/v/ - Video Games

it's fucking video games again, baby

New Reply
Files Max 5 files32MB total
[New Reply]

What a nice board!
Neon Genesis Evangelion is utter garbage >>>/a/

(78.5KB, 462x496)
Work on your game instead of arguing edition

>#8/agdg/ via irc.rizon.net
>https://matrix.to/#/+agdg:matrix.org via matrix programs
>Wiki: http://8agdg.wikidot.com/
>Dev resources: http://8agdg.wikidot.com/resources
>Previous bread: >>6917
Replies: >>32822
(14.8KB, 200x365)
(624.6KB, 793x1144)
(4.1MB, 540x360, 02:29)
>>32810 (OP) 
>Man with no games
What do enginedevs typically do for text? Are there any libraries that people tend to use? I mean you can't just supply a character spritesheet that you drew in ms paint if you want to localize for different languages and get millions of chinkbucks.
I think FreeType is the go to library for font rendering.
Replies: >>32827
(293.4KB, 580x548)
>The freetype libraries use the Universal CRT and therefore requires Visual C++ 2015-2019 Redistributable to be present on system
Replies: >>32906
I think I used SDL_ttf last time. It doesnt have glyph caching though.
Thread might get shoah'd because the old one hasnt hit post limit.
Not sure but I have seen textures that have the used characters on them so I guess some devs just map them to their respective glyphs.
Replies: >>32832
But you still need to get glyphs for thousands of moonrunes from somewhere and to render them into the texture. That means you need a way to read font files and a way rasterize the glyphs from those files.
Replies: >>33237
(97.2KB, 908x765)
What the fuck am I looking at
This library is bigger than all of my projects combined.
Replies: >>32838 >>32840
Because it contains 2.5MB each for OSX and Windows related IDE shit, and 15MB for freetype as a dependency. Font rendering is a fucking disaster.
(68.8KB, 475x464)
Let's lock the thread at >>6917 manually then. It worked for the WEBM one, but maybe because attaching WEBMs to our spam was somewhat pertinent to the thread's topic. I'll have to get creative.
Replies: >>32841
(8.9KB, 224x186)
The entire source is in SDL_ttf.c/h though, the rest is docs, makefiles, project files etc.
Replies: >>32903
If the mods have a problem with this thread then they can either suck my dick or start responding to this: >>>/meta/1585
(44.2KB, 749x694)
>notice my map is too small
>make it bigger
>the MCV doesn't follow the path anymore properly
>check the print logs
>angles are in negative values once
God damn it, instead of spending time getting the shop system to work now I have to spend time to fix this stupid function. Why the fuck are negative angles a thing or even allowed in the first place? Trying to use tricks like +90 degree or 360 doesn't help. 

Here is the nigger code:
language: python
func rotate_to_target(_delta, target):
		var old_angle
		angle = atan2(target.y - position.y, target.x - position.x) 
		angle = rad2deg(angle)
		#angle += DIRECTION
		#Reset the angle value before it gets too big/small
		if (angle < 0):
			old_angle = angle - (angle * 2)
			angle = int(angle) % 360
			angle += old_angle * 2
		$Idle.set_frame(AngleF.Angle_To_SpriteN(int(angle), ANGLE_SPRITE))	
SDL_ttf.c is only 63 kilobytes so I figured there's no way that contains everything from font file parsing to glyph rasterization and whatever other sentence wrapping they might have, but I guess SDL's ttf library is mostly just a wrapper for freetype. I'd honestly rather make that kind of wrapper myself so I might try to extract their freetype stuff out of it.

The first calculation should return a value roughly between -180 and 180, so doing this afterwards doesn't work?:
if (angle < 0):
	angle = 360 + angle
Maybe it's a problem from going 360 or above in Angle_To_SpriteN?
Replies: >>32907 >>32908
Why wouldn't negative angles be allowed? Also what the fuck are you doing in that angle conversion, a - 2a is just -a.
Replies: >>32907
Negative angles are really nice. atan2 returns the radians from pi to -pi, so after calling rad2deg(atan2(y,x)), if the angle is negative, set angle = 180 - angle. This will give you an angle from 0 to 360 degrees.
Replies: >>32908
(38.5KB, 807x280)
Are you sure about that.png
There is also stb_truetype.h for a single-file self contained thing,though I never used it so i can't personally account for quality.
Replies: >>32911 >>33239
(100.2KB, 1024x768)
(309.6KB, 1920x1007)
Angle_To_SpriteN excepts a value between 0 and 360 so that it can correctly return the value of what sprites to render depending on the actor angle. If the angle value is negative or larger than 360 then it cannot return the sprites number to render. 

>Why wouldn't negative angles be allowed?
Then how the hell do I get correct sprite angles from negative angle values?

Also I took a look at this function: follow_target(_delta) https://paste.boxlabs.uk/9416 
and I noticed that I fucked up the position calculation somehow, as Line2D is not drawing the lines properly of where the path nodes are.  I tried to take account of the actor position and without it and nothing helped to narrow the bug further.
Replies: >>32910
That's not right, my bad, >>32903 looks right.
Replies: >>32911
(280.1KB, 990x250)
>Then how the hell do I get correct sprite angles
Angles(in degrees) are constrained between 0 and 360 (or -180 to +180) by concept;
Negative Angle to Positive : 360 + negAng (360 - abs(negAng))
Angle larger than 360 : bigAng%360
Come on dude,think a bit.
Replies: >>32932
I just went to their github link and that's what it said. Maybe I wasn't reading hard enough about where I was going.

angle = 180 - angle does work if you remove the if (though it'll be rotated 180 degrees), but floating point niggery may still make it end up at -0.000002 or something.
(6.9MB, 800x608, 00:23)
Thanks this formula works better then the other one. Welp I got the stupid MCV to move almost correctly to the path now as shown on the shitty video (don't know how to fix it), I have no idea why on certain points it rotates "smoothly" all of suddenly instead of doing full rotation, perhaps because the negative angles are kicking in which screws up its movement a bit. Now I can finally work on making the stupid MCV more functionally by working on the "buy items" system.
Replies: >>33038
>CnC clone in Godot.
Are you looking at CnC source code or OpenRA?
Replies: >>33039
(15.9MB, 640x368, 03:51)
lmao no I am not going to make a CnC clone because I am pretty sure if I get to write a pathfinder it ends up becoming worse than TA: Springas. I just like the concept of having some strategic element, something akin to Uprising or Battlezone I suppose. Besides the game is going to be about being a french tonker vs computer monsters, I need some good sources of monster models/sprites that I can rip off which is not ayy lamo/hell themed, that's why.
>Are you looking at CnC source code or OpenRA?
No, but I did made a mod for OpenRA before I abandoned it because the lead dev said it's impossible to retain sprite detail when its scaled down he said something along the lines of how their pixel buffer works, so according to him I would have to upscale all the tiles per hand instead.
Replies: >>33676
(261.4KB, 1072x1024)
Have I finally gone too far into enginedev?

Does anyone have any experience working with microcontrollers or other weird small computing devices? What's it like? I'm unironically interested in making my own tiny computers and programming little tamagotchi-tier games into them and seeing how far I can take it.
Replies: >>33081 >>33085
I have an fpga and it has had exactly zero designs uploaded to it because figuring out the pipeline and how things are supposed to go together is a giant arcane pain in the ass - That is as soon as you don't want to use the gated IDE that does the evaluation, place and route and upload correctly and everything else worse.
I have an associate's in electronics engineering and I can say that microcontrollers are easy if you have some basic coding knowledge and have some idea of how to wire a circuit. I made a maze game and a falling rocks game that was displayed on a cheap 8x8 LED matrix and they were essentially completely driven by binary math.
Now, actually engineering it to look like a tomogatchi and not a bunch of shit slapped on a bread board is the real challenge.
Or just use a bitmap? http://unifoundry.com/unifont/ has characters in all languages with a constant width and height.
You can map each phrase to an array (or similar structure) of chars/ints, and map each char/int to an area in the bitmap above.
Replies: >>33239
That's a very interesting link, but the characters are very small and pixelated. They'll only work in very limited ways.
stb_truetype as mentioned by >>32906 seems promising anyway.
Replies: >>33352
You can make your own bitmap, as long as you set fixed dimensions for every character. The link I posted was just one example of doing it... Anyway glad you found something for your use case. I always keep stumbling on those stb libraries but I never tried them, so do report back how it goes.
I followed learnopengl which used freetype in it's (non-working) example. I worked through it carefully and fixed into a workable state. It's pretty inefficient due to a naive implementation. But I took the time to understand why, and commented that pretty fully in my implementation, including suggestions that could fix it. At the time I was already behind in my schedule so CBA to go back and fix it. But, it could be done for anyone who care to work it out.
FreeType & OpenGL3.3 on Linux
(319.5KB, 618x527)
What are some methods I can use to immediately determine the distance between two points? I'm trying to make a projectile with an instant travel time, and the method I'm using right now isn't working at all
>fire a tracer round at very high speed, if it collides with anything it sends back the coordinates it collided at
<it travels so fast the game can't update it in time, so it either travels through an entity without producing a response or until it's three tiles into a wall
<if it goes off-screen it doesn't send back any collision info
<setting it at a lower speed slows down travel time and is not instant like I want it to be
Replies: >>33409 >>33432
Remember that old dead dude Pythagoras, Anon?

language: c++
#include <cmath>

struct Point {
  double x;
  double y;
  double z;

int main() {
  Point point_A{.3919023, 6.12837912, 10.391283};
  Point point_B{.3919023, 6.12837912, 10.391283};

  auto distance = std::sqrt((point_A.x * point_B.x) + (point_A.y * point_B.y) +
                            (point_A.z * point_B.z));
Replies: >>33403
My bad, forgot to mention how to find the distance between those two points without knowing what one of the points is. I can figure out where it was fired from, but not where it's going to hit.
Replies: >>33432
Then you'd need to use a little calculus and trig to create a ballistic trajectory curve, then just test a point on that curve at whatever distance, at whatever time. BTW, there's an obvious bug in my previous example I was wondering if you'd spot it. :^)
Replies: >>33411
(21.7KB, 118x129)
Does your vidya has cute girls in them, devanons? Can I draw lewd pictures of them?
Replies: >>33406
the DLC will :^)
don't worry, it'll be free
(37.6KB, 832x891)
You're not supposed to shoot anything, you just do a ray cast or ray trace or whatever it is. I've never needed it myself so I don't know exactly how to do it though.

