/agdg/ + /vm/ - Amateur Game Development and Modding

I'm working on my deformable physics objects right now - trying to set per-poly collision for them but Unreal is being a bitch. – I've heard they prevent ComplexAsSimple from working with physics bodies because it's too expensive to calculate, but what the fuck? Can you not just put a warning there?

Any idea how to circumvent this?

Reminds me of an anecdote about Satoru Iwata from the old days. Shigesatu Itoi's little bundle of joy known as Mother 2 wasn't going well. The game was buggy as hell, and under threat of cancellation. So Iwata stepped in and, after looking at it, said "There's two options. Either we fix what you have, which will take two years, or we start fresh, and finish in six months." They took the second option, and six months later, Mother 2 came out.

Point of the story? In a lot of cases, it's much harder for a programmer to fix another programmer's mess than it is to start anew. Unless your code is neat, orderly and efficient, I'd have to tell you to just get someone else to do it from the beginning.

If you don't like programming, you wont be able to do gamedev. That's the most basic requirement for making a game.

Don't post traps in agdg nigger.

...

Ok, not posting it then. It is worse than that, so I'll clean it up before posting it for mockery.

Hmmm. Maybe then I'll leave as much coding un-done as possible. I have no money to pay anyone, and not enough done to garner interest (I'm assuming I'll get a lot of accusations of "ideaguy").

I wanted to approach it the opposite way, instead of being a coder using coder-level art, I wanted to up my art skills and hope to entice a programmer to my project.

this is what I get for saving my memes from Holla Forums

I'm not the guy who posts Felix pictures though

Please I need help I'm gonna slit my fucking wrists otherwise

I hit a breakthrough in my blueprinting and, in a surge of inspiration, blueprinted some basic actions for a 2D fighting game which include:
Walk - 4 or 6
Run - 656
Back Dash - 454
Crouch - 2
Jump - 8
Super Jump - 258
Air Dash - j.454 or j.656
Light Attack - 5A
Light Combo - 5AAA
Range Attack - B
Low Attack - 2A
Anti Air - 2B
Sweep - 2A+B
Guard/Crouch Guard/Air Guard - 4 while standing, crouching or airborne
Burst - A+B+C (can be used in any state)
Hitstun is also working and can interrupt any action except guarding (will put you in blockstun) and burst.

Best part is that I can implement new actions or button combinations with little difficulty due the way I set up the blueprint. It's a good programming practice regime.

…should I know what those numbers mean? I don't play fighting games at all.

Stick inputs for fighting games are notated using numbers on a number pad. 6 is forward, 4 is backward, 2 is down, 8 is up, 5 is neutral position, 656 is double tap forward, 236 is quarter circle forward, etc.

Don't worry about it.

Fascinating. Seems like a bit of a strange notation, but I'm sure there's a good reason for it.

Why not just say 66?

isn't that hold 6?

Who is this chinigger?

...

Hey guys, help me.
What is it called when you start with a naked 2d sprite and in order to have maximum customisation you add layers on top of it (in game) like a hair layer, eye layer, pants layer, shirt layer.

Composition?

Oh no, procedural?

I think composition is what I'm talking about, thanks.

The word you looking for is "compositing".

Can now steal cars from civvies by killing them first

I'm loving your characters, they look like out of a comfy children's game

Haha they're just an asset pack, I'm but a programmer. So credit goes to Synty Studios for that.

It's a japanese notation standard that got adopted by players of KOF and anime games, because there's regularily so many weird motions in these games that you can't really use a shorthand like qcf for each of them. Also saves a lot of space & typing in notation.

Hold 6 is [6].

Hi hello I'm still alive, have an inventory screen.

...

can you do a drive by?

Because the stick falls back to the neutral state even if it's for a very short time

The gameplay reminds me of Sub Rosa but with NPCs and better graphics. Is this going to be playable without VR?

...

is there any reason why my client doesn't load the online scene? (unet/unity)
this was working yesterday just fine, today i'm elsewhere so i'm using a different wifi connection but it shouldn't matter
debugging and all, the client is being connected, the network is active, but for some reason he doesn't load the scene like he used to

So I found a solution for complex/simple collisions. I'm going to use the list of vertices making up the mesh to generate an N number of convex hulls and use simple collision, instead by adding them as components to the procedural mesh.

Also anyone know how to generate convex hulls for 3D meshes eks dee

Overall actually, if you guys know any useful resources concerning computational geometry for fucking retards you'd make me very happy.

How to make an automatic crash log into a text file? So users can send you the file instead of just saying "oh your game crashed, fix it"

I'm probably missing something obvious here but why aren't you using the built in collision creation?

You are using the Procedural Mesh Component if i understand correctly?

forgot to mention, im using C#

Probably something in here that might help:
tutorialspoint.com/csharp/csharp_file_io.htm

someone help me out with this
i've got two textures generated by perlin noise, one is the colored version, other is a heightmap
they both get generated fine
but then i have to send them to the client
initial problem was that the textures were too big to send in one RPC
so i've got a few functions that send them in chunks of 1024 bytes
individually, either texture gets sent fine.
but when i have to send both textures, only one ends up getting received correctly, the second one ends up as a 8x8 question mark texture

[Command] public void CmdSendTexture() { StartCoroutine(sendTex()); } IEnumerator sendTex() { Texture2D t = (Texture2D)PerlinNoiseMap.refer.r.material.mainTexture; yield return StartCoroutine(ChunkSend(t.EncodeToPNG()));//colored yield return StartCoroutine(ChunkSend(PerlinNoiseMap.refer.heightMap.EncodeToPNG(), true));//heightmap } public int counter; bool waitforC = false; IEnumerator ChunkSend(byte[] fullImage, bool heightMap = false) { //prepare byte array waitforC = true;//wait for the RPC to finish RpcPrepareForChunks(fullImage.Length); while (waitforC) { yield return null; } int numberOfChunks = Mathf.CeilToInt(fullImage.Length / 1024); counter = 0; int d = 0; bool stopSending = false; do { byte[] temp = new byte[1024]; for (int i = 0; i < 1024; i++) { counter++; if (counter == fullImage.Length) { stopSending = true; break; } temp[i] = fullImage[d * 1024 + i]; } yield return null; Debug.LogError(counter + "/" + fullImage.Length); int size = counter % 1024; waitforC = true; //send chunk RpcReceiveTexture(temp, d, stopSending, size == 0 ? 1024 : size, heightMap); //wait for it to be received while(waitforC) { yield return null; } d++; } while (!stopSending); } [Command] void CmdReceiveTexCallback() { waitforC = false; } protected byte[] temp; [ClientRpc] void RpcPrepareForChunks(int size) { temp = new byte[size]; CmdReceiveTexCallback(); } [ClientRpc] void RpcReceiveTexture(byte[] bytes, int index, bool finished, int size, bool heightMap) { for (int i = 0; i < size; i++) { temp[1024 * index + i] = bytes[i]; } CmdReceiveTexCallback(); if (finished) { Texture2D tex = new Texture2D(1, 1); tex.LoadImage(temp); tex.Apply(); if (!heightMap) { PerlinNoiseMap.refer.r.material.mainTexture = tex; } else { PerlinNoiseMap.refer.heightMap = tex; } } }
i'm not getting any out of the ordinary errors, so i'm wondering whats up with the question mark texture

oh no, the I/O part i know, i use external files in my game, what i don't know is how to trigger the log write on crash, should i just try/catch the whole thing?

