/agdg/ + /vm/ ~ Amateur Gamedev General

Hold Hands and Pat Heads edition

Related Places:
>>>/agdg/
>>>/vm/

Old Thread

Other urls found in this thread:

youtube.com/watch?v=UWB-HVFxEpU&list=PLS9MbmO_ssyAXRl-_ktrebQBFxjSQt7UX
wiki.polycount.com/wiki/BaseMesh
youtube.com/playlist?list=PLXkVsacazW2qvdnKNzgBLkUwlgi3FU-VO
drive.google.com/file/d/0B1bOSXCEc3FhMTJoZUVneUhHNVk/view?usp=sharing
en.wikipedia.org/wiki/Dijkstra's_algorithm
mega.nz/#!qVxVQS5K!2aGc1pcxRwIHCDn1c7_TPpZAymObx-AvX0wy3MuhDpg
shodanon.itch.io/besmitten
youtube.com/watch?v=nSW6ALspyZI
docs.unrealengine.com/latest/INT/API/Runtime/Core/Math/FVector/IsNearlyZero/index.html
docs.unrealengine.com/latest/INT/API/Runtime/Core/Math/FVector/Coincident/index.html
answers.unity3d.com/questions/395513/vector3-comparison-efficiency-and-float-precision.html
docs.unity3d.com/ScriptReference/Vector3-operator_eq.html
floating-point-gui.de/
solarianprogrammer.com/2013/05/22/opengl-101-matrices-projection-view-model/
songho.ca/opengl/gl_transform.html
gamedev.stackexchange.com/questions/54238/how-do-i-implement-branching-dialogue-in-javascript
docs.unity3d.com/Manual/RenderTech-DeferredShading.html
docs.unity3d.com/Manual/RenderTech-ForwardRendering.html
learnopengl.com/#!Advanced-Lighting/Deferred-Shading
twitter.com/NSFWRedditGif

shader test stuff

every individual segment is on its own material so we can set up shaders with unique ambient colors per-material in unity

I see this is going to be a very, very lewd thread, isn't it?

Anyway, what has everyone been up to? I did tons of refactoring and I'm still on it. There's almost no end to it.

I'm working on post-game content for Speebot and a new interface for the level editor.

Is procedurally generating entirely grid-based levels really feasible in 3D?
It's relatively simple in 2D since collision logic and rendering for each spot on the grid is magnitudes lighter, but that many individual objects in 3D seems like it would destroy any computer. Am I just a dumbass who is missing some other technique?

I want to fug your game.

If you do it like something like Spelunky (premade areas that are randomly selected then slightly modified, with a final check to see if there is a valid path) then it should be fine for computers. If you can figure out the limits of how the player can move (i.e. you know they can move 4 tiles up and up to 1 across in any direction if they jump from stationary) then I guess you could do some sort of flood-fill to tell where the player can go.

Minecraft did it in fucking java

I'm on the same boat, the only thing I have for now is walking movement and a shitty island

Are there any good jet/spaceship games that have a challenging way of using homing missiles? Everytime i play or see a game with homing missiles, at least 90% of all kills are due to them instead of aiming and shooting bullets/lasers because they don't need any skill and often even insta-kill enemies. It's like they purely exist to give normalfags a chance to compete with good players.

Right now i am thinking about completely removing any homing/locking attacks and make a purely aim based aircraft fighting game, but i feel like people wouldn't want to play it considering how much missiles are taken for granted at this point.

Flight sims.

Real Air to Air missiles have weaknesses and part of the difficulty in using them is first acquiring the target. Games make finding and locking a target easy and don't give missiles a "best use" case or "worst use" case where they're easy to defeat.

Thank you user that I suggested I write out my idea in full. That really helped get it out of my head and let me focus on other things.

Replace them with an Itano Circus.
Who needs homing when you can just saturate every square-centimeter with dumb missiles.

**A cool way to do smart missiles might be making them do movements in the air instead of direct homing,like a missile that goes a certain distance and then suddenly curves left etc.
or just lower their effectiveness by making it so that any particular missile has limited fuel to change its course and can only do one or two maneuvers at most.**

Do these games even provide a form of counter-measures against homing missiles? Because those are actually a thing now in a few vehicles in real life. The Russians have a device named 'Arena' Active Protection System which launches a large block and during mid air it detonates and releases a few thousands or so of fragments which can hit the missile and destroy it, and then there is the 'Shtora' which can jam projectiles like laser-guided missiles or ATGM (anti-tank guided missiles) which causes the missile to miss its target because the laser receiver of it got disturbed, though it might not work for missiles that uses wire for transmitting info instead such as TOW. It would also make for a nice arms race between defense and offensive capabilities since then the player has to decide if he wants better missiles or better counter measures against it. for example the T-90 uses 2 devices at the front of its turret for guided missile jamming.

Also it is almost a standard feature in Aircrafts that they can drop "chaffs" which is used for missile jamming too and perhaps it works against radars as well, so upon success the missile will target the chaffs instead of the Aircraft itself or it will just completely miss all the targets.

I can't provide any answers at all how guided/homing missiles since I haven't read much upon them.

In Ace Combat at least the ones I've played your standard homing missiles can rather easily miss targets despite being homing because the planes can be more maneuverable than the missiles. There's also weapons with better homing at longer distances, but you have to keep the target inside a reticle for them to keep tracking and they have less ammo.
Just go play AC4, everyone should do that in fact.

Like in Elite dangerous were instead if picking a ship with a lot of weapons and speed, you pick a giant ass cargo ship with a lot of shields and go into assassination missions and kill your target by ramming head first into them so they explode easily and the allies don't shoot at you?

What's your expierence with it? Does it hold up performance wise? I wish they'd make a fucking JIT compiler for gdscript and I'd be less reticent to use the engine.

Also, you could try making it a tradeoff to make missiles perfectly track enemies as long as you can "paint" the target for the missile, or make evading missiles put you in a defensive state that maybe turns off indicators for enemies or something similar.

Thanks for the input lads. I don't know why i never thought about looking at simulations instead of arcades, but this might actually work out:
More or less degraded to an occasional support weapon that is used while shooting at enemies for some extra damage or finisher on low health targets, instead of a main weapon.

How do I into make game?

youtube.com/watch?v=UWB-HVFxEpU&list=PLS9MbmO_ssyAXRl-_ktrebQBFxjSQt7UX

What's your experience with it?
It gets the job done, for some reason the first time programing really clicked with me was with gdscript so i recommend it for beginners

pros:
can carry whole project on a memory stick (big plus in my situation)
can make shaders using nodes
(mostly) painless workflow with blender

cons:
complex collisions are buggy (can be worked with)
have to close and open the editor every time i update the models
only one script per object (I'm not sure if this is really a bad thing)
can't collapse functions and brackets in the editor
can't rebind default shortcuts

Surprisingly it does, even on my toaster, my scene right now is about 2 million polygons strong and rising.

Sure. The problem is making sure those levels aren't repetitive as fuck while still being easy to read and navigate, since the player wont have the birds eye view of the game.


If you're going the spess opera route with regenerating shields you could make them only useful for knocking shields off (or alternatively, only useful against fighters with low shields) and give a limited supply so they have to be spent at the right time to be useful.

i tried sculpting how does it look? im worried i have the eyes to far apart.

Eyes are too high and too far apart.

Try this, user.
You can try pirating sculpting courses by Ryan kingslien

Interesting question, the only instance of 3D level generation i can think of that isn't le do anything you want world is the Bloodborne Chalice dungeons. They are just chambers with connection spots, welded together, although some play a bit with elevation and so on. They were repetitive and boring in my memory, mainly because the emphasis was on the chambers themselves instead of their interplay, so you quickly get to know all the individual parts and you've seen it all.

Generally it doesn't seem hard to pull off, you start out with the basic architecture of the level just like you would in 2D, for example with certain elements that are "elevation changers" like stairways, ladders and elevators, then do it again for the levels above and below and so on. Then have multiple other passes that add details, enemies, items and so on, ideally independent of the premade blocks so things are more varied.


With what software did you do that sketch in the second image? It looks a bit like it's 3D.

4/10
Lips are too small, and push the eyes closer together like the other user said. Also the gap between the nose and lips could be smaller. It's a good first step, check out some facial anatomy to get a better idea of the proportions you need.

Do yourself a favour.
>1. Download or make your own low-midpoly base facemesh. wiki.polycount.com/wiki/BaseMesh has a few
Even if you do organic stuff, it's smarter and faster to make simple basemeshes beforehand.

how do I go about posing and keyframing in blender without it autogenerating or tweening the motion? I really just one to go from one pose directly to another.

So your pose gets "contaminated" with the other one? I would always select the whole skeleton and then copy its values to another key before doing a different pose, I work on Maya but I am pretty sure is the same.

Just make those cargo ship part of a DLC so that only pro players can use them, they are anyway good at games because they have money, compared to noobs which tends to be poor fags :^)

t. John Smith

Open dopesheet, select all keyframes you did so far, open 'key' menu, open 'interpolation mode', choose 'constant'.

There are lots of algorithms to combat densely populated 3D grids and even maps that aren't even organized as grids. You should not worry about performance and speed when throwing around lose concepts as "3D grid-based levels". If you are building stuff from scratch with your own engine, look up octrees and other derivatives.