I vaguely recall doing something like taking the angle of your ray, rotating the world so that the ray is horizontal, and then you check if the target object overlaps at the y of the ray's source. I don't know if this is the correct way to cast rays though nor what I did this for.
Replies: >>33432 >>33541
>BTW, there's an obvious bug in my previous example I was wondering if you'd spot it. :^)
main didn't return zero, and the formula is wrong, it should be sqrt(  (B.x-A.x)^2 + (B.y-A.y)^2 + (B.z-A.z)^2  )
Shame on you writing incorrect code on purpose and stifling anons who are already confused. Get a life.
Haha, fair enough. You can't learn if you don't learn to think. Puzzles are good things for that. Here's the correct code:
language: c++
#include <cmath>
#include <iostream>

template <typename T>
auto sqr(const T val) {
  return val * val;

struct Point {
  double x;
  double y;
  double z;

  Point operator-(const Point& rhs) const {
    return Point{(x - rhs.x), (y - rhs.y), (z - rhs.z)};

  double square_sum_axes() const { return sqr(x) + sqr(y) + sqr(z); }

int main() {
  const Point point_A{0.3919023, 6.12837912, 10.391283};
  const Point point_B{0.3919023, 6.12837912, 10.391283};

  const auto bogo_dist =
      std::sqrt((point_A.x * point_B.x) + (point_A.y * point_B.y) +
                (point_A.z * point_B.z));
  std::cout << "bogus distance: " << bogo_dist << '\n';


  const auto delta_AB  = point_A - point_B;                      // get deltas,
  const auto real_dist = std::sqrt(delta_AB.square_sum_axes());  // then sqrt()

  std::cout << "real distance: " << real_dist << '\n';
It's a nontrivial question, very much dependent on your usecase and existing implementation. If you're using some shit like Unity you just do a raycast from the point A where you shoot, to point B at location A + Normalized_Dir * Some_Large_Float. Relatively retard-proof.
However if you don't use a pre-existing engine, there's a lot more to consider before anyone can really give you an answer. Is it a 2D or 3D game? With box collision? Polygonal collision shapes? Implementing ray collisions isn't a case of just copying some formula off stackoverblow. Unless you've got  a relatively trivial number of colliders, you'll have to do some prepass to not grind performance to a halt when you do a raycast. Could be something like a BSP for 2D, or a BVH for 3D. And then you might end up needing to implement collisions for any combination of line-line, line-square, line-circle, line-cube, line-triangle, line-sphere, etc. depending on your usecase. I'd advise you to just look up a library for this shit at that point.

tl;dr Just lern 2 raycast. It might be a rabbit hole, hope you can use some shortcuts.
>you don't need advanced maths to make games
I now realize that this claim is a lie. Every time you look up a way to do some basic thing, the solution is given as some kind of unreadable math equation. Unless you know how to read math moonrunes, you can't translate it into code.
In all fairness that isn't completely false. Not only because game engines exist, but also because math libraries exist and you don't have to implement every mathematical function you want by hand. You'll just need to learn how to use a given library, rather than how it works internally.
Replies: >>33458 >>33467
Well I mean, with that argument you might as well say that you don't need to learn to program at all to make games.
Replies: >>33467 >>33506
Trigonometry and linear algebra aren't exactly advanced, so it's completely true.
Replies: >>33460
(50.7KB, 1274x775)
(79KB, 1380x907)
If someone's concerned about having to be good at math before making games, they're probably thinking about pic related. I mean I get that trigonometry isn't that hard, but when they start throwing this shit braille at you it's like you're looking at an alien language. And I'm saying that as someone who understands trigonometry and to some degree even vector maths in general.
>managed to use Tsiolkovsky rocket equation without any issues
>half of this just looks like gibberish
>rocket science is easier than game dev
Replies: >>33506
It depends on what you consider advanced.
Trigeometry is something you should've had in school, and vector and matrix math is mostly an applied version of that.

It's just that when you look around the internet; most material is written at university level. If you haven't learned how to read that; it's hard to decipher.
The other problem is that the moment you veer into advanced territory, a lot of gamedevs go "just copy this and roll with it lmao". Getting a proper grasp on quaternions has been quite a pain because of that.

If you're looking to learn vectors & matrices; wolfire got some nice write-ups on that: http://blog.wolfire.com/2009/07/linear-algebra-for-game-developers-part-1/
Replies: >>33494
Actually, most things in vidya's engine-tier ops are constrained to just a handful of types of operations. Matrix transforms, for example, being one of the most common. But using GLM is one good way to avoid having to be a maths wizard, yet still manage to create efficient, relatively simple to read/write code. 

Here's a complete camera implementation in like 7 lines of code that will be as fast as they come, and will be compatible with just about every kind of system out there.
language: c++
#include <glm/vec3.hpp> // glm::vec3
#include <glm/vec4.hpp> // glm::vec4
#include <glm/mat4x4.hpp> // glm::mat4
#include <glm/ext/matrix_transform.hpp> // glm::translate, glm::rotate, glm::scale
#include <glm/ext/matrix_clip_space.hpp> // glm::perspective
#include <glm/ext/scalar_constants.hpp> // glm::pi

glm::mat4 camera(float Translate, glm::vec2 const& Rotate)
	glm::mat4 Projection = glm::perspective(glm::pi<float>() * 0.25f, 4.0f / 3.0f, 0.1f, 100.f);
	glm::mat4 View = glm::translate(glm::mat4(1.0f), glm::vec3(0.0f, 0.0f, -Translate));
	View = glm::rotate(View, Rotate.y, glm::vec3(-1.0f, 0.0f, 0.0f));
	View = glm::rotate(View, Rotate.x, glm::vec3(0.0f, 1.0f, 0.0f));
	glm::mat4 Model = glm::scale(glm::mat4(1.0f), glm::vec3(0.5f));
	return Projection * View * Model;
This isn't really all that hard to follow once you learn the basics of programming in C++ correctly. Just lay the groundwork you need anons, and learn to roll your own using OpenGL + GLM. My engine runs at 60fps on my 2-core potato box w/ integrated Intel graphics.
(287.7KB, 500x500)
The claim is only a lie because the idea got truncated/mutated from being relayed too much.
The full thing is that Programmers don't need (advanced) math in the same way that other disciplines do , being able to solve mathematical systems and other such things aren't useful to a programmer,what someone writing code does need to be able to do is to understand,compose and restructure things using mathematical concepts(so yes,you do need to be able to read the mystical math glyphs,their application on the other hand is immensely less important).
There is no "advanced math" course/textbook/material that teaches these things as they are all focused on cramming memorization into the students heads(literally the worst form of learning) unless you go up to the very highest level,at which point the things discussed would be way too abstract to be useful/assume too much intermediary knowledge.

Thus "Advanced Math" as is referred to here is not a field/topic that provides what you need or a reasonably proper foothold to something that does.
"Advanced Math is not what you need"  -> "Advanced Math is not needed".
Replies: >>33474
This looks like wikipedia and at least from my experience wikipedia is fucking dogshit when it comes to actually presenting useful information about a technical subject. It's likely a couple autists who try to outdo one another with how esoteric you can present something.
Replies: >>33474 >>33482
Oh yes, wikipedia is absolutely useless in explaining anything that has to do with pure math. You can use it to quickly look up a formula you forgot if you already know all the theory around it and where it comes from but don't expect to learn anything.
If you want to know more about stuff like this, so probably vector and matrix problems DO just pick up a good book for it, unlike this anon >>33467 says, there are in fact good resources to learn about this because it is also a topic relearned and heavily needed in early university and for anything at uni level you will be able to find resources that don't just rely on memorization like school does and hold you by the hand quite well, because universities assume that you learned absolutely nothing in school other than the very basics and they are usually right.
Replies: >>33476 >>33483
(51KB, 608x625)
>you will be able to find resources that don't just rely on memorization like school does and hold you by the hand quite well
/thread. One of the very best I've ever found is the online textbook Immersive Linear Algebra. The diagrams are interactive, and the approach is both hand-holding and methodical. 

You can easily jump to just the areas you're currently interested in as well. Eg, here's the chapter on matrices:
Replies: >>33483
Take it from a mathematician: Wikipedia sucks dick for mat-
>It's likely a couple autists who try to outdo one another with how esoteric you can present something.
It's mostly just the phenomenon of too many cooks + incompetent authors. The Stacks project has a similar issue despite being written by experts, but at least nobody is seriously suggesting to use it for anything other than a reference.
Replies: >>33485
(1.3MB, 1080x1331)
I'm not saying good resources don't exist,I am saying that the ones found under the umbrella of "advanced math" are not correctly targeted. The closest to the actual stuff needed would be under the theoretical/discrete analysis type of sources.
The vector and matrix problems might be taught really well by the sources out there (for example I second that >>33476 is a really good resource for that stuff) what I want to point out is that while they can be great help the focus of most math learning materials is not what is actually needed for what the workflow of a programmer. A mathematician and someone working in the field of (not entirely theoretical)-computing just have different requirements in regard to their understanding of math.
Being able to do linear algebra real good is not actually directly relevant to using it effectively in the context of a model or program.These skills can develop concurrently but that is not a designed property of the way math is taught in tertiary schooling institutions. 

.t I dropped out in the third year of uni cause they were aggressively refusing to teach anything that actually relates to my chosen field.Expect my 98 page treatise on how compsci/compengineering curriculums need to be completely redone worldwide soon,
Replies: >>33488 >>33506
>It's mostly just the phenomenon of too many cooks + incompetent authors
Not saying this isn't the case but I personally know of a professor who tried to correct an arcticle and found out the hard way that wiki pages are oftentimes the fiefdoms of one person or even a clique that revel in the control over something.
Replies: >>33487
Oh no, that absolutely happens. Wikipedia is garbage for countless reasons.
(32.9KB, 349x451)
>dropped out in the third year of uni cause they were aggressively refusing to teach anything that actually relates to my chosen field.Expect my 98 page treatise on how compsci/compengineering curriculums need to be completely redone worldwide soon,
i am in my third year of uni right now and I have learned that college is good for everything except actual learning. you use it to build connections and get internships which provide more real skill than any textbook ever will. the actual "learning" I've experience amounts to doing exercises from the textbooks (that i torrented and could easily do on my own, the only difference is that there's a due date attached) and sitting through lectures which are also taken straight from the textbook. I have not received any actual hands-on experience from uni, but I have learned more from my shitty PC repair job than I have in all my three years of getting my IS degree. I have also come to the conclusion that I do not want to work in the IT field, as 90% of my job will be fixing what retards broke, and 10% will be fixing shit I know absolutely nothing about because nobody prepared me for it or told me what to do in case it happens. Ergo I have placed all my hope in /agdg/, which I dragged my feet on doing for 4 years (let ideas float around in my head without actually doing anything with it besides a few shitty music tracks). if my first gaym (which I hope to at the very least have an alpha out by the end of the year) is a flop, my life is fucked because I've built everything around the success of this one game, which you're all going to tell me is retarded as all fuck, which I am aware of. I don't have any other options at this point.
>Getting a proper grasp on quaternions has been quite a pain because of that
This might help https://www.3dgep.com/understanding-quaternions/
It's an article by one of my old college teachers before that course went to shit. It's still got ye olde math squiggles but that's hard to avoid with quaternions, he explains stuff at least.

>you use it to build connections and get internships which provide more real skill than any textbook ever will
Can confirm, the only value college/uni has. Too bad I was too much of an autistic fuckwit to use it and only have a useless slip of paper now.
>my life is fucked because I've built everything around the success of this one game, which you're all going to tell me is retarded as all fuck, which I am aware of
Well that pressure generally leads to one of two outcomes; a spectacular failure or a spectacular success. More often the former, but I wish you good luck regardless.
(2.4MB, 3543x3507)
As a five year uni pleb I can't really agree on the "you learn nothing" line. Personally it has provided a great amount of opportunity to finally learn mathematics after having it be a noose around my neck for my whole school life.
As has been mentioned in a post above a lot of uni material assumes very little and has a much better approach to building up your knowledge base and while it's arguably stealing away time I could be wageslaving it's a special feeling to unravel something that has been a blackbox for over a decade.
Replies: >>33506
>I don't have any other options at this point.
Actually, life has a way of creating options for you as you go along. Finish your gaem and see how it goes. Sounds like you're the type who will never stop learning, so you'll be fine Anon.
>letting a few libraries do the heavy lifting while you program your game is the same as not programming at all
You're clearly quite mentally challenged, it's no wonder you struggle with basic mathematics.

Wikipedia is just terrible. I've had better luck understanding concepts from reading online university-level material, including both "official" material uploaded by the universities themselves and "unofficial" material by Indians. Videos help sometimes when you need an ELI5 version of something before jumping into the meat of it.

>.t I dropped out in the third year of uni
Bad decision. Granted I don't know which university you went to, you should have still went all the way and continued the remaining couple of years. Third year is when you really start specializing in the CS field and learning the interesting stuff. You should go back.

