/v/ - Video Games

>>14671171

Are you allowed to use already written scripts for LD or does everything need to be brand new?

>>14671302

My understanding is that you can use publicly available libraries. Assets might be okay if they're public as well. Scripts are "specific game content" in my opinion so it'd fall into an asset, no?

15 days until demo day

>>14670418

>>14670620

if you wanted to be a total hackwitch why the fuck wouldn't you, don't make shit hard for yourself you could use the rigify armature as a reference and build a body around that and then generate weights to that and then extrude armor from that.

>>14671989

I really like it inbetween but full leaves and none looks really weird.

>>14672043

Some anon posted a great tutorial on how to do trees in the previous thread. Just do that.

I'd give up on trying to get the no leaves look cause that shit is always going to look weird. Even professionals can't make those look right.

>>14672007

You become slow, weak, dumb and you never give up. That means you'll probably die 30 seconds into your first fight because you will fight way past your capabilities and destroy your own body.

>>14671290

try to read robert mckee the story and find on duck duck go Emotionering the art of emotions in games.

Good writing is as technical as good music and art, so learn the rules.

>>14672065

The polycount tutorials won't help me do what i need, my trees are not static, they need to lose leaves

>>14672093

Or if you want to go full AAA-level, just use speedtree like everyone else:

https://store.speedtree.com/

>>14672085

Talking about this

https://simonschreibt.de/gat/airborn-trees/

Just have the color switch in between the ones you want. While your solution is more realistic, it might not be the best solution for an independent dev with no budget.

If you're trying to make it look like the tree is dying, then you'd probably even be better off changing the color from green to a sickly dark green/dark grey.

>>14672104

>buy triple A assets

>can't make any new assets because you're not good enough at modeling to make anything else fit in with the style of the game

What a fantastic idea

>>14672109

Its a seasonal leaf loss, unless there is a good way to shrink the cloudy leaf mesh following the branches that method won't help me

Also the trees need to be mid-poly as there will be many, but they need to be decent-looking from close

>>14672043

Put whatever you're using to modulate the leaves' size into a square root, so changes more rapidly when close to zero.

>>14672123

I think the problem is not the speed of leaf loss, but the texture itself, i just don't know how to do it better.

>>14672121

Now ask yourself, do you NEED the leaf loss? Cause you're just making it more complicated on you for something that I seriously doubt is important to the game itself.

Animal Crossing switches between all the seasons with minimal new assets and it looks great because they don't try to make it more complex than it actually is.

The choice is yours, but I wouldn't spend too much time on small details that are really not that important.

>>14672134

Its a farm game, season change is a VERY important feature

We are also doing the season change very gradual, leafs won't lose all leaves overnight like in every game of the genre

trees won't lose all leaves*

>>14672167

I am aware, im just saying lobes would look better because there would be no sharp angles. Looking at your fully grown tree right now the billboards stick out a lot.

Speedtree apparently does leaves falling out by simply removing the leave lobes one by one, embed related, starts at 0:26. Looks strange in a timelapse but if you would have them lose the leaves that fast ingame you could alpha blend them out. I also like that the tree becomes brown from top to bottom, not in a global fade.

>>14672180

so just bend the branch billboards instead of having them be simply flat?

Would anybody be down to make a game about the dyatlov pass incident? I think it'd be good for horror.

>>14671989

I hope you can handle multiple trees

>>14672184

Sort of, here's a good paper on texture lobes

https://www.cs.tau.ac.il/~dcor/articles/2011/Texture-Lobes.pdf

Im sure you can find less daunting tutorials as well

>>14672203

Ok, i will read it, thanks

https://github.com/ndee85/coa_tools

opinions?

>>14672190

-What is the moment to moment gameplay like? Describe to us what the player does.

-Do you have a design document?

-What engine are you working in?

-How many levels will be in the finished product?

-If there's dialog how many lines of dialog? If you're the person writing the dialog may we see some samples?

-What skills can you bring to the table?

-What kinds people do you need? Programmers? Artists? If so what kind of artists? 3d Modelers? Sound Designers? Environment Artists?

-Who do you already have in the team?

>>14672195

Good to hear.

>>14672110

poast her in a swimsuit

>>14672080

>robert mckee the story

ok

>duck duck go

stop using that shit, anon, and use startpage or ixquick

>>14672190

>the dyatlov pass incident

never heard of that b4 look up on wikipedia, shit that's scary

so you going to have a "beast" that you flee while you slowly freeze to death trying to make it down the mountain? fires warm you but draw in the beast, loop around back to camp for some warmer gear or take your luck running strait back to the "village"

>Game like Chip's Challenge

>Have an item called Sun Orb and Moon Orb

>Want to have it so they open their own gates and close the other ones when they are activated by pushing them into something

>Can't think of anything besides a generic hole in the floor

Any ideas?

>>14672261

i havent made it yet, we are trying to get the protag working and moving on the town for the 5/5 demo, so i stopped making the girls for now

>>14672334

The explanation is actually pretty clear. After looking at the stuff that was found it appears as if they were cooking something inside one of the tents, this caused the tent to catch on fire.

The people inside are forced to get out in the middle of the night, half naked, in the middle of a blizzard. They can't find their clothes so they decide to walk towards their friends or something else down the mountain.

This guy does a pretty good analysis. Sorry for ruining it for you.

>>14672353

What? Radiation poisoning is more consistent with their symptoms. Plus they were in a no-go part of the mountain

>>14672356

There is an explanation for the poisoning too. I don't remember exactly but one of the people worked with radioactive materials and many of the clothes were embued with radiation.

None of them actually had radiation poisoning, they were just slightly radioactive.

>>14672353

Huh, the german wikipedia article did not mention a canonical explanation, will watch the video. Still you'd be retarded to miss up on a chance to incorporate all of these themes and instead use a monster. Then again the lovecraftian-type undescribed or nameless horror can't really be conveyed in a video game, you'd have to find one definite explanation for it.