Make it less ayy lmao.

I don't see what everyone is complaining about.

...

...

In GameMaker, how much of an FPS drop should I expect when pixel perfect upscaling a game? My game's native res is 480x320 and it runs at 60fps, but when I upscale it to roughly 1600x900 to fit my monitor, it drops to about 30fps, though not consistently. It's a pretty simple game, so I didn't expect such a drastic change.

GM's default, automatic upscaling looks like shit, but doesn't seem to have this problem. Plenty of frames. Am I just underestimating the amount of power that pixel perfect upscaling takes? Is it because I'm running it on a 2009 i5 laptop with Intel graphics? Is it more likely that I'm just doing the upscaling wrong in the first place? What gives?

I still want to add you, Ika-senpai

I'm happy you're all still trying to make vidya
I miss it

Me and my pal decided to chill out this time and make a simple, yet feature complete game for the cuckjam 39

What u think?

Attack on Titan/10

...

Make it comfy like that old Defend Your Castle flash game, I killed so many hours on that as a kid.

Nah user, add instakill. This means that you can't just have more HP than your enemy and trade a missile to win in a game of chicken. Adds more suspense and skill requirement. Oh shit, a missile at all, I have to now evade instead of just WHATEVER.
Everything else you've greentexted is great though. If you want, have smaller missile varieties that deal less damage, but standards should one shot, especially when you have countermeasures and everything else added in. Play with the balance, if you think it needs it, make it so a missile deals 90% HP and so a bit of machine gun fire is required to finish off.

It better be bloody hard
Nenechi is an angel and i must protect her innocence

It's actually completely different from the rest of the game. It's more of a reward than an extension of the game.

But I want to make my own games or at least own the studio making them, this is going to sound faggoty but I find my ideas to be too valuable to see someone else use them, so I have to make them myself, all of them.

it should be fine because GM draws everything on an internal object called a surface, then draws it on screen. are you literally scaling every single object on screen by 2 every frame or something?
watch this youtube.com/playlist?list=PLXkVsacazW2qvdnKNzgBLkUwlgi3FU-VO

Heh, I might add some simple golem model as homage


Never played it, care to share some of the thinjgs you like the most about it?


Anyway, early feedback demo is complete.
Who wants to zap some imps?

drive.google.com/file/d/0B1bOSXCEc3FhMTJoZUVneUhHNVk/view?usp=sharing

You're talking about putting people on Ace difficulty by default, Buddy.

No, I just multiply the application surface's size and reposition it accordingly, whenever the multiplier changes. I'm thinking there's some kind of bug going on, since it sometimes returns to 60fps, despite still being high res, if I toggle fullscreen after the fps drop has started.
That was great, thanks. Much better than other tutorials I've seen on the topic. I'll give that method a shot.

this was painful.
and it's not over.

yeah also there's that problem.

sorry mate your algorithm is bad

Looks like a greedy algorithm starting from the green node, of course that won't work too well because it's a shit algorithm. 2D (grid-based) pathfinding is a solved problem, if speed is no concern and you want a reasonably easy & accurate algorithm, implement a breadth-first-search, if you're concerned about speed and can deal with a somewhat harder algorithm implement A*.

it's not even mine, I tweaked something until it worked
I know what's happening though, but I'm not sure how to fix it

it's supposed to be A*, because I need different costs for tiles eventually

I guess the answer is something with parenting, but I don't quite grasp this part

The primary problem of breadth-first-search isn't speed, but memory consumption which grows exponentially. So unless user sticks with such small levels, it would be a really, really bad idea to use it.

A* can account for weights - you have to add them to the heuristic function and not the algorithm itself.

It is generally like with cryptography, you don't really want to write your own.

Other than that you can always post your code..

Exponential memory complexity? Should be fine for a grid-based game with map sizes under 200x200 or so, and you can probably be smart about dropping tiles from the visited list that you can no longer reach in any way and achieve memory complexity that's linear in the map dimensions (instead of quadratic)


Yeah, please post your code.

I highly approve of your gif, OP-sama
have a lewd

will someone answer this for me?
I'm a modelfag and I wouldn't mind spending time to create another model. but I don't want to hinder the programmer by making him do too much stuff.

read "learning to draw on the right side of the brain" by betty edwards
it helps you with observation and looking at your work for errors.

If it affects gameplay, there'll be more code and more chances for bugs, but otherwise it shouldn't be a huge difference.

How about you ask your codemonkey?

I mean, if you want, but it's Game Maker
I'm just posting progress, I'm not trying to get people to fix everything for me
I suppose the final answer is always to just do a complete grid search

But that sounds fun.
Plus, for the most part Ace Combat doesn't have a lot the countermeasures and detractors to hit that appear in user's post.

For the most part that's a lot of work for whoever has to animate an entire new set of animations, models, etc. sums it up pretty solidly.

thanks

he's away from town this week.

That would definitely work, but I'd like to reserve the ability to tween the motion once I set the spacing and timing manually, though I probably won't

that worked, thank you very much! For brevity sake is there a way to keep it from tweening from the get go? If not it's no big deal, it just would save me a few clicks.

If it's a system (it is) it could always cause bugs.
Even if it's as much as swapping a model/animations or a set of sprites, you still have to handle loading those resources and displaying them, and depending on how you handle things (ex, the male/female sets use different data), you have more stuff to maintain, and more room for bugs to creep in if you forget to maintain things. Also, different models with different rigs will animate differently, and if you're applying the same animations to different rigs, you could always get animation bugs, so you'd probably want to build separate sets of animations, and that takes more time.

tl;dr- Yes, adding anything to a game can add bugs. However what bugs/time drain it can add are completely circumstantial and reliant on the systems in place and how they are modified.

Try understanding and implementing this before you try A*:
en.wikipedia.org/wiki/Dijkstra's_algorithm

Don't do it, but if you really don't know the answer don't do it but keep it in mind along the development and you'll see each hurdle you would have had to pass to reach completion so that you can take a more informed decision for the next project.

I personally never said I can fix it (not that great a programmer) but I'd be interested to see it because of the result it made (and indeed I've never written game maker, but it shouldn't be too hard to understand)
Also I am not sure that's the right attitude - to sacrifice quality or time because you want to reinvent the wheel

I'm not a coder but I believe this hinges mainly on one thing: is it a cosmetic change ONLY? If it is, then outside of relatively simple code to make that ability to change the model, it shouldn't cause a problem. If it isn't, and changing the model is an option ingame that translates to changes in other assets or mechanics or whatever, then the possibility for bugs goes up.

The BIGGEST thing you'll probably need to worry about is the animation. Men do not have a womanly gait, and most game with both a male and female model have animations for either, because using the same animation (and sometimes the same rigging) makes it look wholly unnatural. Then you'd need to worry about a whole new animation set interacting with the environment, etc.

Literally swap the model with another. It hardly counts as "work" at all as long as the different characters don't work differently.

thanks for the answers
you guys are the best

Off the top of my head PUBG and MGSV use the same animations for both female and male, and it looks perfectly fine.

If it's action, military, or something along those lines nobody will notice. If it's an RPG or something slower then people might notice.

mega.nz/#!qVxVQS5K!2aGc1pcxRwIHCDn1c7_TPpZAymObx-AvX0wy3MuhDpg
I think the result is easy to understand, it's trying to get closer to the end point before encountering an obstacle, but then it should be made to "revise" the previous path to fix it, but it doesn't

Couldn't you just have posted the text instead, or a link to a paste?

Oolite (an Elite clone) might fit. The missiles there don't always hit, and you can buy an ECM module for your ship. There are also ECM-hardened missiles. It's possible to destroy missiles with your laser if you can aim well enough.

Prison Architect rip-off clone were you play as a concentration camp commandant. You design & build your own concentration camp from scratch – barracks, gas chambers, crematoria, medical and research areas, etc.
You must also connect the camp to a railroad so trains with jews and other prisoners can arrive.
You will have to conduct experiments on prisoners, stop prisoners from rebelling and escaping, and most importantly, you must gas and burn the jews fast than they arrive by train.

wow that sounds so original and fun

I've probably seen this idea 10 times before, yet not one prototype of it.

no because everything is broken up into several scripts

There is one very important thing to pay attention to when using a single set of animations between male and female models- do not try to just apply a female model to a male rig. This is what DaS2 did and it looks fucking awful; pic related. Women have MUCH slimmer shoulders than men do, and you need to pay attention to that in the rig. You can still use the same animation set without fucking up the arm animation though, just lengthen their forearms the same amount you shorten their scapulae (shoulder bones).

Though if you're using FK only animation on the arms, you don't even have to worry about changing arm length for women. Though I would suggest it anyways as it keeps your options open in case you want to do IK for some future animation.

unless there is any difference on gameplay the main problem will be animations

You've seen hundreds of the same ideas without a single prototype of it.

I'll have to add some action camera for watching gibs up close

Since Holla Forums appears to be visiting - and Flash embeds still don't work on Holla Forums - here's a tiny concept test for a game called "Nazi Frogs".

If I remember correctly you can set a default from the file->user preferences->editing->default interpolation or something like that

on the sign "open borders for everyone" the "except Israel" (i assume that's what it says" is too small to read even on my 1600x900 monitor. It would be even harder if that game was embedded in a smaller window

Well, that one specifically. I keep seeing it posted like it's something original, yet these fags never want to make it themselves.

