HolyC is best language, language wars are over now edition

those doorknobs look a little low

Faggot.

It's to hide the fact that you'll be playing a manlet.

I'm working on her wedding dress today. So far about ~4k triangles.

I learned that you should avoid using UE4 C++ at all times whenever possible. Don't be afraid to sound pussy because you aren't using C++ (it's C++ under the hood regardless). Blueprints are there for the reason that you don't have to do all the dirty behind the scenes crap. You should only use them for adding very specific functionality like overriding certain functions. The worst thing you can do is your entire project in C++. But to answer your question: How do I do my component communication? With blueprints.

I dunno yet, honestly.
I'm between projects and lately I've been playing a whole bunch of puzzle-horror RPG Maker games. Some of them are neat. They're rarely actually scary, but there's something…I dunno, comfy about them?
Plus, the RPG Maker community has a whole ton of interesting custom tilesets that people are allowed to use for projects. So I decided to take a couple and put them in first-person and do a basic roam-around-and-interact-with-things system.
I dunno if this'll go anywhere or if it'd even be interesting to play, but I like the idea of a first-person not-Ib-i-swear.

I checked some doors and oops, yup, bit low. I've raised 'em up a bit, thank you.
The character is a young boy so you're not far off.

...

Your statements confuse me, last time i've checked nativization is still much slower than C++, see this video for example: youtu.be/V707r4bkJOY?t=1m45s
I would go completely insane when trying to do the things i do in Blueprints, it's literal spaghetti code.

Have you thought about making a genuine horror game? I feel like there's few things coming out lately that manage to get it right or flesh it out.
The last thing that intrigued me was that game named "Walking" Holla Forums freaked out about a year or two back, if you remember that. It was a very simple game but conceptually I thought it was really cool and clever with how it fucked with the player. I'd link it if I could find it again.

I've never really done horror before, but I'd be interested in giving it a shot. I like trying new things.
What would you say are the obvious things to avoid or to aim for, the Dos and Do Nots?

Terry got a drum set. Then he danced to Hava Nagalia in celebration.
templeos.org/Videos/MAH00228.MP4

How do games like Darkwood and Factorio do their top down 2d graphics? Is it some sort of filter they use on top of regular 3d graphics?

Avoid jumpscares, every game after Amnesia was jumpscare heavy shite.
The idea of being able to fight back, or just in part fend off is a step in the right direction IMO.
If you've ever played condemned, that would be a good example of some decent shooter horror.

Do nots: jump scares. You can have them, but extremely sparingly. Silent Hill 2 had like 3 in the whole game. Most of the time the game seems to set them up but then nothing happens and you're left hanging. The result is a constant feeling of dread, that's your goal.
Do: psychological horror, undefined or out of place visuals, breaking patterns and fucking with expectations. Again, SH is masterful at this. It introduces a lot of patterns and gets you used to things, then suddely changes them up for seemingly no reason but to put you at discomfort. The unkown is the scariest thing in the world, so you have to visualize that somehow. The player has to not know what they're looking at or what will happen next but want to try and figure it out.

I don't consider myself an expert on the topic but that's about the gist of what I think makes truly good horror.

Summary of Walking: You could walk forward and backwards. It was procedural, nothing really ever happened, but the game got more and more glitchy as you went on, there was a figure that sometimes followed you or stayed just far away enough for you to catch glimpses, and occasionally when you quit it flipped the fuck out glitching all over your screen with a horrible noise before closing.
It was mainly unsettling, with the exception of that closing thing that could be seen as a jumpscare.

On the topic of horror, I have an interesting problem myself. I don't get scared by pretty much any situation except social situations IRL but jumpscares "get" me. I do enjoy spooky shit and dislike jumpscares for being cheap, but they're the only thing that actually gets a response of "fear" out of me technically, albeit a short rush of adrenaline because my body thinks it has to defend itself.

Well yeah Blueprints are slower but therein lies your problem, you are worrying about performance. I don't know much about your project but from what I saw it looks like you are the guy making that Knight and Cardboard enemy 3rd person game right? You should not go full C++ because you sacrifice too much time now for some performance gains later. Performance gains that you haven't even marked as actual bottlenecks. So you are right now wasting your time. Blueprint scripting is wonderful for fast development and prototyping, which is what any dev should strive for. All of those animation notifies, rotation cache, states, conditions, and functions can be done in Blueprints. It's not spaghetti code. The video has no real context as to the performance gains (which is the point of using C++), it just screams "C++ is faster, don't use blueprints". Look at it using their example:

This is in a 10,000 loop (you going to have 10,000 main characters?) so really you are arguing spending 8x the dev work in C++ to shave off 0.00345ms for your game with 1 main character. It's a waste of time. If you used blueprints you dev 10x as fast and then refactor it later when needed. I too have had times when I want to do things "the naturally fast way", but It took me 8 years to realize that making shit work well is far more important then how fast it works.

I might be remembering wrong, but the game was always windowed, right? But the "glitch" effect went across your entire screen. I found that really clever for some reason.

Jumpscares are a fine tool, but you run into the problem of getting your player used to it if you have a lot of them. While I like Dead Space, the game stops being frightening really quickly because you know exactly what to expect after the 5th necro man pops out of an airvent with the same shrieking soundeffect playing. You absolutely always want to avoid being predictable.

It's true that jumpscares are generally shit, but I think that's because of how they're handled. See vid related starting at 1:37. I don't fully agree, but it's worth thinking about. There always has to be the possibility of the jumpscare. In my opinion, jumpscares are generally shit because they usually don't do anything scary to earn the jump.

I want to fuck Burnbot.

Yep. Also, do you know Irisu Syndrome? That has an even better twist on windowed screens.

Yeah, jumpscares can be used in smart ways making them "good" at times, but in and of themselves they're cheap. One popular example of a good use I've used before is in silent hill, there's a cat that jumpscares you in the normal world jumping out of a locker. Later, in the silent hill version of that place the locker is shaking, and you're terrified of opening it, yet there's nothing in there except blood splattered all over the locker. Then there's a jumpscare when you walk away.You also hear the cat cry after the initial jumpscare and it spawns a monster where it went, making it likely that the cat got killed for an extra bit of subtle horror.
The thing is, jumpscare fear leaves as quickly as it goes, even when done well. I've made people scared and paranoid for days by playing into their fears and acting spooky before, which is a lot more fun than when they're spooked for 10 seconds when I sneak up behind them.
So do I.

Yes

I would still use C++ even if BPs were just as fast, everything slightly complicated ends up in a gigantic mess, i hate using the system except for prototyping. Im not slower developing in C++, it's not a sacrifice to me.

As for performance, this is a matter of ratio, the 10.000 loop is an arbitrary value, in the vaguest sense you can infer that things that take 1ms in C++ take about 8ms in BPs, that translates to all the logic you create. Anyway that wasn't really the point of my original post, im sure it's just fine to use BPs only for /agdg/-scope projects, i agree that switching to C++ just for performance is dumb.

Can we do demo day on September 17th?

I can't out do the cake. So I'm fucked.

It's not just about ratio. It's about the amount you are applying that ratio. It's not worth the C++ for a small module. You get diminishing marginal returns.
No, it's really good in moderation. It's just that blueprints exist for the inherent slower dev speed of C++.

I might be able to squeeze finish up a super-early prototype build for my game if the deadline was Sept. 17th…

Maybe if it was a monthly thing? Or quarterly (once at the end/beginning of each anime season)?

Wasn't there an user that wanted to set up a list of games in development on the wiki?

...

I was not aware we had a wiki.
If we do, that would probably be the place to put it, otherwise a simple pastebin might suffice.

The old abandoned Holla Forums wiki that is, /agdg/ doesn't have one as far as i know

Does the part from 0:30-0:35 count as a low-key jump scare, or is it something else? When I originally saw it in the theater, because there were no screeching string instruments or loud screams, nobody else reacted when it happened, and I was especially spooked because I myself wasn't sure if I just saw what I thought I saw. Out of the whole movie, it's the only part that's stuck with me these years later.

posting terry OC

What about a 2-week long h8chan game jam every 4th month?

I think it could be a good idea to have a wiki of indev games. Guilt can be a good motivator to finish your stuff if you know people haven't forgotten about it, and are silently judging you for being a nodev. Apart from the fact that it would be great to see what everyone is working on.

This does sound like a good idea, there seems to be a halfchan one here: implyingrigged.info/wiki//agdg/

But, I can try registering one on some kind of wiki site.

Yeah, that sort of playing with expectations is great. Though if things like that happen too often, you sometimes wind up trying to figure out if you're being played, and start looking for less likely places for a scare to actually come from, which is fun, but probably not what the creators want.
I think that depends on the person more than the material. Blog: I actually can't consume "good" horror more than a bite at a time, because I can wind up stuck in that state for months, and desensitizing myself by binging it isn't going to happen. I've tried.

Why is its head so flat? Am I not seeing it right? I wouldn't call that a jumpscare, because it never jumps. If anything, it's more like an easter-egg that makes the scene scarier if you notice it.

It's a hat.

Why is its hat so flat?

Anyone experienced in Unity as to say what prevents every other developer from optimizing their game?

its babys first game engine? the bar is low. the developers are shit.

I want to give it a try, the only downside I have heard is how it is not very mod-friendly just in case my game gets popular enough as to get mods.

I'm assuming that, in order to become truly skilled at modeling, I'll have to learn to draw for myself, because I can't assume that I'll have an artist to give me T-poses all the time. But how immediately critical is that step?

My end goal is animation. I love animation, and I find modeling mildly enjoyable, but for some reason just don't give two shits about drawing. I've tried and failed to learn several times, so I know that on some level I'm just avoiding it because I think I'll fail again. But am I overestimating its worth to begin with?

Not really.
You will need to learn about stuff like proper anatomy and knowing the things that make a drawing good may help,but the addition of a dimension changes things enough that a direct relation between skillsets is not cause for concern.

Retards who don't bother with the api or are making their first game and so are just spaghetti coding gameplay features into barely functioning.

Unreal is much friendlier towards mods. ARK's mod tools are front-page in the Epic Games launcher, even.

the modding is a joke for ark and friends 90% of it is just comedic shit. and some 9.5% number "rebalances" and god items.

anywho building your game with the idea of it might being popular enough that people would mod it it stupid. and sounds more like an excuse for not just making a game.

I mentioned it because my project is prone to getting nude mods and all that stuff, even if I make some myself there will always be someone who wants to make their own.

