/agdg/ + /vm/ ~ Amateur Gamedev General

What browser do you use? It's not respecting the image rendering CSS, that's why it's blurry.

You need to click the area where the text is at the top, itch.io for some reason doesn't want to give the window keyboard focus if you just click the canvas.

The original idea was to make it a clicker game, where you click to add sand, and then use the piled up sand to buy upgrades.

I lost interest to it though.

Chrome, I am a good goy

I've pretty much given up on doing a game I was working on, at least in 2D. I was trying to program physics in 3D (falling, height levels, etc.) but it's become so much of a clusterfuck that there's no way I'm going to finish it.

I'm looking at Godot now and I'm trying to figure out how the physics system works. My 2D prototype used callback functions for collisions in Box2D and calculated damage from velocity and mass, and it needs to have the ability to apply impulses and forces. I see there's a pretty robust 2D physics system similar to this but I can't tell if it works in the third dimension from the docs. I really do not wish to use a non-free engine but I might be forced to if Godot is not properly equipped for it. Can anyone confirm or deny that this is possible in Godot?

The most fun I had with a Boulderdash-esque games was Supaplex, especially in levels that made use of the quirks in collision detection. Juking out the enemies by moving away right before they hit you and causing them to consider you a wall and turn around, or riding the enemies and walking on explosions when it was a level with gravity. Those were more common in custom level packs, though.
Just don't make mazes. Mazes suck. Don't make the second level of your game one big maze with nothing else in it. And don't reuse that concept later with pieces that allow you to only go in one direction and turn it into a maze with dead ends you can't get out of, forcing you to try again every time you make a wrong move because not only is it hard to clearly see where you can go because the pipes look similar, but much of the maze is off-screen so you don't find out until after the mistake has been made. Fuck mazes. Shit.

Can't go wrong with trying to incorporate lasers, mirrors and prisms if you need fun stuff to push around and worry about line of sight. The DROD games also had a variety of cool mechanics, some of which might translate quite well to non-turn-based affairs.

The 3D physics should work the same as the 2D physics, except everything has Vector3s instead of Vector2s

Thanks, I'll give it a shot.

Did they guy who made this go down to the ATM Machine and punch in his PIN Number?

what

/amateur gamedev general/ general
Automated Teller Machine Machine
Personal Identification Number Number.

Actually the OP image I used does

What's that make a fun level like this

oops

oh wise agdg, give me motivation to create the hair of my waifu and finish before midnight

I feel so burnt out

There's a fresh progress general in >>>/agdg/ if anyone wants to show their work.

Delete work, start over again.

Not this time mother fucker

This is basically held together by duct tape and prayers.

just leave her bald, it'll be fine

Your loss pal.

...

...

...

cute

cooool

Install a different WM.

ありがとう！！！！

I rather plink around on my Debian 2007 Core 2 Quad machine tbh. I've been meaning to switch distros anyway. Probably Void, Debian or Slackware, if it isn't too meme tier.

Both excellent choices, although Void isn't exactly stable enough for developing on fastest startup speeds I've ever seen though, booted up in 10 seconds on a regular hard drive for me and Slackware's Mesa builds are probably really outdated.

HAHA TRAIT REWORK COMPLETE
Faerie Traits are now fully customizable on the external .txt file
Now i need to implement the two final traits and make a traitNames file for names and descriptions (for easy translates)

see, I'm correct about everything

while we are bashing each other os choices, has anyone here installed xubuntu 17.4 from scratch, any problem with internet or is just me?

Slackware is probably my choice. Though it's no dependency resolution bugs me to no end.

You know what? Fuck it, I'll use void. Much more cool shit going on with it. Have some clang for your troubles.

lolwut
use slackware

But why not void? Slackware has no dependency resolution. Plus, everyone knows slackware, wanna try something new.

Also, I've been planning to make a website. What web framework or way to make it should I do. I heard node.js and PHP are shit.

Last time I used Void, it was really comfy aside from some weird ALSA errors that turned up no web search results.

because i use slackware and love it and you have to too.

Make a static site or use something like werc.cat-v.org/ .

But sometimes I wonder to just use Debian, it's really comfy. I use it on my server.

But what if I want to make a chan?

Also, refer more to sites like motherfuckingwebsite.com/ , bettermotherfuckingwebsite.com/ , and older software library websites as a reference to designing simple websites that load quickly.

Werchan exists, see github.com/kfarwell/werchan

today was a good day.

That's doesn't look like dice, liar.

ca3.tumblr.com/post/160603702874/having-fun-with-unity-and-pbr-shit

Yes, werc is cool, but it doesn't have a lot of documentation. Can you recommend something that has examples/cookbook/docs?

your forgot bestmotherfucking.website/
It goes over
- TLS certification so you are using HTTPS instead of HTTP
- gzipping your webiste so it loads faster
- making use of the cache so it loads even faster the next time they visit a given page

lmao, I need a better PC.

Python/Django. Best choice.

>>>Holla Forums9904233

Why the fuck are you here?

I lurk every thread.

How do I control what happens to an object that gets hit during a Physics Raycast in Unity?

you mean sending code to an object you raycast?

Yep

this is five seconds of google:
"raycast colision tutorial"

>>>/agdg/28961
I need some debugging help. Please.

Do I need to know how to draw to learn to model? I am not very good at drawing but have been wanting to try game dev for a while. I made some simple games like breakout but that did not require any "art".

drawing is the basics of art.

you're asking if learning to play the piano will help you with FL studio.

learn drawing user, is worth it.

>tfw want to shill ur art/muzak on here and 4/agdg but don't want to give any ammo for future autists because you actually want a semi-stable revenue stream

sheeeeit, i didn't know there were people that still remembered VoxelWar! I thought that it'd just suffer the same fate that Iceball did (glacial development speeds)

I really just need to fix a segfault in voxellap.

You'll definitely need it for modelling humans and other natural things.

But if you're just doing artificial objects (guns, robots, etc.), you can avoid it for now.

...

Make it for unreal.

...

What about architecture?

They stalk just about anyone.
Check out their website and blog for some laughs.

It depends how complex you wanna go. Modern buildings are pretty simple to model.

Although if you want good textures, you may want to brush up on the art skills.

watch out for scams my nibbas
archive.fo/nmynC

If this is true, why not get H3H3 or Jim Sterling to make videos about them? Their cancer fanbases will destroy bsm.

H3H3 is being sued by an even bigger fedora faggot because said fedora faggot is more triggered than pic related.

AFAIK, that was for whenever mr. fat asexual british man specifically targeted shitty Steam Greenlight games and the shitty thing(s) around there

tfw no archaeology game where you discover long lost secrets by displacing sand with a shovel

if there were, would 2D like terraria or 3D like minecraft be a better setup?

Pretty neat idea.

minecraft has shitty gamplay mechanics, and terraria is definitely superior; though I prefer exploration in 3D (if it's done well, so probably 2D, as terraria style graphics would be easier).
Also, I'm damn sure terraria uses marching squares, which is super easy to implement; while being very robust in its applications.
Though with the central focus on exploration of old ruins, finding relics/long lost secrets, archeology, and general indiana jones type of badassery (hook shot esque whip usage, cool hat is obligatory, etc).