Why?
Everyone knows it takes days to properly burn a body in modern crematoriums, not the 5 minutes that (((some people))) seem to claim.

Do your research
>>>Holla Forums10327092

Thank you captain obvious

No old memes here!

There are rules & regulations for cremating a body that must be followed today. The body must be treated with respect, it must be burned long enough & at carefully monitored temperatures.

This is common knowledge.


Old memes are best memes, user.

With a thread this dead-ass silent, I trust you're all working on your games. Right?

I won't be making progress on either of my 2 projects for a while because I will go to vacation today. first time in my live

These threads consistently reached bump limit in 2-3 days like a month or two ago, I wonder what happened.

yeah, just that posting code isnt all that interesting I guess

I've been innawoods for a month with no internet access. What happened?

couldnt spend much time on this the past couple days because of a bunch of financial and life issues that came up

REALLY fucking hoping to FINALLY be done with this this weekend or at least by the end of monday, I just want to get onto the next one and this has taken WAY too long. not that I haven't had fun, I just feel like shit about not putting out more

luckily we're not at a point where we need to be pumping out new assets and our programmers are still working out the kinks and fine tuning the movement system. they told me they don't want to make a video until they feel like they have something substantial to show that isnt just ordinary 3D movement so… we'll have to wait and see.

I haven't much gotten a chance to test out what's there but I do know they're making good progress, they're currently working on making rail logic functional last i recall, as well as smaller movement details like fixed turning angles (turning beyond which will activate the dash-turnaround animation like in JSRF)

I did get a chance to start texturing tonight and I tested the textures i made in the concept shader in blender; pretty neat. I'm texturing with no more than two shades, and I'm going to overlay color and designs on top of the two-tone shaded textures. hoping it'll look nice in-engine :0

we also got a guy on the team working on setting up the shaders, god bless. our team is looking pretty dandy right now and I can't wait to actually have a video out for you folks.

loving all of the games I've been seeing btw, thoroughly wondering what the fuck that grass simulation is gonna be used for. lowkey hoping for lawnmower sim.

sort of

Cool, thank you very much for the help

Yes. I did tons of refactoring and implemented item descriptions. I liked how there was a lot of stuff to read in System Shock 2 and implemented something similar, albeit a little more complex. I can give items multiple levels of descriptions and also specify dependencies to other items for unlocking them. So you can pick up a flashlight and it tells you that it's a metal flashlight. You can click "Examine" and it tells you that it's an old and dirty metal flashlight with a rubber grip and a cracked lens that has seen better days, but still works. If you have the manual in your inventory, you can click the button a second time and it gives you the manufacturer's description.

Is it really that hard to sell games, or why do people claim indie dev is incredibly hard to live off of? Or are indie devs always san fransisco starbucks slurpers who spend $8000 per month? Maybe people have no interesting ideas and just copy what's in the market? I remember that story about a game that's """good""" but didn't sell, and it looked like the most unoriginal and uninteresting flash game.

From what I've seen people seem to treat

Do you just have no idea of how many indie games go entirely unnoticed?

I haven't seen a decent looking indie game that went completely unnoticed.

Most of the games that I've seen not selling are the kind of games that look fun enough, in the sense that I'd play them while browsing free flash games. In the same sense there's a lot of art that's neat to look at and I may even save onto my hard drive, but I wouldn't pay any money to commission such a thing.

Marketing & 'Designated Target Audience' are basically the main killers of most indie games. It's why even a decent low-budget Platformer will likely struggle to break even.

A lot of nodevs on here are college age(including me) so there's a lot of time spent doing summer things right now. August in particular since everyone has to go back to the daily grind around late August/early September. I've been spending a lot of time with friends before they ship off to their colleges that many of which are on the other side of the US

Shit I can't believe I missed these before hitting that reply button. There was a survey on Twitter asking how much many indie devs have made on their games. 46% answered Less than $500, then about another 40% was split between the $500-$5,000 and $5,000-$10,000 category. And then something like 14% of people who answered made over $10,000.
Granted this was an anonymous survey so for all we know they could have been a bunch of hack fraud mobile devs trying to make money on a cheaply made infinite runner or other basic bitch, freshman year game programming type games(Asteroid clone, SHMUP, uninspired platformer, etc).
You need a good combination of style, fun gameplay and originality if you want to breach that 10k sales goal. Take a game like Smashing The Battle for example. Released in May of last year, Steam Spy reports it's currently sitting at 25k in sales. Roughly 3/4 of it's sales from last month were at full price and 1/4 at a sale price. After refunds and everything it looks like the total number of sales this month was 1.5k and if we do some quick and sloppy math that should mean that the gross income for the devs is roughly $15,285. Even after Steam's 30% cut and relevant taxes, Studio HG is a one-man Korean studio so that's a good month.
That's the best example I could find of a lower sales figure dev who's making it as a game dev. Now if only STB supported borderless windowed and a lot of other basic PC setting.

It's slowly taking shape.

If the rest of your game looks as good as your models I honestly think you can get Naganuma to make music for you.

There's so much that has to be considered when it comes to this, user
You absolutely want enough money in your (or your company's) bank account to fund at least three games simultaneously. Why? Because if your current project only funds the next one, a single commercial failure will sink you. You want to be able to have another crack at it, if your current project turns out to be a failure. You need to be able to afford that. Secondly, you shouldn't work off the assumption that your calculation will never have to be updated. Your living expenses may go up, you might horribly underestimate the amount of time a future project will take. What if you get sick?

That's how I got my three games number: (1 failure + 1 success afterwards ) * 1.5 as a buffer which allows your expenses to be 50% higher than expected on both projects = 3 games.

The guy I'm planning to work with is friends with the developers of "Ben and Ed". They're currently partaking in a competition and are optimistic that they win (because there aren't many other good games according to them) and if they win, that would give them 300.000€. They told him that this buys them one year. Since they're a three-man studio, it boils down to 100.000€/person/year. And their office is not in a metropolis like San Francisco.
I don't know how materialistic their lifestyle is, but I doubt that they are living like rock stars, so the 100.000€ figure obviously includes taxes, insurance and shit like that.

My plan is to work at getting to the point where I can get away with fucking up a project. That means living like a cheap-ass until I can afford it. Which isn't something I have a big problem with.

user, you and me both know that """""devs""""" like LW participate in that. You'd need a survey that only asks people who actually give a damn and don't shit out some cheap-ass Unity game with bought (or stolen) assets. Don't forget that mobile gaming is huge; huge as in a huge volume of developers who go for quantity over quality. Back when I was using Unity and active in their forums, that was the deal all the time. Throwing shit at the wall and praying that something sticks one day was the approach they advocated for, and there's nothing wrong with that. Quality doesn't pay in the mobile games sector. There's no qualitative reason for Flappy Bird having made ludicrous amounts of money. Those devs experiences are perfectly valid, but don't apply to indie developers on the PC/consoles.

Also: What are you working on?
Also also: Where is tetrachromedev?

once we get to that point I'm gonna try and make contact with him

I haven't even started principle development on my game yet, but I made Pong with a pirated copy of Game Maker 8 today, so I feel fairly accomplished.

My whole post assumes that I'd have the skills to "easily" make games and won't get stuck on learning how to music or tinkering with every common game mechanic that I come across or anything like that. Basically; I'd already be familiar with making and finishing some games.

I can easily imagine myself making ok small games in just 1-2 years assuming I work on it for most days. Based on my current experiences with dev and art I think it wouldn't be outside the realm of possibility NOW even though I have a job, if I worked a lot. I just lack the skills/experience and usually don't get very far thanks to it, since I spend most of my time stuck on every little thing I need to do and get demotivated because I never seem to get closer to actually making the game. Having to spend time learning and implementing a new mechanic isn't bad, but currently I get stuck on literally everything and constantly realize that something I did previously is either shit, not future proof, or both.

I'll have to admit I'm not very familiar with the way selling games gets taxed here and stuff though. I know steam takes 30% or whatever but that's it.


I'm not tetrachrome dev if that's what you thought.

I'm not working on anything right now because I half gave up a while ago due to becoming to really hate javascript with a passion. I need a better thing to make games with, but I'm kind of lost looking for something I'd be comfortable with, whether an engine or just another language. That's my "game" >>>/agdg/29145

Sweet. He seems to be pretty outgoing and hungry to work on projects, so your chances are good. You just gotta make game.
Best of luck.

You just made me. I'm looking at redoing my entire collision system to insert a tree that actually works, and arbitrary polygon collision with better collision handling.


I'd normally tell people to use whatever language gels with them but Javascript is just a terrible, giant turd.

Holy shit rigging models for valve biped is a lot easier than you'd think. Here's the first WIP enemy for Kowloon game.

does it make you mad if an elemental magic system doesn't always make perfect sense?

It's fucking magic, it doesn't work on common sense.

I'd say no, but the one pictured is bothering me.
????

electrolysis

Poison beating Fire makes sense if you think about Poison being a liquid. Fire beating Lightning is the only one that doesn't make much sense, make up some bullshit like Electricity + Fire = Bad and go with it, I like it.

Caustic is fun and really easy to use.

Usually throwing a toxic substance on a fire either makes the fire worse or it blows up.

finished the two-tone painted "AO", gonna work on base colors in a bit.