I see. Then, with your advice in mind, I think the answer may be to simply practice drawing enough to be able to draw things which are adequately understood, since most of what I want is to be able to express character concepts. That's a much lower bar, but there's always the option to change my mind and decide to shoot higher when I get there and already have the building blocks for further progression conquered and at my disposal. That way, even if I decide not to, I should have the shared skills for modeling by then.

True, mods are a strictly post-development objective. Having mods as part of your development plan messes with your development pipeline and is generally best saved until you're mostly done with your game.

Factorio guys do a 3D render and then adjust it in photoshop.

factorio.com/blog/post/fff-194

I don't think a game jam is a good idea. Jams require a certain level of experience that it's easy to think you have when you don't. There's constantly game jams going in itch.io anyway.

Recap type shit in halfchan /agdg/ used to lead to 80% of participants ONLY participating in the recap. The threads would be nonsense shit 6/7 days of the week, and then one day every week people would post progress and fuck off for the rest of it.
It's part of what made me stop going there. Would not recommend.

just like release a version of you game that is a porno

A lot of stuff to think about. Thanks, anons.
I guess I should start thinking of setting and spoops and things like that, but I figure it'd probably be better to get core gameplay and systems done first.
So I did up lighting, flashlights, rain, ambient sound effects, smooth player movement, and etc.

I'm actually a big weeb that loves J-horror and K-horror, so I guess if this pans through it'll probably be pretty heavily influenced by that.

Is your character supposed to be a child ?

Added some animations for turning around. Still tweaking it. Also starting the bully. If you don't kiss her within some time she starts crying. I did the animation really quickly so it's very stiff right now.

Yeah, he'll probably be a young boy.
It's a cliché, but big environments are easier to move around in without getting caught on things.

AUDIO BALANCING MOTHER FUCKER, DO YOU SPEAK IT!!

shouldn't you flashlight be pixelated?

Probably, but I'm a clueless bint and dunno how.
Making textures and sprites pixelated is easy, dynamic lights…not so much.

Lol I forget my speaker is set to low.

Uh… what? your grammar and my internal translator just collided.

I want to make a video game, but I am retarded. Help. Is there some guide I can follow somewhere that will cover the fundamentals? Is there some book that covers the basics of programming specifically for video games? Is it really as easy as just grabbing a game engine and fucking around until something happens?

Messing with an engine can help you a little, but you need to find the path of least resistance first, that is, what part of development seems to be your thing, art, coding, planning, etc.

No. Unless you want to drag and drop shit from the Unity asset store and become unity shitdev #563847384534, you'll need to be able to come up with some original code.

Just look up Unity tutorials on youtube, follow one of those and you should be able to get a decent idea about how it works. You may need to look up some kind of tutorials on C# or coding basics separately though, I doubt you can learn that from unity tutorials.

If you want to make a game without an engine, you need to really love programming, but it's totally doable as long as you're not looking to make the next Skyrim.

learning to draw is supper easy, you just need to know how.
you can go from literal stick figures to "just enough to get my ideas across" in about 3~4 hours, then learn to draw proper in about two weeks and then to photorealism in about 8 months ~ 1.5 years. to put it in perspective, it takes more time to get this good at dmc4
youtube.com/watch?v=vvndA87ZLK8
but you can earn money from drawing

start with "Drawing on the Right Side of the Brain", skip the filler and just do the exercises, once you finish it (or get bored with it) you will notice you can do excellent drawings from pictures but can no longer do it from imagination, then change books to "you can draw in 30 days" by mark something, of course you are going to marathon this one too. once you get done with these you will be the average drawfag, now is time to pick an specialty and stick with it, you said you want to animate, unless some other user wants to step in you are on your own there.

is very critical that you start it before going 3d or you will learn wrong. however is also possible to learn both at the same time and might even be better as they reinforce each other

once you know how to draw, there are a few options that open to you:

Still alive

After a bit of a productivity crisis, I'm back at it.
If anyone is looking for a good looking outline shader in Unity, here's the best I could find.
willweissman.wordpress.com/tutorials/shaders/unity-shaderlab-object-outlines/

I modified the code a bit, you'll need to do this for best results.

Pretty good. I agree with as well. Matter of fact, I think messing with the overall look might be a good idea. Pixel art in 3D is one thing, but with your current lighting it reminds me too much of Minecraft.
The game itself might not look very interesting, but something like the jaggies in imscared might be something you could try, or a dithering shader like Fight Knight. The Silent Hill games tend to have a slight noise effect on the whole screen as well. There's a lot of stuff you could do to set up an unsettling atmosphere.

Depends on how retarded you are. There are tools to make games that require no coding, things like RPG Maker. Maybe find one of those that suits your needs. That said coding is pretty fun and is less about being "smart" and more about learning and understanding a language (though being smart helps with figuring out the logic of code and how you want to implement it).

Also another thought is you could make a mod for a game you like as a starting point to see if you even like developing. Some mods for games offer up whole new stories and worlds so creating an original game in a mod isn't out of the question (just don't expect to be able to sell it later unless you're willing to get a coder or license the engine, honestly though you'll end up doing early work that's free regardless of approach that you might not release, at least with a mod you might have something to show for it).

Using Unity.

How would I go about creating a visualization of my character's jump arc in 3d?
Mathematically it's most likely just about speed/jump force and gravity, but how do I make it into an arc that I can visualize?

This is mostly for debug purposes, I'm planning to use the landing point of the arc as the point in which I'm casting a sphere cast to

You create some sort of clone of your main object that only collides with geometry and simulate its jump. You should have a function somewhere already that takes the position of an object at the start of an update and returns its position at the end of an update. Apply the proper jumping momentum to your clone, then call that function with your clone object and some fixed timestep (probably 1/60) over and over until the clone object has landed again, or too much time has passed. Record its position after each movement step. Then you take that array of positions, draw line segments between the positions of successive frames and there you go. Jump arc.
This is probably badly explained, but the idea is just that you simulate a jump and make the arc from the positions your simulated object would be at after x timesteps.

This seems like it could work, no idea where to find a function that returns the position the object is going to be at by the end of update.

I'll keep searching the docs for it, but in case someone knows please post it

Agdg get

a sort of ribbon would be best for that. Made with either particles or a trail renderer.

In all honesty, as long as you know the flight path in space (likely expressed in a series of Vector3 points), it should be easy to use either of those to achieve the effect you want.

I'm unsure as to what this means, can you elaborate?

Dunno if you need to use an explicit, inbuilt function for it. Just look at how you move the player in your update step, wrap it into a function that takes starting position and returns end position assuming a fixed timestep, and call it until the return position is on solid ground.

This is how grenade arcs are visualized in vidya. I imagine using the same method for jumping would work (a'la Gunpoint).

You can recreate this effect in Unity by using a trail renderer or a series of particles.

I can probably cross my movement vector with my jump vector and then multiply the speed of my rigidbody I guess?
But how do I launch a function only at set intervals of frames?

This seems like it would be much easier but correct me if I'm wrong, it seems to be a visual thing only.

You misunderstand. Don't call the function at set amount of frames. During ONE update step - when you need the arc visualized - you simulate ALL of the movement you need until the arc is completed. Say the jump would take you 3 real seconds or so. You don't want to wait 3 seconds until that jump arc is really complete, instead you calculate - during one update - where would the object be after 1/60th of a second, based on that position and the changed movement, where would the object be after 2/60th of a second, and so on until the object would hit the ground again. Then you have a lot of positions you can make a jump arc from. You can probably choose a slightly less fine timestep, but the method still stands.

you'll have to calculate the trajectory for the arc to be displayed correctly. Something like this dude made in Unity (something I'll likely be using as well)

forum.unity3d.com/threads/projectile-prediction-line.143636/

there's also this thread
forum.unity3d.com/threads/generating-dynamic-parabola.211681/

I wanna kill CIANiggers with him.

You know, that might not be a bad central mechanic for an first person platformer. Replace awkward distance judgements with mouseclick accuracy just like an fps. Then you could have fast-paced, difficult platforming that couldn't be done in any other kind of system.

keep in mind that this would remove any air control during jump, which is something hard to get used to

That's actually pretty clever, user.
I'd love a Turok styled platformer FPS with this mechanic

I'm having trouble finding the Vector3 for the direction of my jump. I'm crossing my Jump Vector (0, value, 0) with my Movement Vector (x, y, z), then crossing the result of that again with my Jump vector, but I'm only getting a vector that goes straight forwards.

How the fuck do I get the vector that forms inbetween my Movement vector and my jump Vector?

You add movement and jump, you don't cross them. Crossing them just gives you a vector to the right with a magnitude based on how close jump and movement are, and crossing that with jump just gives you a vector forwards with a magnitude based on how close that right vector and the jump one are.

Nevermind, I forgot rigidbody.velocity existed I'm a dummy

Keep in mind that unity's physics engine has a drag value, as well as friction for materials. Both will modify the flight path you'll be calculating.

Adding them just gives me a Vector that goes up

Oh I know, but I'll consider that once I get there, cause if I do now I'll never get anywhere

okay, rest of the vector issue then.

do you have it visualized? post screens. Your Y value can be too high.

Ah fuck, my Y value is extremely high, you're right

Can't really change it, if I do the jump will get floaty

are you using Rigidbody.addforce and ForceMode.Impulse?

Also, when I made my own rigidbody movement, I took out my old physics textbook and re-learned newtonian physics.

Just apply the formulas from there into your code.

Yes

Also I don't own any of my old physics text books anymore

well, your X and Z vectors are velocity while Y is force.
Force = mass * acceleration

Pretty much this:

I made my graduation project (non-vidya) 2 years ago on unity and now that I look back at it the fucking monster spaghetti would make the flying spaghetti monster church think it's a second coming of their lord. Unity is good because you can easily make it work, so when you are short on time you go for easy instead of optimal

Okay listen, I have no idea what to search to find this, but how would I calculate this arc by using my position, velocity, mass and gravity?

What is the actual formula for this?

You know what, unless some other user will give you the answer I'll write you a summary later today, after I get back from work and grab a beer, okay?

I guess I'll treat it as training for my own implementation of this. Pls be patient.

don't expect to be spoonfed code, though

I mean I'd really just take how the formula is called or where to find it, that's all I need

But oke

Does anyone feels extra retarded sometimes when programming ?

I feel extra retarded every time I think about just making game

Well, okay. The umbrella term would be "ballistics", understood as simple projectile physics in classical mechanics. Your character is a projectile after jump, after all.
Have fun and good luck.

Time to go home.

Cool, I can try to work with this

g = gravity
v = velocity
p = position
x(t) = v.x * t + x(0)
y(t) = (0.5)gt² + v.y * t + y(0)
This does not include drag. It's also 1st year highschool physics, FYI. Also, mass is useless if you already have velocity. You use mass and acceleration to calculate velocity.