>all of a sudden I can't get a collider with a script to display a simple debug message on trigger stay

What the fuck

>>14672334

You forget one of the investigators being told by the military to call off the search and deem it a "force of nature"

>>14672583

>restarted unity and now it's working

good engine

The Theme for Ludum Dare 41

Combine 2 Incompatible Genres

>>14672167

Hey tree hugger. Have you looked at L-systems at all?

>>14673032

If you finished the other half of the walking cycle it wouldn't be half bad.

>>14672936

Racing TCG

>Literally exists, it's a classic french game called millesborne

>>14673089

>Racing TCG

That was a super shitty yugioh series.

>>14672955

If a teacher can put up with him for long enough he'll be an unstoppable retard.

>>14673032

Does he walk with an invisible walking stick?

>>14673199

lol I dunno.

Maybe I should try to stop making such high res sprites.

>>14672982

Just googled it, looks awesome if you know how to use it, but its probably not the optimal choice for our project

>tfw I've been trying to get godot ui to show up on screen properly all day

>Finally realize that some elements don't show up because the default min size is always 0,0

Fucking why.

>>14673319

Probably because whatever primitive type or tuple it uses is uninitialized to 0,0 by the language specs

>>14673343

Yeah, that's probably how it's set up, I wish it defaulted to 1,1 or something. That way I can at least see something on the screen when I run the scene. I had completely missed the min_size property it seems.

>>14673022

My theme is "vaguely medieval fantasy with some machinery"

>>14673478

>>14673528

I tried this awhile back, and i failed at it.

tried it this time, and easily made hair cards; thanks.

despite that… i do dislike haircards, and am going for a really specific semi-realism style. good to know, but autism demands i conform to my current art-style…

look at this shit and tell me it's not the most beautiful thang you ever seen

>>14674090

>that accent

>that loud ass keyboard

>dude's just fucking punching those keys

>video is fucking 50 minutes long

Like, it's cool, I just wish people that make this stuff wouldn't be so bad at it

>>14672202

Anon, if your grass is going to be that thick, I hope you're not fucking rendering anything else in that game except for large, empty fields. What the fuck.

I switched from Monodevelop to Visual Studio and I'm having a super shit time.

Am I doing something wrong?

VS doesn't suggest me any of the autocompletes Monodevelop did and it also seems to run much worse.

>>14674780

He can always repurpose it into a "Lawnmower Simulator" meme game if the real game doesn't work out.

>>14674788

Fuck IDEs to be honest. Way less headaches (that are not caused by you) if you just use something like Notepad++ or Emacs and compile from the command line.

>>14674838

>negative hp

Spooky is this intentional or just temporary work in progress thing?

>>14674893

>Way less headaches

How so?

I usually use notepad++ for a bunch of other stuff but for Unity an IDE just seems faster and more convenient.

There's a lot of shit that you'd have to remember or google without suggestions, that wasted time adds up in the end.

>>14674898

>temporary work in progress thing

now it's fine

>>14675043

I like it

>>14675043

It's simple but it works. I have nothing bad to say about it.

What is it for anyway?

>>14675336

I want to remake the old popeye nes game in godot.

>>14675043

His right arm bulges noticeably on some of the frames.

>>14674893

i've hardly had an issue with anything from jetbrains

everyone that i know that has a hate boner for IDEs really mean that they hate all the bastard forks of fuckin eclipse there are

>>14674788

VS requires a bit of setup, google "improve VS performance". There's a lot of features that are mostly useless and slow shit down, like CodeLens, Diagnostics and so on.

https://mspi.es/blog/5-Performance-Tweaks-for-Visual-Studio-2015-and-large-solutions

>>14675351

Everyone that i know that has a hate boner for IDEs never had to work on code he didn't write himself.

>>14675572

Try using VS 2008. You might get on with that better. That was the last version before they "upgraded" the UI using their new frameworks (many of which are now deprecated by ms anyway).

>>14674788

Are you using https://docs.microsoft.com/en-us/visualstudio/cross-platform/getting-started-with-visual-studio-tools-for-unity ?

Also try VSCode and grabbing the Unity plugin there, it might run better.

>>14675726

I think is much better than what I was able to make before, sure, it may not be perfect, but is what I can do now.

>>14675594

>>14675726

I'm not an expert on animating, but I think this vid might be useful for you guys.

>>14675755

Could be. Just thought it might be helpful to point out what struck out to me.

Maybe try to really get something solid down with four frames. One for each foot just as the heel touches the ground in the front, and one for when the legs are passing each other in the middle.

You want those to really read well, and you don't want to ease into them.

In-betweens can make everything better, but if you put them in the wrong place, it can make something like a punch look like a push, or maybe even make it look like the guy punching is actually slowing down the closer he gets to contact. At that point you're better to have just two frames, one anticipating the punch and the other in the final position: not contacting his target, but in the place where the target was and the target being already pushed out of the way.

>>14675794

>>14675821

>>14675864

I guess animation isn't about the more smooth and fluid it looks.

I though choppy animation looks bad.

>>14675870

More frames != better, the Japanese realized this shit ages ago.

>>14675870

If the animation conveys what it's supposed to convey, then it doesn't really matter too much. The more frames you have, the more nuance and character you can inject, but you'll also have that many more opportunities to introduce jitter, distracting details and fuckups in general.

Two frames can show you where something was and where something is, and the pause in between will tell you how long it took. Add in a third frame, and you can tell if it's linear motion if it's smack-dab in the middle, accelerating if it's closer to the start and slowing down if it's closer to the end, assuming the framerate is constant. How far it is shows the extent. You can emphasize that energy by distorting, stretching and squashing as well.

But as you throw more and more frames into the mix, you'll have to be mindful of the speeds at which things are moving and how they change shape, otherwise something that's supposed to be accelerating will appear to be jerking around at weird speeds before finally making its way to the end, all while some other completely extraneous details seem to grow and shrink in size for no apparent reason.

