Amateur Gamedev General ~ /agdg/ + /vm/

but how

You can make them as you serialize
You may even use the object's position in memory

Tree ID: T7 has Apple ID: A2, Apple ID: A6, Apple ID: A9

The apples are deserialized first
Then you deserialize the Tree, when you deserialize the Apples array you get their IDs (strings), then you search the global list of Apples and assign the correct ones to the Tree based on their IDs

please, have the IDs be longs, better in every single way than strings.

I am pretty sure you can count the universe's atoms in a long and still have remaining space

So there's 5.42e+60 times more atoms than a long.

Atoms in the observable universe: 1e80
Max unsigned 64-bit integer: ~1.84e19

you are not running out of it, even if I was wrong about the atom thing

...

sry bro, no idea

besides the fact that it seems you are using a language whose indentation have semantic value, there seems to be no problem with the logic.

well at least it must be nice having a friend IRL that understands this place

You were right about using integer references. I'd say that using size_t or your language's equivalent, if available, is a best practice for that, though.

Or you can just not do it, since there are fucking mountains of serialization/deserialization libraries available for every fucking language. Capn Proto for schema-driven, and msgpack for schemaless (or json if you want/need a text format). It's pointless to roll your own serialization unless it's a learning exercise.

You must never have had to deal with unity's serialization

I have not, but you don't have to roll your own from scratch to not use it. You can still use MessagePack or something. You can still use Protocol Buffers or Capn Proto and copy between their generated classes and your own. I never said that all serialization packages were good.

Unity don't like generics or 'not knowing set in stone the type of the thing you are serializing'
When I need to be able to serialize a subclass instance and revert back, I need to serialize as string the complete class name (in C# would be the Type's AssemblyQualifiedName), and then serialize to byte[] all of it's relevant data.
Unity sees string + byte[], but it is actually serializing a instance of a class that extends 'Processor'
Sure I could use a library to get the 'full class name' and the byte[], but I only need it in a couple of places, and choosing a library for that would take longer than just do it manually.

thanks polygon tits bro,
I was just unsure how to code "best fit",
so I ended up with that code after trial and error, without understanding it all

please note that i've wrote seems right, too tired to actually decipher it properly

If you want to do it by hand, that's fine, but you're setting yourself up for a lot of pain when your objects get significantly more complex, such as nested objects, lists, nullable fields, sum types (tagged unions), and the like. Hand-rolling is easy if you are serializing exactly one object that is always the same size on-disk. If that's your use-case and it's never going to expand, feel free.

And there's no reason Unity ever needs to see your serialization. Just don't use Unity's serialization, and don't directly serialize those types. You can use a proxy object, and set up whatever C# Unity object you have to be constructable from the proxy object, so you can load in your serialized object directly and then throw it right into your class's constructor to build it up. There are a thousand ways you can do it so that Unity doesn't have to even be aware of the serialization directly, as long as you know how to store and restore the important state.

i know this is confusing, but that was my brother, the Grass guy, here his stuff
He is the main programmer, im the ideaguy on steroids

...

We live in the same house, and we usually don't get the same ID, but today we have it

We Land of Lustrous now

korioth and kyron?

Applied modifiers
yes i did apply the subdivision, it is necessary in this case, and i do have a backup from before applying it

No, but nice digits

Real life can go crawl into a hole and die

I guess next is working on the pause menu, inventory, and save system.

Here's the in game final results for those who were interested.

that tabletop simulator?

Yep. Made the character specifically for it.

that's pretty neat, did you do other stuffs for it?

So I'm working on an inventory system in UE4. How the fuck do I destroy an actor and keep all of it's data, indiscriminately in a variable to spawn an actor from it when needed? Is there a catch-all method for this shit, or am I supposed to keep it all in a single struct that keeps all of the relevant data?

I dun get it.

That's generally what you want to do anyway, it makes things a lot easier to reason with and it's faster

nice stuff

I guess. I'm just working alone so I wanted to reduce the amount of headache I'd have to deal with if I wanted to add new items.

Like, I was thinking that there must be a way to store the actor's data as just info and then recreate the actor given a class reference, right? I don't know, I'm fucking retarded. I just wanna get this shit out the door in a couple years, not ten. But I guess if you say a struct is the way to go - it is.

this will likely be your only chance to complain about your waifu's typing and actually get me to listen

make sure you don't have any regrets by the time this thread disappears

Gee user, you really like ghosts, the most boring type of them all

i just think there are too many psychic and ghost types

yeah no fucking shit

too many darks too, but there's no types more fitting

Yeah I've got like 70ish models on the workshop, all of them for 40k.

keep up the good work and try not to get sued

first, checked those triple dubs
second, I'm already wondering how I would be able to do this with godot, any hints?
from what I've understood from the docs you can just serialize any object, it will retain its size when reading

Thanks, I've had these models on for long enough to where I think GW either doesn't care or can't really do anything about it. I'm not selling these off or anything.

this is worth reading if you have little formal training in math or programming. If you can overlook the fact that the author tries to be funny

It can't be worse than that one asshole who tried to define in a Python tutorial, "syntax" as "the extra money you pay for cigarettes"

What the fuck is this bullshit?

I'll let you judge.

Imagine you are a contestant in a TV reality game. You are standing in a
clearing in the jungle. Several other competitors stand beside you. You’re
all very nervous and excited because the winner gets to date Cameron
Diaz… and the losers have to watch. Sweat is dripping from your forehead,
your hands are clammy, and you cast nervous glances at the other competitors.
The bronzed, anvil-chinned TV host steps forward and hands a goldtrimmed
envelope to each competitor. He steps back and orders you all to
rip open your envelopes. The first person to complete the instructions will
be the winner. You frantically tear away at the paper. Inside is a note. It
says:
I’m waiting for you in a secret location. Please hurry, it’s very hot in
here. You can reach the location by following the vectors (–5, 5), (0,
–10), (13, 7), (–4, 3).
Cameron
With a smile on your face you watch the rest of the competitors sprint off
in the direction of the first vector. You do a few calculations on the back of
the envelope and then set off in a completely different direction at a leisurely
stroll. By the time the other competitors reach Cameron’s hideout,
sweating like old cheese and gasping for breath, they can hear your playful
giggles and the splash of cool shower water…

It's a pun. "Syntax" sounds like "sin tax". Cigarettes are heavily taxed, and are a vice, so that extra money on top of the base cost is a "sin tax".

Any anons have advice for coming up with puzzles for am rpg? Ideally they'd be themed for each dungeon but first I'd like to come up with some that aren't just 'murder everything in the room' or 'push blocks'.

Push blocks AND push balls. Balls roll until they hit something. Blocks move 1 tile. Figure out how to get balls onto switches but blocks can also weigh down switches so it seems like there's more than one plausible solution.

Miner Bob: Pickin' up a nugget
Miner Bob: Ah'm leavin' the gold mine with mah pockets full o' sweet gold
Miner Bob: Goin' to the bank. Yes siree
Elsa: Walkin' to the can. Need to powda mah pretty li'l nose
Elsa: Ahhhhhh! Sweet relief!
Elsa: Leavin' the john
Miner Bob: Depositin' gold. Total savings now: 4
Miner Bob: Leavin' the bank
Miner Bob: Walkin' to the gold mine
Elsa: Walkin' to the can. Need to powda mah pretty li'l nose
Elsa: Ahhhhhh! Sweet relief!
Elsa: Leavin' the john
Miner Bob: Pickin' up a nugget
Elsa: Moppin' the floor
Miner Bob: Pickin' up a nugget
Miner Bob: Ah'm leavin' the gold mine with mah pockets full o' sweet gold
Miner Bob: Boy, ah sure is thusty! Walkin' to the saloon
Elsa: Moppin' the floor
Miner Bob: That's mighty fine sippin' liquor
Miner Bob: Leavin' the saloon, feelin' good
Miner Bob: Walkin' to the gold mine
Elsa: Makin' the bed
Miner Bob: Pickin' up a nugget
Miner Bob: Ah'm leavin' the gold mine with mah pockets full o' sweet gold
Miner Bob: Goin' to the bank. Yes siree
68 | Chapter 2
Introducing Elsa
Figure 2.5. Elsa’s state transition diagram. The global state is not shown in the figure
because its logic is effectively implemented in any state and never changed.
Elsa: Walkin' to the can. Need to powda mah pretty li'l nose
Elsa: Ahhhhhh! Sweet relief!
Elsa: Leavin' the john
Miner Bob: Depositin' gold. Total savings now: 5
Miner Bob: Woohoo! Rich enough for now. Back home to mah li'l lady
Miner Bob: Leavin' the bank
Miner Bob: Walkin' home
Elsa: Walkin' to the can. Need to powda mah pretty li'l nose
Elsa: Ahhhhhh! Sweet relief!
Elsa: Leavin' the john
Miner Bob: ZZZZ…

Hmm. Well, it's something. I could probably throw together a puzzle where the player would have to use the blocks to set up the balls to hold switches first.
Anyone know how many kinds of puzzles old snes era rpgs tended to have?

