Amateur Gamedev General ~ /agdg/ + /vm/

Don't give up user, go make other parts of the game and when you have more experience and programming knowledge you can come back and solve your problems with the controller. It might be impossible to make the exact game you dream of, but you can damn well try to get as close to it as you can.
Also, if you are having problems you could tell us what kind of controller you are trying to make and perhaps someone has some insight that can help.

I have learned to always search the internet for what I want to do before actually trying to see if someone has found a simpler way of doing it.

aaaaaaaaaaa

where do i put the centre on this blue sprite so it rotates properly with the ship? i tried lots of ways and i cant get it right

theyre both 16x16

kind of like my earlier problem but that was just drawing a sprite, this time its an object

Where the ship's center is, or make the blue part a child of the white part.

To rotate it so it lines up? You need to rotate its position around the center of the ship.

Trigonometry is your friend. Look up how you can make the blue bit always be the same distance away at whatever angle the ship or whatever is pointing.
Once you get a sine and cosine involved in the position, you should be able to make sense of it just by fucking with the numbers. Having an intuitive grasp of what they do will be useful for you later, so I'd encourage you to look up only as little as you need to kind of get the ball rolling.

...

And you've been blogposting this same shit for weeks now. Are you just after a pity party?

You don't even know the scope of what you do not know now. Stop caring or trying to figure out why you need to know these things. You are too dumb and uneducated to make these distinctions at this time. As you gain skills you will understand more of the "why" answers, not just the how.

Power through it. Or be a pussy and cry about it. No one really cares.

So did I
Couldn't find anything

Low-health effect is in. Likewise, now the screen flashes green when you use healing items (except for the burger, it flashes gold).

I really hate the BLOODY SCREEN SO REAL effect but I've found that for players it's unbelievably easy to lose track of how much health they've got. So instead it's a faint red pulsing.
Hopefully it's not too obnoxious, if so I can probably make it fainter.

Why do people even bother with 3d?
You spend so much time trying to be a one man army on the programming your shit is just bland compared to any AAA game by the time you have to give it those personal touches which set it apart. If you even get that far.

Is there something I am missing here?
It just seems like a terrible idea, no offense.
Not quite MMO-tier bad, but still pretty damn bad.

Literally what?

I've just spend a week and half to go from fucking aritmetic to last weeks of precalculus.

high school math is easy as fuck if you're an adult, literally as easy as basic adition and substraction.

I've never cared about math nor I still think is pretty much useless to learn all that, but I think I would be much worse trying to make a 2D game without understanding of at least trigonometry.

It's like complaining about hardware tools and going to the hardware store when you're trying to make furniture.

Because 2D is just not fun for some people.
Depends on how you structure your game. Most people are literally just recreating another game that's already been made. Like Tetra-dev was basically making devil may cry waifu edition. You might be able to add more content to a 2d game, but if it's just another rehash it's no different then a shitty 3d game that's spread out too thin.

Pic related is a indie game made by 2 guys who built their engine to support multiple cameras at once, so they could offer a unique squad/stealth game.

Nigger, Panda3D was made by Disney and used for shit like Disneyland rides and Toontown. It's been around for a long time before Signal Ops.

Oh yeah, ogre3d is one of my favourite game engines. Up there with directx and opengl.

Speed, acceleration and velocity
Fuck if I know, kind of a broad and simple concept the could be useful so you can frame your problems in a mathematical way.
Any sort of "smoothing" uses a quadratic function or more, I myself was using it as a way of giving some acceleration to the run speed of my character
Bread and butter or any sort of rotation your character needs to make. Pay attention to this one.
Haven't come upon this yet, but perhaps they could be used to calculate things like an orbit around a planet.

Most of the stuff you learn in math isn't immediately useful, they are tools for when you have a problem to solve you will remember some obscure thing about math that made what seemed a first look like an herculean effort is now a mere equation you need to implement. If you are studying math I would also advise you to study binary logic.

Damn, I guess this is one of the reasons we should give a light read to the documentation then

user please, stop being dumb. It's all about balancing complexity, content and time to make it. 3D only has a higher minimum complexity necessary to make it work. You see a lot of indie 2d shit because it requires the minimum amount of effort necessary to make a shitty game, but you lose a lot of possible depth to the game, a 2d devil may cry clone can work, but it will inherently be inferior to a 3d one because of the lack of third dimension where you can have enemies coming from. Other concepts only really work in 2d, like bullet-hell.

Damn, you must be smart. Do you watch Rick and Morty?

do you really think math that is taugh to preteens and young teenagers is hard?

Reminder that no matter how bad your game is, it will never be a furfag abomination.

Hopefully that isn't a 3d version of that furfag RPG where you could rape everything. Also the game I posted uses Ogre which is a graphics renderer not a game engine

Signal Ops doesn't use Ogre, look it up.

Well shit. I remembered when I looked into the devs ages ago they mentioned building in ogre, but I must have miss-remembered.

I'm a dumbass who can't handle numbers to save my life. I slam straight into a wall anytime my code needs math more complex than addition/subtraction/division/multiplication.

download the teaching company DvD about algebra I, algebra II, trigonometry and calculus I.

It becames easier if you start from the basics and move up.

Thanks, this worked!

Guess I'll work on the targeting system and item drops next.

You're right, gotta keep that perfectionism in in check. As I said, it just works and it does so without any noticeable bugs, so I have no real reason to complain.

Is it really the case for a single-player shooter? I never had trouble with keeping the track of health aside from few rare moments when action got too fast.

My advice would be to make the red glow work similarly to the green healing glow - it should be a brief effect triggered by injuries suffered below certain health threshold. Your game doesn't have auto-regeneration, so If a player would be out of healing items he'd have no way to remove that effect until he finds something and it would become very annoying, very quickly. Plus I don't think people need a constant reminder that they're almost dead, just as they don't need a continuous green glow to know they're at good health.

bretty good.

I though it was a PS2 game for a second.

Looking good, but

My game can now send information from a client and receive it as a server. The concept of multiple clients isn't implemented yet, though. This is all done over UDP. The next step is to use this system to implement player movement.

I know it depends on use and environment, but do you know of any good introductory texts to networking?

Stuff that conceptualizes it, explains it, different protocols and layouts, etc. In particular, I'm in C#, but anything is useful to know, I guess. Less code that I have to write to learn about it right now, the better

All I have been doing is reading through a couple of code examples that seem to be somewhat self-explanatory. This was the most useful example: binarytides.com/udp-socket-programming-in-winsock/

I'm using the WSAPI which, the way C# works is that everything in it is just a direct abstraction over win32, so you'll be using the same API, just in some kind of object model or something. Windows Sockets seem to have a lot in common with UNIX sockets, a lot of the functions look exactly the same.

TCP and UDP, since this isn't so low level, you can just use the API's to set up a TCP or UDP connection, and then that should be fine. You just "start" two sockets, and then the API's make it pretty easy to send data over it: It's just really simple functions like:
int sendto( _In_ SOCKET s, _In_ const char *buf, _In_ int len, _In_ int flags, _In_ const struct sockaddr *to, _In_ int tolen);
which should explain themselves, I want to send a data through a socket to an address. It's pretty conceptually similar to the idea of using a file descriptor to send and receive data with a file. I haven't actually looked at any sort of conceptual texts about networking, I have just been reading a few code samples.

i have no idea what i'm doing

I feel like your song is a diploma.jpg of music composing

i genuinely have no idea what i'm doing

I see you're making the soundtrack to one of those existentialist and surreal Thailand/Japanese movies.

should probably end it while i'm ahead

Fuck me goddamit I just went full degeneracy modeling a loli vagina in my quest to study anatomy. Looks pretty good for my first try but goddamit I feel so goddam dirty.

Slap fur and a tail on that and you'll be making patreon bank in no time.

I laughed

That ain't happening. I want a comfy waifu VR sim first. I just accidentally let my dick take over the wheel for a couple days.

I didn't make any vagina bones. What are you laughing about?

Working on starting a new project; hopefully something will come out of this one. Right now I'm working on all the simple debug functions and shit I'll need, as well as a coherent codebase this time. I'm using Godot, and the last projects were always a fucking mess because I didn't know how to take advantage of the node tree system. Currently I have a main menu with a file dialog to test various scenes, and a debug function that can be turned on and off.

This isn't really progress worth posting about, but I want to get into the habit of doing so. Hopefully it will keep me motivated.

Maybe you should keep going until music comes out.
Does OpenMPT allow you to play notes just by wailing on the keyboard, perhaps along to whatever you've already tracked? It can get pretty tricky to feel out what you're doing if it doesn't.

Redpill me on writing my own raycaster.

...

how do you get pseudo 3D with unity?

pfft

...

...

...

*97.808219178%

I guess I'll remake this game on unity.

doing so wastes your effect slot, since unless you're right on the note, you add a delay effect.
on top of that, it'll only place notes on the current channel, which means if you play a chord, only one note is placed.
all of this effectively makes it useless

Real men stick with the text

That wasn't really what I had in mind. I was referring to more of the ability to just fuck around and play whatever instrument is selected in polyphony while not recording. To sample and try out ideas. In something like, say, MilkyTracker, you can get however much polyphony you want to just sort of jam around, with or without previously material playing. If you've got jam channels enabled and high enough, whatever delays, echoes and effect you have going on won't interrupt each other, as each note's will be playing on its own channel.
You can record onto multiple tracks as well to capture polyphony, but which of the selected channels it's going to write to can get a bit messy.. It'll cycle through them in order for each note pressed. Works well enough, but you'll probably want to rearrange them later if you want, say, the higher voice in its own track and not strewn about.