So, I'd assume it would be like… Indiana Jones, but with some traditional archeology parts?
Sounds fun tbh

Looks fine to me.

3d easily but I personally have never been able to play side-scrollers on the PC. Plus it depends if building shit is going to be an element. I remember I couldn't stand building in Terraria.

How the fuck do you, conceptually, sync shit over multiplayer? Say you've got basic water physics that effect gameplay. That's hard enough on your own without having to make sure it's the same across 5 clients.

You host it on one persons machine and send updates out to the other 4.

Fuck that. Just send the input to the fluid simulation and run it on each of them.

Where are you sending it from

There's 2 methods of simulation.
One is the RTS route, which is slow to respond, and the other is the fast but inaccurate method.

The RTS method is to sync the gamestate completely (this generally means you group up in a lobby and start, and don't have the ability to join midgame), and then only send over inputs. If your game logic is deterministic (=same inputs produces same results every time), then everything will run in sync.

This is actually pretty easy to set up. It translates to:
Set up a server and N clients.
Server sends out a tick event every 1/fps seconds.
When a client receives a tick event, it performs a game tick.
If a player wants to do something (e.g. order a unit), it sends this command to the server.
Server receives command, and bundles it together with next tick event.
Client receives tick event with commands, performs commands and then a game tick.

The advantage of this is that you don't have to worry shit about synchronization, aside from the commands and deterministic game logic. For anything very complex (RTS, physics, etc.) this is the preferred method.
Downside of this though is that it introduces a delay. That delay is created as follows:
Player does a thing, sending a command to the server. (delay = latency to server)
Server waits until the next game tick. (delay = latency to server + (0 to 1/fps))
Server sends the tick, and the player processes it. (delay = latency to server + (0 to 1/fps) + latency to client)

The result is that if you run gamelogic at 10fps (supcom does this), then your latency will be the ping to the server, with anywhere between 0 and 1/10fps seconds added on top, depending on when exactly the packet arrives.
On a 50ms connection, this means you have 50~150ms delay.

The other method is to spam over the gamestate (or at the very least, the variables that determine the visuals of it). If you have objects that frequently change state (like fluid physics) and have a large memory footprint (like fluid physics), then you can't do this unless you're playing over LAN.
You COULD do this if people would stop playing games over fucking wireless and got decent fucking internet. But we can't have that. Nope, everyone buys fiber for blazing fast 100mbit, and then tries to play games over shitty lossy wireless like a fucking retard.

I see, thanks. So if a client does something, say, breaks a block, I'm assuming it'd happen on their end normally, be sent to the server, then the server sends it to everyone but that original client?

The server. What I mean is, instead of sending the results of the simulation you just need to send the inputs. The server just needs to timestamp them.

Have you ever played a single Korean MMO?

That's basically what I I said, except I assumed it was p2p since it's /agdg/ and servers are expensive.

No. It does not happen on the local client immediately.
You have to wait till the command comes back to you to ensure proper synchronisation. Otherwise you'll end up in cases where 2 players break the same block locally for their perspective, but are told later that another player broke that same block.

What you're referring to is Minecraft's shite netcode which is just a trainwreck all in all anyway.
If you want players to locally (& temporarily) modify the game state to achieve better response times, then I suggest you make some sort of ghost object you can locally interact with, which purely exists for this purpose. If you make a deterministic game, then keep it deterministic. Only modify the deterministic gamestate if a server command tells you to.

Factorio, embed related, is a game that runs the way I describe, whilst still giving you a character that you control.
You can pirate it and play multiplayer with a friend to test the lagg for yourself. The following netcode behaviour is noticable:
You have a ghost copy of a character you control. If you hold space, then you will instead control/see the actual character, which has a noticable delay. (This delay is actually pretty significantly! A 30-man server I just tested on sports about a second delay!)
If your actual character gets stuck on something, (e.g. someone places something in its way) then your ghost gets warped back.
When placing structures (which is the majority of the game), you place a ghost-structure with no functionality. The server then either tells you you placed it (replacing it with an actual structure), or you didn't (removing it).
Because the ghost structure has collisions when it comes to the movement of your ghost character and the placing of more ghost structures, it feels as if there is no delay at all involved. Non-ghost characters can you move through it though. (But if that happens, then the client deletes its ghost copy before you see it phase through the structure.)

The thing with Factorio is that you don't need very precise character movement. If your game contains combat that requires any degree of positioning or dodging, then this method is most likely going to be too laggy.

Maybe you should get into more detail in what you want, as well as the scale there-of.

Sounds like shitty design. I'd certainly notice if each of my actions had a 50-150 ms delay.

still reading the factorio stuff

Welcome to netcode.

You either eat the delay, or you get inaccuracy. You don't get to have both, fuck you.

The only thing you can do is hide it by adding a lot of local ghosting bullshit and letting clients cheat. But this is all band-aid shit that makes it only "feel" better.

You can do prediction client side for things like movement to make it feel more responsive but the danger is a poor connection can then have you rubber-banding as the server "corrects" the client as to where you are. Like 465a5d says, you also have to consider how much of the game-state you want clients to be able to alter because anything they can you have to presume they can cheat with.

It's also a balancing act of how hard you want to work the server, scale/scope of your game, etc. For small online multiplayer I just have clients act as relays for their input and handle all the logic server-side so clients can't touch the game-state directly at all.

Reminds me of this: en.wikipedia.org/wiki/CAP_theorem

Whoops, meant for

see:

Fucking shitpost, but the more I think about this. The more I want to actually do it.

What would /agdg/ the language even go want to do look more like? War?

...

Sounds like you didn't need water after all, better think of a new mechanic.

don't leave us when you become successful

shit, wrong image

To the user that told me Blender has a video editor, and everyone who supported me with valuable feedback until now: thank you.
I'm preparing for Steam Greenlight, let me know what you think of the trailer and the website… I'm still open to suggestions to tweak things but I want to finish up today.
kircode.com/speebot/

...

what did he mean by this?

Sure fam

He wants to copy it and put it on his own channel, I'm fine with that.

Those sad fucks that run windows need to know their place, right, fellow unix user? So much ransomware and bloat up the ass.

Seriously though, windows7 is so easy to run on KVM that you don't even notice the slight performance difference. Just check KVM-capable hardware.

Xubu is my goto backup. It really just works. Xubu is hated on pretty much everywhere, but it delivers a mid tier OS with minimal hardware-impact. Meaning it can run fast on literal toasters. Love it.

Site looks good.
A few animations would spruce it up probably, but it's seviceable.

is the zooming (far) out thing part of the game? I find it jarring

What is the difference between godot and unreal engines?
Anyone tried live 2D?

You hold a button to enable it, it's like a mini map

one is a piece of shit open source software, the other is Free (libre) Software.

Sorry dude, I don't know anything about that. Might help if I knew what computer you are using???
But you better go to:

...

I have a fairly similar experience:

Optimizing that "path finding" algorithm is now my job. I fucked my hobby up. Just.

Is Live2D available to the public? I thought it was a proprietary tech used in a few private projects.

