Amateur Gamedev General ~ /agdg/ + /vm/

...

Old and new bomb vest

rewrote the system and added support for multisized tiles. The snow is a 32x32 sprite that is split up into workable 4x4 chunks so that I can truncate it at the edge with added variability. Hopefully tiles like this wont seem as plain once I add things like water, trees and rocks to break it up. Pic related is a heavily inspired mock up tree that I dont really like and will change later.

new icons!

that image is giving me heavy flashbacks to something
can you post the full res?

they are all for ants
here the reference used

How's everyone's training coming along?

thanks, I seem to remember a dead free mmo where you frighted with birds or something

what is your goal? and how serious are you about it?

not bad, but next do it with your own assets

Unless you already have an artist to make assets for you i mean

That music really hit my nostalgia. Looks pretty good, but keep in my a game like KH are not small, if you make your own assets it will take a lot of time.

Just practice for now. It's more of a hobby for now since most of my time is taken up by a real jerb. I figured that successfully recreating the first world of Kingdom Hearts would give me most of the programming knowledge needed to create my own action RPG.

Saving the art stuff for later. Currently I'm just focusing on programming and functionality.

anybody got a first person rig I could use as a placeholder?

ok guys, how did i do. i just want to make a simple player model just so i can work on a demo. so how does this look, and what can i do to fix it if their are major problems. the only things i think that are problems are the forehead and lips, but i wanted a 2nd opinion first.

...

...

...

...

If you're working on a 3D game, what are you using for an engine?

UE4 or Unity if you're a masochist.

I made the first alternate subweapon for the doomslayer

tbh code up a game about spiders tbh

like this?

REFRAME

Made a little proof of concept project…
I wish I had a 3d modeler. Can only go so far with free assets, and the mech and ship look nothing alike…
My idea for the game is something like humanity fighting against xeno scum, like Starship Troopers, MuvLuv, or the space people from Gargantia. Exterminating some form of aliens in the vast universe, in your transforming spacecraft/mecha.
Something something, humanity will always lose in the end, and then everything resets and you play again.

You can't be anyone other than Labyrinth. Now I'm torn which project of yours I want to see completed more.

Wow, that would make for a nice gamejam project. Cool concept. Keep working on gameplay and try looking for a modeller around here perhaps

I want to fuck that spider

Remind me again how youre processing the tiles? I remember you saying it was slow. It looks great though

How do you find out which tile a player clicked on?

In what? Generally you have to project the screen coordinates to world coordinates, if it's 3D you have to use raycasting

cute buggos tbh

2d isometric
i also want to get the specific tile not just coordinates
say you have a 2d array of tiles, find out which tile the user stands on without iterating through all of them

Easy

Old RTS games like AOE drew a 4 colored image on every tile that was on a separate render target that was never visible. But using a function to get the colored pixel under the mouse, it would determine whether the selection box would tilt into an adjacent tile near the edge or not

I think you're just asking for the math to turn a x/y position into a tile, right?

Vector2i getTileFromPosition(const Vector2f& position) { return Vector2i( static_cast(position.x / tileWidth), static_cast(position.y / tileHeight));}

>return Vector2i( static_cast(position.x / tileWidth), static_cast(position.y / tileHeight));
I want to erase all C languages from existence

damn it feels to good to refactor and shrink xbox hueg classes

Okay, I played with this shit long enough, gotta start from scratch and do something useful with it.

I was about to just write pseudocode but it turned into C by force of habit. The static casts are just there for compiler warning suppression

How can I rig forearm/wrist to rotate properly along with hand in Blender? Webm related, I tried to add constraint to the forearm so it copies local Y rotation of the hand but it only works properly if I don't rotate the hand on other axes. If it's set like, for example in the webm, the quaternion kicks in and fucks up the Y rotation making the forearm jump, not twist enough or even rotate backward. I'm out of ideas to how to fix this, for now.

Despite what you believe, you can't actually twist your forearm/wrist/hand 360 degrees.

I know, I overexagerated the rotation to show the problem I described.
Even when I rotate the hand by 360 degrees the forearm barely rotates or suddenly starts going backwards. And even when I don't make any unrealistic twists there are sudden jumps.

Use a Transformation constraint instead of a Copy Rotation constraint

This works nicely, cheers.

What's your favorite language? What should a programming language approved by you look like?

Nope 8agdg.wikidot.com/gondola-the-experience
Thanks for the flattery though, Aethyr looks cool as shit.

There are 256 possible types of tiles defined by the green channel in the mask image. Each pixel in the mask image is put into an array which keeps track of each pixels x coord, y coord and rgba. The array is sorted in accordance to each pixels red channel which is used to create a draw order allowing me to choose what tiles overlap all within the mask. This is necessary because although the tiles are 4x4 pixels, I choose to draw sprites at each position that are up to 8x8 which creates things such as the nice border effect around the grass.
A second array stores the information of what each pixel means and how to handle drawing the requested tile. The stone path tile for example has entries for its requested sprite sheet, the sub image within the sprite sheet to draw with capabilities for variation and the possibility for random offset.
With all this information, the program starts drawing to a sprite in order of red channel.
Each tile also has information for how many parses the requested material needs and all tiles with the same parse on the same layer are drawn at the same time. To explain: the base sprite of the grass is an 8x8 brown texture which is drawn at every position where grass exists on the first parse for all grass on the same red channel. It then repeats for all required parses and red channels. Think of it like layering a cake, you need to put the first layer down for the entire cake, not do it segments or you might get gaps or overflow of icing.
If the requested tile is a tile that is greater than 1x1 than it requires a special case. For these I define the sub image within the sprite sheet in terms of the modulo of the tile size times two. For a 4x4 sprite (32x32pixels) this can nicely be defined as half(x coord mod 8) + twice(y coord mod 8) where the x and y coords are that of the pixel positions on the mask. This gives a number between 0 and 15 which corresponds to predetermined sprites that are drawn on the even numbered coords of the background sprite that is being drawn on.
The arrays used to define tile type and position will later be used to quickly discover the type of tile at a position at request, therefore they are not discarded. They are massive though.
There are 256 possible tile types each with 16 possible parses all being drawn onto a 480x270 resolution grid. This totals to a huge amount of possible draw calls per 1920x1080 pixel resolution room and most tile types require a call for a random number for offset variation. I'm not that good with hardware but I think most of it is being done on cpu. My i5-7600k manages to get a 480x270 resolution grid with 12 types of tiles and 8 parses each complete in about 10 seconds, increasing the number of tile types doesn't seem to increase load time that much. Although it's not too bad, 10 second load time for a background texture isn't exactly favorable.
An alternative is to have the game load the background sprites as it plays, swapping in the random tiles from the normal tiles in a discrete method. This unfortunately is pretty difficult because of the layering system. Another way of alleviating the pain would be to save the created background sprite as an image on disk to be referenced later, meaning it would only have to load once per game. Another method, the most cheesy one and kind of undermines the whole point, would be for me to save the image once completing the room and use that for release. That last couple of method are pretty shitty because I eventually want to use this system to create dynamic rooms. Saving the image prevents this or at least makes it much more difficult.
The alpha channel and the blue channel are currently unused. I feel they may be used in the future to possibly define things such as entity placement, collision detection or something else like camera movement information.

wew thats a lot of text for a simple question, I don't blame you if you don't read it.

g-guess i fixed it then, right anons?

haha time for valgrind

sorry, bro. I'm still damn torn.
Why haven't you shown anything from the gondola game recently?

Taking inspiration on Nintendo's design philosophy of deciding on a strange gimmick and rolling an entire game on it and thus having one button that literally does everything

What sounds the most fun:

* A gun that pushes you backwards when you shoot
* A gun that lets you rocket jump without taking damage
* Both
* Can't tell without a prototype

Literally 50% of all game jam entires ever made do this, on top of another 80 million other games that do it, and it has never been fun.