Maybe give some other software a try. MilkyTracker's pretty good once you get used to it but the UI's certainly a product of an older generation of trackers. Tick some boxes in the I/O part of the config relating to multichannel and instrument playback things, take a look at some hotkeys and you're off to the races.

oh, you can do that in the instruments or samples tab.

...

Fucking Unity. I was using animation preview on my player gameobject, but then during play when my script called gameobject.find to get a gameobject that was a child of the player during awake() it was getting the clone created during the animation preview, and then it was throwing a null reference exception because this clone didn't have the script I wanted.

The clone itself has a different name from the original, so you can't easily get the wrong object, but the children of the cloned object are also cloned but have the same name as the original. To make it worse this clone is never displayed on the hierarchy tab, but if you get the name of the parent of the child object you wanna find it will say something like "Parent_object_name_here (Clone) AnimationPreview". Made me waste a fucking half-hour thinking I was going mad.

so change the names or don't use gameobject.find, you dummy

Making even a turn based game with proper save/load is such an eye opener.
My entire gaming life is being replayed and I'm like oh that's why I despawn enemies in every modern fps by going through a checkpoint and reloading.
But also being really impressed by all the devs who didn't. Every game that allows me to save during action now gets a 20% boost in my mind.

Hm. Well, I guess it's just a matter of playing around and writing down what seems to work. If you're not using an external keyboard, I do hope you're using a key binding that has the regular tracker format. You might lose a bit of range with just two and a half octaves, but having nothing but chromatic rows can make it very tricky to keep tabs on what key you're trying to play in. Having some duplicate notes in ,./ and QWE gives you some breathing room for when you need to cross that gap between octaves. Keeps the octaves lined up as well, as they're lined up right on top of each other.

You didn't understand, the animation tab on Unity has a preview option, it creates a clone of the object you are using to preview your animations on and adds the "object_name_here (Clone) AnimationPreview", changing the name does nothing, the object also never appears on the hierarchy tree so it's not like you can manually delete it, or even see it. It also clones the children of that object, but does not change the name of them. For some reason the preview was not clearing these objects fast enough so that when I pressed play and the awake() condition started these objects where still there and one of my scripts was getting it instead of the real one. This is completely on Unity, these clones should never appear during runtime.

Surprisingly, yes.
I've seen a lot of people on Twitch play earlier builds of my earlier projects, and when there was no obvious low-on-health indicator there were a lot of moments where people would not be paying attention to their health, die suddenly, and go "Wait what the fuck? What killed me?".
As soon as I added a low-on-health sound (for DemonSteele) or changed health text color (for High Noon Drifter), there was something instantly changed that they could notice, and the problem stopped.

I really hate the Bloody Screen So Real effect, but in terms of telling a player to pay attention, it's probably the most obvious. Having a lasting glow that fades away over time might also work, though, and it'd let me get away from that goddamned strawberry jam smeared all over the screen.
I'll experiment around with it, thank you.

I had the same criticism in one of my games (red sky) , I think that lots of games having these indicators have made people less able to just use the hud like normal.

you really shouldn't be animating in unity in the first place

it's best to assume anything relating to animation in unity is just broken. most things work, but the few things that don't are infuriating.
try to keep most of your asset creation outside of the program whenever possible.

I'm not animating on it, it's just the preview, you also have to use the animation tab to put events on it. Apparently it's not triggered by the animation tab preview too, it was just a coincidence. answers.unity.com/questions/1012246/findgameobjectswithtag-finds-an-object-that-doesnt.html this is exactly what was happening.

or you could just go to the import settings of the animation, assuming that you made it outside of unity
and that shit gets previewed in the bottom right of the editor, not in the scene view

oh wow, that is fucked. I wonder if that affects built games or if that's just an editor bug.

I know, but if you want to compare size of your animation to objects on the world you need to use it on the scene view.

I think it's just an editor bug, I believe this is the piece of code that's guilty: github.com/MattRix/UnityDecompiled/blob/master/UnityEditor/UnityEditor/EditorUtility.cs#L259

I was thinking it wasn't related to the scene view preview but I changed my mind, this code piece of code seems to be the one that creates the objects with the (Clone)AnimationPreview name, like I thought, it's instantiating the object on scene view for preview but failing to clean it properly. Restarting Unity fixes it, but using the animation preview seems to cause the bug again, it just happens randomly. I don't think I can reliably pinpoint when it happens.

For me at least it won't be a problem in the end, I already have a workaround and I plan to remove all gameobject.find when cleaning for release.

the cpu that will never run any interesting code

GitHub's gamejam has ended:
github.com/blog/2483-game-off-2017-winners

learn about dictionaries
now don't quote me on that but it could be the reason many games use scripting languages for their characters. for example in godot, just scan for nodes in a group, then just store the transform, forces, and data in a dictionary (hell, i could use the same dictionary for gameplay), and parse it through (built in) json. then in reverse to load game.

Dictionaries aren't magic. They're just one of many ways to store a collection of objects. What would be more beneficial in this case is serialization

How good is SFML?

Pretty good, what do you need exactly ?

Remake SS13 in a year or two from now

Hate VHDL nearly as much as UE4 Blueprints, thank god i only had it for one year.

Why not use the Godot engine ?

I always pictured engines as being tailored for simple games, or generic projects

Well, go for sfml then.

I almost got to do image processing on an FPGA for a robotics competition. It was one of the last times I got super excited about doing a project. Unfortunately, the person who took over the team decided to dedicate it teaching useless retards how to do things instead of actually making robots. Ignoring another team member and I's insight that useless retards will always abandon you at the first sign of resistance, and anyone worth while will pick it up along the way.

The useless retards couldn't even do our condescendingly easy sample project before abandoning and the team fell apart very shortly after I quit

i've been stuck for weeks, just finished classes awhile ago now im starting back up on my projects. but here is my problem and i've post this last thread but no answer, so right now im sculpting pic and before i move forward i need to know three things.
1. clothes; do i sculpt them or not, and how do it work with the topology that i will face later?
2. anyone got a proportion sheet, i saw one on loomis a long time ago that measure the body by heads and i can use that.
3. i want to do a slender male character about in their early 20's however i know most proportion sheet is going to have the ideal body so how do i translate that to what i want?

thank you all, and have a wonderful holiday
and yes i know there are somethings wrong like the neck but that's another reason i need a proportion sheet

Either sculpt them or use Cloth simulation in Blender or Marevelous Designer. Could also do it in combination. You could also paint displacement/bump maps directly.
Copy parts of your body lowpoly basemesh that are directly below your clothing, and then cut off the parts that stick out, then get is as close to the highpoly clothing as you can for example with shrinkwrap modifier. If you do this after rigging your basemesh, you don't even have to re-rig much the clothing too
Pinterest is filled with anatomy sheets, proportion stuff and real life examples.

Wonderful holidays to you too.

What kind of game would you expect to hear this in?

also does it suck?

MMBN vibes, sort of

Sort of reminded me of Black Rock Shooter: The Game, although maybe it's not quite as similar as I thought it would be as I go back listen to it again.
youtube.com/watch?v=8zR69wrzqpo&list=PLADDC01EC004022DE&index=1

Also sort of reminded me of some Yume Nikki music.
youtube.com/watch?v=_GrzwtHfZxg

It might not be bad as an asethetic/menu/overworld map theme, although it doesn't have any sort of memorable melody that's going to keep it stuck in a listener's head.

I'm starting a modeling blog in a thread over on 8ch.net/agdg/res/30078.html

1. I found it easiest to copy topology of the parts of the body it covers and then just extruded outward (soildify). That way it copies the topology and the weights so its perfect and wont ever clip with your body.
2. I'll be posting my folders of reference I found later
3. Idk

It feels very tense. Something bad is about to happen, but it isn't happening right now or if it is it's happening somewhere else. Point is there's no action, just tension.

If this is going to be a overworld theme it would need to be one during a race against the clock.

Is there an example of a good nu-low poly scene. Not just isolated models, but an entire scene with low poly. I always see things like the TF2 low-poly models, the knight and the JRPG chick, although I never see them in an actual scene.

I say this because lately I've been playing a ton of retro games and they go from looking amazing at low resolution to shit when upscaled. Pic related. I was playing dungeon keeper 1 for the first time on dosbox unmodded, and fell in love with the graphics style. I looked up some screenshots and they all looked terrible just because of how you can see the polygon edges crystal clear for everything. And it's like this with almost every retro game I play. It got me thinking about how most games now are going to be on a 1080p monitor, have so much space to fill and can't hide their flaws as easily. It looks like nu-low poly isn't really that viable.

Food Fight is pretty good too.

that's not what you do. You learn things as you go, and when you need them. I've been modding games for years, and I'm currently developing my own FPS sandbox game - I still have no fucking idea how Trig works - I just look up how to get something I need, and if Trig is involved, I use it.

I'm not proud of my own ignorance, mind you. I should probably learn it. But my point is - you don't need to know it to use it to get the effect you need.

Because 3D is fun.

Probably because you were intimidated by them as a child. Math is very simple, it's all about understanding it by heart, and that's done by explaining things properly.