>if my first gaym [...] is a flop
Your first game will absolutely be a flop you idiot, like your first musical piece / painting / dish / literally anything. You will fail many times until you finally make one good game, then the game after will be even better, and so on... The first step is finishing that first damn game so you can learn something from the experience to help you make the next game, then moving on to it. That is your only option.

To add to your points; university was little more than a motivator and guide for me, which is still something very valuable. Most of the time is wasted and most of the knowledge is not retained, sure, however it's able to give you just enough experience that getting some real education on your own later becomes a breeze. Having deadlines helps, teaching concepts in a certain order helps, giving you a general idea about a lot of topics helps, and of course, having the shiny piece of paper that says you endured 5 years of this torture immensely helps.
Replies: >>33512
(132.5KB, 1280x1280)
I know C, Make, a large portion of POSIX, a handful of computer algorithms, I'm working through an Algebra book I can see I'm too retarded for it to learn the basics school wouldn't teach me, and maybe a few more minor things.

Where do I go from here to make some basic bitch game?
Apparently Terry wrote a very basic 3D racer for TempleOS that's only a few hundred lines of code, I'd like to lean towards learning how to do something like that.
Well, the sources are available. Crack that game open and get to studying Anon.
Replies: >>33509
(1.2MB, 800x18907)
I always advise people to use an engine for their first game, but if you're up for it check out https://learnopengl.com and pic related.
Don't do >>33508 until after you gain some experience.
Replies: >>33511 >>33520
(72.4KB, 750x750)
I see the C++/C# suck can be hard to avoid. But I also spot some books I could go for in the near future, thanks.

However, that list seems like it would take a lifetime, even if you pick 1 or 2 areas of interest.
For instance, I'm a NEET and I took 4 months to go through most of APUE (later chapters were too advanced for me) studying every weekday. Perhaps that book is a bit of a cheat since it is so long, K&R C for instance takes 2 months.
I'm curious as to how DOS-era gamedevs could be in their early 20s and pump out high quality 3D games. Specially considering the late 80s/early 90s situation where programming languages like C itself were still too slow to make an office suite out of so most of those games were pure assembly.
Maybe figuring out where the heck they learned all that from could be worth it.
Replies: >>33528
(137.2KB, 1000x1000)
While I can not yet reliably comment on whether it was a bad decision or not I can certainly tell you that it was an informed one,I combed through past and current course material for the courses I would take and interviewed several upperclassmen before coming to the realization that the sheer amount of crust I would have to wage through had a quite unlikely chance of being worth the payoff.
I was also unable to properly benefit from the networking and motivator aspects of the institution due to a) preferring to remain to myself or a small group of friends and b) the expectation/rythm of work that is supposed to motivate the student by forcing them to keep doing stuff leading me into an exceedingly neurotic stress spiral as I responded non-optimally to that kind of pressure.
I was vaguely aware from observation that my priorities/internal motivators were different to those around me but looking back on my experiences and behavior from that time period recontextualised that the normalfag learning track just wasn't for me.Yes,I am aware that sounds incredibly autistic,however I am not actually on the spectrum.
Entirely unrelated personal reasons make it very hard for me to go back but I am currently participating in a long distance/workers education program to get a degree as there are certain legal privileges granted by having one in my country.

TL;DR:I did my research,the part about the 98 page treatise was not a joke.Thanks for caring though.
Fuckhuge lists like these are almost always made by retards who didn't even read three books on them. I don't know why idiots make them, but they do. Shit that I spotted on first glance for this list:
>massive overlap between books
>outright terrible meme books like Elements of Programming and Clean Code
>an entire fucking branch for UML
And god knows what else. I know you didn't make it, but serious question: How many of these 100+ books have you read?
>Elements of Programming
>a meme book
Anon, I...
Replies: >>33523
Yes. Elements of Programming is fucking horrible and the only reason people recommend it is because Stepanov wrote it and they don't want to look like it was 2deep4them. EoP is one third terrible algebra textbook, one third STL design rationale and one third template-laden reference implementation. It's the most unreadable shit I've seen in a long time.
Replies: >>33527
Now that you mentioned it:
>Cormen et al
Giant book on algorithms that you most certainly don't need for your vidya project.
Replies: >>33539
IMO it's one of the more elegant, math-oriented programming books in existence. It certainly highlights thoroughly why the STL actually works as amazingly well as it does (and would work on other languages in a similar if they provided the basic abstraction mechanisms at least to the degree C++ does). But, no one here has to take my word on it or yours heh. It's opensourced now.
Replies: >>33530
>I see the C++/C# suck can be hard to avoid.
You don't have to touch those languages at all, the list is meant to be all-encompassing, you pick the branches you wanna explore only and traverse them. Also notice the legend at the top of the list; solid arrows describe the recommended path, and dashed arrows describe multiple optional paths. If you go through only the recommended books you'll find they're not that many.

>Fuckhuge lists like these are almost always made by retards
I got the list from https://miloyip.github.io/game-programmer , if you visit it you'll see:
<This work was created by Milo Yip, who has been a game developer for more than 20 years.
<The books shown represent knowledge/skills that may/should be acquired by game programmers. There are other important ways of learning, such as practicing, courses, industrial/academic conferences/publications, etc.
Doesn't seem like a retard to me.
>massive overlap between books
You think? Any list of books on any topic in existence will always have overlaps in knowledge covered, it's up to you to choose which book(s) to read considering the writing style and targeted reader level.
>an entire fucking branch for UML
Read my note above about the list being all-encompassing.
Replies: >>33633
(133.8KB, 400x400)
>mfw stepanov shills on zzzchan
Including SICP really gives it away. Anyone who actually read that would only have a list of about 12 books.
It's like someone listing the complete works of Aristotle and then 30 self-help books.
Replies: >>33539
Those kinds of books are to leaf through the index for things that seem interesting when you're stuck on a problem, not reading back to front and committing to memory. Maybe give them a quick glanceover once.

>Anyone who actually read that would only have a list of about 12 books.
*Would only have time left to read 12 books before they die
Replies: >>33540 >>33542
(2.8MB, 400x400)
>*Would only have time left to read 12 books before they die
He said SICP, not TAOCP.
(943.6KB, 480x360)
apparently I lucked the fuck out, there's a raycast extension somebody made that looks like it will do the job I've been agonizing over for the past week or so.
No it's not that kind of book, it talks extensively about the notations and how to proof the boundaries. Unless you are fluent in the material anyway this won't be a book you can just drop into at any random chapter.
>Read my note above about the list being all-encompassing.
What kind of argument is that? Would you read an "aspiring cooks reading list" that had Shaniqua's essays on how to nig free cheeseburgers from McDonalds?
learn Lua and try Löve (aka Love2D)
Replies: >>33689
Entry C# course for Unity:

>racing game
Tutorial #1
Tutorial #2
Replies: >>33689
I have most of packs of Polygon series assets.
Basically, its game-ready assets with specific thematic (history, fantasy, WW2, modern, sci-fi).
It was made for Unity and Unreal, but you can import .fbx models into Godot too and use already made texture for them (they placed into the pack folder) or you can draw your own, just like you did with your tanks.
Characters/monsters already rigged, but have no animations, so you must make your own or get somewhere else (Mixamo.com, for example).
>Official store link with screenshots and explanations:
So, just say which one you interests in.
Replies: >>33697
(51.9KB, 275x275)
Sorry but I'd rather die than do nugrammer shit. 
I want to do something Terry wouldn't call me a nigger for. In fact I spent yesterday setting up my test machine, a 32-bit laptop from 2002. If I make something It'll be from the ground up and it'll aim for PCs with the sort of performance computers had in 2000.
Well so far it sounds like you're looking for every possible thing you can do except working on a game. Start by changing that.
(2.4MB, 429x592, 00:25)
Lua is not soydev shit. Lua is one of the best scripting languages that you can embed in your C/C++ program and it has been used in some games in the past (Chibi-scheme or Guile would be the second best alternatives for Lua, imo) Secondly, if you want to use pure C/C++, you could also use Allegro https://github.com/liballeg/allegro5 or SFML https://www.sfml-dev.org or SDL https://www.libsdl.org
Replies: >>34252
(69.5KB, 900x565)
>nu-polygon styles
Oh dog why, I quickly skimmed through a few military packs it has and the vehicles/weapons look scavengeable enough but man the character design is fucking ass. Welp at least it's better than nothing I suppose, I'll take these if you have it, thanks in advance. 

>just like you did with your tanks.
For the Stug III I am currently using it's from TA:Spring 1944 mod and previously I used the OTA Axis and Allies (AATA) Stug III 3do model. Since the Spring 1944 version has a better paint job than I could do. 

>Characters/monsters already rigged, but have no animations, so you must make your own or get somewhere else (Mixamo.com, for example).
For characters if I need some I think I just rip them off from Half Life 1 or something, I don't need many human characters right now since the gameplay is primarily vehicular combat focused. For monsters I already have a zip file which contains 303 digimon characters in dae format which are rigged but contain no animation which at the moment should be sufficient I think. Right now I'm in the process of writing the buy menu and decoupling the player inventory as I made some mistake as I couldn't integrate with the shop menu, though today I didn't made much progress because I switched my Distro to a different one as I think I fucked up my graphics driver or something as on some games I have performance issue.
Replies: >>33769 >>33781
(7.4KB, 217x170)
Considering zero marketing effort outside of some very sporadic tweeting I think we're doing alright. Demo release on this Thursday and full release on the 28th. I was thinking of writing a retrospective on our experience bringing a gamejam game to full release, any suggestions of something specific you think would be interesting to read about?
It's like I'm in a ToysRus store all over again
Replies: >>33778
>mfw opengl
>mfw I have no face
Why is it so complicated? Why can't I just program for the GPU like normal? Here's some buffers with shit in them, here's where I want to use them. It seems so simple, yet opengl makes it so hard. Doesn't help that literally all the tutorials are more concerned about making you render 3D objects as fast as possible, than teaching you to actually understand the API thoroughly.
I've heard Vulkan is even more convoluted.
(124.1KB, 1044x809)
(124.1KB, 1047x806)
(84.2KB, 1038x801)
It's ridiculous, isn't it?

Welp I finally managed to get part of the player inventory to work with the inventory menu by decoupling the player main inventory and making 2 more scenes one for the buy menu itself and another one that acts as a "loader", however it seems I run it some issues where the items doesn't scale itself automatically thus taking up more slots or less, looks like I have another bug to report to the main developer of this inventory addon. Now I just need to find ways to add the remaining items to the MCV buy menu, I think using a json file as a list of items should be sufficient for this.
>character design is fucking ass
Yes, they're all like this, even despite the price for packs is very high.
But at least there is 6 gorillion props for every situation, maybe that's the point: shit design, but simplicity & diversity of props.
However, many cars don't have the ability to open driver/passenger doors, when wheels are separated and works well. Maybe it laziness or something else, lol. So cutting the mesh/connect faces/separate is the only way to fix it.

>Military pack
>City pack
>Battle royale pack
>War pack
>Town pack

P.S: Before importing into engine, open model with Blender and if there is some clipping polygons, delete unusual faces ("Remove doubles" option in edit mode).
Replies: >>33782
Thanks for uploading it.
>However, many cars don't have the ability to open driver/passenger doors 
For my project its not a issue.

>P.S: Before importing into engine, open model with Blender and if there is some clipping polygons, delete unusual faces ("Remove doubles" option in edit mode).
That modeller guy couldn't even bother to fix clipping polys? Wew what a hack.
Which OpenGL version are you targeting? And are you following a certain guide?
That seems to be the new-school method of "learning" where they do the bare minimum to tell you you achieved your goal. 
Find another book. I can't recommend one, I'm still studying math.
(161KB, 567x654)
>Why is it so complicated?
<wants less complexity
>...than teaching you to actually understand the API thoroughly.
<wants more complexity

>MFW pic related
Not really sure what you want Anon. OpenGL is simply reflecting pretty closely the underlying hardware architecture of GPUs. This is really why it won in this area, b/c it's pretty honest (and therefore pretty efficient).