A mix of waning enjoyment, realization of scope and cant be bothered moving unity to my new pc. Also I had a few weeks of some important shit in my life so I had to take a break from game dev which instigated my move away from it. It's probably in hiatus hell.

your front looks decent but the side isn't like the others said
remember, you're working with 3D, try to fit the proportions and accuracy in all angels. Constantly turn around the model at various angles to see if you're getting it right.
The ear looks too big and high up (pic related)

Go to sketchfab to look at examples of heads or here
posemaniacs.com/tools/handviewer/ (3rd option from top to bottom)
And follow along tutorials and try to copy them at the best of your ability.

I think I've seen your earlier works here and you've improved quite a bit compared to your first ones. Keep going user.

also feel free to correct my redline. I'm sort of in the same boat but I wanted to give the advice I can

Flattery all around B)
Times are tough for ur ol' boy, glad to see there is someone remember me :')

Anyways-
Your tiles are really good, wizard is cool too. The tree needs quite a bit of work though, perspective is a bit weird. The bottom half is fine, but the top half is a bit wonky.

...

of course I remember… what's up, still sick? Hope the tough patch will end

yeah, i have been using tutorials this mostly(youtube.com/watch?v=I6K7x1tX9rk&list=PLvPwLecDlWRBDOl57qaLMScvH4OQX9MOW) and this just for motivation (youtube.com/watch?v=zlfRNVe1kmQ) but i don't think i can be that good.

i have very little time to model anymore so this took my way too long to make then it should. but i'm glad that i'm getting better hopefully i can get 2 models done one player, and one boss. is this any better after this i going to take a break for a day or two black Friday and thanksgiving. however one question still bugs me and that is clothes because when i was trying topology everyone was talking about the body but what about the clothes. are they modeled on top or with the body, do i have to sculpt it or no?

well at least i made some people laugh.

Those look almost exactly like tiles that came packaged with old versions of RPGMaker.

:3c

I have 3 more chapters to go with the python for scrubs book, is it normal to feel pretty stupid about python? i.e. I can't write small-ish scripts with it yet.

isn't that one of those archaic retarded languages where indentation have semantic value?

...

Well if you mean that it uses it then yes it does, it doesn't uses brackets at all. Which I have to admit I am sometimes confused from it when some of "my" scripts behaved a bit differently or it did not worked at all just to later figuring it out the indentation block was not correctly set.

it isn't remotely archaic, but yeah, scoping is indentation

what tutorial did you follow to get armature like that
also this is in blender right?

No longer sick really, just a string of unfortunate happenings. But it's all good.

I dunno why I've heard people say it looks like RPG maker. I've only seen their tiles retroactively, they aren't pixel art, and they really don't look very similar. Unless I'm not looking at the right stuff. I mean other than that the area looks bland and uninteresting, which yeah, but it's very WIP

yep, blender using riggify, it does almost all the job for you

comfy as fuck

fuck i can't stop playing with this, help

a lot of pixel stuff kinda gets lumped into RPG maker or any top down rpg game for that matter. kind of the consequence of shovelware maker 2016 existing at all.

Shit, I guess I didn't realize how garbage their tilesets looked. Yours looks way better (the ice crystals not needing to be grid aligned goes a long way).
I don't have any of the newer RPG makers, but your tiles look like what I'd expect an eventual update to their assets to look like I guess.

you are not using autotile for the ground and probably not using the cliff tiles correctly either, but yeah, the user's is way better

Yeah, that's what I was thinking, but I've gotten multiple people saying rpgmaker, so I was starting to think there was substance behind it.

Ah I see, I never saw any ice crystals when looking through rpgmaker stuff, so I can see the comparison. I know know it's not a unique idea, but I thought I had a fairly unique implantation of ice crystals. I guess it's more or less the same as rpgmaker. rip

After the Demo Day, LAN party, fixing the fucked boiler that caused the house to go cold for a week… I'm kinda burned out.

I need another quick, weekend project. Something like tic-tac wars or that flightsim I made.

I know there will be Ludum Dare next week but I'd like to have some solo/learning fun.

I'm thinking of prototypes for:
1. some sort of TB game, like modern XCOM or Valkyria clone
2. a simple first person RPG that would take place on a very small planet
3. a simple assymetrical FPS vs RTS multiplayer game
4. my original idea for aprevious LDjam game and making an endless runner kinda thing

what do you guys think? Which one should I choose?
Or maybe there are ideaguys who'd have a simple enough idea for a quickie project?

my nigga

sounds the most fun

well…. you could look at it as your game actually looks like a professional product ?

Hmm, I could use a gamejam, it's been a while and I'd like a change from my current project.
I am a programmer. Anyone up for a collab?

what u need?

at least art I guess, since I can code and anyone can be an ideas guy

You like it? I was imagining something more… strategic, I guess, when making Tic Tacs. As in placement of units and environment being important.

That would likely be some basis for the "simple asymmetrical FPS vs RTS multiplayer game"

We did one more RTS project for the 38th Ludym Dare but that wasn't 100% what we wanted even if the basic concept got some recognition. Still, good times.

I'll either be working with my usual codebro or going full retard and doing the 48 hour solo thing. Hope you can find someone to help you tho. you can go solo too. programmer graphics ftw

Well, I learned that an idea guy who can make creative concepts on the fly is pretty useful in a jam. Our ideaguy friend helped us a lot during the previous LD, we couldn't come to any consensus over what we should do without his chiming in.
Unfortunately, it seems that most idea guys are cherishing just one dream game and that's it.

itch.io/jam/lewdjam2017
they should've called it lewdum dare

u wanna make a stronghold remake in first person?

...

you have to remember that it's in good tone to keep your game related to the jam's theme and that is only given at the beginning of the jam. The common practice is to discuss what you want to make in a few hours and get going with the jam after some sleep or just like make game at once.

Question about game design, particulary about JRPGs. What makes an JRPG fun besides the story? What could be done to keep people playing just for fun?

Or at least something I could do with RPG Maker or Gamemaker.

Yeah I will consider this, sounds like an interesting experience that has the potential to kill me.

True, true. I don't mean to actually under-appreciate idea guys (it varies wildly how useful they are), but indeed especially during a jam it's great to have people to brainstorm with. Just for me personally, I would say an artist is more essential to have working with me, or visual representation is going to really minimal.

I've never played Stronghold tbh, though I think I can imagine what you mean. Would have to discuss it more, but like the post above me says it should follow the theme of the jam (if we want to be part of it). (I imagine it wil be hard to scope the idea for a jam though)

So basically you have a byte array thats interpreted as RGBA plus XY, and a second array that defines the style information and how to interpret it?

Also why are you doing random stuff? When I need random tile fluff I make a byte array of similar size, then give it a random value, then use it as a seed for effects. Its not perfect but if you have eg 5 subtile variants, you can just use rngFluff[n]%5, for example.

Does your system support "brushes"? Eg the way Starcraft map editor would let you paint tiles and have it randomized but with certain logic so edges always appeared in certain spots?

Enemy variety, spell / weapon variety.

The enemy variety I have in mind would be big, probably making the size of the game too big because of the many PNGs, but I mean stuff that can get people hooked and you could do with little resources.

Basically, something you could have people play and not just see a Let's Play video instead.

lemme tell ya, even the 72 team jams are exhausting. Still fun doe

first song that came to my mind while making that trailer thing

...

never been in a game jam but i have noting going on right now so i would be game as much as i can be for art/idea guy i'm game for whatever

i have a discord you can contact me with if you want i know they steal information (who doesn't), but its free
discord.gg/utrSNa

No autotile for these dungeon tiles, no boundaries between the stones/snow to place manually.

Will enginedev break my soul?

yes

Soul, mind and body
Even that autistic yiffe dev wanted to finish his game quickly and rushed it after 5 or 6 years developing the engine, resulting in quite a lame game despite the great tech

How the fucking what is wrong with me

feature creep is the worst enemy of the indiedev

Sometimes you need your soul to be broken. Sometimes you need to the passage of time to unrelentingly slip by you while having nothing to show for it. Sometimes when you think you've reached the lowest of lows, you only then gain the insight to the vast underscape of new lows you're destined for.

you could make your own engine, or you could just use uppdogg.

Don't use uppdogg, use Hupweth U.