That actually looks extraordinarily good. You got that puffiness just right.

sketchfab.com/tags/low-poly

It doesn't really look like it works. Although pretty much all of those do that style where they show off the limitations rather than conceal them and cake tons of shaders ontop. I was thinking something more along the lines of pic related.

Here are some relatively recent games off the top of my head with low poly assets (I specify assets because the actual scenes portrayed in the games use a lot of polys collectively). Two aren't even released.
Modern low poly, at least the type that actually makes it into a game and is not just one of the hundreds of useless artist mock ups, tends to skimp out on textures. Personally I think devils daggers is the best example of modern low poly which probably has something to do with the fact that they keep all the other old rendering limitations. (vertex jitter, low color depth etc.)

Everytime I play old 3D games and then I see those screenshot it fucking pisses me off how those assholes are missing the point of low-polygon. And now when I want to search for actual low poly screenshots then it is cluttered with those trash. Thanks assholes.

After a week of fucking around i managed to make my weapon bob around as the player moves. There is still room for improvements, but i tried to make it look like HL2's weapon bobbings.

Look at this shit user. Let me tell you if you're watching: that this character controller is nearing perfection week by week, and still got 2 bugs left to solve. And let me tell you how much i fucking suck at anything related to mathemagics. But i didn't had to use much maths. Only a tween and a curve that forms an 8 shape, and script the weapon to follow along this curve while the player moves. Second webm is for you.

Weapon bobbing is tricky to get right, too slow and it feels like it's not moving right, too fast and it looks weird

Also don't forget the weapon moving more into the view if you're looking down, and less into view if you're looking up

good shit though

Thanks user, i had your idea in mind too to make sure the weapon bloats the view as less as possible.

that isn't low poly, thats just a chunky simple style, low poly objects without textures are just very simple shapes with no definition, low poly art is all in the textures

mega.nz/#!ouhGXQiZ!Uz7Oe2qY_PQTqbrd0OoZ9ostDqMDvhaawIKD60f3lro has a whole bunch of low poly examples.

I'm alright with the distinction, but if the lad was asking for "nu-low poly" these are the closest examples I know for actual games and not artists wasting time.

You sick fuck.
Nice references, quite a few duplicates in there though.

Please tell me this isn't real.

Making 3d models is fucking hard dude. Sprites on the other hand can be made in fucking paint.net if you're desperate enough.

Correction, making good 3D models is fucking hard. But the same is true for making good sprites. And making a lot of animations with good sprites is way more work and harder than making a lot of animations with a 3D model.

PROGRESS!
The character controller is almost done, I finally fixed the rotation of my animations so now I'm not gimbal locked. I think the next thing I will do is polish those animations.

How's your character moving? Root motion? Or whatever that thing where the movement is dictated by the animations is called?

That's some impressive shit, I couldn't really understand it myself.

I've been checking out the limits of my wall movement script. After a few fixes I think there's no more bugs and it seems to work well. Now I just have to rewrite a completely different control scheme for it.

Did the robot turn into a giant piece of shit?

...

since you are going to use godot you are bound to head into animation tree player territory.
play the animations through the tree (by changing the parameters of the tree), not the animator player. the main animation player option is your library AND you can load other animations from other players, but they need to be their own scene/files(you can save animations as files).
this is what my tree looks like, in the upper branch i'm feeding walking animations and in the bottom one i'm just manipulating the seek of and aiming_upwards_downwards animation, then use a bone attachment bone for the guns and stuff
in case you want to take a peek: github.com/ostoru/vcsona
gun bone is broken for now

I know the animation tree is coming for me, i'm quite impressed by it, and have no clue how it works. Thanks for the ressources, i'll study them tomorrow.

trying to figure out how to align accesses properly

That giant piece of shit is a placeholder for displaced dirt, meant to make it look as if the robot is not actually walking on the wall but instead buries into the ground and drills a tunnel with its nose? Am I getting it right?

There's also the wonderful world of multi-user dungeons, for when the edginess of HellMOO just isn't enough. All the same sperginess of rapey fetishistic furfaggotry, but in text form. Those places are where some of the furfag-specific vernacular came from.
In my brief visits many years ago, looking for disturbing or cringe-worthy material, I remember the people in those places managing to come across as either almost normal or completely deranged, taking very horrifying fantasies very seriously.
If you want good autism, stick with visuals. Or hope someone is kind enough to curate for you, so you don't have to dig into such cesspits.

On a more topical note, pandering to furfags can be a valid marketing strategy. You'll alienate the larger potential audience for a small, fickle but dedicated fanbase with much, much lower standards. Guaranteed income as long as you don't say the wrong things or have the wrong thoughts. It'll leave a stain on your track record if you choose to keep using the same name, and a stain on your soul if you choose to keep living.

Look up "world space texture shader".

Some animations like the walk, run, 180turn and climb I use root motion, others like the fall and jumps I use simple gravity and speed accelerations. I made it this way so I can create things like a feather fall potion or a long jump potion.

While I like the old games, I feel like dying if a game doesn't have widescreen support. What's a good resolution to use?

I like the idea of having a square or semi-square playing field, with a sidebar taking up the rest of the area. What I've done on my own game, for example, is is have a 320x320 grid at 32x32, then upscale it 2x, finally centering it and slapping UI on it

I use 640x360 for my pixelshit. A slightly bigger window would then be 854x480.

Doing displacement is a lot of work honestly.

That's a placeholder without animations and particles. The final one will have moving textures, rock models rotating around and will deform or stretch depending on the players action.

It'll look better once it's done

Fucking mutexes, how do they work?

I tried to make my con_printf function work asynchronously by adding a mutex, which would just block a thread from writing to the console until every other thread was done, but it looks like the threads are just ignoring it and writing over each other.

The mutex is a valid one, and the error condition doesn't happen. So, I'm not sure why the console output is messed up like this. Each frame on the server thread at 40FPS, this call is sent:con_printf("Sending message: %s to socket protocol %s:%d",buf,inet_ntoa(from.sin_addr),ntohs(from.sin_port));
and this is sent from the client thread at 60FPS:con_printf("receiving message %s",buf);

All references to con->buffer are using this mutex, so I don't understand the issue that's being had. Is the issue that I am accessing con.write_mtx asynchronously? In the example of how to use mutex objects on MSDN they don't do this. msdn.microsoft.com/en-us/library/windows/desktop/ms686927(v=vs.85).aspx

I'm not sure. I too have taken note that having a square work space is 10x as comfy as having a wide screen. Everything is just in the center of focus. At the same time though no one will ever go for it because having widescreen is more convenient to work with because you have more to see at once. Some screens are so wide now that you would need to literally cut the screen in half to get the desired effect. It also acts as a feedback loop. UIs can have so much detail that it's hard to make them bulky, meaning more space is available, meaning it's harder to cover half the screen.

Well, fuck

nehe.gamedev.net/
OpenGL!

The #ifdefs are actually defined? Write up a minimum working example of using the mutexes to ensure you're using them correctly.

Are you using Linux? Right now MonoGame is bitch to install on it since MonoDevelop decided to use flatpak and MonoGame doesn't seem to want to make the switch.

Just fuck around with Unity and GameMaker for a few days each to see if they actually aren't suited.

It's just like SFML but more barebones

Yeah, the ifdefs are all defined.

OpenGL has more capabilities, for sure. Think of all the AAA games that used legacy OpenGL…

nuDoom uses OpenGL 4.x and Vulkan. AAA devs don't ignore it because it's incapable, they ignore it because they aren't familiar with it.

GLFW, create OpenGL context, go to town. Build off of the demo programs that come distributed with it. Cmake to build, so not all that hard once you have the compiler installed. About that…

I actually recommend nuwen.net/mingw.html distribution. Get a 64bit mingw system with pre-compiled SDL, zlib, ogg/vorbis, libpng, jpg, GLM, glbinding.

From their, start adding things like Bullet3D, LuaJIT, whatever else you feel like exploring.

I just had trouble installing dependencies, why the fuck do you think I would be better off with a C compiler? And why is every C compiler I can find like a 250 mb download? I thought it was supposed to be lightweight. No im sticking with C#

I would say at this point, a majority of games are using OpenGL (and GLSL shader code) for rendering. It's just not talked about much. Most people are using an engine, any many of those are using OpenGL. Why? Because it's cross-platform. Engines want to easily allow released on consoles, Linux, Mac, and Windows. OpenGL allows this. Modern OpenGL is essentially bindless and stateless. The worst thing about OpenGL is all of the legacy BS around the web. Most tutorials teach outdated OpenGL, or outdated methods.

A couple weeks ago I stuck several notes on my monitor to fight my depressive swings, mostly reminding about my tendency to see everything in a negative light when depressed, and insisting that if I feel like my game is garbage and I'm bad at things - it's just my depression. So far it worked beautifully. Like, it really works as a reality check. I recommend it strongly, especially to unmedicated anons around here.

When you get frustrated and feel like you can't do anything - that's because you're too agitated to think straight. I recently realized just how fucking tense I am on a day-to-day basis because of my fucking baggage. Relaxing and slowing the fuck down is what you need, so then do meditation.

No, not the spiritual kind. It's sort of like cleaning your cache. Just sit there for 15-20 minutes, with a straight back, close your eyes and breathe with your nose. Focus on your breath, and specifically on the air coming in and out of your nostrils. The more agitated you are, the harder it's going to be to pay constant attention to it. Don't think about anything but that, and when your timer goes off - you just get back to what you were doing.