Speebot is on Greenlight now:
steamcommunity.com/sharedfiles/filedetails/?id=924909034

What's a good tool for recording webms? OBS is a whore.

Sound effects are a little on the low side. Which wouldn't be too bad, but the little brassy hit when you pick something up sounds too, uh, musical. Really stands out as something strange when it's so quiet. That's all besides the music just not being to my taste.
The trailer's long with not much really going on, and I've lost interest by the time you get to listing off all the features to be had. To the point where I'd consider cutting it down to just 30 seconds and having separate gameplay footage if you're so inclined. If you don't have anything else suitable to show, mockups of the editor, some short spedup video on working on design or something, then it feels like endless filler.
Title in the first few seconds, smaller cuts of the various gameplay elements. 200 levels? Sounds like a big number. Good. Show a bunch real quick-like. 4 worlds sounds too specific and unimpressive. "Different worlds" and either rapidly or simultaneously showing some different aesthetics would likely work better. Maybe something goofy and/or memorable happening near the end, like doing one of those spinning grapple jumps and just failing to make the distance, going to a zoomed out view, blur and fade out to title and contacts.

Do you have any plans for social media? You'll probably want to have at least a devblog or shitter to stir up hype and keep people engaged. Engaged people bring your stuff to other people's attention and do your shilling for you. Ideally, you'd have been drumming that stuff up for quite some time before actually making the announcement that you're going to into Greenlight, so you actually have an audience beforehand.

As for the website, it's serviceable. The smaller text gets a little muddy with the dropshadow and some of the lines in the background, and horizontally, it feels only implicitly structured. Kind of begging for some indication of where the column of content begins and ends, or at least something that doesn't always start on the left, like a centered heading, or some remarks aligned to the right. And while I don't appreciate the general style of page, the colors matching the game is nice.
The text is the worst offender there.

Here's to hoping you didn't shoot yourself in the foot.

It's pretty much that, except you're forced to choose between either consistency (delay), or availability (inaccuracy), as partition tolerance is a requirement, especially with all the people playing over shoddy wireless.

Sounds like you don't need to worry about netcode at all. :^)

Thanks a lot for the feedback, all of it is very valuable. The trailer for the final game will be different and I'll most likely follow your advice and make it shorter.
I do have a blog (kircode.com/) and a shitter where I've been posting content regularly.
I will consider your suggestions for the website and try to tweak it. Thanks again.

OBS is such fucking trash, all these features and the only way to record a part of the screen is the godawful 'crop' filter

Well, that's as much as I can do for now. Time for some soul-crushing wage slaving.

Whoops, better slow that down. Not sure how I'll make it look like not-shit.

You mean visually or what? Some simple bitwise tilemapping for the water would help.

OBS is pretty much the best free option. Its simple and flexible enough. The reason why its set up that way is because its intended for steaming so you slap a facecam on your game stream

Your water looks good. Does it randomly go left /right when it hits something? How does it know how far to fill up; does it use a parent/root system or does every block track the source's height? Or does that only apply when a source is pouring water?

Yeah, that's a good idea.

I don't really have a game-window, it's more like trying to capture a small subset of my entire unity project.

For now, it just heavily favors left. I'll have to fix that later. It only fills upwards when its original source (essentially a linked list) is higher than it, but I'm trying to figure out how the heck to add pressure to it.

are u me?

DF works by teleporting a water tile to the end, in the case of full water pressure

hey, me, what's the point of living?

VIDEOGAMES fam

To crush vidya enemies, see them coded before you, and to hear the lamentations of filthy casuals.

I recommend you read this post about indie marketing by Adrian Chmielarz:
theastronauts.com/2013/12/one-lethal-mistake-indie-marketing/

You have the same problem he describes. You focus to much on dry facts that are meaningless without context.

You tell that there is a story, but you don't give me any idea what sort of story this is. Is it a story about Robot fugitive/hero/explorer? Is the story funny/dark/mysterious? If the story is worth learning then give a summary because if hits the right spot it may change someones opinion.
I'm shit at this but for example: "Robo-villagers found mysterious derelict robot made of mix of carbon and calcium while digging. Guide Speebot on his journey to find a constructor that will be able make the poor guy operational again!"

You list possible activities but maybe try to present their meaning and what experience they are supposed to give. Is the game challenging or laid off? Will I use my brain or agility? Or mix of both? You probably made a game that you yourself would like to play, so why exactly do you want to play this game? What is fun about it?

Also your game reminds me of some 3D platformers I used to play as a kid. If you were inspired by something you should mention it for nostalgia points.

Nice, saved. Will try to think in this direction and improve the descriptions. Very helpful, thanks!

Wait, shit, I didn't do half of it and I don't even get how this is working right now

Did you mean Love2D? That's more of a framework for 2D games scripted in Lua than a game engine. Godot's main draw is the Node/Scene system. It doesn't have an equivalent to UE4's blueprints, everything must be coded in GDScript (3.0 will add C# and GDNative). There's a comparison between unity and godot in the documentation if it helps.
godot.readthedocs.io/en/stable/learning/editor/unity_to_godot.html

Best of luck, SpeebotAnon!

You did one thing that no other sand simulator has done so far

waifu is cured from leukemia
all I have to do is to finish the ponytail

why does she have tumors growing on her eyes

it's how animu eyes are made in 3D
there's some examples on sketchfab or MMD

Don't most anime style models leave the facial details up to the texturing?

I think it depends on the game.
the one I'm planning to make is sort of like a VN or simulator, where the models are just there for show and the player doesn't control them fully.
so the models are focused on animation.

of course other games would probably use textures for limitation.

but I'm not an expert, I haven't even touched any engines yet.

You'd be better off rendering the pupils as textures that move around the socket

also because non-textures look funky as all hell
but good luck

Add more detail to the enviroment. It seems really empty.

Also, add lot of customizability, like manipulating the volume of the sounds and music separately, the velocity at wich the text apears, etc.

Add a speedrun mode if there isn't.

Add a leaderboard for it. It will make it live longer.

also, the sequel needs to be two players, with lots of traps, to cash in on that youtube LPer cancer :^)

Also, Speebot is not fuckable, androgynous and progressive enough. It needs to be a blackbot with an effeminate robodong or two.
Also, it's too hard, maybe get rid of the water.
Also, does it have to be a robot? Machinery and sharp angles are too masculine.

Promote yourself with youtubers.

obs is fine just reduce the bitrate to a 500 or so value for 720p

>>>/clang/ and would strongly disagree with you.

You need to have the player's shadow immediately beneath the player so that they have a better idea of their position in relation to the platforms.
Look at Mario Sunshine, the player's shadow is always right under the player.

so if i wanted to make a city builder strategy game like stronghold how would i generate the terrain? basicaly just squares with elevation where squares of similar heights would "tilt" to mark them as passable? like making a grid with random heights and making squares where the points would be an average of the neighboring squares but if the difference is to high then generating a cliff face.

I've been following this game since it's conception, and I gotta say man, congrats on prepping for Greenlight! I hope it turns out well!

better example yet is 3D world

youtu.be/AxI_aQCRO2w?t=2m20s

The player's shadow is independent from the rest of the shadows and despite being dinamic and not just a black circle it's always pointed under the player.

