variables you're using instead of that 500*500*500 array of custom types.

Debug.Log and similar functions actually cost more than you'd expect, remember to remove them once you no longer need that.

If you're going with dynamic lighting, every Point Light will force the render to draw another lightmap, so keep these types of lights at a minimum.

Similary if you're using dynamic environment, Unity doesn't really do occlusion from what I've seen. Could be wrong, but you're better off making an algorythm to hide stuff you don't need to see.

Only use Update when you really have to. They do have a cost, even if they are entirely empty and you're often better having a method that something else calls for in specific circunstances than several random Updates waiting on a condition.

Other than that, don't worry too much yet. Be efficient while you write\draw and optimization follows. Once the game is done, check if it's running well and fix it at this moment if you really have to.

Is there a program that counts pixels? Like you draw pixels and then the program tells you how many of them there are.

I just realized how useful that would be for autistic building games like Terraria and Minecraft and Stardew Valley so I wanted to make a program like that, but maybe it already exists.

>>16902718

I've never heard about anything like this.

>>16902718

You're literally just iterating a T[] and tallying unique elements. This is a trivial problem

>>16905114

Nevermind, there's a whole thing about this

https://en.wikipedia.org/wiki/Connected-component_labeling

>>16905114

Use whatever movement algo you're using for NPCs (like A*) to try and connect to the first room you've generated. If there's a way to every room from there everything is connected. If a room can't be reached from there, it's not.

Alternatively, start with every tile being a wall, then carve out rooms, then go from room to room and carve a path to the next room in sequence. That way you'll always know that everything is connected.

Is it a good idea to use the source engine to make an action rpg or should I use Godot?

>>16905149

so you just want a connected layout but necessarily a straight line but no maze either?

if it changes over time you could check if tiles getting blacked cuts off the path, else pick a different tile. another idea would be generating layouts till you find one that works (probably faster than trying to optimize one, but I'm still wrestling with graph theory myself, so someone else might know more in that regard).

www.redblobgames.com might have some math that might be helpful, least did for me.

>>16905240

both c++ so godot, if you wrestle with an existing engine might as well pick the more open one that also gives you more freedom. never used godot but I assume it also works better for an action rpg too.

>>16905240

I see no reason to use source unless you're either making a mod or really want to use Hammer for some reason.

>>16905307

>>16905367

I was wondering because I'd like an engine that doesn't require a lot of power for the 3D.

>>16905478

That's more down to your implementation than the engine you pick.

>>16905522

If the AK mags can't just be a straight line block they should be a 3 block L rather than a 4 block Z. That bottom block barely has any magazine in it and disappears when loaded into a rifle regardless.

>>16905623

all the shapes are auto-generated, those few pixels on the bottom make all the difference to the generator. That can be tweaked, though. I'll play around with it, though stacking them this way is fun

>>16905671

That makes sense. I suppose as you keep testing you'll figure out what works best. Good to hear it's adjustable so you can fine tune problem areas if they arise.

I would suspect that during gameplay if the inventory is full most of the time you or your play testers will find a small section of an item taking up an entire block as more annoying than fun. Then again, you have the game in-front of you, and I don't.

>>16905689

next Demo Day in less than a month. It will be fun to read the feedback on this and all the other stuff that will be added by then

Theme slaughter round is live for LD46, anyone here going to join? https://ldjam.com/events/ludum-dare/46/theme

>>16907446

>someone finally reminds me of LD before the event starts

>it's during a lockdown anyway so I have the time

Thanks anon. Thanks to you I can actually get my tools in order beforehand for a change. I'm joining.

>>16908954

Full thread about the game here >>>/agdg/>>33710

shmup anon here,

Last time I think I just fixed trails and blur effect (somehow)

Next I worked on camera movement which made me realize I needed to differentiate between ground units and player layer. I went with adding a unit (vertex) attribute and moved camera as well as lighting in their own uniform buffer.

Then for some reason or along the way I noticed the framerate would keep at 50FPS, so I re-arranged the delay calculation and it's more stable at 60FPS. While I'm at it I also tested the windows build but there's something fucky that I would need to fix later.

Then I was thinking about how I would manage background scrolling and other elements, but it should follow the same process than regular units:

Using the common model buffer for rendering, then have the specific background data buffer for element placement, UV shifting, etc… That I would need for background.