Couldn't you just always produce a log, and print to it the data you want as the program goes along? Just have it remake the same file for each time you start up, so you don't spam the folder with log files.

It's not quite what you were looking for, but it should work in the meanwhile, right?

i already have a normal log, i want one for crashes to print the callstack

Haha found a way, for anyone who wanna do it too

First add this to your main

AppDomain currentDomain = AppDomain.CurrentDomain;currentDomain.UnhandledException += new UnhandledExceptionEventHandler(crashHandler);

And add this method
static void crashHandler(object sender, UnhandledExceptionEventArgs args) { Exception e = (Exception) args.ExceptionObject; File.WriteAllText("user/crash.txt",e.InnerException + "\n ------------------ \n" + e.StackTrace);}

The most useful part is the InnerException, also change the "user/crash.txt" to the uri of your crash log file

Amazing how he already completed more progress than Sub Rosa and it's VR compatible. Also the Sub Rosa dev is a sjcuck.

I'm using Runtime Mesh Component - it's a third party plugin, and it's free - but apparently it's also twice as cheap to compute. My game revolves around the concept of fully destructible physically convincing environments, so I decided that I need to get all the efficiency I can get.

The issue comes when you take into account that PhysX does not support complex collision for physics bodies - meaning that it does not support concave bodies. It can easily handle a shape made up of many convex bodies, but not concave. I won't settle for generating a simple, singular convex hull because what's the fucking point then? You can crack, slice and smash any object but it doesn't behave realistically? What's the point in that? – I can already enable complex collision when the body is at rest - now I just need to find a solution to make a pretend-complex collision when it's moving and we're in business.

The difficult part is that I'm fucking stupid and I can't figure out an algorithm to break the concave hull into multiple convex hulls.

this is what I mean by the way. You might as well ignore the first picture because the second one is already what the mesh technically is - a concave hull of a point cloud.

Now I just need to be able to get Figure fucking 3 in 3D space given an array of vertices.

github.com/kmammou/v-hacd

you ever had your dick sucked before, user?

Does anyone know a good way to record gameplay in the ue4 editor?
if i try to record the window it gives me nothing or only the small popups.

Just use OBS.

beat me to it. I use Bullet so if anyone is interested here is the relevant Bullet3D info: bulletphysics.org/Bullet/phpBB3/viewtopic.php?t=7159

Okay so now that I know the face of the beast I'm facing I've got to figure out a fast algorithm that can do this at runtime and be efficient enough to do it every time an object is deformed. Thank god it's a fucking videogame and not a shape-recognition or a 3D printing implementation. All I've got to use this for is good collision.

Sub Rosa probably won't ever be completed until he hires more people or cuts out the autist-tier procedural movement and driver and traffic simulations and damage modelling. Two dorks can barely handle an intensive simulations under ASCII characters, so I'm not sure if some nerd in Austin who's only claim to fame is co-developing fucking Gish can finish his own projects.

looks like I'm fucked

What, you afraid of competition?

Why not just send the seeds and let the client generate the same thing?

Yeah. Worked for me.

Does anyone know if it's possible to create a 'gradient blend zone' (not sure what the actual terminology is) where two submeshes meet so that the transition between materials is smooth? In Unity, specifically.

twitch.tv/sgthale

sauce?

Have you looked at Bullet3D softbody code? also there is dynamic fracturing in the standard demo set for Bullet.

youtube.com/watch?v=FIPu9_OGFgc
youtube.com/watch?v=9sh-D0nejUY

What about it?

youtube.com/watch?v=qzpR4QOal4s

For anyone who missed it.

lmao, he did it.

For fuck's sake.

...

This was really well done, especially the ring and the leading by the hand.

The absolute madman.

I mean my waifu is an eldrich abomination too but holy SHIT my man.
Either way I'm fucking proud of this autist. He did the impossible. Truly an inspiration too all anons.

It's a metaphor for the beast within him.

well, congratulations. that was without a doubt the most autistic thing I've ever seen.

Proud

You fucked up the sound. Here is a better one.

Your timing's perfect, I literally JUST sent it to you.

Where's the sound fuckup?
I downloaded the vod straight from Twitch. So any would assumedly be on Twitch's end.

3:12 audio is offset. I made sure to animate every single line she spoke. Here's the raw file from my channel: a.uguu.se/Ssx6TxrOyvWQ_user'sWedding2017

Well, sorry about that. I can't imagine how that would have happened, all I really did is download the vod and then upload it.
Twitch wants to suppress the marriage, I suppose.

How does it feel? When not even the wrath of God can stop your sinful hand, I must ask you, How does it feel?
There was a third earthquake in mexico today, two hundred and three persons are dead now and not once did you stop to ask god if what you were doing was right, not once did you stop to ask God if you should continue, and now you must face god and ask for forgiveness on your immoral soul.

www.reddit.com/r/PewdiepieSubmissions/comments/7232x7/sgt_hale_marries_his_waifu/

upboat my reddit post anonymouses :^)

...

this made my night and sent me to bed happy. Well done.

Of course he is retarded, he sent the earthquakes to the real mexico instead of the virtual one, but now we have evidence of its existence and a clear message that says
GOD HATES WAIFUS

I kept thinking For Whom the Bell Tolls was going to play at any second.

Congratulations, user.

It wasn't user's waifu that got Sodomad and Gomorrahd, now wasn't it?
Sort yourself out, bucko.

Even more dev drama:
When does it end?

until demo day.

Never. People will never stop being dumbasses, and dumbasses will never stop letting these people get in charge of things to run them into the ground.

Is that meant to be a negative?

Something i've been working on this for literally less than 24 hours. It's not exactly impressive or even original but it should teach me how i need to do certain things.

Fuck grammar!

i already am and at first it worked, but after maximizing the viewport once it doesnt work anymore and i have to capture the whole screen.

well at least he actually made an interactive experience

This is what i'm doing instead of working on my game: software rendered flat triangles! I can also do lines using Bresenham's algorithm, I used a macro to generate the code for all of the cases.

I am behind on math and I need to catch up, that's why I can't get work done, because it's studying time.

i got the perlin noise map working
green terrain is normal speed, everything else slows you down
although i should probably also add some damage bonus if you have the highground

Streaman is not pixelman

Didn't bounce back to me with an error, so something's eating it along the way.
I'm just about to head to bed but re-sent, I'll check if it arrived tomorrow.

...

It worked now, thanks user.

Sad I missed it but holy shit. Truly nothing can stop a man dedicated enough.

Hang in there.

Do 16x16 and get a bigger monitor.