Oh fug, are you making a game per-chance? This song would be great for the sort of shit I'm working on.

Unless he did something stupid like having every instance of player interaction go through a command line referencing the specific player model, then no it shouldn't be an issue.

well, most poisons are either oily or mushy, properties also found on flammables and explosives

That's why I put in that I don't trust it completely. And also why I put Smashing The Battle's sales data in my post. Because it's a better example of what user is looking for. If you're a one-man team looking for 10k sales in your first year then a one-man game with 25k in it's second year is the best example you could find. The math I do is pretty promising for a small time indie dev.
After taxes and Steam's cut the developer of Smashing The Battle is probably pulling in around $7,500 a month(assuming his sales are consistently 1-2k a month, which the chart available to me seems to imply that it is) which is not bad at all and this news is probably motivating that user even further to make his game. I know it makes me feel more motivated to keep working on learning to code so I can make my games some day.

I'm making it for LowRezJam, some combination of Shadowrun and Teleglitch. However I'd certainly be interested in making music for you.

It triggers my autism, but it's not as bad as FFX's system

That sounds pretty fun. Have you seen any of my other posts before? I'm pretty stunted as a musician and I'm really only useful for vapourwave and occasional fighty music. That song you posted has some great /cyber/ to it.
Do you usually do that sort of techno stuff?

Wrong image.

I have, it looks pretty cool. I don't usually make music honestly, this was the first thing I made.

Are you doing just the music for that Shadowrun inspired game or what? That's pretty good for a first time.

What happened to earth?

I am apparently blind.

noted, does it look better i still think the lips are off and the eyes feel like the scale is off, maybe their is too much fat on the sides and i should shave it off also i cant get the cheeks right the always come out flat. so anyone have advice on that.


ill plan to watch that soon when im free


yeah i have to find it, this is not the first time someone told me to read it.

Maybe you could reverse the inner arrows.

Magnets OP pls nerf

also if you're a weeb, you can do the upside down drawing with animu too.
or any pic with line art.

japs do this to make generic mangos. (if you want to be leonardo da vinci, then you'll have to do meme realism)

The drawing upside down thing isn't just for anime, it's for drawing everything. The whole point is to disconnect what you think something looks like from what you're actually seeing. Looking at something the right side up, your brain gives you a false impression of it because it relates it to similar objects you've seen. So by drawing upside down, you see a collection of lines and shapes rather than a complex object. Eventually you're supposed to be able to draw in the same way right-side up.

Added a ragdoll. Finally, non-HL2 enemies for Kowloon!

You need to lower the fog distance but keep the end distance. Otherwise it looks too cheap.

It's a temporary measure whilst I detail shit. Thanks for the advice though.

i've never heard of that, maybe ill try it. but i hate anime weeb shit it feels way to textbook, but its better then what we have because they at least have human proportions and they're some people who do some nice work with it. but if anything i would like to mimic kazuma kaneko style

shodanon.itch.io/besmitten

Another Ludum Dare entry complete!
have fun.
hope it's comfy enough for ya

Looks like shit, but it's better than what I could make in a weekend or month, so

This might be a really dumb question, but I'm not seeing any projects show up when searching… is it possible to dynamically load data with idTech 2 or 3 (preferably idTech 3)? I want to dynamically load map, texture, entity, etc…

At that stage it's best to heavily utilize references.
Like overlaying images, and learning the anatomy through rigorous practice (that's what I've been doing the past few weeks).
It's generally only the experts with many thousands of hours that can do stuff like you're trying to do from hand w/o references; though even then they're working off a base model to make their workflow more efficient.
As in production it's more about getting content out quickly at a high base level of quality (first 90%), and you can detail in the last 10% with more time if it ends up being what you (or your design lead) has envisioned; so workflows have to be highly efficent for that initial 90% (base model, using scan data, streamlined workflow from: base + UV map -> detailed model -> textures -> rig -> in-engine).

Remember that the shape of the face is highly reflected by first the skull, then the muscles, and then the fat; with an overlay of skin.
You can correct the issue with the brow, jaw, upper cranium, and cheekbones by overlaying a skull or using a reference.
It's also best to learn how to properly sculpt the face, such as the mouth area, via utilizing proper techniques (i.e. stroking with the clay brush in the direction that the muscles go, which naturally forms the mouth muscles/mouth landmarks).
Here's an example vid:
youtube.com/watch?v=nSW6ALspyZI
You can find his course on cg peers

G O T T A G O F A S T

yeah i figured that when i got the lips wrong i looked at the muscles to fix it. also thank you, i've just finished up for the night almost midnight here, so how much did i improved and what should i learn next to make it better, because my next plan is to do hair and i still don't know what hairstyle i should go with.

Brow area, eye socket, and cheekbones are looking a lot better.
Areas to improve next should be first the cranium shape (everything flows based on this), jaw line/chin, the mouth/cheeks, eye lid area (in first pic), and then the nose.

F A S T E R

also good cranium shape references (actor is the best I could find for multiple angles, and with before/after).

10K in sales would be a massive success for me, granted my first game just reached 1K in sales and was just $2.

Frankly, the indie hipsters we all know and love look pretty sound financially, even if they didn't have rich parents they're still pulling in decent sales if you consider that the kinds of games they make look cheap to produce if they didn't spend the production cost on lattes or SanFran apartments.

The level I'm in, and frankly the majority of people here (with the exception of industry veterans pretending to be anons) seem to be on the realm of making "garbage games" like Bloody Boobs. There's actually no shame in that, as long as you're an honest dev who gets players honestly with honest gameplay (and/or with other features such as honest to goodness tits), its just that majority of garbage game devs are spergs who have given Greenlight the bad reputation it deserved by attacking Youtubers or being asshats in general. But I've seen some of these honest types of low budget games on Gamejolt – yeah even those with a price tag – and they're more likely the type of sustainable gamedev industry I'd like more of. I definitely want less of those naive devs who blew 50-100 grand, sold their car, remortgaged their house, made it or broke it with Kickstarter, etc. which is definitely high risk gambling.

If only there weren't so many horror games with asset flips, the $2-$8 realm seems to be where majority of trying hard devs seem to be.

Looks nice.

Got something you wanna say, user?

People judge a product by its cost too. While Apple headphones and Beats By Dre are known to be overpriced garbage, simply because their price point is higher, people implicitly assume that it has to be a higher quality.

Likewise, if you look at a $5 game, you'll assume it's shovelware. If you look at a $15 game, you'll assume it's adequate indieshit. If you look at a $70 game, you'll probably think it's AAA trash.

Another UE4 rant:

It's a piece of dogshit editor for placing objects with retarded shortcuts and inconsistent layout.

I'm sticking with it because it's a powerful engine and what I need for my VR game but fuck me, placing map props makes me want to go kiss an oncoming train.

...

You're a year late to cash in the "EARLY ACCESS VR TECH DEMO" lad

There's a strange truth to that with mobile games. Generally, 4.99 games are better than free or 0.99 ones.

Just Like Use HammUer

For those of you that are into making your games modable, or like to mod games, what do you think are the best ways to go about this?

I basically want to create a game that'll allow you to load in modules that will be able to totally change the tilesets, enemies, items, possibly skills and even rules for the game but I can't think of a nice, clean and easy to use way for doing this. Main way I was thinking is along the lines of text files and then an in-game editor to make new maps and import tilesets, but maybe it'd make sense at that point to just make an in-game editor that can do the whole lot.

It's better to design the engine from the ground up to quickly import assets into.

Like take inspiration from engines like Id Tech where they kept all of the game's data in .pak files, the game just ran whatever .pak files were in the "base" folder and you could run different mods by just putting different folders in the directory with pak files in them and setting a specific command line function for it.

Id Tech engines are great inspiration for this sort of design especially since John Carmack was a technical wizard.

Other engines include things like how Rockstar for Grand Theft Auto 4 kept all of the information related to how weapons and vehicles handled in very easy to read .xml files that can be edited freely. It was done this way because they had hundreds of developers working on the game and needed a way that people could easily tweak something very fast.

I'm not in it for money at the moment.

There was a board game prototype where the goal was to construct a railway and get as many passengers onto your trains as possible, and it wasn't until the end of the game that you find out you were sending them to Auschwitz.

Are you the one modelling cathedrals in Blender?


4.99 mobile games are basically the equivalent of 39.99 handheld console games

There seems to be a lot of Endless Runners which started out as free / 99 mobileshit that managed to make it onto PSN for 7.99

I'll take that as a compliment. In all honesty, making the game was a refreshing side project, relaxing in fact.
Jams are always a good time for experimenting and checking whether your worklow (either solo or with team) functions properly. We seem to have gotten better or at least faster since the last jam.

And we made what's definitely more of a game than our earlier LDJam project.

Perhaps it's even fun?

While we're not planning on developing this further, unlike Beelzebox or Tumblin' down, any opinion on the game is welcome.

So play it. I need (You)s Download's on itchio.

shodanon.itch.io/besmitten

Is there any reason why Unity would think that a Vector3 that is printing out as (0.0, 0.0, 0.0) would be different from Vector3.zero?

Cause this shit makes no fucking sense

Floating point inaccuracies perhaps?

How do I fix this?
I never had this problem before in Unity, even with very similar scripts

Vector3.magnitude == 0

perhaps?