I'm almost feeling how to re-arrange the code to generalize this, but it's still a bit fuzzy.

>>16910829

Should the worst happen, where can we find progress on it? Or at that point will it just be dust in the wind?

Just want to give a heads up to all my fellow aggy daggiers that there are gonna be hard times ahead.

See also >>16910985

>>16910900

>Should the worst happen, where can we find progress on it?

my suggestion is that everyone bookmark the wiki and just coordinate there if shtf.

>>16911122

The wiki could use an update. Thread archives haven't been updated since before the domain change. The projects list is very dated. There's no page to list bunkers.

>>16895791

>the cube engine lets players edit levels in real time

>not a single cube engine game has a destructible environment

>>16911471

Wrong thread?

>>16911447

You have no idea how much I'd like to update it but I don't have time. I even wanted to upload and host the archives somewhere as html

Cos yeah, I have been archiving that and all bunkers

>>16910900

>Will it just be dust in the wind?

I will continue to work so long as the UT streams continue. If they stop, I'll probably do what Die Totenmaske did and make something original using what I've built. Though, I will probably remake everything in 3D and overhaul a lot of the systems simply due to the fact that 2D animation is really difficult and I don't think I can afford commissioning animations right now.

>>16911122

>my suggestion is that everyone bookmark the wiki and just coordinate there if shtf.

I had not considered using the wiki. I think I'll go make an article on the game now.

>>16911501

How do you update a wiki?

>>16911555

You go to the page and click the edit button. It helps to know wiki syntax. Your IP or part of your IP will be shown in the change log unless you use an account.

>>16911587

Ok, I'll do a couple of updates later.

>>16910997

How many of you anons are gonna stick around here? I've mostly been using the /agdg/ board as a slow archive of progress, and these threads to show off / ask questions. Is it worth going to the bunker or another site??

>>16911696

I'll stick around to see what happens. This place was already pretty slow and I don't think we will survive another split.

>>16911696

I'll go wherever the community is. It would be nice to have a working site but we can't enjoy it without users.

>>16911696

I'd be fine with moving to another site, 8kunt is a laggy piece of shit and jim is an even kike than cakejew.

>>16910997

>>16911696

If you really cared about any of this then you'd have left long time ago.

But if you want to explore options then there's currently 2 in the webring.

https://julay.world/v/res/4470.html

https://anon.cafe/agdg/

>>16911696

I'm staying here unless the community moves somewhere else.

>>16911799

Oh, forgot about this one. https://smuglo.li/vg/res/129.html

>>16911824

/agdg/ also mentions this

https://endchan.net/v/res/11733.html and 2chen (ded), which was our primary bunker untin vch.

there really was some heavy splintering going on. Both smug and endchan threads (and boards) are inactive.

>>>/agdg/29079

>>16911863

Didn't Endchan go down at the same time as 8chan? That's not a good sign for a bunker.

>>16911875

isn't it also run by some tranny buttbuddies of cripplekike?

What about Josh from kiwifarms? Doesn't he have a board somewhere?

>>16911863

This is the only non-4chan /agdg/ with any semblance of real activity. julayworld's thread is mostly shitposting but it's the next most active.

>>16911905

If we're migrating then we should take the opportunity to become free of all these autistic jews, not move from one to another.

>>16911875

it wasn't. didn't live long but it was (and now apparently is) there.

>>16911714

>I don't think we will survive another split

and some devs never came back after the last exodus. No idea what doctard, Sentiencedev or godot stralthdev are up to

>>16911916

>only non-4chan /agdg/ with any semblance of real activity

I'm aware. There's /gd/ but it definitely isn't the same thing

>>16911905

I thought Josh was codemonkey, or at least someone on the team. Christ, there's so many name I know but don't understand, since HW buggered off those years ago.

>>16911956

Jim is the owner of 8kun and 2channel, and the one who's most interested in /qresearch/ and pandering to them.

Ron (codemonkey) is Jim's son and the only actual developer we hear anything from, though we don't hear much from him either.

Josh (runs kiwifarms) is some retard who was hired long time ago to create new software for 8chan from scratch because they could hire him for peanuts, but it failed. He's been mostly irrelevant since except for attempts to create 16chan/nextchan or whatever, and failing those too.

I wouldn't characterize any one of them as stable or trustworthy people.