Is that what I think it is?
Is that motherfucking Kriegspiel?

Time to die.

uh it's more like this

(A certain form of) Kriegspiel is exactly like that.

oh, well in that case it's just like it but a lot worse

excellent use of homework time! extra credits: Maybe take a look at the assembly files generated by your compiler. What happened to your if-then-else statements? Does it change under different optimization flags to the compiler? What happens if you write your macro as a C function? Does it change the assembly? how? Does it affect the execution time?

alright, damage is now also affected by the height difference
minor side effect though, your velocity depends on the terrain, so the lower ground ends up being able to push the higher ground (mountainous terrain is .6 speed, mountain top is .1). not sure if i should keep it like this
i'd also really appreciate if someone can test if the multiplayer actually works. the one time i've tried this on a computer other than mine, the client wasn't able to load into the scene for god knows what reason
drive.google.com/open?id=0B1laio2k4HufX3JpUU1QRWdOaFk
just run it twice, one host and one client. units spawn when both players are in

Moving up takes more effort and is thus slower, moving down is faster. There is no difference in terrain by going up 1 cm, terrain and its speed differences should be calculated with some biome generation instead.
See pic

u wot
are you extracting the entire thing or just running the exe? should be extracting
reuploaded it
drive.google.com/open?id=0B1laio2k4HufX3JpUU1QRWdOaFk

That what I got when I tried to either open or extract the rar, and it's still happening with your re-upload.

huh weird
try downloading it as a folder
drive.google.com/open?id=0B1laio2k4HufNmo0aVdJWWhhZnc
it will still zip it, but at least it's not me doing it
if that doesn't work, there might be something wrong with your archival thing

It's more likely there's something wrong with what you used.
Anyhow, that worked and your multiplayer worked as well.

i'm using winrar, and i'm able to extract it just fine
go figure

Got it again. Unity once again just not doing valid language features right. For fuck's sake. No exceptions, anything. Unity gets to the end of the first if/else block and quits the method. How can people work like this?
Everything is fine; no nulls, etc, and it executes OK in the immediate window*.

What the fuck is a man supposed to do.

*that is to say, when this even works. "Failed to find a match for ___" is extremely common also.

...

What's the value of x and y?

21 and -1, so there isn't a real tile in the main array (negative index) and we want to pad the generated one with a dummy; both checks return false correctly when I step through them.

Attach the debugger and step through?

That's precisely what I'm doing. I step through, both boundary checks false, it goes into the wrong block (if instead of else) and then jumps out of the method. It's just…

Does Mono treat && differently or something? Is there some fundamental division between bootleg .NET 3 (Mono) and C#6 that is causing grief?

"Expand" the code then.
...bool test1 = Tiles.IsInBounds2D(x - 1, y - 1);bool test2 = TileExists(x - 1, y - 1);bool result = test1 && test 2;Debug.LogFormat("{0} & {1} = {2}", test1, test2, result);if(result){ ...}...
See if it still happens.

Well, now it goes to the else block but still dumps out of the method prematurely. It's an improvement.

Does anyone else experience this?

im pretty sure 7zip doesn't support .rar
at least last time i tried it it didn't work and i had to download winrar

...

It doesn't work now, why would I optimize and compact the code this early?
I'm still improving my service methods for the map right now.

what's the point of getting the tile you're on? isn't this suppose to get 8 tiles around you?

It'll be an optional arg in the end.

Here's a scenario:
Tiles are four verticies, and if you set the primary tile's tile height, you need to adjust all the neighboring tiles' corner verticies accordingly to match to ensure there's no gaps. So, we loop through the nine tiles and set the vertex Y values.

In that case, you may as well get all nine.

If you had written it as a loop it's possible the compiler would have unrolled it and generated the same thing, but you wasted the time typing it up.

then remove the stupid if else crap for the tile you're standing on and learn to use the debugger or write perfect code

To be fair he could have just copy-pasted the conditionals and changed the tiles accordingly. A loop still would have been faster though.

I made it verbose because it wasn't working.
When code isn't working in a way you can't anticipate, you don't just double down and compact the fuck out of it to make it both sexy and broken. It's all very well to tut disapprovingly from the comfort of your armchair, but it doesn't make you a better programmer.

Its literally my homework. I'm not trying to optimize it because it's on a deadline and it is "fast enough". I cant change my macro into a C function; see the arguments in it like "v1.y++"? They actually don't get evaluated when "calling" the macro- during the macro expansion they get substituted for "chx" and "chy"; thus allowing me make the line drawing iterate up and down the for loop from the macro call itself.

If I were going for the fastest code, I would crack open "Michael Abrash's Graphics Programming Black Book" and start reading up… but this is good enough. My assembly skills aren't good enough to take advantage of SIMD.

That's begging for a typo or oversight to fuck him up though.

Honestly, I don't know shit about SIMD, and I'm using OpenGL.
I've written a pretty feature-complete game engine, uses a shitty 3D math library I wrote myself (so no SIMD optimization there, either), has real-time inverse kinematics and other fancy features, except that the render engine is still a bit on the bare-bones side (has shadow mapping, HDR, Bloom, still needs more work).

My frame-times are less than one millisecond.
This is why pre-optimization is the root of all evil.
On a 7+ year-old PC, my computer is literally waiting 14-15 out of 16ms just waiting to swap the window/display framebuffer.
Why bother with optimization then?
I mean, I'm interested in mastering Vulkan, but clearly my game isn't going to need it until I get a lot more rendering techniques and much more complex geometry modeled.
And even then it might run at a smooth silky framerate without Vulkan.

As independent developers, we typically aren't trying to make the next open-world game (if we are, we're stupid), so half the time you can get by even without optimization features like mip-mapping, LoD geometry, etc…
Make the game function first, performance optimizations should be done late in development or as necessary during development if your game screeches to a halt all the sudden.

On a related subject: You might think you need Unreal Engine or Unity to succeed, since you don't have time as an individual to implement mip-mapping and LoDs and Physically Based Rendering and all that jazz, but you would be wrong.
I've already gone over why the performance optimizations are not necessarily a must-have depending on your game.
Regarding the latest and greatest rendering techniques, the most important thing in graphics/art is not quality, it's consistency.
(Now, to be fair, Physically Based Rendering helps ensure materials look consistent under all lighting conditions rather than just some, and this is why PBR is having such a big boom in the game industry right now. Makes work for the artist a fair amount easier. Next-gen real-time ray/path-tracing renderers will make the artist's job even easier)

Yeah, I wasn't trying him to get to optimize the code. I was trying to motivate towards and understanding of what the compiler is doing and why the macro approach was kind of naive. Use a function passing in pointers and the compiler will inline it out for you, but the code becomes much more concise and avoids issues with the macro preprocessor.

In terms of modern gfx performance using GPU accelerators - it's all about state changes. Knowing to collect and sort on texture and shader binds is a good optimization. Approaching a problem naively means rebuilding your materials and shader subsystem at least once. So no, not optimizing up front is a massive mistake. The key is knowing where to do it and why.

to be honest, I'm making more shovel-headed anime gurls
I'm practicing on using better topology and not just create a bunch of floating/seperated meshes.