As I posted at the tail end of the last thread, I realized graph theory can be used in a lot of places. Now I just need to figure out how to like, make a nice, flexible graph framework

Not many people know this but all the puzzles for RPGs are literally exactly the same. They're all just poorly made sprite swaps.

Pro Tip:
If you're a beginner and want an intermediate course in c++ try to implement a poorly documented toolkit with an inactive community into your code.
It's like climbing a staircase but I'm 2 inches tall, also gif related.

UGold RPGdev,

Started taking apart old code and putting it back together with more sanity checks and optimizations. 300 lines on one went to like 120 or so, due to a lot of copy-pasted checks that could be done through method implementation to do the same code 6 times. I'm now putting the code for a mutator, a player replication info, and a level up item together to create a functioning leveling system. The stats/leveling system works like this:

In case I need to spawn a new puppet, it should load all of the information that needs to be loaded into the struct array(s). The reason for this is because if I lose a pawn somehow or need to spawn a pawn for the RPG combat mode, I can load these variables from the .ini or struct array to spawn a new puppet. This is also useful for drawing menus to display stats. More has to be done, but that's where I'm at. New HUD screenshots when I get home I think, if it works.

That guy whos making the strategy game with the rain webm,you still around? hows it going?

Also I've been mulling over ideas for how high stats could potentially get. Base of each stat is 15, multiplied by a focus number. Focus starts at 1, focus is modified by subclass selected, so for example a strength/dex subclass plus a dex class, so that would make strength's focus number 3 and dex's focus number 2. Also I only want the player to get up to level 10, there just is no reason to have a stupid high level. Also only 10 stages in my mod, just gotta theme 'em properly. So 1 level per stage, no grinding, levels gained from bosses a la Metal Gear Solid, levels shared amongst the whole party because individual leveling just pans a game out and makes it a boring grind, no exceptions. Level up is caused by getting a thing from bosses and doing… something with it, not sure yet. Also each level unlocks a new ability, abilities of your second subclass are only unlocked one-by-one after level 5.

So the highest a stat can get would be 35, due to a subclass restriction of only 2.

Stat = (statfocus*level) + stat.default

I should also point out subclass 2 can only be assigned once your party is level 5.

i hate unity's bullshit

You can ask, if you want any help and I'll turn it into a plugin and open source it once I have a new PSU.

nah, i have options, they're just really tedious
either i make duplicate classes for all of my visual scripting just so i can save stuff, or when saving i go through every single node and check it's fields to see what i have to save/load
and then i still have to mess with id's for object references

Good luck then. Serialization is annoying. I had luck because I adopted a really good serializer for Unity that the original developer abandoned. Although over-engineered for most purposes, I learned a great deal and it made the development of the UE4 one relatively easy.

I'm working on something you guys here might like: A Blender plugin to copy pic related and make texturing architecture much, much quicker.
Unreal has BSP tools, but you're supposed to convert them to meshes (because BSP generated geometry is awful). The problem is that I found that manual texturing took way too much time. So now I'm writing a plugin.

really activates my almond

you have to use Unity's serialization if you want to have hotswap during playmode

Now my rudimentary piece-of-shit zygote of a game actually feels like a responsive shootan when looking around. Warm fuzzy feelings.

Let me test it boyo. I've played video games since I was 2 years old and I've developed a good feel for shooters. A good sway is the basis for a good feel of movement. Don't fuck it up!

...

oof I fatfingered a reply button

Anyway, yeah once I got a playable demo with the basic gameplay cycle, I'll post it on a demoday. I hope I won't get torn to shreds too hard.

more improvements on the hands

would hold gently

Guess not,if you're around let me know how your project is going.

This is actually how I learned c/c++. Running around like a madman trying to get x11 to work with opengl/at all.

Why is this allowed?. Is shit like this system's or Unity's fault? I'm unsure which one to blame.

maybe what is happening here is that Unity is processing buffered I/O. So, if you press and hold 'x', this will start generating a ton of keypress messages, and buffering them in a message queue. So, what happens is. you will generate a ton of key press messages and then eventually, the program will handle all of them, ensuring that it doesn't lose any keypresses. For example if you have ever used a program that has a text box, and you are typing into it, and then the program hangs and stops updating the text box, if you keep typing eventually it will stop hanging and then process all of the queued key presses, putting all of that text into the text box. When without this system everything you typed during the hang is lost.

So, let's imagine that in your unity editor, it undoes things slower than you are generating the buffered key press messages, so that means the longer you hold down ctrl+z the more messages are queued up, because it can't process them at the same rate you send them, and so when you let go, it still has all of those key presses in its queue, and none of them have been lost. So, it continues to process everything and no input is lost, causing the scenario that you just experienced.

So, while this design is very useful, and its very reasonable for it to exist as the default way to handle keyboard input and this is generally the way that window managers like win32 and X11 handle it (programs would be pretty bad if dropping keypresses on any kind of slowdown was the default behavior), Unity should have just read the state of the buttons directly to see if it should have performed an undo (so that it ONLY does an undo if you are holding the buttons RIGHT THEN) instead of checking the buffered I/O queue for the keypresses.

both, the 'undo' action took longer than your keyboard spammed the ctrl+z command (this speed probably can be adjusted in the keyboard options is the OS), so it added up more and more keyboard events.

Unity can only either do that or miss input events from time to time (Which I consider worse). It can't do neither

Interesting. Could Unity somehow differentiate which ctrl+z's in the buffer were "held down" and which were "pressed once"? There is no sense in executing undoes from a buffer if the user just held down the keys because there is no chance he knew where exactly he'd end up like in my case and undoing should be cut short when keys are no longer pressed. But if the user was pressing ctrl+z fast enough to add, for example, 10 presses to the queue, it's obvious he knows exactly how many undoes he wants so no problem with executing those.

nope, or so I assume, you would have to research how (I assume you use) Windows' events work, but i'd guess it just repeats ctrl+z presses without telling it was because the button was held down.
Also, although there is some solutions to your specific situation, they would break other's.
It's a trade off

alright, someone tell me if this is a stupid serialization solution or not, because it's gonna take forever to make
base class, inherits from scriptable object which is why i can't directly save it to a file
public virtual List Save() { return null; } public virtual void Load(List _sf) { //load whatever values child classes have from saveFields }
example from a child class which represents a bool
public override List Save() { List saveFields = new List(); saveFields.Add(value); return saveFields; } public override void Load(List _sf) { value = (bool)_sf[0]; }
and then in the manager script for all the save stuff, i save an object which has a ton of List for each character*node*value
and when loading, for each list of lists of lists … i set the values to the appropriate existing object

yes, the input and output are not byte[], so I'd rate it stupid/10

its good that godot made this Q&A website
but it kinda opened the gates of vidya hell once again, hasn't it?

Yeah this problem is completely solvable, my map editor that I am writing myself doesn't have it for example

it has its own series of subtle bugs and problems, of course

well it worked, at least for the bool node
now i just need to do it for around 15 or so classes, and every time i make a new inherited node class
thanks unity :^)

Just use these typings from second chart there.
en.touhouwiki.net/wiki/Touhoudex/Type_Chart
And use this for reference.
en.touhouwiki.net/wiki/Touhoudex_2/Touhoudex_2

...

Easily
Ones that are followed by a key up event are pressed, ones that are held are followed by a key down event.
The problem of holding down can be then be solved by having a set max iterations of undos per second when in a held down state.

For the load method can’t u just use a generic? No need to manually do type conversions

why don't you use version control
also, why are you using unity

don't i need to know exactly how many types i'm working with? either way i still need to specifically choose which variables get saved or not

Sounds like a job for reflection

doesn't matter, i'm done by now and i don't see much of a point in changing it

Adding bones to butt and thighs for extra softness

at least until i implement marriage in the game

i'm too much of bitter and stubborn old man to relearn a new set of types and weaknesses

...