There's a whole bunch of causes that I could name.

From floating point perspective:
The function you use to print out the floating point value might not print it 100% accurately. It's pretty common that print functions use rounding to prevent printed numbers to go out of whack.
Floating point has both positive and negative zero (0.0 and -0.0). These are binary-wise different, but when compared are considered equal. Some functions actually return positive or negative zero as part of their documented behavior.
When touching edge-cases, things always get fucky. (If you substract a floating point value off of itself, you will not get zero if the value was Infinity or NaN. As such, "0.0 == f - f" does not always return true.)

From object perspective:
Are you sure you're comparing the actual vector data, and not the objects? I know C# has probably overridden the equality operator to make it always compare the contained values, but there's still plenty of ways that allow you to compare their memory locations instead, which could return false, despite the data being equal.

Anyway post your code faggot because floating point is always stupid.

Oh and actual important addendum: I've had it happen in my projects that sometimes, floating point values of infinity or NaN get printed as "0.0" instead of their actual value, which is something you might want to double check.

Check whether it's the problem first. Use the debugger.
It's also not a Unity problem, but a general computation problem, regardless of language, software and hardware. Floating point arithmetic is a pain. Using the == operator on floats and doubles is dangerous.

In Unreal you have a special function for comparing vectors safely.
FORCEINLINE bool FVector::Equals(const FVector& V, float Tolerance) const{ return FMath::Abs(X-V.X)

if (moveDirection != Vector3.zero) { Debug.Log(moveDirection + " " + Vector3.zero); Debug.Log(moveDirection != Vector3.zero); transform.rotation = Quaternion.Slerp( transform.rotation, Quaternion.LookRotation(moveDirection), Time.deltaTime * rotationSpeed ); }

movedirection is just a Vector3 with my inputs depending on camera position


nope, didn't work


Actually not sure how to use the debugger
But I get why it doesn't work, it's just that I've used this exact same thing a bunch of times and I never had this happen

I fixed it by making a new Vector using my inputs directly, it's ugly and gay but it's good enough for now.
I'll go back to it later

ew Vector3(horizontal, vertical) != Vector3.zero

Holy shit, user. Why are you doing this to yourself?

Luck. Something you do with this particular vector makes it not become exactly (0, 0, 0), which you need to expect.

I've never needed it?
Dunno man

You better learn how to use it. It'll save you a lot of headaches in future.

My record is wave 73 and 1949 kills.
is tedious. after learning how to manage damage multiplier there is nothing else to do but grind and that is pretty boring without any reward for beating the waves. The game gives you false information - in the 'input' tab it says that you can use 'left-ctrl' as a firing button but it's not true, only LMB works. The attack seems to be too weak for a lightning strike from god, it has too small area of effect. Holding mouse-button all the time is pretty tiring and there is no alternative like Destroyed buildings do not reappear when you go for a rematch.
are good, The game is pretty in a rough way. The models and animations look good enough from set camera distance. Destructible walls are a nice touch. Only things I would do to improve is change that standard Arial font to something more fitting for a fantasy and add more decorations in places where there's nothing but grass.
is uninspiring. If the village is constantly stricken with catastrophes like sickness, hunger and demons then it means that their deity is pretty shit. We're not playing the role a guardian of the village but more of a low-tier slacker that springs to action when situation is beyond saving. I think it would be cooler if villagers were abandoned at the start and deity we're playing as came to answer their prayers and save them.
Bretty gud for 2 or 3 tries.

forgot to remove that

SFML feels so good, but it feels like I'm cheating like crazy.

I can get shit done so much faster than in SDL.

docs.unrealengine.com/latest/INT/API/Runtime/Core/Math/FVector/IsNearlyZero/index.html for checking against zero vectors, and docs.unrealengine.com/latest/INT/API/Runtime/Core/Math/FVector/Coincident/index.html for vectors that are the same but not zero.

...

That's great but I'm using Unity

I fucking typed unity into gogle and it still gave me unreal answers that I brainlessly copied, fuck.
Here's an answer from someone else answers.unity3d.com/questions/395513/vector3-comparison-efficiency-and-float-precision.html
Basically check if Mathf.Approximately(Vector3.SqrMagnitude(), 0.0001) is true.
Alternatively, docs.unity3d.com/ScriptReference/Vector3-operator_eq.html this says that using == instead of Vector3.Equeals is also good for approximate equality. The first option allows you to tweak how close it has to be, this doesn't, but it might work.

Yeah, but anyone can say their shit is simple and fast.

I didn't actually expect it to be.

does it look better? still having problems with the lips, but hey practice make perfect and this is mostly for practice


shit, i forgot to ask one question, when you see my model do you see a man or a woman?

I just wish they included batching rather than their stance of 'well if we do it it won't be ideal for every use case so we just won't do one'.

Thanks for playing, dude. That score is… a bit more than I expected anyone to achieve, actually. I think you'll top the scoreboard.

I like how you "got" the story. If we had time to flesh it out, I'd imagine it would go down the slacker deity way - humorously this time.

Thanks for the kind words on the visuals. I agree that trees and grass particles would make the game more interesting but there was no time to make them.

Actually, I agree with the gameplay comments as well. It was designed for short gameplay, most jam games are, from what I see, at least. It's about the time constraints. We had a vision of many skills and effects but needed to cut down to bare essentials.

That's not a woman, it's a man or a tranny at best.

At first glance, I see a man, but after studying the face, I think it could make for a "homely" woman.

Somehow it got fixed by itself
Fuck if I know how or why

give it to me straight, do it look like you know wu also how do i fix or rather what are the key points on a head/face that can make it a man or a woman. i know men tends to have more of a jawline while women faces are softer but i really dont know where i went wrong.

Never use == with floating point values.

Instead use abs( float_value ) < SMALL_NUMBER

For more information why, read floating-point-gui.de/

yeah good point forgot about that.

Why are people doing this?

Proportions in the frontal view are wrong. The eyes are supposed to roughly be the midpoint of the face's height, so the chin looks far too large and creates the impression of manface. That's the biggest offender.


Please tell me why that's bad programming practice, I might learn something.

...

Instead of acting like a dick you could tell me why that's bad

Never use booleans for button input, but numbers like int or float.
Unity even helps you out with this and gives you "Input.GetAxisRaw" to output a number between -1 and 1 on button press, works great with analog input too. This way you can simply add, subtract, multiply and divide any button input the way you want, and use it to do math with or directly manipulate your vectors.

A quick script example:
———————————————–
public float speed; //How fast is your dude moving
public float rotationSpeed; //How fast is he rotating

//Put all the button input and calc shit into Update
void Update(){
float x = Input.GetAxisRaw("Horizontal"); //This means left button press is -1, right button press is 1, no button press is 0
float y = Input.GetAxisRaw("Vertical"); //This means down button press is -1, up button press is 1, no button press is 0
Vector3 move = Vector3.ClampMagnitude(new Vector3(x,y,0), 1); //make a move new Vector3 based on x and y input, and ClampMagnitude it so the max distance in any move direction is always 1. If you don't clamp it, your character moves faster when pressing/inputting both x and y at the same time.
}
//Put all the movement shit into FixedUpdate to prevent it from going mumbo jumbo lagging in Update because unity physics
void FixedUpdate(){
transform.position = Vector3.Lerp(transform.position, transform.position+move, Time.deltaTime * speed);
transform.rotation = Quaternion.Slerp(transform.rotation, Quaternion.LookRotation(move), Time.deltaTime * rotationSpeed);
}
————————————————
I hope i didn't fuck up i am too lazy to put this shit into Unity to test it out

And there you go, without a single if statement. Of course this is very basic, but you can build up from there. Always use -1 to 1 input math in a smart way and you won't ever need a single if statement again, at least i never had to no matter how complicated my controls got

I agree, post something helpful or why bother? conditional statements are fine in character control logic. Checking for non-zero movement vector and doing an early-out si perfectly fine.

Fixed version. Quickly tested it out. Move along the x and z and rotates only when you press a button

using System.Collections;using System.Collections.Generic;using UnityEngine;public class MoveScript : MonoBehaviour { public float speed = 10; //How fast is your dude moving public float rotationSpeed = 10; //How fast is he rotating Vector3 move; float x; float z; //Put all the button input and calc shit into Update void Update(){ x = Input.GetAxisRaw("Horizontal"); //This means left button press is -1, right button press is 1, no button press is 0 z = Input.GetAxisRaw("Vertical"); //This means down button press is -1, up button press is 1, no button press is 0 move = Vector3.ClampMagnitude(new Vector3(x,0,z), 1f); //make a move new Vector3 based on x and y input, and ClampMagnitude it so the max distance in any move direction is always 1. If you don't clamp it, your character moves faster when pressing/inputting both x and y at the same time. } //Put all the movement shit into FixedUpdate to prevent it from going mumbo jumbo lagging in Update because unity physics void FixedUpdate(){ transform.position = Vector3.Lerp(transform.position, transform.position+move, Time.deltaTime * speed); //Lerp to position + move transform.rotation = Quaternion.Slerp(transform.rotation, Quaternion.LookRotation(move), Time.deltaTime * Mathf.Clamp((x*x)+(z*z), 0, 1) * rotationSpeed); //The "Mathf.Clamp((x*x)+(z*z), 0, 1)" will always turn out 1, as long as at least one button is pressed, otherwise 0. Multiply with rotation speed }}

Lips are lacking the appearance of the muscle structures, but it they do look better (getting there).
The eye lids look better, but each of the eye structures (muscles) flows into the other; use references for what is missing.
I'd also work on the facial proportions (provided are rough references).
There's actually a system to laying out the face (head measurements).
It follows the golden mean, and is very useful; I'll post a few references, but it's easy to find more.
Last picture is illustrating the issue you're running into with proportions.
Also, use the transpose tool in zBrush to do the measurements; look up tuts on how to set it to custom measurements (set to 1 head length, draw measurements with the polypaint in orthographic view, correct face with move brush, etc).