Anybody know if there exists a wireless device that uses RDP-like connection to a desktop computer on the same network and has a monitor, touchpad and a keyboard? I've been shitting a lot lately and it's fucking up my dev trance. I wanna dev while I shit.

Anybody know what this sort of device is called?

...

...

You could leave a light-weight cheap thinkpad in your bathroom, and just ssh into your desktop from there.

Yeah I guess this is the point at which a laptop with RDP would work best.

Optimization is fun, that's why I do it. If you are compiling on a modern compiler it will use SIMD when it can. The reason I like it is because I take pride in it; I think that it's more fulfilling to know that you are writing good code.

My optimization target is my Thinkpad T41, that's a laptop from 2003 running windows XP.

I'm not in the pre-optimization trap though, I can write Vulkan code but it's purely for academic reasons because i'm not competent enough to use it to get good performance.
I can't tell how much of this post is directed at me, I feel that you've just gone on talking towards the end of it.

It's not naive; I just explained to you why I used the macro, and you still think that I could have written that as a function… You aren't realizing what i'm doing. I'll post the code so its clearer to you:

pastebin.com/raw/AbPQvKDd

Now, if you're right you will be able to rewrite this in a "non-naive" way. You should know how to set up a framebuffer at 32bpp on whatever OS you happen to use.

Had a dream were I picked up Thinkpad "T480" and it was produced by Nokia. It looked similar to a T420 but it had a PowerPC processor. It also had markings by "Marvin Electronics" on it. Fuck, man that sounds really cool.

The top of a head is literally just a hemisphere what the fuck are you doing user

make them exaggeratedly niggerly

ok. ur money.

Is your work the one on the left?

Yes, the other picture is Nykra, the game the guy wants me to make it look like.

That's gotta be frustrating, but hey, it's like you said: It's his money.
I wanna see more of your art though, I really like your choice of colors, it gives off a cozy underground vibe

I make this garbo game.
It's no sweat off my back, and the style he wants takes a quarter of the time, but I'd prefer to make stuff that I like. I guess that's why I'm the one who is being hired, and not paying people to do art for me.

Oh fuck, it's you. I should have guessed.
and fuck you I don't think your game is garbage, it looks fuckin neat and I want to see it developed it further

Thanks. I'm tryin', I ran out of money so I gotta' do some commissions. I should be good soon.

A cent is a cent, and you can't say it's a half.

I'm kind of interested in Sonic Fan games. There seems to be a large amount of available methods of making a game engine. 1. Romhack/Mod existing sonic game. 2. Use one of the Sonic game engines like Sonic Worlds

Anyways, what's your recommended game engine for 2D in general? Gayme maker? Unity? Make your own?

What about 3D? Unreal? Sauce Engine? Just modding existing 3D games?

What's the best FOSS 3D & 2D engines? I haven't really noticed anything. - Specifically for 3D engines. It looks like Richard M Stallman would be disappointed with the current state with game engines.

Fuck it I'm posting a video anyway. Sub-sector walls are now properly shared between neighbours and have front/back textures. I re-did the projection math and got sector heights working the way I want. So now I can draw the spaces above/below differing heights properly with the top/bottom textures for that wall. Looks funky without ceilings and floors to cover them but I guess that's my next big task, along with more reading up on texture mapping.

I've also been flip-flopping on whether to draw back-to-front or front-to-back with clipped portal views into other sectors like Build does, but it doesn't seem like drawing each column would perform well nowadays. I guess the lack of overdraw might be a benefit but maybe I'm vastly overestimating any impact drawing a few dozen extra surfaces would have.

Fuck me.

Anyways, been reading up a bit on music theory and experimenting more. Wanna get the six "main" songs right. Need to finish writing the last one and adjust the other 5. Analyzing K.K. Slider's music has helped me pick up a few new 'tricks', so to speak. Just hope that I'll have something worth showing off one of these days.

well after going over this thread i'm not sure i want to post it without taking the time to catch up first but shit here i go…

fixed the rotation issue not applying to the object b4 a translate happend. and i got the jump "working" gain after a bug poped up after the fix. will update the wiki once i get textures and attack animations working.

any advice for how to animate the jump? and from the ground up how to do melee attacks? as far as programming goes?

Melee attacks are typically done by simply creating a 3D area in front of the character and pushing a "receive hit" funtion call in whatever it finds. You can also check if the area intersects with the environment to thump against walls for example. You can also do a more complex style in the same vein where the area follows the actual bone transform of the character's armature.

This looks like a comfy game to clone. What's the deal with the black ice in later levels, though? Is it indestructible?

Making progress with the documentation for Hellbreaker's level editor.

Still need to make a video tutorial for making a level.

Your game looks pretty fucking rad too, the fact you're adding a level editor to go with your game is pretty cool.
What's the rest of your arsenal gonna be like? Standard doom fare? I've seen your pistol with infinite ammo and shit

I couldn't resist.

Honestly, i'm not shocked your code isn't working given the needless complexity to it. Maybe it would still not work if you were using a loop, but it would be easier to read and edit.

I have more sage advice for you. Use standard code blocks after if & else statements, even if, you're running only one statement after them. This is less important, as you're not likely to run into problems by skipping this step, but it's less cumbersome to support multiple statements at the start, than shoehorn it in later, when you find out you need more than 2 statements.

Alright, so looking at your code, it looks like your boat load of if, else checks should all work the same way. I wouldn't sake my life on it as any mistake could easily be hidden with all this noise. So assuming it's not this code that's the problem, we should test something. Click on your IDE, or notepad with this code open, and hit ctrl+s. Do a quick test, and figure out if still doesn't work. If not, my guess, is, the error lies in the "Tiles.IsInBounds2d(int x,int y)" or "TileExists(int x, int y)" methods. We can easily test this by putting "print("I2B: " + Tiles.IsInBounds2d(int x,int y) +" \n IE: "+TileExists(int x, int y))" before your if statement, or after you else. Maybe also have a "print("passed")" in side your if statement.

tl:dr Learn to code, and stop blaming your tools

Godot is FOSS 2D+3D engine, though it's not quite as good at 3D as things like Unity and Unreal yet. It's usable for 3D, there's even an user here making 3D stuff with it, but not as good.

Here it is on X Windows, because the person who grades it uses Linux, so I had to port it. I was able to use the knowledge gained from Sigma II to allow me to do most of the work, but the real trouble was making the window actually display the bitmap, the way you do it is :

XImage *img = XCreateImage(fbc->disp, fbc->vis, 24, ZPixmap, 0, fb, W_WIDTH, W_HEIGHT, 32, 0);XPutImage(fbc->disp,fbc->wnd,fbc->gc,img,0,0,0,0,W_WIDTH,W_HEIGHT);XFree(img);

I have made no attempts to optimize this; it took me hours of reading man pages and tutorials to get this far, eventually I found a code example on a forum post that actually worked. I also didn't know how to record my screen on X Windows

I'm going to make convex decomposition my bitch. Watch me make the most efficient algorithm in the history of ever. I'm locating 12 vertices relatively equally set apart, and generating convex clusters using those vertices to produce, hopefully, not a complete fucking mess.