Does anyone know of a game that takes the "billboard" 2D sprite kind of look (a la classic Doom) and just gone fucking crazy with it?
I was thinking, the reason the billboard effect looks like shit (see Mario Kart 64) is because the artists were limited by memory. There was only so much room on a cart or CD, and buy the time storage mediums had caught up the industry had more or less moved onto 3D. So at best we're talking maybe 8 angles of rotation that were captured for the sprite sheet (front, back, left, right, and the four diagonals). But what if, using something like Blender, I captured 16 angles? 32? I mean, I should be able to have megahuge sprite sheets these days in most engines, correct?
So in the same way Donkey Kong Country was able to display 3D models that the SNES hardware could never ever render in realtime by pre-rendering them into sprite sheets, how many angles would it take to become fairly convincing? Because we can still pre-render models that even high end PCs would struggle to render in real time (and bake in certain effects that would be expensive to do real time).
I was also thinking that with the overhead this might save, could we have actual, serious crowds that are more than blobs of units like in Total War games? Could we actually have 50+ independently acting billboard baddies?
Obviously I can test this stuff out but I wanted to hear from the community before wasting my time.

You can look at gameplay of Sanic 3D Blast on the Genesis, that uses 16 directions for at least Sonic himself. If that doesn't look smooth enough for you, then 32 would be enough in my opinion.

Once your billboards reach a certain resolution and you add a shitton of angles and frames, they'll end up using way more memory and GPU processing power than a similar, decently optimized game using 3D models. Alpha transparency is also surprisingly taxing on memory bandwidth and it gets worse as your rendering resolution goes up.

easy solution to that, just break your images up into smaller images with very little transparency

Use billboards for distant LOD. 32 angles times 7 vertical angles, plus 90 degree verticals.

hooktube.com/watch?v=44KZ0neSz7E

These guys got 65,536 enemies onscreen at once on a goddamn PS2.

Nice example, thanks user.

I assumed there would be issues, the industry must have not gone this way for a reason and I'm hardly the first person to think of it.

Impressive as fuck, damn.

I'm retarded, I thought this was a scene at a monastery or something at first glance. Like I thought the leg bones were candles or people, and the ass curvature was a fancy window overlooking a blue sky or something

In black I have the 3D representation of my closed polygon
In red I have the 2D representation of the same
In green I have the 2D result of a deflation algorithm.

How do I convert the 2D green polygon back to 3D?

Would you rather have a 4k texture that only shows some of the possible actions, or have a 128x128 texture with a few extra bytes of memory dedicated to describing vertices, transforms, and an animation, and be able to view it from any side, angle, or action?

Hint: You can also do an in-place render to make it look like a prerendered sprite, then draw it to the scene

Well how did you to do the transform to project the black guy into a 2D object? Do most transform models not have an inverse transform thing you can use to undo it?

user…..

Yeah, these days billboards are mostly used for LOD. They're a useful trick if you know where to use them, but using them carelessly can absolutely murder performance in memory-bound situations.

public static List Planify3Dto2DPoints(Vector3[] points, Plane _plane, out Vector3 centroid) { Vector3 sum = Vector3.zero; foreach (Vector3 mv in points) { sum += mv; } return Planify3Dto2DPoints(points, centroid = sum / points.Length, _plane); } public static List Planify3Dto2DPoints(Vector3[] points, Vector3 centroid, Plane _plane) { List m_points = new List(); var rotation = Quaternion.LookRotation(_plane.normal) * Quaternion.FromToRotation(Vector3.forward, Vector3.up); for (int i = 0; i < points.Length; i++) { var vec = RotatePointAroundPivot(points[i], centroid, rotation); m_points.Add(new Vector2(vec.x, vec.z)); } return m_points; } public static Vector3 RotatePointAroundPivot(Vector3 point, Vector3 pivot, Quaternion angles) { return angles * (point - pivot) + pivot; }

Let me tell you this was mostly done in try and error without understanding the underlying maths behind it.
I tried reversing this, but got some very strange results

I'm not looking to replicate the billboard style/look as much as my thoughts were, how far could it be taken? Would the net end result to the player be improved if we pre-rendered a bunch of sprites ahead of time instead of rendering 3D models in real time? Could a modern engine handle 1000+ billboard baddies better than 1000+ 3d models?
Like the billboard method could, in theory, use real photographs, similar to attached. Boom, if we have enough angles to be convincing we just successfully navigated the uncanny valley, right? Can't be more realistic than a photograph, right?
Which would look better: an attempt to 3D render a model wearing some cool costume, or actual real photos of a human being in a quality costume? Cosplay has gotten crazy good these days.

That's what I was worried about.

my attempt at the reverse operation
public static List Spacefy2Dto3DPoints(Vector2[] points, Vector3 centroid, Plane _plane) { List m_points = new List(); //var tmp = centroid.y; //centroid.y = centroid.z; //centroid.z = tmp; var rotation = Quaternion.LookRotation(_plane.normal); for (int i = 0; i < points.Length; i++) { Vector2 v3 = new Vector3(points[i].x, 0, points[i].y); var vec = RotatePointAroundPivot(points[i], Vector3.zero, rotation); m_points.Add(new Vector3(vec.x, vec.y, vec.z)); } return m_points; }

"Lighting artists lay out all of the light sources in the game world, both
static and dynamic, and work with color, intensity and light direction to
maximize the artfulness and emotional impact of each scene."
Can you imagine having a job this mindless?

The monastery is not for lewd.
Except you'd be limited by the framerate you capture the sprites, and the sprites' lighting wouldn't change at all.

Lighting is actually a pretty huge part of making a game look pretty, and it takes a bit of effort to get it right.

Good lighting is one of the big factors that separate an amateur looking game from a professional looking game. It's more difficult than it seems on the surface to get right.

Thanks for thinking through this with me, anons.

Heh

Let's not pretend they're doing anything other than moving pre-built sliders back and forth until it looks nice

wait until she is fully animated and bouncy

Its indeed a holy thing

I tried a thigh bone too but found it less realistic than using butt bones that have gradually lower weights towards the thighs (thigh jiggle really is subtle, and goes in conjunction with butt jiggle).
It also made it easier to work with, as its fewer bones, and fewer touch ups I had to do for my jiggle system (spring damper soft body dynamics).

Why are you even posting here? Video games are about designing interconnected systems and aesthetics are a major part of this which can greatly affect gameplay. If you cannot see this and only view each piece of the system in isolation, you shouldn't be working on video games.

Small correction, sometimes they also have to choose from a color dropper.

Holy shit this is really cool

yeah, im considering working with just the butt, its being pretty hard to handle the weights on the leg movements

But at least i figured out how to improve the hands, rounder nails makes it cuter

Any action sounds dull once you reduce it to the raw motions: pressing raised surfaces in the correct order, wiping a stick with wet fibres over a canvas, and so on. If you view every action in isolation and ignore the wider context, everything you create will be a disjointed mess.

Here's an actual lighting artist talking at length about his job.
hooktube.com/watch?v=avKlUcx3TlE
Skip to 1:20 to get the actual guy talking, skip to 9:15 to skip to him telling you what he does.

Except some actions ARE essentially just their raw motions. Artists should be treated like assembly line workers, because the majority of them are completely incapable of being anything else. It's just nicer to move sliders around than to work in a cannery.

Ah, you're the "artfags should be mistreated" retard. Enjoy being the sort of hyper-specialized assembly line bugman you accuse everyone else of being and what media Jews dream of shaping mankind into. Incapable of seeing anyone else's job as anything but a series of raw motions, a cog in the machine overseen by by a mastermind (((chosen one))) who drip-feeds his goyim instructions and ties their isolated work together into his creation.

...

Whether the user I remember is you or not, enjoy being the splitting image of the AAA cancer Holla Forums despises. You're not going to make it with your mindset unless you give in and join the AAA workforce as a disposable drone.

You need more images.

A reaction image folder grows quickly, unlike your bugman mindset.

I have as much time and autism as I need to make a game without wage slaving or hiring a guy to move sliders around.

Assuming you're 18-25, you probably only have about 30-40 years, or approximately 350,400 manhours, or about 116,800h factoring in sleep, meals, life, etc. It's a very finite amount.

thanks user, when I said I had as much time as I needed I wasn't implying I was immortal.

Then learn it yourself, faggot. There's a lot more to getting lighting right than moving some sliders even if you're baking everything or using some fancy global illumination system.

Gamehut, the TT guy who makes videos about various tricks he pulled off on the Genesis, just started a series explaining (in very simple terms) how to actually develop for the Genesis.

That channel is 90% clickbait titles and patting himself on the back he's a good programmer, though, so I'm optimistic

You are just jelly.

...

Well, I am sure that many folk wanted to know how he did it and he gave you the answer while getting ad revenue from idiots with an eyecatching thumbnail. As I said, Jell-O.

is this that reddit crossover I keep hearing about?