Out of curiosity, how would one go about making a game for the Genesis or SNES nowadays? Would it have to be done purely in assembly code?

Hey do you guys know if there's way to graph variables over time in UE4? I'm trying to recreate oldschool movement and I just managed to make bhopping work but I've got some movement quirks that I need to figure out and I can't do it without having some visual data.

You could write Genesis stuff partially in C(excluding hot rendering loops, etc.), but for the SNES, you're probably going to be working in ASM, since the 6502 sucks ass at running C code.

Thinking of making a homebrew eh? i'm curious what your plans are.

You need to make 2 control bones for the forearm and upperarm. These do NOT deform the mesh so you need to uncheck "deform" for them. Then for each control bone on another layer you make 2 bones that go on it that blend between each other rotation. The wrist control bone copies Y rotation of the hand bone so it twists with it.

docs.unrealengine.com/latest/INT/BlueprintAPI/Rendering/Debug/DrawDebugFloatHistoryLocation/index.html

heil'd

You're awesome, thanks.

What's wrong with unreal?

Tencent. That's what's wrong with Unreal.

They're awful. I hope Godot gets more or less on Unity 5's level in 3.x. I don't see them getting anywhere close to Unreal though. At least not any time soon.

Elaborate. I'd rather hear it from you than read dilluted versions online.

Gravity altering gun.

Tencent has the power and reach that Google has, but in China. It literally powers their internet and it responsible for enforcing all the government-mandated censorship on its citizens. They pretty much have a monopoly on cell phones and such, too.

Being Chinese, they are all too happy to take western games and repaint them (if you're lucky) and rebrand them as a superior Chinese product that invented the genre first, creating anti-American sentiments amongst the young Chinese gamer demographic (eg., "America is such a ripoff, copying our games, lol")

They're also major stakeholders in Vivendi and Actibliz, which means a lot of the bullshit that's getting pushed in their games (eg lootboxes and treadmilling) is done to satisfy their investors. They also purchase land on the west coast under shell companies to fuck with our housing market.

And remember! Every sale of a game you publish with Unreal feeds into their beast

china

Yeah thats pretty much exactly it. I don't know why that took me 600 words to write.

Because a quick and dirty implementation of random tile fluff was all I was aiming for. Thanks for the tip, I'll implement that later.

I've never used the starcraft map editor so I dont really know what you're talking about. The layering system I use allows me to pick different sprites and variations per parse meaning that a single tile can have elements of variation and static points. My implementation of tiles larger than 1x1 allows me to create patterns that appear based on world coordinates rather than per tile. Later, if required, I will implement systems that create borders around tiles so I can make things like cliffs easier.

I rewrote the raytracer with entirely my code, and now it can render trivial stuff correctly.

user, frankly speaking all markets are in one way or another being tapped into by some huge corporation. We're living in a proto-cyberpunk era, and this is the harsh reality that we have to face. I'm being given an incredibly handy, user-friendly and retardproof engine that is constantly updated and improved. I then can publish that game anywhere I want, and I will only have to pay 5% of all income to Epic, and only after I make the first 3000$ dollars.

In the perfect world, you wouldn't have to pay anybody anything, but this is not a perfect world. Only having to give them 1/20th of what I make sounds ridiculously good in this day and age.

renderslice does appear to be a little confused being called multiple times on one framebuffer

Okay guys I'm having some trouble with my movement so I'm going to ask a silly question. Do I add acceleration vector to the velocity to apply it, or does some other math operation take place there? I'm trying to replicate oldschool movement mechanics in Unreal and it's behaving a little fucky, but right for the most part.

...

Listen up here you fucking nigger where the hell did I say to vote with your wallet, or mentioned the invisible hand? We're consumers. We eat shit coming out of these corporations' asses daily. We're fucking powerless.

I want to make games. I don't give a fuck who takes the fee that I need to pay to be able to do so. I'm not going to hold back from using an industry standard just so I can pat myself on the back for not giving chinks money. I don't care about the world economy, politics, the chinks getting money or whatever else. They're going to be shoveling that dosh no matter what.

The best thing the normal people can do is to steer themselves in the torrent of diarrhea coming down on us from these corporations in a desired direction.

blenderfag here, I can work really fast within my zone of comfort (buildings and clothing), I can animate (badly) and know how to code a little so i could do non crucial stuff like gui and trigger events.
this is an open offer for anyone here applying for the ludum dare and not only for . I should warn you that I often end up going AWOL.
going to sleep for now but here is my email
ostoru [at] cock.li

Okay, then consume something else? This is exactly what I'm talking about.

I said before, I'm not going to pick an inferior engine just because I have to pay 5% to some corporation.

pathetic.

I'd like to release within this century, thank you very much.

code a game about underground battleships tbh

Made a pushable ice block, thoughts on how it looks?

freeze water tbh

Working with placeholders speeds up development so much but damn everything looks ugly

Added some GUI and made the transition between flying/mech have a bit more flow, as well as adding more mobility options for the mech (JumpJets, as well as forward+side boost jumping that fully overheats the jets)

I'd have to look at it in context of the whole game, but frankly speaking it looks pretty slick as is.

what kinda mechs are you going to have?

I don't really have any right now. I've had it in the back of my mind ever since I found out about the NES Assembler the Retro City Rampage guy made, but I don't actually have any ideas for a homebrew right now, nor do I have the skill to create an engine in an antiquated language that nobody uses anymore.

Any anons here familiar with font editing? I've got something I need if you're feeling philanthropic - a font conversion. UE4 does not like bitmap fonts, so I need to be able to import this standard old Windows 1995 font as .ttf, fixed at a particular size to get that crispy pixely look.

what is fs Tahoma 8px?

This is MS Sans Serif at 14px. Actually I'm not entirely sure if the image has it at 14px, but my concept does. I tried other fonts but they just don't carry that technical oomph very well.

I'll see what I can do using FontCreator. I'll give you a tutorial on how to do it once I figure out a method for what you need.

Here's some (very outdated) mech transformation footage of a game I worked on. We showcased the game at TGS and got positive reactions. Second video is what the art looked like last time I saw it.
God I hope I'm not breaching my NDA this way, it's from the tutorial and a public youtube video so I think I'm safe.
Funnily enough we also had (omnidirectional) boosting, jumpjets, and flying transformations. Sadly some design choices, a few bad team members and some other unfortunate circumstances means the project died out.

Well that first video sure came out wrong. It starts after 8 seconds.

looked like it had potential

That would be absolutely fucking amazing, user. Thank you.

It got a lot better slightly before that second webm was made, but at that point I was the only programmer left and had a giant bloated terrible codebase, and we couldn't find another programmer willing to work with us so redoing the code from the ground up wasn't feasible with our allotted time. Wound up having to jump ship, and then the project slowly died being just artists polishing stuff largely for their portfolio, and the lead designer.
At least we had fun, and I think that lead designer is currently working in Japan which was his dream, which he managed to do partially because of this project. So it wasn't all for naught.

why was the code base a mess?

aw, fuck. I'll only be able to help ya in… 10 hours, maybe? Sorry about that, I don't have FontCreator on this computer and don't have electricity back at home.

But here's what I wanted to attempt, maybe you can try on your own:
1. Open the .ttf version of MS Sans Serif in FontCreator. You'll have a window with all the faces available in the font. Open the one you want to edit.
2. Open paint and write the letter you need at the size you'd like it to be 8 px is more pleasing to me, personally
3. Select and copy the raster image of the letter, paste the image inside FontCreator editor.
4. FontCreator will then turn the raster image into vectors. Resize the image so that it overlays the already existing letter.
5. Delete the original vector image of the letter, leaving only your pixellated version
6. Repeat the process for all the letters you need, until satisfied.

It's a painstaking process, but I can't really see any other way of doing this and this is why I wanted to prepare a tutorial instead of doing it for you

Thanks! This is enough. I'll get it done and post the results here.