These two things are pretty obvious to well-adjusted people, but I'm pretty sure there's going to be a few people like me here, too. I know you hear about these things a lot, and you think they don't help - but although mental illness is real, it's still just in your head and it works by warping your perspective. These two methods work by letting you "remember" what your normal state feels like and helps break the cycle of depressive swings and anxiety.

A lot of never-ever games come from lack of persistence. If you feel like you're not good enough - it's just your sad self talking. If you feel like you can't think straight or can't figure out what to do - it's your anxiety talking. Learn to recognize your mood patterns and perspective shifts and combat them.

Everyone here can make it. You just have to learn how to never stop moving and never doubt your ability to learn, improve and persist.

You really should be using gamemaker or unity.

100% great fucking advice. I'm glad it's working for you. I myself have fallen into the traps you describe. The notes are a good idea. It's easy to forget these simple facts at first and fall into familiar, negative modes of thought.

Inability to recognize mood patterns is what really hurt my teenage years. Both academia and shit like drawing and animation I'd be much better at if I knew how to handle my brain's tard wrath back then. These self-control methods have reduced my downtime significantly. I've been doing a lot, a lot more work lately and it feels great, even when I'm feeling down.

Also eat breakfast, drink tea or coffee in the morning and at least three cups of water before and after you wake up. Your body is a machine, and unless you're fed and hydrated, it pulls blood away from your brain to power your body to get sustenance. The difference is staggering once you really start doing these healthy things. I used to roll my eyes at that, as though it's placebo - but honestly, it's amazing. I even perform better in vidya that require attention and reaction time.

I was talking about the type of OpenGL that my link teaches, which was used in the 90s mostly, and what was AAA in the 90s. Fixed function OpenGL is great.

What compilers have you found that go over 250 MB? You've probably just been downloading software suites that have a bunch of bloat in them.

I can show you a few C compilers that are small:

bellard.org/tcc/ - Tiny C Compiler - 274K
smorgasbordet.com/pellesc/index.htm - Pelle's C Compiler - 9.6M
openwatcom.org/ - Open Watcom C/C++ - 80M

I Would recommend Pelles C since that is the best software here, the others are hard to use.

Great for making retro vidya and toaster-friendly stuff, but OpenGL 3+ can be a lot faster on DX10-tier hardware and up. Unless you're working on an actual low-poly game anything earlier than OpenGL 2 is bending over backwards to support an API for ancient hardware that probably won't run your game anyways.

All of my thinkpads have GPU's that do not support newer OpenGL standards. So, it's just a necessity for me that they support those older API's. Otherwise, I can only develop on my desktop- and that hurts my productivity. I think that it's actually not that big of a deal to support them, since its a much simpler and easier to use system than any new API, so the code is much quicker to write. I also like using the older computers as an optimization target, to make sure that my engine isn't running like shit and hasn't broken compatibility with windows XP. I am making an actual low poly game, but to make sure that its a "real" low poly game, it should at least be able to run on a Pentium M and Radeon Mobility 9000…

You're sure your Thinkpads don't support OpenGL 2.x? Your points are valid otherwise.

I asked in /agdg/, but I might as well ask it here.

How do I into python?

I am doing it on code academy at the moment but I feel as if I am bruteforcing the exercises. Is this something that I have to work towards? Or is programming really just an exercise in bullshit?

Sometimes it can be, but when you're starting out it shouldn't be bullshit and you shouldn't be bruteforcing solutions. You aren't expected to solve every problem on your first go at it especially when you're just starting out, and experimenting with possible solutions to see what works is an ever-present part of programming.
Maybe if you give an example of what you're doing and how you are doing it I can give better advice.

Not to mention most consoles use OpenGL. Aside from the xbone.

do sicp you retards

Truly this place can be ressourceful with encouragements and your post is one fine example. Was it true that the greatest artists known to man, such a Van Ghog, were depressed as fuck when he was creating his masterpeices? If so, then he used his shit to create quality shit, and that's so fucking badass in my book that i might just do the same for my game. In fact i will do the same. I prayed to God so much so he could give me the intelligence to code and finally exteriorize my shit and now i am able to code a character controller with it's weapon bobbing around. And i thought i wasn't smart enough to code. In other words, follow suicidal artists example to free my mind and God helps me do it.

Witness me as my reputation gets flush down the toilet all over the internet including on here.
[spoiler]>yfw you pray to God to get 1337 coding skills and gives it to you.

i forgot spoilers doesn't work on totally accidental reddit spacing

My machines have the varying levels of support, for: 1.3, 2.0, and 2.1. On Linux it's somewhat worse because I don't know how to do that many things, I know that my distro (slackware) came with Mesa and it supported maybe 1.1 with no extensions? Not sure, but it ran fine.

I don't do any development on the 1.3 machine, but it's my optimization target. The game should be playable on low settings on that machine for me to be satisfied with its performance. I do plan on using shaders at some point (although it will be optional) in the OpenGL renderer. On any modern computer however, the user can just use the Vulkan renderer. I'm not neglecting more modern things, for example the engine can run in immediate mode or use VBO's for everything on the OpenGL side.

I like OpenGL 1.1 because of how easy it is to write the code, so my map editor is just entirely done in OpenGL 1.1 because I can write the graphics code really fast and I just run it on a powerful enough machine anyways (not that its slow…). So, that's why I think its good for beginners, it's like this middle ground between advanced modern OpenGL and libraries that are easy to use but hide things from the user.

I'm going to post a webm soon (maybe tomorrow) of me going around in the level. I'm implementing movement slowly because movement has to be done by connecting to a server that has all of the authority over where you move, so that the client is just a dumb graphics engine that isn't really aware of very much. So, I've been working on this. Right now I won't implement a working client-server, just the bare minimum to move around with no collision detection.

Alright /agdg/. I'm about to make my new year's resolution.
THIS YEAR I AM GOING TO START MAKING MY GAME!
I AM GOING TO FINISH MY ENGINE BY THE END OF NEXT YEAR AND BEYOND THAT FUCKING FINISHING IT!
BY THE POWER OF SHITPOSTING, THREE MINUTE NOODLES AND A FUCK TON OF ASPIRATION I WILL FUCKING DO IT!

...

i understand where you're coming from, and i agree that you should channel your negative energy into positive output. that being said, van gogh was aun untalented hack, as were the rest of the modernist/abstract/postmodernist/cubist/dadaist socalled 'artists'. shit, the lot of them. gl on your game tho :D

Why start next year?
Start this week.

a = 0b1110b = 0b101print a|b without using bitwise operators

glot.io/snippets/ewqkxic6gw

That is as far as I've gotten.
please no bully

Again, this is really brute-force-y, but what I was doing was;
1. Converting a and b into a list
2. Removing 0 and b from a and b
3. Reversing the order of a and b
4. Comparing if the xth bit in a or the xth bit in b is 1, adding it to c if it is, 0 if false
5. Converting c back to binary by reversing and adding 0, b to the end
6. Converting the list of c back to bits

I am unable to get past step 4, lst.popleft() should do the trick, but it doesn't work. Still reading up how to's.

Does anybody have some good documentation/tutorials on how to not make your particles in Unity not look like complete trash?

This shit is confusing

glot.io/snippets/ewqnh5aa4m
Things to note: Binary is a representation of powers of two, same as octal is a representation of powers of eight, decimal is powers of ten, hexadecimal is powers of sixteen. We call the power of the representation the base. To extract the least significant digit of a representation, take the modulus of the base ie to get the 9 out of 1239, do 1239 % 10. Now we need to remove the least significant digit from the number. Do this by dividing the number by the base and discarding the fractional component. 1239 // 10 gives us 123. Store the number extracted via modulus in a string for display and then reverse the string since we put everything in least significant first.

glot.io/snippets/ewqnsvow6m
Another implementation that doesn't bitwise or the number before printing it out. Uses the same logic as above and then uses logical or to check the extracted numbers.

If this is how you are practicing programming, you are making a huge mistake. What the fuck kind of problem is that. "Do this very simple thing in the most obtuse, roundabout way possible." Do you want to be training your brain to come up with crazy multi-line solutions for something that is a single operator? What good does accomplishing this do? Find exercises that are actually practical, or better yet, start doing real programming.

I was following code academy, on the section on bitwise operators, they suggested I try it without using the operators.
The exercise was literally typing print (a|b) into the console
Where can I go to start doing real programming? I've tried project euler but the problems scaled way too fast.

That's way over my head, but it works, thanks for that. I'll try to understand it.

Find something you want to make, get an engine and try to make it.

Read the docs, google and research stuff that you need to know as you go along. OR you could buy a 500 pages book, read through it, and only then actually start making stuff. I don't think that shit works, but it depends from person to person.

Do SICP. Then just find something you want to make and then make it.

I'll just leave this here
graphics.stanford.edu/~seander/bithacks.html

use sprites that soft blend around the edges

reddit.com/r/dailyprogrammer/

No I mean as far as how to position and set them so they don't look weird, I'm having trouble with the different shapes I can use to create them

how does texture animation work in Unity?
I'm a modeler and I can't code for shit

For example, what I want to do ideally is to split the eyes and mouth of the face mesh and be able to swap them in an animation.
I think I'm suppose to create 1 texture map for the eye and mouth mesh, and having different expressions on that one map.

what I'm asking overall is what I should prepare for the programmer.

also merry xmas and i wish for everyone

Long term goal > Video games
mid term goal > machine learning
short term goal > python

Thanks. That's a buttload of text, do they really help?

Thanks. I'll give the easy ones a shot.

Why?
Why?
Neither python nor machine learning have anything to do with games

This isn't the right place to ask for that, go on /g/ or Holla Forums

just like make game, its like you forgot the most important rule of gamedev.

It's something to consider if you need to make a super low level efficient system. It'll be fine without it.

Should i add metal shoulderpads to match the boots?

my /agdg/ thread is at bump limit, but i got attached to it and don't want to make a new one ;_;

in godot, i have two objects (yaw and pitch) that in conjunction need to look to another

返事ください

What I would do, and I'm saying this without knowing the engine you're using or what exactly you want to do, is create one image that has every single variation of the texture you want

Then the coder can just change the X/Y position of the texture to swap between the different images.

Don't over design him. He looks fine the way he is.
Although if you're going for a cowboy look, give his boots some spurs.

He used to travel by foot, also spurs make it hard to sneak by goblin hordes due to the noise

you know i forgot to add some more detail to what i wanted to say and what is my plan. i want to do about 7000 to 10000 triangles, because i would like to imitate 5th - 6th gen. that gen had simple models with very good textures, i dont like what people do nowadays, with models so im taking what was in my opinion the best and using it. as for normal maps im not sure maybe, i have not thought that far ahead i mostly just want to get to retopology then worry about that. also forgot to talk about hair because im drawing at a blank their.

o yes you, thank you, your attention to detail was great

yeah found one i liked and could work with(see pic 3). however the clothes are going to be static, no animation or cloth physics so im thinking doing what said, but i should tell you what i have in mind as for design. what im thinking is dress pants,dress shirt and a dress vest, but the color is bugging me because its black and white, i think i need more color so it does not look so bland or monotone.

just done putting the bases for the arms going to Boolean them and rotate them for T-model form, when i do the bases for the legs, so how does it look proportionally.
o yeah its the 25. Merry Christmas

Merry Christmas, everybody!

I succumbed to temptation, I'm taking the day off. I'll get back to working on stuff tomorrow.

...

The one thing that's keeping me from doing an RPG is how to have persistent areas, as well as querying arbitrary game variables via scripting to see if dialog options appear, etc

I've got some of that done but it's total spaghetti code

There's multiple ways RPGs tend to do persistent areas. You can set progression variables and then on the world map decide to send the player to a different version of a town map, which works especially well if you need big changes like a town burning down. Otherwise you just add variable checks to actors themselves, for instance if you talked to Dude A and he walked away to some desert to find loot you set DudeALeft to true and check if that's false before spawning him.
Sure it requires a bunch of manual planning and adding variables, but it beats having to serialize entire maps.
Just like use debugger?

In Gamemaker how do I have one object follow another?

Like if I put in Object_A in the Step event a line that says Object_B.x = self.x and the same for the y, or maybe put a line in Object B that says self.x = Object_A.x

Problem is, every way I try it ends up with lag, where it doesn't follow exactly.

What I want to do, is have this bosses pincers at the front, work as shields, that bounce projectiles back at you. So I thought of taking them out of this sprite, and having them be their own object, but I need it to move perfectly in sync left and right with this boss, and stop whenever this boss stops, and maybe move down the screen at the player when it does, but it needs to line up exactly.

pretty sure for that, you could move the pincers in the boss object. so, in the boss object's step event, move both it and the pincers. they'll all move at once, then.

Use End Step to move objects before they are drawn.
IIRC the order for relevant events for everything every frame is

Begin Step
(Built-in input events)
Step
(Application of built in movement)
End Step
Draw

I mean shit like using Lua to query say, C# variables that may or may not exist on an object

Not too hard. If you set up your map data correctly, the tiles should call for only a few a bytes of data each. You might have a few local map variables/states to hold onto; and maybe simple object and actor states.

Its obviously much harder if you have eg Minecraft terrain

You can do that with C# itself, there's this nifty thing most modern languages have called reflection…
msdn.microsoft.com/en-us/library/system.reflection(v=vs.110).aspx
In C# you can query any type for fields/properties/methods/constructors/indexers that may or may not exist, all at runtime.

Actually you probably want to refer to the methods on the Type type
msdn.microsoft.com/en-us/library/system.type(v=vs.110).aspx
Important ones are GetProperty/GetMethod/GetField

Heres a good set of examples of how they are used
dotnetperls.com/reflection

It would be pretty easy to use reflection to write a very simple game scripting language on top of C#

I was gonna write out a little thing for you, but this website does it better.
purplemath.com/modules/numbbase.htm

It's not that hard. MC stores the changes to the map along with some information like the seed. When you load the game, the seed is used to generate the map in its initial state. After that the changes are applied. That's the reason why savegames in MC get bigger over time.

Meanwhile, on Notch's Twitter…

Why is he such a fan of shitty languages?

Hey friends, I have a question. I'm making an engine for my never ever game, which I'm making with C and Lua, and I'm wondering how to do networking for it. I wanted it to be calculated by the server, where the clients send input and the server sends state. How should I do this?

Is this a technical "how is this even done" or a general "what's the best way to structure this"?

If you don't mind, Both please.

I can't answer the second since I've never actually done it. But generally if you want to do network programming in C, you have to use the socket functionality of whatever OS you're programming for. In Windows that'd be winsock.h, in Linux it's sys/socket.h I think. I'd suggest reading up on that so you get an idea of how to actually send data over the network.
Implementing the actual server/client structure of your engine could probably be done in many different ways depending on the exact type of game it's for. If you're too afraid to get experimental though, you can always study the tried and true methods in something like the Quake 3 source code.

I'm working on this right now, with winsock on windows and soon-to-be POSIX sockets on Linux version. Windows is basically using POSIX sockets since its API is very consistent with the POSIX API, so I doubt porting it will be a big deal.

I have been sending simple messages so far, there is one client that the server is aware of and the server can receive and respond to some basic messages, for example a "say" message that, when the server receives it, it will echo to all of its clients (which is just one right now).
See my previous post:

Merry Christmas my fellow autismos.

I need some critique on my animation. I animate the weapon first, and only then the hands - but I gotta get the weapon right first.

I'll preface this by saying that I know that you don't slap guns around like that in real life. This is meant to be exaggerated and aggressive. Also, the guy who made this model pumps them out quickly so have placeholder assets for our game - it's missing certain details.

Other than that - please tell me what you think. If there's any aesthetic improvements or cool things I can add - I'm all ears.

One day i wish to get as good as Hyper or Mr Brightside

the butt slap is too noticeable, but not worth redoing just for that. but there are more important questions.
1) is your gun zeroed? this is really important, more if you are doing fps or want the npcs using them
2) can it be played in game?
many engines have limitations on how you can setup animations starting gun first can lead to serious problems. also the gun itself its the least important part, the juice is in the hands.
3) are you supposed to put this much effort in this stage?
projects fall all the time and there is no point in pouring your soul if you are bellow the minimum viable product, even if you lie to yourself that it is for your portfolio, portfolios are worth nothing and nobody sees them.

1) Define zeroed

2) Yes. I'm animating with a constraint to hands. I'm then going to be exporting the animation of the hands and removing the master bone keyframes from the gun before exporting it. This way I can have UE4-friendly animations, and this is actually a much quicker and controlled way to animate since all the focus is on the gun and I can make it move the way I want before I figure out how hands need to move to fit that movement.

3) If we're going to attract a proper artist willing to work with us under revshare, we've got to make this shit look good with what we got.

By pouring my soul into the projects I work on, I get better. Even if this whole thing falls through (which I don't think it will - I'm way too personally invested in the game's concept) I'll still retain these skills. There's absolutely no point to halfass anything. This animation took me a few hours, start to finish. It's not that much, considering that it's for one of what, a couple dozen guns we intend to have? It's a drop in the ocean of the hours I'm going to have to pump in.

1) that your animations can look where you tell it through simple commands. a problem that i'm facing right now
2)and have you tested that it actually works? what about simple modifications/updates? you might think that it is not a big deal when doing just one animation but that time and testing will eat you once you start expanding your library.
I went through great trouble to make a work flow for my animator to export once and sit back. that way he can spit out animations and test without breaking anything
3)i do not know what stage are you in, but if you still need a proof of concept be sure that it actually proves your concept first. this means placeholders only, spaghetti code, and tackling the most unique parts of your project first, whether it is multiple weapons, swappable characters or a mix of turn based system with parkour shooting, you need to be sure they can be achieved with your current team and tools. anything beyond that is wasted effort, and you are better off working on other projects on the side until your team catches on.

...

Now on topic I really like it. It's expressive animation. It's got flair.

motto of every aggydaggy dev

Not sure if this will be specific enough to help user, but in the past I've handled it like this:
>Client sends input changes to the server when changes occurred but capped to a maximum number per second (if viable for the type of game you're making)

Thanks, user

How do I make shadows in Unity? So far for dynamic objects I only got shadows to show with spot and point lights, but they look like shit for a sun, directional lights simply do not produce shadows. I have been messing with the mixed lightning and managed to bake shadows for static objects, with the light probe group I can make the dynamic objects receive shadows, but they never cast any shadow.

Mixed Indirect mode produces no shadows for either the dynamic or static objects, Shadowmask mode produces shadows on static object, but Distance Shadowmask produces no shadow if the shadow distance is high. Didn't even try subtractive because I didn't think it would be worth it.

I have been trying combinations of forward rendering or deferred, changing random settings but not even realtime GI produces shadows for dynamic objects. I'm completely lost and don't know what else to do, I tried to copy the settings on the video but nothing works, since the mixed lightning is a recent thing I haven't even been able to find proper tutorials on this.

did you check your shadow distance? quality is usually the problem for this

Just like make netcode.
First thing you need is a protocol for sending the information between servers and clients. The protocol and the code handling it are probably the most important piece to get right up front. Decide where one message ends and another one begins, and how the computer knows this. Decide if/how you want encryption done. Decide how messages are received/stored/handled. Etc.
Suggested thing: Multithread the netcode. Having multithreaded code is good, but a pain to get right. You won't want to stall the rest of your game. It's easier to avoid doing multi threaded code if you have lower level control over the sockets. You also want to do this early, since it will can be a big change to the structure of your netcode.
Next thing: RPCs (Remote Procedure Calls)- Have a layer that allows client/server to pass messages and call functions on eachother. Lowest level thing you can do which gives you the ability to basically do anything you want. Consider using information about the source of the RPC (what client sent it) as information the RPCs themselves use, which could be used to prevent any player from faking messages to move objects they don't control, for example.
You will definitely want a set of RPCs that sync data between client and server, with the server sending deltas (only what changed between ticks) to connected clients every tick. You don't want to send actual delta positions, but new information about objects, if it changed (can send both position and velocity of an object)
After that, it is easy to write any system you want in the game as a collection of RPCs and shared data. Entities, chat, inventories, etc. Just structure any player action as an RPC.

I've been writing a custom netcode system in Unity, using C#4.6 Tasks to "multithread" it. So far it's turning out pretty well, hardest part was getting the tasks to play nicely with Unity's main thread and play/pause/stop state. Thankfully pretty much all tasks end up getting killed every time scripts recompile.

Well FUCK ME. I had set the camera far plane to something like 10 million and I discovered that setting it to more than 2 million fucks up the shadows.

After about a week of full commitment and not really knowing what I'm doing turns out airstrafing isn't hard to implement. There is no reason a FPS game shouldn't have it.

pretty not fucking bad i like it

Here it is in motion Connecting and disconnecting from another comp, with a little bit of motion interpolation.

What did you expect setting the clipping plane that far away?

Neat. I wish that we actually had more games with mechanics like bunnyhopping/strafejumping.

I've implemented it in my side project remaking Hidden: Source a while ago. It's not hard to implement, but it's a tad difficult to figure out how to slow down the speed gain.

I was messing around with it sometime ago and forgot to set it back to normal levels

Faggot.

Yo this looks rad as fuck.

How do you do low poly hair physics in blender for unity? Do you make an armature and rig the hair?
I never worked with more than one armature on a single character, how would that work?
I have been looking for tutorials, i only find that hyper realistic hair crap for unity
Also what do you guys use for character clothing? Do you use the Unity cloth system?

This shit is why video games have stagnated

My melanin enriched brethren

i don't use unity but i do this, then shape key the fuck out of it (anim-all add on might be involved)

for clothing i use this other and also shape key the fuck out of it too.
it takes age to render so i made a computer out of old components and leave it running overnight/a few days.
keep in mind that for it to work properly your humans need to be like 10 times bigger there is probably a setting to change scale somewhere else

thanks i will check it

Games usually avoid realistic violence against children, but is it just because they want to avoid controversy or is there actually some law about it? I don't want to get in trouble.

Notch is shit at game design. He just likes making little toys. He could make another game based on randomly generated cubes but it would still be a shit game even if the cubes were generated better.
polite off-topic sage

don't forget your hair also don't learn everything just because you can, pick one or two things and learn them good, then participate in other people's projects

depends where you live

Well I'm Canadian so I already know I can't have any child diddling, but I want to sell my game in the US and Europe too

I think that it should also be noted that while notch is definitely more popular, Ken Silverman's work on voxels from 2003 blows anything notch has created out of the water. Imagine what this engine could do if it was taking advantage of modern hardware- it's not even using the GPU, it's designed for the Pentium 4, and yet it's still comparable to minecraft and notch's demos.

advsys.net/ken/voxlap/voxlap05.htm

The only problem that it has is that ken's code is insane and impossible to read. His programming style involves using two letter variable names for everything, and his code is littered with inline assembly blocks. The engine is one 12000 line .c file (it's C++, but he said in the readme he names them .c "to save typing".) Basically every single "style rule" that you would normally see has been broken. It looks like the code that you get when you try and generate C from a disassembled program.

Trudeau will probably kick down your door. Come down south and enjoy some freedom.

Honestly, I wouldn't bother. That sort of shit sounds like it's a recipe for the wrong kind of controversy. Then again, I don't know what you're working on.

Reminds me of Ace of Spades.

I don't have enough money to immigrate… and when I do I'd rather go to Japan

If it's just a controversy I don't mind, since the rest of the game is controversial anyway, so I figured I might as well go all out. I'm just wondering on the legality.

You DO know porn needs to be censored there, right? Just saying.

I'm not making porn tough

That's because ace of spades is using Voxlap as it's engine. See: buildandshoot.com/download/

I just don't see why you'd trade one type of authoritarianism for another. I guess it's just the way of the leaf to be a bootlicker.

You guys don't know nothing about freedom

crazy bump but free and for linux

Whether I want to deal with a controversy very much depends on what the controversy is about. Anyway, you will have the rating agencies of all countries breathe down your neck. Since they don't all subscribe to the same universal book of rules and have a tendency to be quite arbitrary (at least here in Germany), it means that you will have to fight a lot of fights.

I knew that, user. It still hurts how badly the dev got fucked over.

SILVERMAN!
SILVERMAN!
SILVERMAN!

Lighting looks good enough to me now. Gonna start reading up on terrain modeling on blender, unity terrain sucks, can't even make a cave on it.

if you are doing caves in blender don't go sculpt head first, use the boolean modifier to chain entrances and chambers together then sculpt that. also don't overdo the sculpting, most of the detail should be done painting cloud.blender.org/p/creature-factory-2/56041520044a2a00caa7b06c
also enable the tree generating add on.
also learn about duplifaces, dupliverts and groups. it will save you lots of time when you are trying to place 9000 trees/lam posts/houses at once

If you're going with a photorealistic look, turning your terrain into a normal mesh is a really bad idea. Texturing and LOD are going to be a royal pain in the ass. It doesn't surprise me that Unity's built-in terrain system doesn't support holes. People have been asking for it for ages. You'd need a third party solution or do it yourself. I don't know if there are any tutorials.

i think you'd have an easier time just using unity for most of the terrain, converting it to a mesh and then adding caves in blender
there were alternatives like rendering caves over the terrain, but i never managed to get that to work, and i'm even more confused how you'd work out collision that way

...

Like said, using meshes only is likely to be a royal pain in the ass. You might want to look into implementing something like MarchingCubes so you can just edit it in the engine. There's some hits when you type in "Unity Marching cubes" so you might not even have to implement it yourself, or only have to change it a little to add in-editor editing.

Heh, I used to have a teacher who programmed like that. He also wrote a raytracer that fit on a business card.

I just wish I could read it, but between the two letter variables and the walls of SIMD assembly I cant.

Anyway I have been working in my mouse input, I implemented mouse input using the raw input API's, I also will implement it using directinput as well, I think that will be enough for windows. I don't really have a reason to implement the directinput method yet because all of my windows NT computers support raw input, though. I don't have pre-NT windows versions installed anywhere.

Isn't Ken Silverman a Jew?

Nof course you idioto

But I guess he's not a *prime* jew since his name is only Silverman instead of Goldman, right?

Exactly, that's why he actually produces things instead of jewing goys.

I could probably translate and comment it a little, might do it if I feel like being not lazy.

God dammit user give me my sides back.

I've met plenty of really stereotypical jews, but all the autistic ones I've seen have been really unkike-y. We should ask China how they made those monkeys autistic and force it upon the jews.

It's not so horrible, its just that I dont know assembly that well and translating that code would take a while before I could understand it. The engine is 12000 lines of that, so it's really not going to be easy.

I know one who made a game that crashed on launch and jewed his way into getting it accepted for a game jam and graded by four judges, none of which even mentioned the game wouldn't fucking load by default. His game was the blandest SJWeeb jrpg knockoff you could possibly imagine and even with all the grooming from the judges and excusing him from the jam's basic quality standards they still couldn't give him an award.

is it feasible for me to write a 3d game in C, rather than C++?
I'm rather hesitant to learn C++ solely to make a game, especially if I can do it in C instead.

Ika is writing his second 3D C engine as we speak. It's uncommon these days but still doable.

All the Id engines until Doom 3 were written in C. Yeah it's possible, but

Whoops I fucked that up. It's possible, but I can't imagine going back to C personally.

Personally I'm taking an approach of C-with-classes, in other words the only C++ feature I use is (non-template) classes (so no STL, no template classes, no 'auto', yes some 'new' and 'delete' statements for class pointers).

I think this C-with-classes approach works out really well, but if even C++ classes leave a sour taste in your mouth, there's nothing wrong with a 100% C approach.
It's worth noting that Ocarina of Time was written in C.

make a video game about space ship to space ship boarding

Totally doable. I don't think that using C++ will help you write your game faster (although I don't know- the last time I wrote C++ was in 2014…) than it would be if you just used C.

There are tons of old engines that are written in C as well. You could write your own engine, or use someone's engine, or upgrade someone's engine to support features you want.

Nice, thanks anons. I haven't really decided on an art style yet, I'm a programmer, not an artist so that will take a while for me to decide on something. I'm leaning more to the simplistic/clean design with vibrant colors but not low-poly. For now I'm just gonna continue working on low-detail stuff and later I will make the decision, perhaps with the environment modeled it becomes easier to test out different styles.

And then compiled into assembly for the console.

My dick has hijacked my comfy loli waifu vr game

you do realize you could make much more money if it was just average 3D porn and stay safe from jail?

Only leaves get jailed for polygons.

I'm not doing it for money. I do my thing for fun.

why pedo shit?

sure, I can see the appeal.
but unless you jack off to loli shit I don't see the appeal, and fapping off your own I just can't do it.

loli is ok when is rule 34 I think.

Do you have any clue what you just said? 'Cause I sure don't.

Thanks. Do you think "Serious Velocity Encounter" would be a too silly of a name for such mod?

Not sure what you mean. The speed gained in the air is worthless on its own if you can't carry it. TF2, CS and other games on source don't have a problem with that. They just don't let you carry the momentum between jumps. If you meant speed gained in the air, that is adjusted with multipliers, sv_airaccelerate in source, for example. Have you looked at their code?

I like girls that are 8-10 younger than myself, so when I was 16 I was into 6 year old girls.

now I want to fuck 17-18 year old girls.

also, most males aren't pedos.

what

C is compilable to pretty much any CPU Instruction Set's assembly. What is your point?

well, if you're into girls that are younger than you, at some point you'll hit pedo shit when you're a teenager.

user that made no sense. Again, I'm doing it for fun. I don't live in a shithole anti-loli state and the appeal of my sim is 100% for my tastes.

post screenshot of your 3D loli.

...

wow

Am away from pc right now but here's a render I did a couple days ago.

...

Oh yeah that's an old shirt. I redid the tits you can see my progress thread here: 8ch.net/agdg/res/30078.html

Hey it's Shinobufag. Have you impregnated your loli waifu yet?

Texturing will be a huge pain in the ass. Collisions, too, because you can't just take the terrain mesh and use it as collision mesh. Collision meshes have to be convex so good luck with holes that lead into caves.
Texture blending will require you to write a custom, layered shader unless you want to draw every single texture variation into its own texture maps, which will take forever and run like ass. And you will need blending unless your terrain uses very simple materials. Basically, if you want to use more than one texture per mesh. Shit like the Spyro terrain other Anons posted works fine, but if you want more than that…

Ok.

You know what actually they were too low barely. I just raised them up a little bit and double checked with reference. Thanks.

Eventually I'd like to go ham on my model with an onahole and my Vive but I haven't formally started my project I'm just doing basic assets. Also learning about modeling is fun.

That's what I was commenting on. I think they were a little low even for the blonde in this picture.

So what is the whole project about? Is it just to have a private little virtual world with you and your waifu?

Daily reminder that lolishit is low T. High T is big tiddy and ass.

Hey y'all niggers need to focus on gamedev, this shit's too off topic

Pretty much. Ideally I'd like a sim where I can just put my headset on after coming from work and see her smile and play small games with her that give me credits to buy her cute clothes. Kind of like an interactive Steam VR Home world. Sex is not supposed to be a focus but I mean she's too beautiful to miss out on having missionary while handholding.

So by that logic, inflation fetishists are the manliest of men?

A true patrician enjoys the best of all worlds

where were you yesterday

I gotta say, that sounds pretty damn comfy. If I ever met you in the real world, though, I'd buy you a good beverage of your choice for your consumption tbh.

This.

Epic argument dude.

Even you have to acknowledge that if the blonde in had any larger tits, they'd look like she stole them from an orangutan, or Michelle Obama.

I'm talking about anatomy and art. That's tangentially related to game dev. Games need aesthetic characters.

At your mum's.

Ever play Pong?

Pong is the gayest and most boring game ever though.
Not an argument.

Not much anymore. It's boring as shit. Now, if it had more porn, that'd be another thing.

You're free to post progress whenever you want.

...

It wasn't even the first game though, fags.

A better argument would be Asteroids or Space Invaders, with regards to "simple" aesthetic - which still managed to have defined entities, not just blocks and ASCII art despite their severely limited hardware.

Visuals have always mattered in games, that's why it's called fucking VIDEO games, not Bleep Bloops for Blind Niggers.

That's the point. People kept playing those games despite no super appealing characters in it. Because it was fun.

Or because there wasn't shit fuck all else as an alternative. Most people don't play pong now because its gameplay and visuals are outdated. That being said, graphical fidelity =/= aesthetic. A lot of people get the two mixed up - an example of aesthetic and general usage of simple artistic methods such as color theory would be pics related.

So you're telling me if Minecraft would've had a round aesthetic rather than a blocky one it wouldn't have sold millions because of its gameplay?

I know that's a rhetorical question, but yeah probably. The blocky aesthetic helped its "simplicity" vibe, causing it to surge in popularity with autists (see: Wurm not doing as great).

Also Minecraft has TERRIBLE gameplay.

That may be so today, but it could draw you in for hours and hours even after you've seen all there was to see aesthetic and environmental wise.

Personally I'd say rather than a game needing aesthetics or aesthetic characters I'd go with "it can only benefit" from having them when done right.
And to me done right is when it doesn't interfere with the gameplay but rather supports it.

I think that Persona 5 is a great example for that.
The style has changed quite a bit over the years.
The aesthetics have become a bonus, a cherry on top and because they're done so well they make you forget that you're just playing another persona where you go around collecting personas to kick ass.

tl;dr:
The gameplay is there to have fun with while aesthetics take care of the atmosphere and emotions.
They are no must but depending on the game they become invaluable assets.

Persona 5 is boring as fuck.

Animated the hands. Thoughts?

...

This was a test of camera controls, its a feature that needs a lot of work since I just wrote a very basic version of it to test it out. It's running in 320x200 because my computer isn't that great at encoding videos, so this makes encoding faster. The engine supports this resolution, even if it looks grainy, I did especially fix the font rendering to work in this mode a while back during the summer.

terrible games benefit more from an aesthetic.

fingers needs more attention, they don't seem to be pulling the gun but the palms, also lacks force.

How is Gamemaker Studio 2? I'm familiar with Gamemaker from years ago but not at all with Studio 2. How is it?

Can you have an armature inside another? I want to animate hair and cloth with their own bones, but how does it would work? can i anchor an armature (hair-rig) to a specific bone in another one (main-rig head bone)?
I need each cloth/hair to have their own armature because i want to allow players to customize their character and allow NPCs to use different clothes based on the season

I have seen my artist do it, so I know it's definitely possible.
Wouldn't know how tho

could you ask him for a link or at least a search term to look for a guide? i have been trying anything from multiple armatures to nested armatures, cloth rigs, etc… and i can't find something that explains how to do this

Gay.

Most models I've seen just put everything into one like image related.

Yes, but in my case i cannot do that, as i would have many unused bones on each model if i did that.
I need to make it so the bones will only be there if certain cloth/hair mesh is present

and now I feel dumb because I reread

The only way to do what you want is to do each item as a separate model, and have a separate renderer for each, unless you want to write something that will combine the skeletons
Customizable characters are really hard to do correctly, especially if you want clothes to be included in the customization.

Also let me clarify for
They made a mistake using HTML5 for the editor, and it sucks. The only good thing about any changes they made was that now projects play nicely with source control, but they screwed almost everything else up.

I talked to my bro, we decided to make like you said, each moving part will be a separated model, and he will make sure the model is attached to a bone on the main rig when you equip it in Unity
Like a hair model will be attached to the head bone, even though it will not be rigged to it through skinning

What I like to do is just temporarily combine the hair skeleton with the body skeleton and animate. Then when it's good I disconnect it from it and keep it as its own skeleton.

That's probably the best solution, and for simplicity's sake, try to keep every customization down to a single model, attached to whatever bone you want on the character model. Each thing can be rigged/animated separately.

There's also the question of how to animate it-
The mechanim system makes it hard to duplicate the state of animators, and you'll need a lot of animators that are sync'd up. The legacy system is much better suited to this, since you can just play animations by name at the same time on every animator in some list.
With mechanim you may need to have 'uber animations'/controllers placed on every single animated item, and pump the same parameters into everything. I could see doing it that way become a nightmare.

We can also try that, i will search how its done

Im not sure how it will be animated, but my bro understands this part better, i think he will be using mecanim for the main rig and something else for the other parts

First baby steps of learning to program.

But I see Linked Lists, so you're starting with fundamentals at least

Quick! Stop using Eclipse! Anythings better- Switch to IntelliJ or even fucking Netbeans would be an improvement.

New thread when :^)

page 13

God damn it, I've been having so much trouble adding a native method for tracking player deaths in his shitheap game, and he goes and does this.
Too bad you can't use that for anything without microsoft suing you for what's left of your 2 billion simoleons, you swedish fuck.

PROGRESS.

Well, good luck anyway.

some places dont allow webms

Okay? Good thing this isn't one of them.

I usually webm if possible since it's more convenient than stuffing my jewtube channel full of shit nothing-progress, but a 10 minute video in 16mb ends up being mediocre quality, so if I make a bigger update or something, I tend to jewtube it.

Its cute to see user's first steps before the despair hits

Wouldn't it make more sense for the values to be inside the boxes, and the indexes to be outside?
Also STOP USING ECLIPSE NOW AND FIND A BETTER IDE. Trust me. You'll be switching away from it soon enough anyway.

why are you using Java? that language will just impede your thinking

I think he was half shitposting to prove a point

Okay so I have a huge grid of tiles, and I want them to animate based on a common timer. However, depending on their tile type, they animate at different common speeds.

Pic related is the most elegant way I can think of doing it, without having a Timer[] with a non-descriptive index/hash. I can't think of a better way to do it. And no, enums aren't the answer here

why not just use the total elapsed time instead of ticking a bunch of timers that could end up drifting?

something like
int frame = ((int)(Game.time / rate)) % numFrames;
at the site where you need to determine what frame to display, with rate/numFrames being determined on a per-tile basis?

Apologies for the blogpost, but I just need to vent somewhere relevant.
Fucking neckbeards.

I don't have enough autism for this.

What, like getting inaccurate over time? My timers shouldn't do that, since floats are accurate with whole numbers, and they'll automatically update and reset very predictably.

My maps are 16,000 tiles in size (though I'm only drawing about 160 tiles in total each frame). This is how my Sprite class looks (in particular the GetFrame method will accept a GetPercent() value from a timer)
hastebin.com/fopujaxihe.cs

...

Are you modding for GZDoom?

Nah, something that's being held together with a lot more duct tape.

Timers may still drift if you're not properly accumulating and consuming time, since I don't know the implementation of your timer, I have to assume you're not ticking multiple times if needed, and just setting a timeout to zero every time it ticks, like most (even pretty experienced) programmers end up doing.
And even past that, why use a ticking timer for these, when you can calculate the desired frame efficiently, ad-hoc?

Do you have one instance of the sprite per tile, or are they batched to some extent? Can't really tell just from the Sprite class itself- it would be more useful to know how the timers provide frame indexes for the sprites when they are drawn. If you're just using a percentage, you can just use something like

float frame = Game.time * rate % 1.0f;

and completely avoid timers. (Changed from / to * because reasons)

If they're batched, calculating the frame index per individual tile can be avoided, and the system can be made more flexible as well (never fixed to what rates timers provide, or could have custom animation functions like Ping Pong (0-1-2-3-4-3-2-1-repeat) to replace the modulo.)
If they're not batched, you can still find a way to cache that information in a way that is accessible to each tile with a given animation.

Can someone help me with making the game harder GameMaker? It's a SHMUP that runs endlessly until you die and I need it to get harder as time passes, with more enemies spawning, and with harder enemies spawning, but I'm not sure how I'd go about doing this.

As time passes? Easiest ways would be to have a global variable run throughout the game that counts either how many enemies you have killed, or just how long the game itself has gone on for. The former, I think, would be better. It should probably work in waves, of "hey, if you've killed this many enemies, you should start spawning some harder versions of them"

That sounds fucked up. Where is the changelog of it?

Timer is here
hastebin.com/exetafutef.cs

Basically if time>=duration, then time=time%duration, so in the case of multiple ticks or ticking by a large value, it will roll over and set the elapsed value to true for one frame/tick/whatever

Sprite is only used by Actor objects, it didn't make sense to define all that extra weight for 16,000 tiles. Instead, I have a TileType that has a reference to an Image object (contains a texture and its texture coordinates). Then, it's just a matter of looking up the tile's image and drawing it on the fly.

well nigger, time to master calculus and linear algebra.

Is not imposible to learn it, you just need to actually be motivated and do it.

Is like those people who make millions on the asset store from people who are too dumb to learn trigonometry.

yeah, just have a variable called dificulty or levels that take into account how many points you have (score) and time.

After weeks of radio silence, I finally have something to show.

Tomorrow I will probably be back with the terrain generator

This user is right, also linear algebra is like 1000x better than trig, in general you want to try and replace everything you use trig for with linear algebra because the process is so much better

Faster too.
Some languages implement a lookup table for trig functions but they are inherently slow

I guess I need to bite the bullet and start learning.

Realism is gay as fuck, do what works in your game. Wasting time on complex algebra so that a bullet is 1% more realistic is a fucking waste of manpower

you know this to be true

wow, just spend your money on the assets store, there's no need to remake the weel.

kys

Hi guys,

This is with SFML and C++.

I have a base game object class that my other objects inherit from.

I'm running into an issue with player input and command handling. I'm not sure the best way of applying the command queue to the game objects and how I would be able to call specific functions inside those game objects.

I plan on my players being able to essentially control more than one GameObject based on their input. If I hold down control and press wasd it'll move the secondary unit instead of the main character.

So I was thinking of serializing the input with the entity id. Which I like since I want these secondary units to be essentially swappable.

The main issues I'm having are:
How do I call a specific function of a derived function when I only have a reference to the base class?

IE: When I'm handling the input from the command queue and I've entered cntrl + 1. Say my secondary unit is a wizard and his #1 ability is to cast fireball. Or its an elf and he shoots an arrow out in front of him.

How do I make it generic enough that I'm not ending up with a yandev-tier conditional tree for input.

I guess to simplify it down, what is the best way to map input to a function and how do I callback a derived class's functions when I only have a reference to the base class.

Learning linear algebra to add pointless realism isn't going to make your game better

fuck off, come on now

What I am doing is catching WM_CHAR or WM_INPUT on my message pump, not sure what SFML does to handle windows/Xlib messages, but anyway once I get the key pressed from the message, I then have an array of "keybind_t" structs that the key press is tested against in a linear search. The array of keybind_t's is generated by calls to the "bind" console command, usually in the default.cfg file that loads console commands on startup. What "bind" does is it binds a key so that when you press it, an associated command held in its respective keybind_t is immediately processed by the command processor, this is then interpreted into whatever command I associated the keypress with, so if I tell the console "bind x quit" and press X while the console is closed, the game will shut down. Input only happens as an extension of message processing, so I don't call some kind of "input_update" every frame.

You cant even do anything in a 3D game without linear algebra. Also linear algebra != realisim. Just set the rotation of the earth and whatever else to "1" if you don't want to consider it.

lmao unitykid.

I handle the events SFML generates in my games loop and then have to add them to an InputQueue to then be processed by the "world" later on.

Before I had held onto a reference to the specific player object and would just look for the WASD keys being pressed and would just call the move() function which just modified the x y positions.

I kind of want to be able to have my main character "posess" teammates or baddies and move them around and use their skills n shit.

I might be retarded.

I could map keys to specific generic actions
such as GameObject.perform(int action) and pass it the keycode as an int.

And then when the command queue is being processed it would just invoke the action of the specific entity id of action of integer whatever?

I guess I was overthinking it and stuck on invoking the derived objects specific function. Such as if I'm controlling the wizard and i press 1 call Wizard.castFireball().

When I could just have the Wizard Overwride a perform(int action) function and handle it that way instead.

I process input immediately, this is then converted into a console command, and if this is to move the player around, that command then alters the packet flow from the client into the server- the way movement works is that the client is constantly telling the server which movement types are active, in a packet that is sent every frame. Then the server actually interprets the movement and tells the client what the position of the player is.

You can do it that way, but You'll need to make a translation layer that has a set of keybinds (to support rebindable keys) and translates those into actual in game "commands" like casting a fireball. Posession of other character shouldn't be an issue since, the game knows what character to send "commands" too, and thus the input doesn't need to be aware of what character is being controlled, it just spews out commands like "move forward" at whatever is being controlled, which can then process those commands in their own way.

PROGRESS

What's the big deal with linear algebra?
Isn't that like middle school shit?

You don't need any more math than PEMDAS to understand how it works. All of the more useful functions of it for programming involve trigonometry in combination with it though.

I'm going to keep telling you, stop using Eclipse. Better yet, stop using Java.

It's easy enough that middleschoolers could probably do it, but modern people are so dumb that it gets put off until college, as a more advanced course.
Matricies and Vectors and Spaces, oh my.

eclipse is fun because the other IDES are even worse.

also java is fine for 2D shit.
also I'm just learning myself it.

Eclipse is the worst Java IDE out there.
IntelliJ is one of the best, but even Netbeans is preferrable to Eclipse.
Hell, I'd rather use the command line exclusively than ever touch Eclipse again.

...

Sooner or later you will run into a brick wall with eclipse, fighting it to get things to work properly that used to work just fine before something somewhere changed. Switch now, you will save yourself tons of time and headache later.
I speak from experience when I had a pretty large project that I could just no longer get to build successfully in eclipse. I even installed Eclipse fresh on a new computer, and it still refused to build that project properly.
I migrated the project to Netbeans, and it worked 100% of the time, and never had any odd issues since. Eclipse is one of the worst, buggiest pieces of software ever written, heed my warning and switch asap.

i'd make one but i'm pretending to work

I don't give a fuck.

everyone is triggered by some reason by java.

NEW THREAD

Sometimes it's just a good idea to say to a project "good enough" and give it a rest. Step back and give it a rest for a while and when you begin work on a new project you can look back over you're old approach and may be able to look at the problems with a fresh perspective and figure out a better way user.

Because it's system design is a steaming pile of shit. Even LUA is better in comparison.