Audio math is scary
archive.is/wVGXm

user, I found the perfect game jam for you!
itch.io/jam/non-game-jam

>itch.io/jam/non-game-jam/topic/191000/assets-pack-for-sale-you-can-use-it-in-your-jam

I feel like this was just a contrived way for him to shill something

I'm upset.

once I realized that nearly every tutorial or event is just a way for people to part goys with their shekels, things started making a lot more sense

wew

Somebody submit a walking simulator.

I've got an idea for that, but I'd need something that can easily host some audio files for streaming. I'd have a small game file that would load them in as needed. Pic related.

Still making animation assets. I'm up to 30 animations.

Very nice and soft looking!

I would not do that. If you are gonna weight paint with that topo, keep your modifiers on. You are doing yourself a disservice. Applying subdiv = 4x vertices to weight paint = 4x harder = 4x longer. NOT 4x easier. I also once thought that more vertices to paint means better deformations. Which is true sometimes but your triangle count doesn't actually change by applying the modifier. It's the same. So really you are just making yourself weight paint 4x the number of vertices for the same topology and triangle count. it defeats the purpose of a modifier. You will also get uneven non-smooth weight paint all over the place that will be unnoticeable by you. Your mirror and subdivison both are powerful time savers.

I've worked with my model for months now and I still keep my modifiers on. Even my face, which uses mirror and subdiv, has shape key controls for both sides of the face that ignore the mirror modifier by using a vertex group blend. Pic related. This saves time, makes edits FAR easier to symmetrize, and overall much saner. I would go back to your backup and transfer your weight paint so far to that one and work off of that one from now on. You will save days of time.

The only time you should apply your modifiers is when you are ABSOLUTELY DOGGONE ready to finalize it for Unity. This means post UV-unwrap and post-weight-painting. Or you have a rare shape key that would collide with the middle of the mirror modifier.

You need to look at illustrations a little more. Your hand is missing a general curve to the knuckles.

Making videogames isn't hard, getting my lazy ass to stop worrying and just fucking work is hard.

Better that then never.

I know its harder with them applied i applied them on Adam too but for me its worth as i can make the weights more precise

By the way Shinobro, are you going to rip her voice and lines from the animes? I think it would add a lot to her, i wanna hear the Kaka when she smugs

its prettier without it

gamedev is weird,
sometimes you can't just push through either,
take me for example, yesterday I was struggling with deciding a object design, I was only getting sleepier and more tired, unable to get out of the thought loop
now a new day, everything seems clearer, and I basically know exactly where to go
on the other hand, you need a bit push through to get momentum, specially when starting a new feature cycle

user, if I was to go with mother nature's opinion or yours, I'd say I'd go with mother nature

What do?

C++ is almost entirely backwards compatible with C, so a C++ compiler will eat almost any C code. As a result there is nothing forcing you to use all of C++'s features you consider bloated. Just restrict yourself to a sensible subset of the language.

You should ignore what says and use C11.

bitsquid.blogspot.nl/2009/10/two-way-serialization-function.html

Why though? Haven't looked into newer versions of C yet. Currently have to learn Python.

C#

Sounds like excuses to not get anything done.

Yeah, mother nature also made these, pretty nice eh?
I prefer to keep the anime style, thanks

Rust of course!

sure thing, forget the knuckles, but hands do have a shell like curvature to them

No thanks. Fucking Ada will give you similar safety without Rust's retardation.

What's the legal situation of using the old Unreal? Are you bound by the terms of the modern engine or is it something entirely different?

You won't get the better precision you are looking for and it will be a waste of time.

Show us that in motion because it looks absolutely terrifying

No that would be a nightmare to clean the audio up. Probably gonna hire a voice actress but way later in the future. Also I really really think you should not apply the subdiv modifier. You don't really get extra precision. All you get is 0.5 more weight more control on the middle of a vertex. Which you won't notice if you compare them side by side. You also defeat the purpose of topology by applying it. I really think you should work on the subdiv version.

Yeah those are pretty fucking gnarly.

Serious talk though - I'm not trying to be mean, but pretty hands are at the very least anatomically correct. I might be treading pretty close to being a straight-up autist, but I have a bit of a fixation on hands and I get extremely spergy about them when artists draw em to an unhealthy degree.

I'm not a fan of rust but that is not humanly possible
Also if you think C is bloated or convoluted you might have brain problems

Have you tried changing the output types to Vector3 and returning the result of RotatePointAroundPivot directly, instead of Vector2(vec.x, vec.z)?

I didn't think you were being mean or anything, i always welcome critics, and im very spergy about hands too, but the style im aiming for is very different from yours

does someone have that $2000 commission that was floating around here a while back? whatever happened to that guy?

...

...

what features do you need that C doesn't have?

This is false. Rust is in general slower than both C and C++. This benchmark doesn't even use the fastest compilers you can use.

She's a big girl.

...

I didn't mean to imply C was bloated, probably should have stuck with convoluted.

Where's the Kira "I had a boner" image when you need it?

Yeah, but if you look at Mallow's hand from above it still has a curve along the knuckles

oh, so that is what you mean
Mine has it too, its just harder to see

I'm not the user who pointed it out, though I agree with him that it'd probably be better if it were more pronounced. I also think you might need a bit of a tweak for the pinky side of the hand. Even without visible knuckles, there's a bit of a protrusion where it would be.

Goddammit user.

I gonna modify the knuckle a bit, also that is the old hand, the pink nail one is the more up to date, the pinky side of the hand has been fixed already

there will be insects to collect Animal Crossing-style and their extracts will be used for Alchemy

1 3 and 4 aren't spiders
from somebody who is anal about their arachnids.

Do they feel good in there?

Cheeky bastard,you know what i meant,i love tarantulas.

I never said they were, Camel Spiders are still nasty even if they aren't real spiders

I'll have you know camel spiders are beautiful too,they just need the shade from you.

Even if they are harmless, I still prefer these

Look up hyllus giganteus,the biggest jumping spiders around and yet they're still small as hell and adorable.

Okay listen here, smart guy

I fully confess to being Kira-tier autistic about hands and also being a nitpicky, perfectionist asshole.

I wish I could say I was also a master manipulator and had a cool stand but the similarities end there

books
worth reading for anyone using C#, very through in covering the language

There's probably a bit of good information hidden in here but it's mostly for laughing at. Features lots of poorly drawn scribbles. I think this guy worked on spiderman, or something.

Griefing. Griefing is when a player will harass another
player by constantly killing them. Design systems and create
a rule structure for your world to combat this bad behavior.
Make sure your policies are clear to the players, and a
warning/zero tolerance system helps keep the peace.

Griefing can be a serious problem. I know it's popular to be le dank gaymers around these parts but when you're tired as fuck after a long day, and you just want to play a game - the last thing you want is having to deal with an autist who completely exhausted a game and the only way to have fun for him is to ruin fun for others. I don't exactly know what you mean by laughing at the document - to a layman who's just recently getting into the industry (such as a middle-aged artist who doesn't know games) such a book would probably be immensely useful. You gotta understand that gatekeeping and preventing people from getting into games by being elitist is just lame - same goes for people getting into gamedev for various reasons.

Myself 5 years ago would barf at what I wrote above. I guess I'm finally mellowing out.

I'm uploading the book for everyone to read, nigger, I'm not gatekeeping anything.

I was assuming you were implying that what was contained in the book was all babby tier laughable shit, and I assumed you were in particular referring to the griefing section.

Some of it is good, like explaining how to add landmarks to stop the player from getting lost. If you're tired after a long day, user, don't go online and put yourself at the mercy of strangers. It's just not good for you.

Waste of quads.

I'm addicted to videogames and sometimes when I'm tired but restless I keep playing because what the fuck else am I gonna do? Devving when tired like that results in garbage code that needs refactoring right after anyway so there's no real point.

It's a good thing we don't have a limited supply of those, right?

This is the lowest level book on C# I've come across; It covers a lot of nuances in the CLR. It focuses heavily on performance. It spends 70 pages just on the garbage collecter, and frequently examines the actual generated CLR.

Large objects (85000 bytes or more) are stored in separate heap that is collected infrequently

If it's worth anything, I've looked through dozens of books that are absolute garbage. I haven't come across a Unity related book that's up to the standards of any other industry. Game development books tend to cover a very rigid series of steps without actually explaining anything, which is great for pajeets who want to clone existing games, and outright harmful to everyone else.