(Not really, this probably isn't going to work. God this is such a stupid idea.)

Is 12 collision meshes even acceptable for a single object?

If it's a terribly large and complex object not really.

Oh, I meant more like is it too much? I can always gauge mesh complexity based on vertex count and create more, or less.

I'm talking about, say, a torus.

If the torus is so small you never have to worry about anything going through the hole in the middle you should just have one collision mesh.
If the torus is somewhat large and you need somewhat accurate collision, then 12 collision meshes may be what you need and in that case, no, it wouldn't be too many. It may infact be too few if you need more accuracy than what 12 meshes can provide.
Though I'm wondering, why aren't you using the built in functionality in UE4 for generating collision meshes?
docs.unrealengine.com/latest/INT/Engine/Content/Types/StaticMeshes/HowTo/SettingCollision/

Should be a value you can tweak per mesh, it depends less on polycount and more how many convex things there are on the mesh.

We've done that you fucking retard. How about you read the thread and get over your autistic need to refine code that isn't being executed correctly - the very fucking reason I decomposed it in the first place.

The very first thing people suggested was to print it out. The code evaluation isn't the problem. Unity is literally jumping out of the method at the first if/else block which is incorrect. I don't care about programmatic accuracy right now, it's literally executing code incorrectly.

I wanted to make progress. But I also wanted to try modeling a gun.

Yeah. I think they're supposed to be stones or something. Doesn't melt, doesn't slide, can't be destroyed. You can still push it into enemies to destroy them, but the block itself will remain, rather than melting like ice.

Yeah the arsenal is standard, tho weapons are pure script so anything can be made with modding.

Jesus christ user, you really are going all out for this aren't you?

Have wanted to try my hand at gamedev for some time now. My first idea is a 2d text adventure. Should i use an SDK like godot or gamemaker, or am i better off learning code and building my own engine?
Think along the lines of Policenauts or SNATCHER, in terms of gameplay.

Yeah, I try to keep all gameplay related things script side. Few parts are C++ side for performance or access.

Can still do better with mod management.
The game is missing auto including scripts so you won't have to modify the main script to add #include's.
Also need mod packages, so adding mods won't pollute the directories, making them very hard to remove. I think Urho3D's packages can be used for mods too, need to test it.

your game looks like shit, you should be ashamed and consider suicide immediately, you gay black jew

I don't think you saw it in motion. Shit's awesome

Sounds like something a gay black jew would say

(((hellbreakerstein)))

Should I make my player character a furry for those autism bucks?

I think there's some resources to make text adventures easily with Ren'py, but don't quote me on that.

Dangit Shodanon, you should know better than to reply to such obvious bait.

To get loads of autismbux on patreon for a shitty furry game you first need to have attention from furfags. Draw some fanart of some popular faggots' OCs, suck their dick a bit to join their group, and then you can start raking in the dosh.
Furshits manage to be clique-ier than game journalists, somehow.

He's just just a buddy of mine being an ass.
Also I need to make a new video, levels been updated and added secrets.

Hey, I needed a distraction. I'm used to them not even replying back.

Why don't you listen to your buddy and add some black gay jew in your game? :^)

What's modding for? :^)

To make sure for every game that exist there is a Moonman mod for it.
Also I think Hellbreakerstein sounds pretty catchy to me, could use it as a mod title and claiming it is inspired by Wolfenstein 3D (^^^: . But for real thou I am busy procrastinating with 3 projects already.

Anyone remember me?

I stopped working on this for a month and came back to it yesterday to at least get the combat finished.

Of course most of this is placeholder and looks like trash, but the important parts are that

To add

Once that's done everything is actually working and events or maps can be worked on.

In case you haven't seen this it's basically just a porn/rpg about a slutty elf traveling the land, lotsa rape, pregnancy, humiliation and shit
Making this in Javascript so there's no downloads

I remember your NPC

Haven't really bothered looking for sprites, probably just gonna steal a bunch from RPG maker and call it a day

Whipped up some sector based triggers to test changing sector heights. Doing the same for wall based triggers shouldn't be a big deal.

I would say you achieved both.

Aint no guns in my shitty gondola game.

Because the functionality UE4 uses is too resource intensive. I'm going to need to generate these at runtime - I'm using this for my deformable objects.

no guns YET

Why don't the demons have massive juicy cocks?

...

Because they're cuntboys?

because youre homo

Fucking disgusting, yuck

Are you kidding? I guess you haven't been around /prog/ or /g/ style cultures for very long but the trap thing has been there for quite a while

finally got this ugly son of a bitch under 1k polys. now onto texturing.
btw, if you use mipmaps or filtering, you're gay.

Why even bother user

Is that supposed to be a derivative of a scopedog? Make sure your textures can at least show some of the detail, that model looks like a fucking mess compared to your design.

Beautiful.

Oh, good, I thought he got lazy with the voices or some shit.

Thighs are too skinny.

Project somewhat ported to GZDoom. Special thanks to Pixelman for helping me set this shit up.

Already updated it on the wiki

Gondola uprising when?

Gondolas do sometimes appear near guns.

have some shit oc.

That's the least likely way a gondola would hold a gun

Made some placeholder textures. This'll make mapping a lot, lot, lot easier.

Alpha release! Please play and tell me how shit it is so i can make it less shit
drive.google.com/open?id=0B0AmP3jVSfMheTJRSGJINTBxRjg
will probably only work on Windows, no need for installation

I forget, is this porn?

no

random download
seems legit

What would be the best file hosting service anyways? I wanna boycott google, but everything else looks worse

Mega is free but since Kim Dotcom doesn't own it anymore, might be iffy. I host a bunch of sharethread files on multiple Mega accounts and haven't had any takedowns though, so YMMV. You could also probably just stick files on catbox.moe if you don't care about file management.

I like mixtape.moe

Oh guys, if the game crashes there should be a file called "crash.txt" in [game folder]/user
please post it in the /agdg/ thread please
>>>/agdg/27922

I was thinking about using Dropbox, but i don't know how to make a folder shareable by link, only to specific accounts

i just use google drive, works pretty well in the long term
i've completely given up on trying to hide my shit from the jewish overlords by now though, so do it at your own risk

I like catbox.moe, but you have to keep track of links. It's only really good for short-term, one-use links.

I for one won't download anything that requires me to connect to google, so you may be able to reach more people with a less jewish wesite.

For some reason, Games made in Game maker like Iji don't seem to work on this machine. It just gives me an "unexpected error occurred when running the game "error, No crash log or anything. Anyone got any ideas as to why?