I'll be honest I don't remember anything about high school physics

What are X and Y supposed to be in that formula?

There's two ways I can think of, one easy, the other hard.

Quantized light. In your fragment shader, where you compute the illumination value you add a step where you quantize the light to a set of discrete values. This means that instead of getting smooth, continuous lighting you'll get concentric circles with sharp transitions. I think this is called toon shading.

If you instead want the light to align with the texels of the world geometry, things are more difficult. You could maybe use the UV coordinates to figure out where on a surface you are, then compute the light from the texel center instead of the fragment position. That would yield per-texel lighting that looks like a lightmap, except real-time.

Speaking of lightmaps, this is totally insane, but you could do real-time lightmapping. Basically, when the light moves, recompute lightmaps for all surfaces affected by the light and reupload to GPU. Since you're working at such low texture resolutions, you might actually be able to do this in real-time at 60fps, even on modest hardware. Another option is something something compute shaders.

I remember this girl in my high school physics class had like FF tits when she was 14.

My dear good sir, if you are worried about people desiring that your game be pornographic (containing Lewd material) then simply provide said degenerates what they desire in perhaps a different release of thy game.

For simplicity's sake I kept it 2D, so x(t) and y(t) would be the position you are at at time t. V.x is the horizontal velocity, so basically your movement vector assuming you move at the same speed the whole time. V.y is your vertical displacement, which is not your jump vector but your actual vertical velocity at that time.

I think your doorways are too tall and close to the ceiling.

I'm making a stealth game in the style of Thief in Godot, and I'm trying to find a way to determine the light level where the player is. Anyone have any idea how to do this?

Draw multiple circles where the light is supposed to be and depending in which circle the player is in and whether he's crouching or not your light level percentage gets higher ?
Also send pics.

What if the player is standing in natural light, like the moon or the sun?

More circles, I have no better idea sorry, I've never done that before.

What I would do is probably start a raycast from every light to the player's current position, and if any of them manage to reach the player without being interrupted by something else then the player is in light, if not they're in darkness.

No idea how to do anything that isn't true or false though

find out the way Godot light atenuation works.

add multiple "sensors" to character's hitboxes

when the character steps within the light's radius cast rays from the light to the sensors to check whether the light actually reaches the character.

if the "sensor" is in range, start calculating the light value (atenuation * intensity) for each light source

Average the values from all the sensors

Sounds good, I think I can leverage Godot's instancing to make this simpler:
Light (w/ shared script)
RayCast
CollisionShape

I dunno, this is where I managed to get by hacking together some code.
Seems almost fine from the side but when looking at it from the front it's clear it's absolutely fucked.

I think I might be too dumb to get this to work, I don't even know where to start

In the first 10 years I studied modeling, I would agree with this. Then I started drawing for fun and my 3D modeling improved A LOT.

While you can do good 3D modeling without it, the improvement it gives you is so great, I would say it is a requirement.

Is this a projection/prediction of the character's motion?

It should eventually be an arc that the character should follow while jumping.
I'm hoping to use it to check the ground where the player is supposed to land but at this point I doubt I'll be able to get anywhere with this

Hundreds of Worms ripoffs say its possible.

Yeah in 2D

don't get too discouraged too easily. In all honesty, I'd be surprised if I could do it in a day.

Save/Load done
(popups and tooltips do not show in the vid)

Is any Receiver extended clone in development somewhere? Love the gun mechanics, not so much the drones and levels.

the everything 1 hit kill

any tutorial book in particular that helped you get into drawing?

To some extent Andrew Loomis books, but I never followed them strictly, more like adapted the techniques in it to help my own shortcomings.
And a lot of stuff I picked up from other people's sketch sheets/pads, some drawing sites have sections just for practice sketches and stuff which have a lot of hints, helps you to figure out how to simplify and break down complex shapes.

Conceptart.org forum has a lot of this stuff, example:
conceptart.org/forums/showthread.php/325846-Mustacheberry-s-study-attempt

I'll take a look at how Imscared works, thanks.

I'm not good at programming at all, so yes, I feel extremely retarded every time it comes to programming. Especially when it comes to calculation-based stuff. I'm not really good at anything at all except for mixing sounds together.
But, hey, as long as something fun comes out, then that's good. Making game isn't a competition about "who does it the best", but if the player has fun with it. And if it's fun, then you can always improve things later.
But don't use this as an excuse to make really shitty code that bogs the game down.

I'll save this in my To Do list and see if I can edit the lights a bit more with that. Thanks.
Lightmapping would be especially interesting and would open up an opportunity for shit around a corridor–you can see light going into a hallway but something might be there. I know a couple other guys are experimenting with lightmapping right now, I'll probably check out how they're doing it.

Finished her wedding dress. I went more for a quinceaniera dress because loli.

Made LOD for the grass

...

Oh shit I recognized that map layout. I totally forgot you existed. Looks good.

upper half of the dress looks too, thin. Like textures applied right to the model.
Not sure if that was your intention, but it would look a lot more cute if it didn't hug to tight.

Thanks.
Recognize it before the texturing?

Yeah I'll add one more layer of frill above the top one.

...

Remove blur, tone down Ambient Occlusion, remove COD tier bloom, add a sky dome light so the shadows aren't pitch black.

>Have trouble learning to code in a self-taught manner because I am a sperg who probably also has ADD
What do I do, Holla Forums?

Do computer science instead.

They are computer science degrees. They just have an emphasis on game and software development. Most of the schools I looked at would teach Java, .NET, C# and C++(and other languages based on the school, but that all shared these languages) and would cover developing normal software as well as video games. Data structs, linear algebra, all that stuff.

Try online courses for learning Pre-calculus and/or look into financial aid for college. Pretty sure a "game dev school" is a bad investment, but pre-calc is useful regardless.

Yeah if you're going to take precalc see if they offer CS.

Enginefags: To treat effects as entities or not to treat effects as entities?

Seriously user?

What the fuck do you think a gamedev classes are going to teach you? Nothing. They're there to scam gullible kids who don't understand how the world works, don't do their research and haven't realized school is a joke. It's in the same league as gender studies. Every degree is watered down so any dipshit can pass, this goes for triple for "fun" degrees.

Whats your plan when you graduate? You're going to have a ton of debt and a VERY hard time finding a job that isn't flipping burgers. I don't even know if non-indie studios hire game design degrees. I'm pretty sure they just go straight for CS, art, etc. On top of it you probably would have just only barely scratched the surface of all the topics you need to work independently. At least if you did computer science you would have a very strong skill for game design and could probably make enough money to contract the other stuff. Even then, with computer science you just barely scratch the surface of most topics. It's only there to get you a job at companies the school is connected with.

That depends on what the effect is.
If the effect spawns and lingers in any form, then yet, it needs an entity of some sort behind it.
If the effect is simply an instant effect being applied or a buff that falls within whatever buff system you have implemented, then no, it's just a piece of code that runs once and is done, and it doesn't need it's own entity.

Most of the custom level editor functionality has been implemented. It's happening!

You're back!

I treat a single particle emitter as a single entity because I batch particles of the same type to a single draw call.

Programming is one of the best skills to learn on your own, all good programmers have learned it on their own. Even if you take CS they won't teach you a lot about it beyond the absolute basics, you're expected to study and practice it in your free time on the side.

There's not a lot of difference between self-taught and university except for guidance, better learning materials and peers for motivation. University isn't school, if you read documentation or some paper on a topic and then fail to process that information and learn or apply it, you won't go far in university.

Stop diagnosing yourself with bullshit nu-age illnesses and better yourself user, spend less time on the internet to recover your attention span, read books or good articles, stop watching videos or only watch calm, longform explanatory videos without skipping. Get off your chair ever so often, take a walk without listening to music, do sports.

If they are, why don't you take regular CS and spare yourself the shame? If they aren't, don't take them. Either way post the curriculum.

Thanks user, got any idea about seriously low-tech effects in UE4 like limiting the color palette or lighting resolution? If possible without post processing shaders.

...

My Gondola Can't Be This Nightmarish

looks pretty cool dude

New Speebot demo: drive.google.com/uc?export=download&id=0BzIMurAb1QxDdUNPbERtYTBVRG8
Dedicated thred:

thanks

holy shit speebot is already 1.5 years old? time goes by fast.

I'm more amazed we somehow didn't drive speebot dev away

Well Speebot dev never gave us reason to. He just made a game and was open to criticism. Nearly all of his posts were progress posts. He wasn't a huge autist like cuckdev and Tetrachrome Dev

free demos help

Yeah I dunno, I can't figure out how to get past this

I asked an engineer friend that's good at physics and math for help and he'll probably give me a hand during the next week.
Still fucking pissed though

Looks like shit tbh

Thanks to the anons who helped me figure this out.

oooooh, that's smart
Didn't even think to use the distance from the hit to determine the intensity of the light
nice work

Good job m8.

why does nothing i make ever get finished

Learn to do one thing everyday. As small as that thing may be it'll help you take the habit of getting shit done.

It's the distance to the attenuation value (d^a).

thanks

Because you set your goals above your abilities.

I'd rather get the shadowmap to know the shadow level. Don't know how to do that, only that it's possible

You should always set your goals above your abilities, that's the only way to get better

Actually that's not right. It's a value between 0 and 1, where 0 is the max range of the light and 1 is the same position of the light, taken to the power of the attenuation value. ( ((max_range - dist) / max_range) ^ attenuation )

Godot's documentation doesn't have anything about how to access them.

how hard would it be to make a ark knock off? as in some open world with other players shenanigans?

want to make a mad max like game where players run/raid caravans from settlement to settlement. collecting scrap gas food and water. (produced at different settlements necessitating trade)

yeah. in favor of a better language (like rust) not fucking blueprints.

since you want multiplayer and probably a large map: very hard.

well a large empty mostly flat map

Starting with an open world multiplayer game is a great idea, those are the easiest games to make

Should take be about 19 min

You didn't even take the the time to check what the video was about did you

We have a salty ideafag accusing Speebotdev of copying his super original idea and code theft over at >>13342563

This is why you shouldn't leave agdg threads
This is literally the last bastion of quality threads on Holla Forums

Calm down, he's getting bullied to hell and back

I finally have a steam page up now. All there is left to do is wait for the release date and fix bugs I missed and shill

I don't think I've ever seen your project

I'm usually on the discord of the original AGDG from halfchan, but rarely I'll post here or in the "other" thread, usually because I never got much reaction here

Oh that's too bad, would love to see it in action if you have any webms

yeah, that's the way you do it! good luck with future developments, my man!
we need more light/shadow stealth games

don't equate the two. I owe many improvements in my code to Tetradev. I wouldn't expect or want any code help from yandy.
And all of tetra's posts were either progress, constructive criticism and good advice until his unfortunate venture into the Nier thread.

