/agdg/ amateur game dev general

Assuming Unity here.
The former will essentially freeze them, so that they have to be moved manually, while the latter makes it so they don't collider, but overlap (which can be detected in code).

I don't want to use physics because I need only minimal interaction between objects, stop the object from moving like hitting a wall. In 2D. I don't want extra complication or use of force to move them.

How would one use triggers? I've just seen them being used for players entering a certain location. How to use it actively? Pls halp

But yes, kinematic rigidbodies are probably the solution here.

you don't have to if you do what said and make them kinematic.

what are the disadvantages of using godot? i've just started working with it and it seems pretty straightforward, so i want to know what to look out for in the future

It's 3D systems aren't quite as sophisticated as the big two engines, but they're being worked on all the time. Or at least that was the case the last time I checked out Godot. Not to say it's not usable for 3D, it is/was just better for 2D games.

well, i'm doing 2d so i guess that's a plus

alright I'm gonna try to stream some game development again this time in a higher resolution and lower framerate.
twitch.tv/rebellionrtsdev
once again I can't actually watch my own stream so tell me if somethings off.

You're focusing on dick, but the tits are at least as important.

Has anyone messed with Xenko yet? Especially Unity devs. Uses C# but is open-source and developed by Silicon Studio, who have done middleware and games for at least a decade. I intend to try it out sometime but haven't gotten around to it.

multi threading is impossible unless you know c

...

Multithreading in Godot he means, nothing to hmph about.

pixel art, or non pixel art?
If I went with pixel art, I would lower the framerate of all the animations I already have done, and be able to animate more in much less time.
If I used non pixel art, I could add a bunch of post processing stuff and vfx without it looking disgusting and clashing with the artstyle.

There are advantages to either, but I'd like to hear your opinions.

i'm really considering whether or not I should make this game lewd.

It just looks like you've put a filter on it or something. Making appealing or at least readable pixel art will necessitate more work than that: look how her eyes vanish or completely change shapes at best in half the frames.

Downscaling a high res animation looks like utter dogshit. See your webm. Either don't scale down or start fresh at a lower scale

What type of game is this? Because if it is a fighting game my answer will change drastically.
The two questions would revolve around what is your definition of lewd and would it work inside the context of your game?

One VBO+IBO for all model data is fine. You don't need BaseVertex, you can just offset your indices manually (when initially filling the IBO) and draw a subset in the draw call. If you use uints as indices you're basically guaranteed to have enough index space.
It depends, do you need different vertex formats for different models? I assume not.

why her legs are so long?

Why would you lower the framerate too? One of the only benefits of downscaling or using render as a base for pixelart is that you can get high framerate animations with a lot less work.

I've done render->pixelart before but you need to design the models and proportions around it and you also need to manually clean up the results or you just get a bunch of indistinct shit.

In some physics engines, kinematics, collision detection and collision response are separate thing that can be individually disabled. Perhaps you can do that in Unity. If so you could disable response and handle all collisions manually (by just setting velocity to zero).

Oh. I only looked a few quotes back. Missed the first one.
Mistakes were made.

i accidentally fucked my last post. alright.

That's fair. I honestly didn't realize how noticeable it was until it was pointed out, thanks. I've sat on these animations for a while so my opinion is pretty distorted.

generic actionrpg sidescroller. I know how overdone these games are, but it's really just for fun and to learn. Nothing too serious.
I don't really know what my definition is lewd is. I was just considering it because I have the means to, but no real reason to.

Not sure. I could fuck with the proportions I guess if it's too strange looking.

Mostly because it's less work for any future animations, and that most pixel art I see is done at a much lower fps than 24.
It would be one of the biggest benefits, since a single animation can take me about a day to make, half of which is cleaning up curves between each key pose, If I lowered the FPS I could possibly completely eliminate the need to spline anything.

Eh? You cant use font freely?

For some reason the offset in glDrawElements doesn't seem to work. I basically took code that worked in OpenGL 3.0, modified it to work in ES and changed draw gldrawelementsbasevertex to gldrawelements and used the offset value while casting to (void*).

wrote some code for allocating and free'ing Vertex buffer objects

i tried.

Why are you using windows xp to make music?

I didn't know Windows XP was a tracker.

probably the best "my first tracked song" I've heard. it's a little dull, but effective.

Man, I love these chords.

Go high-res, it's less work than pixel art in some senses and it looks really good already.
Though her stance (mostly foot positions) looks a bit weird, looks like you tried to line her feet up on the ground but the rear foot should be closer to the camera for it to be more natural and balanced.

Also, you have 10 seconds to explain why she's using a katana instead of superior European blades.

I really like that discordant ethereal sound. It does a super good job of setting atmosphere.

The normal animation looks the best.
Pixel art is a lot more than just downscaled images, every frame has to be designed with the animation in mind, otherwise pixels just jump in and out of nowhere. Like in, well, your webm.

They remind me of white noiz

duh, most fonts have licencing fees. Most anons on /agdg/are quite copyright concious so we tend to do most shit we have to from scratch.

Exocet font family costs 100 usd for a single licence as well… and I don't see any option to use it in vidya so most likely its owner would want a much higher fee for such use.

myfonts.com/fonts/emigre/exocet-ot/

note that in the blog post I even mentiones specifically using an old ass paint shop pro version to be extra safe.

Thanks, I'll experiment with it to see whether I can add flames and stuff to make it look more like D1

That's odd. I just tested and it seems to work fine for me. I set indices to 300, subtracted 300 from count and roughly 100 triangles at the start of the buffer did not get drawn. Sanity check, you do realize that just like with count, the value should be the number of indices to skip, so for GL_TRIANGLES it's triangle_offset * 3. Also, it's a count offset, not a byte offset. So it should not be offset * sizeof(GLuint) either. What kind of platform are you doing it on? Can you do a minimal test, like allocating a buffer and experiment drawing with different values for indices?

First I've heard of it. Sounds interesting, my only concern is if they'll abandon it if ti doesn't attract enough of a community. Havok had a pretty interesting free engine for a few years that they actually sank money into promoting and they still dropped it after a couple years.

although they are working on implementing C#, I heard the gdscript is not really favored much due to it's lack of speed for reasons I don't remember. Also lacks some proper documentation and tutorials in general but not as bad as it was about a year ago. So far I made a pong clone and working on some Tetris clone for practice.

Its 2D systems are better than Unity. But yeah its 3D is unfinished. There are noticeable broken/missing things in the 3D classes but things still mostly work. I seem to recall two different rigid-body objects that both use unique, non-primitive engine-generated collision meshes will not collide. Don't know if that's been fixed or not but that was the biggest problem I can across. You can mostly work around the "holes" in the engine. For example, the animation transition nodes don't take into account the current animation frame when transitioning to another animation, but you can emulate this functionality with blend nodes and a bit of script. In general it's not ideal but it is usable for 3D games despite its unfinished state. Apparently they are breaking backwards compatibility for 3.0 (whenever the hell that comes out) so lots of old problems are being fixed. But on the same coin that means updating 2.x projects might be a pain in the ass.

is it component based like unity or more object oriented like unreal?

If I make one how much will I make

kill me
and it wouldn't hurt if i could actually run the engine, i'm stuck here working with visual studio only and praying that shit ends up working later. wouldn't have taken the course if i knew it was with UE4

I'm doing WebGL and I basically spent all last night swapping out indice values with the draw.

One thing I thought it might have been was I was adding the amount of vertices in the buffer at the start of buffering, to the indices when I added new indices. So it would pull vertices from where they're actually stored not the beginning. Now that it's not 6 in the morning anymore and I have a clearer head. Maybe I should have divided by 3 since I was probably adding how many floats instead of vertices.

Good, it is best to do it sooner rather than later. The more projects built with the thing you want to make breaking changes to, the worse off it is for your adopters.

Looking over their docs, they have a "Xenko for Unity Developers" section which is nice. doc.xenko.com/latest/manual/xenko-for-unity-developers/index.html

alright guys I'm gonna stream game development again for a bit twitch.tv/rebellionrtsdev

no linux support >:(

Also, now that I'm looking over the docs, it seems Xenko is component based,
doc.xenko.com/latest/manual/xenko-for-unity-developers/index.html#entity-components

Hey guys. I am sorry if this is considered asking for spoonfeeding but I can't find a guide anywhere else no matter how hard I tried.

I am making a game in Python, and I wanted to make a system where you could create your own character, but in a visual sense. Like, having a menu with different sections for the hair, eyes, skin type, body type, etc. With arrows on the side that you could scroll through. Something like pics related, with a paper-doll kinda feel. Is there anyone who can help me or might be able to point me toward a guide or something?

Yep. I forgot to divide by 3. It's always the dumbest shit.

One thing that always stresses me out about packing tons of shit into one buffer is when you have to remove shit, which will probably happen a lot in my game. Basically it seems like you'd have to delete the object, then move everything behind it over, which can be a huge operation.

Seems my internet is to unstable to stream right now

Hey anons I have a question. I want to start moding a game as a way to start learning how to model, and code. What programs do you recommend. Are there any things I need to be aware of?

Fits right into demonic theme.

How are you using it on Windows 10? Do you have some drivers I don't know of?

At least it supports half A presses. I'll wait until they fix it, looks promising.

That's a pretty comfy setup, user.

It is, isn't it?

If you by remove mean stream vertex data into and out of graphics memory, that is a fiendishly difficult problem that you should avoid tackling as much as possible. Trust me, I've tried. If you can, just upload all the data you'll ever need and leave it at that.