What were you thinking, cooking user? Do something like this.

would work for now since im only releasing this for you guys, but i want to use a more permanent solution for the official releases

will check it, thanks

i understand the feeling, i will put more links later

im remaking the UI little by little, won't be ugly like that forever
My artist is not available since his house was visited by two hurricanes

Click the share button on the folder, then select "Can View" in the dropdown box, then create the link. Just leave the name/email textbox blank.

holy shit im dumb

I use itch.io for deploying builds, also supports delta updates.

I tried 5 different free file syncing programs before like Google Drive, and they all sucked in some way.

i just made fermented fermented fermented yogurt. and i'm wondering why i cant boil anything. i have stock but can i soup?

Does anybody have any idea what tileset Violated Heroine uses?
I thought it was the default RPG maker tileset but it's not.

if you use recursive recipes like your yogurt you lose quality and deliciousness, use some variation

You need to use a high water/other stuff ratio to make soups, like 50~70% liquid, too much water becomes stock, too little stew, none and you have fried food (bonus from using oil)

I'll do the head later so I can think through it more carefully
right now i'm 'blocking out' the main shapes of the body

whenever I create nignog characters, I just create a generic animu girl and color her brown

wait i can pick how much of x i cook?

right click
double click to consume, feed faeries or destroy

i thought those would be intuitive, should i make a mini tutorial?

BOOOOYSSS

I report with a rather significant success! This right here is the product of my runtime convex decomposition algorithm. It's pretty simple, and sounds pretty stupid but holy shit it actually made a delicious collision bound for a torus.

The reason I can't use the normal algorithm for this is because it takes way too much time and resources to do, and so I've come up with a cheesy ass way to approximate collision by breaking a mesh down into 12 vertex clusters relatively equally spaced out, instead of using the normal volumetric approach.

i don't understand what the fuck you two nerds are saying, but good job

I am one nerd explainin what I did.

PhysX does not allow complex meshes to be used for physics simulations. That's why a mesh is usually broken down into several convex shapes to simplify it.

picrelated.

I'm working on an algorithm to do it faster than most standard algorithms, because fuck voxels.

You lost me at the voxels.

He ended his sentence with a period, which is usually used to end a sentence without emphasis or questions.

I made a video outlining how to do basic IK rigging in Blender, for the wiki. It assumes you already have a weight-painted model bound to an FK armature to start with (I'll try to make a separate video on that later).

Dropped

what if i don't have an nvidia card

i would add the click and drag to feeding faeries and yourself as i was not aware of a nutrition meter as i assumed that ferries ate as needed from your food stocks. also why is one of my faries refusing to work?

When's Demo Day™ again? Nov 11?

Yep.
8agdg.wikidot.com/demo-days

it's on dubs of every season.
11/11, 2/2, 5/5 and 8/8. I think it's on the wiki that I still haven't used. I know someone put my game on there but I never cared to do anything with that.

They eat from the food stock, when you give them food you increase the stock, if the food is tasty also makes them happy

Mouseover the faeries will give a tooltip, they will tell you if they are too tired or unhappy to work

It's more comfortable than 2 buttons since you don't have to move the mouse.

They eat from the stock, if you give them food the stock increases, and they the amenities bar increases or decreases depending on the food quality

Mouse over the faeries and they will tell you if they are tired or sad, let them rest

...

Glad to know that this shitpile of a website isn't broken for just me.

Fucking hell sorry for the multipost, im trying to delete it, but Holla Forums is shitting itself, or maybe its my browser

ah a typo, I'll fix it

If you can run any other game made in UE4 you shouldn't have any problems with mine.

Can someone tell me how the fuck to sort arrays in blueprints in UE4?

I'll stop once I take some time to learn Unreal coding standards.

Done with implementing auto script reloading, and Urho3D packages seems to work.

Time to go to sleep

Fuck me this really does make mapping so much easier. Now I don't have to worry about alignment or sizing or anything of the sort.
Well, I mean I will later, but for now I can focus on the layout of the map first and foremost.

Why didn't I do this earlier for other projects? This would've saved my ass in so many things.

Is it doing something special with them? Like, do you mean you take any given blueprint and take an array from it, and sort the values in the array? That's basic shit

Took a break from learning physics to come up with a better way to modularize my plane setup. I was over-complicating things beyond measure, but I think I found a sensible way to get things to the prototype stage (and so that integrating the physics side of things won't be as much of a headache).

The idea is that I have two dummy components that handle most of the work: one that handles things individually, and one that handles the combined net forces. Everything else (collision boxes, static meshes, etc.) are attached to these. It sounds obvious in hindsight, but my original way involved a ton of arrays and an unnecessary amount of variables and function calls. I should have just done this in C++, I regret it more and more.

It's quite remarkable how limited UE4's blueprints are when it comes to arrays. You're stuck with only 1-dimensional arrays (unless you use vector or struct arrays, but still). You have to find strange work-arounds just to do basic shit, when the whole point of blueprints is rapid prototyping what ought to be basic shit.

...

I guess it'll make my life much easier to just go ahead and learn how C++ works in Unreal.

Just so you know, there's already a game called Helldivers, and it's trademarked.

That's Helldivers, user's game is called Hell Divers :^)

actually a pretty fun game

God I'm fucking tired of physics in games feeling so floaty. I have to set the gravity to twice the normal value in UE4 for it to feel anywhere near real life. Really hampers stuff when you want impacts and falls to feel dangerous.

change your bounciness

I disagree. Game physics should be like a bounce house. In every single game. Name a game that wouldn't be better with bounce house physics. I dare you, user.

Starting work on a special set of weapons now. Thankfully someone had already made a wand pistol hybrid weapon I could use as a springboard for my own edit for it.

How do I get motivated and stay motivated? I want to start coding my game, but everytime I start I get writer's block and end up playing other vidya or watching anime.

First off break down all the shit you need to do into a todo list of smaller tasks. Consider taking notes when you play vidya or watch anime and see if something sparks inspiration or an interesting idea.
It also really helps if the shit you're doing is something you find really cool and interesting, that's how I've been keeping my own momentum.

Is it hard to code? is it hard to come up with ideas? is it hard to design game features? is it hard to translate your ideas into code? What's so hard about it?

Unless you don't actually like making games in which case stop and don't try again.

By writer's block do you mean you don't know how to do something, can't decide what to work on or can't choose how to implement something?

Thanks, I'll have to make myself a list like that then. And the game I want to make is basically everything I liked or found interesting in RTS games, with zombies added.

It's not the coding, because I've got that pat down.

Well, I suppose it's more or less implementation and over-worrying. I'm a bit too much of a perfectionist autist to just make gaem.

As someone who suffers from the same issue, I can tell you right now you need to figure out where to draw a line and get the damn thing done. Adhering too hard to wanting to make things perfect will mean you never get that shit done. Striving for quality is one thing but make sure to not go full retard.

(nice dubs there)
Thanks user, I'll do my best with that.

I had the exact same problem. I'm trying to break the habit by not over-designing stuff and intentionally doing things in a rough prototyping manner and then tidying them up afterwards. It feels faster because I can see visible progress sooner and by the time I get to tidying up I have a much better idea of how things should work.