F yea. Great stuff. I think I am one of the people who prefer top-down shadows

I have some badly converted mp4 to webms. I can't get the settings right to retain the quality

Here's one of the Hell bonus levels

You can post webms here as long as they're encoded in H264

>selling your game at all
Congrats on becoming yesdev, but please vacate the premises.

I meant mp4s

I am absolutely certain that was bait. Why anyone is falling for it, who knows.

...

My heart seems to be utterly broken at this point, leaving me unable to continue working on my game for now. I still hope to pull myself back together, but in the meantime let me at least warn you all, so that you might avoid what I've had to go through.

I have not been stumped by programming bugs, nor am I lacking in artistic skill or experience in music composition.
I have not been limited or rushed by time or by money.
What has ruined me, wasted mountains of my time, and ultimately black-pilled me, are people.
Nasty, terrible people. The worst of this world.

If you try to do something that is difficult, that is risky, there are certain people who, knowing nothing of you or what you are capable of, yet believing they are acting in your best interest, will tell you day-in, day-out to stop, to give up, to do something, anything else other than the one thing you are doing right now, the one thing you actually want to do.
These people are the absolute worst scum that walks this earth- they will not stop at just nay-saying, they will dedicate their meaningless lives to ruining yours, intentionally wasting your time to run down the clock, trying to have you thrown out of your residence (if they think they can do it), threatening you with violence, etc. to try and ensure that you will give up on your game dev and go do something else.
YOU MUST REMOVE THESE NAY-SAYERS FROM YOUR LIFE, OR THEY WILL KILL YOU AND YOUR GAME DEV.
You must get away from these people, and you must do it swiftly. With this particular kind of person, you cannot give in to kindness, you must not feel sorry for them no matter how much their feelings will be hurt or how pitiful a being they are or how they are maimed or how they rely on you, you must not give them any time in hopes that they will change their nay-saying ways- they will not change, as nay-saying is far too easy, and changing, learning, or actually supporting you far too difficult.

Game dev is one of the hardest things to do in this world.
Nay-saying is one of the easiest things to do in this world.
Tell the nay-sayers to go kill themselves, walk out on them, secure yourself a location where you have control over who you interact with, slam the door in their face and lock it.

Unless you physically had people coming to your house that you knew to stop you from developing because they didn't like your game this whole post sounds like the ramblings of someone with severe problems.

Is your mom mean to you again?

Telling people to go kill themselves because they don't believe you can make a video game is crazy.

go home phill fish

I guess this is the normal response when you can't even fathom the sort of insane circumstances that would lead to my initial rant.

I'm referring to multiple people in my life-
my mother, who went crazy from a change in depression medication and divorced when I was around 16-17 (but not before going on a trip to have sex with someone she met online),
my brother, who has punched holes in walls and broken down doors when I have not done as he demanded,
my stepmother, tried to have me thrown out by my own father when I asked for a single year of self-study instead of four years of college
(this ultimately netted me a really good job at a big game company, far beyond that of your typical entry-level job, so don't try to tell me skipping college was a retarded thing to do. I went to college for one semester anyways at my father's recommendation, skipped several years to highest level of Japanese they taught there, didn't learn a fucking thing. In programming, I didn't bother to skip years, but could have. First end-of-semester project is group project, one guy is hospitalized from drinking too much, other guy couldn't program his way out of a wet paper bag. I did the whole assignment then left the college. All of my Japanese and 3D programming is completely self-taught, school never taught me shit about any of these things and never was going to teach me shit based on what what I saw they had planned for me for four years.) ,
and my grandmother, who I went out of my way to try and help after leaving that big game company, only for her to constantly bitch at and ridicule me (feminism taught her that this would get her whatever she wanted) while thoroughly using me to save herself dimes at my expense of my time, EVEN THOUGH my mere presence
at her residence was saving her thousands and thousands of dollars (without me there she would have no choice but to hire maids to help, which she was doing before I came and is doing now that I have left).

It's not hard to see how this crew of irrational people might prove to also be irrational when it comes to things like education and game dev.
I have both the time and money to make this game, I already landed myself a good job for a good while on my self-taught skills.
Yet they have all told me to go back to college incessantly over a long period of time, and they have all told me to do something other than game dev, incessantly over a long period of time.
They're all fucking retarded at best, violent at worst, and for all the grief they have given me I really would love nothing more than if they all would just off themselves.

...

I like fucking with armatures.

...

I think my game is almost ready for early access/beta, though I did no marketing and got 0 followers. What should I do?

To finish giving context, I'm away from all of these people now, in an apartment with both time and money to work on my game.
But all of the bullshit these people have given me seems to have finally broke me- I am completely and thoroughly black-pilled, I'm tired and weak all day long and can't get shit done.

Thus, my initial rant.

Idk, but something with those screenshots because they look pretty awesome although the hud sticks out like a soar thumb

...

he could've at least tried to wait until the overwatch faggots got tired of playing it

what is your game about

...

Also,

Release a demo. Sink or swim, baby.

Idk if this is bait or trying to point out how blatantly corrupt youtube reviewers are.

So I should take no HUD pics for maximum ａｅｓｔｈｅｔｉｃｓ?

Even with a demo still need to market it so people will learn it exists.

A soft launch on itch.io + trying to contact small youtubers for reviews is a good start?
I want to get some feedback before attempting
Steam and to get on bigger YT channels.

Either that or update the hud if you can imo.

I could have sworn there was an entire subreddit dedicated to marketing your indie game to let's play faggots. I know that the gamedev subreddit is generally helpful if you make a post asking about it.

>gamedev subreddit is generally helpful
It's not quality help though

Hey guys, indie dev here. I'm making an Earthbound-inspired RPG in an SNES/n64 style. What filters should I use to make the game look more retro-y? Are scan lines a good idea? How do you get the proper chromatic aberration?

Shaders
shadertoy.com/view/XsjSzR

A tiger does not care for the opinion of sheep
Seems like you are a sheep, if you are independent and hate them so much, cut all contact.

I don't have any ideas right now for a better design. Any suggestions/tips?

Found this: reddit.com/r/letsplay/comments/3z3fr6/where_to_find_games_to_play_how_to_get_them_for/

I'd have to agree. Most of the time it's just lengthy posts about the kind of work philosophy and discipline you need to become an independent game dev. Which is good for beginners, but there's very little on there about actually learning for newcomers and posts about actual code help seem to be a once-a-day occurrence.

...

The best advice I can give really is to look at examples of dark games like Thief, Blood, etc and just to see how they do it. Right now what you have kinda reminds me of those custom gmod huds. It's just really clean and simplistic, which contrasts a lot with the rotten corpses with glowing eyes.

...

Even if it wasn't reddit I'd still be extremely skeptical about massive gamedev communities. It has the same problem as playing guitar. A bunch of retards get into it because they think it's cool. Then they do shit like, watch extra credits, and regurgitate whatever shit they hear without actually thinking about it.

This is why I was so critical on game design classes. Most of the people going to those classes don't even consider what they'll get out of it, and schools in general are extremely mediocre at actually teaching anything but can get away with it. I remember taking a programming class that overlapped with gamedev. The kids there would argue about semantics constantly to sound like they know what they're talking about when they fucking don't. I remember one dude was like "Minecraft is a good sandbox but it's not a good game", like, Ok I guess, but who the fuck cares.

This is a "too little, too late" sort of situation, where I'm in a good situation now but the damage from the previous series of bullshit is so great that I can't do much until I've recovered.
I'm free of these people now, but I made the mistake of trying to help out my grandmother until about a month or so ago, and tried to keep good relations with most of these other people for the past several years.
Thanks to that about half a year of dev time just went down the drain, and while I still have time and money, and have cut all contact with them, I'm just dead-tired from it all and am losing even more time as a result.

I definitely fucked up, I should have been more cold-hearted, my kindness allowed these people to make me absolutely miserable, and I should have cut ties with all of them a long time ago.

Fighting game first (hoping to do a release by end-of-year 2018 at the very latest), then there's a really interesting concept for a 3D adventure title I want to work on.
I'm hoping to work towards scaling up the amount of artwork I can do, like Ryuukishi07 did going from Higurashi (7~8 characters) to Umineko (~40+ characters).
I don't know if I can scale up to 40 characters, since 3D character modeling is a lot more work than just character illustration, but 40 character sprites for a visual novel is pretty crazy too, if one human can pull that off then perhaps anything is possible.

what sort of fighting game a street fighter knock off or more of a smash bro casual fest?

Game engine is ready as soon as I pull myself together for prototypes, but prototype development hasn't begun quite just yet.
Depending on the result of that prototyping, it might end up like something not quite like either of those games.
But I am hoping to make it fast-paced and responsive, which is probably closer to Smash (Melee) than Street Fighter.

I hate retards like that who will constantly point out flaws in things but never provide solutions. "Minecraft is a good sandbox but it's not a good game." Please, tell me how you'd make it so that it's also a good game, oh holy armchair critic.
It's like, congratulations fuck-face. You've made a simple observation, but are you capable of the deeper analysis required to figure out how to solve the problem? No? Then fuck off

You're giving the incorrect implication it's somehow our fault for driving someone off rather than their own for acting like a gigantic faggot

That's not the implication. The implication is that we only drive out gigantic faggots and that it is their fault they got driven away.

...

meant to reply to

If you learn anything in one day in an AGDG thread, is that AGDG is fucking nothing but hipster programmers. Java this, Ruby that, C#, C is god… But the moment you want to use C++, OH FUCK C++, C++ IS THE WORST FUCKING THING EVER, HOLY FUCK YOU ARE COMMITTING GRAVE SIN EVEN CONSIDERING C++ AS THE RIGHT TOOL FOR THE JOB!!

And they have all kinds of reasons to argue with it, and it all comes down to .. they don't like the syntax. "It doesn't look as pretty to me."

Tell anyone who tells you not to use C++ to go fuck themselves.

"YOUR BIGGEST PROBLEM IS YOU'RE WORRYING ABOUT PERFORMANCE IN YOUR PERFORMANCE INTENSIVE APPLICATION"

No kidding, bud.

You realize C++ = C + OOP right? Also you know that OOP introduces branching, so if you mindlessly spam objects in the wrong places you rape your performance. There's a time and a place for each tool.

Low level network guy here, peeking in. C++ is often used as C with RAII, not using OOP. Used well, C++ can eliminate a lot of the error checking spam in C code that is prone to error because it's so hard to test without introducing bad ideas and the slows from OOP languages.