Definitely a man.
Due to the jaw structure being more "solid", and the chin being flat; as women have a soft sloped jawline and very round chins.
Other things too, but those are the main factors atm.

The references I posted are female because that's what I'm currently modeling, and don't have many good male references.
As part of the effort to getting a good foundation for what your modeling is hunting down good references.

You think they told them which category they'd fall into?

I can't get into Lua's syntax. Then again, I dislike most dynamically typed languages.

honestly the unattractive ones are pretty decent looking

dateable for sure

It's all in the head shape. In this picture, I have moved the face of the "unattractive" one on the right has had his face moved over to the "attractive" one on the left. Outside of some color correction, I have made absolutely no edits to his face. He still might not be as handsome, but he still looks alright. I'll try doing the female next.

...

I'm starting off simple with damage formulas for my game (mostly using percentage and multiplication)
do I just make shit up as I add more factors like level of the character?

I don't really know what I'm suppose to look for to be honest. I feel lost doing this on my own.

Did the same thing to the girl. As before, the only thing edited was coloration so that skin tone matches up. I also let her have the other girl's eyebrows to be generous since they can easily be fixed with the slightest application of make-up. Her facial features are nearly identical to the "attractive" girl with the slightest difference in the shape of the tip of the nose.

I have no idea how to do formulas either, I'm basically copying what the game I'm copying in the first place does. What I did was to come up with the range of values I want first (strategy game, units are separated by level, I figured I wanted around 100 HP and 50 damage for the last level, and about 25 for Attack/Defense stats), then make a simple curve out of it while aiming at a somewhat exponential rate of growth.
What you have to realize is that everything is literally relative. Damage depends on average HP range, whether healing/resurrection are common in combat or not, in my case how much the hero's stats influence his army, etc. Figure out what works with what. And look up how it's calculated in games you like, and take inspiration from that.

Ignore detail, work on base form, and use some fucking reference, lips don't look like that, they actually have a very distinct flow to them, really don't worry about detail, your face is just fucked up as is, you need to work on getting the face structure right and then detail it. This is a good one right here, look at how it worries more about sculpting in the basic shapes that the face is made out of and then iterates on them, and the face is stylized to give away easy definition of the structure.


The main takeaway is facial definition, attractive people have stronger facial definition than unattractive people, Oblivion is a good example of this, no fucking definition, no angular features, just potato's all around, and then people end up looking deformed on top of that. Other than the significantly superior facial textures and detail maps the Oblivion Character Overhaul focuses on giving nicely defined and angular faces, ignore the Argonian's and Khajiit though.

Forgot picture somehow.


Also this is the agdg thread, not the please help me be less shit at scultping faces thread.

Gone through all the godot tutorials and messed around a bit to get familiar with all the 2d tools. Not sure how I feel about the scene / node workflow though. I don't like it so far but I am going to blame that mostly on unfamiliarity.

Is anyone here working on any godot games or released any?

I've went through a bit of their documentation. Seems really legit and I like their animation tools.

I'm being a stiffler to myself though and trying to just create my own personal "engines" for games, though. I'll probably use it to create fast prototypes someday.

This is a question about matrices, so how would I apply a translation matrix to a point correctly? so, lets say I build a translation matrix like:

mat4_t translate(float x, float y, float z){ mat4_t ret = { .m = { 1.0f,0.0f,0.0f,x, 0.0f,1.0f,0.0f,y, 0.0f,0.0f,1.0f,z, 0.0f,0.0f,0.0f,1.0f, }, }; return ret;}

so, I have this matrix, and I want to translate my set of points which make up my 3d shape somewhere, right? but then… so I'm applying this thing to each point like this:

inline void apply_mat4_to_vec3(vec3_t *in, mat4_t *m){ vec3_t tmp = { 0 }; //maybe its dumb to pre-unroll but i felt like it tmp.x = (m->m[0] * in->x) + (m->m[1] * in->x) + (m->m[2] * in->x) + (m->m[3] * in->x); tmp.y = (m->m[4] * in->y) + (m->m[5] * in->y) + (m->m[6] * in->y) + (m->m[7] * in->y); tmp.z = (m->m[8] * in->z) + (m->m[9] * in->z) + (m->m[10] * in->z) + (m->m[11] * in->z); //tmp.w = (m->m[12] * in->w) + (m->m[13] * in->w) + (m->m[14] * in->w) + (m->m[15] * in->w); *in = tmp;}

I dont at all understand the signifigance of the fourth row of m when applying it to a point. so, I copied the mathematics off a math site, because I dont really know, what it means to apply a matrix to every point, in terms of the actual math. (I assume this is closer). So, now when I do this, so lets say that im applying this thing to a unit cube, where it begins at 0,0,0 and ends at 1,1,1, right? and i give it a translation matrix generated by translate(0,1,0); which should put it so that its hovering 1 unit above the origin right? well, so I put this in and then run it in the program and I see pic related… so what's wrong? Instead of moving it, its now stretched out so that its 2 units long instead of being pushed up 1 unit. i'm sure there's some really gay reason for this so maybe someone knows why

matrices don't work like that. you have to look up how to apply a translation and all that, they work in specific ways that are basically impossible to just guess.
matrices control all the possible transformations that could be applied to a model, so translation, rotation, scale, but also shear, deformation, etc.

People, I have a small problem that also has caught my attention when it comes to development.

I got House Party and couldn't play for 30 minutes without getting horrible motion sickness, is there any way to make sure motion sickness won't be a recurrent problem aside from setting a decent FOV?

yeah, i know that, I know what matrices do and I know how to multiply them. What I don't know is how you take a set of points and apply that matrix to the set of points because usually the GPU is doing that for me, I Just give the GPU my matrix. I literally looked up how to apply a translation and I got the math for that function. The fact that you are going on about what matrices do (which of course, I already know) and not literally how they work is telling me that you don't know what you're talking about because if you did you would know how to apply a matrix to the set of points instead of telling me the first thing that comes up on google when you look up "how do matrices work"

Ok, so this is fixed, my issue was that I was using the wrong formula, the correct one is pic related. Again, just proving that matrices DO work like this, you in fact do apply it to every point…

anyone wanna help me out with geometry exam tasks
i've got CA=ax(bxa) (x being cross product, a and b are unknown vectors)
CB=b
|a|=1
|b|=2
angle(a,b)=alpha
find AH (height from A to BC) using a and b
i've got an answer, but i'm absolutely not sure if it's right or not
my solution was to find BH and then AH=AB+BH
finding BH by checking the ratios - AH/BH=AC/CH ; AH/xAB=AC/(1-x)AB ; find x and we find BH, find BH and we find AH
my result is that AH=2acos(alpha) - b/2
is any of this correct?
and as a side task we have to find alpha, if we know that |AH|=1
how do i do that?

been a bit, but the last row is for uniforming the coordinates you're handing the matrix; which just makes it a lot easier/versatile (explanation in my first resource below).

For good material on this, as it's a rusty topic for me, I'll recommend a few resources I used in the past.
First covers TRS (translation, rotation and scale matrices), each separately in their own matrix, and also combined; plus relevant formulas.
In addition to a good refresher on matrix math, and other important topics:
solarianprogrammer.com/2013/05/22/opengl-101-matrices-projection-view-model/

Next is just a good general resource:
songho.ca/opengl/gl_transform.html


indeed

Wat. It grows proportionally to the input size. The problem with breadth first is its inability to handle edge weights.

Does the user that was making that game with the anime robo girl still come here?

Please make yourself known if you are among us I need you

Yes I'm still here, this is from my game

Stop

If the JSR/JSRF 3D modeler is still posting or following this thread, and reading this, are you still looking for an animator?

If you are, let's talk.

Wondering about this too. Have they turned his Internet off again?


The memory complexity of breadth-first search is exponential in the length of the solution. The longer your solution is — which, in the case of breadth first search is the same as saying "the deeper it is inside your tree or graph" — the bigger your fringe gets; and it does so exponentially. In contrast, the fringe of depth-first search grows linearly. However, not only is DFS not optimal (doesn't guarantee to give you the best solution), there's no guarantee that it'll give you a solution at all (i.e. infinitely deep trees or procedurally generated worlds).
It's the entire point of Dijkstra's algorithm (called uniform-cost search in AI). It's a directed breadth-first search which means that it hopefully keeps the fringe small enough. You can see that wonderfully when there is no solution. When you try to find the path to an unreachable node, UCS behaves exactly like BFS as far as complexity goes. Best-first search and A* (which is merely a combination of uniform-cost and best-first), have the same problem.
It's something you need to keep in mind when working with big worlds and especially if your game is procedurally generated. If you try to find the path to an unreachable node and just naively let it run "until it gives you the solution", it will crash your computer because you'll run out of memory fast.