This book's useful if you're not already comfortable with certain design patterns. Even if you don't use them I think it's good to have different ways you could structure your code in mind. It does consider performance, but obviously a book on design patterns is going to want you to add an extra layer to your code. There's an interesting section on creating and compiling your own bytecode for implementing game scripting.

and this one is for masochists
2004 - Gender Inclusive Game Design (Charles River Media)

Though there was much excitement
over these new titles, there was also disappointment on playing
them. One of the women mentioned that she had purchased a popular
fighting game for her PlayStation® 2 (PS2). She considered herself
pretty good; but one day, her boyfriend challenged her to play.
He promptly bested her.

She was disappointed, particularly when
she noticed he was using all sorts of moves she had never seen before.
There was nothing about those moves in the manual, so she asked
him where he had learned them. He replied that he had learned these
moves just by pushing the buttons a lot. This made her quite unhappy
with the game, so much so that she ceased to play it. The machine
as a ‘foe’ became a barrier to her enjoyment.
Thus, to avoid creating these barriers for a percentage of the potential
audience, the game mechanics should be intuitive and easy to
learn.

I gotta second this. A lot of people believe in learning by doing, and somehow they think that "doing" is replicating rigid steps.

I'm personally having a fuckton of trouble with figuring out how to properly make my inventory work in my sandbox autism simulator and I have not come across a single tutorial for UE4 that explains how to save an actor's current state as-is to "pick it up", destroy the original and spawn a carbon copy at a later time. Most tutorials I've found rely on using a rigid data structure to reference an item. However, this implies a standard and classification for every item. This means that if there are deviations from the norm, I'll have to make a new struct specifically for a new case, which is silly.

I fucking know there must be a way to save what's currently in memory, save it as a binary blob and then spawn a carbon copy of that actor - given you specify the class. I'm probably way off the mark as to what's actually done there - but I'm tired as all fuck of trying to figure this out. This particular area of knowledge is so wildly out of my understanding (data management in particular) that I don't know where to start.

Anyway, this turned into my crying about my woes. The point is - replicating clones does not help me learn how to solve unique issues.

Related.

Are you referring to the concept of a copy constructor?

...

I'd settle for his /fa/shion sense, honestly.

...

I don't think so. Through osmosis, I think I'm supposed to be looking at serialization but I've honestly got no idea where to even look. Maybe I'm too wishful, really. Maybe I'm wrong to expect that this would be possible - but I wanted to write a class that would play nice with an inventory class, and forget about it; Inheriting this class whenever I want to add another item and just expect it to play nice no matter what kind of data I slap on top in the children of that class.

That last bit is starting to make this whole thing sound a little too convenient to be possible, I guess.

Creating a system that effectively punishes griefers and that cannot be exploited by them to abuse other players is almost impossible
Griefers are usually the most ingenious and creative players in any multiplayer game, you need to think like a grifer when you create mechanics, "how can this be abused?" must always be in the developer mind

Here you go, user.

That doesn't sound unrealistic to me, I think you could leverage inheritance in a way that enables that

So I'm learning that creating fun, turn-based, rpg battle systems is actually really hard. I've been playing other RPG's recently and I can't seem to think up a system that's a good balance between strategy and tedium. For example, most jrpgs become tedium once you figure out the optimal strategy.
Once you find the strategy in these types of games, the fun just kind of disappears. Usually the enemy can do things like call for backup or land a crit, but not much can keep you on your toes once you've found the way to victory. I've thought of a few ways to counter this, but I'd like to hears some thoughts on what people consider fun turn-based combat.

Thanks my dude.

Are you talking about JRPGs specifically? I really liked the pseudo-rhythm combat of Mother 3

How the hell can you prevent people from getting into games? Is there anyone whose first game was a multiplayer one where you can grief people?
Sage for offtopic

i'd sage too

Not necesarily JRPG's. For example, Darkest Dungeon has some satisfying turn-based combat, and has this rare issue where the combat itself isn't tedious but the overarching gameplay is. Mother 3 is decent, the combat's not very deep but learning the rhythms of the background music was a cool way to incorporate some realtime elements into a turn-based battle. Same with the rolling health meter, it's always a great feeling when you don't heal someone who's health is rolling to zero and you manage to get some hits in and end up saving them in the nick of time.

Oh yeah the health meter was cool. I agree with the Darkest Dungeon sentiments as well, I liked the all the positional elements in it.

Team building is usually a lot of fun for me in RPGs. Have you played the Labyrinth of Touhou series? It has team building - which you can change on the fly and adapt to what you need, as even inactive chars get like 85% of the active party's XP - some degree of meaningful resource management in dungeons, and a good deal of strategy in boss fights because of its active/reserve party system where you have to cycle through who's actively fighting to keep your characters OK on health and mana and have everyone who's low rest up, doubly important since the only way to revive team members is to return to town so you better fucking keep them alive.

Sounds interesting, I'll try it out. I always liked how resource management worked in games like Persona and Etrian Odyssey. SP/TP management always leads to interesting behavior, and it can change the difficulty of a game depending on how stingy you get with it. Although these games lack good dungeon design, and I never liked most maps in dungeon crawlers. EO tries to fix this in later titles where the direction of monsters matter when you get into battle, or enemies can move while you fight and spoil your fun, but level design just seems to be an area where turn-based RPGs always struggle. I'd like to keep the mechanics of my game as "levels". I made this shitty pic related to try and convey what's missing from most turn-based RPG, and highlight the things I want to focus on in my game.

I remember playing an old flash game a while back, on the hardest difficulty you only had a short time to make a move, and on top of managing mana, cooldowns, and a shitton of buffs and debuffs, and also how you built your character was important too.

any recommendations for a good starting point to program enemies in 3d in godot

Any idea what I should be looking into to figure this one out?

Morning: studied math
Lunch: wrote my story/script, as well as sketching up the bad guys (left is the admin and right is the basic goon)
Afternoon: worked on my base meshes

Tonight and tomorrow I'll work on studying more math because I'm pretty lacking in that area.

I can't really say anything because I'm still studying so my words are a grain of salt.
I feel like there will always be some kind of exploit in rpg turn taking games, whether it be something like badges in Paper Mario or something else.

So what I've been thinking is: First make some battle system, list out biggest flaws. Then try to fix the biggest flaw (by changing the damage formula or adding moves etc) which may create another flaw, the goal is to make the flaw unnoticeable, or not significant as possible.
like I said, I'm saying this mostly out of what I observed from turn-taking battle games and I haven't done the math.

Why do you need to clone something out of memory? I can't guarantee help but I'm interested

I own a hardcopy of this book. Very good read; it's not for beginners to programming, nor does it explain general programming patterns. Instead, it covers the design aspects of the language, the needs of C# and how and why it works the way it does under the hood. It gives you a good appreciation for applying best practices.

Notable C# experts on Stack Overflow, whose answers are usually fantastic: Jon Skeet, Eric Lippert, Hans Passant, Marc Gravell, and Reed Copsey. Skeet is basically autistic about C# and more or less works exclusively in it and knows it very well, and Eric worked on the C# Microsoft compiler team. Any of their blogs/articles/publications are absolutely worth reading

The fuck kind of object would ever get that big? The only thing I can think of is a non-managed texture object, that gets cleaned up by you (or your engine) anyways, or an array that's huge.
85,000 bytes is over 20,000 integers, or roughly a texture of 145x145 or greater, using a 32-bit format (eg RGBA).

Older versions of Unreal are completely free to use. The Killing Floor 2 devs intentionally used Unreal Engine 3 and reverse engineered features of 4 to get around having to pay Epic.

Have you ever generated a mesh at runtime?

Okay, so here's the end result I need to achieve. Keep in mind, I've only been coding for 7 months so I might be approaching this completely wrong:

Let's say a "thing" is an actor class of anything that a player should be able to pick up.

1) Character has an inventory, it's an array containing "things".

2) When "thing" is picked up, I want to somehow store every single variable it has as a "footprint" of precisely what it was at the moment, in the inventory, and then destroy the actor.

3) I then need to be able to produce it out of the inventory by spawning it using the data I stored above.

The thing is, it's not certain exactly what this data will entail when things get more specific. It could be an array of long-ass strings in case if the item is a book; Or it could be calories, hydration, and culinary properties of a food item, if it's food. It could also be a weapon, in which case it would have properties pertaining to a weapon - ammo count, condition, etc.

I'm looking for a catch-all solution to maintain object persistence when they are picked up by the player, basically.

Nigga jesus christ just use 2 classes
When the layer picks up a WorldItem, its InventoryInfo is added to his Inventory and the WorldItem is destroyed.

Okay, well I might be retarded. How would this work if my InventoryInfo is wildly different depending on the item? What if the item has, say, a simulation of a pocket universe that has to be precisely replicated when the item is spawned back into the world?