>>16912057

You're supposed to respect all of these rules or just a few?

>>16912064

It's just voting so we get 3 rounds of different themes then a final round voting between the top themes from the first 3. Only one is selected at the end and you don't have to follow it but it's encouraged, especially for compo.

>>16912057

>turn a violent genre non-violent

Could be interesting.

>1 minute to prepare

Sounds suitable for a game jam, but also pretty restrictive, like everyone's going to make some wave defense game in essence.

>fragile

>side effects

I can imagine some interesting ideas for these.

>>16912082

They just all feel like they would deliver samey games to me, which would make a gamejam really boring. But we'll see, I have no ideas myself so far at least.

>>16912057

Decay could be interesting if you have "less" of things as time goes on. For example, reduced stats or enemies, or the tile-based level literally disintegrates. For Garden, I'm thinking a SimCity clone

>>16911527

>I think I'll go make an article on the game now.

I have created an article on the game. I probably won't post any story updates there, but I will hopefully remember to keep it up to date with any mechanical changes to the game.

If the worst comes to pass and this site goes down, I'll probably start putting story updates there as well. Would be better to have a place people could comment things, though. I much prefer having someone to talk to about the game, even if it's something as simple as "I hate everything you're doing."

>>16912057

>turn a violent genre non violent

I always wanted to see a Survival Horror game that had "happy" undertones, in the sense that YOU are the horror and THE ENEMY must survive you.

>>16912082

I know there was some youtube letsplaybait game a some point where you had 60 seconds to gather shit before the nukes dropped, and then had some roguelite elements as you tried to have your family survive in a bunker.

Could also make a game where your character goes through a dungeon himself, but you have one minute to gather some stat things or equipment.

Or a game where you get to read instructions for solving some elaborate puzzle for one minute, like Keep Talking and Nobody Explodes kinda.

I don't like that theme, but more because it'll lead to short-ass games not because there's no interesting ideas.

>>16912240

>and THE ENEMY must survive you

That's violent.>Sounds suitable for a game jam, but also pretty restrictive, like everyone's going to make some wave defense game in essence.

>>16911991

Is Josh the guy who made that new imageboard software that shat the bed when running on 8ch's servers? real fucking shame, on the test website that shit ran like a dream.

>>16912350

It was because of the moderating bot or something like that.

>>16912357

What was because of the moderating bot? I just found a website that uses Infinity next and it's source code repository which still gets regular updates, still runs like a dream, kind of makes me a bit sad.

>>16912350

The software was fine until you put a lot of traffic through it, that's when it shat the bed. If I recall right he never properly fixed it even when he continued developing it on his own website, but that was a long time ago, maybe it's been fixed since.

>>16912357

I think you're thinking of the /a/ mod bot. That's completely unrelated, it always caused a lot of stress to 8chan.

>>16912378

He's still developing it based on what I've just found, I wish Codemoney would consider a test of that software again, it still at least on the test website I checked runs very smoothly, although it still also needs a lot of work.

>>16911696

>How many of you anons are gonna stick around here?

Abandoned this shithole months ago to one of the alts that was posted. I thought it was bad before 8chan went down, but good lord. Only came back to see what you guys were up to.

>>16913091

Is there a problem with this specific thread?

>>16913195

>4chan

>discord

What the fuck

>>16913507

We already knew he's a crossposting furfag, why does discord surprise you?

>>16911474

I'm implying someone should make a Cube engine game with a destructible environment dummy

>>16913524

I have a hard time keeping up with individual devs to be honest

>>16913524

aren't you confusing people

>>16913567

See >>16902869

The furfag is the same person working on Plutonia map edits. Just because his IP switched doesn't mean it's not the same person.

Here is the final Congo edit.

>>16913507

They are more active than here. More progress posting. I only post on the discord though, their thread sucks about as much as this one. I tried hopping into the irc for this AGDG and no one is ever on, so I stopped. I'm here to post progress.

>>16913601

This.

>>16913657

Well of Souls edit.

>>16913657

Aztec edit.

>>16913661

>>16913659

It's the same video.

>>16913686

Fixed.

>>16913731

bigger tits

>>16898977

He's like 450YDPM (YandereDevs per Minute) of fast.

>>16913768

The ethics committee has said they're too big now, change it.

>>16914880

Why does she have two testicles but no penis?