If you just have a few frames and they're passable, it'll give you all the illusion of movement you need. Or if you really need that flair and nuance, buckle up and do what you have to to get it.

Or, uh, more frames make for a double-edged sword.

Bad extra frames will take away a lot more than good extra frames ever could add, and there are infinitely many more ways to make bad frames than there are to make good ones. Hedge your bets accordingly.

>>14676010

I think that's just the nature of programming as a skill. There's just so much to learn that you will never be at a point where you can be satisfied with what you write.

>>14676077

That may be true, but there is a fuzzy line where over 50% of programmers would agree code is maintainable and therefore decent.

The biggest problem is the naysayers who advise caution because they themselves are uncomfortable with language features. Or they advise that you go in and refactor afterwards and write bad code up front, because running through all your integration tests to determine whether your drastic refactor to your now-brownfield code broke anything is a totally sustainable practice.

>>14676144

refactoring bad code exercises your ability to utilize your better knowledge

the advice to make a mess up front and fix it later is to the people that are scared of the water and won't start or the people that over-architect everything and then never actually build it

>>14675043

That's a lotta wibblewobble for a sprite with mostly static body-parts.

>>14676172

This is starting to look pretty good, although I would try to put a bit more motion or focus on Popeye's arms. Maybe have them swing forward a bit more? They're his defining feature after all.

>>14675351

I’ll second jetbrains, and VS is a nightmare even with performance tips.

>>14676172

Starting to look good.

Look at the 12 principles of animation. The Alan Becker vids are really good, and Popeye has Disney style animation. The principles are from the horse’s mouth, as it were, and are useful for even 3D too.

>>14676178

>disagreement noise in video

What the fuck is wrong with you

>>14676521

Looking good, but his right arm shouldn't snap forward so quickly.

>>14676521

His torso should be moving up and down.

>>14676588

Forgot to mention that these happen to showcase what happens when you have separate limbs or entities occupying mostly the same spot in the same pose. If you look at the right spot, it's as if one of the legs doesn't move at all or they kind of swap places or bounce around at random. Which is generally something to avoid in the final product.

>>14676434

dishes like you belong in the trash

>>14677205

also treasure chests

>>14677205

Anon what's this?

t. newfag to agdg

>>14677356

>"i'm not sure what genre it really falls into"

>Visible, unspoilered skeleton.

You know perfectly fucking well what genre it is.

>>14676633

The arms lack the same snap that the legs do. Maybe exaggerate their motion a bit?

>>14676172

>>14677170

>still just shilling

>thinks agdg is meant for only shilling

You have to go back

>>14677440

bitches and whines instead of producing a commment worth posting

>>14677443

what if i told you all the waifus were boys?

>Nigger are you making valkyrie chronicles?

>Are you not aware that this already exists and is an established franchise

>>14677502

>>14677495

Please go back to CDPR

>>14677495

How's Cyberpunk 2077 coming along?

>>14677495

>>14677502

Would've thought that an AAA dev would never out himself here because you could get in trouble for associating with the darkest reaches of the internet. Guess things work differently at a polish studio. Either way i hope you have a lot of different tanks in your game to reach the uncountable WW2 autists . Focus on the german and russian tanks especially, you have to have the Königstiger and IS-2.

>>14677546

It's /agdg/, eventually you'll have to reveal part of your identity, that's not the problem

>>14677502

I'd really appreciate if you would be quiet

>>14677629

How are you going to deal with cute animu girls getting shot and blown up in your game?

>>14677696

LAWSUIT

>>14677848

>9 days

No anon, it's two weeks. I only have 9 days off during that time.

>>14677852

Woops, probably should have read that post properly.

Now I can put out half a demo instead of no demo

>>14677356

Helmet HUD… grainy graphics of 3D space… dungeon… I like it.

>>14677546

dude who the fuck cares? at least he is making a GAME, go work on yours

>>14677557

I hope you're not fucking around with us, i really want to believe that there are still AAA devs out there that aren't high on soy.

>>14678154

>shill on 8ch after shilling on reddit

>wow they're so cool

What's a better method of input handling? Polling or events?

Personally, I like polling, but events seem handy in some situations as well (eg, handling text input). If you click outside the game window with events, it will act "sticky" even while the game is out of focus.

Usually I process the input, then map it to a game state (eg direction=move_left, or action_1=true) so I only have like 6-10 meaningful action flags that get set when input is done

>>14677205

what is next a spell of sealing that if it fails it opens the door?

>>14677553

>>14678093

I for one also agree that one shouldn't reveal their identity and talk like a redditard. We have plenty of devs that abide by these rules. Don't make exceptions.

>>14677376

>>14677495

>professional

>posts in an /amateur gamedev general/

>writes like a redditor

Get the fuck out and go post on reddit or shut the fuck up

>>14679026

I have three criticisms.

1. Her ass has a "plastic surgery" look to it.

2. Her head weird.

3. She's walking like a damn bimbo.

Besides that, great work anon.

>>14679133

>Her head *is weird.

>>14677516

> Focus on the german and russian tanks especially, you have to have the Königstiger and IS-2.

what about Maus, Ratte and Monster?

>>14678310

maybe

>>14679203

sick aK

>>14679203

The guns look somewhat realistic yet the gameplay looks quake-ish

How do you guys get out of ruts? I've been slaving away at my game doing everything as my friend who is "Helping" me is also a programmer but doesn't know how to do anything and isn't willing to learn Blender or anything like that.

I've gotten to a point where I feel like I am stuck, I did work for an entire sound manager, an inventory system, and have basic quests working but I just feel so overloaded and don't know what to do.

>TL;DR Feel like shit as I do everything, how to stop being in a rut and stop slacking?

>>14679570

cut content from the game so its easier to manage

>>14679570

Get a friend that compliments your own knowledge. I'd be in a rut too if I'd have to learn blender while not knowing how to draw, rig or animate.