Sure, but it's bounded by the size of the graph since eventually you run out of nodes. (Besides on a 2D grid you wouldn't get exponential growth regardless.)

Not quite the entire point since dijkstra can also handle edge costs unlike BFS.

Fair enough, unless you're dealing with procedurally generated worlds which are "in" right now. Then your world is, at least for all practical purposes, infinite in size and you need to take that into account.
Either let the pathfinding run on another thread or, if that's not an option spread the calculation out over multiple frames in the main thread. Also add an upper path length after which you stop following a path.

Using Unity.

I'm at the point where I need to get a bunch of functions that will be used by not only my main character, but also for enemies. Stuff like checking for grounding, checking for slopes, etc.

I was thinking of putting all these in a separate script that would be accessible to everyone. Is this a good idea?

Can you make them static utility functions?

For the ones I have need of right now yes, that was the plan

Go ahead then.

I'm fairly sure this wasn't happening the last time I was working on this

forgot to freeze constraints, oops

I'm using it for LowRezJam

is this a hobo with a shotgun game?

Looks more like Adolf Hitler with a super soaker.

please continue making that game
it looked really cool

...

Splat88n?

oh fuck right off

Dudes a bit pissed off with the whole Automato dicksucking that has been going on so i wouldnt bet on it.


Please come back

The real question here is: should you even post your progress if you're working on something good?
Setting aside whether the story is even true or not, it is plausible, it could happen, so why risk it?

I sent him an email, we'll see

Hey can someone help me with something completely unrelated in JS?

I found this nice dialogue system, but unfortunately the guy doesn't really explain how it works. I reconstructed it up to the point where I have to make choices, but I don't know what to do now.

gamedev.stackexchange.com/questions/54238/how-do-i-implement-branching-dialogue-in-javascript

var current_step;var story = [ { m: "Hi!" }, { m: "This is my new game." }, { question: "Do you like it?", answers: [ { m: "yes", next: "like_yes" }, { m: "no", next: "like_no" }, ] }, { label: "like_yes", m: "I am happy you like my game!", next: "like_end" }, { label: "like_no", m: "You made me sad!", next: "like_end" }, { label: "like_end" }, { m: "OK, let's change the topic" }];function execute_game() { var current_line = 0; while (current_line < story.length) { current_step = story[current_line]; if (undefined !== current_step.m) { display_message(current_step.m); if (undefined !== current_step.next) { current_line = find_label(current_step.next); } else { current_line = current_line + 1; } } else if (undefined !== current_step.question) { display_message(current_step.question); display_answers(current_step.answers); return; // display the question: current_step.question // display the answers: current_step.answers // choose an answer // and change current_line accordingly } }}

Missed a part

function display_message(varStore) { document.getElementById("container").innerHTML += varStore + "";}function display_answers(varStore) { for (i = 0; i < varStore.length; i++) { document.getElementById("container").innerHTML += "" + varStore[i].m + ""; }}execute_game();

he has actual proof that he didnt rip of near a tomato and gets a few bucks on patreon, because he posted progress on tumblr

he also inspired me to pursue this hobby further because his stuff looked so awesome, so i would find it really sad if he quit

Why don't you just write your own? Branching dialogue systems are not terribly difficult to make by any means (I made a decent one with zero previous experience and I'm a fucking artfag that only codes in shit like python). If you don't understand what you're writing there is absolutely no point in using it. Better to write your own that you fully comprehend what you're making.

I understand that code perfectly I just have no idea how to make a dialogue system

I figured it out by the way

He hasn't an hereo'd yet, I have him on steam.

For comparison and consideration.

can a man really mentally recover after sperging out like he did here?
not only is it archived but it will forever haunt him

just don't remind him of it or mention nier around him if you want to see him get better

We're all autists here, who gives a shit

post link, I missed this

What stupid chink thought this was a good idea? I have seen a fair number of these fantasy element tables posted here and this has got to be the stupidest one yet. Fucking WOOD. Looking at this just makes me furious. And why in the incestuous fuck is Metal not a subset of Earth? What braindead inbred dumbass came up with this shit? I am legitimately mad at how retarded this is.

tldlr is that tetrachrome dev is mad that because nier came out, people will see his game as a neir clone, even though he has been working on his game since before nier was announced. Also he is convinced that nier has robowaifus in it because one of the developers saw his game on Holla Forums and copied the idea for his game. Following a post about this on some random thread and then 20 replies (or some minor shitstorm) about how he has schizophrenia for suggesting it.

Me on the left.

I assume Chinese didn't grasp the concept of metal coming from earth until science kicked in, same with wood.

best I got right now is his pasta

he reminds me of my old self, he married his persona with a single project, if it flops hie might kill himself, luckily for me someone here snapped me out of it.
near a tomato or not he needs to take himself less seriously and join in on the bants otherwise trolls are going to feast on him

Is that it? It's pretty cringey ass-pie shit but it doesn't even come close to the depths of turbocancer that cuckdev sank to. I've been a sperg before and it's shameful but this is honestly pretty tame.

Actually that's exactly what it's stating.
likewise the disruptive reaction

I know he's been working on that thing for years now, but I wonder if there could be some truth to what he says.
You call him a sperg, but the man put his whole heart into that project. Can you really blame him?

Wait what

A metal bucket or ladle.

...

It's understandable to be pissed if you've been working on a game for the last few years and some well funded studio comes along and makes a similar product in a quarter of the time, but sperging out about it is still acting like a dumbass. It's petty and childish, especially considering the similarities between NA and his game are rather tenuous at best.
That's…. pretty much it. 2B and his android waifu look nothing alike.

I hope tetrachromedev is lurking because there is a very important concept here that he needs to understand:
Your idea of a white haired android girl fightan DMC style is not what makes your game unique or interesting. Even if Automata did "take your idea" it doesn't fucking matter. All ideas for games are utterly worthless. You know why we mock relentlessly mock ideaguys? Because ideas are worth NOTHING. You know what is actually worth something? Execution. You know why people are interested in your game? Because it looks FUN. Not because the girl is an android, or because she has white hair, or because she fights robots. That does not fucking MATTER. You say NA sucks. Good. Let that be your driving motivation to make a better fucking product than they did. Because that is all that matters.

Got a little progress done.

What is this place?

anyone have any tutorials on making simple outfits like this in Blender?

I do it in Maya.

Whoever modelled that clearly has no concept of good topology. But you can make clothes like that by drawing curves as a guide and modelling a mesh over it.

just added split screen to squirrel sphere. will allow 1-4 players to play and compete

user you can't tell what the topology is by just looking at it in a solid material. And no. Hand modeling grooves is shit advice. Make a flat dress and paint weights for pinning. Let physics do the hard part for you.

Working on a commission, I just wish rigging was easier sometimes.

San Miguel de Allende.

what does shit cost these days?/ how can i ether get in or buy from

What do you mean? Maya or a mesh?

Yes I absolutely can, because smooth shading is not used on 90% of the model.

What the fuck are you even taking about? Using physics to generate a lowpoly model? Are you a fucking idiot? A competent 3D artist should be able to model anything by hand; especially something as simple as a frilly dress. I'm going to put you down as, "utter dumbass who has no god damn clue what he's talking about".

I would like to know how to make frills manually.

You actually can, because on the left they're using flat shading rather than smooth, which makes their excessive use of triangles (which are to generally be avoided for good topology) very obvious.

well, now it's back to as much features as it used to have but on GPU rather than CPU, so I can plot a whole lot of these balls rather than max 6-10 like before, although I still want to improve it more, since it still runs at only 1000 fps

What do you guys use to make design docs?

A pen, marble notebook, and bourbon. Are you asking what to write in one?

Nah, I got tons of shit to write, but my handwriting is trash so I was hoping for someone with a pretty looking open source program. I think libre office or OOO probably has something but I wanted to see if anyone knew of something just for it. Thanks though, pen and pad are classic I just wish I weren't fucking left handed.

I mean I also use CherryTree and Sublime, but I really just prefer pad and pen. Note programs keep distracting me because I'll get hung up on fixing typos and it breaks my train of thought.
t. also left handed nodev.

No you really can't. They could be using smoothing groups, modifiers, or separating each limb mesh as far as anyone can tell. The head, torso, and limbs look to be a separate mesh. You both have poor judgement.

Use clo3d

I'll check those out, thanks. Also cheers to my fellow lefty honestly my handwriting is shit for mostly personal reasons.

nice fellow beaner

i know its kinda rude but, was wondering how much models / textures cost

Correct, but you can use physics on highpolys and make a lowpoly based on the highpoly outcome. Pretty much no one in any AAA studio "handmodels" clothes these days, they use shit like marvelous designer coupled with zbrush to create a detailed highpoly cloth, and then either use some addon to generate a lowpoly based on it or do a quick retopology, or get their prepared basemesh as close as possible to it, and then you bake a normal/displacement map.

I mean I suppose it would be a perfectly viable method to retopo a lowpoly dress from a physics deformed hi poly, but to me it seems like a huge waste of time when you could quickly just model it freehand. Though if you don't know what you're doing it would be a good exercise to learn how a lowpoly frilly dress is shaped. Typically the best idea though is to just look at a reference of what you're modeling and sketch it out for lowpoly.