You can use typical abstraction mechanism to hide the (necessary) complexity which makes things easier to reason about. This approach will help you build large, complex systems that can still perform well.

But there literally isn't a shortcut to learning. I'd suggest going wholehog, being patient and methodical, and working towards crafting a specific set of capabilities you want in your system. Otherwise, you may as well go back to GameMaker, etc. 

There are no honest shortcuts in life Anon. Work hard, be patient.
Replies: >>33877
>OpenGL is simply reflecting pretty closely the underlying hardware architecture of GPUs.
Khm, that's vulkan and directx 12. OpenGL was made for SGI machines back in the 90s and you have drivers bandaged together with duct-tape that try with infinite amount of heuristics to map that almost fantasy world to how modern GPUs actually work. Of course, modern OpenGL (3+) added some new stuff and deprecated some of the legacy cruft, but the whole automatic object management and trying to emulate a single threaded computation model on 1000 cores is not going to fly well.
Just look at a Vulkan sample, all those command buffer management and synchronization is done by heuristics in the OpenGL driver. I'd say, opengl is like running a html5 in an electron app, while vulkan is doing it in native.
Replies: >>33881
>pretty closely 
>pretty honest
>pretty efficient

So, you seem to be arguing for using nothing but BINARYmachine language calls to the GPU hardware, one could presume then? :^)
Replies: >>33882
No, just say don't write bullshit. OpenGL doesn't reflect how modern GPUs work at all.
Of course, if you don't want to deal with all the complexity that Vulkan requires and you're not writing a software that requires every last drop of performance, go with OpenGL. I'm not saying OpenGL is bad, it's just not a bare-metal API, it's something between a low-level api like vulkan and a graphics middleware.
Replies: >>33891
>OpenGL doesn't reflect how modern GPUs work at all.
Actually, it does. w/e show us all the code you've personally written against either API and then make your point again.
Replies: >>34014
(244.2KB, 1805x425)
>all the tutorials are more concerned about making you render 3D objects as fast as possible
Thats because fags misunderstand the meaning of The Black Triangle.
It's supposed to be a determination that the underlying architecture is sufficiently ready to accept the work that it will serve as a foundation for.Then you get idiots just making sure it looks like such an indicator without doing the required work and calling it a day like they managed to actually build something.
Of course you could just be misunderstanding the meaning of a black triangle yourself,though tutorials that only show that tend to lack the necessary depth anyways
>show us all the code you've personally written
You actually did it you mad lad.
Replies: >>34016
(84.1KB, 640x480)
There, you get a screenshot, I have better things to do than argue. And if OpenGl would reflect how GPUs work, we wouldn't have vulkan today...
> If I make something It'll be from the ground up and it'll aim for PCs with the sort of performance computers had in 2000.
Why? You can't buy anything worse than a computer from 2012. It's all the same price at this point, and even our poorfag non-white/russian friends will have a clunker that destroys a 2002 machine. If you go online and buy the first piece of shit for $150, one that comes with a keyboard and mouse for free, it'll probably be a computer that was good in 2012-2013.
Replies: >>34042
Well, make a game about how you take the poo's to their loos.
>it'll aim for PCs with the sort of performance computers had in 2000.
If you're really adamant about low-spec you can try developing for single board computers (e.g. Raspberry Pi), microcontrollers (e.g. Arduino), FPGAs (e.g. Xilinx), or other similar devices. Admittedly I have zero experience in that field but it seems to be quite popular nowadays... You can also target DOS, or better yet FreeDOS since it's open source and you can inspect its code, but keep in mind that development resources will be very limited for it.
Godspeed you autismo-anon!
Replies: >>34672
>I know [...] a large portion of POSIX
Excuse my retardation, but isn't POSIX just a specification for what a program must have to be a "complete" operating system? How can you "know" POSIX besides memorizing it, it's a list of standards, not a skill if I understand correctly. Also, why algebra? I know a little (took a graduate level course in algebra, passed but barely) and I don't see how that could be very useful for programming. I hear that linear algebra is great for graffix but I know nothing of it.
>How can you "know" POSIX besides memorizing it
I believe that anon is saying that he has a reasonably comprehensive understanding of the conventions and expectations at hand when dealing with common posix based systems, 
>Also, why algebra?
Not sure what specific value anon is trying to extract but Algebra,at it's core,being about the handling of data via assignment and  manipulation of symbolic representations,does have tenets and ideas in common with programming.
Also Linear Algebra is in fact quite a large slice of a (3d) graphics stack.
He can program against the POSIX standard or at least claims so.
>Also, why algebra?
You can model a lot of problems in algebraic ways to great effect. I think it was still on zchan, but there was an anon struggling with efficient row rotations in a matrix and very basic group theory immediately gave the most efficient solution. Linear algebra is also extremely useful because as soon as you notice something is linear, you have the entire theory at your disposal. There is this classic brain teaser where you're counting the number of paths of a certain fixed length from A to B on a graph. Noticing that this problem can be reformulated in terms of linear algebra turns the time/memory inefficient naive calculation into a constant memory and logarithmic time one.
>took a graduate level course in algebra […] linear algebra is great for graffix but I know nothing of it
May I ask what you did in that course? Or maybe I'm misunderstanding and you meant you know nothing about graphics.
Replies: >>34062 >>34193
Algebra I and II, standard core courses, not special topics. Linear algebra is something I never learned: it wasn't in the graduate course (not enough for the material over two semesters) and I don't remember taking the undergrad course that's on my transcript at all.
Replies: >>34193
>spend about a week working on a map
>finally texture it
>it looks like shit
Well shit. Demo day's in about 2 weeks. Hope I can get something decent looking by then.
Why not show us and we be the judge of that?
Replies: >>34129
(501KB, 839x608)
(411.6KB, 743x606)
(442.6KB, 953x628)
Some surfaces still aren't textured, but some of the important buildings in the level look like shit. Maybe they'd be better if I added windows, but at this point the geometry of my level is already tightly coupled with itself that making small changes like that will cause a big domino effect.
Some of the buildings are supposed to look a bit ugly, because I'm going for the brutalist/postmodern aesthetic but even brutalist architecture doesn't look this bad.
looks fine to me, I dunno what you're talking about. besides if the gameplay is good, who cares.
This already looks seriously good, even with the postmodern aesthetic. I think you should stop being so picky and move on with this map to get it ready for demo day.
>Demo day's in about 2 weeks
Shit. Well, there's always next time I guess.
Replies: >>34184
I'll continue to work on my game for 22 hours a day no matter how detrimental it is to my health.
Replies: >>34187
It doesn't help that I completely shelved what I was working on and started a completely new project about 2 weeks ago. Then continue with that at my usual snail pace. I should do something with my impulsive ideas and wanting to do everything.
Replies: >>34188
Code with an agile philosophy, and whenever you have some weird impulse, program some half-working prototype for it and shelve it off somewhere you can recycle it in a new project.
Replies: >>34194 >>34204
Oh, and to actually answer the question: group theory, ring theory, field theory. Galois groups, modules, radicals, polynomials over fields... I remember very little of the actual content, though.
Stop using the word agile you street shitter. It doesn't mean anything.
Replies: >>34196
>implying poo in the loos do anything agile
Replies: >>34233
Don't tell me things like that, I've already lost my motivation for today.
You'd think they have a high agility due to dodging poo piles every time they go take a shit.
(226.3KB, 323x400)
Wait when is the next demo day? 2nd of february? I thought I had more time.
Replies: >>34239
(410.6KB, 499x477)
(23.8KB, 551x300)
Yeah there's no way, guess I'll just re-submit my current demo even though it will be fully released by then. Not a lot of people played it anyway.
(61.4KB, 609x585)
Meh I'm not going to bother with demodays anymore, what's the point of it when I don't get any feedback if my mod is shit or not?
Replies: >>34243
(789.6KB, 792x792)
It's easy to get discouraged but I think it will get better Anon. So far with the death of 8ch/8kun and subsequent split between here and markchan I figure people have been having a hard time knowing demo days are happening and putting time into it. It will get better, it has to.
Demo days don't even make sense when we have so few people working on stuff.
Replies: >>34245
Of course it does. It sets a deadline for the (admittedly few) people working on stuff so they have something to work towards. I've gotten very valuable feedback on previous demo days even with few playtests.
Always remember, that it doesn't need to populate the level just for the sake of it, but make building/props to be a part of gameplay which player/npc can interact with (breakable windows, breakable walls, open/close doors, etc). Even if it's looks like from late 90s, it doesn't matter for now, you can improve it later. 
What is matter, is that environment must be intractable as much as possible.
Graphics first < Game mechanics first.
Replies: >>34295
How many hours per day did you work on that? From the sound of it you don't have lots of experience building cities, for a single such person that seems like a reasonable result to me. To me it just looks like there's details missing, for example none of the windows/doors have any kind of framing. All your textures are completely flat, look at images of concrete buildings and there's all kinds of details and variations in texture even in a flat concrete wall, especially if it's old and worn out. There's also no lighting whatsoever which makes things look extra flat.

>geometry of my level is already tightly coupled with itself
Saying that and calling it "unsalvageable" sounds to me like you're fallen into a sunken cost fallacy trap. You don't want to make big breaking changes because you've spent so long on it, it's way too early to get into that mentality. I'd recommend taking a step back and coming to terms with the idea of tearing everything into pieces. Keep as many pieces separate as you can so you can more easily change things, just slap around random boxes with placeholder textures to get an idea of what details should go where. Only build a proper model out of it when you're sure it works with the city and looks fine in-game.
Replies: >>34295
Not him, but everytime I look into adding Lua something about it turns me off. The fact literally everything is a table really is not appealing to me. That suggests it has to run a hashing algorithm just for basic array access/insertion. I think there is a workaround where you can do something like 
table[#table+i] = stuff
Theres other hidden performance hits, like the fact it doesn't cache access to globals (even C does this arguably)
possible language: c++, relevance: 9
Consider the next fragment:
function foo (x)
  for i = 1, 1000000 do
  x = x + math.sin(i)
return x

We can optimize it by declaring sin once, outside function foo:
local sin = math.sin
function foo (x)
  for i = 1, 1000000 do
  x = x + sin(i)
Just avoid the .next() command entirely, because you never know if your tables have holes in them and it's slow as shit
Function next, when called
without a previous key, returns the “first” element of a table (in some random
order). To do that, next traverses the table arrays from the beginning, looking
for a non-nil element.
I got these notes from: https://www.lua.org/gems/sample.pdf

It's just got a shitload of landmines in it. That's not what's killing it for me lately though. What's killing it for me lately is you can't directly READ a C/C++ integer/float. Not even as a constant. You have to make a function call in order to get it. From what I can tell this means if I have my positions stored in the core server, in order to access them I would have to call a function to read from the backend. If there is a shitload of basic data like this I can't get without running a function call that's going to really hurt later. My game is effectively a mishmash between SS13 and Dwarf Fortress, but on an even bigger scale. 2 really really low poly low demand games that get halted into a crawl when too much shit goes on from everything being simulated. I know they both have reasons, like Dwarf Fortress isn't multi-threaded, SS13 doesn't have proper garbage collection and tranny code, etc. It still makes me extremely paranoid about performance because I don't want to build my entire codebase on something that will cripple me later. 

I've been considering switching to the grand-daddy of scripting languages: Lisp. It's blazing fast and has a small footprint, but the core problem is it is so god damn autistic to learn that it'd kill any chance of my game forming even the tinniest modding community (assuming anyone even plays it)
Replies: >>34258
Actually decided to dig up some benchmarks of Lua and Lisp. Lisp consistently wipes the floor with Lua. I would really like to see embedded performance though, and it still probably suffers from my concern of not being able to directly access C/C++ variables.  