>>14679570

How do you manage your project? Do you have a GDD or something similar that outlines what exactly has to be done? Do you use a project management software of any kind, so you can actually track progress?

Having a plan actually helps, but having it "in your head" doesn't count.

>>14677170

>Also subscribe to m filthy degenerate mailinglist so i can send you an email at some point when this shit is finished.

I subscribed because I want to know when games from here get finished, but you really don't need to try so hard. A simple "Hey guys, if you want to get notified when my game is finished, you can subscribe here" would be enough.

>>14677170

Neat. Always cool to see people from the wider world.

>>14679915

Kirkcode released his 3D platformer which I think is the most "game" game that's been released by 8/agdg/ so far.

>>14679915

Guy who made Deadbolt was from AGDG.

>>14679915

Speebot.

>>14679935

Who?

>>14680140

I think he founded or works as part of Hoopoo Games. Deadbolt was suggested to me because another anon was telling me an anon made it, and it's a pretty good time. It's like Gunpoint and Hotline Miami were mashed together and it feels pretty tight. Can recommend.

>>14680129

it looks great!

>>14679915

Hellbreaker and Speebot were released on steam.

>>14677502

>what if i told you all the waifus were boys?

That would explain the hip-shoulder ratio

>>14680561

It is super cheap I agree with you there.

We've so far looked into a bunch of solutions for it.

>Crash 2 version

This was the first one we thought of but it just didn't look quite right and we couldn't get it to pop out enough.

To fix this we thought we could put DB's hat on top of the dirt pile but I still don't think it's enough.

>crash remake

This is what we currently want to go for, unfortunately I cannot understand how the hell the particle system in unity works. I tried a few months ago to get a similar effect but I could not get it to look right.

I'll be honest I'm not even sure if I should use the particle system to make that effect.

That's the next thing to fix though so I'll try to get it to look right. Thanks for the video references anon

>>14680607

I've been avoiding it for very long time but seeing how the other method doesn't work half the time and the time it does it's buggy as shit due to sheer amount of special cases, it's time I bite the bullet.

>>14680607

the fuck am I looking at here

>>14680635

fuck man thank god I'm too dumb to enginedev

>>14680129

I love the way this looks, I think this can definitely go somewhere. I think a demo level full of just those little cool sequences of getting a good drill mobility combo will really feel amazing.

>>14680647

Yeah I mean fuck me, implementing an actual physics engine is way easier than this.

>>14680635

>>14680607

Why not call lerp functions and test the circle against that? It'll probably tidy the code up a ton

>Inb4 muh performance

>>14680659

It's all about accuracy and consistency. But also you'd need to run these tests possibly millions of times per second in my application to at least get acceptably shitty accuracy.

Other than pirate, surfer and vampire what are the options for silly accents that can be rendered in text only?

>>14680635

wouldn't it suffice to simplify the circle to a bounding square oriented with sides parallel and 90 degrees to the line? find the clash between the vector from circle's starting position and the line, then figure out which point of the circle trajectory line is radius away from the other line and redraw the circle there again.

>>14680899

nevermind I'm a fucking idiot

>>14680909

bit weird, looks like he's sliding

>>14680774

>Vampire

That's basically Romanian, isn't it?

There's also Irish, Scottish, British, bad Chinese, etc. Also look at the DQ series, they have lots of that sorta thing

>>14680912

is a backward walk.

>>14680909

EXAGGERATE. HIS. MOVEMENTS.

The sprite is tiny, an his feet need to LIFT off the ground, but from what I can see, they're just sliding.

Does anyone know how to destroy single particles but keep the emitter itself?

I can't find anything about it

>>14681163

Using what? GameMaker? Godot? Unity?

In general, you can set a particle's life to 0

>>14681163

I'm not entirely sure I understand you. Couldn't you just keep track of the emitted particles and call their destructors?

>>14681174

Unity.

>>14681175

Well I have all the particles, I'm cycling through them and setting

m_Particles[i].remainingLifetime = 0.5f;

Once the loop is done I launch

m_System.SetParticles(m_Particles, numParticlesAlive);

However nothing changes. I'd rather change the lifetime so I can fade them out rather than having them just disappear. Should be the same thing either way, problem is that I can't apply changes at all.

>>14681190

Full script for reference I guess

        int numParticlesAlive = m_System.GetParticles(m_Particles);

        for (int i = 0; i < numParticlesAlive; i++)
        {
            //m_Particles[i].velocity += Vector3.up * m_Drift;
            if(Vector3.Distance(m_Particles[i].position, this.transform.position) > 2f)
            {
                m_Particles[i].remainingLifetime = 0.5f;
            }
        }

        m_System.SetParticles(m_Particles, numParticlesAlive);

I've literally copied the unity example from https://docs.unity3d.com/ScriptReference/ParticleSystem.GetParticles.html

but it still ain't workin

>>14681206

>>14681190

Nevermind I discovered what I was doing wrong.

Even if I set remaininglifetime at 0 the particles won't destroy themselves immediately and will just keep getting reset to a remaining lifetime of 0 essentially making them infinite.

>>14680845

As you should see, the intersection points are defined as the points where distance from point interpolated on vector to line interpolated between a pair of points equals to some specific constant. Geometrically it's the point where perp cross product between circle point to line's starting point vector and line direction vector, divided by line direction vector length. It's basically doing exactly what you said except without an extra layer of complexity of using a square.

>>14679026

make the boobs bigger, you know you want it

>>14681269

i have a bone for that

>>14677557

So it's not coming along well. Noted.

>>14681510

>Gun literally can't shoot, the barrels are completely solid

I know it's a placeholder, but this is somehow endearing in a way

>>14678959

>Look at Yanderedev.

But he was a huge faggot even before he started posting here. Where do you draw the line of "avatarfagging and crossposting"? For one, you're discarding your anonymity when you start posting your game which is why I stopped posting mine. And two, plenty of anons here have devblogs on tumblr and twitter. If you want your game to reach more people, using other platforms is simply unavoidable. There's a reason why a lot of devs who start out in agdg stop posting, both here and in the cuckchan threads. That's been a phenomenon as far back as 2013.