What's the difference between "every single variable of a thing" and a thing itself? If you have all the information representing a thing, why would you not have a thing? As for the concept of things, can't you just have an array of references to a class that you subtype for different types of data?

Does a hand fetishist masturbate using his hand, or does he hold hands in such high regard that he wouldn't dare masturbate with his hands since he thinks that's too lewd?

Do your own hands count or do you need other people's hands to get off?

It sort of sounds like you just need genericization on the part of the information and a regular interface for the inventory to access it. That's a strange example

I don't care about my hand the same way I don't particularly care for my dick despite being 50% a faggot. It's just a part of your body, and you feel desensitized to seeing it. But for example handjobs are my personal favorite.

Well, fuck, well I mean - the thing I'm trying to store can have different variables depending on the actual purpose of the thing derived from the class. If I have InventoryInfo and I store it, wouldn't I only store things that are generic, and lose the data specifically pertaining to a special use case I have on hand?

I'm sorry, I'm just not very flexible in my programmatic thinking yet, or maybe I'm not explaining my particular issue well. But I'm just not quite getting how having what essentially amounts to a "dogtag" indexing items to make them palatable to the inventory system help me save the individual, unique info that is contained in classes derived from the basic item class.

interesting adgdgdgdgdgdgdgdg this time

That doesn't apply to everyone, I know someone who admitted she gets off to sucking her own tits.

Well I mean it depends on what you're doing with it. If you're just jerking off there's not much fun to be had in focusing on your dick only.

game developers are almost all sperglords, and not even the likable kind

Yeah I don't really know what's up with that. I'm personally a massively abrasive, snooty cunt and I've been trying to work on that for a while now. I'm not sure what's the common trait that gamedev autists share.

She has taste.

shid

you dirty fucking cisgender christian white neurotypical ableist males how dare you for some of us walking is a challenge in itself how dare you games should be playable even if you can't see the mouse behind your fat rolls

Is running a gameplay?

which amounted to "look at our cookie cutter configuration no.4"
which amounts to "random npc spouts hey I know you did that thing"
arguably this is wrong as I'd say that can be a game mechanic which could be made into gameplay but no one bothers figuring out something fun
no one has ever done this correctly
it's the other way round, make the player compliment the companions since the player has greater control of their character and thus would be better fitted to change based upon how the companions handle situations
kinda a obvious point
usually amounts to "a key for a door or take damage/do damage"

But user Gone Homo has a great walking gameplay mechanic :^)

walking gameplay can be summed up as movement gameplay which is the foundation of fps games, sidescrollers, platformers and fighting games. to exclude it out of design on the whole does nothing but ruin the potential of your game tbh

what kind of game? also seconding this for an arpg
look at the available nodes and their signals, then use whatever seems useful raycast, animation trees and visibility notifier

If you figure it out, post the answer.
This has happened with entirely different sets of people too many times to be a coincidence, so there's clearly something wrong with the way I conduct myself. The current theory is that I'm abrasively opinionated even when not trying to be, and try to simultaneously ingratiate and emotionally distance myself from people through humor and ribbing that are too harsh for most people's tastes.

You at least got called out as an asshole IRL. I on multiple occasions have conducted myself as a completely irredeemable asshole to a degree where even anons called me out. That there was a wakeup call.

*called out on being an asshole

english iz hard comrade

Third person is the only definite thing right now. I'm just learning Godot and have player movement, a basic camera, menus, and some more. I've got a box that damages the player already. Now I need to figure out a way to add weapons and attack.

try having everyone you know and think that are your friends suddenly turn on you because you were so unaware of how obnoxious you were and end up with a rep for being a massive faggot but the worst part of it is being made out to be a pedophile because those same people used a lie against me spreading it around to such an extent that fuckers who weren't even apart of the whole drama bullshit know of me and to stay away due to it

I was stupid to do so, but at the time I felt like anons were to a degree a "friendly" presence. Like, I totally get what you're saying, since I'm sure it's much more of a shitty thing face-to-face, but at the time it felt like the only place I felt like had people with a similar mindset just ripped into my ass like hounds. I was shocked at my own reaction, really - it honestly hurt, like a long-time friend turned on me.

Did the lie have any truth to it? What caused the lie?

Ah, so just the classic zero-tolerance smear. Cunts. I know it might not feel like it, but they were not worth being friends with.

it's less that I lost those people but more they prevented me from gaining new ones

Made some progress on my engine lately. Completely rewritten the collision system, particle system, integrated it all into my map builder, made some improvements to the renderer. Added support for multiple shadows at once - the idea is to have directional shadows for static objects and top-down shadows for the player, enemies and bullets.

I dunno how to even deal with that, man. Sorry that happened to you.

what game are you working on?

Autistic sandboxy type of deal. Working name is Sæglópur (Say-glo-poor) but I noticed people are too confused by the pronunciation and there's also a song by Sigur Ros with the same name so I'll probably be changing it to "Lost at Sea" or something to that effect, instead. Doesn't look nearly as cool when written out but at least people can pronounce it. I feel like the fancy name will hurt marketability by word of mouth.

hey guys I heard c8f4dc is a pedophile

Really? What a sick fuck. Don't be friends with that guy.

QWOP

I'm very sorry to be bringing my problem up to the top again, but I'm kinda bottlenecked with this inventory shit right now - I can't continue devving until I figure it out because I gotta add weapons and equipment and shit

i read the whole book a couple years back. surprisingly good read; it basically explains why women suck at games. they want the game to handhold them all the way through. i might be misremembering it, though.

Have you considered just not rendering the object anymore, and making the inventory just pointers to objects?
Say you have a Gun:pickupObject, and instead of deleting it when picked up you just remove it from the world but keep it in memory, and then taking it back out of your inventory you just place the actor back in the world at the new position.

You're right, that's definitely a way to do it. Any idea how to properly achieve that in UE4? It seems that destroying the object actually removes any trace of it. Should I just stop it from ticking, make it invisible and let it sit in space like a ghost?

rule number 1: be attractive being attractive and being good looking are two different things and most people can be attractive

That usually boils down to being rich and being able to buy things for people, or just being a doormat, always being there for the people when they need you but never bothering them. At least from my experience. I've always been too inconvenient for people, so I checked out of the whole system. No benefit from it, anyway.
Polite sage for blogpost.

only if you can jump and there are obstacles to jump over

user, that's absolutely not true. Charisma is a skill that can be learned.

people will shorten the name if they can't pronounce it

that's really neat but the time stuff and heal are displayed at the top of the screen separate right?

well even temporary things like getting dirty but also it's more a player character doesn't have to be a realistic human in a realistic setting so it could be something like "a dick raises up on your character which means a player can do a different action"

Do you really think "Sæglópur" is an acceptable title? I absolutely love the way it sounds and looks and I'd love to keep the name. I've been told it's a mouthful by a couple folks but man, I really wanna name it that.

my natural inclination would be to pronounce it as singapor though it'd come out like slagopor, what you could do though is have a sub title that's easier which could also be used as referring to it specially if it can be a good abbreviation on

A subtitle is fun. That's a great idea, user, thanks!

Alright /agdg/, I want to get into this game dev shit but I'm retarded and know only the bare basics of technical stuff that comes with development. Tried reading through SICP but my acoustic brain can't process beyond a certain point. Basically I want a game engine capable of the following with little experience in gamedev aside from fucking about in Doom builder, if that even counts.
Standard game engine shit aside from that, which I imagine is stuff I can easily edit. It's just that those are the most complicated things from a technical perspective from my point of view. From my research I think Unity is the best candidate but I'm open to suggestions. I'm just tired of being an ideaguy.
Also, can someone learn C# as their first language with little difficulty compared to something like Java or Python?

it could be "Lost at Sea" too though I'd have it something like "Lost of Baston" so it'd be shortened to SLOB

C# is often lovingly referred to as "retard C++" so you shouldn't have much trouble. I know I didn't have any when modding Unity games.

Well the problem there is that "Sæglópur" already translates into "lost seafarer" or "lost at sea".

I'm certain you'll find something you'll like then

I sure will! Man, I didn't even think of a subtitle, though - thanks for nudging me in that direction.

no worries!

That's reassuring. Any advice for pitfalls to avoid while learning or general gamedev advice?

wrong, wrong and wrong
being rich doesn't make you attractive (at least not to the good kind of people and even then they are not there for you but your money), being a doormat doesn't make you attractive, it actually pushes people away, and when people want you there they make room for you.