What kind of question is this? I use UE4 and I'm constantly making components for behaviour shared between multiple actors. I think you may be using Unreal the wrong way.

Sounds to me like you'd be better of learning how to program (better) instead of reading some guide, that is really not something that should be challenging by the time you have the skills required to make a game.

I wouldn't even know where to start with the logic. Everyone needs help sometimes user, you can't teach yourself to do everything. I've got a good 60 percent of the game done, but I can't apply anything else I've learned to this situation. You can have the skills to make a game but still be in the dark about certain things. Thanks for the super cool input though.

I'm familiar with both and user is onto something.
In Unreal you have actors and actor components, whereas you have GameObjects and components in Unity. These are not similar at all. The fundamental difference is that actors have functionality, whereas GameObjects do not.
In Unity your GameObjects only have a transform and can be turned either on or off. Aside from that they are merely containers for various components. You could say that the behavior of every "entity" in a Unity game is realized by the sum of its components. Everything is a component.
In Unreal on the other hand, the behavior of your entities is a part of the actors themselves. Actor components merely serve the purpose of letting you reuse certain functionality like rendering a mesh or playing a sound. However, how and when these meshes and sounds are used is defined by the actor, which is some class that inherits from AActor in some fashion.

It sounds similar enough, but it really isn't. I can say that as someone who started porting his shit from Unity to Unreal.

I also made a pong clone and posted the webm in the last thread. I'm probably going to keep making little games until the release of 3.0. I don't know what I'm doing right now but this far i have movement with a little sliding

the drum loop sounds wrong looping at e-4, anyone know why?

Your mistake is in thinking AActor is the base class in Unreal, when it's a child of UObject. You could technically work exactly the same way in Unreal as in Unity, it's just a lot more of a hassle and for no good reason in almost all cases.

...

that's because it's changing the speed of how it's played based on pitch. You might be better off using the tempo lock on the instrument for the drum loop.

I know this, that's why i changed the tempo of the song from 101 to 127.

My desktop has suffered a severe autism attack, I fear it may not live many more days.

what happened? my old build's psu literally exploded, and refused to boot with the new one, a few months pass and it turns on as if nothing had happened, that was 2 years ago and is still kicking to this day

The motherboard is kind of fucked, and I have to jiggle the shit out of the ram slots when I start up to actually boot.

Reported.

Reminds me of my PC.

I think it might be linux. Something is really fucked up with Ubuntu or Mint, where it will just freeze on writing.

I have those same floor tiles.

Newbie curiosity:
Does the _t in vbo_t and gfx_t stand for type? Does that mean this is your own created data type using a struct?

I think is the other way around, windows likes to pretend that everything is fine, when an old drive failed it was ubuntu that keep warning me about write errors, faulty voltages and heat; while windows only did it about a week before.
I'm still mad at that incident because i lost of files that were corrupted before my new hard drive arrived and could back up them. I'm not mad at windows or anyone for that, it would have happened anyway, but the damage could have been a few magnitudes worse with windows only.

while I'm at it, there was a sound file that i lost that was called mystery-song.mp3, I'm not a sound guy but it was a very unique piece of music, if i had to describe it had class, it was a very nostalgic song in part due to the recording quality and part for the song itself. it had no vocals and was probably played with real instruments, the main instrument was a piano and had some corded instruments in the background. originally some user from 420chan found it in his computer or phone and could not remember where it was from, the file went from there to 4chan and reddit and no one could tell either. from what i remember it was 2010 and a few thousand people were trying to search for it at one point. if somebody has info on it I would really appreciate it.

I took a break for a day because my eyes felt twitchy
now I can't get myself back to work, everything feels hard and exhausting

what do I do?
I want to make it this time instead of repeating the cycle

I don't think it's the drive going bad. It's a case of if you use too much ram eventually out of nowhere linux will start writing shit like crazy and lock up the whole computer until you kill the game. I have a CPU monitor on my taskbar, which has dark blue as writing, and the whole box will just be dark blue with a few specs of the black background at the top.

I'm not the only one I know who has this issue either. It might have to do with the partition boundary not being set correctly.

I guess I should clarify a bit more. On my shit computer, it will slow to a crawl when this happens and sometimes I need to kill the X server. On my good computer it will just die and not want to reboot, even if I change hard drives.

progress

Holla Forums wouldn't accept my webm. "Longer than 7200 seconds" which is total bullshit

I don't know if this makes me sound like trash but maybe you can try doing something like a pomodoro setup until you feel you're in the groove again? Work for like 25 minutes, break for 5, repeat 3 times, then give yourself a long break.

why is her hair made of caramell now? i liked her more with the nutella hair

I asked this a few threads ago, but anyways

All my containers in the game will have a maximum volume, maximum weight, and list of items it contains. I also want to give them a length/width/height so they can calculate the volume.

When allocating shelf space, it's easy enough to figure out how much product a shelf can store, because it's basically rectangular. However, for top shelves and skids, they have no top, so there isn't a reasonable "top" to determine volume, and I'm not sure of a sensible way to resolve this… otherwise I'll get skids that are like 20 ft tall and a massive heap of items out of reach on the upper shelves

Just playing with modifying materials via script. By default the hair is still gnutella.

nah you sound ok to me
I'll try that, thanks user

Just give them a top. There is only so high you can stack stuff.

Right, but what should the "default" be? I agree that you can only stack stuff so tall, but deciding the limit is the hard part.

Like, our vacuum aisle has the bottom shelf, and a second shelf, the top, around as high as your collar bone. Vacuums are on display, and are like 3 ft tall.

Another case is skids, which are 40"x48". Going with twice the larger dimension here, would be 96" or 8 ft. That seems a little too high, while 6'8" seems a little random, for the smaller dimension (40").

Apply that same logic to the shelves, the average shelf is like 24x18 or so, that'd give the top one a maximum clearance of 3'

I'd opt to just make it modular. A space has 6 available feet. If I put 1 shelf on that aisle, I got 6 ft of clearance. 2 shelves, 3 ft. 3 shelves, 2 ft. Skids sound like they are an unique display, and can be given whatever clearance you want.

Hm.. I guess I could do it that way, since the racking usually extends up a little bit past the shelf brackets anyways, so that might not be a bad thing.

Delivered skids full of products are probably going to be a set height, but I still need a sane upper limit, if they get product piled onto them. I'll figure it out

I took a gander.
Most interesting parts is the Vulkan support (in Unity too, but this is a good sign), HLSL shader extension, and it being similar to Unity; while being open sourced.
Although certain parts needed to make a game don't look quite done… like the animation system; i.e. no auto-rigging equivalent, no skeleton visualization, etc.
In addition to lacking cross-platform compute shader support is a deal-breaker for me; as "windows only" features are a big turn off.
Also sorta sucks that I'd have to port a lot of the Unity centric shit I'm using to its engine API, and I'd rather not have to do that.

All in all, it's probably best to remain with Unity if you want a large team updating the engine consistently, don't want to worry about it being abandoned, and want an already finished engine; the only pragmatic reasons to switch would be one driven by principle, or by extension a need to access the source code.

You've never used pre-2008 engine, have you?

I took a shot at an over under shotgun. I still have to fix the normals but the base is pretty much done.

those are editor centric tools to make certain steps easier, and to make the model/rig to animation stages faster; as otherwise you waste a lot of time, or need to use your modeling program's equivalent (I'd rather animate in-engine).

It will not break the game you dolt, wtf are u talking about; it's editor centric, and any breakage is your own fault.

You have no idea what you're talking about, do you?

You dense motherfucker I was quoting you with hyperbole:
I was pointing out in my post that older engines didn't feature either of the things mentioned in your post. If you wanted to import models back in the day you had to make the animations out of the engine and there wasn't really anything like auto-rig.

Did you model this in 3D first? Why don't you just make a 3D game? This isn't the old days like during Starcraft 1, Diablo 1 and Final Fantasy VII where you make prerendered everything. Proper pixel art will trash your game, which is a shame because I quite like your 3D model, long legs and all.

You can try out Ren'Py, the visual novel engine that's built on top of python. Making switchable image maps are a cinch, there are even a few visual novels made with dynamic CGs (with layers switchable depending on what the player selected as the main characters). You can apply your python functions for the other stuff.

I can barely program beyond editing values but managed to shit out a very basic test of a VN when I was like 16. If you're just doing something minimal than Ren-Py is great but it has more than enough potential if you can actually program.

Thank you very much user. I have a basic idea of how to do switchable maps, my big roadblock is finding a way to keep the image as the player selected it through the entire story, since they will be looking at themselves a few times. I would imagine the logic threads would stem from something like what it takes to specify and keep a selected gender but obviously it's a bit harder than "if BLUEHAIR show BLUEHAIR.jpg." It will feel weird to use a visual novel editor to do what I want, but if you say it will work I will trust you.

Just store the assets that make up the character in a list or dictionary or whatever sort of data structure you want. Even a data object would work well here, character.hair = "bluehair.jpg"; character.shirt = "red-shirt.jpg", and then have a character.show or whatever that returns the charactere rendered complete according to the settings

I feel awfully stupid now. I can't believe I didn't think of putting them in a dictionary. I guess I always thought that was just for values, I didn't think to apply it to something like an outfit catalogue. You're a real human bean. I'll work now, until I decide to grossly over complicate something else.

does anyone know about an image format which has better compression than WEBP? I would use libFLIF however this is not an option since I cant build their source code and they don't distribute binaries.

Are there any formats which are better than WEBP that are not FLIF?