>>14681510

This really hurts my ears, it's all over the place

>>14681557

well you can't do much more for a jam game. we only have like 72 hrs to make this

>>14681577

but there is time for teaks. Is it about the volume or the track itself?

>>14681617

it's the fucking violin and all the really high pitched sounds

>>14680607

What exactly are you trying to do? Usually quartics aren't involved in gamedev and a serious sign that you misdesigned something.

>>14681667

Instead of Java, use literally anything else. C# looks basically the same and isn't infested by pajeets

>>14681667

If you don't want to engine dev then you might as well get one of the engines that uses c# or just write it in c# yourself.

>>14681702

It's why I'm planning on making a long-form video explaining the process of what went into making my game, and post that there, alongside maybe the actual trailers and stuff like that, but never otherwise be involved. Seems like Twitter is probably the only other place where I could make progress updates, though it's impossible to get a following there without already having a game under your belt.

>>14681722

I've heard good things about Gamemaker, but I've also heard a lot of shit about it.

There's also Love2D but it might not be in c#.

>>14681722

>>14681732

There's also SFML which is good and what I'm using, but it's not an engine, but a low level framework.

>>14681732

Yeah, Love2D is in lua, but it actually seems pretty interesting, I think I'll use this to start. Thanks.

>>14677170

>So i made stuff for a tank cleaning minigame!! hope you're ready for fucking tank-cleaning minigames, washy washy brushy brushy. Also subscribe to m filthy degenerate mailinglist so i can send you an email at some point when this shit is finished. go here you fucking weebs: www.shimapanzer.com

So is that what people over at reddit actually like to read? You're still shilling, if you actually lurked longer you'd have realized that you're the only person posting mailing lists, links to their websites, and promotional material exclusively and regularly.

>>14681682

What it says on the tin. I need to check for these 4 points where the circle touches a moving line, and that's a quartic equation.

>>14681942

To be more specific. The circle is an enemy's hitbox and the beam is the weapon's hitbeam.

>>14671290

>>14672080

I'm not a fan of cinematic experiences, but this shit is useful to me as a 'writer'. I already have The Story.

>>14681949

>>14681942

If it's a circle, then that means if any part of it could intersect with the rectangle (whether AABB or not), it will be guaranteed to be a distance less than or equal to its radius

>>14682002

A simple approach would be to define your rectangle as rABCD, then check AB, BC, CD, DA to see if the enemy distance is within radius units from that line segment.

No derivatives or roots or whatever it is you're doing.

>>14682002

No look the object is to find exact lerp coefficient T that corresponds to the time where circle touches the beam. These are points in time where the enemy starts and stops taking damage from the beam. I'm not doing frame by frame calculations, instead it's an event-based logic and frames are only there to render shit. Put otherwise, it's variable tickrate logic when ticks only occur when something changes. That might not sound like a good approach but it's perfect for handling mid-frame occurences, which would be plentiful.

Also, clearly just computing cross product is simpler than doing all that shit you're suggesting.

>>14679786

I had GDD, kinda lost it, I also have a mindmeister map that I have a bunch of stuff written down on. Most of what "Needs to be done" is written down on that, but things like the feel of the game, the gameplay, and some psudocode is in my head. I guess I should put that on paper.

>>14682058

I personally recommend a program called WikidPad. It's free, and it allows you to keep wiki-like notes about various shit. It's really great, I have my whole GDD written out in that format, and it's super-organized.

>>14682064

I'll second WikidPad. The ability for it to have a bunch of self-contained references lets you rapidly link ideas and pages with minor amounts of formatting and style. I keep all my various half-baked ideas in there

>>14681713

I think you still need a large central mass for the smaller rock trail to emit from, since it's the large body that is leaving the wake. The heading on the cubes is a little bit hard to gauge because the cubes spawn from the player's center and not his frontmost position so that will be a weird thing.

Other than that it looks pretty good.

>>14682064

>>14682099

Thanks, I'm trying it out now, I'll have to watch a tutorial or something, can't even figure out how to add a new page to a wiki

>>14682191

>Watch a tutorial

>For a productivity application

Jesus anon, just open the help file. It tells you that it automatically creates and links pages

>>14682092

Doesn't make any sense and looks like troll code to be honest. How can any member of Asset do anything with T if it has no constraints on T's interface? I know this without seeing your code because you don't use the where keyword. You badly need to learn how OOP and generics work. Go right back to the beginning.

>>14681800

Dubs of truth, anon.

Well said

>>14682334

https://hastebin.com/noculuvoqi.cs

Here's the code in full. And no it's not troll code. I couldn't put a type constraint on T, because I can't (or don't want to) guarantee that T should implement IDisposable.

For example, most of SFML's objects derive from ObjectBase (which is IDisposable), so it would be called in this case. But what if I (or someone else) makes a custom type that doesn't? It suddenly shouldn't fit in a simple POD object because it's not IDisposable?

Asset itself doesn't do anything, but it's a container that goes into an AssetManager<T>, which holds 3 dictionaries https://hastebin.com/ejokecesit.cs which is mostly just a helper to hold shit I load in (textures, sounds, etc), but without putting any conditions on what it can actually hold

>>14682350

Just popeye giving a punch to bluto and receiving the hearts until you win.

less like the original game and more like a fighting game.

>>14682382

what the fuck is this

>>14682442

See >>14680635

I believe it's circle/square collision code.

>>14682456

Ah, makes sense. Interesting stuff.

>>14682382

do you got some sort of article on the math of this? I need some refresher to keep up here, my lineal algebra is a bit rusty

>>14682571

No, it's all derived from the linear interpolation equation. Circle's motion is p = p0 + v * t, where p0 is origin, v is direction and t is time. Line's two points are a+(b-a)*t, same exact thing except it derives direction vector from the two endpoints. Then it those are used to compute points where the value of perp cross product between direction from circle to line start, and the line direction, is a specific value adjusted by line length, to convert it from barycentric to unit coordinates. And from there you get more and more complicated shit until you're left with something that looks like a standard solvable equation, quartic equation in this case.