Thanks!

More detail will be added later. Sound channels can already be separately adjusted, and there's a lot of other options in the settings. There will be also be some sort of statistics screen, not sure about a leaderboard, but I'll consider it.

I should finish this one first!

Will try!

Thinking about it.

Thank you, I hope so too!

Your right that is a better example.

Is this video games yet?

Pretty sure my water system isn't going to work with all the features I need. Which is a damn shame. Any other algorithms I could try? Rather than my hand-made bullshit I hacked together.

I want each square to have a deterministic pressure, which I can then use to decide if blocks should move up. Pumps can go over a square of water, too, decreasing pressure on one end and increasing it on the other.

How much will it cost? Also how do I vote for it on Steam Greenlight? It looks pretty cool.

yes

What's the timestamp good for? Throwing away older messages?

I found this after a bit of searching.

please use archive.is/view/feature/131954/interview_the_making_of_dwarf_.php?page=9

politics was a mistake

Thanks! I'm thinking around 20 bucks…
To vote for the game on Greenlight you're supposed to sign into your Steam account and press "Yes" on Speebot's Greenlight page. Just to be clear: this is free and does not require you to buy the game or anything, it just tells Valve that there's someone who thinks this game should be on Steam.

That's 5 cents more than Braid at release!

I think this sort of thing would usually be less

needs a few action packed levels
rising water or projectiles fired at you, idk

Piecing the timeline back together on the server.

The player can proceed with actions and as long as there are no conflicts then it will appear as if the player's actions are occurring in real time, but actually the server gamestate is behind that of the client. It removes the lag from gameplay, at the expense of potential divergences which need to be straightened out later.

Timestamps can also determine latency.

My netcode averages player latency then computes the size/duration of an accumulator buffer for inputs, running inputs through the physics / logic ticks only after a delay to allow other clients to respond with inputs.

I don't use timestamps though, I use tick counts.

The last recieved tick count is recorded to throw away duplicates. And for greater reliability I pad smaller packets (

You can have the same effect with bumpmap textures.

It creates the illusion of uneven surfaces, without modifying the geometry.

For the people here to use Unreal Engine, I found a series of videos that go over the optimization tools available and how to best use them to find problems and optimize them away. A lot of useful information and you'll come out of it understanding the ins and outs of the rendering side of Unreal Engine better.

youtube.com/watch?v=H9Yb8Y2-Kng&index=1&list=PLF8ktr3i-U4A7vuQ6TXPr3f-bhmy6xM3S&user=UCs2RyUGngZIZYdZaQ5p2mGg

Your game reminds me a TON of Frogger: He's back!

It only clicked now though.

Is it more important to have results and that I am overthinking how proper coding should be?
Or is proper coding very important?

The most important thing is to have something functioning and visual. Then you can optimise/tidy up.

Do it in cycles for every feature.

Updated based on inflation would be never buy a game for more than 36 dollars and 52 cents.

Sounds great but what about cheese pizza?

Gave you a vote, good luck man. You might want to link the demo in there too.

Well that was entertaining for all of about 20 seconds.

Speedbotanon, did you make a new account just for posting your game on steam or do you use your own gaming account? I ask because I don't know which one is more convenient.

I'm not sure of the best way to do that. Like I'll have both floor tiles and moving objects, so I can't just use a simple array, because there might be more than one thing in a cell, or none at all.

For security reasons you should never use your personal account for anything business related

Is there a limit to the number of things that can be placed on a cell? If so, perhaps the simplest way to do it would be to use multiple arrays. So in each update cycle, you'd go through the numbers, but on each one you'd check multiple arrays. GridA[1], GridB[1], GridC[1], etc.

Obviously that's impractical if you're planning on having dozens or hundreds of objects occupying a single cell, but if it's only two or three this might be the simplest solution. (Unless the programming language you are using allows for three dimensional arrays.)

But don't you get v& if you have multiple steam accounts?

I think at most there would be two objects (eg water tile and the player for example). I want to cap level size at 64x64 I think, though 256x256 (a short) would be the hard engine cap. I just dont want to have to create and destroy a million objects every frame to track animation or update states

game developers get special treatment

How the fuck do I do AI in godot? I am retarded

Professional opinion here, go lower. Like, WAY lower. The game may be WORTH $20, but you have to find a pricepoint where the customer sees the game as WORTH taking a risk on and buying. If you throw it out there for $10-$12, you're going to get a lot more people seeing it as a bargain and picking it up on a whim than you would at $20.

The ultimate goal when setting a pricepoint is to strike a balance between feeling like your employees (you, in this case) and your customers are both getting a good deal. It's always gonna be leveraged towards the customer in any reasonable situation, but you'll be surprised at the income you see at that pricepoint compared to the $20 one you're suggesting.

Rate my title screen
Laetus is the world's name, all my games will be in this world

The feature demo is Everyday Lite
The full game is Everyday Life

I support 's notion. It's easier to make one dollar each from a million people than a million dollars off a single person, and I know quite a few people who would buy Speebot if it was around the $10-12 price range, even more if you include poorfags who only buy things on sale including me.

Dont lowball it either. If its $5 people will think its shovelware because its priced like it. Gotta find the right balance

i need to find a new artist for these later, this is out of my current one's speciality

The fairy one kicks fucking ass, what's going on in the blue blob one though?

Its the INN where the main character will live until you select a spot to build a farm

What the hell is blackshellmedia anyway? Last year I kept seeing them reblog a bunch of posts about a shitty game I tried to make, even though those posts were from all the way back in 2013. No clue how they got ahold of them unless they're a bot scouring game dev tags.

They're scammers, apparently:

Doe anybody have a copy of the last /agdg/ thread saved?

archive.8ch.net/posts.html#!v/12524303_#12524303

Have you considered reducing it to the clock tower on the inn? It might be more recognizable and give you room to add some flair to the clockface

Holy crap, the archive actually works now?

I'm trying to look up other threads now but apparently they don't work.

Sometimes you need to reload the page a few times to get the content to show up, but yeah. It's been back for at least two, three months.

better than mine

Archive works, search doesn't.

still got that 9000 polygon sign ?

I hope there aren't many more places the game, otherwise your game won't be finished by 2025.

If it's good it's worth archiving, right?

I read it as Lactus at first glance.

The green-blue gradient on gold shadow looks grody.

The Fables text could stand to be a lot closer to the big text.

...

I'm getting tired of making the same shit again and again so how do you organize your blender projects?
is there a way to save selected objects as their own file? or do i have make a library file then crawl through each file?

i might have answered my own question

I don't normally post in these threads but i do computer drafting and I save various items I use a lot as their own files in their own folder. If the program supports it, I save it to the in program object library and copy from that.

There are also useful addons like:
blenderaddonlist.blogspot.de/2015/05/addon-asset-flinger.html

I keep individual objects each in their own folder, with a thumbnail set for the folder icon. The + key when saving an object advances the file number, so I keep versions that way and then do a zip archive every now and then and burn to disk.

To get objects into my scene I just have two files open, select the object and ctrl+C then ctrl+V into the other file.