>>16914893

https://www.invidio.us/watch?v=uyNJZ-2Dod4

>>16914916

Your animations look good, not so much your textures.

>>16914925

They're PSX style pixel art. Maybe I'll get some dedicated artist to spruce them up for me.

>>16913948

And round 3 is as well, but you can still vote for round 2 if you missed it.

>>16915276

Super fun to play around with but not a totally straightforward game to use. It heavily relies on logic circuits.

So Ludum Dare voting is supposed to be 3 unique batches, then the most popular ones from the 3 rounds in the final round right?

Then why is Deja Vu in all 3 of them? A meta shitpost?

>>16915437

>Then why is Deja Vu in all 3 of them? A meta shitpost?

I think it's a joke but I can't really tell. Seems like a dumb thing to shit up the voting "for the memes".

>>16915594

Give him a chance, he's probably in the early days of the making of his game.

>>16915594

Well that's cool and all, but I like my games to look good, AKA not made by a chinky monkey. Those examples look like utter numale pixel shit and are probably from a trash game as well.

>>16915595

Nah. About a year off and on.

>>16914916

You should really be basing your textures off of real photos which have been edited and shrunk down to 256x256 or below. 128x128 and 64x64 were common.

Most ps1 art like that was drawn as textures first at high detail usually based off real photos then applied to a model.

>>16915594 This anon gets it. Color choice and contrast are everything with older style textures.

>>16915595

Stop being a fag enabler. This attitude does nothing to help him improve. Who gives a fuck if he's in the early days? If he can't take this much and improve on himself he will never make it to begin with.

TLDR: Look up what actual PS1 games look liked you faggot.

>>16916287

I don't think >>16915594 was enabling shit textures, though. He can't improve if no one tells him his textures are not "PSX-like", which you are also doing.

>>16916242

>For now, I can just fuck around with free assets in godot perfecting game mechanics and honing an story, but I know it will never happen, and if it does, it probably won't look anything like I really wanted it to.

Do it, and teach yourself how to draw/make pixel art. You have to start somewhere.

>>16916311

Time is the enemy. I've been trying to improve my art skills to. One thing I did is try to make a 3d model every day within 1 hour for a while. Regardless of how shit it ends up I still timeboxed it to an hour. Maybe if I could cut it up, like doing the prims one day and the textures another, but it was still an hour. Right now I kinda accepted I'll never achieve my game while working full time, so I'm trying to make changes to switch to either part time or just quitting and living extremely rural. Problem is corona fucked up the economy so I can't switch jobs easily.

>>16916331

>but even then and will all the free time I have, I barely manage to improve

The fact that you're probably worried about not starving or ending up homeless most likely has something to do with it. Sometimes I feel the same, but I think of it that way: How closer to my goal would I be had I not wasted time waiting until everything was smooth?

>>16916349

Maybe focus on the project you think you can monetize and publicize, that way you'll be able to make some cash at one point. Also maybe you should hang out on some other game dev forums who aren't too pozzed so you can build a fanbase.

>>16916296

Well I should've clarified. I want it to look like GOOD psx games. I know there's tons of shitty looking psx games out there as well, and well- obviously I shouldn't try to copy those aesthetics.

>>16916560

I know that I have rustled your panties, nodev, but I typically think games with style and not garbled numale japmonkey games are usually the best looking. Quake and Journey come to mind.

>>16916331

I was mentioning the 3d practice I do more as something you could try with 2d for practice. I actually started doing it because of Bob Ross' videos where he'd paint an entire landscape in 30 minutes. Honestly just have Bob playing in the background when you do shit. Dudes re-assuring as fuck and it'll show you the main way you paint shit in art is not by stressing the tiny details.

Honestly you just gotta grind. Maybe you can try doing pixelart for other people as practice and eventually replace your odd jobs with that if you get to a point where people start paying you.

>>16916511

Your game doesn't look like a PS1 game at all though.

>>16916630

Nightmare Creatures was made by a western studio and your gun leans way closer to the Jap style.

>>16916360

>Also maybe you should hang out on some other game dev forums who aren't too pozzed so you can build a fanbase.

Trying to find a dev group in general that isn't pozzed would be a miracle. I wanted to go to the local indie dev events to see what would come of it but then corona broke out.

>>16916746

Stop arguing and get back to work.

>>16916770

Were you born retarded, or did your mom drop you on your head as a baby?

So far you've done nothing but say dumb shit, claim you didn't say said dumb shit when it's visible to anyone who scrolls up, and show garbage models and taste. As the other anon said, get back to work. Or fuck off.

>>16916770

Your gun looks hidous. Take another look at it. The colors make 0 sense and aren't even a pleasing combination >>16914916

>>16916770

What a fuckign sperg

>>16916770

Jesus fucking Christ how new do you have to be holy shit

>>16916713

My strategy is once I get the base of my game working, I'm going to post on some C++ forum, the SDL forum, and maybe I'm going to go on gamedev.com and a roguelike/roguelite forum/

>>16916770

good lord you sound like your average passive agressive cunt on reddit trying to be funny to deflect any criticism. funny thing is, that is a world model (that still looks better than the shit in your game) and not a first person model that has a higher detail level, if you are trying to make a game inspired by old pc games why the fuck would you look at psx games?

>>16917168

>What are you doing? I like what I'm seeing.

I'm just developing a graphics engine for now, to update my C++/graphics programming skills. I'm considering porting over my lua game engine framework to work with this graphics engine but I haven't decided yet.

FUCKING CODENIGGER LET ME POST

Final LD voting round is live. Options are:

>Dreams

>You are the power source

>Turn a violent genre non-violent

>An abandoned world

>1 minute to prepare

>Decay

>Side effects

>Fragile

>Outdated technology

>Ruins

>Momentum

>Automation

>Keep it alive

>Garden

>Connections

Better go vote for the ones you like and maybe try coming up with an idea for everything.

I think I'm gonna try make a game with ct.js. I doubt it, but anyone got any experience with it to know if it's shit or not aside from javascript being shit in and of itself, but you generally want things playable in browser if you want people to actually play your LD games?

>>16918600

>schizo

Next you'll say "Take your meds." Christ, you shills really do follow a script.

>>16917909

>>16918600

>haha cope ecks dee

Get the fuck back to reddit.

>>16918652

I want both of you to fuck off.

Block IDs and stop giving people like this attention. Just make your game, if you get good criticism or shit criticism, consider it and keep devving. Please don't shit up the thread, it never solves anything.

>>16918707

>Interdasting. I kinda gave up on my project when C++ gave me a bunch of retarded errors after I organized all the code into objects. If I had just used structs in the first place, there wouldn't have been any problem.

That doesn't really make sense to me. Classes are just structs that can be inherted, have operators overloaded, and have functions inside of them. The reason to use C over C++ is mostly for reasons you can't see form the code alone. Things like avoiding system calls, avoiding cache-unfriendly methods and structures, avoiding branching, avoiding slow compile times, etc.

>>16918707

>>16918850

I'd honestly recommend looking into Data Oriented Design more if you're seriously considering C. It's pretty much the only reason to use it. Even then hardcore performance wizardry like that only makes sense if you're pushing the hardware you're on to the absolute limits. Usually console developers who really push the graphics hardware like Insomniac and Ubisoft are the ones worrying about this stuff.

An example of why C is better than C++ for this: Dynamic Memory (new/malloc) causes a system call because you're asking for more memory. Your game effectively has to go the operating system, say "I want this memory", the operating system will then queue up your request and get to it when it can. This makes it slower than branching. With C++ the STL has a ton of objects like vector and such which rely on dynamic memory, where as in C you pretty much have to build every data structure yourself. It also makes it easier to make niche specialized data structures that solve the problem at hand well, rather than overly generalized ones that might be overkill.

You can work around it in C, but it's honestly simpler and safer to just cut out the dangerous elements all together.

>>16918885

>You can work around it in C, but it's honestly simpler and safer to just cut out the dangerous elements all together.

I meant you can work around it in C++. I think you can give STL objects custom memory alligators, but I'd think if you're going that approach it'd be better to get rid of the STL all together.

>>16918885

>>16918889

>Even then hardcore performance wizardry like that only makes sense if you're pushing the hardware you're on to the absolute limits.

So much this. You could spam every thread with it and it still wouldn't be said enough. Most of the "slow language" slowness comes from idiot code, not lack of microoptimizations.

>custom memory alligators

And they say C++ doesn't bite.

>>16918908

>Most of the "slow language" slowness comes from idiot code, not lack of micro-optimizations.

Usually micro-optimizations refers to tweaking one line of code or maybe even a small block of code to get the most performance possible. Structuring an entire program to not to use dynamic memory is for sure not a micro-optimization.

>>16918912

I meant it as "optimization with miniscule performance gains, especially compared to the involved effort" here. I would say that redesigning your entire program for preallocation (instead of just moving heavy parts outside of tight loops, which you can do in any language) counts as that. Doubly so if you consider what technical league most /agdg/ games are in.

>>16918934

>fuck this shit I'm out

>but before that let's get the last word in with this text wall

This is top tier faggotry, cease.

We are not developing for consoles or embedded systems here and Dwarf Fortress is neither simple nor typical. The average /agdg/ project isn't bottlenecked by memory like DF is. There's nothing bad about knowing your target specs, if anything you are agreeing with me on that one.

Just use whatever gets you to make a game, but don't cry when it performs like shit and you don't know how to fix it because you used Java++ OOPajeet shitting street paradigms that you learned from (((college))) graduates in stack overflow, but also don't cry if you get stuck eternally rewriting your engine over and over because you're too autistic about every little detail and thinking about the memory of your variables when you should be thinking about making a game.

>>16919252

>language wars started in 2017

Wew.

>Everything you have said so far is so vague and is based on assumptions about me.

I said nothing about you at all you fag. Well I guess I called you a fag just now, but this is highly factual.

>The only way to win with you retards is to walk away.

Which you insistently refuse to do for some reason.

>>16918932

If you're only getting a few groups (And you're using that data multiple times per update) it might just be easier to store it in a global to that script variable and update it at the start of _physics_update() (or fixed_update() if you're using >3.0)

I used to get pissed at having to type get_tree.root a lot so I always stored it in a variable called root at _ready() for instance.

>>16919252

>2017

Newfag.

Any yuropoors doing LD? Are you going to start at 3am as scheduled or wait until the morning? Currently betting on "1 minute to prepare" or "You are the power source" winning the vote.

>>16919393

Yewrofag here, I'm waking up a little before 3 AM because I was going to watch Melody stream anyway. Gonna get a few hours of sleep in a bit.

I hope one of those two wins, and very much hope the one about turning violent games pacifist doesn't.

>>16919412

>waking up 3 am

Jelly, wish I could go for it as well.

>very much hope the one about turning violent games pacifist doesn't

Probably won't bother participating if this one wins honestly.

>>16919428

If it does, I might just turn a nonviolent genre exceedingly violent out of sheer spite. The rules do say the theme is only a suggestion. Hell, you could just pick the theme you wanted to win if you feel like it.

>>16919393

>Are you going to start at 3am as scheduled or wait until the morning?

won't be doing LD this time but when I did, I used to wake up at 3, look at the theme selected, brainstorm a bit with whoever I was doing it with and go back to sleep, usually with a fresh idea in my mind or work started by ameribros already.

>>16919938

these look great, remind me of the old days.

how do you create them?

>>16919938

That architecture looks sweet, are you basing it on something in the real world?

>>16919938

looks like something id see in fallout or underrail

>>16920131

Made a video of what I've been working on.

Just learned binary linking, get this…

EVERY SINGLE FUCKING TUTORIAL IS WRONG!!!

All of them instruct you to use the object file's symbol names directly without removing the prepended underlines and it doesn't fucking work.

I can't believe so many people didn't even test their code. Is my installation of gcc fucked up or something?

>>16920352

What are you trying to do exactly?

>>16920360

Embed binary data into a C executable via the linker. I've got it working but all of the tutorials have the external symbol names incorrect.

For instance, if I used ld to link a binary file called "data" I would get an object file with 3 symbols "_binary_data_start, _binary_data_end, _binary_data_size".

To use those in C you'd reference them as "extern const char binary_data_start, binary_data_end, binary_data_size;" deleting the prepended underlines (As all symbols have underlines prepended to them).

For some reason, every tutorial instead references them as "extern const char _binary_data_start, _binary_data_end, _binary_data_size;" and I you get undefined symbol errors as the names are incorrect.

>>16920375

>As all symbols have underlines prepended to them

That doesn't sound correct. Are you sure you aren't accidentally using some weird linker options that cause this behavior?