Is this from RAII or something in the C++ language like exceptions? I also have been using C++ but avoiding OOP, because of branching. Although, my error checking comes from assertion spam, so I might be missing something huge.

nito

Would messing around with the Unreal Editor for UT2004 be a good way to dabble in game dev?

Look into things like std::unique_ptr for examples. The goal is to reduce the 'goto out' spam you'll see in good C code. It's boilerplate, voluminous, and hard to test and winds up being the source of most errors.

Wouldn't using a tool like valgrind be able to handle things like dangling pointers? I guess it doesn't solver accidentally overwriting a pointer though

Reposting semi-fresh Terry OC.

Did he make a new channel/stream yet

Ignoring what it can and can't check for, we're talking about code that is only invoked in error conditions. Many of them you'll never see on your valgrind box unless you spend far more time writing tests than you do writing code. And in C, in many domains (e.g. networking), that error handling might be 50% of the code you write. So cutting code by a half and massively reducing the number of bugs as well with no performance overhead is seen as a good thing. There's stuff like cleanup in gcc/clang but I'd not recommend using it.

ReactOS is a powerlevel?

They deleted speedbot user's thread

random posts are missing, just a moment ago i changed from phone to pc to answer a faggot and couldn't see new messages past

Yeah it was just the board shitting itself. False alarm.

first of all you sound like an edgy faggot, not only that but your previous posts highlight how much of a spineless bitch you are, you are so fucking spineless that they could make you an anime protagonist, actually someone already did, is called tokyo ghoul and is shit, if you still insist on watching it skip to the last episode. That is right faggot, your real life character growth driven by real life events gets outdone by character growth driven by shit writing in less than 30 minutes. you went through all that shit and still learned not a thing from it and the one thing you think you did learn is wrong.
1. coldness is not strength, cold people can take difficult decisions and pushing people away is not hard, is cynicism, instead of admitting that the problem lies in yourself you closed up on people instead, worst of all once you stop being mad at them you will either return to your old self or (ab)use people just like they do now (now you realize why your family is full jerks? that kind of people breed that kind of people)
2. you are mistaking kindness for weakness, weak people operate on impulses, kind people do on empathy, you don't help people on what they may think about you but rather on what you think of them. people gravitate towards kind people not because you can use them but because they can understand you, you should always aim to be kinder.
3. i'm pretty sure that if you showed signs of being miserable people tried to help but you pushed them away with "is nothing, i'm fine" as to not to bother them, BREAKING NEWS, people want to be bothered. just like you, people will go to great lengths for you to care about them. they want to be part of your life your problems and your secrets as much as they want you to be part of theirs. but unlike you, people stop caring about you as soon as they see you not care about them, which happens when you stop telling them things, let them out of the loop or putting them on hold (not saying that you should stop doing these things, just letting you know what you are loosing by doing them)

this wall of text is brought to you by someone that has walked this path before

to be a better game minecraft needs tighter rules and tougher objectives, as it stands now is more a setting/activity than a game. however i don't think it needs to be a game to be better, right now kids can gather their friends, build a castle of dirt and have one of them play as the dragon with blocks of tnt which then the others try to kill, once is done the one that killed it gets to be the dragon the next round. with the previous example, if i wanted to make minecraft better, i would have to allow to make character models with blocks, to animate different groups of blocks and to play their own sounds to simulate the dragon. then to allow the players to make their own rules, make their teams and even change rules based on teams, also to limit or encapsulate the world so no outside objects get destroyed and to save these settings as presets for every time they want to play "the dragon". at that point minecraft becomes more a game engine than a game but in the process it allows for better minecraft games
If instead of making minecraft better my goal was to make shekels, i would do all that in studio, allow no one to alter or modify anything and sell different children games as expansion packs

This is 100% true despite what chanculture teaches you.

I wouldn't say that imageboards teach you to be cold. Rather, I would say that imageboards teach the importance of eschewing empathy when practicing empathy is not the right decision. Realizing that empathy is the right choice far, far less than people would have you believe is a good lesson that does make you stronger, but people do sometimes confuse that with "coldness = strength," which is patently false. I actually find that warmth you do find on imageboards is always genuine specifically because this is the case.

Coldness is far better than being a gullible fuck that is easily manipulated through his emotions. Most of the mess we're in wouldn't have happened if society as a whole was cold and calculating instead of fickle and malleable. In one's life and different situations a balance of both is the best but if one is forced to choose one and forgo the other, logic should always trump emotion.

Is this bait?

Well, even if it is, it gives me food for thought.

As we know, scanlines added a lot of depth to games way back when. I wanted to describe this but I guess pics will tell more.

Your typical indieshit devs don't even care about scanlines and their games look like crap. Classic DOS games look best on CRTs like here

People who care about the classic effects apply filters and shaders, as wants. All of them look like crap, despite making the games look better it's still not what the classic game was supposed to look like. And some are so overexaggerated, they miss the point to the level that makes me want to puke (gif related).

I never really saw a GOOD shader achieving a scanline effect, not to mention that with different CRT technologies (aperture grille, shadow mask), various possible resolutions and different build quality (chromabs were fucking rare, don't use them), moiree effects and different convergence, linearity and purity settings, no one tried to recreate how the effect actually was achieved, on a technical level. But with modern screens and faster GPUs it should be possible.

Hell, why not make degaussing, magnetic anomalies and other shit a part of the gameplay?

If you plan on making a retro pixely game, If I were you, I'd put a lot of effort to create the first ever good CRT-emulating shader. Perhaps you'd make the first retro game that actually looked acceptable.

Speebot is also the most boring and least offensive project out of the three. Think about that.

Except for the fact that Speebot is a lot more fun than Cuckdere Shitulator.

i've got this in unity
public enum UnitType { infantry, artillery, cavalry, none } private UnitType _type; public UnitType type { get { return _type; } set { if (_type == value) { return; } else { _type = value; SetSprite(); } } }
and i'm setting type from a custom inspector since it doesn't show up otherwise
[CustomEditor(typeof(Unit))]public class UnitEditor : Editor{ public override void OnInspectorGUI() { DrawDefaultInspector(); Unit myTarget = (Unit)target; myTarget.type = (Unit.UnitType)EditorGUILayout.EnumPopup("Type: ", myTarget.type); if (GUILayout.Button("Reset Images")) { Unit.types = null; Unit.Init(); myTarget.SetSprite(); } }}
this works fine
but when i press play, the type resets to the default one (infantry)
how do i fix that?

C++ is much more than OOP, and it's optional. Maybe Java is C + OOP.
C++ is more efficient than C because you can do things at compile time. Template meta programming is one of the more interesting and powerful tools C++ has.

This whole convo was actually something Sam Hyde brought up in regards to being a productive, healthy person. For the sake of practicality, you should cut cords with people who daily manage to make your life miserable, because they sap your productivity, put your under stress and you're very unlikely to change them for the better. But that doesn't mean you should cut out anyone who happens to be a bother sometimes, because that makes you an actually cold person.
You definitely sound like you've gone through years of stress, but judging from your posts you're still pretty angry. I also felt anger once I realized all of my faults were due people around me fucking up, but eventually that anger subsided and I just accepted it. Holding a grudge doesn't help, it just prolongs that stress. You said yourself you're in a better spot now, so you're in a good position to start fixing those faults.

...

Nice, now time for you to implement boids to make it better.

All you need is a decent level of maths, a bit of knowledge in technical mechanics and whatever programming languages you intend to use. And even the technical mechanics part is pretty much optional due to the fact that it is only necessary if you are doing physics-related shit, although it does help tremendously if you intend on adding flashy features like body deformation. For example if you take the time to learn about the finite element method then you can do shit like in embed related at 4:11, which I can GUARANTEE they won't teach you at a gamedev school. In fact, I doubt they would even teach you entry level technical mechanics, just "haha this level looks funny" on shitty Powerpoint slides.

well i added colliders and rigidbodies, but i have no idea how else to get them to avoid each other

www.cs.umu.se/education/examina/Rapporter/PalmqvistDimberg.pdf (someone's master thesis on this exact topic)
please use archive.is/view/feature/130232/modeling_opinion_flow_in_humans_.php (gamasutra article, wanted to archive but it's 8 pages)

Boids are quite simple. You generate a couple of velocity vectors independent of one another. Then you add them up and scale them by a factor (that is preferably adjustable at runtime). Making the boids avoid one another is done by taking each boid, determining the closest neighbors and generating a vector that would move them away from them, if they're too close. For performance it's absolutely crucial to restrict the number of neighbors you take into account. Assuming you keep the boids in an array, you get a runtime complexity of O(n²). This is good enough unless you want hundreds of them. Then you would have to look into some better data datastructures like quadtrees, which massively speed up finding the neighbors.
If you're smart about it, you determine the k nearest neighbors for each boid first, then calculate the avoidance, matching and flocking vectors at once since they all need the neighbors' positions and velocities.

Of course each vector has its own factor.

I forgot that existed, guess I'll go through all the pages.
archive.is/BIqes
archive.is/FhB8w
archive.is/xGfZj
archive.is/XjAoF
archive.is/NCUd4
archive.is/p4TP6
archive.is/2wTr8
archive.is/EYS4a
That's all 8 pages, not even sure if they'll be helpful from a quick glance. That master thesis would probably be the most useful thing, but it's a thesis so it might be hard to get through.

It isn't, really. People who're hardline agreeable will be alright around good people. People who're hardline disagreeable end up miserable and alone.

The point is, of course, that there's a higher aim here, and that they're both necessary. You need to be a monster, when you have to be, but otherwise be kind.

Don't settle for either. Settle for a good life.