I made a test VN in Renpy, where can i upload it so you faggots can bully me about it?

I like to use uploadfiles.io/ and for a test VN, 5 gigs should be more than enough.

A year ago or so I tried to ilstall it on Win10, I think I got some other company's drivers to semi-work, as in: the computer "detected" a tablet, the clicking functionality worked
… BUT …
The cursor was jittery as hell. Nothing could be drawn with the thing.

I actually thought the tablet was broken. But when I plugged it innto a WinXP machine… it worked okay. So I gave up and as the blog post says

The top screen is in fact the XP laptop I mainly use for music playback (SRS audio doesn't work on Win10 either)

Thanks, my man. It's only a fragment of the whole thing, but this one is relevant to the actual work I did on the font (the beer helped a lot).

trying to stream again this time on youtube since it seems to work a lot better with slower internet youtube.com/watch?v=lrZPhZhx8Yk

my computer bluescreened and apparently when you restart your stream the url is different youtube.com/watch?v=TXvr2aG55UY&feature=iv&src_vid=lrZPhZhx8Yk&annotation_id=channel:58e27640-0000-23d7-bdf3-94eb2c05ad42

Calculate it to stretch to the ceiling. That way you could have a ceiling height property on buildings (or have buildings consist of a set of rectangles with individual height properties), allowing players to go full IKEA with shelfs that are several meters in height, requiring a forklift to remove items from. As for dumb shit like placing vacuums on the top shelf just because you have space, handle that with a combination of two things: Shelves have a weight limit, one hard limit for instant collapse, one soft where the chance of a shelf randomly collapsing when customers interact with it, the chance increasing as you approach the hard cap. Additionally, if you place heavy items anywhere but on floor level, there is a chance customers won't pick the item based on their body strength (ofc if a couple is visiting the store, no check is made, since the woman will tell the man to take the item and it's not an option for the man to say no). Last, if a customer overladens themselves, there's a chance they'll drop the item over themself, possibly bringing the entire shelf with them, depending on customer body mass and shelf structural integrity. So if you put 10l containers of peanut butter on the top shelf, some fatty will make a huge mess trying to grab one and hold your store liable for millions in damages.

Did someone say full IKEA?

I haven't made myself clear there. I didn't mean to say that AActor is the base class in Unreal. What I meant to say was that AActor is the closest equivalent of a GameObject in Unity. It's in fact one of my main points of criticisms of the Unity API when compared to Unreal's. You don't have a UObject equivalent in Unity and, by extension, classes like "GameInstance", "GameMode" and "PlayerState". In Unity every piece of code is somehow "attached" to a level by means of a GameObject, even if there is no reason for it from a functional perspective. If you want to have an object not be a "part of the level", you still have to use some sort of management object that is in order to instantiate it.

my mistake, I misinterpreted your post

I changed the drum loop to play at c-4, back from e-4, and removed a pattern.

sometimes when I'm lying in bed half awake trying to fall asleep. I hear music, music that I've never heard before and it sounds good. How hard would it be to make this music?

Are you me?

Hello me, sometimes i also hear music and it sounds as if i was hearing Bethoven reincarnated and delivering a concerto in my room, then i proceed to think that i am probably a genius composer but lack any skill to recreate said music, in reality i think it's just sleep paralysis setting in

Admittedly, my own inspiration for making music comes from coming up with music in my head, and wanting to make stuff that I would listen to myself. I've had some formal musical education I guess, though I don't really use it while working on this stuff, and it isn't too hard, but I'm still not a fan of what I create(because it isn't as good as I want it to be), and ultimately that's the most disappointing part.

I don't see what sleep paralysis has to do with it really though what causes it doesn't matter as long as you can remember and recreate it.
The reason I mentioned is because recently it became electronic music and I thought I might be able to recreate that whereas before a lot of it involved orchestras and singing etc.

ok here it is, bully shitty code.
ufile.io/5e7f9

I've decided to call my character " int dir[5000](); " and broke your game

...

i named my character [] and got the same thing

Why are you using wierd names?

Think their point is he should sanitise user input or else people can fuck around and break things.

How can you do that?

You look for weird shit and escape, remove or otherwise disallow it.
It's not absolutely critical in single player/single user environments, but if your programs talk with other clients or your own server, you've got bad times a-comin'.
Are you familiar with the concept of escaping characters?

lol

I had implemented trigger functions on my objects, where I put ontriggerstay as translate=vector(-speed_it's_currently_moving) thinking it would repel player from going into the box. Now it is behaving like this. I tried ontriggerenter but it didn't do anything. What would be the correct method?

Is there any reason why you need the collision to be trigger based?

You usually have a set of characters that are acceptable; for example a name might consider only of [a-z] and [A-Z], no spaces or underscores. Or other input might also allow for numbers. What you're doing is just accepting anything, and leading to broken code.

Like others have said, no it doesn't really matter in a single player context since it's basically someone breaking their own toys, but it can reflect poorly on you if you don't fix it, since you will inevitably come into problems with it later.

It's up to you whether you trim whitespace, remove characters, or just shit out an error message, but you should at least acknowledge somewhere, somehow, that the user is an idiot

Physics based collisions were too realistics, I just needed simple stop movement or rebound back if attacked and less taxing. More importantly detecting collisions when jumping, I plan to make collision a function with order in layer, so that when I jump, it checks within a certain range of yaxis and doesn't collide if it is outside that, such as an object far behind it.

No

something like this?
python: MC = renpy.input("What is your name?", default = 'Mario', exclude = '{1234567890!@#$%^&*()_+-={}[]|\:";'',./?}') MC = MC.strip() if not MC: MC = "user"

Anyone knows how to stop the line [MC "Looks like I'm the first one"] from appearing if the [jump hello] is triggered?

menu: "Say hello": jump hello "Ignore her and go to class": jump classeslabel hello:scene benchwith fadeshow juana: xalign 0.5MC "Good morning"J "Oh, Good Morning"MC "I guess we are in the same class, Calculus right? You know where it is?"J "Ah... Yeah, lets go togheter there"jump togheterlabel togheter: scene classroom show juana: xalign 0.5 J "Here it is Calculus classroom" MC "Thanks for the help, ehh..." J "OH!, I havent introduced myself" J "My name is [J], and you are?" MC "[MC]" "Both shake hands"label classes: scene classroom with fade MC "Looks like I'm the first one"

You don't have a jump between 'label togheter' and 'label classes', so I presume they're just running sequentially.

user please it's "together"

Escaping is what you did with the single quote in the exclude string, to make sure it's read as data rather than some structuring element of said data.
If you don't escape characters, the string ends at the second single quote and then everything else beyond that point is just read as instructions. The same thing can happen with arbitrary input.
I found something on renpy about how it handles text. Maybe you can find something of value there.
renpy.org/doc/html/text.html

That might work. Probably not a complete solution, but it should work. If you want to be on the paranoid side, you'd either escape fucking everything or use a whitelist approach, tossing out everything that isn't safe.
You just want to make sure your strings can't be interpreted as anything other than, well, strings.

If that behaves as I would assume it'd behave, then labels don't actually isolate anything. Once it reaches the end of "togheter", it's just going to keep on trucking.
So the "jump togheter" is redundant. You might want to jump somewhere else at the end of the "togheter" section.
Also, is that supposed to be 'together'?

more like get yourself together, user

forgot pic

According to the docs
So, I think by giving it the negative value of the current speed, you're telling it go in the opposite of it its current direction on collision.

I'm still real new to Unity so I not sure exactly how to do what you want, but I'm not sure if translate is the best option. I would think that you would want to set the current position with getting the current position on collision.

OnTriggerStay sounds like it only triggers on the second frame of the overlap with the trigger, if you did it on OnTriggerEnter it should move you back out. That said, you were already told to just use kinematic rigidbodies, not triggers. Did you even try that?

Epic Games just released this for Unreal Engine

docs.unrealengine.com/latest/INT/Resources/Showcases/PhotorealisticCharacter/

It's worth the read even if you don't use UE4, as it goes into some of the nitty gritty of what goes on behind the scenes.

working on my VBO system. right now vulkan can draw VBO's but openGL needs some work. This code is pretty unfinished.

Probably limitless amounts of money if you do it right, provide the modding tools and decent defaults that are memorable, and make it work on even the most decrepit of toasters. But you really do have to put your heart and soul into crafting those penises, with additional foreskin animations that sync with hand movement penetration etc with the option of people animating the foreskin themselves, you see, it's basically the difference between those nice sfm animations of a shemale zelda with a decently rigged penis with foreskin and those ones that are blatantly using the fucking model found in Oblivion mods, that are NOT slofs work or anything.

Then, there's the differentiation between normal breasts and "Filthy backalley street whore entry trap level" balon Implant tits, with optional Surgical scar textures when you really want an actor in the game that looks like a Rock Bottom loss of Maculinity sense of self and being whore, you know?

Then there's making those dicks that just…look…Chewy, like a fucking Windgum haribo treat but it's a glistening Po-ju esque penis.

Also, Vulva puffiness is a must for hermaphrodites, No ball futa almost never looks good unless you model it in the same sense as the Bible Black artist would, because it actualy looks and feels like a sexual organ opposed to silly ammounts of hyper penis cum dumb vein dicks that you'd see at someone's cumbath yiff party, you've got to make it look like when that dick penetrates and you look at it from the bottom angle as if you were filming the porno that it doesn't look or feel like a Strapon without the whole Uber grinddyke ultra horny sweatgoblin mode sex when you get to the Non-soft fuck your true love yuri, no, it's got to look like the work of a fully functional Greek Aphrodite sex-goddess worshipped for her superior form of being a girl with a dick with no balls but there are balls and that's why the vulva is puffy having just the right type of sex contact noises from thrusting, hitting the cheeks, varied lubrication effects for friction based noises and much-much more.

You must be the sexual archetect of a Generation, the workings of a porn game are basically you, the Renissance, and a need to waste some seed.

...

so back to work on modeling rigging and animating after taking a unexpected break over the site being down and all, still cant understand materials or how blender handles them but i guess i should just work on how unity handles them but i'm stumped on that too. this mixed with me being absolutely revolted by my own textures. i guess ill just skip it for now and work on a character controller.

...

I'm trying to figure out why my song is technically wrong.

It depends, are you adept at any instruments? Do you have any knowledge of music theory? Both of these are fairly important. If you don't have theory knowledge it will be especially hard since it will involve a lot of trial and error to get what you want, and you may lose or forget what you had in your head in the first place. Sometimes this works out and it ends up being even better than what you had planned, but most of the time you just get frustrated because you just wrote the same phrase you had in a song 3 weeks ago but it's in Eb this time instead of C. This is coming from someone who put in a few years as a trial and error musician before finally learning real theory. The difference is huge. Three years ago I tried to write an "epic" (that is, a song with a 3-act structure with different instruments representing "characters") and I slaved over that shit for at least a year and the end result still leaves a lot to be desired. A month ago I tried to write something similar and I was able to write the whole thing in one sitting (5 hours long, but one sitting) and it was much better than my other piece.

Regarding escapes: Some special characters (notably quotation marks and backslash, \) affect the string parser in pretty much every language. So say you wanted to to do something like

The quotes in the message would fuck it up. So you'd use an escape character (usually backslash) around each offending character to denote the parser to treat it as a symbol, and not do anything special:

GDscript is basically an NIH/in-house scripting language. Despite being p. similar to Python syntax-wise, the devs mention that they had more dificulty using Python or Lua for scripting and just made GDscript instead.
Thank god for C# scripting coming in 3.0

is /agdg/ still broken? can't open the pinned threads.

Report them?

I can open the progress general, but not the help thread.
The latter is pinned and cycling

I will report you faggots every time you show up.

Is the only way of selecting multiple polygons in blender to click shift and then right click on each one of them separately?

Surely it can't be this unintuitive

B for box, C for circle select which is like "painting" your selection, and then Ctrl+Right Click will take the shortest path from the last selected item to the next selected one, doesn't always work the way you want though.

Thanks

Blender sure shows it was designed by programmers, it's so powerful but goddamn is it shit to use

How do I into quad trees?

I understand that it basically chunks data by representing it over a large area, but like, if I have a setup like pic related, how can I query it and give it a value like (50,50), and it "knows" that it falls within that range?

oke

isn't that just ctrl+d?

Now that I look at it CTRL+C says that something is getting copied
but CTRL+V is just opening the vertices menu so I have no idea how to paste

Are you trying to copy the vertices of one model into another?

I'm trying to copy polygons to the exact same model because I need another identical version to edit

...

har

I should mention that I want to use this structure for tracking temperature flow, and possibly pathfinding(?)

so i need to use "jump" to jump the dialog that i dont want to trigger? any other way to code it?

anyway to isolate the options? i dont want to abuse the jump thingy

LOL

DELET THIS

shift + d if you want a back up
alt + d if you want to keep booth objects synchronized

ctrl+c copies the current animation frame
ctrl+v pastes the copied animation data to the selected object in the current frame
but doesn't saves it to the time line for that you need to press "i"
you can navigate the animation time line with the arrow keys
ctrl+shift+v to paste animation data but mirrored (very useful for walk cycles)

This doesn't seem to work for me. I tried searching for paste and I got the option to paste from the space search menu, however that copied the whole object.

In the end duplicate was what I needed either way

What would be the best tool in blender in case someone wanted to push out a chest to make breasts?

Something that can push out and also round at the same time

Extrude?

Kind of but more controlled. Extrude seems to just push it out without any control
Should I just sculpt it?

You're not gonna make, it at this rate.

Yeah no sculpting is definitely the way to go

I'm more than ready to sculpt shit, I've just never done it in blender and I'm scared

Now that's the place to be.

I have the basic types of rendering worked out on the openGL side of the rendering engine. Im still working on the vulkan side which can render the cube.

Seems like when I sculpt stuff and then go back to edit mode, the shapes are usually completely different from what I had before.
They're fine if I go back in sculpt and object mode though, weird

How the fuck do I show a sculpt mask while I'm sculpt mode? Right now the mask is working but nothing is showing on my model making it impossible to see what is masked and what isn't

How is vulkan? I was going to get into openGL 1.1 for some 2D, but bleeding edge could be cool too

Don't do vulkan. Do openGL 1.1, its really easy and its really fast to write openGL 1.1.

This is openGL 1.1 code:
gluPerspective(90.0f,(float)gfx->g_width / (float)gfx->g_height,GFX_DEFAULT_ZNEAR,GFX_DEFAULT_ZFAR); //do something with polar coords to say: what angle am I looking from? into a vector form float radius = 4.0f; float theta = rads(gfx->cube_theta); float x = radius * (float)cos(theta); float z = radius * (float)sin(theta); gluLookAt( x,-3.0f,z, 0.0f,0.0f,0.0f, 0.0f,-1.0f,0.0f); glScalef(1,-1,1);

This is vulkan code:
void vkr_init_uniform_buffer(vkr_t *vkr,sys_t *sys,jmp_buf jump){ //create uniform buffer VkResult res = 0; vkr->mat_projection = mat4_prespective(GFX_DEFAULT_FOV, (float)sys->vidmode_list[sys->vidmode].width / (float)sys->vidmode_list[sys->vidmode].height,GFX_DEFAULT_ZNEAR,GFX_DEFAULT_ZFAR); vkr->mat_view = mat4_look_at( (vec3_t){2.0f,3.0f,5.0f}, (vec3_t){0.0f,0.0f,0.0f}, (vec3_t){0.0f,-1.0f,0.0f} ); vkr->mat_MVPC = mat4_mult(vkr->mat_view,vkr->mat_projection); VkBufferCreateInfo buf_info = { .sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO, .usage = VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, .size = sizeof(vkr->mat_MVPC.m), .sharingMode = VK_SHARING_MODE_EXCLUSIVE, }; res = vkCreateBuffer(vkr->device, &buf_info, NULL, &vkr->uniform_buf); if(res != VK_SUCCESS){ longjmp(jump,vkr_error_depth_view); } VkMemoryRequirements ubuf_mem_reqs = vkr_init_allocate_memory(jump,vkr_error_uniform_buf,vkr,vkr->uniform_buf,&vkr->uniform_mem); void *ub_data; res = vkMapMemory(vkr->device, vkr->uniform_mem, 0, ubuf_mem_reqs.size, 0, &ub_data); if(res != VK_SUCCESS){ longjmp(jump,vkr_error_alloc_uniform_memory); } memcpy(ub_data, &vkr->mat_MVPC.m, sizeof(vkr->mat_MVPC.m)); vkUnmapMemory(vkr->device,vkr->uniform_mem); res = vkBindBufferMemory(vkr->device,vkr->uniform_buf,vkr->uniform_mem, 0); if(res != VK_SUCCESS){ longjmp(jump,vkr_error_alloc_uniform_memory); } vkr->uniform_info = (VkDescriptorBufferInfo){ .buffer = vkr->uniform_buf, .range = sizeof(vkr->mat_MVPC.m), };}

And of course, the utility functions I had to write to make this function work:

//obligatory unreadable matrix multiplication functionmat4_t mat4_mult(mat4_t m1, mat4_t m2){ mat4_t r = {0}; float tmp[4] = {0}; uint32_t row_offset = 0, col_offset = 0; for(uint32_t j = 0; j < 4; j++){ col_offset = 0; for(uint32_t k = 0; k < 4; k++){ for(uint32_t i = 0; i < 4; i++){ tmp[i] = m2.m[col_offset + i * 4]; } col_offset++; for(uint32_t i = 0; i < 4; i++){ r.m[row_offset + k] += m1.m[row_offset + i] * tmp[i]; } } row_offset += 4; } return r;}//this uses DEGREES because people have to set the fov and you dont think about it //in radians unless you are some kind of NERD HAHA wow//copy of GLM prespectiveLHmat4_t mat4_prespective(float fov, float aspect, float near, float far){ float angle = (float)tan(rads(fov / 2.0f)); //ok this might be important... is GLM library doing its matrixes //row first or column first? ask nogames if anything breaks also mat4_t r = { .m = { (1.0f / (aspect * angle)), 0.0f, 0.0f, 0.0f, 0.0f, (1.0f / angle), 0.0f, 0.0f, 0.0f, 0.0f, ((far + near) / (far - near)), 1.0f, 0.0f, 0.0f, (-(2.0f * far * near) / (far - near)), 0.0f, }, }; return r;}//copy of GLM lookatRH (used to be a copy of LH)mat4_t mat4_look_at(vec3_t eye, vec3_t center, vec3_t up){ vec3_t f = norm(sub(center,eye)); vec3_t s = norm(cross(f,up)); vec3_t u = cross(s,f); //unsure- does line 560 of matrx_transform.inl mean that the matrix is initally 1? (SEE GLM GITHUB) //I am making this assumption, refer if this is broken code mat4_t r = { s.x, u.x, f.x, 0.0f, s.y, u.y, f.y, 0.0f, s.z, u.z, f.z, 0.0f, -dot(s,eye), -dot(u,eye), -dot(f,eye), 1.0f, }; return r;}

Now think about it… you wont use perspective in a 2D game, but you will have to do hundreds of things like this if you use vulkan, when in openGL 1.1, you just call a few functions and its much much frendlier and easier to figure out why your code doesnt work.

Now with vulkan… my ~1500 line pile of vulkan code DOESNT WORK AGAIN and it worked THREE HOURS AGO and i didnt even change that much but now its BROKEN and I have to fix it because it crashes for NO REASON AGAIN FUCK SHIT

...

just going to follow up to this post i did ONE FUCKING THING and then it PUT IT ALL BACK and its STILL BROKEN and i even have the vulkan debugger on and it wont tell me what the error is FUCK YOU why didnt i make backups

...

im going to say that something is wrong with my computer because it cant even create an openGL context for my program which NEVER EVER fails… i'll restart it and check again

You niggers should team up

...

Oh wow, good point there.
I'll stick to it then. I'm using SDL to open my window, if that makes any difference.

but, why?
If you think raycasting is cool, search up sparse voxel ocrtees and and real time raycasting
Here's an example research paper nvidia.com/object/nvidia_research_pub_018.html

okay, so this is embarrasing… usually to test the vulkan renderer i make it so that flags = GFX_USE_VULKAN

so i was setting it up to use openGL and instead of changing line 40… i changed line 43… and broke everything and was completely clueless for about 15 mins

Sometime back then, I surmised that I should just use real 3D instead of raycasting.
Sure enough, I was able to do a simple dungeon crawler map-traversal using Unity, but now, I got interested in learning raycasting, so here I am.

Yeah, if you are new to graphics APIs, there is not a reason, Vulkan is designed for big companies like valve and epic games and whoever makes unity because they probably have a team of guys whos job is just to crank out Vulkan code and to know everything about the graphics API and optimization for the GPU… Vulkan is barley within my capabilities and sometimes I have to ask a friend for help when I get stuck.

If you look at the stuff on nehe.gamedev.net it has a whole load of tutorials on legacy openGL, its got lots of cool stuff that will help. It really does not matter if you do context creating with SDL, until you want to call on windows_functionality_9001 which does not have an SDL wrapper around it, and you would have to call into win32 API. You probably wont ever want to do that.

People might say NEVER USE OPENGL 1.1 because its old but it is compatible with basically every graphics card ever, and you don't need all the new functionality in the later versions.

Right now I use the win32 API instead of SDL, its not really a burden to use platform speicifc APIs for me since its like… 1000 lines of windows / linux specific code by the end of it? 1000 lines is not a lot of code and I like to learn about how they work. windows and X actually work somewhat similarly, except X has a way cleaner API.

By nature the (quad) tree data structure is hierarchical, and makes this query for which quad the position lies in really fast; this is best done with a type of bounds operation starting from the top of the hierarchy i.e. the root node (I'll run through that below).

So you have a root node, which encompasses the entire space, and the root node has four (quad) children nodes; all of these nodes have bounds (various configs, I prefer: min position, and size for bounds checks).
Thus, you start from the root, check if the position (50,50) is within the root using a bounds check (easily searchable online), if within the root bounds, then check which child it's within (0-3 children, i.e. 0 = (0,0) to (127,127), 1 = (0,128) to (127, 255), etc). Thus you get which child node (0-3) of the root node the position is within, and you can keep going down in depth if you have additional subdivisions if desired with a few passed parameters/logic.

So to put it another way, I cant possibly know the answer until I hit a leaf node, and the purpose of the bounds checks helps me drill down jnto an appropriate node?

I only want to use openGL for pure performance reasons, and the simpler the better for my retarded ass. The plan is to write a 2D engine for an open source harvest moon FoMT clone with some modding friendly abstractions, or maybe even a scripting language.

What's that? Or a better question; in what context would I end up needing to call it for?

its an example, the windows API has an incredibly large amount of things that it can do and maybe you would want to do one of those things? SDL will always be a wrapper and it will never ever have as much functionality, however you don't need 99% of the stuff SDL doesn't implement. If I want to do something like automatically open up the developer's website in a web browser from my application, SDL probably doesn't do this but win32 API does.

forums.libsdl.org/viewtopic.php?p=4923&sid=07288b71133380260270057d206b3cde

You probably wont ever run into any problems with SDL unless you want to do something obscure that nobody else wants to do, like that. What if I want "feature X" and SDL doesn't have it? Win32 might.

Really as long as you are using openGL instead of say, software rendering it will go way faster. I cant imaging optimization issues with 2D on modern hardware where quake 3 can run on anything built in the last 10 years.

Thanks Ross.

I don't see myself needing much more than what SDL/openGL will provide me with. But on the off chance I do, good to know where to look, thanks.
the faster it runs, the better it will run on older machines I guess. I hope to use it as a CV example when the time comes, hence why i'm taking
speed and engine architecture into account. that and why the fuck not

>tfw trying to draw a single sprite in GIMP haven't installed Krita yet
kill me pete

He's right.

yup

if the node you're checking is a leaf node, then yeah
otherwise it depends on your per node logic for recursively checking node bounds (and when it returns); as each quad is just a bounds cuts into four pieces.
so… the accuracy needed determines when u can return

Yet I find good finished games at a rate of about 1 per 3 years.

That thing has been in development for 3 years if not more now.
It looks good, but I wonder if he'll have the strength to finish it and not rush.

You have to redesign the characters depending on the resolution you're going for so the face and other important parts are clear during action. It's more complicated than adding a filter.

I doubt it's because they don't exist. Most of the newer indie games that Holla Forums brings up on here already have a pretty decent following. It just seems like nobody here really wants to try anything unless they've repetitively heard good things about it already. This isn't including shit like the fact most people on here will drop a game instantly on some small detail that triggers their autism. The only thing that really gets a free pass is stuff trying to basically recreate an old dead game or weebshit.

I'm more afraid the Polish Mafia does something to him.

Have only certain different vertical spaces the shelves can have. Take a tape measure to the grocery store next time so you know the sizes.
The game will be much easier to program and play if everything is in a cubic grid system, down to the height of the ceiling. The more automatic it can get while providing variety, the better.
Also I would advise titling the game something like "Cleanup on Aisle 5" if the game's appeal is like Surgeon Simulator. Also have a couple of threads for anons to invent products for your shelves if the scale of the game is like Surgeon Simulator.

Is it inspired by Food Fight?

I want to use a simple easy to use scripting language like Javascript or something, but I also like to enginedev to a certain point. I'd like to do most things through code rather than some UI engine. I can't do graphics though, programming for GPU shit is 2hard for me. I've also had performance problems when trying to code my own path finding method.

What tools/methods should I use to just like make game?

HE won't but that doesn't matter if motivation doesn't return somehow. That one game just flat out broke me, sorry.

What could happen is that I get shipped off to an unofficial debtors' ghetto out of town. The mafia thing is a meme but I'm not sure what else to call them. I could also get randomly stabbed, because this is possibly the worst part of the town.
Also, the druggies next door nearly burned down the building a week or so ago, I yanked out my hard drive and was ready to leg it but in the end didn't have to.

Implemented kinematic objects. But I wanted triggers because in webm related, the jumping also collides with the box. Trigger would have checked the "depth" of the player (order in layer) and only "collide" with those objects. Can I change behaviour of kinematic objects to corresponds with that, or back to triggers?

That has nothing to do with triggers, and they won't fix it. The problem seems to be in how you do depth logic (at least based on bservation of the behaviour in that webm). It looks like you're emulating 3D movement in 2D in a 3D engine.

Keep working. Automata won't be fresh in people's minds anymore by the time you're done. Also, it doesn't hit my bodysuit fetish buttons the way your game does.
I believe in you, even if I can't do anything else for you.

I know a little basic music theory and I play the piano

It's already a meme. They took my plan and executed it perfectly. I have nothing left now, you can't get motivated if you already know the outcome. I'm sorry.

Maybe try changing up the aesthetic or thinking up new story elements to distance it from Automata and thus bring you a bit more motivation?

just give the project to someone else than would be a shame to let all that work go to waste.

That's the exact opposite of motivating. I shouldn't have to. I don't think you would do the same if asked, either. The insult to irony is how C&D happy Square is known to be. I really just want to go back to work, but you can't work like this. It's not one single thing that makes me want to quit, but the whole picture. Some of it will go away, some is irreversible, and some things can't possibly be confirmed.
I'm desperate enough to confront them directly at this point and clear up at least a tiny bit of it, but I don't even know where to start.

reported, It seems the mods are inactive

Shit, I had not considered it from that perspective. The game is simply too similar in some aspects and only through sheer coincidence then again, Moonman does possibly browse here. Square has been known to do scumbaggy things to shut down competition. Shit man I don't know what to say. I wish you luck either way.

Anyone here who has some good tuts, forum posts, pics or meshes about lowpoly deformations like pic related? For ~2k tri characters. Or games that did deformations really well with lowpoly characters that i can look at?

...

Your picture comes from Polycount's LimbTopology page, start there. Halfchan's /3/ and possibly our /3d/ should have stickies with info.
For the most part, that gif already shows you all you need to know. You need to have two vertex rings if you want a joint to deform correctly. Anything on top of that should only be for the silhouette and shading's sake.

No. That would be true for DMC5, but it's missing the point entirely here.

So lemme get this straight
Your game is too similar to Nier automata, and because of it you don't want to work on it anymore?

It's funny because your game is actually nothing like Nier, aside from the most high level ideas (you play as a sexy robot girl and it's cuhrayzee).
Nier is an RPG
Nier has multiple characters
Nier has edgy (for lack of a better term) storylines
Nier is an open world game with quests
Nier has fishing
Or basically, Nier has a lot of things that take away from the standard cuhrayzee game experience, and is more tailored to being a Yoko Taro game. As long as your game is mre focused, there's no problem as the design will be so completely different people won't even compare it to Nier.

Nice, this is exactly the kind of shit i've been looking for, i've been browsing polycount forever but somehow never came across this. Thanks.

it's visually similar to Automata, which
>has a very loud fanbase
The game is fine but I'm not. I judged my chances as pretty good, all of it seems gone now. Can't get motivated for a failure. Can't get motivated for anything else either, if you did everything right and something like this just happened anyway.

If this is the same game I think you're talking about they don't look even close to similar.
Your art direction is not nearly as good as Nier.

Seems like you're just being a big baby to me. Your games are completely different, they're just fucking action games for fucks sake.

Stop being a faggot user

Finish your game, at least for the sake of having a finished game under your belt.

Nice dubs but come on, man. Are you a fan of DoD? DoD fans play Neir for Yoko Taro and, in your own words, the new game is a meme game. Cuhrayzy fans play cuhrayzy for cuhrayzy and it IS a dead genre for the most part.
You were never going to attract DoD fags unless you were edgy and you were never going to get normalfags attention.

You should know yourself that your fears are blown far out of proportion for silly psychological reasons, like the worst outcome is the only thing that can happen. Think about how to market your game to highlight its differences to Automata instead of rolling around in your bed screaming about the similarities. Or just find some fun games to distract you for a few weeks, maybe you'll start feeling it again. But don't keep on being the bitch you are right now.

Mate, half of modern japanese games look like nier automata.
Most of the gameplay isn't something new either.
It's a genre and style you can clone with barely anyone complaining.

For example i am planning to make a beast fighting game, and i already know people will call it a pokemon clone, even if every single detail except monsters fighting each other is different. It's barely as bad as kids calling every block game minecraft. Anyways, this also spells a chance for you, because fans of certain games wish to play similar games when their done with that game.

You could gibe .it file.

jezus fucking christ man you will you stop being such an unbearable fucking faggot and just finish your game already.
Like how long has it been like 3 years or something? How much longer is it gonna? another 3 years you'll probably be competing directly with the next nier game that comes out instead of this one.
The genre of game you are making is not that niche there is a good chance several games will come out in the same genre before you are done and you should have known that from the start.

Nigger finish your damn game! i want to play it

Add cute undead please.

??????

And look man. If you don't want to complete it AT LEAST release the source/project files. I'm Kowloon user and I'd love to make some environments for that glider.

Seriously. I intend to make a cuhrayzee game in the future too, and having the source of one on hand would be useful, I'm sure.　That said, I still think you should finish your game, or at the very least, implement its content/features into another one.

Happy birthday to me!
Will post a playable demo as soon as I make a song for levels 11 to 20, probably next week.

Hey, I made something a little while back.
Released it. People liked it. First-person crazy-action game. Features a protagonist wearing red, fighting demons while using a sword and a gun, doing stylish stunts with a rating meter.

Do you know what everyone thinks of when they see this? "First person Devil May Cry". A lot of people use "it's like Devil May Cry" to describe it. Even a lot of the louder critics, just saying "oh he ripped off Devil May Cry". I haven't played Devil May Cry.

I also know an acquaintance who made a top-down shooter, run around through an area and shoot dudes up, everyone dies in a couple hits. Retro soundtrack.
jmickle.itch.io/coffee-mafia
Everyone keeps saying he ripped off Hotline Miami. He hadn't played it, either.

Similarities and comparisons THRIVE in a medium as broad as ours. Everything that ever exists or is made can universally be boiled down to the same static building blocks and archetypes which have been used to build hundreds of other things, and in these we get the "Like X But With Y" comparisons.
It's been there since the start of gaming (starting with the reign of "Like Doom But" or "Like Mario But" in the 90s), and likely will persist until the end of time, when the last keyboard is put down and the last shitpost is sent.
Whether intentional or not, people love to compare as a shorthand, and trying to avoid these is futile. If people think you're a ripoff, nothing's going to change that. Even if Nier had never come out, people would say you're taking from Rodea the Sky Soldier or Metal Gear Rising With Tits.

Make your game, mate.

Happy birthday Speebotdev you cute son of a gun.

Oh shit, this looks amazing. Great job!

After I'm finished with GigaMaidens I want to look into making a platformer. I'm leaning towards it being 2D and sprite-based, but I also have ideas for it being 3D as well. If I went the 3D route I'd want to make it like Super Mario 3D World, and if it's 2D I want to make it like a combination of SMW and Wario Land 4.

I made this sprite a long time ago. I did make an engine test with a friend a while ago but it was extremely basic and was nothing to look at. I was in the middle of creating animation sprites for her when I stopped.

This character is named Kasha. She's a catgirl with a magical tail that can shapeshift into tons of different objects. She can use it as a grappling hook, a flashlight, a power drill, a hammer, a coil spring, and all sorts of different things. She can only have one "transformation" equipped at a time and you press a button to switch between them.

You worry too much. If the game is good, there will be people who will like it.

I've played Coffee Mafia! I liked the aesthetics of it more than of Hotline Miami, actually. I just wish it was longer.

Thanks!

Shut up and dev, faggot

Monthly reminder that I've been devving for almost 7 years and haven't even gotten far enough to release a demo and constantly remake projects

Happy Birthday, user. Looking forward to the demo.

It's always great to see your progress and I'm really looking forward to trying your demo.

Happy birthday! Really looking forward to the game.

Right, and happy birthday to you too

FOOL

It's also Holla Forums's birthday.
Mark will eat your cake with no survivors

Seems like your project's coming along nicely.
Happy birthday (w/ or w/o dicklicking).

Happy birthday user
I wish you the best for your game and deving skills

Working on this again after about two months of nodev. Not quite sure whether to keep finishing the PC moveset or look at enemies/platforming tests next, though.

Full disclosure, music/sfx are stolen placeholders.

oh well, time to redo this shit with less pixels.

spends 3 days on wifu
its a nigger

wut?

Looks like the torso is a completely stiff block, independently shifting in position and constantly breaking whatever shards of bone attach it to the hips. And there's some kind of strange wind singularity moving that skirt around.
Hell, most of the body parts seem to not really be connected, just moving on their own with no respect to the rest of the body.

Also something something race war.

why do black people trigger retards so much?

yeah, learning to do animation and was copy pasting and moving areas without redrawing them, like tween animation.

You should probably learn how to draw first.

Will I get caught if I use a pirated software to create my own models and make money to pay back and buy the software?

I want to use Zbrush but
there's no fucking way I can pay for that jesus christ

also I can't vpn because I already bought it + I don't want to look suspicious because I'm not an avid pirate.

Fuck off user cmon

It's much better anything I've seen from the mugen comunity that isn's stealing sprites from capcom.

pixologic wont give a fuck if some starving nobody pirates it's software retard.

sorry
I just wanted to play safe because I heard on /3/ last year that some companies are nazi jews who will sue anyone if they can

How about you try using FOSS software instead?

they will sue other companies retard, not some random nobody who doesn't have money.

just don't be a retard and tell anyone you pirate the software.

And it's much worse than any given doujin fighting game, aside from being bad from an objective point of view.

which doujin fighters?

I can only think of Vanguard princess and the DBZ mugen fighting games with better sprites.

confirmed for never having to work/live around niggers

I'm colombian bro, and my city is 30% black people.

Sorry, but americans need to leave their thirld world nation and meet blacks from other countries.

It's like claiming white people are trash because you have only met rednecks, slavs, jews and gipsies.

And yes, there are black countries with less crime than germany, ghana and madagascar have less homicides than the average european nation.

Sorry, but I've read more about black people than 100% of all pol retards.

E's Laf, Akatsuki Blitzkampf, Wonderful World, Eternal Fighter Zero, Melty Blood (even just counting the early characters), any Touhou fighting game. Fearless Night if you want to count indies. That 2D pokemon fighting game. Pick any given game and it looks better than what you put out. Your art has no style, no sense of proportion, extremely stiff animation and just looks bad, I don't care if you try to justify it by pointing at even worse art.

I'll try to make sprites inspired from other styles and post a poll to see which people like more.

How can A* ever return multiple paths? Like if I have a fixed grid and two points, I get there are multiple best paths, but wouldn't the function only ever return the first set of valid points it finds?

Don't humour this guy. He's a massive newfag who refuses to take any criticism and will gladly call people racist for disliking him.

A* star returns the list of nodes from the last node to the first.

I already said I will make another animation following other games styles.

Also, why do black people trigger y'all fags so much?

Work on your artistic fundamentals, no style you try will look good without fundamentals.

I was copying scwu style, but it seems it triggers anons so much.

Yes I know that, but if you're going between two points with no obstacles, there are many many equally valid paths. How do you know which one is taken, and does it matter?

theory.stanford.edu/~amitp/GameProgramming/Heuristics.html

It returns the one that has the smallest score.

Bullshit. Also we have I.Ds here, retard.

Notice how the artist behind it actually knew how to draw. Your art is nowhere near as good as that image.
Because it helps to get rid of fags like you.