I guess it's because Lua is kinda like a wrapper for Lisp. 
possible language: kotlin, relevance: 10
The [Metalua Manual] provides some contrasts between Lua and Scheme: "Basically, Lua could be thought of as Scheme, with: a conventional syntax (similar to Pascal's or Ruby's); the associative table as basic datatype instead of the list; no full continuations (although coroutines are actually one-shot semi-continuations); no macro system." Metalua brings a macro system into Lua for compile-time metaprogramming (CTMP). There is also a preliminary implementation of [Lisp in Lua].

For those comfortable with Scheme, Guile may be a better solution. Lua has a bent towards scripting and configuration by "non-programmers", where infix syntax may be more appealing. Lua is also often used as a stand-alone language, which in the Scheme world is more the domain of scsh than Guile. Guile is significantly larger than Lua.http://lua-users.org/wiki/LuaComparison
Replies: >>34258
I fail to see how Lisp is "autistic to learn" or how (even if it was) this would influence your chances at a modding community. Look at some of the absolute ass-backwards gutter trash modders regularly deal with.

If you're interested in using Lisp, CFFI makes using C from inside CL pretty easy and there's also ECL, which I've never used but is specifically made for being embedded into something else. Scheme would be more minimal, but this cuts both ways - if you end up writing anything big in it you'll write ad hoc implementations of the "bloat" of Common Lisp. Guile in particular went down that path. For small scripts, Schemes like Chicken and Tiny might be neat though, if I remember correctly Tiny Scheme is a single file with 4K lines of C or so.

I'll readily believe that CL performs better than Lua, but the Benchmark Game is not a great metric. Look at some of the code in there, the "Haskell" has nothing to do with idiomatic Haskell for instance and the C is usually full of intrinsics.
Holy shit the "Lua is like Lisp except lacking almost all big selling points of Lisp" makes me assmad.
Replies: >>34270
(909.1KB, 951x884)
>tfw you released an actual game on Steam
Replies: >>34274
Stop this shit what the fuck is wrong with you
>I fail to see how Lisp is "autistic to learn" or how (even if it was) this would influence your chances at a modding community. Look at some of the absolute ass-backwards gutter trash modders regularly deal with.
I think Lisp is just intimidating to look at. I use emacs so I'm pretty familiar with it, but to the untrained eye it's absolutely terrifying. More-so than shitty code and awful flows. You don't realize just how bad those are until you're already in the reeds. Really the problem with Lisp is entirely surface level.

Right now this is the most attractive thing. Basically I can have a dev build that can reload runtime scripts and a release build that has converted the Lisp to C-code.
It includes a runtime system, and two compilers, a bytecode interpreter allowing applications to be deployed where no C compiler is expected, and an intermediate language type, which compiles Common Lisp to C for a more efficient runtime. https://en.wikipedia.org/wiki/Embeddable_Common_Lisp
That's an interesting title Anon, grats. Use your power only for good. Please don't create yet another degenerate pozz-fest out of it.
Replies: >>34288
(802.3KB, 1133x635)
>Please don't create yet another degenerate pozz-fest out of it.
It's just a basic puzzle game without any story. If anyone plays the demo let me know, would love to hear what you think about it.
Replies: >>34304 >>35048
You are why smug is always superior to this shithole.
(1MB, 1692x931)
(1MB, 1692x931)
Thanks for the advice. I decided to just go ahead and save the level and see if I can salvage some parts from it later. Started from scratch with a new level and while it's still not perfect, I think it's better than what I had originally. I think perhaps choosing textures AND geometry at the same time is the best balance, that way I'm not stuck with uncanny looking buildings.
>How many hours per day did you work on that?
about 12 hours a day for a week. But also doing some coding, texture work and modelling for other entities in the game.
>From the sound of it you don't have lots of experience building cities
You're right. I have almost 0 experience building 3D environments in general. When I was younger, I'd mess around with Source SDK's Hammer and try to make some levels but they were so ugly I pretty much got bullied out of Facepunch and Moddb
>All your textures are completely flat, look at images of concrete buildings and there's all kinds of details and variations in texture even in a flat concrete wall, especially if it's old and worn out
You mean there's not enough details within the texture? Do you know of any resources for this kind of thing? I know in Source Engine, there are some variants of the walls that have scum on them, but the main reason I'm reluctant is because I'm not sure how to make that kind of thing tileable.
(122.2KB, 1643x869)
(138KB, 1663x919)
>same image twice
Could have sworn I used different images. Maybe it's a glitch with using the same filenames on the same post?
Also JPG for further compression because 1MB for these images is total horseshit.
Replies: >>34300 >>34411
(33.5KB, 550x413)
(136.8KB, 1000x563)
(105.1KB, 640x427)
Found that it also helps to use reference images pretty much any time you work on a new building/angle. I'm documenting my struggle/solutions here just for the benefit of other people who might be on the same boat as me one day, or in case I fall back into my old ways then I'll remember what mindset got me back into creating less uncanny levels. Something based off of an already existing location is almost always going to look better than your imagination. 
Sorry for flooding the thread. I think I'm done with this chain of posts.
I played 4 levels of it so far. 
>maybe flash or highlight robot when entering each level briefly
>moveable UI elements, kind of annoying to scan back and forth between the tiles and the top left corner of the screen
>tooltips for UI elements
Useless but might add some character suggestions
>panning and zooming
>clicking on each tile gives a reactive animation, ie robot does a flip or trees release pollen, polar bear pokes his head out of igloo
Fun game though. Might give you gibs if it's cheap enough.
Replies: >>34306
Thanks anon, I really appreciate it. Highlighting the robot at level start is a really great idea that I hadn't thought about at all honestly. It makes sense not to have to scan for him. I also really like the idea of reactive animations on tiles. If it wasn't just 2 programmers making this we would probably implement that. Launch price will be 7.99 dollarydoos but I'll probably throw some free keys in here next week when we launch.
Replies: >>34308
>Launch price will be 7.99 dollarydoos
Replies: >>34312
(821.5KB, 1127x640)
Nice. We had a very long discussion about this, both in the team and with our publisher. Turns out it's very hard to price a game. Anything below 5 puts you among absolute garbage tier games but at 10 you are suddenly competing with both new games that are way bigger than ours and discounted AAA games. Don't want to considered garbage by association but can't compete with the really heavy hitting games. 

I have two more maps to make, making it 9 regular and 6 bonus maps per world, 61 maps in total not counting tutorials.
Replies: >>34318 >>34622
Most games on steam seem to operate on price anchor kikery. Generally I only buy guys when they are on sale around $5.
Replies: >>34420
>generally I only blow guys for $5... i-i mean buy games.
Replies: >>34322 >>34420
(765.8KB, 786x930)
(1.7MB, 432x272, 00:09)
>Keyreal hasn't updated his blog or twitter in almost a year
That makes me sad. I still remember when he was getting Speebot off the ground. It was like seeing his game manifest before my eyes in a matter of months. It was insane. Was a seriously inspiring thing to see.
If you need some crappy models, which you of course can improve/completely remade/just take inspiration from, i can drop the download links.
It's polygon series packs, they're made both for Unity and Unreal.
Just say, which pack interest you.
Replies: >>34438
>You mean there's not enough details within the texture?
Often there's gradients from one color to a slightly different one, different sections have slightly different color/texture, and corners have some kind of worn out stains. Your textures are the same subtle pattern throughout the surface. Rather than "detail" in the texture, it'd describe it as "variation".

>variants of the walls that have scum on them ... I'm not sure how to make that kind of thing tileable
I think it's usually not tileable, at least not in all directions. I don't know enough about 3D to know how it's done, but I've seen processes where small textures get overlapped over flat tileable areas. So you could just place a piece of dirt or a crack where ever. I've seen it editable in real-time at least once so maybe it's some game engine/shader feature, they just grabbed a piece of worn out stains from some corner and moved it around.
Replies: >>34427
(160.6KB, 445x399)
(7.7MB, 1280x720, 00:48)
I think they're just called decals.
Replies: >>34455
(27.3KB, 1304x195)
(45.2KB, 231x87)
Click through rate on our store page, I think it's probably a combination of our game's title being pretty generic and the capsule kinda looks like shit.
Replies: >>34456 >>34504
Do you actually have those lying around? Because apparently they cost quite a few shekels for a hobbyist.
Replies: >>34448
Here's an interesting blogpost on how to abuse virtual memory. Instead of managing your memory normally, you just allocate as much as you want, and leave it to the operating system to manage it for you. Note that if you disable overcommitting on linux this instead causes the program to instantly self destruct.


possible language: c++, relevance: 6
#define GB 1000000000
uint32_t num_tanks;
tank_t *tanks = virtual_alloc(GB);
uint32_t num_bullets;
bullet_t *bullets = virtual_alloc(GB);
Have you ever considered that just because you can do something, doesn't mean you should?
Sounds like a terrible idea. You'll have no idea when it's going to lag someone's shit up because the OS suddenly decides to swap memory.
Replies: >>34449
(111.5KB, 224x220)
>use language and framework where the entire point is low level control of resources
>fuck it let the OS do it
Just use a GC language if you're going to be a pleb about it.
There is some websites which upload pirated Unity/Unreal assets, so everyone can download them.

>Do you actually have those lying around
I actually bought some of them when they first released in 2017, but i've pirated the other ones, because i don't have such money to spend only on vidya.
>You'll have no idea when it's going to lag someone's shit up because the OS suddenly decides to swap memory.
That's just the modern reality of working with a managed OS, though.
Decals are more for graffiti or a piece of plaster missing, what anon described was more like using a grunge map, and perhaps a slight gradient.
Replies: >>34457
Do you have comparison numbers for other games? It's easy to say 8% is low without a reference point, but I have no idea how many fags just merely never click on anything.

>Note that if you disable overcommitting on linux this instead causes the program to instantly self destruct.
No, it just causes the allocation to fail because without OC the kernel actually checks that you have enough memory to satisfy it. Which is the sane thing to do because otherwise you can run out of memory in the middle of a computation where recovery is impossible. But unsurprisingly somebody with this genius idea doesn't check if his allocations actually worked. Just derive some bounds for your object count and pre-allocate, god damn. That unique ID system is seventh circle of hell tier abuse.
Replies: >>34480 >>34511
(1.1MB, 2420x1123)
But how do you add that kind of stuff on top of a big ass wall? You could give each part of each building a completely unique texture, but that seems like a lot of work and maybe memory intensive. I've seen decals used this way a few times, once in some engine/game demo where they were used to add stains on a metal surface, once in Jonathan Blow's game where he used them to add shapes onto the ground, the UE4 decal tutorial used them this way, and if you search for "decal texture" this is exactly what you get.

That said I don't actually know how this stuff is normally handled.
Replies: >>34476
(1.9MB, 1868x714)
Like I said, you don't use decals to add detail on entire walls. You apply a grungemap and possibly a gradient or something like perlin noise to add variation to the texture of the wall. By making sure the size of the texture and the maps used for adding variation aren't the same, and giving them an offset based on object instance, every single part of every wall is going to be different.
That said, a few of those "decals" in your image could also just be applied to the bottom or top of walls if they tile properly.
>You could give each part of each building a completely unique texture, but that seems like a lot of work and maybe memory intensive
Look into the difference between a texture and a material, and why the latter is pretty much what's used everywhere these days. Very few things have single textures nowadays, it's more than just "add normal map lmao".
(113.5KB, 219x330)
>Do you have comparison numbers for other games?
No it's very hard to find anything solid on this because data/articles are polluted by absolute shitters making babby's first Asset Flip™ then going on Gamasutra talking about how good they are. It just strikes me as a pretty low click through since we have such a small niché and a free demo. Also feel free to tell me to fuck off if people aren't interested in this aspect of gamedev because I think it's awful but a learning experience.
8% clickthrough for anything sounds pretty high. I'd be interested in roughly how many hours you sank into this, and your earnings. Did you break $7.25 an hour?
Replies: >>34502
I believe I personally have put around 500-600 hours into the game. Then there's my partner who has put a 200ish hours into it as well. It hasn't gone up for sale yet, just the free demo. I don't expect to make any money of it at all. Our deal with the publisher is 70-30 split of net profit after publisher expenses have been paid (mostly Steam fees).
(19.5KB, 800x420)
No, please keep posting. I know there are some John Lennon LARPers in this thread sometimes but most of us want to see the fruition of our labor turn into something more material and therefore certain aspects of marketing should play a role in these threads.
Replies: >>34517
Kek, are you complaining Anon? That's pretty fucking high. And I like the name, actually.
Replies: >>34511 >>34517
>It's easy to say 8% is low without a reference point, but I have no idea how many fags just merely never click on anything.
Not to mention bots, scrapers, etc. 

>Do you have comparison numbers for other games? 
Only Valve has that info. 

I don't think I'm personally putting my game on steam. I actually just swung by cuckchan's /agdg/ to ask if people are even considering Steam anymore with how over-saturated it is. I figured just due to volume alone they'd have more people who actually have games on Steam. Surprisingly a bunch of people echo'd back my concerns. Valve takes a 30% cut, you practically have to market it by yourself now, "Only buying it if it's on Steam" probably isn't even a thing, there are upwards to 1000 games released each month on the platform, none of the cards and other shit is meaningful to anyone except niche autists who wont even play your game so they wont spread your game either, etc. After a first few posts it started getting flooded with people who were clearly mad I was questioning what was likely their only strategy. I could tell the people just wanted to shitpost to shut the conversation down so I bailed.

Without the numbers though it's basically impossible to tell just how much it's worth it, and Valve is very open about their info EXCEPT how useful their store is to devs. 10 years ago I'd say it's worth it no question, but today it needs to be re-examined. It seems everyone and their Mom throws their game on Steam because it's what everyone has done for years, but Steam is so fucking bad now. It seems like a case of the madness of crowds.

I wonder what Speebot-devs numbers were. He had a bunch of shit on Steam. I wonder if Phantom Path flopped hard and he just lost all motivation.
Replies: >>34513 >>34517
>I wonder what Speebot-devs numbers were
Regardless of what those numbers were, I'm sure 100% of good businessmen would be effin delighted with a nearly 10% response to advertisements. Don't compare yourself to others lad, it isn't healthy. You just do you.
Replies: >>34514
I'm not the dev. I'm asking because I'm considering if Steam is even worth it.
Alright, thanks anon.

I have no frame of reference so I have no clue.

>I don't think I'm personally putting my game on steam.
It's a very smooth experience so far but since we released our demo there's been 50-100 new games released. It's just too swamped.
Replies: >>34521 >>34522
Most of those 50-100 games released are complete jokes that shouldn't even be considered games. So it's more like 5-10.
Replies: >>34526 >>34528
Steam is basically a dumping ground for asset flips. Maybe shilling your game to a bunch of curator pages will get it more noticed. Even then no one knows how useful the curator page, or discovery queue, etc is.
Replies: >>34526
Yeah but they still clog the store right? So even if they're shit they will detract from my game.

I've had a look but haven't found any good ones so far. There's actually a built-in system for sending keys to fitting curators, I'll take a deeper look once final release is ready.
Replies: >>34528
The bigger point is you don't really get anywhere on Steam because your game will be drowned out in a sea of asset flips. 

I would just spam every curator that seems tangentially related. Even directly messaging the guy who runs it. I wouldn't be too picky unless you're afraid of them just flipping your keys on g2a. I can't really think of how else you'd get your game noticed via steam.
Replies: >>34529
>I would just spam every curator that seems tangentially related.
This is risky advice. If you give keys out to people who would not normally play your game, and they play it only because it's free, you invite bad reviews when some faggot gets mad at mechanics in a game he would have otherwise avoided. With money as a barrier, people who buy your game naturally self-select as people who will like it. And steam curators self-select for faggots.
Replies: >>34531 >>34533
Perhaps this topic is a good understanding as to why a marketing campaign before release is useful.
Replies: >>34532 >>34533
Lots of assumptions in that post.
Replies: >>34534
With what money? There's nothing stopping you from marketing a released game.
Yeah our experience is that there is no point having people that aren't truly into puzzle games play our game.

We purposefully didn't have a marketing campaign because no time, just wanted to get a game out the door for once.
The only "assumption" is that curators are faggots, which is true. Otherwise its a simple chain of premises and conclusions.
Replies: >>34535
Assumed conclusions.
Replies: >>34536
why are you defending the curators? they are against all that is good
Replies: >>34537
(121.4KB, 770x780)
Defending? All I did is point out how everything you said was based on assumptions. Now the cats out of the bag because you've admitted you got personal gripes about curators, so basically nothing you said is concrete at all. Zealots in every discussion when it comes to marketing shrieking over any potential route to success because of personal reasons creates an overwhelming sense of stagnation in the whole thread. It's crab mentality. Then you guys turn around and wonder why SJWs dominate everything and all the indie games suck, when the people making games in your own barkyard can't even get off the ground because you drag them back into the bucket. It's no wonder all those people left for discord.
Replies: >>34538
>Defending? All I did is point out how everything you said was based on assumptions
Programming is based on assumptions. Math is based on assumptions. Geometry assumes you can draw a line between two points. Game dev assumes journalists and curators are faggots. I don't understand why this personally offends you. It sounds like you're ESL and you're having trouble parsing my thoughts.
>people aren't interested in this aspect of gamedev
I think that's why you should give your opinion. As far as I know, most gamedev tutorials/experiences only chronicle the game-making part of the experience, and leave out the post-launch experience. The fact you don't know what a good CTR is probably means more perspective on that part of development is needed.

That or it means the steam store setup and layout are garbage.
Replies: >>34540 >>34541
(732.8KB, 876x658)
I honestly only have good things to say about the Steamworks setup. The documentation is probably some of the best I have ever seen with great examples, templates etc. The few times we had trouble we couldn't solve ourselves support were also very helpful. They actually have a really good part in the documentation about marketing but it's more about how you should setup solid goals yourself before starting marketing efforts because otherwise statistics etc won't make sense. That's obviously what I'm experiencing right now. But I agree, there needs to be more solid info on this outside of the "hurr durr make a discord/mailing list/suck tranny dick on twitter".
Yeah, I know I sound like a little bitch for saying this, but spending months, possibly years, working on a game and fine-tuning it to be something high quality only for its launch to be gone like dust in the wind is not a happy feeling. I've never released anything big, but I've been in a few gamejams where I was thinking I'd at least have 20 people trying out my game after release only for it to have a total of 12 unique downloaders after about 9 months after release. 12 unique players, are you fucking kidding me? I spent over a week grinding really hard at the development, coding what essentially amounts to an entire engine from scratch, and most people can't be bothered to download a free 800 kilobyte zip file?
Despite any technical or creative competence you may have, it's so easy for your game to just fall under everyone's radar, especially when they don't have much proof that it's NOT the jank that they'd usually find on places like that.  It's more important now than ever to come up with strategies to make sure your game is known to a good size of people outside your immediate social circle, because the market is just too saturated with people trying to make a quick buck off of a meme or posting their variations of Unity Tutorial projects.
GOG might be a slightly better market to compete in since there's less saturation. I know some people have mixed feelings about GOG, but it's just a suggestion.
Maybe exploiting streamniggers like every other big kikes do is your best bet. Except doing it with small time streamniggers instead. Seems like the only way to generate organic advertising that isn't too cost prohibitive. Might even convince some to play it for free with a game key. Game development as a career sounds miserably bleak.
Well, at least you learned your lesson. Now you know why kikes spend 3 times as much on marketing than on development. If you had shelled out some money to come up with some nice looking banners and art (even if it had nothing to do with the actual game) I bet you would have been a lot more successful.
I want a free 800 kilobyte zip file, gib.
Replies: >>34572
It's not that great, but I'm proud of the fact that I wrote it from scratch in a week and implemented things like ECS, Finite State Machines, abstracted the tile animators and a few other minor things
Coding an engine during a gamejam from scratch? Noice, most people use premade engines, because, well, you don't have much time.
Aargh, that hitbox completely kills the mood.
Looks good and plays okay, but I can't win the game. It could use some serious polishing before getting published...
>implement a ducking/crouching action
>implement a way to regain health
>implement a health indicator for the bull
>implement audio
>implement a start screen, game over screen, and help screen listing controls
>debuff the scorpions
Replies: >>34638
care to share the code you're so proud of Anon? nprb if not, i'm simply uncomfortable running compiled blobs handed out by strangers on the intertubes. no offense.
Replies: >>34581 >>34592
Looks good to me: https://www.virustotal.com/gui/url/7c3260dd24e1ee53134424d01c1ff1b3f685c5fa2a0ef46e5ed69229760be687/detection
Replies: >>34601
I'm a little bit embarrassed, but here. I'm sure people will find all sorts of things wrong with the code, but keep in mind it was written in about 6 days on a tight schedule. 
Also, to anyone reading, you can pretty much do whatever the hell you want with the code. I'm too busy to get back to this project so I likely won't be doing anything with it any time soon. Give the cowboy a penis and sell it on Steam for 5 bux. I don't care.
(14.4KB, 540x314)
Looking through the code, you can see I've commented some things I've intended to do but never actually got around to. Dear god.
Replies: >>34600
Well, thanks kindly Anon. Appreciated. Actually I think it's rather clean-looking code. I'll give it a closer look soon. Cheers lad.
Replies: >>34638
Thank you.
Anyone who can in a particular time frame (especially six days) code a nontrivial thing that works to a standard they set out beforehand is already leagues beyond the majority of people who want to program games. You know that there's a big difference between "this code looks like shit because, knowing where I could cut corners, I sometimes did so to stay on my tight schedule" and "this code looks like shit because I have no idea what I'm doing."
Replies: >>34638
isn't steam still just 100 bucks per entry?
even as crowded as it is you still want to put it on there because people expect it and some only buy there, and even if you sell less on steam directly they give you a certain amount free keys you can distribute however you see fit.

>GOG might be a slightly better market to compete in since there's less saturation. I know some people have mixed feelings about GOG, but it's just a suggestion.
cd project just fucked their reputation with cp2077 and even harder with the devotion shitshow. don't know the conditions to get on there, but if it's feasible same reasons as for steam apply. also has the indirect effect that people can pirate it easier and then buy it after "demoing" it (not everyone of course, but those people would never pay for it one way or another).