It's worth noting boids were used in Pikmin to great effect. See vid related.

alright, i think it works
avoiding the closest unit in front of you
also using a material with 0 friction for good measure
when a unit has finished moving, he can be pushed away after which he will return to his position
private void LateUpdate() { Vector2 pos = transform.position; float dist = Vector2.Distance(pos, desiredLocation); if (dist > 0.1f && (moveTrigger || stoppedMoving)) { Move(desiredLocation - pos); } else { stoppedMoving = true; GetComponent().velocity = Vector2.zero; moveTrigger = false; transform.rotation = Quaternion.Lerp(transform.rotation, Quaternion.Euler(0, 0, -90 + Mathf.Atan2(desiredRotation.y, desiredRotation.x) * Mathf.Rad2Deg), Time.deltaTime); } } public List units = new List(); public void Move(Vector2 dir) { stoppedMoving = false; Vector2 curPos = transform.position; Vector2 endPos = curPos + dir.normalized; Vector2 avoidanceVector = Vector2.zero; if (units.Count > 0) { Unit closest = units[Random.Range(0, units.Count)]; foreach (Unit t in units) { if (Vector2.Distance(t.transform.position, transform.position) < Vector2.Distance(closest.transform.position, transform.position) && Vector2.SignedAngle(Vector2.zero, t.transform.position - transform.position) > 0) { closest = t; } } if ((endPos - (Vector2)closest.transform.position).magnitude < 0.27f) { //potential collision avoidanceVector = ((Vector2)closest.transform.position - endPos) * 1.5f; } } rig.velocity = (avoidanceVector + dir.normalized) * Speed; transform.rotation = Quaternion.Lerp(transform.rotation, Quaternion.Euler(0, 0, -90 + Mathf.Atan2(dir.y, dir.x) * Mathf.Rad2Deg), Time.deltaTime); }
the only problem now is if there's a wall of units, you can't avoid those and have to go through them

I can barely get 50 fps rendering this scene on my intel integrated graphics craptop and it goes down to 40 when I make the sphere move up and down

Can someone help me enable VAO in my code?
My previous code (without VAO) that works fine:
public function addVertices(vertices:Array, indices:Array, calcNormals:Bool = false) { if(calcNormals){ calculateNormals(vertices, indices); } resource.size = indices.length; GL.bindBuffer(GL.ARRAY_BUFFER, resource.vbo); GL.bufferData(GL.ARRAY_BUFFER, Utils.createFloatBuffer(vertices), GL.STATIC_DRAW); GL.bindBuffer(GL.ELEMENT_ARRAY_BUFFER, resource.ibo); GL.bufferData(GL.ELEMENT_ARRAY_BUFFER, Utils.createIntBuffer(indices), GL.STATIC_DRAW); } public function draw() { GL.enableVertexAttribArray(0); GL.enableVertexAttribArray(1); GL.enableVertexAttribArray(2); GL.enableVertexAttribArray(3); GL.enableVertexAttribArray(4); GL.bindBuffer(GL.ARRAY_BUFFER, resource.vbo); GL.vertexAttribPointer(0, 3, GL.FLOAT, false, Vertex.SIZE * 4, 0); GL.vertexAttribPointer(1, 2, GL.FLOAT, false, Vertex.SIZE * 4, 12); // 4 * 3 floats GL.vertexAttribPointer(2, 3, GL.FLOAT, false, Vertex.SIZE * 4, 20); // 4 * 5 floats GL.vertexAttribPointer(3, 3, GL.FLOAT, false, Vertex.SIZE * 4, 32); // 4 * 8 floats GL.vertexAttribPointer(4, 3, GL.FLOAT, false, Vertex.SIZE * 4, 44); // 4 * 11 floats GL.bindBuffer(GL.ELEMENT_ARRAY_BUFFER, resource.ibo); GL.drawElements(GL.TRIANGLES, resource.size, GL.UNSIGNED_INT, 0); GL.disableVertexAttribArray(0); GL.disableVertexAttribArray(1); GL.disableVertexAttribArray(2); GL.disableVertexAttribArray(3); GL.disableVertexAttribArray(4); }
My attempt at using VAO, which results in nothing being drawn:
public function addVertices(vertices:Array, indices:Array, calcNormals:Bool = false) { if(calcNormals){ calculateNormals(vertices, indices); } resource.size = indices.length; GL.bindVertexArray(resource.vao); GL.bindBuffer(GL.ARRAY_BUFFER, resource.vbo); GL.bufferData(GL.ARRAY_BUFFER, Utils.createFloatBuffer(vertices), GL.STATIC_DRAW); GL.vertexAttribPointer(0, 3, GL.FLOAT, false, Vertex.SIZE * 4, 0); GL.vertexAttribPointer(1, 2, GL.FLOAT, false, Vertex.SIZE * 4, 12); // 4 * 3 floats GL.vertexAttribPointer(2, 3, GL.FLOAT, false, Vertex.SIZE * 4, 20); // 4 * 5 floats GL.vertexAttribPointer(3, 3, GL.FLOAT, false, Vertex.SIZE * 4, 32); // 4 * 8 floats GL.vertexAttribPointer(4, 3, GL.FLOAT, false, Vertex.SIZE * 4, 44); // 4 * 11 floats GL.bindBuffer(GL.ELEMENT_ARRAY_BUFFER, resource.ibo); GL.bufferData(GL.ELEMENT_ARRAY_BUFFER, Utils.createIntBuffer(indices), GL.STATIC_DRAW); GL.bindVertexArray(null); } public function draw() { GL.bindVertexArray(resource.vao); GL.drawElements(GL.TRIANGLES, resource.size, GL.UNSIGNED_INT, 0); GL.bindVertexArray(null); }
VAO, IBO and VBO are generated beforehand and saved to resource.vao, resource.ibo and resource.vbo.
What am I doing wrong?

Glad to see you're following my advice.
1. You need to enable the attribute with enableVertexAttribArray, just like you did before. This state will be saved in the VAO.
2. Which element buffer is bound is not saved in the VAO. This state is entirely separate from the VAO.

I tried following your tips but could not get it working. Here is my current code:
public function addVertices(vertices:Array, indices:Array, calcNormals:Bool = false) { if(calcNormals){ calculateNormals(vertices, indices); } resource.size = indices.length; GL.bindVertexArray(resource.vao); GL.bindBuffer(GL.ARRAY_BUFFER, resource.vbo); GL.bufferData(GL.ARRAY_BUFFER, Utils.createFloatBuffer(vertices), GL.STATIC_DRAW); GL.enableVertexAttribArray(0); GL.enableVertexAttribArray(1); GL.enableVertexAttribArray(2); GL.enableVertexAttribArray(3); GL.enableVertexAttribArray(4); GL.vertexAttribPointer(0, 3, GL.FLOAT, false, Vertex.SIZE * 4, 0); GL.vertexAttribPointer(1, 2, GL.FLOAT, false, Vertex.SIZE * 4, 12); // 4 * 3 floats GL.vertexAttribPointer(2, 3, GL.FLOAT, false, Vertex.SIZE * 4, 20); // 4 * 5 floats GL.vertexAttribPointer(3, 3, GL.FLOAT, false, Vertex.SIZE * 4, 32); // 4 * 8 floats GL.vertexAttribPointer(4, 3, GL.FLOAT, false, Vertex.SIZE * 4, 44); // 4 * 11 floats GL.bindVertexArray(null); GL.bindBuffer(GL.ELEMENT_ARRAY_BUFFER, resource.ibo); GL.bufferData(GL.ELEMENT_ARRAY_BUFFER, Utils.createIntBuffer(indices), GL.STATIC_DRAW); } public function draw() { GL.bindVertexArray(resource.vao); GL.bindBuffer(GL.ELEMENT_ARRAY_BUFFER, resource.ibo); GL.drawElements(GL.TRIANGLES, resource.size, GL.UNSIGNED_INT, 0); GL.bindVertexArray(null); }

Sorry, I was actually wrong in regards to element buffers. They are included in VAO state. I just messed around with my own code and will go ahead and confirm they are included for my setup.
Some older shitty intel drivers didn't actually include them, which is probably why some tutorials don't say anything in regards of it being included.

Probably a stupid question, but did you make sure to generate a VAO?

Yeah, the VAO is generated…

I can't really find any issues with it.

Your code isn't effectively all too different from mine.

Does glGetError() spit out any errors?

C can to with pre-processor directives
They also rape compile time. Boost is a shining example of this. Also I don't think templates have anything to do with program efficiency. That's more for minimizing source code.

Calling GL.getError() at the end of draw() sometimes displays no error, sometimes - GL_INVALID_OPERATION.

You should probably track down which call exactly causes that.

I debugged it properly now and there are no errors coming from addVertices() or draw().

Alright, spill it. Which one of you is developing a VR compatible game?

I got no clue. Maybe try just create a separate test project and try to get VAO's to work, and then try to find the differences between that code and the speedbot project.

>Using attribute((constructor)) to initialize static vectors

i was thinking about making one but i no longer have a vive to use.

VR is a meme

Opinions on making my own editors in libgdx to make a small engine on top?

If your goal is to make a game, that's quite a waste of time.
Use a popular engine or an open source one.

Pre-processor isn't C. Might as well write a programming language that expands on top of C… oh wait.

having the compiler spend few more milliseconds is better than having a programmer spend few more hours or days writing, debugging, and maintaining repetitive code.

Also they don't rape compile time anymore. Boost is an outdated example of pre-C++11 that probably heavily relies on macros and hacks. Modern TMP libs scale quite well, see vid at around 31:00 for benchmarks.

And you don't even need them that much anymore. Useful and fundamental TMP practices are added to the standard.

most engines are shit tier for fighting games.

I'm interested in having a poke at gamedev, main thing I'm interested in is making a turn-based tactics game in the vein of Final Fantasy Tactics or Dofus. That being said what engines should I look at using? I don't want to use a bloated piece of shit like Unity, I always hate playing Unity games because they're usually needlessly large and demanding. I have some experience with C and Lua so engines that are friendly with these two languages are preferable, as is one with development tools available for Linux.

then use Love2D

Template metaprogramming still absolutely fucking destroys compile time, code size, and performance as so many aspects of the language go completely haywire and are literally uncontrollable because of it.
The example I usually give people who don't want to believe that is to try and implement a simple variadic printf and evaluate it. It takes 1-5 minutes depending on skill (for Holla Forums, possibly a few months) and relies on recursively peeling off the list lisp-style via templates. When you look at it you'd think everything would just boil away at compile time and it'd be no slower (or even faster) than printf, but then you put some pressure on it and start to discover the horror - impossibly massive bloat as you increase the number of arguments to functions and disastrous compile times as you vary the combinations of arguments fed to it.
Of the many problems you'll discover, the worst is the bloat. It's caused by compilers attempting to limit the depth of recursion of templates to prevent locking the compiler up. Even if you don't hit the hard limit, most compilers (none of this is standardized) will attempt to stop the recursion early. They do this by emitting the method in the binary to end the recursion. Even if you declare it 'inline', that will eventually be ignored and emitted. Since the arguments and their types remaining will vary, you'll end up with a ton of nearly identical copies of it in your code.
In attempting to make the simplest and most commonly cited example of variadic metaprogramming actually work, you'll find yourself adding lots of compiler-specific hacks and attributes to prevent these problems.

Is anyone here unironically making a TempleOS game? It has a sort of beautiful simplicity to it. No libraries, no frameworks. It's like BASIC but better, which I guess was Terry's goal.

This is the only project I'm aware of.