>>14681800

>Avatarfagging is another issue.

>As long as the other posters recognize you by your content, everything is fine.

>It's when you try to ensure people know it's you for no reason that you're met with hostility.

I agree. And the same thing goes for flagfagging, it's basically the same. We've got some posters here that do it since forever, get rid of that shit >>14679850

All it does is turn this thing into IRC/disagreement cancer sooner or later, there's literally not one good reason to do it.

>>14682627

I speak french, spanish, italian and english.

But man, as someone that knows languages, I gotta tell ya, translation games really don't work

>>14682618

You seem to know your stuff and be helpful enough so I guess I'll share my confusion here, maybe this makes others see it as clear as you do too. My math education is all in Spanish so there's that.

You define the circle's position all right, with p being a position vector, p0 being another position vector (center of the circle at the beginning of the simulated interval, I guess) and velocity being another vector, t there is scalar. Now, for the line. You don't define an equation so the tree ways I studied to define a line (point and director vector, implicit and parametric) don't apply since they require an equation with an equals somewhere (several in the case of the implicit and parametric forms). Now, interpolation means roughly "get given two positions and recreate the position in between", so far that I know. That mean that the line is defined by vectors position a and… (b-a)? that'd be… a being taken as the point in the line and variation from a to b as the director vector? it being multiplied by time hints that the direction of the line varies with time and that you are interpolating the positions at which the circle would hit the wonky line that moves around anchor point a then. Fair enough. Now comes the hard stuff.

"perp cross product" is something weird to me. I was taught vectorial product (the one that gives out a vector perpendicular to both used as factor, with length equal to the product of the length of the vector factors by the sine of the angle their form), scalar product (first coordinate of first vector multiplies the first coordinate of the second, and gets added to similar products for all coordinates to equal product of vector lengths by cosine of angle, the result being a scalar number only), and the mixed vector doesn't apply here so what gives. To make matters worse, the perP I've researched lightly apparently means per point, but then has "cross" in the name, and our mathematical notation used point notation for scalar product and cross notation for vectorial product. Hence my confusion. But I think you mean vectorial product since you later on talk about converting it from barycentric to unit coordinates. So now, no idea why is it useful to get the vectorial product here for anything other than defining a plane in 3d space that contains both vectors (since the result is perpendicular to both vectors, useful for defining planes). In the case you mean the scalar product, I seem to remember a useful way to compute projections of vectors over other vectors with it alright, but that's all I remember being useful for calculating distances to stuff (we just used a^2 + b^2 = c^2 for all our distance needs). I hope I didn't bore you to death with all that.

>>14682687

Line is defined as a+(b-a)*t = 0. The b-a is simply a vector of direction in which the line extends from point a. So you can define it as a+v*t = 0 and that makes no difference whatsoever. The feature that separates line from ray from segment is the range t allowed to have, for practical purposes it's left open here, so technically that's an infinite line.

>perp cross product

I meant perp dot product. Cross product is only defined in 3 and 7 dimensions. Perp dot product is simply a dot product where one of the vectors is treated if it was perpendicular to itself, i.e. rotated 90 degrees. And coincidentally it shares the same looking formula as cross product, it computes vectors' matrix determinant, except in 2 dimensions rather than 3.

>>14682730

Oh, and perp comes from perpendicular then, OK. I didn't know.

So, you choose the circle center direction vector, flip it 90 degrees by choosing any same-length vector that dot product'd by the first one gives 0 (I think the trick was to exchange the coordinates and change the sign of one of them, for 2 dimensions) and then you dot product it by the vector of the target line. I understand the procedure but fuck me the logic is not so easy. So the value would be (assuming unit-length vectors) sine of angle shifted 90 degrees. Which would give you cosine of either the right angle or the angle shifted up/down 180. So far it does seem you are going for the "projection of vector A over vector B" (A being maybe circle center direction and B being direction of target line). So now that you know how oblique one projects on the other, you can use a^2 + b^2 = c^2 to add the condition of the radius distance, and solve that way… maybe? sqrt( ( p0x + lambdapx - ax - lambda(b-a)x(perp_dot_product(p, b-a)) )^2 + ( p0y + lambdapy - ay - lambda(b-a)y(perp_dot_product(p, b-a)) )^2 ) = radius, maybe? Been brushing on my maths a couple of months now, I need this

>>14682848

aw fuck no, you gotta choose an anchor point for the position and find the perpendicular point in the opposite line to have my equation work. Replace ax and ay with… fuck, err… you gotta do another equation here, equal the expression of a + b-a*t to p0 + 90_degrees_shifted_p and solve for the new p0_projection_x and p0_projection_y, then put these in replacing ax and ay and I think you are golden.

>>14682879

The only problem is that the line is not unit length so you need to divide the whole thing by square root of sum of squares. To get rid of the square root you square both sides. Then you get quartic equation.

Hey AGDG can you make me a sandwich?

>>14683155

Fuck off, Carlton.

>>14683222

Bottom left is clearest. Bottom right's corners are a little too rough in comparison, I think.

>>14683315

I mean I'll still have a few to choose from, but I was hoping that'd be the case. Thanks.

>>14684080

I mean you don't have to be fat to enjoy ice cream. Most normal people buy carb nightmares like that when they want to treat themselves, which is fine.

>>14684080

>being a soulless husk that can't even enjoy ice cream

fag

>>14681800

>Crossposting isn't the problem

I'll disagree there, anon. I came here to ESCAPE the bullshit mods on 4chan, why on Earth would any self-respecting anon want to post there and have to deal with them?

>>14684878

How much flouride did you put in?

It looks delicious

>>14684897

I need to learn more about life in Eastern Europe. That stuff is fascinating.