That or go to "file", and hit "append", then go into "object" (shift+F1)

That's fine for simple models, but you'll mis out on blender specific info, potentially rigging and what not.

You can also import lamps and materials this way. I have a folder with .blend files just for materials and lighting arrangements.

I'm pretty sure you can export everything needed with the correct export file formats.
Otherwise try this:
blenderaddonlist.blogspot.de/2015/01/addon-blend-library.html
With this one you should be able to directly import Blender objects and groups.
There is an Addon for nearly everything if it's not already in Blender.

Forgot to mention, you should read up about managing data blocks and linked libraries.
docs.blender.org/manual/en/dev/data_system/linked_libraries.html

These are handy because you can edit things externally, individually, then link them into the scene.

I have a custom python kludge script to export data from the main .blend into my engine's native format.

In some ways this sucks because each editor I use (not always blender) needs its own export plugin, but in other ways its good because the storage can be more efficient and more closely match the data structures used in game. The external editor format acts to normalize assets and the plugins modify them for the different engine / game versions.

The clock tower is pretty small, can't barely fit a bed in there

lowered it to 1000 triangles :^)

This will be the largest building, i also did it while relearning how to 3D model

I will eventually get an artist to make a few adjustments

I need to check what type of Ground the player is standing on to see what action he can perform on it, how should I do this?

I currently have a raycast going down, but as far as I'm aware I can only return the name of the 3D object with RaycastHit.Collider.

Should I call all the grounds I want a particular action to be performed on the same?

Don't listen to that fag, the gradient looks awesome

I think he has a point, it could look better than that…
I might have to play with other colors later, i will keep the gradient though
But right now i have to start implementing events and shit and finish the demo already, once the faerie system and the cooking system are fully integrated i will start releasing it here, so you guys can give that delicious harsh merciless feedback

Here's my first functional enemy for the game. I call them "Nutcrackers," because that sounds better than "Ball kickers". Pretty simple pattern.

I'll show off the boss fight in the next thread.

You'll have to excuse the 20 FPS. My computer is a piece of shit and when I'm not recording, it's an average of 45. Still not very good, but I'll worry about optimization later.

This stupid fucking website didn't upload my mp4 because the feature got removed or whatever, and then posts this anyway. Fuck you Jim.

How the fuck am I supposed to upload my video, now?

Maybe this time it will work.

You could try converting your video to a webm.
Or post better.

I don't know how. I got it to work this time, but I had to compress the video to hell and back because of my slow ass internet

You control the ball?
What will your game be about?

The ball is supposed to represent the toxic masculinity which encapsulates "gamer culture", and the red color is supposed to represent the Republican/Alt-right takeover of the mainstream internet. Over time, the ball slowly comes to terms with how toxic he is to society and changes his color and shape. It's gonna be one of those "the bad guys are right" games, and I'm really proud to be part of the progressive movement to silence these nazis once and for all.

Nah, just kidding. Just gonna be a fun action/platformer-without-jumping hybrid. Literally no story or sense behind most of it

So you roll on ramps to get over obstacles or something like that?

Sorta. I'll have to pump out more footage for people to understand. The AI was pretty pathetic in that video because it's out in open plains. They'll usually be found in close quarters. Once I actually finish a functional level I'll show it off.

nah, I like the first idea better, do that instead

Depends on what your "Ground" is, and also how you define this.

You can do so much more.
You could raycast the collider, get the attached GO (via getting the collider, then collider.gameObject), and then do anything you want with the scripts (components) attached to the GO via a simple get component call.

I don't know what u mean by this, as this is a horribly broken sentence, but if you want to be able to have some sort of functionality determined by the "Ground", which is located via a raycast; I'd do the following…

Get the attached GO, and grab the "ground" component (a script w/all your settings attached to the GO), and then call a method from there.

Or, you could define each "ground" part with a physics material, and then just grab the defined physics material; then perform your "ground" specific actions based on this distinction (i.e. each ground has a different physics material).

Imo, the 1st option is better, as it's sticking to the idea of splitting functionality into components, and you can manage all of the actual update ticks, and logic calls inside of a "Ground_manager" (like your "system" for the "Ground" component functionality); all you'd need to do is awake the component from sleep (due to that raycast), and the manager/system would perform any logic that's required (what that is, I have no idea).

So everything we post on 8ch gets saved forever and ever and all eternity?

no, only the embarassing posts that you want everyone to forget

Hey, anons. Remember that really embarrassing post I made in the Fallout thread last week? Oh man, I don't know what I was thinking. Glad we're past that. BTW, where is Kyle? I haven't seen him in a while.

Can I get an Array of all the colliders that have a specific tag in the current loaded level by using a script in Start()?

I can't really take just the GO because I then have no way of identifying it, I think getting colliders is the way to go.

Then I can just check if the thing the player is walking on matches with any of the entries in the array of colliders I have.

Suddenly I'm questioning even posting on this fucking website anymore, man. Fucking Christ, being able to move on from mistakes and arguments is the reason I use this place for my share of human contact, of sorts. With that gone, I might as well just never speak again.

I doubt anyone will care enough to trace anything back to me, honestly, but it still is a shitty realization.

What the fuck are you talking about

Too bad there aren't many people who sympathize with this anymore. People used to pride the fact about imageboards that everything is temporary, you "had to be there" when something happened, but now whenever it's brought up people start defending total archival of everything in imageboards.

The temporary nature and anonymity is exactly what drew me to image boards. Both that and how everything is sorted by time rather than votes or some other algorithm.

yeah sure that could work, but it's not how I'd do it personally.

I fear this too.

No presence or reputation to uphold. If you are not a regular on a forum people will just ignore you in favor of known users. On image boards you just join in whenever you feel like it.

Move the "Fables of Laetus" down and add some sparkles so it looks like the fairy conjured it from her fingertip

It's disheartening to see what sites like Know Your Maymay are and have been doing to the internet.
To see any unit of culture, fluid, organic, context-sensitive, tempered through reuse and reiteration, deeply intertwingled with other similar items and sometimes existing in a few different dialects reduced to one incomplete, incorrect conglomerate of an interpretation the length of a paragraph or two, set in stone forever. Then fucksticks not in the know take that paragraph as the complete and impartial guide to the whole phenomenon and base all future iterations on that. This is worse than just making new garbage. This is burying the old, taking its name and slapping it on something new, effectively memory holing everything good about it.
You try to bring up some of the history involved? You're just making shit up or are talking about people doing it wrong, because the irrefutable authority of the archive doesn't say anything about those details, regardless of how recently said archive entry was made.
Same thing with regular post archives. People see one sliver of one person's participation in something, and come off extrapolating that as the average or normal for said thing.

Then you get to the posts and the generally ephemeral nature of human experience as a whole. There's something beautiful about a spark that comes and goes, the only evidence of it having existed is in those who were there to witness it. Now those affected could drum up new iterations and retellings, mix and match, recycling and adding value organically. And sometimes you end up meeting those people in new threads. Despite not having names, the current setting having nothing to do with where you met before, something about that original spark shows itself, and you know for a fact that you had met before.
This can't exist in a world of cookie-cutter threads and verbatim retelling. You weren't in the XYZ thread, you were in a XYZ thread.
That is not too much different than what goes on in social media, where everything good has to be completely stripped from the people and intimacy involved, and everything bad has to be permanently glued to anyone remotely connected to it, given a single unified and crystallized existence, to be shared like wildfire and memorialized forever.
Anonymous cool things happening somewhere! Whoa!
Horrible man (John Smith, see picture, contacts and social security number) did something stupid! Contact your local news outlets.
What good is an experience if yours is no different than anyone else's, unaffected by what makes you, well, (You).

So ends my inarticulate rant. Fuck.

Here's a meme UUUU

Christianity was the know your meme of religion. Seeing how that ended up you may as well surf on that wave or die like the old religions did.

Isn't Jim ex-FBI or something? Either way he's a kike so this should seriously come off as no surprise. You think datamining threads on Holla Forums and Holla Forums are bad? This whole fucking place is a datamine.

In fact, I'd argue the risk of anonymity loss is worse than it is on a place like Reddit, because you don't post on Reddit under the pretense that you're untraceable. Now we have an archive of
EVERY
SINGLE
post ever made, and autists, spooks, and autistic spooks can use that information to trace your identity across the entire SITE because of some semi-personal info you may have dropped in another thread, typing patterns or reaction images that only you use/have a unique title for.

Our anonymity is gone now. We're living in an illusion, and the competition of reality is fought by delusion.

It's an arms race. There's no going back to the time of innocence.
In the halls of quiet conversation, when you've got someone yelling and blasting airhorns to drown out everyone else, you can either restrict entry or start yelling and using megaphones, and neither is favorable.

We are living in the September that never ends.

(checked)
Beautiful get. So where I am supposed to shitpost then if not for this place n-no where at all?

Is there anything wrong with using the fixed function opengl to learn and quickly get going? Everyone keeps telling me I am an idiot because I need to learn OpenGL 4+ because current year.

...

There are tales of a BBS that is still alive.

You can use whatever you want to use. You're probably going to want to understand both eventually, but if you learn the new API theres not much reason to learn the old one except for legacy code. Personally I find shaders to be radically more interesting than the fixed function way.

No one cares enough to cross reference shitty anime faces in an imageboard archive, there's no profit. It will probably die a quiet death like most do anyway

Yeah, not until I get famous. Then you know everyone will be trying to dig shit up.

Not until the moment someone has a bone to pick with you. If my memory wasn't terrible, I could probably get a few examples in. But it's generally been either obnoxious furfags, tripfags or outsiders persecuting themselves, the occasional gamedev working on some robot game in a friday night thread looking for /cuteboys/, or just shills. Outsiders and shills tend to trip up on IDs and not really require any reconnaissance, though, so that's always a blast.

Is this subscription based only?

sure buddy

See, even your ID is screaming in fear at the prospect!

...

I mean for learning I get it, but isn't it a bad idea to do sell a game made with pirate software? Doesn't the company contact you asking for license proofs when the game is out?

no

1. nobody cares.
2. piracy is legal in my shithole.
3. unless you're american nobody cares.
4. Don't be a retard and scream that you pirated.
5. Make enough money and then buy the software.

In any case, it will cost more money the company to try to sue you overseas and hire a lawyer than what they will make from you.

unless you're american, in this case, move to a real country.

nobody cares period.

So what you're saying is, I should just shut the fuck up and liek maek gaem?

yes

pretty much, companies don't make a lot of money selling to random people.

All the posts are public, anyways. If not for an archive, whose to say there aren't several hundred scrapers reading the site at any one moment?

Why would you post something online that you don't expect everyone to read? Even if its private, assume a server or account will get hacked and have its contents accessed.

Why do you think whoever posted the TOR announcement on the front page of every board?

I've been thinking more about what my water physics really needs. If I could just somehow calculate the pressure being exerted on each 'tile' of water, in each body, everything else would come naturally.

My spagehtti-string method simulating pressure works, but it won't do much for, say, pressure pumps or shooting out of the side of a wall with a velocity based on pressure.

How active is this thread? I have an idea for a Fantasy Business simulator and I want some feedback.

just ask you cunt

this is something more people should get, computers are devices that let you interact with the outside world, nothing that you do in one should be assumed to remain private

Just remember, everything you post gets saved to a database for everyone on the internet to see
FOREVER
so post wisely.

that's not the important question.

the more important is:
can you make it?

There will be some people who will find your shit cool, hell, I saw once a faggot who think kitchen tiles are sexy or some shit like that.

I hate that I recognize where this is from

Fine
So the game would be about buying up businesses in a fantasy world and running them. It has some RPG aspects in that you have skills so you can run a shop alone during the hobo phase and so good/shit employees can exist in the latter phase. It's basically Fantasy X3.
I have a good idea about how the Alchemy system would work and how the economy would work. I have a vague idea how cooking would work but for smithing, tailoring and such I have no clue yet.
The alchemy system works a bit like both the TES alchemy system and the Witcher one. You have ingredients with various elements within them. These elements activate either automatically, when combined with other elements or under certain temperatures/liquids. You create a recipe and which is a set way of making a potion, this allows you to teach it to employees so they can make the potion for you. You can also hire adventurers to find you ingredients.
Cooking is similar, every food ingredient has certain properties which describe flavour and texture like Salty or crunchy. Every customer would have a preference and making a meal which matches it more makes them tip you more and come back more often.
As for the economy, people will have needs and they will fulfill them, like the Oblivion AI but less retarded. People will try to buy food when hungry, will look for potions of healing when sick and so on. A particularly unscrupulous merchant could infect a city with plague and sell it the cure for profits. There would also be a way to set up contracts with other businesses or ruling governments and agree to supply them regularly with something. The tradeoff being that if you fail to, the deal is off and you get a bad reputation.
The major problem with this idea is that it could be overcomplicated and make it either impossible to program or to play. Especially considering how many stats the computer would have to keep track of, just making the potions system alone might take up all of the RAM.
Any comments?

can't give opinions until you implement a prototype.

some ideas sound fine when you're ideaguying, but they suck in a real game.

download game maker and make a prototype.

This game maker?
yoyogames.com/gamemaker

How the fuck are some potion recipes going to take up 'all of the RAM'? It's 2006, every computer practically comes pre-installed with 2 GB

Sounds a bit like the guild.

Customer preferences, needs, income. Don't forget it's also simulating an economy in the background, like rivals who are doing similar things.

Never tried it before, which do you reccomend?

What do those have to do with potions, though? You said potions alone might eat up all the ram.

optimize it retard.

Your ideas seem a bit disorganized, but nothing really stands out to me as terribly bad, it's all in the implementation. I agree with . Try to condense the game into the core gameplay loop then build that to see if it's any good. Don't worry about building assets.

The 2nd one is alright, I don't have much experience with it to be honest, it's fairly fun but it's definitely clunky in areas and a little buggy, I think if you grab it with all the expansion packs either pirating or whatever method you prefer, would be the best.