5 programmer tech team, with not a whole lot of experience with the engine when we started, working on something for a year and never having the time to refactor anything. And at least one of them was just so incredibly fucking incompetent he became a meme after he left. Same thing for one of the artists.

well, I don't want you to waste your time so post results for the first few letters as soon as they're done

seems like a complete waste of time to work with something the team don't really know alot about

I will! I'm going to bed pretty soon so I'll have to hold off until tomorrow, though.

sure. hope the results will be satisfactory. If not, we'll try something different

Yes, college tends to be a waste of time.

certainly was for me

don't ever take a video game design course kiddos

but how will I know how to design games then, user? How will I able to challenge the archetypical male power fantasy and create meaningful interactive experiences if I'm not taught to?

by wearing girls cloths and taking T blockers tbh

It's less about learning how to do shit (I learned most of that during the free team projects, not classes), but the piece of paper it gives you can be important if you pick the right college. Mine happened to be held in high regard in the industry and had a lot of people hired straight out of their graduation or even during their internships, so I put up with it. But most of them are a fucking joke.
Of course, most people here just gamdev as a hobby so don't even bother thinking about taking such courses, it's completely useless if you don't want to wageslave as a dev.

I've always wanted to make a game, shit job gets in the way though. How do you nigs find the time?

be NEET masterrace

I'm a NEET. I'm moving in with my girlfriend soon and she's got a job. I told her I wanted to get a job but she insisted that she wants me to stay at home and work on my game because she wants me to do what I love and believes I can make it big.

neet mind

you better not squander that trust or you'll be pretty fucked.

the wageslaves among us just gotta use every available free minute on devving.

No way I'd ever do that. I've been working on this shit like it's the cure for cancer when living under my mother's roof - there's no way I'll allow myself to be a lazy fuck once I move in with her. I mean, I thought I'd be a fucking deadbeat wage slave when I decided not to go to college - this is a one in a million lucky circumstance.

I spend less time on devving now compared to when I was working. Granted, the job wasn't really exhausting.

Thanks for the help, but either something is missing in your advice and tutorials I saw or I'm simply retarded. It works fine enough in some poses (like in your webm), but there's still a lot of this jumping even when I try to do very simple poses. The worst thing is even though this problem occurs in all methods I tried so far, I have no idea how to call that shit in a search engine

yeah, a routine is good. Every time I have some vacation time I tell myself it will be time for devving.
Fuck nah. You gotta go to the garden, visit family, go on a trip, invite friends and other such shit.

you ooze normalfag from all your pores

that's what I am, unfortunately. Had to leave the NEET life behind but the other option was tfwnogf, poverty, alcoholism and death. I'm glad you still live the good life, my friend. Get the most out of it. Hope you won't end up like me.

wew

more like an excuse for not making progress

great potential on the ideas, for a different kind of game though. I was expecting to eventually make a megaman legends clone where you have actually good movement

Have you tried drivers?

I just bought this movie on steam and want to share with the internet, but Gaben is a fucking jew and I can't download it. What's the best method to fuck valve? Download OBS and capture it?

seems so

Dayum, it's gonna take some time doing this.

MFW

But how come no jump scare at the end? Or was the lack of a jump scare a jump scare?

No idea. Whatever artist I end up with will ultimately decide, but something more along the lines of gundams than battletech. Just can't find a good model with that design that's free.

That's kinda sad. I'd like to keep this a small scale project for now to avoid that kind of stuff happening. Mostly I'm interested in solidifying a core gameplay loop, and then seeing what makes sense to add from there.

Nah probably his four watt setting is too high.
He needs to turn it down.

i got my doom-style sprites and local avoidance working
anyone got sprites i can use as placeholders?

spriters-resource.com/
have fun

Finally getting stable again but I'm getting more done than ever by sheer force of wanting to rub it in her now-indebted whore face. Spite is one hell of a motivator.

Can you make an RPG with it? or a loot system? or a location shooting system?

did you use stencils? or some other method?

From the looks of it, just billboarded quads with sphere/capsule colliders or character controllers.

No, I tried to avoid using them in this case, as it would probably require math that is beyond me. For now I'll satisfy myself with doing it by hand, if it proves too bothersome to use I think I'll try drivers.

Lighting!

That's the thing I guess - my girlfriend never even once suggested the possibility of failure. She's the kind of person to congratulate you for getting up at the right time. She just has that maternal streak about her.

Are you actually going to make a game with raytracing?

I really want to!

How are you getting 60 fps in your raytracer? How did you go about writing it? I did one with my own (slow) vector math and SFML for blitting, although I only did a few shapes and basic materials/lighting

sorry to hear that, dude. Hope one day you'll find a girl like
I always mention my wife in these threads- there are girls who can appreciate and support devs.

godspeed. That shit has never been done properly from what I know. Maybe one day this tech will be like the voxels of old (in games like Delta Force or Outcast). Very rare but things that use it ase very special.

I wonder how the lighting works on those, if at all

I'm happy to hear that other people have found success.
Literally every single girl I've been with has hated my dev habits and tried to get me out of them. Usually it's what I ended up breaking up with them over.

Can't be helped, really. In the end, doing what I do takes higher priority than sex and titties.

The trick is to not date normalfags. My girlfriend is old 4chan blood.

How do you find non-normalfag gfs? Just by pure luck or did you actually meet up over 4chan?

She either has posted nudes on the internet, or has no respect for authority. Either way, I'd be wary.

Met up after talking for a while on a completely different website.

She doesn't have any respect for authority. Neither do I, really.

Apparently I am now writing Battle City type game, hence this horrible level editor. I think that the tank waves will be generated randomly, and right now I am kinda lazy about spawn points, and since Godot works with JSON, I will probably just add them by the hand. So yeah, now I only have to implement saving, and editor is done!

I wouldn't really call this "doing it right", since the lighting is kind of complete hackery.

That's not at all what I asked.

Sorry, please forgive me! I was going to reply to your post too, but truth be told, I'm not sure why it's "fast". It's multi-threaded, and uses glm for vector math, with everything at float precision, and I really try to avoid division/sqrt(though you can't in sphere intersection), I'm going to attribute it mostly though to me only shooting 1 ray per pixel(really hacky).
Technically means it's a raycaster though.

Mine was a raycaster too technically, though not multithreaded. I thought I should go for a faster intersection test before I added shadow rays or reflection rays. Are you casting every frame? Good luck with it, I love how they look

Both not true and not hacky. You shoot one per pixel, and then one per light source from the location that first trace hit. Classic Whitted-style raytracing. Just because you only have one type of material implemented does not make it hacky. Merely unfinished.
Wrong, raycasters only do 2D collision and then draw walls based on those collisions. You're probably thinkinf of the difference between raytracing (send rays from pixels into the world) and pathtracing (spawn a lot of rays at light sources, have them bounce around until they die based on chance or by hitting the camera).

I've read the distinction between raytracing and raycasting is that casters only do one intersection test and tracers check secondary rays like shadow or reflection before coloring

I am, and thank you!

I swear that's what I read from wikipedia.
Yeah, the materials don't reflect yet.

Nope, raycasting is only one ray per vertical line of the screen. Wolfenstein is an example of this, as is this video some other user linked to a few threads ago.

Not having additional bounces/reflecctions per pixel yet just means you haven't implemented other materials, what you have is a perfectly diffuse material. Adding normal randomisation would add a coarseness variable, and for blinn-phong you'd use two bounces and decide how much of each result to use based on the angle of the trace hit. But saying that not having a blinn-phong material means you're not raytracing is like saying the same about not having a refracting glass shader; it means you're missing common features, not that you're doing something else.

That's very interesting, thanks

Here's my other shit so far

class MultiTriObject : public Intersectable {public: MultiTriObject(std::vector tris) : tris(tris) { inter.reserve(tris.size()); } glm::vec3 getNormal(Ray &r, float &t) { return intersect(r).normal; } Intersection intersect(Ray &r) { inter.clear(); for(Triangle tri: tris) { Intersection i = tri.intersect(r); if(i.intersected) { inter.push_back(i); } } std::sort(inter.begin(), inter.end(), [](Intersection a, Intersection b){ return b.t > a.t; }); return inter[0]; }private: std::vector tris; std::vector inter;};
This isn't thread safe.
I need to figure out another way to do this, my current guess is register MultiTriObjects as Triangles in my raytracer, and give a Handle for every Object, which will """link""" to the Triangle Objects in the Object array.