Having a list is also very helpful but be careful not to overwhelm yourself by listing out every idea or plan. Keep them to short term goals or broad over-arching ideas until you're reading to dig into them.

how do i get a name which acronym is KEK? preferrably one which has "kingdom" as one of it's words

Kingdom's Excalibur: Kinglike

Kingdom Enriched: Kangz

Gamejams help with this problem.

Kingdom: Eternal (K)Night
Your call if you put the k on depending on what kind of game.

RNG is hard

why not just use WELL or a mersenne twister?

Kingdom el Kek

There's already a dive bomber called Helldiver, from which my game gets its name, and I'm pretty sure that's public domain by this point.

That didn't stop Bethesda suing people for the word "Scrolls" just because that word is also in "The Elder Scrolls".

Just subvert the name SLIGHTLY and it constitutes a new name.

I.E: Calling it something like 'Hellbringers' might be enough of a difference.

Like when scientology tried to trademark the Cross?

Nah, if anything I'll just name it after a different dive bomber, since that's what I pulled it from originally.

What are you dividing? Are you dividing by RNG, or is this an output from your own custom PRNG function? In either case, if there is a division of some sort, best to add a small constant number to the denominator to prevent that, like 1/(0.001 + rand() ).

I'm doing procedural terrain generation and >1x1 sprite tiling, and the 0s were related to tile coordinates. I was dividing stuff to get seeds for the pseudo-random number generator.

Could you offset the equation so as to avoid division by zero?

Kal-El's Kingdom
Why does it have to be KEK? That seems almost as bad as naming your game M.E.M.E. or something.

The coordinates go both ways from 0 so it's a bit difficult. I think I found a solution by adding +0.1 inbetween there like you suggested, but my textures are still misaligned so it may not work in the end.

0.1 will probably misalign more than say, 0.0001 would, but you could (if not already) round up/down afterward to overcome that.

many thanks everyone, i think i'll go with a variation of

Super Smash Bros. Melee.
You can see what bounce house physics did to it with Brawl.

WHAT THE FUCK

Added some fog and lightened up the sky. Looks nice now.
Also added a tentative player sprite. Might need to do a little more work to him.

They probably have shit servers. You could probably host your own repository, but that'd take a bit of effort and is probably overkill if you just need to share a single project.

what is bitbucket nerd
although they do limit you to 2gb per repository

Are you fucking gay or something

Getting rid of the white background, how is it? too woody?

better

I don't understand how to upload files to gitlab, I just have no fucking clue what I'm doing

I want to make a game, but before I begin, I want to make something transparent: I want my game to be commercially viable. Looking at pic related makes me salivate
I don't need millions. If I could just sell 10,000 copies at $5 each, I'll have enough to set my retirement/passive income plan in motion. Yes, I'm looking to sell out on my hobby, but at least I want to make a fun game.

I want to make a side-scrolling shooter, similar to Metal Slug, (because I like the games a lot) and sprite graphics. I only know C++, but I don't mind learning new languages if need be. What sort of engine/environment should I make it? Should I just download a free C++ compiler and some DirectX libraries, or is there a simpler way to go about this?

I can't draw, and barely know music, but since I plan to make money on this, I can't ask you guys to contribute. I'll just have to bash my head against those walls until I have a product I won't feel shame for selling.

First you have to learn how to use git.

Git gud.

Gamedev really isn't a good get rich quick scheme unless you're absurdly lucky, unusually talented in many fields including design and perception, or know how to suck a lot of dick and get connections.

I would say that there is no market for platformers or side-scrollers in general unless you can make the next Shovel Knight or Knight

First find your niche

or Hollow Knight* i meant

Yeah that's the problem
I mean I managed to get all my files, and I know what I should do, but then I go

And it tells me my folder is already a git

No idea how to fix it

git-scm.com/book/en/v2/Getting-Started-Git-Basics
git-scm.com/docs/gittutorial

Git hub is only one of the sites that host git repositories, learn about Git itself

I'm not looking to get rich. I'm looking to get into investments a bunch of rental properties as my long-term retirement goal , but I'm just an Uber driver now, and I'm basically unemployable. I have nothing but a computer and some personal affects, so I was hoping to make some starting capital this way. And even if that doesn't pan out, at least I could have some software I could use to make myself employable.

I can believe it. I played a little bit of Shovel Knight, and it was very fun. I don't know if I can make something of its caliber, but I've got to try at least. If I can't make something that can entertain a person for more that two hours before they refund it, then I'll call it quits.

wow how exciting

I finally have enough of a 3d skillset to make the horror game i wanted to make since forever. There i just one design decision i am really unsure about.
Pretty much all horrorgames have a main character that leads the story by speaking to himself and giving the player hints, but i plan to erase that. In fact the story will be optional and the clues about what is going on will only be available if the player searches them for himself if he cares, but other than that it's up to the player how he "escapes" his predicament, you know like when you're in a game and you think "why the fuck is the character not just hopping over the goddamn fence/going through the window" or similar, i want to give a player such options and it won't be completely linear. Technically you could finish the game without ever finding out why shit hit the fan.

I want to make that kind of game, however i also want it to sell at least decently hoping to start my own game company someday. Do you think that's feasible or do players need a forced story thread to follow? I really have no idea how it would be recieved.

Please make this happen.

You can't really predict how it'll be received, but if your story is subtle enough while not being autistically cryptic, you might get the attention of bored theorists which, I assume, do push at least a couple units.
What I'm doing, and I think this is what you're thinking of doing too, is establishing that the MC is in a particular (spooky in your case) predicament, and you need to actually interact with the world to figure out why and what they're there. It's a way of subtle storytelling that I personally prefer to the overdone "Person I know went into spooky place I must find them" scenario.
Basically, do it, faggot.
Worst case, make free DLC that explains the story better or something.

I have started my own puzzle-platformer clone of this

Like a Souls game, I quite like that approach myself, as it doesn't lock the player on long cutscenes that drag on forever if he doesn't care about the story, yet lets the player that do care about the story enjoy it to it's fullest, not a short, neutered version tailored to the ADHD players. I'd say this is a great idea.

This is where it gets complicated unfortunately, if you go that way, the "you can do whatever" way, your game will branch out exponentially, forcing you to either make a minuscule game (in the same time that would take you to make a game with average length), or have a lot of premature end-game situations (like this user said )
, which gets really annoying after getting 2 or 3 and having to start again. Linear is easier and more efficient to make, this is a fact.

Fuck.
I'll have to pirate it once you're done.

It's not a direct clone, but I'm thinking that you play as a witch with 4 elemental wands, and a "get to the exit" style gameplay, after collecting things

This user know what he wants and just goes and fucking gets it. Godspeed you magnificent bastard, you are what we all should strive to be.

Godspeed to you, Hitler.

Feels pretty fucking good, this is way above what I should be able to make

>making Stanley's Parable: Haunted House Edition
Go for it, but just come up with better excuses for why a player can't do certain actions (so as to avoid the exponential work load that talks of). Spend time on the "core" aspect of the story first, THEN add all the other extra non-linear gameplay and ending aspects after. Also, make damn sure you implement a Save/Load feature, so players don't have to start all over again if they accidentally an ending.