What the fuck. Goddamn CIA.
Seriously though, it was up yesterday. This is really bad. This guy made the only program that I know of that can create TempleOS compatible ISO.C files. Fuck, I should have downloaded everything when I had the chance.

Not everybody can suckle on mommy and daddies teet forever you know.

This can't be so hard: I want the camera to rotate in the direction the player is facing while he is moving, depending on how fast he's moving. So

so you know the answer but refuse to implement it just because is not elegant?
just like make game you fucking retard

WHERE'S THE GODDAMN PROGRESS!?

Post yours faget.

Would do it, but it's in another computer in the moment.

Yeah I'd post my progress too but my dog ate it, maybe next time

A font.

hehe

calculate the transformation you want to do, then linearly interpolate between identity transform and that the desired transform based on player speed.

if you dont give us a name soon manlet simulator will stick.

Why? What's the special problem, and what prevents you from fixing it within an existing engine?
How's that problem affects every domain in a game engine (graphics, input, audio, physics, path-finding, networking, serialization, …), that it requires you to write a whole new engine?

literally linked u a video with benchmarks
If you mean executable size - templates can reduce size because they only instantiate what you use.

If you mean code size as lines of code - u wot m8? it eliminates tons of repetitive code.

Doing things at compile time instead of runtime means better performance.
Hello world level example of this is C++'s std::sort being faster than C's qsort.

What's your point? That you don't know what you're doing?
std::cout doesn't use variadic templates.
And if you had a legitimate use case you could use optimized TMP libraries like brigand.
Also C++17 adds fold expressions that avoid recursion.

Why the fuck won't UE4 allow me to call functions via UObject pointers in UStructs?

//UEffect_Masterclass inherits from UObjectUEffect_Masterclass* Effect;Effect->SomeFunc();

Second line just causes "use of undefined type 'UEffect_Masterclass'"/"pointer to incomplete class type not allowed". I can declare variables as pointers to the UObject class no problem, but as soon as I try to access them it's fucked. I'm working inside of a custom UStruct, which seems to be the important part here. Does UE4 not allow that? I can do all this in another way but would be nice to know if I'm doing something wrong or if it's forbidden by design.

Yes I've tried including the Effect_Masterclass.h, and I've tried forward declaring it.

Nevermind, it seems to be that you simply can not access stuff via pointer from UStructs, regardless of what they inherit from. I assume that this is by design.

Is there a comprehensive guide or something to using compiled languages?

Like how do I properly manage my code and files, how do I link files, can I use libraries without compiling them every time, how do I use variables and data stuff without making everything globally accessible, how to use debuggers, etc etc.

I don't want to go to school but I feel like it's the best option, since they have to walk you through and explain everything even if you don't know anything about any of it.

You compile your libraries only once.
With the debuggers you learn as you go, make an hello world, decompile it and read the ASM, do some not so complicated if/else and decompile it too then move onto more complicated things and decompile as well.

Read a introduction to C programming. There are some good ones. Go look at some good source code examples of the types of projects you would like to do. Have you looked at the Doom source? How it's organized? How they designed their data structures?

MITOpenCourseware has some decent online learning too. So do a few other universities.

Are there any good resources for someone who's familiar with C and wants to learn C++11? I don't want to go full OOP, but I want an idea about what all this new shit is for and how to use it right.

You can watch the Keynotes from CppCon for a simple introduction.

I may sound like an insufferable faggot, but just go for Java/C#. The advantages of not having to deal with all of the shit you described outweighs the VERY SMALL performance gain. I am talking about 0.001%ish performance gain, since any CRITICAL code can just be implemented in fucking assembly if you want to and then call it from Java/C#

Do check these ( ) anyways since they look like great references, regardless of what language you are using

You do, and you're going to give this guy the wrong impression with that. The fact is it's OOP that slows shit down, and while you -can- do Data-Oriented programming in Java/C#, it's a huge pain in the ass to do it properly and actually results in more complicated and slower code than if you simply do it in C/C++.

Really, the amount of misinformation about this is staggering, the industry has everyone fully convinced that the performance difference is minimal when it is anything but.

Watch Handmade Hero, it's amazing and will demystify a lot of the seemingly complicated parts of C/C++

please elaborate then

Writing things in assembly is retarded because:
1. it's hardware specific
2. it isn't maintainable
3. the compiler will do a better job than you

and it isn't VERY SMALL performance gain, or 0.001%. It's more like 500% on average, which means you'll be years behind your competitors in terms of performance according to Moore's Law, so you'll lose in any domain that's performance critical.
(note that a lot of shit-performing languages fall back to C to implement performance critical stuff)

Note that hardware is so powerful today that most games can get away with shit performance.

Also "Data-Oriented programming" is a meme created by people with most basic understanding of data layouts, that fail to identify that that "Data-Oriented programming" is just a single case of locality of reference and it isn't a silver bullet.

I just quickly watched it and it looks like it's really good. You can check that too, it's a let's remake of cave story :
youtube.com/watch?v=ETvApbD5xRo&list=PLNOBk_id22bw6LXhrGfhVwqQIa-M2MsLa

Thinking back, it's retarded because you can just make a compute shader instead if you really need
nice source

lolno, unless you're processing +10,000 objects since only the memory transfer overhead will take more time, and your algorithm needs to be parallel.

lmgtfy.com/?q=programming language benchmark

good choice. I recommend strongly against trying to do everything using the OOP concept. Many things are just better expressed functionally.

decent top-10 of C++11 here: codeproject.com/articles/570638/ten-cplusplus-features-every-cplusplus-developer

"auto" is very nice when you're using throw-away stack variables and complicated iterator types. Ranged loops like "for( auto x : itr.second ) are so much easier to read and use. Very good feature. Smart pointers are probably the most useful and important additions to C++.

There are some nice tools for concurrency and threads too. I'm not sure if they are good or not yet. I've been experimenting with OpenMP and haven't used much else besides pthreads.

In GameMaker how would I go about making an object disappear partly? I want the player to be behind the object and for part of it around the player to become transparent wherever the player moves.

C++'s concurrency model with futures really simplifies things.

easily run a function asynchronously. Can be used with std::launch::deferred to achieve lazy evaluation.

Easily attach a continuation to a future that'll be executed once it's ready.

The performance gain is just a nice benefit. It's actually much greater than what you are implying. The ability to, for instance, manage your particle system's memory by-hand yields a pretty massive performance gain over a GC language. It might just be something as simple as a fixed, heap-allocated array with a wrap-around index implementation. Maybe with a priority level. Whatever. With C/C++ it's your choice. Given all of the boilerplate with java, it's probably fewer lines of code as well. And it isn't crowbared into an OOP thought process. Benifit: no random frame-drops for blackbox GC "doing something important".

C/C++ is basically cross-platform assembler with some modern language features. It's fun. You are exposed to how a computer actually works and it's not a bad thing. I find it endless amusing that when you combine well-written cross-platform libraries with C/C++ it actually delivers on the cross-platform promise. Unlike Java and C#/mono.

It feels like it would be something done with draw_set_blend_mode( substract), using a circle object (or whatever shape you want your object to transparent around the player sprite) and playing with the depths. docs.yoyogames.com/source/dadiospice/002_reference/drawing/color and blending/draw_set_blend_mode.html

check it out and see if you can make it work, I've never done anything like this myself but it feels like this is the way it should go.

They're very different languages.
Sure, C++ is generally a superset of C, but only pajeets write "C with Classes".

also checked.

the google search gave me no source for your crazy claim, the best bet was this website, which have a nice page for you to read: benchmarksgame.alioth.debian.org/dont-jump-to-conclusions.html

(checked)
Yes, and for these very small black box systems, you can just write in freaking assembly if you have to, and call from Java/C#, that's my whole point.

I am not taking about performance gains on core parts of your code. If it's really critical, fucking make it super low level.

I am talking about the other 90% of your code that is freaking extensive to code but the machine spends very little time on it. The simplicity of Java/C# outweighs the performance gain on those 90% of the code, where the critical parts of your code, can be just tossed to a native plugin or compute shader, like the particle system, which everyone should make run on the GPU due to it's parallel nature

It's probably because you are one.

what the fuck is the point of these "mosaic" reaction images if the images in the mosaic arent actually what contribute to the colors in the image?

it's clearly just a bunch of vaguely-properly colored anime reaction images overlayed over the reaction image theyre supposed to be forming

this is fucking low effort and worthless, it's not even doing what it's supposed to do. if you're gonna make a fucking mosaic like this then do it right, fuckin pussies.

Removed top down shadows from Speebot, replaced them with this. Looks much better.

Ask the faggot who made it, not me.

nice

Kids are pretty short, bro.

I've thought about making a program that takes a bunch of images, checks their average colors, and then creates a pixelart from them by using the image with closest color compared to each pixel of a given reference image.

OK. I guess if I have a point then it's that Java or C# is not any more simple or compact than C or C++. C/C++ isn't that hard to write.

Naaaah. C++ is modernized C. Classes aren't even required or beneficial in all cases. Half of C++ features shouldn't be used by anyone but a library developer. A few others are dubious at best (operator overloading).

Pajeeet-tier C++ would be massive overuse of inheritance and classes in general.

Added a head bob with matching (WIP) sound effects. I think the roll is a tad too much. Also WIP light indicator.

I love it, please keep posting these videos.

That's a point I can't dispute due to lack of experience with C and C++.

Since the previous argument will not go anywhere, I will make a new one:
I propose that:
a) It's easier to make games on game engines rather than building your own, thus increasing the chances of you ever finishing it
b) Unity is the best game engine, especially after they finish this job thing (embed), which will make it super easy to make those 'highly specialized black box systems on plugins' but not really, since theoretically you are still running on c#. I will go ahead and wildly and baseless claim that you can get about 70-80% of the performance of writing super specialized code in C while only having to bother with 20% of the effort needed
3) In Unity you have to code with C#, therefore making C# the superior choice of code language if you intend to ever release anything.

Good goyim

Which other engine can compete?
The most popular engine will always be the best one due to the second point, loads of high quality free stuff

looks great! I downloaded your demo but haven't had time to play with the level builder yet.

The best engine is the engine that individual anons make the most progress in.

Looks great.

For almost everything I'd like to work on, another engine can do it better. That, and I use Linux as my main platform and Unity's Linux support still isn't the greatest.

I'm afraid the level builder is not in the demo. It's not finished, too. I'll get to that later this week.

Unfortunately, all of you are wrong - the best engine is YUME. Now with angular deadzones!

Sorry Keyreal, but Yume won't be the best engine until it's ported to HolyC, and everyone who was fighting over languages ITT needs to reread the OP. But thank god for angular deadzones, though. That was about the only improvement I felt was necessary.

lol what a bunch of manlets

Is it possible somehow when baking a High-Poly model to a Low-Poly model where the light is distributed evenly? Since I don't want it at all that there is such a strong shadow at the bottom of the rock. I tried to look for some solutions online but the search engine didn't gave me any relevant results at all. I tried also using 2 hemisphere lightning but it just made the whole baked texture look pretty weird or so, its hard to tell when the details are hardly visible when doing that way.

Am I autistic?

Watching something come together is always exciting.
It's the same appeal as watching cooking videos, it's fun to see the assembly of otherwise-unrelated stuff come together.

I like your games aesthetic. Reminds me of ape escape

Way to take it out of context. You sure shitpost more than you dev. No application is performance intensive until proven. Don't be a retard and optimize before you need it.

under bake change full render to normals for bump maps or textures for just color

I've yet to see a real one.

...

...

What does this error message even mean?

Nothing gets baked into it when I just use "Texture"

Working on some animations.

"texture" imprints the unshaded colors from the model, if you have no textures on that model it will only bake the default color

Anyone try using a kinect for mo-cap? I never animated before and probably will have to resort to using it, because there is a good chance I'll end up having to make a ton of animations.

All i remember is trying to set the drivers up way back when it was still relatively new, the 360 one. Can you actually mocap with Kinect?

There already is being software sold for it which machinema's use. Also I've seen people use a library called OpenNI for it (vid related). I was thinking of using the later to rig some sort of mocap system together to avoid paying out the ass for licenses.

Is that the old 360 kinect or the new Xbone one?

If they showed the depth data I could tell you. The only thing that would matter between the 1 and 2 is the shadowing effect. Example is pic-relate, the shadow on the kinect 2 is like 1/5th the size of that. Personally I'd go with the kinect1 just to test it out because, last I remember, they're dirt cheap and plug straight into the computer. For the kinect 2 you have to shell out and extra $80 for the developer adapter that lets it plug into the computer.

Mocap with PS3Eyes is cheaper, more precise and versatile

but more of a bitch to set up and it's most likely harder to master

it's here

8agdg.wikidot.com

add to thread template pls

You still have to use a 3rd party program which prob makes you use a license to sell shit. With the kinect I think you can use a combination of opensource libraries where something could be rigged together for free.

Good work user! How can i create a page? I've signed up but i can't find the option, im thinking an article "Active Projects" listing every game actively developed

nvm i found it

Started the list, my descriptions suck and the screenshots are what i had laying around, get to work on your game entry anons

I'll be browsing my saved /agdg/ threads for games that appeared here.
these descriptions are breddy good tho

Format it like this instead. I don't know if any of this info is correct for Speebot though, it's just an example.

Someone is editing the page right now, but i agree with that change. Rename "Image" to "Screenshot" as well.

Done.

Halfchan /agdg/ thread has also these links for resource:
Website: tools.aggydaggy.com
Weekly Recap: recap.agdg.io
AGDG Steam Games: homph.com/steam
Fanart and stuff: drive.google.com/drive/folders/0B6j4pcv3V-vfb3hKSlhRRzlLbFE
New Threads: >>>/vg/agdg
Archive: boards.fireden.net/vg/search/subject/agdg
AGDG Logo: pastebin.com/iafqz627

GameMaker: yoyogames.com/gamemaker
Godot: godotengine.org
UE4: unrealengine.com
Unity: unity3d.com

opengameart.org
blender-models.com

freesound.org/browse
freemusicarchive.org
incompetech.com/music
fantasymusica.org

obsproject.com
gitgud.io/nixx/WebMConverter

I don't think we want any of the 4chan/agdg/ links either.

There's also this >>>/agdg/29080

It would be better to have a fully comprehensive list of all options for everything, like /loomis/ has this page: 8ch.net/loomis/hub.html

well, starting a page on the wiki with this would be a good idea

Ehh I am wrong, the texture baking does work when I apply the "Matcap" Material to my High-Poly model but the result is a few "blotched paint" and it doesn't solve my bottom shadow issues either, which sucks.

Welp I forget that the /agdg/ board here had few links too.

Well, that went places I didn't expect.

Meh I guess for this one I am going to stick with Ambient Occlusion for now then, I didn't know that it even accepts diffuse colors.

like?
Does not sound like sour grapes AT ALL

welp i'm sold. i'm going over to the blender game engine

opengl 1.1 > *

this is true

meant for

id Tech engines and forks, PLIB, Irrlicht.
Assuming you want a 3D engine, 0 AD's engine or Spring. There's a lot of room for improvement here and if the scale is large enough you probably need to write your own engine.
Probably UE4 or Cryengine if you hate documentation. You could also wait a little while for Godot 3.0 or Armory3D.
Tesseract or any id Tech engine/fork unless you're going open world, then see the above.
Godot is probably the go-to choice these days. There's also Love2D, Anura, Orx, and a multitude of others if you're willing to sacrifice the fancy engine IDE.
Unless you're doing something really unusual, Ren'Py.
I dunno, look into LambdaCube 3D, Nu, and Hate (Haskell clone of Love2D).
Crown for 3D stuff and Orx for 2D.
OpenSceneGraph or some other scene graph
GemRB (re-implementation of Bioware's Infinity Engine), EasyRPG (re-implementation of RPG Maker 2000/2003), and Solarus (designed for churning out 2D Zelda clones and similar games). If you need something more unusual, fuck around with one of the generic 2D engines listed above.

There's a fuckton more, but I'm not trying to list everything. Point is, Unityfags and Gamemakerfags get really defensive over their chosen engine, declaring it as God's gift to indie devs in the form of a black box that makes their dreams real. Different engines have different architectural advantages and a generic "one size fits all" engine will usually be outperformed by a well-designed, more specialized engine.

You listed a lot of alternatives, but not any explanation why any of them are better than Unity.
I will ahead of time concede that Unity is not for 2D stuff though

When you don't use unity, you don't have to program in awful languages like C# or JavaScript, and nobody will instantly discard your game on account of it just being part of the flood of shitware that gets made on Unity. Also, I highly doubt Unity has performance anywhere near what other engines have for the same thing, because you have to run your code though a Microsoft VM with garbage collection, and whatever else heavyweight shit Unity puts on your game.

I too can make non arguments:
When you use unity, you don't have to program in awful languages like anything but C#

The fact that C# is:

Ahah, well, damn.
Looks like this is gonna burn a fire under my ass and force me to finish my game, lest I get publically humiliated.
Time to get to work.

Dude, have you ever played it since 2005? Do you know that whole dev team did change 4 times? Do you know that since years this shit became terrible showcase of resource hogging, pathfinding is completely broken and units constantly keep colliding, merging and getting stuck with other units/buildings/even terrain?
All of that because current dev team "couldn't understand" the code that previous teams wrote and decided to scrap it, rewrite it themselves and since 8 years now it has been unfixed mess.

Don't ever recommend that pile of shit you fucking faggot.

Thanks for the warning, user.

You will have to prove that java is bad before that becomes an argument
Have it's pros and cons, you don't have to make objects if you don't want to
And why is this intrinsically bad? Also, I am pretty sure that there is a OpenJDK equivalent for C# if you really want to be a bitch about

meanwhile
www.assetstore.unity3d.com/en/#!/content/17660

There is also the free version which is more of a tech demo than a toolkit

First of all, Java has a truly awful syntax. The way Java works is that you kind of tie together bits of the enormous standard library in interesting ways, which means that you spend most of your time looking through the documentation for the piece of the standard library, and then when you find it writing an enormous initialization statement and then, when the boilerplate is done, actually doing what you wanted to do. Compare "System.out.print" to "printf". Opening a file means finding the right thing to use , and the standard library duplicates functionality constantly, so you have "FileReader" "FileInputStream", and an endless amount of other duplicate objects that have an unncessarily verbose way of being typed out. Also, you have to make Objects for everything. Objects are cancer because explicitly associating your data with your functions is totally arbitrary and only makes sense in specific cases, not for every single thing in your program. This just makes your program bloated as you try and force this idea onto everything. Also, you have to waste your time writing "getters" and "setters", which are the most retarded language feature ever, instead of the normal (C) way where you write "structure->feild", you have to type out an entire function to do this basic functionality, and then write "structure.get_feild()", which is verbose and unnecessary 99% of the time. Also, For every object I have to make a new file, which is just aids and means that I need a stupid amount of files for my program. Also, to point out some more "anti-features" of java, lets consider:

I have a world that I want to get written down and flesh out. I have stories, histories and characters but I'm having trouble forming my chaotic thoughts to put down in a more organised manner.
Does anyone have experience with something like this?
Where should I start and what's the best way of going about this?

I manage everything in plain text. I may split more complicated concepts in several files if there's something particular that requires a lot of text on it's own.

Code folding is my favorite features in text editors, it makes everything much more manageable and I can easily have thousands of lines of text without getting confused. Line wrapping on tab position is also very useful for it.

Pic related is my todo/notes file for one of my projects. here's hoping I didn't capture anything embarrassing in the gif

I'd imagine uploading the image would help.

I have experience with this. You can use a desktop wiki to keep things relatively organized as you write them down. Personally I use Zim. The main thing you will want out of this is to be able to find any paragraph of text in a timely manner. If you put everything in a single document this is going to be hard.
There are two ways you can start. You can either start with the furthest reaching aspect of your world, or the part of your world you value the most. Starting with a far reaching aspect of your world (like the economy, the laws and government, the society and it's customs)
will make keeping things consistent overall easier. Starting with the part of your world you value the most means that you will build the rest of your world around that. If you don't know where to start, start with the geography.
Internal consistency is more important than realism, but having some level of realism will make your world more accessible to people and keep you from having to explain every little aspect of your world. Keeping both internal consistency and realism in check minimizes immersion breaking. Keeping internal consistency, realism, theme, and story all in check is a balancing act that no two people tackle the exact same way.
As for how exactly to organize your world, that depends on what your world is like. If you have a lot of factions, you may want a page that explains what factions are and some general facts that fit with all factions, and then you have a list of links to the pages that explain specifics about a faction.
The ultimate end goal is for all of this stuff to become routine enough that it just gets out of the way and you can focus on your world.

I don't think there's much advice that I can give you besides just sitting down, open a word .doc and just do it.
Start giving a short summary (2 pages at most) of the events that thanspire during the game. Once that's done, start by explaining the setting, from the biggest to the smallest details, so if you have a world with particual history of properties, start from there and once you are done start with the countries, then communities, etc (this is going to be long as fuck if your world is fleshed out, make a lot of coffee).
After that go for the Characters, making a clear distintion about who is going to be a main or support character.

New thread?