I think your style would benefit from going smaller. Think 16x16. It's also a good idea to restrict yourself to a palette of 16 or so colors. You'll feel more comfortable within the limitations, and things will automatically look better as well. I used to do a lot of texture pack stuff for Minecraft back in the day.

I'm going to do it, Holla Forums
I'm going to hardcode my game objects in a strict inheritance chain instead of ECS or node/scene or system-based shit

...

Just use TGA, it's a much nicer format to write a parser for. Just don't support compressed TGA.

You can also write a simple BMP parser if you only support power-of-two textures, in BGR format.

Code related is a little retarded because I wrote it in 2015, but it works.

//load a 24-bit bitmap image as an openGL textureGLuint loadBMP(char *filename,int w,int h){ GLuint texture; int *data = malloc((w*h*3)*sizeof(int)); FILE *file = fopen(filename,"rb"); fseek(file,54,SEEK_SET); fread(data,w*h*3,1,file ); fclose(file); glGenTextures(1,&texture); glBindTexture( GL_TEXTURE_2D, texture ); glTexEnvf( GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_MODULATE ); glTexParameterf( GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER,GL_NEAREST ); glTexParameterf( GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST ); glTexParameterf( GL_TEXTURE_2D, GL_TEXTURE_WRAP_S,GL_REPEAT); glTexParameterf( GL_TEXTURE_2D, GL_TEXTURE_WRAP_T,GL_REPEAT); gluBuild2DMipmaps( GL_TEXTURE_2D, 3, w, h,GL_BGR, GL_UNSIGNED_BYTE, data ); free(data); return texture;}//load a 32-bit Truevision TGA image as an openGL textureGLuint loadTGA(char *filename, int w, int h) { GLuint texture; int *data = malloc((w*h*4)*sizeof(int)); FILE *file = fopen(filename, "rb"); fseek(file, 18, SEEK_SET); fread(data, w*h*4, 1, file); fclose(file); glGenTextures(1, &texture); glBindTexture(GL_TEXTURE_2D, texture); glTexEnvf(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_MODULATE); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT); gluBuild2DMipmaps(GL_TEXTURE_2D, 4, w, h, GL_BGRA, GL_UNSIGNED_BYTE, data); free(data); return texture;}

a slight battletech gundam has been done alot tbh

Y'all bad boys like programmer art? What about programmer 3d modeling? How about programmer animation?

Making a horror game I see.

...

Nah, this is just a placeholder for the game I'm working on. I've noticed that I have a perfectionist streak, so I've intentionally got an ugly ass model from makehuman with a lowpoly topology, and I'm going to be rushing a playable alpha state with shitty assets instead of trying to make it look pretty for now.

Alright, we put up a page for our game on the wiki
8agdg.wikidot.com/die-totenmaske

It's not an amazing game but it is a start and we're still proud of what we did.

If we can do it so can you!

Congratulations. Even if it's not amazing, each game means you've learned more and can do more.

Thanks, we're going to continue working on version 2.0 eventually.

First though we're gonna go back to working on Teagan Quest.

Oh nevermind, it did save correctly

Any good programming guides for retards? Tried reading through SICP but my acoustic brain can't process chapter 1.1.8 onwards.

Accept the fact you weren't made to be a programmer and move.

make a video game about making video games which make video games tbh

Programming books are a meme.

Read through multiple online tutorials until you start to understand. More or less simultaneously you should try to make something, because that'll make you ask the right questions.

Start with something simple where you can easily see results without too much groundwork, i.e. not C/C++.

I'm working on a "survival" sandbox game. Survival is really just a label for these types of games, in reality they're just resource management sandbox games with an everlasting gather/utilize cycle (one of the reasons why they're so fucking relaxing).

What are the main issues with this genre? I'm going to use a transvoxel algorithm to allow for a high level of precision with the sandbox bit, but the rest of the gameplay is still up in the air. What are the golden bits, and what are the things to avoid? I've got my own list of do's and dont's, but frankly speaking I'd like to hear what the folks around here will say.

I love the actual zen feeling of playing these games, but I can't help but feel that the entire fucking genre is cursed.

700 femtoseconds in Adobe After Effects.

one massive issue with surbibal sandbox games is that they're usually very simple mechanically. you get a crafting system which gets like 8 or so meterials three starting, two middle, one late and one useless that builds everything in the game which is some really awful looking cookie cutter object/item/weapon/clothing and then nothing to really do anything with those things with bar expand till you uninstall. multiplayer survival sandbox games adds a whole other different kinda cancer to the game because market research says it allows devs to ship a empty game since "players will be the content of our game"

what you want to do is make sure you have a base game with solid mechanics and then add all the "survival" sandbox shit along side it and if you add crafting to the game make it as autistic as possible with different tech trees for different things you can craft into or research that do way different things for the player so that if you do have multiplayer there'd be reason to have lots of people doing different things rather then having 20 or so wooden houses all grinding up the same skills because multiplayer was an after thought, also on the note of multiplayer make sure you have some basic measures of protection for players and their shit, like in eve online not only does it take awhile to get into null sec and find players pos' but it takes two days to destory them and even longer to remove them from their space entirely.

I suggest you look at various mods which completely overhaul their games. Tremor mod for terraria, thaumcraft mod for minecraft and bob/angel mods for factorio.

speaking about game play though, I'll have to know more about what you want to do with your game

do. not. delegate. mechanics. you. can. do. in. the. environment. or. with. characters. to. the. fucking. UI.

Sounds like you got memed by Holla Forums
Start with an online tutorial for a simple language like Java or maybe even Python, do as you go and try to have fun with it, it'll come naturally from then on.

I'll try it out when I can, user.

Any of you fags is a 2D artfag?

something optimization

My normals seem a little bit retarded.

...

I'm paying thread initialization costs 4 times every frame.

where can i find a codebro for free i don't wanna bother with programming

Post your art so codebros become interested

i can't do art only low quality models

>>>/ideaguy/
we don't need you

git gud

Okay..
i'll come back in 5 years when i'm good enough

You should learn to program too, so you're not dependent on any codebro. If you don't, you'll never wind up making exactly the game that you want to, since programmers will always have the final say. That's what happened to Kojima, which is why he learned to code.

post your models, bro. It may be salvagable. And there IS a team for the next Ludum Dare being completed

Show your stuff. Also, you won't need 5 years to become good enough for /agdg/, if you put in the effort.

I implemented a work system and now I'm spending most of my time in the mutexes.
Great.

I'm another programmer who could help out with this. On top of programming I can model hard surface objects like weapons, furniture, tools, and so on. I can kinda do 3d animation but I'm slow and bad at it. I am capable of doing sound design but it's really not my strong suit.
I'm most familiar with Unreal Engine.
Are we going to have a discord/trello? We should set that up soon along with a gitlab/gitgud repository. Ya'll know how to use git, right?

The moral is never multithread

Don't multi-thread a friggin' raytracer?

I'm bored and hate myself, and would love to contribute my lack of expertise.

When I wrote a simple software raytracer a couple years ago I just used fucking SDL threads, creating them every frame, which was retarded, but for some reason worked pretty well. I later wrote a much better one in GLSL, but that wasn't as fun, as the most challenging thing was coming up with a good way to pack BVH trees into a GL texture.
You should read up on work queues and scheduling, maybe you'll find shit that suits you best among all of the algorithms out there. You could also try using OpenMP and SIMD shit to go really fast.

As far as gameplay, I'm trying to take best of both worlds from Minecraft and Rust. I've played Rust quite a bit and there was a period of time when it was good - but then I watched the devs fucking ruin it update after update. It had a feeling to it that I would like to recreate, that I've never really seen a game produce. I think the one thing that made me extremely into Rust was the building. Not the building system itself, mind you - it was pretty rudimentary and still is - but the concept of using various building layouts to perfectly utilize your surroundings in a way that gives you an advantage over other players. That gives you a shelter where you know you will be more or less safe, because you've put effort and dedication into making sure your building is secure.