How long of a game do you plan it to be, or have you not thought everything out that far yet?

This looks cool as fuck user, now make game

Kek, might be a fun easter egg to be honest.

That's pretty much what i plan to go for. MC is a normal dude who just does activity x, then gets trapped in a predicament accidently. From then on it's all about staying alive and getting the fuck out of there, with enemies and obstacles in between. Finding out why the fuck certain things happened, is entirely up to you.

Yeah, that was the first major problem i realized too, so i decided to choose a location where i can "lock up" the MC as much as logically possible. It's going to be a place like an underground facility or similar where final escape routes are extremely limited. Basically MC ends up at one location, the main exits are somewhere far away, and in between is a more or less open game map filled with dangers and many typical "puzzles". However, finding out where the exits are, what route/puzzles you take and what you explore is up to you. A playthrough should take somewhere in between 3-7 hours, based on how smart the player is with solving shit, finding quick routes and getting an idea bout where the final exit might be.
I haven't chosen a final location yet, but space sation, underwater lab and underground facility are my current top three. Mc discovers/visits it due to his work or hobby, then shit hits the fan and he gets trapped inside and can't exit from where he came from but must find another way out.
Absolutely, i'll go with resident evil style saving with 3 save slots.

Thanks for the input lads. I am starting to get a better idea of how i should go about it.

Could be ye old abandoned mine, tunnel collapses and all that shit. Could either be there out of hobby (spelunking), or job (miner), tunnel collapses and you're left looking for another way out. End up in tunnels that "shouldn't be there" or something, you get the gist.

This is optimal for non-linear games, you actually want an exploration game, alike the old Sierra graphic adventures more than a straight up open-ended game like I thought at first, which is much more possible on reality.

So despite dicking around a lot, I've never tried to implement a state manager. I guess it's going better than I thought it would, but I need to remake it.

I managed to get smooth tile-based movement, and changing animations based on state.

I think my delta timing is desperately fucked. Everything moves faster in negative directions and slower in positive ones. Granted, I know this code is absolute garbage, but still.

millennial pong

...

Implemented the idea I had earlier here . Now I'm trying to approximate the right sort of data, like drag coefficients and shit, based off what little data seems to exist and fudging the rest. One step at a time, I guess.

Ahh shit baby, I have 2 sets of walking animations now. Gotta work on back-right facing walking now, then forward and backward then I can test my textures in game. what whaaaaaaat

is this the Unreal RPGdev project, or something else?

I revisited the wand pistol concept, I think adding a longer barrel helped

I guess this is my interest now

Somehow I missed that obs was recording audio

I was about to comment about how the arm being diagonal and the pistol being straight looks weird but I just noticed that's how you hold most guns

The video was better for it.

This is a new feel

Dialogue automatically changing depending on the age of the protagonist and their pregnancy state, each has fallback to a default message in case a variation for the current state of the player is in doesn't have a special message

where do I place my preorder?

Planning to make it free to play and put a few ads on it.

Also options are

Each has different base stats, and is affected differently by pregnancy.

Can you cast magical spells to turn yourself into a loli when you are and adult?
good luck with that

I mean I guess I could but it seems fairly pointless when you can just choose to play as a loli

There's no actual depictions of lolis, I doubt that'd be a problem. Not to mention I'm planning to just mention the character as a "young flat chested girl" so it shouldn't even be specific.

Those animations are for the RPGdev Unreal project, yeah.

I have a specific question, not sure if anyone here can help me out, but here goes:
I want to make a game. I've not chosen a game engine, I don't have any of the specifics down at all. In fact, I'm just starting out. I've got a book on C++ programming, and I'm gonna read this 700 page monolith in order to better acquaint myself with the language before I get into the specifics of building a game with Unity or whatever the fuck. I figure that I may as well understand the general application of object oriented programming and problem solving that will go into making a game before I start fucking around with an engine.

My question is this; I've downloaded the NetBeans 8.2 IDE, and I need a compiler. Can anyone tell me which compiler is better between Cygwin and Mingw? I tried to install Mingw, but all of its necessary packages are hosted on Sourceforge, and the installer can't access the files it needs to download because Sourceforge is currently having some kind of problem with its servers or something. I don't want to wait to start doing exercises out of the book, so I thought I'd go with Cygwin instead, but I find myself not knowing anything about either of these two components. Should I wait until Sourceforge makes Mingw's files available again, or should I download and install Cygwin? Hell, do I even need to install these things, or is there some sort of built in compiler that comes with Windows that I can use?

You are looking for Visual Studio. Don't bother with cygwin or mingw on windows unless you really want your game to be on linux too – and if you're just learning to program, you don't need that additional complication.

Should I just ditch NetBeans then? Also, my book came with a copy of Visual Studio 2010 Premium Edition. Should I use that or the 2017 Community version instead?

If you're writing c++ on windows absolutely do not use netbeans. I don't know a lot about what changes have been made to visual studio, but I'd say try the one that came with your book. Community edition is free, you can just try it if you find out it has some feature you need.

Good to see this project is still alive. I was worried.

Took a break from programming now that I have level persistence. Begun to convert the BSPs from Unreal to normal meshes in Blender, because BSPs are less efficient and flexible.
Just have a few more parts to unwrap, but I'll do that tomorrow.

I wondered about why BSPs should be less efficient and had a theory which turned out to be correct. They don't scale very well in terms of how much detail you can add. It doesn't take a lot for it to produce quite a few unnecessary triangles and a godawful topology. I reduced the polycount from 4129 to 834 in pic related.

Changed the buttons too, how is it?

Colored text on colored background is a big design no-no
Especially when you put yellow on brown

should i darken all text?

Try it out

ok i will edit it later

Yesdev here I come.

I blame ffmpeg. Are there any good screen recording tools that work on a toaster with opengl2? I'd use obs but my hardware doesn't support it.

I would say darken the background and lighten the text.

Throw on some electronic music and you've got a goty

It got stuck in my head, now post the music

Oh no, it is never dead. I've just had trouble securing money lately to keep food in my mouth so I had to figure some shit out. Also I think I have some sort of cyst (nothing serious though) in my hand so it really depends on how my hand is feeling.

Nigga, working minimum wage sucks but I'm stuck until all the marijuana is out of my system, used to smoke a lot for creativity for coding. It worked but now I'm past the phase where it is remotely necessary.

JUST

I've tried doing that a few times, but it really doesn't work for me. It makes it hard to concentrate, and even if I have cool ideas it's difficult to actually implement them under the influence.

You definitely have to do a lot to get in the right mindset. I had to eat right, take vitamins, exercise properly, get some non disruptive music, and HEAVILY MODERATE smoking. Too much and I'll jerk off or go to sleep. On top of that, never study high, instead take work you've already put some hours into based off knowledge you already have. A lot goes into it, or rather did. I came up with the player movement relative to the camera's position system high, it was fun.

Bread soon

...

FRESH