>>14684910

I'd love to answer more shit, but I think we should keep to just one extra post, because it's offtopic. Let's call it, uh, discussion about how vidya can be authentically set in Eastern European settings. Pile on whatever you got into a single post. I love remembering the special things that made my life in Russia great. I'm just past the point now where nostalgia is starting to set in and it's super comfy to think about.

>>14684915

Who can blame them? There were really just two camps of people when the USSR fell apart - those who loved it, and the simplicity it brought, and those who realized that it was a commie cesspool of corruption. The latter really pushed hard for modernizing, swapping out the old and stagnant shit for the magical wonders of the West. When I was a kid, America to me was like a wonderland - I think I fell in love with the idea of United States when I watched Home Alone as a child. Shortly after, I went to a friends party - she was an American immigrant (I think she was an ambassador's kid) and they invited our entire class to the Embassy for Halloween, imported candy and all. The sweetest day of my childhood. Anyway, I digress - but imagine that sort of impact on every single citizen of the USSR when they were first exposed to what the West had to offer. No wonder there's barely any old stuff left.

>>14683772

>>14683971

what did you see anon, what did you do

>>14684940

Your entire project is a single file?

>>14684994

I just want to create a function without having to link and update it everywhere. I feel like it would be easier to put everything into a single file.

>>14685043

Easier almost always means broken and fucked.

>>14685061

Your shit is almost certainly still only one file no matter how much text you write to obfuscate the idea.

>>14685095

No it isn't. Each file compiles to its own binary object, linker then stitches them together as necessary.

>>14685182

>each file

lol

>>14685345

It's not about how you act incidentally as much as it is about how you act on average. I'm sure a lot of us had autistic fits, I know I personally did - but Yandev didn't only have spergouts and fiascos, he refused to recognize his own mistakes and blamed it all entirely on his audience, the anons, and etc. If he just came back online after the first, second, third time and was like

>"Okay guys, I was being a complete faggot and this is why I was a complete faggot. Apologies to specific x, y, z etc.

Then anons would still tease him but he would at least get points for self-awareness and ability to reflect on his actions. He'd recover and he'd grow from the experience. But no, he went the Phil Fish route.

>>14685407

whoa, rip

>>14685581

I should add that any of the Rotator values can be between 65535 and -65536.

>>14685581

How much game do you actually have?

Cause I swear to god I've seen you post in /agdg/ for the past year+ and I'm fairly sure I've only seen sprites and animations despite you talking a lot about code

>>14685658

I have what I had written in Unreal along with the newer Godot stuff. I rarely post. I've translated a player, a controller, control binding, the console UI provided by demo, a "map", learning lighting, created a debug HUD, and learned the basics of Godot 3 the past 3 months. Progress is slow because I don't dev every day because I run into issues I don't know how to solve and have to wait for answers. Or I'm busy or just don't dev. This is a hobby at this point because I just don't have the education for how to do advanced stuff in game engines. Also have you not seen my latest video, utilizing physics, collision, and lighting? Even better, if you want me to advance and have any knowledge on my coding question then feel free to share.

>>14685714

>actor

You're using Unreal? If so, it should only generate one hit event. If not, and you're using your own engine, the depenetration should happen in one step and only one collision should be registered. If that's not an option for whatever reason, you could try scaling the damage by the impact force (or the speed of the rock at the time of impact). Of course you need to define a threshold below which no damage is applied. You could also give your rock object a "CanDealDamage" property and set it to true after the damage was dealt. That way each rock object can only deal damage once until it's reset (which would happen when you throw). The last one is probably the easiest solution.

>>14685737

I have not seen your video, feel free to post it.

>I can't figure out how to do the math that would determine the facing angle of a node relative to my camera

I'm going to suppose from the code you posted that you need to find the actual angle in numbers?

I've got the solution on how I did it for my game, however that was almost a year and something ago, and I didn't comment it, so I only kind of remember how it works.

You want it?

>>14685789

>>14685737

I mean it's not exact of course, I'm using Unity, but I can try working it out for you.

Gimme 30 mins

>>14685762

>You're using Unreal?

I am.

>it should only generate one hit event

It does, but not always, if the rock bounces around on the actor (hits chest then onto arms) it returns two, this isn't really an issue since the second hit is weak and is filtered by the threshold but when the hit actor moves as well, then it gets spotty as the movement causes the second arm hit to be pretty forceful. Mostly I'm worried about what happens when you start mixing thrown actors, movement, and level geometry (e.g. the rock being stuck between the actor and static geometry bouncing back and forth at the considerable depenetration velocity).

>treshold

Yeah have that implemented, bumped it up a bit and it seems to mostly fix the issue, I guess I'll just leave it as "good enough" for now and if, in the future, double hits are still a problem, I'll just implement a timed ignore or boolean for it.

>>14685799

Use my second solution. Add a property to the rock which defines whether it's used up or not. When you throw, you set it to false, when it impacts, the damage is applied only if it's false. After the damage is applied, the hit actor sets it to true.

>>14685581

Okay so here's what I got.

I tried doing it using Godot's methods but I cannot understand how the fuck you read godot's docs, so you're going to have to translate that yourself.

All the logic is there however so you should be alright.

    public void HelpingAnon()
    {
        // AxisSign is what we use to differentiate which side the camera is facing, it's not THAT important if you don't want two sides of the sprite to be
        // different but you might as well have it.
        // The resulting Vector will either have it's Y value positive or negative depending on the direction.
        // Ex. (0.0, -1.0, 0.0) if we're standing at the left of the sprite

        // To find this we use the Vector that represents the forward direction the Sprite is facing and we Cross it with the forward direction of
        // our camera. Of course, if your camera is pointing down a bit, you'll want to zero out the Y value to make sure that's not being considered.
        Vector3 gameCamForward = gamecam.transform.forward;
        gameCamForward.y = 0.0f;
        Vector3 axisSign = Vector3.Cross(referenceBox.forward, gameCamForward);

        // Now, since we only use Axis Sign to determine if we're facing the sprite's left or right side, we can just map whatever values we get to either
        // 1 or -1. So if Y is bigger than 0 (say something like 0.5) we just say it's 1, in the other cases it's -1
        axisSign.y = axisSign.y >= 0 ? 1f : -1f;
        
        // Then, we transform the directions our character and camera are facing into angles by using the same forward vectors we used during Cross
        // This gives us 0 if we're facing the same direction
        // It gives us 180 if we're facing straight on
        // It gives us -90 if we're at its right
        // and 90 if we're at its left
        float facingAngle = Vector3.Angle(referenceBox.forward, gameCamForward) * (axisSign.y);

        // We then take the angle and dive it 180/7 (180 is the highest value, 7 is how many values you want to have in total)
        // This will give you a range from -7 to 7 depending on where you're looking, which you can use for the sprites
        // If you don't want a range this high you can just remove AxisSign and you'll have a range from 0 to 7
        int spriteToDisplay = Mathf.RoundToInt(facingAngle / (180 / 7));
    }

Hopefully Godot3D has something resembling forward vectors?

Let me know if you can't understand something

>>14685789

This is my work in Godot 3.2 so far, I'd been having a hard time grasping the concept of how to search for instanced nodes in Godot's tree system. I still don't know how to instance a new node that isn't already in the tree before the world is loaded or whatever calls func "_enter_tree()" on the very first node of the tree. Also I'm not sure where the root is, if it is before the first node or is the first node. I've heard it said as one or the other and I have no idea. On top of that, I had a hard time understanding how transform worked for nodes until I had the idea to just group certain nodes together in a packed scene, instance them at vector3(0,0,0) and just create a spatial node for them to move to and become relevant at. I'll need to find out how to instance nodes on the fly in the future, or at least learn how to create a spawner.

My current issue is this angle math, but I'm learning. I have yet to get to cross product, it is later in the script I'm working on. Here are some notes on modifying player movement without the rotation code, which controls whether the player applies forward velocity complimented with gravity and can have an anti-gravity force applied:

# ================
# Handle Player Physics
# ================
func _physics_process(delta):
	var lv = linear_velocity
	# velocity

	var g = Vector3(0,-9.8,0)
	# gravity
	
	lv += g*delta 
	# lv = (0,-9.8,0)*delta
	# linear velocity, gravity applied
	
	var up = -g.normalized()
	# against gravity

	var vv = up.dot(lv)
	# Vertical velocity, dot returns -1, 0, 1, or between

	var hv = lv - up*vv
	# Horizontal velocity

	var hdir = hv.normalized()
	# Horizontal direction

	var hspeed = hv.length()
	# Horizontal speed

	var dir = Vector3()
	# Modified by input

	acc_forward *= 0.4
	acc_strafe *= 0.4
	acc_up *= 0.4
	# Input values, assume acc_ whatever is 0 or 1.
	
	dir = Vector3( acc_strafe*get_transform().basis.x +
		acc_forward*get_transform().basis.z + 
		acc_up*get_transform().basis.y )
		# Where does the player intend to walk to

	var target_dir = (dir - up*dir.dot(up)).normalized()
	# target_dir = ( input - anti-gravity * input.dot(anti-
#				gravity).normalized()

Also gonna test something on this post:

[cpptext]test[/cpptext]

[c++]test[/c++]

[java]test[/java]

>>14685714

just reduce the damage by DesiredDamage/NumberofHitsOnAverage

>>14685970

I shall take a look at this, check >>14685989 .

Also the docs are written so annoyingly and full of this standardized, college-age young faggot appealing writing style that can be easily found in most tumblr-approved web comics. It is almost written like each page is a PSA for impressionable young people with how soft it's presentation is, or at least that is my impression up until a little past "babbys first game". They are useful, but god I hate that writing style.

What's a good way to do collision (2D)?

Currently thinking of simple bounding boxes.

>>14685989

It could be because I'm used to unity but Godot seems so needlessly complex for no apparent reason.

Cool shit though.

>>14686028

It's not even that, it's just that they literally seem impossible to navigate. Like, why is all the Vector Math in one single page? Why not have different pages for each?

I dunno it doesn't make any sense to me

>>14686109

It could be worse. Try looking up documentation for Unreal Engine 1, use keywords "unreal" "wiki" "ut99" "Unrealscript" "ut99" "beyond". That information is scattered to the wind, dead links for eons.

>>14686074

Basic axis-aligned bounding box collision's easy enough to implement, if you're looking for something simple.

If you want more detail with character-on-terrain events, you could look into using hotspots. Basically a set of spots attached to the character that act as its collider. You get the benefit of knowing when, say, one leg is on a slope or off the edge of something, or when jumping up and barely touching an overhang on the corner, or in general know which hotspot or edge is doing the touching. Easier to get proper corner response and walking up slopes working with a setup like that.

If you want to get fancy, you could try sweeping out the motion for every frame and get the exact position where one entity were to stop just as it hits something. Might be useful for very fast movement or projectiles. May involve raycasting and reading a whitepaper or two.

Maybe start with AABB and be mindful of off-by-one errors if you just want something working now. You'll figure out your needs later, and how much optimization you might want to cull out unnecessary checks.

>>14686471

Entry database exam? Like details about implementing your own database, database design, interfacing with an RDBMS, or what?

>>14686614

they give us a 90% working website where you can add/edit/delete cats from a database

you have to figure out what's broken and fix it

i did php and c#, tried javascript but couldn't figure out why the cat wasn't initializing, and i didn't have time for java

i do in theory understand them though

>>14686737

Looks like homer with Bluto's face lmao

>>14686737

I don't know why but the mouth looks amateurish. Maybe it's the uneven width of the teeth.

>>14685714

i would go with white with no clear outline, if you had a brown or grey just to have a transition between beard and mouth outline and take the uneven teeth down to an off white yellow color