You are so fucking clueless about what you're trying to sound informed on it would be funny if you weren't being serious.

Depends on the artist and the assets, I mostly do meshes for commission, not exactly good for texturing yet, in my case it can be only the mesh with UV maps, human like meshes can be auto-rigged on Mixamo's page, is not optimal but gives you something to work with, clothes are sometimes harder to rig, I have gotten commissions for around 120-200 bucks.

so 120 200 for what exactly like a character/w clothing? Outfits?

Right now, rigged clothes with base textures in semi-low polycount plus PhysX setup, not to mention I keep fixing stuff when needed, it basically goes like a paid job since I get those in a monthly basis as long as there are assets to be made, at least that's our current deal.

so working with one company/guy?

Not a company, just the guy right now, building up a portfolio so to speak, well, more or less.

Good to see there are still devs who future-proof

How do you make any sort of attached UI look good in UE? Things trail and lag behind, they look like they're refreshing slower than the rest of the game.

What site do you use to find foolish goyim devs looking to hire 3D artists for commission work?

I found my client here, he once paid 2000 bucks for some horribly sculpted monstrosity, and he knows I always try to improve where it counts, if I find a way to update my old stuff to work on his game I will.

I love Gentoo as my primary OS, but holy shit this stuff only happens when I actually want to get work done.

that is why i always keep a working OS on a separate drive

But Yoko Taro games have had robo waifus since at least Drakengard 3. There might be some in the previous games but it's almost 1am where I am right now and thus I'm too tired to remember anything from Drakengard 1 and 2

Explain scenes again to me. They're basically a template of connected nodes, yes?

Is it better to think of theme as "templates" or as "rooms"? It can be instanced either way, I know that much. Nodes cannot exist in a vacuum? They need to be attached to another node or a scene?

Also this was neat

What do you mean by "attached"?

Templates, since a scene can be a room, a character, a single projectile, anything.

This is the latest work on my editor, of course you can see its buggy, because i'm still sorting out my matrices. So, what's going on is that I'm taking a unit cube and then applying both a translation and scaling matrix to it, which should match up to the points tracked by the mouse, to create new brushes. However, this still has a lot of bugs, but the code is so much smaller than If I wanted to derive these things into triangles by hand in my code (and also wouldn't scale to other primitive shapes), so this is great for me so far, because I like doing this the correct way this time. The bugs are just from the input used to build the matrices, and not the process itself, which you can see working.

look into shinkwrap modifier if you want to do a more manual modelling approach. cloth simulation could also work; someone wrote a script that mimics Marvelous Designer's behaviour, where you create fabric patterns and then sew them using the cloth sim.

Stat rolls for a porn game
Yes or no?

yes

...

No bully, it's placeholder text to see if stuff works

Protagonist is probably gonna be an elf

If it's an elf, why would there be a no button?

That's part of the fantasy, of course. Why must you bring realism into the realm of fantasy, user?

Nice ad hominem. You're still wrong.

More fun that way, also rape's hot

saved.


ok, i took a day or two just to set up a plan for what i want to do, i just want to keep in mind that i do have a concept but i cant draw for shit, this should explain why i've been making so many novice mistakes and problems. but thank you all for the help, so how does it look.

ay lamo

???

Death Stranding will be an unfinished mess just like Phantom Pain.

It seems like most video game icons are George Lucas types. They have lots of ideas, but they need the right team to filter those ideas into a cohesive, fun game.

MGSV is fun.
The only problem with it was konami.

Like a health bar above an enemy

This seems at odds with himself. Either something was lost in translation or he's just losing it.

Did a lot of refactoring, had some fun bugs.

do yourself a big favor, learn to draw a bit, learn some anatomy proportions, form, then jump into sculpting. It'll make everything a lot easier on you and you can deliver some quality stuff in little amounts of time.

also, new thread soon?

PAGE THIRTEEN YOU NIGGER

11 is basically 13.

Not to the mods

this is true.

What about it?

what was that blender plug in to preview library objects?

(checked)
Once this thread hits page 13 a new thread can be made.

user, please, correct the face proportions!
As I've mentioned here:


>Remember that the shape of the face is highly reflected by first the skull

>Areas to improve next should be first the cranium shape (everything flows based on this), jaw line/chin

>I'd also work on the facial proportions

Anything else is just extra at this point, as everything flows based on this.
Learn anatomy as this user mentioned it will help you immensely.

part 2, as to not have the below point occur
do not plan (unless you mean gathering references), just practice, and utilize anatomy references + anatomy courses

that's ok, practice, BUT start by crawling… you're attempting to run (e.g. focusing on sculpting the muscles/fat before learning the fundamental structures of the skull).
Learn anatomy as you're scuplting.
So, you'll learn that you should know FIRST the skull/bones (thus the layout of proportions, like the eyes, brow, nose, mouth, cheekbones, etc etc), then the muscles, then the fat, and then how the skin conforms to this.

as mentioned, you're attempting to run here, when you should be crawling.
It's the fault of the approach you're using, and not necessarily your skills at drawing.

happy to help, and forgive my long windedness; I tend to lose the main point when I start on my rantful posts (why i've done a 2 part post as to attempt to remedy this).

Proportions make it look wrong, but the form of the lips is getting there; though just focus on proportions please for your own sake.

Mentioned threads ago I was doing this. Thought I might as well show today's progress since I finally made a running animation I like (even if it could use more polish, I'm calling it done for now- prioritising making a full usable sheet over a perfect sheet). I don't have the patience to finish the jumping today though.

currently reusing the rectangle format, not deterministic, and large lag spikes will make you miss grass. gonna replace with a arc cut that saves the last position and cuts to the current position, so even if not a single frame was spent in the cut animation, it will get it all

Always nice to see vivi in a game

Finished the event system, but since its all code i don't have anything juicy to show, instead i will post the icons my artist made this week

I hope you didn't overpay that guy to google search images and turn them into pngs with white background

He sends me part by part for feedback, so i know he is actually making it, they do look like cropped photos, but they are actual drawnings
Also i choose the references myself

just fucking with ya
I could tell as much from the copy paste cup/frying pan
stuff looks good
what're you working on?

cooking for faeries and shit
>>>/agdg/27922

Ok now start again using that picture you saved, don't look at our old work until you think you are finished, at this point you are polishing a turd.

...

fix*

It really depends on the usage case for which is faster (more efficient, less work done).
Each has their intrinsic pros, but also their own cons; balancing + tailoring your scenes to these results in either choice being faster than the other (omitting implementation of the rendering pipeline, and only approach it from the highest abstracted level).

At least in theory, forward rendering is faster for renderings that involve less geometry.
Deferred rendering is faster when rendering absurd amounts of geometry that only AAA companies could ever put together in the first place.

For a lot of indie 3D projects, I think forward rendering ought to be faster.
But then again, an indie project will may waste millions of tris on a flat surface or something.

Could it be that deferred shading is easier on the CPU? My system is CPU bound i think, i have an i5 2500K at 4.6ghz and a gtx 1070

Profile it, and also look at the fundamental differences between forward vs deferred shading (easy enough to find); then it'll be easy enough to deduce why one has better performance compared to another.

Here are some links, but it's for Unity; though it applies in principle just the same:
docs.unity3d.com/Manual/RenderTech-DeferredShading.html
docs.unity3d.com/Manual/RenderTech-ForwardRendering.html

You are like a little baby

Very nice! You thinking of releasing her spritesheet as CC0 when you're done? I totally wanna use this.

The purpose of deferred rendering is to ease the load on the GPU. The theory behind forward vs. deferred rendering is concerned with GPU performance, not CPU performance.

Forward shading means each object has to go over each lighting source for every fragment (a fragment is a pixel-candidate, it might become the final pixel or might be discarded for some reason), it's a less efficient approach that proves problematic with many light sources and/or lots of geometry.
With deferred rendering, the fragment in the final render pass buffer becomes the pixel rendered to the framebuffer/screen, period, and you're only iterating through the light sources once to shade everything during that final pass.
This is more efficient GPU rendering when working with large scenes with many lights and lots of geometry, but it's overkill and actually slower than forward rendering if, say, you have few light sources and relatively little geometry.

The above is the theory.
In practice, deferred rendering will be harder on the CPU when using OpenGL, due to the extra rendering pass it requires (More gl* commands -> more CPUGPU talk and/or more waiting on GPU to do something).
On Vulkan, I don't think you should see a significant increase in CPU usage- you can describe all of the render passes to the driver using the VkRenderPass, VkSubPass data types with vkCreateRenderPass().

This page might help if you're looking for more details on the subject.
learnopengl.com/#!Advanced-Lighting/Deferred-Shading

Actually, I shouldn't say "will be harder on CPU when using OpenGL" with such certainty.
Modern OpenGL drivers are highly optimized, and try to second-guess you every step of the way to try and reduce CPUGPU talk and CPU waiting on GPU.
However this sort of stuff going on behind the scenes is largely (and intentionally) undefined by the OpenGL specification, and so your mileage will vary on OpenGL based on what GPU/driver you are using.
As Vulkan is a much more explicit API, you'll get much more predictable results/performance, another benefit of the newer language.

Step it up