Picrelated. Rust has monuments that players USED to be able to build on. This provided them with a unique advantage, since unlike player structures, monuments were indestructible. It took a lot of trial and error to be able to build a base that used a given monument 100%.

Minecraft, on the other hand, provided me with Redstone. I think I've spent thousands of hours tinkering with it and coming up with mechanisms that would let me do various cool stuff.

I've got a GDD that outlines most of the PvP, stylistic choices and other generic shit, but the sandbox stuff has thus far been boiled down to high concepts. I've been mostly exploring what makes other games click and how to possibly expand on those things. I've included an example of my improvements to the Rust building system, something I'm going to be working on after I get the UI out of the way.

whoops forgot to add the building concept

inspiration

Oh God is rottenredditor here?

33 fps with a threadpool.

Yes, I've been doing /agdg/ for a long time. Don't derail the thread, just shit on my steam page instead.

I know you have, for as long as you've been shilling overwatch. I was the one who found your reddit account.

I never shilled Overwatch. I said it wasn't necessarily a terrible game, but I never outright shilled it.

Sure, that's why you were posting your own gameplay videos. I know how you write, I know which posts you made in the Overwatch threads. It doesn't matter, good luck with your game and your shit taste.

You aren't using threads for something like pixels or individual traces right? In raytracers, you want threads to work on sections of the screen, raytracers are "embarrassingly parallel" as it's called, it's really hard to not gain almost a multiplicative speedup per thread if you're doing it right.

looks like it's fun to code, and probably serves some good practice.
if you're serious about raytracing performance, then do it on the gpu using compute shaders or better yet openCL, but even a simple naive fragment raytracing shader would get u 60fps with your scenes so far easily.

don't use the gpu then :^)

raceconditions and deadlocks are bad, but avoid the possibility of them in the first place; so then u avoid using locks n such… as you've seen they can cause quite significant performance hiccups.

Can you be a little less edgy? I've talked about OW because I play it, and I posted gameplay vids because I had them. Boy do I fucking wish I was being paid to shill OW, at least then I'd have money to pay artists for assets. That being said, though

Thanks.

I love when idiots unmask themselves
don't take it personal, is very obvious you lack social skills, just push through and own it

My RX480 that I brought with me doesn't seem to… work. I'll open my desktop up and reseat it in a second, hope it actually does something.

The threadqueues avoid locking afaik.

Calls to RenderSlice draw from yfirst to ylast, which at 720p(current res) draws 720/4 vertical pixels.

please no bully for the extra space

I'm not sure what you find edgy about that statement.

To be fair, most people here lack social skills myself included. that's why i cut down on socializing, to avoid it backfiring in the future, but it might anyway, since there's cringey stuff out there that i can't delete.

The fact that you think you can tell who is who simply by the way they type. That's some graphology tier shit. I understand if you just called me out as rottenhuman for liking Rust, but saying you did so because you somehow recognized my writing comes off as pretentious.

it's dead.

not really edgy at all
i can identify regulars here from their grammatical habits, and so could anyone who lurks enough

>The threadqueues avoid locking afaik.
ah, so u use a workqueue for your threadpool

i bet most here consider themselves to have "bad social skills", but the reality is, is that most don't have anything to talk about with nodev+novidya people that feels like a worthwhile use of time; so there's no effort made to do so well and thus an image of having "bad social skills" forms due to a socially propagated image of lacking what is normal conversations (e.g. talking about trivial and uninspired bs like events, or people).
I'm sure most here have wonderful social skills in the correct setting, and could formulate their ideas, perspective, and input/retorts in a highly cohesive manner.
Though, in reality, what is considered "good social skills" is just a matter of being adept in jew speak (fluent in lying/saying anything to please another in a way they want to hear).
Which myself, and many others intentionally avoid due to various reasons; however, it is a useful identifier for people who are pieces of garbage.

It's not pretention. I've argued with you in overwatch threads, I've read very many of your reddit posts, I've seen you in this thread several times. You're not hard to recognize by your writing and methods of argument.

500IQ Holla Forumsack logic

Now I'm just interested - what gives me away? What are the methods of argument I use?

to that note, why not instead of a human survival sandbox make it into the insect world survival sandbox?

kinda like the Insects Infestation mod for half life 2 cept merged with fort source and some natural selection in

you could have it so that those monuments provide more then just shelter and something like area control that spawns strong npcs for that "hive" which can only be gotten if some kinda resource intensive thing is built in a certain spot that needs a resource supplied to it once a month or something to keep it functional and around that area there's certain resources that only spawn with a controlled area of which you can build up to produce more or different resources to use.

you can have some kinda evolution tree for the players of which needs those various resources to progress in which could dictate how the player operates like you could have it with afew different evolution trees which the player can switch to once a month or something

these evolution trees could be
and then into more focused roles

also I'm

Look, stop this now before you regret it.

You too, I don't know what the fuck is going on but just talk about making shitty 2d shooters and platformers.

I don't care to interact with you further. You have access to your own writing, examine it yourself. The way you brush off and then attempt to earnestly engage with your detractors is a distinct and relevant trend, you exhibited it in the thread you were outed for example.

could you fucks not drive away another fucking developer, what the fuck is up with almost every autistic fuckers on this site I swear that this is the reason why the site is dying tbh

You're acting autistic.

Yeah, fine. Vidya from here on out.

I think I'd be down with this if I wasn't already invested in the idea of beating another player's face in with a blunt object. There's a whole lot of very specific things I wanted to be a part of the gameplay. I mean, there's not a lot of real reasons why I wouldn't want to make an insect-themed survival game, but I'm just not really feeling it, you know? I also don't like bugs much.

You asked; I answered

Have anyone ever heard about this Orlygift?
I received this email a bit after setting up my steam (comming soon) page

in godot:
how do I get an object position relative to the camera plane

I found an awkward girl around my age in the wild today but was too beta to ask her out.

I mean yeah, fair enough. But at the same time the way you did it had me cringing. You just come off as a pretentious person who always thinks he's the smartest guy in the room.

well if you don't make it stand out from all the other survival sandbox games, your game is going to be lumped in with them and if that happens it's just going to be over looked due to the fact that that fad was stillbirth basically

And you come off as a buzzword dependant redditor without an ounce of self awareness. Feel free to keep it "vidya from here on out"

I think I would rather make it stand out by having rich mechanics and fun, long-lasting gameplay than go the novelty route. I feel pretty safe knowing that if I have doubts I can just come here, drop a demo and talk it out with other anons until make the right decision. From what I've seen so far, every single survival game out there fails eventually, due to the developer failing to listen to the userbase more than anything else. Rust, Minecraft, DayZ, you fucking name it. They're so far up their own ass with their concept and the idea of their game (rather than the actual game they have on hand) that they ruin it with every update beyond that initial release. I think just knowing what the game actually needs vs what I want to add would be novel enough to stand out from the pile of dead sandbox games.

meant to type "universally good social skills", it was ever so slightly implied, but alas you lack the IQ to discern such an implication :^)

oh shit you got me

SICP is a wonderful book, but most people can't jump right into it. Start by learning Python first, it's a lot more beginner friendly.

The main reason most games in the early-access survival genre suck are because they're obsessed with realism, which means they'll never be truly complete. Long-term success of an early-access survival game only happens when you're willing to abandon realism; "low-fantasy" early-access survival games (ARK, DayZ, Rust) tend to do worse long-term than "high-fantasy" early-access survival games (Don't Starve, Minecraft, Factorio).
I'm not sure whether that above paragraph is an actual insight or me just explaining my own biases though.

Perhaps it's your choice to respond to an observation that the vast majority of normalfag social interaction is conducted with the goal of signalling social status and maintaning appearances instead of communicating meaningful and important information with "500IQ Holla Forumsack logic."

That's actually a really interesting observation. Maybe it's the willingness to commit to game balance instead of grounding everything in reality? Can you think of any examples of low-fantasy and high-fantasy games dealing with same problems differently? Maybe there's some wisdom there.

you can do both but I guess I'm really talking about the things which has kept me from completely dropping a game like this and those things are the look of the game, if I'm playing with friends and what I can do in that game.

maybe what you could have is a sort of biological machine theme much like genesis rising if you want humans and are finding bugs interesting for your game, your world could be one of flesh robotics where various bug like suits biological exoskeleton and mechs are what's crafted by the player to survive then?