And that "style" is shit and you need to focus on fundamentals first before doing any sort of "style".

If it helps to identify how much of a fag this guy is
ca3.tumblr.com/

I don't are about pol cucks.

I'll try to make a more realistic style, then.

Im vanillafag

Im thinking of adding mommies

...

...

ayy lmao i remember this fag
he made a terrible pixel art thread and complained about how much effort it takes to animate it
ca3.tumblr.com/post/158819194534
archive.fo/H4QVS

That's not what anybody said either, you dumb cunt. Learn how to read: develop your artistic skill set and then create a style. You have to learn fundamentals before you can start bending and breaking rules without it looking like shit. You don't build a house without a stable foundation. What you're doing right now is slapping together 2x4s and making a shitshack.

What's wrong with that link?

Isn't that the natural inclination of his favela-dwelling kind, to be fair? On that note, would it technically be a favela outside Brazil, or merely a Colombian shantytown slum? The fine distinction escapes me.

tumbler of the Colombian mutt with a thing for niggers

I still can't understand why black people trigger pol so much.

Oh this is rich. Please, never stop posting. We all need comedy relief.

I already said, I'll try again.

and I'll try again until people like it.

You'll never improve with your attitude and we don't want you here if you're such a newfag that you complain about racism.

Guess that means I'll go with the Switch if I ever get around to console development. Fuck paying over two thousand or even whatever the Xbone's current price is for a glorified x86_64 PC.

No, try again until it's good. People liking shit doesn't make it good. People love eating shit all the time.

I'll try again and I'll try again.

I'm not complaining, but It gets tiresome to hear the same pol BS everytime I draw a black person.

if I want a hughbox I'll post it on facebook, and normalfags on facebook told me looks good.

that's why I came here, to be shit on.

Honestly, what kind of mental disorder do you have to think you belong here? Fuck off. Go back to where you belong. You're not welcome here. Are you too stupid to understand this?

I've been on the chans since 2010, newfriend.

just because I'm interested on blacks doesn't mean I'm from reddit.

And yet you won't listen to anything any of us are saying. You just want some sort of feeling of superiority.

...

You act like that's a long time, kek. That year puts you right in the middle of the start of the reddit invasion of 4chan and even other chans.

Colombia was a mistake.

...

No, I already know I need to work on my fundamentals, I guess I'll practice them this week.

I like blacks

I don't even browse reddit or facebook.

Then act like it.

If you like niggers so much, you should learn how to draw them instead of somehow making them look even worse.

yeah, I need to go.
I'll try later to make better sprites.

How do you manage to contradict yourself in under 10 minutes?

Well, you got two options. You could get better, maybe read up on Loomis, The Animator's Survival Guide or whatever else you find, and spend some time actually integrating to imageboards, which feels less likely to happen with each post you make. Alternatively, you could cater toward the diversity-lovin' folks, where you get unwarranted and exaggerated praise for anything and everything you do, as long as it's diverse enough and you avoid ever saying anything that could be considered offensive by anyone. Easy enough to get a following, but it's going to be brittle. One offense, fabricated or otherwise, and you'll be dropped like a lead weight.

For someone with that many years behind you, maybe you just don't have the aptitude to integrate or express unpopular preferences and opinions. Maybe consider the latter option.

ZETA MALE

where did time go?

which one is it? I was going to trow you a bone but your story doesn't add up

Your post makes me wonder if we should make a chart of what AGDG devs are like by nationality.

you do realize there are black girls that are models, right?
and I posted an african, so you can't give me the she's mixed argument.

I have a facebook account to talk with my normalfag school friends, but I don't use facebook nor I care about it beyond an indie developers group.

when I say I'll make a more realistic sprite I mean I'll make a sprite taking into account loomis and all the ic meme books.

I have a facebook account but I don't use facebook.

Just stop posting user, you're sperging out massively

yeah, I'll go.
bye

You still here?

We don't give a shit. Racism is an integral part of imageboard culture, Mr. "2010 but acts like he started yesterday".

The nightmare finally ends.

L-look, I found a path

Nice. Why did you need all possible shortest paths instead of just one?

No I didn't, I thought it was a complication that came up.
Also you should be able to tell this is a very incorrect solution since it goes straight up, then back down for no reason

It's still also not returning the idea path .. like I have a collection of visited points, but I can't really prune it to one path? If that makes sense

Maybe your problem's looking at the wrong data. The way I remember A* is it returns you an open list of tiles it checked in the process of finding the shortest path with each tile having a parent. Then you take the parent of the goal, parent of that parent and so on until you get to the starting tile and I think that was your actual shortest path. Are you doing something like that?

It works but I have no idea why
I know it's broken together and works, but it's all wrong

I stored my visited points as a Dict and it generates a single path with no revisited points, now. The key is the node, the value is the parent node.

Still crashes if start/end is on a wall. Not sure how to test if a path is possible, or exit early if all nodes have been checked (it just continues forever for some reason). Lots of other shit to consider, too. Also need to generalize it into its own class, sadface

Also my heuristic function does not fucking work. If I change any of it, it breaks extraordinarily

From what I gather A* looks like this:
Check and double check whether your algorithm is actually doing this.

Your code looks like it's only considering the heuristic as distance measurement, so it's a bit short sighted. Also seems to stop as soon as it finds some path to the goal. I honestly can't fully understand the language or your implementation, but this is my feeling.

The crash is probably as easy to fix as checking whether start and goal are pathable before going into A*, and returning "no paths" if one of them isn't pathable.

Any linux users here who can preach recommend me a decent Code editor? And if it's one of those meme editors the laundry list of addons. I have come to the conclusion that I am way too shit at this as for me being able to code even the simplest things with a mere text editor.
Nice games yall

Text editor or IDE? If the latter, what language?

Is that not what the heuristic does?

Yes, it exits early when the goal is found.

But how can I even do that without finding a path between them?

i use qt creator

I kinda want to get the behemoths that most IDEs are out of the picture for early code learning though that's not a set in stone thing. A text editor works as well as long as it's reasonably supportive(detect the language, detect some errors, helping with style).

Oh right I totally forgot about that one, do you have a good tutorial or something for it? The few YT videos I tried were all rather horrible.

The heuristic is an estimate of the distance to the goal, but you need to also consider the path you have from the start to where you are, or else the algorithm may take a path that looks like has the shortest remaining distance at every step but actually doesn't because of walls or something, so it ends up looking short sighted.

Exiting early is unacceptable, if you exit early you just find some path to the goal, not the shortest. Point is in open list = there is some path to it, point is in closed list = we know the shortest path to it. Your algo breaks when the first case is met. Try to implement it again, I think.

For the crash: If start or end are a wall get out early. Done.

just install it and use it it's not hard

KATE for text editing, QtCreator for C/C++ (you don't have to touch Qt with it. It's just easily the best C/C++ IDE on Linux).

This is why I asked about multiple paths earlier. And if it found a path, taking the shortest step every time, it should be guaranteed to find the shortest path.

Sure, but "islands" still make it hang forever

I tried that. That one hacky part where I call h_a=heuristic() is the distance to the start, and h_b is the distance to the goal. Except if I try to incorporate h_a in any way, I end up with something like this, if it doesn't hang/crash… With only distance to the end considered, I get 367 points, distance only to start, I get 6732 points, and if I take the minimum between the two, its still 2894 points

h_a isn't even the distance to the start, it's the heuristic to the start. I have no idea why you're doing it, you know what distance to the start you have. You already have a path.

Hanging shouldn't happen. Are you sure your visited list works correctly? Are you sure you're not adding points to the open list you've already visited? You should eventually have visited all reachable tiles, get an empty open list and exit the while loop unless something's fucked on that front.

My heuristic function is to use Manhattan Distance, so it gives me the distance to the start. What?

I don't have a path yet, because it hasn't been calculated. That happens at the end when it does a traceback from the end to the start.

Also, calling the min between the two actually did see an improvement when I did diagonal movement. But it's still checking far too many tiles (and I can't draw the visited ones unless I destroy the path I calculated and return an entirely different set of values)

The Manhattan Distance is a HEURISTIC. It may underestimate distance, oekaki related.

But I don't know the distance until I calculate the full path

Also, I can't really tell it to find an alternate path

You get distance to the start if you're keeping track of the parent chain and reducing distance to summing up distance of neighboring tiles (should be 1 each for you), which you should do, because otherwise you're not implementing A*. The distance of any point in the closed list gotten that way is shortest, because that's how A* rolls. That that just werks is the basis of the whole algorithm, and because you don't implement a closed list anywhere your algo is shit and doesn't really work.

Honestly, at this point I think you should just throw everything away and implement the algorithm exactly as I have greentexted it some posts ago. What you're doing is a giant mess.

But it is exactly what you've described

And somehow even though it explicitly checks that a point hasn't been visited before adding it, it still has duplicates and fucks up

Also I don't understand why it works when I take the manhattan distance towards the goal exclusively and ignore the distance to the start. By definition on this page (redblobgames.com/pathfinding/a-star/introduction.html), I need both in order to avoid concave traps, but they don't affect my implementation, it works

Nevermind, I'm just going to fucking play VIDEOGAMES before I wreck something

Give Love2d a try then. It's used with Lua which is really simple and flexible to use.

In order to find out whether you can reach point B from point A efficiently, you should pre-calculate that data. There are a couple of ways to go about that and they vary based on how your maps tend to look like. They also only work on static maps. If your maps always change, you have to solve the problem differently (multi-threading, limiting the max. path-length of the algorithm to put an upper boundary on how long it takes in the worst case, etc.). I'll assume a static map here.