Just read a lot about other people's experiences. You're gonna make a lot of mistakes, just make sure to learn from them.

man i'm really annoyed with what adobe did to fuse
what's the point of buying an IP that makes free products just to make it useless?

please make this the next thread's title

the thing is, we think we behave some way, but it actually translates completely different on other people's perspectives
like first user, thinking he was being nice while actually being a jerk to people,
one example I have is friend of mine who is pretty nice, but sometimes says something bad and starts overthinking on everyone else disliking what he said, when in reality no one gave two fucks about it
on the other hand some people have the talent to say just the right stuff to upset other people while thinking it was just a nice conversation

I have a question
on game design paradigm….

so here is the thing, currently developing in godot,
I intend to have a lot of data loaded from json plain text files on my game, so it takes them through a long spaghetti loading process and create scenes out of them for use later,
so the "scenes", in terms, are a custom Control node class with the pertaining data and behavior for it, good enough right, that's what classes are supposed to be after all, data + behavior

here is where things got complicated, I've recently came into a dev cycle where I needed to read from another kind data, and this data being something that could be used by different scenes,
so I needed a different object class just to store this data, I also needed a different node class just to be able to get this data and show it as visible in the game space,
the different "scene classes" would just need to put these two together and show it on game, but it got me wondering if the other stuff of these scene data should also be kept as their own object class instead
for example, there definitely are elements on these scenes, that would be common in structure, but used differently on different types of scenes,
the problem I'm having to decide on this, is because, then I would be separating data + behavior, where as, some objects would be just data a while scenes being just behavior
so it feels kinda pointless and redundant in some ways

another big point to this, is that having data objects classes with custom creators, I could move part of my spaghetti code of the loading process onto the class itself
which would indeed help a lot in visibility, so the code pertaining to creating a object type is only visible on that own object, cleaning it up a bit,
but this brings on yet another potential problem for spaghetti, part of my loading process, is some scenes taking data from other scenes,
like if they're referring some other scene I can just read data from it instead of having to manually write the data, if that makes sense
why would it be messy? because then on the object creation, now I'm not only passing its object data but the whole world with it, or I would need to make the checks beforehand which would make it pointless,

does any of this make sense? I just want to hear opinions
and I admit, its pretty hard talking about abstract programming

oh and I forgot reaction webm

It seems like some of your issue is related specifically to UE then, is there no way to stop an object from rendering without destroying it?

Oh, I was asking the question myself. I'm learning new shit every day, I'm honestly just trying to pin the solution to this inventory nightmare without having to set myself up for a headache in the future.

There's nothing wrong with having classes that are only data. In some languages these are structs but if you're using some hardline OO language then a class just for data is fine.

Yes, that would work fine, and is probably the easiest way if you want to keep the objects state intact and don't care about actual saving/loading (yet).

C++, BP is probably similar
Item->SetActorTickEnabled(false);Item->SetActorHiddenInGame(true);

and the opposite for equipping it

I literally did that as what I thought at the time a temporary and hamfisted solution. Is this really, really an acceptable way? Because if so, I'm a blessed man.

if it just werks, it's acceptable

I've done it like this myself because I didn't want my gun states (cooldown timers in particular) to be reset when switching weapons, which happened with spawning/despawning on equipping/unequipping.
Of course a solution will always depend on the situation, in my case, I had only a handful of weapons that I wanted to be able to equip and unequip. Just disabling and hiding them was probably more performant too, compared to dynamically spawning and destroying them all the time.

Though reading back your older post, sounds like for your situation it would be more appropriate to make more of a separation of your items data and their instances as someone recommended earlier. (This will make your life easier too if you're going to implement saving/loading).
I would create a generic struct for every type of item, in which you describe/store the info you need for it.
e.g. struct WeaponInfo, struct BookInfo, etc.
Maybe have them inherit from a base ItemInfo struct, so you could store them in a single array with dat polymorphism.

I can't really describe too much in detail as I haven't worked it out, but the important idea is to separate the essential data you need to re-create an item from an actor object itself.

The only other way I could think of is creating a custom C++ base class for your objects, and having an actor that contains a reference to such a class, do all the visual stuff based on some info you get from it, and manage your custom classes yourself while being able to destroy the actors that have references to them.
However, that would be quite a bit more work I think, when the only benefit to it would realistically b taking away the overhead memory usage inherent to the AActor class, which I kinda doubt would be significant enough to do all that.

So yeah, probably best to just do it the way suggested.

Thank you! I feel like I know what I need to do now.

thanks mang, I was honestly thinking no one would even reply
the funny thing with these blog posts, is that when you write about it, somehow the answer just comes to you

so writing even smaller classes for data keeping, makes things much better on readability
and even the class creation, I'll just pass the world to it, if it needs so,

Yeah there's the concept of "rubber duck debugging," stating the problem is the first step of its solution.

so anyway, sorted the things from this dev-cycle and pushed
here are some goodies of what I did

...

neat stuff man, what kind of game are you doing next?

I spent five hours modeling a mug - random background prop. It happens, user. Be proud. You care more than most do.

An action adventure game this time.

Anyone else here doing gamedev because there is nothing else to do? Other than dig through endless trash and hope you find a gem.

...

I can finally satisfy my bodyless nipple fetish, thanks user.

Using your time wisely is important.

what language is this and how are you taking up so much time? To find out if something is a verb for example, do you just compare it to a giant list of verbs in a linear search? Since, that is the only thing that I can imagine would be making your processing take so much time. Is your program multithreaded?

not gamedev related but does anyone knows of a setup to access another computer files that will be using both windows and linux?

sometimes the answer is outside of the 4th wall, for example, I've been playing ff XII and one of the things i didn't like about is the tedium once you make the perfect gambit setup gambits are up to 12 commands fo the kind "if X do Y" and most of the time you will automate the game into boredom, the problem can be resolved in two ways
tlr is made by the same team, and uses many of the design choices as a base, once you are familiar with the mindset required to play tlr you start using gambits as an aid instead of a "do nothing to win" option.
when i first heard that they added that option in the remake at first i didn't think much of it, now while playing i wish they could have gone the extra mile and made the speed changing automatic and added enemies that purposely fuck with it

Which is why I don't understand why Animal Crossing like model user applied modifiers. His workflow does not make sense. Even if he wanted to have extra precision for weight paint (which you dont get!), you would first paint with the subdivision and then apply it to refine the paint afterwards. Bad time management.

That looks very promising

I'm not exactly sure what you're asking, do you want something like Samba? By "access" do you mean over a network?

What's up with (you), posting the same thing twice with slight differences? You webm related or something?

basically i have various computers and want to keep my music/movies in only one

Nah I realized I was talking to the wrong person.

and all of the computers are on the same network

This is in c++.
I'm using the NLP section of a data science toolkit called MeTA. It uses machine learning and "a fast shift-reduce constituency parser using generalized averaged perceptron".
I don't know if the toolkit is multi-threaded, but my section of the code isn't. It really doesn't need to be as my code is making up a fraction of a fraction of the total processing time. Just as an experiment I commented out all of my code that wasn't directly involved in running the parser, it went from taking 10.7672 ms to process a sentence to taking 10.5497 ms.

The parser model I'm using is large and complex. It is 173 megabytes in size and takes 12 seconds to load into memory. After I'm done working on the dependency parser I may switch back to a smaller and faster but slightly less accurate model, which is a third of the size and conveniently three times as fast.

but seriously though, who cares about 10 milliseconds? A text based game doesn' t need to run at 120fps

A strange question, but would it be strange to have a 3D platformer with no voice acting whatsoever, including no Banjo-Kazooie-type voices? Most have some sort of grunts or noises, with the only exception I've seen being Crash Bandicoot, and even in those, Crash is the only character to not be voiced in some capacity.

I've been mulling over the things I'll have to do for the game, in a very broad sense, and so far, this is the one thing that I'm physically unable to do, and I don't really wanna pay anyone for voice acting, nor invest in actual sound equipment.

...

Plex works well for movies, not sure about music.

Probably. Very rarely does lack of VA make a game significantly worse, but it does mean if you want characters to interact, they need to express themselves in other ways.

use weird oboe samples or trombone noises

You could get creative and eliminate all dialogue. An npc can instead point in a direction and have a white balloon appear with the icon of what's in that direction. I'd still have some kind of noise at the start of the interaction, though, like what says.

It's a legitimate worry. I'm going to be building quite a bit of functionality on top of these processes that take 10 milliseconds, which means 10 milliseconds is my lower limit not really, simpler sentences take less time and more complex sentences take more. Milliseconds makes 10's of milliseconds, and 10's of milliseconds makes the game feel unresponsive or sluggish.
The one thing I want to avoid is input feeling sluggish. When someone inputs a command I want them to be able to almost instantly start reading the result of that command. I said it in an earlier thread, but under 10 ms is best case scenario, under 20 ms is my goal, and under 40 ms no matter what.