You're not going to run out of RAM. You should try programming before you plan to make "Oblivion AI but less retarded"

An exaggeration, I was saying that one element could be able to use up ram because the idea I had was a bit ambitious.

n-no bully

Sure, but I don't know how to. I am literally just an idea guy right now. Which program should I be using? I do have some experience programming in C though, and some basic programming knowledge

download the following:
A book on data structures (any language)
A book on algorithms (any language)

You could probably do pretty well with Godot if your doing a 2D game.

Of course it would be in 2D, it's a business simulator not an action game.

hey now, we're the only ones who can be cunts

I'm still trying to figure out the best way to structure my level grid.

Like say if the player is adjacent to a monster, you explode and have to start over. But while you're moving between cells, you'd technically occupy both cells at the same time, but wouldn't react to monsters until the movement finishes. So there's like some sort of weird "locking" I need to do to prevent two objects from trying to enter the same cell (though I know that I want the top left object to have higher priority, as a game rule)

dude, like just make a test scene and test if the idea is better/good.

lol

user… you've been in a coma for 11 years

Im now implementing active effects of

...

You make us proud Speedbotdev-kun

oh, god

I can't wait to play all the new sonic games

Go back to the hospital and sleep for 11 more years user

explain this maymay.

...

There is not time to explain, JUST RUN

be a kyle, occasionally drink and shitpost as i try and make me game werk god i hope i don't get tracked down. but as is i'm about to snap at me family if they start talking politics again, dealing with bullshit? take a drink.

progressu.

I seriously hope the face is ironic

needs textures and hair.

So is windows 10 a pain in the ass to pirate in?

is windows 8.1

dunno, is windows, disable windows defender and antivirus when using the crack.

No

Here's the model for the UMP 45 if anyone has any criticisms/critiques/suggestions. I still have to model the picatinny rails but I wont really need feedback on that.

Kinda feeling burnt out as well. I'll be glad once this model is done and I can start working on code again. The GAU-8 will have to wait.

Anyone know if there's a way to specify an int type in lua? Like unsigned short (16-bit unsigned)

Not in vanilla Lua. LuaJIT has the FFI which lets you declare variables mostly like in C, and easily call C functions.

This is probably a stupid question, but how do you keep a camera from going outside set boundaries in Godot? I have absolutely no fucking idea what I'm doing in this engine.

Shit.

...

if (player.x + speed < screen.width - camera.width / 2)
{
Do code.
}

...

I don't know godot, but i would do something like this:
If (playerposition.xleftlimit){
cameraposition.x = playerposition.x;
}
If (playerposition.ydownlimit) {
cameraposition.y = playerposition.y;
}
There is probably a better way to do it, but this way the camera would only follow the playerposition if it's within your set camera limits, otherwise it sits still.
Also research if Godot already has inbuilt camera/position clamping or limiting.

Why do the vertices on those edges seem hellbent upon merging as I scale? Can I fix this problem, or am I just going to have to manually adjust the vertices with each extrude? What can I do to prevent this in the future?

(This is a mirror modifier with clipping enabled on Blender 2.78, for what it's worth)

...

I was more talking about the edge flow, removing clipping just gives me a bigger, worse problem.

...

Okay, I'll try it and re-loop cut if it gets to be too obnoxious, thanks.

Okay, I tried and I think I'm just going about it all wrong. I should probably be asking how to set boundaries in a world, since I don't think simple CollisionShapes are gonna do it.

Is the camera limit option built in, or is it an export to which you have to write code for?

You could also be a mega jew and just put invisible walls and scenery out to the maximum camera bounds.

It's built in, but it looks like it's hardwired to the world coordinates.

Didn't really wanna go this route, since level transitions would probably look really bad like that.

…as in, it's based on the coordinates of the scene that it's imported into? You could probably write a script to correct for that. It'd be pretty damn tedious, but it'd be a solution.

Looks like it to me. The left image in this post stops at 0,0 of that particular scene.

Y'know what, mega jew it is. I thought the platformer example that comes with the Steam version of Godot did exactly what I wanted my game to do, only to find out that it just had really long walls on the right side of the screen. It's probably fucked.

honestly if ever single moderator of every single forum dropped dead no one would care

So 0,0 is the corner of the grey nothingness then? In that case, just attach this pseudocode to your level loading script:

# when level is loadedlevel_spawn_coords = get_transform(level)camera.limit.left += level_spawn_coords.xcamera.limit.right += level_spawn_coords.xcamera.limit.top += level_spawn_coords.ycamera.limit.bottom += level_spawn_coords.y

You brought this upon yourself

I'm getting desperate; I made an account just to post on their board. Who are these people and why do they hate life so much?

Well progress is progress, I guess. Maybe tomorrow I'll actually implement what I learned, and make a plane that can fly semi-realistically instead of the magical ufo it is now.

Why not just make a magical UFO game?

...

Pic not related.

Mine was
please tell me you got the joke, user

I think so.

That, and the ayy lmao ending in the same game.
Dogs and aliens in one place always remind me of SH.

Please try programming literally anything before you start telling people that computers don't have enough ram for your game idea.

You sound like you have no idea how this game would work in practice, just a generalized vision of something inconceivable with a hodgepodge of random game types jammed into a single porridge. Textbook ideaguy shit such as "food has stats like salty/crunchy" mixed with out of control scope such as economy simulation or something.

You don't idea guy hard enough, buddy.

Or, you know, you all could just download more ram.

Looks like I'll have to, give me a second I need to google this

Realistically simulating fluid physics is just not really a thing that is possible right now

Except simple simulations like gravity or other directional forces, simulations within realtime applications are extremely restricted, especially fluids and hair.
You'll lose half your fps for something that the player will barely notice or give a shit about, and the simulation won't be very good looking either.
Even if you somehow best every single current realtime simulation magician and manage to create a good simulation, that time could have been used to make a game.
Good luck though.

REMINDER: CGPEERS REGISTRATION IS CURRENTLY OPEN

nigger

What do they offer that others don't?

literally millions of seeded torrents that are not on public trackers.

Don't sweat the haters. Your questions are moved because the people there don't really know shit. Go read some SIGGRAPH water sim papers if you want hardcore realtime sim info – but those are sometimes too academic for agdg gamedev.

I used an atomic approach to fluid and pressure simulation.

The cells of the grid had stats such as temperature, pressure, average direction etc. and had a pointer to a linked list holding particles of matter.

I had state changes of matter where N particles of E element at P pressure and T temperature would transition to a single particle at a lower phase, but having greater individual density. This allowed liquids to condense out of gas and become ice, and ices could even sublimate.

Each tick the temperature would be transfered into adjacent buckets of less energy, the bucket's velocity and remaining energy would be attributed to the atoms, the phase changes would be processed and the atoms would be processed, possibly turning temperature into pressure (more particles + velocity) and be summed into the next state's buckets. Collisions added the reflected velocity vectors proportionate to the density of the bucket collided against. This allows water to move against wind, and atoms to pass by each other due to density affecting partial reflections. The trick is not modifying the current state's grid/buckets and propagating changes into the next state instead, so you have an unmodified reference to look back at as you process cells. This is a staple of functional programming, but such immutable states can be beneficial in other paradigms too.

In the sim with only two elements, water and nitrogen (having lower boiling / freezing point), I could add / remove heat from the simulation in spots and weather would emerge. I would see rain and even snow as the colder nitrogen would sometimes freeze the rain before it touched ground.