I'm going for a semi-lovecraftian feel, like Stalker did. Things that look ordinary but on further observation aren't. Uncanny shit. Diving deep and finding creepy looking fish. Digging too deep and stumbling into a cave filled with fractals and lovecraftian monsters. Finding a field of gravitational anomalies. I've got some pretty abstract lore written to justify this but it's far from being complete.

I've got a pretty good idea of what I want the result to look like, and I think it'd be worthwhile enough to stand out by itself, but I'm still in the process of finding the method to this whole thing.

ah alright

You could say it's a comparison between games that focus on building around their core mechanic, and games that are split designing around multiple "core" mechanics.
These can be split into design philosophies which are "focused vs unfocused".
Following this logic it can be said that high fantasy in a survival setting allots for more design freedom (the ability to wisk away the design constraints of a realistic low-fantasy), and thus they're, potentially, more focused.
F.e. rust initially fell into the former category of being focused, but has strayed into the territory of being unfocused mechanically.
This reminds me of a talk done by a nintendo dev, and him mentioning that they utilize a related design philosophy, which in short, is creating/perfecting a fun core mechanic, and building up from there (f.e. splatoon, mario, LoZ, etc).

sounds like a neat concept tbh

instead of crafting being the primary mechanic you could have extracting spooky mcguffin shit be it and you could have it be a situation like in stalker where there's lots of factions of various backgrounds staking a claim. the player before spawn joins a faction he likes after character creation which puts him in the out skirts of this not the zone but when the player gets closer to the center shit gets much tougher which in turns hopes to make cooperation more important.

...

oh and I forgot to say that the player can make sub factions within the not zone but only claim land in the worst area possible

you could also have player buildable not zombie emission towers which are designed to protect areas from people not in a players subfaction but they have limited range and are slow in not zombifying players

how it could work is a player needs the right algorithm of frequencies to block using a personal signal blocker which the subfaction who owns said not zombie emission towers gives out to it's members and if you wish to invade the place it's protecting you need to set up a frequency scrambler which has to "calculate" for 2 to 5 hours in one or two days

It's human and everyone remotely successful does it to some degree. Jews just perfected the craft.

I don't know about 'oldschool movement', but yes acceleration is applied to velocity, which is of course applied to position.

wew, I've seen your project at igad. I gotta be honest, I thought your project looked like shit from the moment I saw it. I think especially your (lead) designer seemed like a typical weeb with too high ambitions trying to satsify his own anime fetish. I was really surprised you guys got to show at that jap convention. Anyway, I don't mean to hate too hard, it was an ambitious project, especially when you have to work with a new engine, but it just looked very clunky overall. It could have had potential perhaps, but it went for publicity way too soon I think, it needed to be more polished for that. I hope you don't take it personal (but it looks like you are aware of the problems you guys had), and of course I don't know specifically who you are (or even who exactly were on the team I am very curious now though which one of you is an 8chner to be honest. I knew one of the programmers personally a bit, he joined the project mid-way, but I don't know for how long he worked on it.
I agree igad is (was?) pretty good for game dev education, especially for artists and programmers. Can't vouch for the v2 system, but it looked like it was going downhill when I was still there.

I disagree. I'm not a big reader but "Beginning C++ Game Programming" was pretty helpful to me when I just started out. Later on, "Effective Modern C++" gave some really good insights. Actually, that's all I've read so far related to programming (other than some random pages of "The C+ Programming Language"), but someone recommended me "The Pragmatic Programmer" so I might read that some day.

I went into the discord this other guy linked to discuss the jam. I don't know if he minds you joining but I suppose it should be fine?

The invite has expired.

oh right ok, i'll get back to you or that guy himself might. I need to sleep.

No, I get that. The stupid question I'm asking is - when applying acceleration to velocity, do I just add two vectors? Like… Just add them? I know it's fucking stupid and of course I should just add them but I've been trying to understand what's causing this bug for over a day and I'm losing my fucking mind and I'm just questioning my own sanity at this point.

discord.gg/uMEgCG
^ this one?

In XNA it would be:

Vector2 position, velocity, acceleration;public void Update(float time){ velocity += acceleration; position += velocity}

And of course, multiplying the delta by the time step

Thanks. Yeah, what I've got is definitely right. Weird.

Proof that UE4 is the one true engine. It holds the right things in it's memory.

Thanks.

(checked)
I don't think that's it honestly. DayZ was very much "focused" on creating a realistic zombie survival sim, they just went into developer hell because making a realistic survival sim is really hard and they never actually fixed a ton of the bugs that plagued it because they were too busy adding new features for realism. In comparison, Minecraft doesn't really have any focus; that's part of the reason Minecraft modding is so wildly successful imo, because there's no real goal or core mechanic that a mod could conflict with, so they have a lot of freedom to do whatever they want with their mods.

Perhaps the high-fantasy/low-fantasy split has to do with the goal of the gameplay? Don't Starve is essentially a top-down action game masquerading as a survival game; Factorio is a survival sim at first, but eventually becomes a management sim and engineering puzzle game; Minecraft is virtual Legos; and 2D Minecraft Terraria is a metroidvania with crafting, fort-building, and a day/night cycle. In all of them, survival is secondary to the actual goals of the game itself. On the other hand, in Rust, DayZ, and ARK, "survive" is the only real goal, so instead of making the mining/crafting/fortifying/exploring/combat aspects just mechanics used to reach the goal or as a conceit to mask what the goals actually are, that is considered the entirety of the game and actually having fun is left to the players (see: "emergent gameplay" bullshit). I think this explains why No Man's Sky failed so massively (other than that the hype train was unsustainable). When you look at it through the lens of a low-fantasy early-access survival game instead of a space sim, all the survival sim mechanics were incredibly bare-bones or nonexistent, the single-player goals were incredibly boring and underwhelming (compare e.g. Don't Starve's plot to NMS's Atlas plot and the galaxy center), there was nothing interesting to explore (compare e.g. Don't Starve's biomes and wildlife to NMS's "quintillion planets of shit") and it didn't even have "emergent gameplay" because the multiplayer was a fucking lie (see: "My mind is blown").
Actually, I'm quite liking this theory, because now that I realize high-fantasy includes Minecraft and Minecraft: 2D Terraria, both of which I despise, it no longer just matches my personal beliefs.

So what's actually going wrong?

running into walls causes rapid acceleration alongside the wall. Changing directions midair also accelerates more than it should.

test for collisions in the move direction and only apply the full force when is free, do the same for air.
incoming pseudocode
keep a different variable for cur_speed_var and increase it every time your input movement matches your current_direction hint: "dot_product()", check cur_speed_var against max_speed_var and clip the excess.
once you have your (inputaxis.normalized() * cur_speed_var) variable apply movement one axis first then the other. also use different max_speed_var values for air and ground.

Thanks, I'll try this.

I'm recalling it from memory, so i can't guarantee it to work, but should point you in the general direction of fluid yet tight movement

Yeah, I'm just looking over Source movement code right now and slowly grafting it onto Unreal CharacterMovement. It's gonna take some time until it really feels like source, although bhopping already works.

Do these seem "neon" enough?

Here's how it looks in action, I guess

I can imagine them being neons, but only because you told me. Light sources are mainly just shapes of pure color, what indicates they're source of light is their effect on space around them. You either should introduce them somewhere in the environment so the player can see the light they give off and remember that these are light sources (and neons at that) or add some symbolic effect associated with light. If you could use a little more space per a sprite you could add glow around them or if that's not an option, use a shader that will do this.

You should work on the way you color your neon. You probably used multiply or something like that but that makes whites into the color you applied which darkens the whole neon. In reality the "inside line" of a neon is very bright, almost white, slapping multiply on won't do.

That uniformly thick "glow" border doesn't look good. Look at anti-aliasing chapter if you have that pixel art pdfs.