Ebin maymay.

That makes sense. I'll have to invest a lot in facial expressions and body language.

I'll have to try that as placeholders and see if it'll stick. Might be pretty funny, actually.

Move them onto one and install a file server or media server, it isn't hard

no but he does woah when he dies

Ahh, true enough. Didn't come to mind, for some reason. I was thinking of the jump and spin sounds, mainly.

I mean everything is stock audio

thanks, it works fine, it doesn't support custom tags for music but i can just slap them on gender

Yeah but I think that short of a duration, on text input really doesn' t matter that much and what little delay there may be can be quite easily masked by animation and/or sound probably if you really feel that it's bothersome.

You only parse a whole sentence right? (not real-time while typing)
10 ms in that style of gameplay, I really can't imagine it being very noticeable.

(me)
actually okay, if your game is purely text (e.g. in a console) then sure you're probably limited with sound/animation (and perhaps so by design as well), but I still think the delay wouldn't be a big deal.

In any case, "premature optimization is the root of all evil", just make sure to focus on the game first.
Interesting mechanic by the way, I haven't played much text-based games, is it common to parse the grammar of a players input?
I do remember being surprised by well some of those games react to (many different possibilities of) a players input.

I'm not optimizing right now, just keeping optimization in mind. I appreciate your concern for me getting things done.
It's almost purely text. I know a guy who does music and once I get far enough along I'll bring him on to do sound. I may add a GUI for quality of life improvements, but that's still a "maybe" at this point and if I do add one it won't have animations or art.

The history of grammar parsers in text-based games is actually something I find pretty interesting. Modern text-based games use engines like Inform 7 that almost always parser grammar and do so very well, better than I'll be able to manage.
Older games relied on a "verb noun" structure of input like "get flask", "go north", "attack goblin" and so on.
Zork was unique when it came out because it did a pretty good job at parsing grammar and was leagues above everything else, even though by modern standards it's so bad it can't even be measured.
My goal is to add at least as much functionality to the parser as what you get in Zork, though I'll be handling combat and talking with people differently.

No, it's actually one of my weaknesses (subdividing geometry, etc). But other than arrays, I can't think of a single object that would get that large.

I mean, Lists and stuff would count as arrays, because internally they use an array that resizes as needed

Yo that is really cool

I have very, very little done on the actual game but I'm working on really good modding support. I'm going to add most of my game content through mods that the user can see and edit. It's also nice to be able to work on either the engine, or content, but rarely both.

hmm

If I hire a cunt who ends up stuffing political agenda into my game without my knowledge, can I sue? I was just thinking about the future and having to work with other people, and I'm kinda concerned with the idea of someone sabotaging my shit with their liberal/pol garbage.

oh boy

Ask a lawyer. You shouldn't be asking here for legal advice.

You can always sue, doesn't mean you can win.

literally just search through all of the content to make sure that there are no references to anything you don't like.

Unreal RPGdev,
wrote a bumch of code but waiting for people to get back to me if it's any good. Also just thought of a way to make legit sprite sheets: make a font texture out of my sprites and call them as if they are charcters between 0-9. I'd need 24 fonts though, or just 1 big one and use ASCII letters as well. That'd be more difficult though.

Goddamn you're a faggot

Are you unable to use a texture atlas in Unreal 1?

I worry a lot about the future. I'm just really trying to understand what tools are available to me to prevent SJWs from infiltrating my craft.

Fair enough.

Why so aggressive?

I remember your UI stuff. Where's your wiki page tho?

I do not currently feel like what I have on hand is worth making a wiki page for. I won't make it until there's a tangible demo that proves the concept and has substance.

Animated particles.

can you make a tail of particles?

The canadian who made the relatively popular indie game 'Celeste' just released the source code for the Player class.
github.com/NoelFB/Celeste/blob/master/Source/Player.cs
why does this seem familiar

There is nothing wrong with that. I also thought about this, although I haven't yet sit down to actually create a specific plan. Even though these are all hypothetical future problems of my hypothetical future self that actually made a game and succeeded, there's nothing wrong with spending some time thinking about such stuff.
In any case, the other user was right when he told you to not rely on an imageboard for legal advice. Even if some honest to god lawyer is trying to help you here, you don't know if he lives in the same place as you do, so his advice might not even apply.
I won't give you legal advice, but share some general thoughts I had on this. Hopefully it's useful.

While there might be places where you can hire & fire as you please, that can't be done everywhere and, even though I don't know much about the laws in detail, it wouldn't surprise me if there were legal hoops you'd have to jump through first (and I'd rather not). Worker's rights were specifically made to protect honest employees from exploitative employers. If you're an honest employee working for an honest employer, you don't need those rights to protect you. The problem is that the tools that an asshole of a boss would use to exploit you, are the same tools that an honest boss would use to defend his company and himself from an asshole of an employee. Simply put, there is no way you can write a contract that covers every single way in which an employee who seeks to do damage to your company can fuck with you (assuming said employee is sophisticated about it and doesn't just start smashing company property). And if you add catch-all clauses, they can be used to drag you to court over, which is a fantastic thing for such a person. They don't have to sue to win. If all they want to do is to damage you as much as possible, dragging your ass in front of court is a fantastic way to go about this, because regardless of the outcome, you'll loose a lot of time and will be under a good deal of stress.
The only thing you can do to protect yourself is to ensure such people never get hired to begin with. Luckily, with SJWs it should be straightforward for the most part. Simply asking for the social media accounts when they apply should do for the most part. There is a reason why virtue signalling is a thing: everyone has to see that they are a "good" person. They might lie about this by either giving an alt account or by claiming that they're not on social media though.
Since I want to have a small company where I can stay relatively close to my customers, I won't have a dedicated PR department. But I don't want my employees to do every social interaction through me, because I will have more important things to do with my time. Basically, I need to be able to trust each and every single employee to know how to behave on the Steam forums/Twitter/etc. To ensure that, I thought that a small recruitment test would be nice. Basically ask your applicants to come to your office to do a small test (wouldn't want them to cheat, after all). The test itself was something I came up with in a different context. I wasn't thinking about SJWs, but of PR people on social media and moderators on forums. If you've spent any time on the Internet, you know that there is a tendency for them to go on power trips. I don't want any of that shit, so I thought how I could weed them out. The method is easily adapted to SJWs though.
What I had in mind was something like this: The test could consist of giving the applicant a printed version of your company policy or the discussion rules of your Steam product board, along with a variety of "difficult situations" and ask him how he would resolve them with extra space for explaining himself. Give him 30-60 minutes and then collect it to see how he did.
Basically, you'll want to craft the situations so that the test will expose hypocrisy. You know how a regressive would "resolve" a situation that involves someone arguing against Sarkeesian or for GG on your forums. You can also imagine how a closet-SJW would try to suck up to you, so you can play on the boogeyman that exists in their head. Don't forget how delusional those people are. If you criticize Sarkeesian, you hate women. They actually believe this shit.
After that, there's trial work. Keep new employees basically as well-paid interns for a couple of weeks, and keep a close eye on them. (You should do this anyway)

Jesus christ, also what's the point of releasing only the player code?

Thanks, user. I'm sinking my soul into my game right now and it makes me shudder when I imagine an infestation of this sort occuring. Nightmare-tier.

Need some help, anons. My character's friction don't work right. Something's up with the way I calculate it - my character slides down ramps when standing on them, and I really don't like that. Any idea how I could implement some kind of budging velocity, maybe? I don't know, I'm a fucking caveman. Help would be much appreciated if you feel like it.

hastebin.com/midocayahi.php

Please no bully

OH NO

I swear to god that clamp is there instead of the ifs I just forgot to remove the ifs aaaaaaa

"Education" or something. The only thing that people learn from this is that it's possible for shitty programmers to be successful.

Readme files are useful sometimes

From what I've read, you cannot publish a game using Unreal Engine 1 code, due to licenses and assets used.

This looks to be the closest I can get to one.

Fuck. My. Life.
I just spent a couple of days on learning Python and Blender's scripting API to do this . Meanwhile there already is a plugin that does that. It's called "Magic UV" and it does everything I need and more.
I mean, I should be happy, but I still rather not waste my time like that.

No problem.

if no player input if |velocity| < threshold velocity = (0,0,0)

Wow, it would seem as though I'm retarded.

that's basically half the engine, user

you work fast user

is this already at limit?

Page 13.

new thread