The physics sim did not directly affect visual sprites or particle fields – visuals were generated by a pass over the buckets using a modified 3D version of but only for visible elements (air wasn't rendered, but I did render temperature differential to a shimmer layer)

That 3D project is top secret, under wraps.

However, I initially began prototyping thermal mechanics in JS in browser on a slow ass HTML5 canvas, and I can share a demonstration of that:
vortexcortex.com/game/vibes-01/
This rough prototype demonstrates thermal transfer and phase shift, but has a very crude physics and pressure system, not condusive to water flow (though liquids do flow over time). Only two elements are present, nitrogen and water.

Press the numbers 1-4 to change views of the sim – I forget what 2 & 3 do, something about thermal and pressure transfer, 4 is supposed to be a scaled view. Left click and drag adds temperature to cells, right click removes temperature. Even with this very simplified 2D sim you can see weather happening. Just freeze one corner pretty good and watch as the temperature gradient causes rain and snow flurries at the atmospheric front. A constant gravity drives the atmospheric layer.

Anyhow, my point is: Fuck the haters and physfags. They can't help you because you're going beyond them, trying to actually make something rather than just be correct about some bullshit. Just code shit how you imagine something might work, refine/scrap and repeat, and ask your fellow /adgd/ devs if you get stuck.

found the mod

Thanks a ton, this is the best piece of advice I've gotten.

I was thinking about that earlier, with potentially just a velocity and pressure, but couldn't figure out a way to make it work. Atoms are interesting. I need to look into it some more after I finish up today.

Okay, I guess.

Worked for me :^)

sorry for the off-topic, but I figure with the thread headed to 404 it's no big deal- couple threads ago, an user started developing a board-tan for /agdg/, and said he'd come back with the design refined and using the traditional /agdg/ colours. When that happens, would any of you anons be so kind as to drop the image, a link, something, over at >>>/btan/ ?

Wew, glad to see people jumping on the CG train, just remember not to let your tutorials sit on the desktop collecting digital dust.

Here's an archive of the last thread.

I don't think he ever posted Aggy in /agdg/ colors, though.

...

...

if owo.WhatsThis(0,0) == true: print("Unzips dick")

Is there a new tread?

Not until page 13

Why? Is page 13 cursed?

8ch.net/v/rules.html
Rule number 6

comfy.

At the speed of posts right now, it'll probably be about … 4 hours or so? Maybe more

Sounds like a meme game.

It feels like Holla Forums is a lot slower than it used to be.

Alright, I've got a better idea on how to handle things for my puzzle game.

While any object to the top/left will have priority in the case of a conflict (eg two objects trying to enter the same cell), the bottom/right object will be able to override behavior as needed.

Basically, a rock would be a sort of basic object that either falls or rolls left or right. A balloon would push a rock upwards, or fall downwards if it has two rocks on it (and coincidentally, the upper rock would roll off the bottom rock unless there's walls nearby). But, the way it would work is that the rock sets its cell's behavior to push it down; but then the balloon overrides it and pushes it up. That way, the rocks don't have to care about the fact that a balloon exists; more complex objects can kind of do their thing

is an autodesk software for shoemakers, being planned on exporting this baby to autocad then export to 3D coat and retopo.

baby easy 3D.

New bread soon?

Oh yeah, what I was getting at, this system is still flawed.

If there's a "falling object" with the balloon, and the object somehow interacts with something to the left or right of it during its update, that will still happen, and then the balloon will tell the falling object to move up (but not cancel its other interactions)

It's probably because there's nothing to post about and it's monday so half of Holla Forums is probably either tired from work / going to work or tired from college/ going to college.
Everyone is waiting for the cringe compilation from the next E3.

You mean

The extra == true is redundant and wastes a check

if owo.WhatsThis(0,0) == (not false): var dick = "Unzips dick" var final_string = "" for i in dick: final_string.append(i) print(final_string)

my plan worked.

easy as fuck 3D shoes for my characters.

Those look like some really expensive leather shoes.

will make 3D coat to automatic retopo it, then use substance painter to paint it then import on maya and animate before exporting to unity.

literally piss baby work.

I meant that they looked like they would cost a lot of money if they were real.

I dunno, I'm not a shoemaker, lmao.

Hm

what do you mean?

My objects have a "Position" and a "Target" variable. This is so they can smoothly interpolate movement over the next N steps.

In the case of sliding floors, ice, and portals, they don't really have a way to "talk" to the moving objects to tell them where to go

My solution seems to be to also track a "Previous" position on the objects as well. That way, the ice tiles can see for any entity that it has a previous/current position, and infer its direction. Eg, a rock moving from left to right across ice would slide two tiles. If the the right is blocked, then the ice just tells it to bounce back to the left. If the left is somehow is blocked too, then it would tell the rock to fall downwards.

Do it in whatever manages them, or give them the ability to check what tile they're on if you want them to do it themselves

for sliding shit you need to separate your speed from your forces in separate variables, then when you're not moving, your force should move to 0.

I don't have force or acceleration. I don't even have speed. It just … kind of happens.

That's the thing. It's going to be somewhat customizable so I can't really tell the "rock" that "ice" exists or the other way around.

if you want ice ground physics, you need to add basic 2D physics user.

go and google how to do 2D physics.

I think I've got it figured out in the time it took me to make a snack instant noodles

Basically, I have a pPrevious, pCurrent, pNext for an entity instance. My Move() function takes an entity and a speed (N) and over the next N updates, sits it in its current tile and draws it interpolated between pCurrent and pNext. This is what I already have.

Since my maps can be 256x256, having 65000 tiles all constantly updating would be a problem, which is why I wasn't sure how to approach it (and rocks would have no guarantee that ice exists as an object/tile type).

So, since all entities and tiles have behaviors, the solution is simply to have the tiles delegate the behavior, but only if it has an entity on top of it. Or rather, have the entity invoke the tile's behavior when it is finished moving. In the case of ice tiles, it looks at pPrevious and pCurrent to see which Move() it should call, and in the case of water tiles, destroys the entity.

use a message manager retard.

Please, use newtons equations, with velocity, acceleration, and position to move things.

the way you solve the ice problem is simple. When you have velocity, you have friction. Friction is a form of acceleration which attempts to make the velocity go down to zero. When you are accelerating something faster than friction, this will override friction, however when you stop doing that friction will kick in and the object will slow to a halt.

The way ice works is that ice has a lower friction value, which means that friction slows things down much slower, which means that it slides across the ice.

It's a deterministic tile-based puzzle game, physics has no part of it here

...

I wasn't asking for physics, I was asking for the best way to logically store data across multiple frames

Look a new thread!
Now I can hide my shame until I post again

COME ON, STEP IT UP!

class animation
{
List animation;
//Make your animation data here
}

Way 2:

class Animation()
{

struct Frame
{
int frame_index; // this references an index of the Data_Frame list with the data you want to repeat.

// Other frame data value
}

List frames;
List data_frame;
}

Or way 3:

Use a map.

In any case, read on data structures.

holy shit, just how much do you weight?