Okay but this kills the bhops. Basically I'm trying to keep the bhops but remove these two other bugs. vidrelated

I've looked into it, and apparently wallstrafing is a popular bug used in speedrunning games on Source engine. This makes me happy in a way, I guess. Source is a strange beast and it feels very particular. The project I'm working on right now is a remake of Hidden: Source so it's kinda important to get these little quirks right. That being said, there's clearly some measures taken in existing Source games to ease the crazy effect the bug has, yet I'm not sure what they are. I assume it's a pile of bandaids all in different places and it hurts me to think I might just have to actually look through the whole thing.

Kill me Pete.

Needs to be brighter and unlit parts should dim rather than disappearing completely. has a good point about association too.

Alright I'm losing it here:

In UE4, I have an actor component that is attached to my player pawn. This component has a TArray to which UObjects are added at runtime. For some reason, the elements in this array are not reset when I exit the game (only tried PIE so far but I doubt standalone does much different). So I play in editor, 3-4 references are added to the array, then when I quit, they are now present in the Blueprint class defaults (which just inherits from my C++ class that inherits from UObject) as null references, but still elements in the array. Starting PIE again I now have those 3-4 elements present (extra annoying because they are null so they like to cause crashes even with range-based for loops). Here's the thing: I also have a similar array in a non-component class and that clears it out just fine when I exit PIE, no null elements are left in the array. So what gives? Is it something Actor Component specific I am missing?

Googled plenty, found nothing.

Forgot to add: Tried making the array and later the whole component transient, didn't help at all.

Welcome to UE4, kill yourself now or forever suffer

Pretty much what said. It's an okay engine if you can learn to diagnose and work around the swarm of bugs infesting it.

Strange, I would have expected Transient to work Have you tried manually calling Empty() on it in EndPlay?

Did you mark the TArray as UPROPERTY()? I remember the garbage collection doing all sorts of terrible shit if you didn't declare that.

But how am I going to make game if I am dead?

Also with some research into the workings of it it: It somehow seems to be adding the elements to the class' array itself rather than the instance of it. So I have my player class (which is a BP class that inherits from my C++ character class which inherits from ACharacter) which I cast from an AActor*, I get the component from that reference (simple Get Component in BP), then construct the UObject-derived object and add it to the array via a function called on the component reference. Standard Blueprint 101 shit, nothing weird. Somehow the Array.Add is adding the element to the component itself in the defaults rather than the instance of the component, and get this: they don't show up in the instance. So if I PIE, press LMB to do the above process, it will add x elements to the array (x being how many times I ran it aka hit LMB) of the class, not the instance. Then if I quit PIE they're still there in the defaults as "expected", but then if I hit PIE again, they are not in the instance, even though the instance of the character is spawned from the class that has those elements present in its array in the defaults. It acts like I am calling and spawning stuff on a "shadow" class and instance, which is identical but separate from the class/instance that I want to use (but there's no such thing, I checked all the references, path names, etc. and I am getting the component from the correct instance at runtime and all, so that's not it).

It's not like I'm new to C++ or UE4 at all, but I haven't had this shit happen to me before, and it's not the first ActorComponent I use. I'll go through my code and BPs again and see if I'm somehow doing something wrong, but it seems weird since it's a super simple thing so I'm not sure where something can go this wrong. I suspect it might be some weird quirk (or bug) of UActorComponent so if I can't find anything wrong on my end I'll go ask on the Unreal answer hub and pray I'm the 1 in 100 that gets an answer.

Yes to both, neither worked. Actually BeginDestroy rather than EndPlay since EndPlay does not seem to exist for UActorComponent, but no dice, it is being called (I checked), so the Empty() is also being called, but it doesn't do anything (and this is all in C++). Again it feels like the component I see in BP (and get a reference to) is somehow detached from both the instances created when I play and the base C++ class.

Looks like it was some sort of fuckery deep down in the engine, no wonder it made no sense. I did try this yesterday but didn't go as far in my cleanup, I guess I had to really wipe everything I could (without killing the project) clean for it to go away.

Lesson for anyone having some eldritch bug in UE4 I guess: regenerate, clean, rebuild.

That is fucked up

One time networking wouldn't work in games built on my machine, because I ran windows 7 with almost zero updates. Racked my brain over what was wrong for a week, machine broke down so I got a new one with preinstalled updated windows, it worked. To this day I still don't know what really caused it, some .dll somewhere missing or being outdated probably.

What would be a good engine for a simple 2d DQ-like game?
I'm torn between Godot and Unity, but have you guys got any other suggestions? I don't wanna use gamemaker. Also, languages shouldn't be an issue as I know python and C++ and I believe everything lets you use at least one of those

That image is retarded because that problem isn't even a problem nor would it be a question in any test.

Unity doesn't use either of those, so you're only left with Godot (which can use both of those). And if your project is actually 2D it more than suffices.

Doesn't unity accept C++?
Also, I'm sure I can pick up on C# in less than a week anyways

Kind reminder to not change the specifications for your programmers too often

true;

You do realize your arm sways the opposite direction of your legs right? They do not sync.

There is the nanba walking style, but yes, most people move the opposite arm when the step forward.

Is this widespread in japan? It goes completely against the human anatomy

I honestly have no clue how widespread it is, I just remember hearing about it. I can't find any good sources on whether it's even real or not. The NYT ran an article that mentioned it, but then again it is the NYT.
As for "going against human anatomy," I feel like it's kind of similar to how you breathe. It's natural to breathe from the stomach; babies do it, and so do toddlers. But at some point, a lot of people stop breathing from the stomach and breathe from the chest instead, and they get into this bad habit of not breathing properly that they can't seem to break without conscious thought. I think walking is the same way; you don't need to swing your arms and twist your shoulders when you walk, but eventually you start doing it and don't ever think about stopping.

Boys I come to you as a stupid artfag trying to learn basic blueprints.

I am trying to make roll-a-ball have a toggling first and third person camera. I have two problems.

1. The cameras are physical objects that get flung around with the ball when it rolls forward. I dont know how to keep them independant.

2. The ball spins madly but has very little friction. I want to be able to build up speed and //gofast// and build up momentum but I also want to be able to stop at a moments notice. Currently it takes forever to accelerate then cant stop efficiently. If I up the torque it just spazzes out more.

Are there any guides you guys would reccomend?

Yeah I know but I wasn't thinking shit through when I made that placeholder.

What's all this about the Kinect being able to use motion capture to make animations?

...

Thanks

Attach spring arms to your ball and attach said cameras to said spring arms. Under the details of the springarms, you can then edit if you want to inherit rotation from the root or if you want to just use the pawn control rotation instead

WHERE NEW BREAD

page 13 newfriend, as always

But Holla Forums is a 1-based index which means we're actually on page 13 right now

you go and try to tell that to a retard vol like pewter when he nukes your premature thread

No, that would be the case if Holla Forums was 0-indexed. Since Holla Forums is 1-indexed, Page 13 is the 13th page and Page 12 is the 12th page.
In any case, we still need to
AWAKEN THE BAKERS

See

what about it? I don't recall a thread about griffons being a general thank fucking god

Oh you just missed them purging all the posts then

oh was that the furry posts or whatever?
I haven't seen the images, but if they were super gay as furry art tends to be and unspoilered I don't see an issues with them being nuked

but you know our vols, doing the legwork isn't part of their mantra

Anyways, looking at how I'm approaching my behavior handling I need to change things a little bit. Right now I'm just using a delegate that take an Actor object and refers to a GameState, but what if GameState changes to a different game instance?

Because variables are captured when the method is created or first invoked or whatever, that means that they'll always hold onto the original GameState object, which means that if I eg make a new map, it'll still point to the old one. My solution seems to be making a BehaviorArgs object and shoving the current gamestate into it

They were spoiled, but looking at OP, it looks like he's just a sperg and whining to the mods because he can't hide by ID.

I'm ready to join the TempleOS master race

eat shit and die in a dumpster

can't be anymore by the book

Sure it could, he could be moving out because evil gamergators and Trump harassed his landlord.

New OP didn't bother posting here, guess I'll do it for him: