/agdg/ - Amateur Game Development General

Where's that link that explains what happened to the agdg community after the exodus? About splitting into groups and whatnot?
I can't find it.

>>1776
for (everything) {
thing.x = -thing.x;
}

a.png
[Hide] (473B, 256x256)

b.png
[Hide] (473B, 256x256)

>>1777
But there's a road running along the x axis. If I flip it on the y axis (i.e. do what >>1778 said), then drive the road, it won't flip the left and right side of the road. It will be like going reverse, not mirrored. See my shitty drawing.

flip.png
[Hide] (71.9KB, 1543x1175)

>>1779
????

>>1780
I guess I shouldn't try gamedev then...

>>1779
Unless there's text on the road, I don't see what the problem would be.

distorted_zoom.webm
[Hide] (3MB, 1910x1190, 00:15)

I wonder if this would still work in a real environment. It might fuck up vertex normals, or cause weird triangle inconsistencies because this is a vertex shader.

In real life you can focus on a particular spot and squint or whatever to try to see more clearly, I've always hated how in 3D games you can't do the same and you're hugely limited by the pixel density of your monitor, you have the least detail at the center even though you'd want the most precision there, meanwhile everything at the sides of the screen are large and elongated and just waste all that detail. You can't squint to try and see what that small object in the distance is.

Games like DayZ allow you to "zoom in" to look further, usually by aiming a weapon, but I don't like how it just zooms in your entire vision, that kind of effect always feels too powerful to me, especially when it's a weapon scope with noticeable magnification that just causes your camera to fly forward into the distance. What I'd really want is the ability to focus in on the center of your vision and get more detail only there, and sacrifice detail/precision at the edges of the screen

noise.png
[Hide] (1.6MB, 1920x1200)

dithered.png
[Hide] (527.6KB, 1236x1236)

no_dithering.png
[Hide] (16.3KB, 1160x1272)

I've seen a lot of games where the sky has ugly color banding and shit, especially at night, but I just realized that it's very easy to do random numbers in the fragment shader. Pic 1 has all pixels replaced with random noise, all you have to do is divide this by about 100-200 and add it to the output color, and color banding will almost completely disappear.

It works because RGB values on the GPU are floats so shadows and gradients technically have super high precision, the banding only appears because that precision is rounded down when the pixels are put onto the screen. But if you randomize the brightness of each pixel by 0.0-1.0 or so, the high precision gradients start to create a dithered gradient when they get rounded. Pic 2 is a gray object, I increased the contrast in an image editor so you can see the "dithering". You could do monochrome dithering too but I assume multicolored would be smoother. Pic 3 is what it looks like without dithering, I could see the color banding before I even increased contrast.

shadedmonkes.png
[Hide] (1.7MB, 1914x1184)

I've been wondering why my lights aren't working properly, turns out I never called glEnableVertexAttribArray for vertex normals. Also the normals were exported with flat shading mode so it wouldn't have looked right anyway.

a0b45ee1f8c7853ecf82594f56da722084b47857b831ab47cc0a481d7da91720.jpg
[Hide] (134KB, 1280x720)

Why is art so complicated. I want to make a Rune Factory clone, but there doesn't seem to be a good way to do the art.

I want characters to be able to wear many clothing items (changes with season and location and maybe you can gift them some), but I think it would be stupid if the portrait doesn't show them. I also want half-body portraits because of reasons, Stardew Valley style head-in-a-box isn't enough. Problem is that if I want both of those things, it would be extremely laborous to edit all the character portraits for all poses of all characters for all clothes.

Option 1: All characters have the same poses like pic related, so you can draw clothing once and dynamically paste it on top of the character sprites. This is interesting because it allows you to add new characters without much cost since you don't need to adjust the clothing. It seems very boring though, and you'd have to make exceptions anyway for different body types (you can't put the same t-shirt sprite on a guy and a big boob lady).

Option 2: Each character has their own unique pose, and any piece of clothing has a separate sprite per character. This allows maximum compatibility with different body types and allows characters to be a bit more expressive. If I keep the amount of clothing small and only key characters can swap clothing, then it maybe wouldn't be too bad.

Option 3: Each character has completely unique sprites. This would make a much much better game than the other two, but the am

1668279773470546.png
[Hide] (3.8MB, 4800x2715)

>>1774
It gets even worse. One of the most disappointing parts of these games is the children, but I want that to be the most fun part of the game.

- I want children to look similar to the wife (maybe the eye/hair colors can change based on player character).
- Limiting your children to 2 (boy and girl) would be pretty lame, if you really like your wife then it feels like not enough.
- It would be strange if all of your children looked like clones of each other, so they should be at least slightly different.

So basically, on top of unique sprites for each NPC, you would also need unique sprites for 4 children for each wife. Maybe you could skirt around the issue by having twins, then it's fine if they kinda look like clones. If the game only had 1 marriageable girl, this still seems doable, but the game really should have at least 3 options to choose from. In any case, NPCs definitely cannot have swappable clothing because it just requires too much art, should be fine to have a couple variations like seasonal or work/casual clothes though.

BUT, that's assuming 4 children is enough in the first place. I can imagine wanting a really big family, but at some point it just becomes completely infeasible to have none of the children look like clones of each other. Maybe you could do it with 3D somehow by slightly altering the face/body proportions, and it would be easier to make universal poses and hair styles

>>1754
AI usage conditions you to stop thinking.

I don't think godot is a good engine at all.

>>1755
I try first then use the AI, it's just an easier google search

>>1758
That says more about the current state of search engines than anything else.

I finally got some time to work on stuff but I'm so drained.

>>1733
Yeah, it's received grants from big companies like Epic and Meta/Facebook. Redot is unlikely to get that kind of investment, especially because of PR risks.

>>1736
As long as a fork actually comes out and it's forked from godot 4 or later, I'm content just staying on one version of Redot forever if I have to, just don't want to be associated with poz

>>1736
Then the best case scenario for Redot is being downstream forever.

>redot's website is now dead

oh_no_anyway.png
[Hide] (1.4MB, 1536x1571)

>>1753
it was a kneejerk fork with zero thought put into how it could possibly differentiate itself from godot. every new feature would be ported from the main godot branch anyways, created by the same people they despise so much
so like most forks where the original project is still active, a complete waste of time

1424132085914.jpg
[Hide] (42.9KB, 353x353)

>>279 (OP) 
Is there an archive of threads, past "fan" games/content, and/or an archive of the 2015 stream(s)? I've only been able to find a handful of those. Thanks.

It's dead, isn't it? 
A shame. Even cuckchan could finish their SJW parody game.

1418378294572.jpg
[Hide] (24.9KB, 280x300)

>>279 (OP) 
None of the Undertale AUs /v/ermin and quasi-normalshits have made have been even close to the original in quality, so I doubt this will really make Toby Phoenix eat hir heart out, but they're still fun to play I guess.

bump

16a35496bb3f9c8f855cd0befd83384b19a057ab20415f983a08e0ff6f936451.mp4
[Hide] (554.6KB, 800x450, 00:06)

>mezzo is a trap
i thought you were going to REMOVE the gay from the game

9f3812d6ffa854431ccb3e4256266dec8a61a4caaa39c4784f5b4340dac73cc6.png
[Hide] (20.4KB, 512x512)

Consider this your ANCHOR POST.
Link your submissions to this post.

sleepy/v/ thread is up: https://zzzchan.xyz/v/thread/253345.html
Make a duplicate post over there with your submission.

1293033708075.png
[Hide] (40.4KB, 313x197)

I was not prepared!

>>1631
Take your time. There is no rush.

maptest.png
[Hide] (342.8KB, 1160x1236)

I want to make some assets to put into my game, but I don't know what to make so I decided to try to design the home map first, and I got demotivated super fast. It's harder than I thought.

Making a map by itself isn't hard, but it's hard to figure out what I want the player to do. I have a vague idea of the player finding items for some upgrades/unlocks (like more storage containers or a better crafting table or something), but I'm not sure what. And when I started drawing a house, I feel like there's way too many containers, I want the player to feel very restricted by inventory space at first and then build/unlock more containers. It's as if I need to make the game first and then design the map, but then I'm running into a chicken and egg problem: I don't know what content to put into the game because I don't have a map design, but I don't know what kind of map to make because I don't have content.

Maybe I can just design a "final" map, and then remove most of it? Like I can put a fridge into my house design, but in-game it's not there and you'd have to first unlock it. Or an entire room is blocked/broken and you need to build it. Or you start out with a bathroom (which feels pointless, but also weird if there isn't one), but can upgrade it into an chemistry lab later.

There's also the problem that I kinda want the player to be able to build as many storage containers as they feel like gathering resources for (I don't think having a hard storage cap is very fun in

ClipboardImage.png
[Hide] (72.6KB, 408x206)

>>1587
Are you trying to invent cataclysm:dark days ahead?

>backpacks
Why not just make everything into a storage? All existing items have to be stored somewhere, and a lot of items are both storages and "items". So simply register every item, by ID/pointer into its storage space. On loading, recursively load all used storages, starting from "the world". 
Its actually important part of the interface. Depending on number of items/buttons, game might start lagging, if every button pressed checks every other button location against it. You will need to separate them into containers anyway. 
And with everything being a container of their own, you can have "hidden" part, where item stores materials its made of.  

>how to ID
You either use half assed system, or make a proper all encompassing item ID system. 
> I want to refer to the inventory by it's name
You dont mean, you literally search all existing items via std::string or something like that?

items.png
[Hide] (34KB, 752x483)

>>1587
>>1588
>create_item()
>create_inventory() 
I just realized that the way I explained it doesn't make any sense because I left out context. Those functions in reality are create_iteminfo() and create_inventoryinfo().

The way my stuff is organized is that I have objects and infos. The info represents the object's type and has all the static information about it, like name and description and sprite, you only modify those at game startup. The object has per-object information, like how many items are in an item stack, they are actual objects that exist in the world.

Pic related are the actual, un-edited data structures for my items. What I'm talking about is creating an Iteminfo, and connecting the relevant Inventoryinfoid into it. I don't need or want to store the inventory info's string ID into the item info, I just want the numerical ID, but the item info may be created before the inventory info so the numerical ID wouldn't exist yet.

>Are you trying to invent cataclysm:dark days ahead?
Not sure what makes you think that.

Test.

c0f9b45354cd49fbcb8f0395a73230637a76596de76df37bb70041c11c947e7c.png
[Hide] (60KB, 1200x1000)

New bread.
>>1605
>>1605
>>1605

>>1194
It just I did everything alright the first time, but it didnt work, so I have to redo it over and over, until I learned that I simply used radians instead of degrees. Not the first time angles fuck me up, its like every fucker invented his own measurement system for them. Yeah, I should have known that this thing uses fucking quaternions, and not degrees, and I am totally heard of them before.  
I spend few hours trying to figure out why textures dont work, and it turns out directX cannot keep textures in memory, if you resize the window. Do you know how hard it is to fix a problem, if you did everything right, but something slightly related has some weird quirk? 
But the good news is, its basically done, and I needed that stuff anyway.

>>1194
>Unrust on your math skills a bit and you'll be capable of pretty awesome complexity.
That reminds me, does anyone have some relevant math resources they'd recommend? I ended up using a function for circle-line distance in >>1160 that I don't understand and I figure developing a stronger foundation will be important.
>>1195
>Do you know how hard it is to fix a problem, if you did everything right, but something slightly related has some weird quirk? 
The last missing piece of the puzzle always seems to take the longest to find! Looking forward to playing your game anon.

ClipboardImage.png
[Hide] (30.7KB, 683x610)

ClipboardImage.png
[Hide] (5.1KB, 378x206)

ClipboardImage.png
[Hide] (8.5KB, 303x238)

>>1196
>maths distance
Its always the basic geometry and right triangles. 

//lazy square distance
if (abs((int) _a1.posX - (int)_a2.posX) <= 2 ) {
	if (abs((int) _a1.posY - (int) _a2.posY) <= 2) {

You basically check x1-x2 first (and mod it to get positive values), if its below threshold("radius") it collides. And than do it for Y too, because its faster to separate them. 

//advanced distance
/*	if (sqrt(pow(TargetX - posX, 2) + pow(TargetY - posY, 2) * 1.0) <= 5) {
It just works.

ClipboardImage.png
[Hide] (27.5KB, 673x601)

>>1197
In case first image is 100% transparent for you too.

>>1197
>>1198
Ah, thanks for the explanation.

>no point in 3D 
Nigger its asteroids. There's no player movement anyway.

>>1112
You haven't played asteroids.

>>1080
Surprised noone responded to this post - this is actually pretty impressive. Care to share the code?

Also, are you the guy who did Super Don't Give Up!! in the last jam? You got the same pixel art/pixel font thing going on. Would be to cool to hear you were still around.

>>1187
Yep, I did SDGU!! last jam. I'll try and find time to post the code with bit of an explanation.

>>1189
Nice. Glad to hear you weren't lost in the move.