The important question is whether you have a mostly open map with some unreachable spaces, or a map that consist of many separated sectors that are open in and of themselves.
Dealing with the former case is super easy. All you need is a set and the floodfill algorithm.
You now have a set that contains every unreachable point. Before running A*, you simply check whether the target is part of that set.

If your map consists of sectors, you need a dictionary (also known as map) and the floodfill algorithm.
You now have a dictionary that allows you to check which sector a given point belongs to efficiently. Before running A*, you check if the start and end points are in the same sector.

Every time you pop a value you need to add it to the closed list, with its parent and cumulative distance (parent chain + heuristic to goal) as additional info, you're adding neither of those. Every time you push a value you need to also define its parent (the point you're currently examining) and the distance (same as above), only add it if it isn't in the closed list or already in the queue with a shorter distance value. You're not keeping a closed list or parents and that's why your implementation fails, you absolutely NEED to implement those concepts for A* to work.

Have fun, I'm overdue for bed anyway

But I don't need to know the cumulative distance, since I can just figure that out when I'm backtracing everything

you're right, but you're also on a board that's been populated with 4/pol/s rejects for about as long as it's been around.
learn to use the filter function and keep going

You can also go back. Anyway, does anyone here have Source experience? Any idea how to program in a dialogue box E..Y.E style? I want to begin adding pretentious writing and depressing NPCs into the city. Sorry if my request is a bit much.

This is stupid

I don't have experience with source in particular but I do know about what it takes to make a dialog system.
Before you start actually writing any code you should figure out what exactly you want from this dialog system. Do you want to be able to reuse sections of a dialog tree? Do you want to block off certain options based on the state of the world?
After that you have to figure out how to work with the UI system Source has. Figure out how to display lists of dialog options of arbitrary length. Figure out how to change the contents of a text box when you click on an option.
Next is the issue of walking through a dialog tree. You can either use XML, YAML or JSON to store the options but how exactly you go about parsing those is up to you. Look up how to parse those with C++ and figure out which one you like the best.

Or if you just want to steal someone else's dialog tree and adapt it for Source Engine then there's this

etodd.io/2014/05/16/the-poor-mans-dialogue-tree/

VIM

VIM, if hotkeys aren't your thing nano, if you want a meme editor, atom with code minimap
and of course, there's always microsoft notepad emulated in wine if you're feeling dangerous.

I messed around.

I like this a lot.

Now that's what I'm talking about.

Since rempy uses python is it a good idea to get this?
humblebundle.com/books/python-book-bundle

Welp

But see now it's working again, I stripped everything off, and completely ignored the cumulative distance and heuristic

Like I still have the distance to goal, and it works so??

The algorithm has learned to ignore the boundaries you set up. Kill it while you still have time.

No, it's a shitty wrap-around because the data is in an int[], rather than a Point[] or what have you

Oh nevermind, it was only reporting the optimal path, it still has a ton of overhead

If I swap the start and end points, is this sort of thing normal?

Automate the Boring Stuff with Python is available for free online automatetheboringstuff.com/chapter0/

The first 10 chapters should be more than enough to get you familiar enough with the language to use renpy without to much confusion.

You've implemented a breadth first search now. Works, always gives you the shortest path, but has tons of overhead. Maybe try a few other algorithms from the Introduction to A* page, see if it helps?

*always gives you a shortest path if all passable tiles have the same movement cost.

The sort of thing in your last post is completely normal for the algo.

yes you do that's how the algorithm works.

no it should be symmetrical

please tell me you are backing up your stuff off site.
and do you really think people will just say "oh its just like nier and i dont ever want anything like that again?". its a singleplayer game and by the time your game comes out most people will have finished nier, so your game being perceived as similar would probably help you more than it hinders you.

Im still having issues with grasping the heuristic. I get that its a distance estimate and that its used to weight between favoring the goal vs the endpoint but i dont know how to determine the weight

it's a distance estimate from the current node to goal not a distance estimate of the path.
Honestly I would just try to implement dijkstra if I where you it's literally just A* without the heuristic
heres a great source redblobgames.com/pathfinding/a-star/implementation.html just go through all the steps slowly and make sure you understand whats going on.

If your heuristic is just the distance to the begin point, you have breadth first or pretty much dijkstra's algorithm.
If it's only the distance to the goal, it's depth first or naive A*.
Both together is an exact middle point between the two. Usually that's not the ideal weight for your particular needs, so you multiple one or the other with some number, which is almost entirely just trying shit out to see what works best on your maps.

...

anybody know where I can find a game built in SDL 2.0 that is open source?

You could build a bot that scrapes github for projects that include the library

Tesseract is the first to come to mind, but most newer or still-maintained 3D freetard games use SDL2 so it shouldn't be hard to find more examples.

Bewbs,

also what am i supposed to do with the hands?
like shit, what do hands do when you run? anything else i can do better?

When you sprint you're throwing your hands forward and up to help momentum and balance. Yours are moving too linearly and with too much horizontal motion.

Hand movement while walking can reflect a lot of mood/personality. You can be cocky, defensive, passive, flirty or aggressive just by the way your arms are positioned and move while walking. A shy person walks differently than a catwalk model or a 280lb meathead.

Yeah, so like it's basically lerping it from f(a) and f(b) in the range of 0..1 as a scale, but I don't know how to determine how far along the path you are (since it could be a long way away, and the position, start, and end aren't guaranteed to all lie in a line)

There is no "how far along the path you are", as you do not know how long the path is going to be. You just take the distance you have traveled away from the starting point, which is the distance from node X's parent + the travel cost of going to X (if you have terrain with a travel cost > 1). If you find a tile that you have already visited before, you check if the cost with this path is lower than the cost it previously took to get there, and if so you just found a shorter path to that node.
When I made an A* search myself, I had an unordered_map Costs, with TileID as the key and the distance it took to travel to that point as the value.

Okay, so say I have points A, B, and P. What if I took the euclidean straight-line distance between A and P, and called it "estimated distance", and then I took the cumulative distance actually traveled to get to point P, and call it "actual distance".

Then, if I go [est / actual], it should generate a number in the range 0..1, which can be used as a multiplier against my heuristic … which is the distance, which I've already used?

See this is what confuses me

see 3,4,5, and first part of 7 for applicable principles of animation to 3D.
also general references attached, and embedded is a great series on the 12 principles on animation; which covers what this user mentioned w/examples.

huh, i guess the embed made my images not post.

should mention that the motion principles 1&2 are generally handled automatically by the animation program… but are still applicable; though not as directly important compared to 2D for making a clean animation.

Interesting, could really see a simple Matrix[] doing lots of shit to a quad via transforms

I've never used estimated distance as a variable in my pathfinding, it is not part of any standard form of A*. You only use actual distance traveled, and manhattan distance of node to goal. Those two combined ARE the heuristic.

Depends on what you want your run cycle to convey.

If it's a frantic sprint, you could over exaggerate the arms and have it look alright, but if it's a light jog you'll want to keep them more at the sides of your character.
I think the issue you're having is not moving the arms enough vertically. People tend to "pump" their arms towards their chest when they run.
You might also want to add more hip and chest motion if your rig allows it, but that's more personal preference on what type of run it is.

is it just me or is the one step longer than the other

nah, you're right. I need to redo that cycle because I don't like the leg movement, but there's something off about the timing, too. Was just using it to demonstrate the arm movements.
Probably should've posted reference from someone who's better than I am. Something like this.

her legs are too long

nice animu gondola

i spent all night configuring a big fucking kernel

prepare to configure it again because wine just got dx11

You don't have to recompile the kernel to install a new version of Wine.

fuck

I want to cuddle Long Legsu-san.
Got any advice for a beginner looking to learn animation?

...

i think it may be the perspective it's rendered at. I still have to experiment with that some more.

This post is going to be a blogposty wall of shit vaguely related to gamedev, but we're at the bump limit so I hope it's okay and people don't get mad. I'm still a beginner myself and I'm not really comfortable giving advice out, but I'll give it a shot.

The best advice I could give is to study other people's animations and learn from them. Analyzing animations frame by frame really helped me when I was starting out, because it helps you get a grasp of timing, how weight should be shifted around, how to pull off certain movements, what good poses look like, framecounts of certain movements, etc, so long as you're studying from people that have a strong understanding of those things (not anything i make).
The best animators out there will be aware of how each frame in their animation looks, even in 3D animations. My favorite internet animator to do this with is Stephen Vyas (webm related), even though he doesn't animate for games. His animations are smooth as all fuck.

Read books if you're a complete beginner and you need information on the fundamentals. Timing for Animation by John Halas and Harold Whitaker is good. The Animator's Survival Kit is ok too, but it leans more towards 2D animation than 3D, whereas Timing for Animation is more universal.
3D specific animation advice, the most important thing I've realized is 3D is more like stop motion than traditional animation, specifically when working in pose-to-pose with stepped tangents. If you treat it like that, you'll have an easier time.
Above all, you've just got to practice. Bouncing ball exercises are pretty important to nail early on, since they teach you a ton of the fundamentals, despite being so simple.

Also, animating for games usually requires an entirely different thought process than animating for non interactive media.
You've got to consider everything you've learned about animation up until that point, while also keeping in mind that above all, it's much more important the animation feels good and responsive for the player. I'm still trying to figure that out by studying fighting game sprites.

tl;dr, read Timing for Animation, study other animations, treat 3D as stop motion, and practice.

NEW BREAD