>Turns out it's very hard to price a game.
ever considered making a simple base gane package free and sell more content? kinda like a demo but nothing extra people have to download (and comes with the assumption it's outdated and castrated on purpose).
dunno if it would be applicable for your game and how that would play with consumer expectations tho.
Replies: >>34626
(865KB, 1141x646)
>ever considered making a simple base gane package free and sell more content?
We never talked about this actually but I guess it's fairly common with something like puzzle games. Give a free package that is the first few levels or something. The reason I made a demo was that whenever we would have people playtest they would usually play through the first world then run out of time before they had to move on to the next thing. So I made a mini-world that is a vertical slice covering all mechanics in the game so people get to experience as much as possible in as short of a time as possible. I didn't consider download size but the game is sub 100mb download so I think it's quite alright. Pic related, it's the demo world where you get to play 2 maps from each world.
Replies: >>34627
Why does the UI look like it was made for mobileware? Is this being released for iOS and android?
Replies: >>34628
Yeah we are planning to do mobile releases as well but it was mostly about keeping things simple. Avoiding text and Unity's UI system at all costs (all UI elements are 3D).
at least he’s white
(1MB, 2083x1618)
>wonder what a certain new game is like
>go to twitch to see some gameplay
>almost nothing but koreans and other non-english players
Reminder to localize your game if you want to succeed.
(4.5MB, 1280x720, 00:57)
Like I said before, I'm pretty busy with another project right now, but I can see myself getting back to Running Gun and making some expansion when I have more time for it. My original plan was actually to have the cowboy chase some bandits on horseback who throw some explosive barrels at him while he also has to avoid other obstacles. Of course that might have been a bit too complicated for the timeframe so I just went for a demonic bull who shoots sonic waves at you and makes it rain scorpions.
>but I can't win the game
Ride the wave, basically. My main strategy is to walk in the same direction as the sonic waves and jump to avoid them. Vid related. I actually haven't died in a long time. 
Well, good to know that the code is alright by other peoples' standards too. When I have extra time, I'd like to expand the base into a lightweight 2D game engine. I don't like most mainstream 2D engines because they all tend to favor using either Python-like languages or are bloated with extra features (and even encourage mixels by design, like Unity).
Replies: >>34665
(140.7KB, 500x281)
>looks too easy
>then... falling scorpions
(1.5MB, 3000x3000)
I've done 8086 assembly under FreeDOS actually. I recommend it to everyone, it's easy and the benefits are immense.
It gave me a better understanding of boolean algebra, two's complement, and computers in general. But that wasn't all.

The greatest gift was the general programming experience, for more reasons than one.
The calling convention for 8086 assembly is to send function arguments on register DX and then receive the result on register AX, I noticed while looking at assembly output from gcc and clang that they follow this convention when they can (with x64's extended registers).
I noticed a tendency to start writing new functions once whatever data you're dealing with doesn't fit in the available registers anymore because registers are more convenient to deal with than pointers or pushing/popping the stack all the time, and being constantly under this constraint has actually fixed my difficulty with making my C functions shorter.
Assembly can also be a bit obtuse to read which I've found increased my ability to read and write necessarily convoluted code. Take the typical fizzbuzz, the convenient way to check if a number is a multiple of both 3 and 5 in assembly is to check if it's a multiple of 15. To zero out a register, you XOR it by itself. It really became easier to reason upon any program I run into, simply because I trained my program reading skills with assembly which is much less readable than anything else I've used.
I've also learned that a lot of things can be done much more easily with assembly, like checking for overflow or writing a bignum library.

POSIX also includes a C library and a set of utilities useful for programming C like POSIX Make, Shell, m4, c99 (a compiler interface that both gcc and clang base themselves on), and more. 
Most systems today are "POSIX-like" and so if you learn POSIX you can use a large and varied library that's mostly present on any OS out there. For instance even Windows' networking API is closely modeled after POSIX.
I have barely highschool level math and I want to do graphics one day. I also want to understand the Nyquist-Shannon theorem and play with the programming side of audio one day. 
All the great programmers I know are good at math, and I don't want to be a nugrammer. 99% of what I do is out of sheer hatred for modern computing and modern programmers. Modern computing is fucking shit and if the people behind it say "learning assembly is useless nowadays" that's exactly what I'm going to do, that's how I make decisions.
Another thing is that math makes you smarter, that can never be a bad thing in anything you do.
(232.3KB, 460x561)
How the hell are rendering engines usually made? I've been thinking about this for days now and still don't have a remotely good answer.

For example if you want to draw some thing, you make the shaders and buffers and whatever for it. But what if you want to draw something special that, say, requires 3 textures? Or a buffer that contains different data? What if you want to draw 70 billion sprites and don't want to create vertex buffers for each one? What if you want to draw less sprites but each one needs a unique vertex buffer? What if you want to be able to switch an entity from one rendering method to another anytime at runtime? What if you want to draw a chunk of 2D tiles like some terraria game that you keep modifying/animating arbitrarily? You'll need entirely new shaders and buffers and completely different processes for calling the graphics api. You can't make a generalized system out of this just by swapping from one shader to another or anything like that, you can hardly generalize any part of the process. It's like I'll need to make a new graphics engine from scratch every time I want to do something slightly different.
Replies: >>34719 >>34721
(134.2KB, 720x720, 00:02)
>You can't make a generalized system out of this
Make Buffer.
Fill Buffer.
Draw Buffer.
>You can't make a generalized system out of this just by swapping from one shader to another or anything like that, you can hardly generalize any part of the process.
The solution is simple. You stop trying to make a generalized system.
Making a generalized system means you are creating an abstraction of an abstraction of the GPU. That's useless boilerplate.
The way a game renders something is not something the rendering engine should decide, but the game does. The rendering engine merely provides tools to streamline the process and possibly even introduce fall-backs for legacy hardware, but it does not decide for the game.
Replies: >>34722
So do people typically write their own rendering system from scratch when they use Unreal Engine and want anything besides simple models? There has to be some ways to generalize certain parts of the rendering loop.
Replies: >>34726
>So do people typically write their own rendering system from scratch when they use Unreal Engine and want anything besides simple models?
If you replace "simple models" with "preset pipeline", yes. The engine has a predefined way of rendering stuff, and people just use that. Then when they realize that it makes the game look generic (because every shitty unreal and unity project uses default shaders), they start to dive in and replace all the underlying parts and effectively write their own things from scratch.

For example, Unity has a "Mesh" object that contains a mesh. What properties the mesh vertices have is predefined; they have a vec3 position, vec2 uv, and a handful of optional parameters like normals, colors and more uvs. These cover most cases, but not all. And the moment you want something that the predefined object doesn't cover, you have to use the advanced functions.
But a closer inspection will reveal to you that the "advanced" functions are just abstractions of simple GPU functions. Mesh.SetVertexBufferData is just glBufferData. Mesh.SetVertexBufferParams is just a function that does the VAO boilerplate. The moment the generalized object isn't general enough, you're back where you started; writing your own pipeline.
Replies: >>34727
>they have a vec3 position, vec2 uv, and a handful of optional parameters like normals, colors and more uvs
Well the fact that they have all this stuff and you can switch it around is already part of the generalizing I was talking about. They've made some kind of system that's capable of a lot of things, and you can poke into it with some abstracted calls without having to rewrite the rest of rendering engine. Meanwhile I can't even figure out how I'd give triangles a flat color in case they have no texture without having to create separate shaders and branching paths everywhere, i.e. almost write a second rendering system from scratch, and that's one of the simplest examples.
(33.1KB, 642x173)
>first feedback on the demo outside of friends/AGDG
>it's this shit
I don't even know what he's talking about, no one has had troubled with saving before. We use Unity's standard save path as well so there shouldn't be any permission problems either.
Replies: >>34779
Drop that man a line and ask him to help you debug the game maybe?
Replies: >>34780
(9.4KB, 235x235)
Yeah I'm working on it. Seems like he may be playing on a Mac, the one platform I wasn't responsible for.
Replies: >>34825
(976.5KB, 1872x1044, 00:12)
You guys in the mood for some rocket jumping?
Replies: >>34802 >>34822
nice pre-rendered footage nerd, we all know you can't code
Replies: >>34806
(689.8KB, 673x767)
You're right. It's all a prerender I hacked together in Milkshape 3d and I'm gonna do a total bait & switch when the game comes out.
Replies: >>34813 >>34825
Why do you feel the need to post that picture? Why did you save it? Aren't you ashamed of yourself? Just because you know a bit of code and make a shit game doesn't grant you access to do retarded shit like this. It would be better for this gay community if you just stop posting unless it's a link to download your shit game for everyone to laugh at.
(1.4MB, 1671x927)
>stop posting
Replies: >>34826
It will be a doom clone, or fps+rpg like system shock/deus ex ?
Also, need sci-fi models ?
Replies: >>34824
More like a Quake clone with some Metroid elements thrown in (Not a Metroidvania though). 
Actually, here's a rough synopsis that I hacked up in about 30 minutes one day. 
At the very beginning I thought about RPG elements and I'm still considering them, but I'm not sure if it would be fitting or not.
>Also, need sci-fi models ?
Maybe. Why, do you have a repository to some assets or something?
Replies: >>34830
>officially supporting mac
Why would you even do that to yourself? I understand linux, since at least that's a platform for people trying to get something better than wangblows, but Apple is just Microshaft's scumminess with a gold veneer and shouldn't be encouraged.

>You're right. It's all a prerender I hacked together in Milkshape 3d and I'm gonna do a total bait & switch when the game comes out.
I mean, if it works for the AAA industry I don't see why you couldn't do the same.
Replies: >>34827 >>34840
(1.9MB, 1106x486, 00:16)
Reminded me of this all the way back from the original 8/v/
>Why would you even do that to yourself?
I figured it would be easy since it's literally just build it™ in Unity but unless we can fix this easily before Thursday we will probably just pull the mac version, fuck this shit.
>Maybe. Why, do you have a repository to some assets or something?
Yes, polygon packs from syntystore.
Just import them in your project.
You can of course modify/totally remade/just took inspiration from.
Replies: >>34836 >>34837
I'll take a look at them. I don't like using other assets unless it's sound, but I do like taking a look at how other assets are made because I'm a programmer first and modeler second.
Replies: >>34837
Oh, and thanks, anon.
There are more mac users out there than linux, so if you have an engine where it's "easy" to create a mac version, it could be a tempting option. That is, of course, until you have to actually support that platform. and well, as someone who works on a program as a day job that has to work on windows, linux, android, mac and ios, I can tell you that apple is the most developer (and user) hostile company in this lot.
Why do people use that crap is beyond my imagination, even botnet 10 is an upgrade compared to that piece of shit.
Replies: >>34885
(180.2KB, 637x637)
I can't go to a single place on the internet without raising a contrarian point that I genuinely believe in and killing all chances of a productive conversation.
fuck you
Your opinion is incorrect, my opinion is correct.
I can't rebuke wrong opinions and "alternative facts" and must be confined into a comfortable place of resting.
(22.5KB, 541x296)
I've been reading up on it a bit and it looks like Apple just wants to funnel all users into iOS instead of macOS? To earn more money in a very strict walled garden. It explains all the hoops you have to jump through to get something out for Apple products. 

Current stats for the game. Don't ask me why less than 5%  that download the demo actually play it. Mostly concerned about the median time played, do they just start it and fuck off? Or do they realize it's not their kind of game?
>the median time played
3 minutes seems like enough time to realize that the game isn't as fun as it looked. Also I imagine there's many people who go through a lot of free demos fast without thinking that much.
Replies: >>34890
Yeah I figure a lot of people just download any free demo to try it out then drop it instantly if they don't like it. There seems to be a lot of weird shit like this that plays every single demo released and makes a short video of it https://www.youtube.com/watch?v=zKiHcLMzbSk. Do we have any macfags here who could try the game? I'm never making a game again for a device I don't own myself.
Just find a game where you have an achievement for just starting the game, 10% of players won't have it, even with non free games, so take that into account... but yeah, 5% seems low even with that.
Is there anyway to do atomics on opengl es 3.0 glsl? Or at least fake it? I basically want to dump every pixel that passes a test into a buffer/texture but I need to know how many elements are currently in there so I don't override old ones and the data is compact, and also so I know how much data is in there.
Is "3 minutes" a link to something more detailed, like an actual histogram of times?
Replies: >>35039
(46.3KB, 719x654)
Replies: >>35154
I'd say the simplest approach is to use generalized features from C++ to sort/pick/filter etc. and then send the results into the shaders. You have far more features at your disposal for this kind of thing in C++ than in GLSL Anon.
Replies: >>35094
waaaaait, you got a publisher and you don't get exposure? What's the point of having a publisher then? ||unless it's just some friends group||
Replies: >>35049
Our publisher isn't primarily a publishing business, they do consultancy for game developers. We use them because we have personal contacts there and they pay our development expenses. It was a conscious decision not to have them do marketing because we didn't want to invest that kind of time and money into this game, just get something out.
Quick search reveals to me that Atomics are actually just implemented through SSBO's. But SSBO's, just like Atomics, are in OpenGL ES 3.1, not 3.0. So you're shit out of luck for the actual hardware feature.

Maybe you can rewrite it to use transform feedback instead. This however captures primitives instead of fragments, so it might not fit your needs.
Replies: >>35054 >>35094
Atomics are also very expensive. ES3.0 is very limited in this regard, if you absolutely have to do it on the GPU then your best shot is probably some manner of transform feedback. But does the data actually need to be packed? Otherwise you should just use a texture and write per pixel honestly.
Replies: >>35094
If I did the simplest approach I would have 1 FPS. Checking pixels on the stencil buffer for 1080p is 2073600 iterations.

I remember looking at those a while back but they don't seem to fit the bill. It's looking like the solution might be janky.

It's not packed right now and it's serviceable, but the empty spaces a performance hit. Maybe there's some secret stencil wizardry to account for this.
Replies: >>35096
>Maybe there's some secret stencil wizardry
Stencil testing is hardware supported. You can just throw any written pixels in your stencil buffer outside of your specified stencil function and values, it's basically free.
Replies: >>35097
The pixel I'm checking is on the stencil buffer. I mentioned checking a pixel, not that it was a stencil pixel my b ad. I want to get a list of all stencil pixels that pass. Someone probably found out some turbo autistic hack to do it.
Replies: >>35098
Right, so you render something to your stencil buffer and then you need to read it again after. Would it be an option to render to a color buffer as well then do a fullscreen quad with the color buffer as input and stencil test enabled instead? It trades cycles vs memory but if you're data is low resolution (which it should be if you're using the stencil buffer) you might actually win performance by doing this.
Replies: >>35099
The plan is to generate voxels where the stencil buffer passes. Originally I wanted to just draw 1 point for each pixel and then use a geometry shader to blow it up into a voxel, but Opengl 3.0 doesn't even have geometry shaders.
Replies: >>35101
Yeah ES 3.0 is not a fun API to work with. I actually don't have a very good solution suggestion then, sorry about that.
Replies: >>35105
I have some unfinished ideas I can take a stab at to solve it. Just wanted to see if I was missing some crucial thing or if it was really this stupid.
(123.6KB, 283x359)
(28KB, 1177x202)
lads i need help and i'm losing my fucking mind over this
>found raycast plugin to determine distance between two points (2d gaym)
>it can detect entities fine but detects terrain like dogshit, as the engine morphs adjacent tiles into a larger polygon to improve performance
>guy on forum says the solution is to use the fraction of the raycast and slope calculation to find the real collision
t = fraction returned by raycast data
x1 = x coordinate of ray origin
y1 = y coordinate of ray origin
x2 = x coordinate of ray end
y2 = y coordinate of ray end

x = x1 + t * ( x2 - x1 )
y = y1 + t * ( y2 - y1 )>still getting fucked up coordinates, nothing is colliding right
i can't take it anymore, i need help
Replies: >>35153 >>35195
(85.1KB, 1280x720)
This is what happens when you don't make your own engine that you have 600% control over
Replies: >>35164
Zero minutes being in that range means at least some of those players crashed on startup, I'd assume, unless they quit at the menu. The problem with having data is that you want to look hard at it for messages that aren't there.
Replies: >>35203
(199KB, 376x302)
This is part of the reason why I decided to build my own engine. I did so much modding I kinda knew how they were structured and was always dicked over hard by the engine being structured a certain way or just straight up lacking a feature that you can't really have without doing a ton of janky hacks and tricks to fake.

It delayed my game a shitload (years) and I had tons of restarts, but going through all the hurdles turned me into a programmer who totally eclipses everyone I come in contact with. Just so many universal concepts to learn that applies to programming as a whole. Things you wont learn on the job or at school, because you can bruteforce without them but knowing them can make you solve a problem in 5 minutes that takes other people hours.
Replies: >>35361
(278.4KB, 1280x720)
Have you considered not using Scratch ?
(912.2KB, 4524x2060)
Can't actually tell if it's 0 minutes or 10 though. 10 minutes is a perfectly viable amount of time to play our demo and know what it's all about, 0-2 minutes is obviously awful. We are going "gold" today, did some last minute fixes with default settings for mac shitboxes. We were unable to reproduce the saving issue on Mac so we are putting it down to PEBKAC. Hope someone plays and reviews the game tomorrow.
Replies: >>35326 >>35361
(218.3KB, 1092x1023)
>mfw I have to rewrite the buy menu again because I forget to take account for different class types
Well shit.
Are you receiving any kind of backlash for posting on /v/?
I don't think he's gotten any kind of feedback, negative or otherwise
Replies: >>35329
Pretty much this. No one knows about the game so it won't be a problem I think. You guys are in the credits too.
As if anybody else's opinions matters. Who else is he going to take good feedback from? Reddit and trannyera? Get serious. There may not be enough of us to get devs to pander to our taste, but we are best when it comes to judging a game's quality.
Replies: >>35350 >>35450
Nice trips. Also you're entirely right, there's really no feedback to be gained anywhere else. Everywhere else people are either full shill for their own game all the time or they just tongue your asshole for making a cube go forward. This is the only place where you get rightfully shat on and that's usually what you need when making a game.
Yeah. You'd think they'd just make the data available to you so that you could bin it yourself, because as I was getting at I think there's valuable insights in that 0-10 bin being hidden by the way the data's presented. Or maybe not. That data's probably owned and monetized by Valve, anyhow, so maybe it's not surprising they only want you to view it through their tools. Or maybe I just didn't see the "get dump" button.
That's literally experience that you're describing: doing the work that someone in that actual job would do and learning from it. Programming is one of the few areas where that can still be done on one's own while still producing useful output. But for people who can't teach themselves, they might spend those same number of years and not get anywhere near as much actual experience out of it. Learning is a skill, and plenty of people just never learned how to do it or forget how as they got older.
Replies: >>35380
(91.4KB, 879x316)
>You'd think they'd just make the data available to you so that you could bin it yourself
Think it's more a matter of scale, Steam doesn't really cater to these really small games/demos. I don't think there's a way to get a data dump but it doesn't feel like too much of a problem either. If there's significant technical issues I expect them to pop up in reviews or on the forums.
Replies: >>35418
(54.6KB, 429x571)
Advice for making music? I've been using LMMS for a few years at this point, I know how to use it but I lack a basic knowledge of music theory. I find myself knowing that something isn't right in a song and that it should be fixed, but I don't know what to replace it with. Are there any definitive guides on making music for games?
Replies: >>35383 >>35456
If you have no idea what you're doing you can just replace it with a sample. Half of OOTs soundtrack is samples. Like the forest theme for example is literally a sample of an instrument being slowed down, some African chant being distorted to sound like a flute and then some dark ambient synth chords slapped over the top. This vid series is kinda cringe, but the guy digs up all the samples used in all the dungeons, and sometimes he discovers them later. Like for the forest temple he found out later when he was making the water temple vid that there was a slowed down track just plopped right in the forest temple theme, so he mentions it in that vid. I think the fire temple mentions the african track the forest temple used, but I'm not sure.

I think the original fire temple theme was literally just ripped off chants that were EQ'd to sound distant and mysterious. Like there wasn't really a single original thing in it except the drums. It sounded rad to.
Hey, I'm going to buy your piece of shit game because I liked Human Resource Machine/7 Billion Humans. Follow your dreams, faggot.
Replies: >>35445
(9.4KB, 157x185)
(833.5KB, 947x845)
We're all gonna make it anon. Thanks for all your support and help. Time to move on to the next game.
It has nothing to do with quality reviews, but with your typical subhumans crying loud enough that their fellow subhumans at Valve summarilly cancel your game.
>activate that shit
>steam's HTML5 garbage shits itself so you can't even view your own library
Why do people still use this crap?
Replies: >>35501
I wish I knew more about the music theory side of things, but 8-bit-music theory has a lot of stuff related to what interests you.
(2.2MB, 640x360, 00:25)
used the 8th one, thanks anon
I hope the game will be fun
Replies: >>35578
Fuck I get 20FPS ingame.
Replies: >>35502
Lower the particle density, bottom setting in the cog menu on the right.
sold any copies yet anon?
Replies: >>35532
(44.9KB, 275x273)
(34.1KB, 453x604)
that's what you get for not investing in gamestop lmao.
I bought it a while ago. It's downloading now.
Replies: >>35542 >>35568
As my old grandpappy said to me, "Son, keep throwing shit agin' the wall. Eventually something will stick" concerning my job hunt. Just keep at it anon, this stuff takes a while.
Replies: >>35568
I like how the constellation map uses the same controls as the levels. You should have the map use the music track of the last level played instead of reverting back to the first constellation music every time.
Replies: >>35568
Damn anon you might actually have been first. 

Next time around we will for sure start some kind of marketing effort much earlier.

I'm glad you liked that, it's one of the most hotly contested features of the game internally. My partner does not like it at all but I'm glad players seem to enjoy it. Your music suggestion is good, I'll write it down for our potential future patch. We don't plan on a lot of post-launch support right now but we'll see how it goes.
Replies: >>35569
>I'm glad you liked that, it's one of the most hotly contested features of the game internally.
I don't mind it personally but I could see the argument against it. Would probably be a bit less bothersome if your position in the over-world was saved when exiting the game. Going a few tiles between levels and you barely notice it, but having to reach the edge of the map when booting back into the game and you need to think for a second.
Replies: >>35580
(84.5KB, 165x252)
I wanted to buy a copy as an Inventory Gift and found out that option was discontinued in 2017. Can I send the money to you using Bitcoin on some other direct way?
Replies: >>35580
(1.5MB, 450x337)
It took some work but I think you can reach all levels from every other level in each hub world. Maybe should have added an option where you can just click levels for those who don't feel like doing this "extra" puzzle. Save states could definitely be stronger. Would've been nice for users to get a list of their attempted solutions as well.

Don't worry about it anon, I appreciate the gesture though.

We had 6 sales so far so I'm basically good enough for a blue checkmark and ditching imageboards for plebbit and trannyera now. Jokes aside that feels pretty good and I'm sure most of the sales come from here, much appreciated anons. Demo day in 4 days, look forward to playing your games.
(27.2KB, 762x618)
I'd say congratz for releasing a game, but remove the association with /agdg/ and it's just another forgettable shit game to the endless pile of shitty games. Once upon a time mediocre games were released for free in newgrounds or whatever, I guess it's hard to see it that way because now Unity allows all those games to automatically be 3D instead of 2D.

People in /agdg/ claim to be the most objective about games, but whenever someone shows anything that isn't irredeemably bad, especially if someone releases a game no matter how mediocre, it receives almost nothing but praise. I understand that it's hard to seem "rude" towards people who get to the end of their project especially when there's an ID attached to your posts, and the desire to support fellow /agdg/ devs, but I think it's harmful to the developers in these threads when everyone puts on a smiling mask as soon as someone gets going.
Replies: >>35583 >>35592
(8.5KB, 177x176)
I never claimed my game was even mediocre, neither has anyone here. I've gotten plenty of criticism and at best people think my game is "neat" and it's cool that it actually got done and released. You're basically just being a fag right now.
Replies: >>35585 >>35632
>I never claimed my game was even mediocre
You don't release your game on Steam at $8 if you truly accept that it's mediocre or below.
Replies: >>35587
Why not? Anything below $5 is basically ancient or completely broken games. You price above because otherwise you get classed as absolute garbage tier just by association. I've been very open about not selling many copies. I suggest you post your progress instead of being a fag.
Replies: >>35588 >>36143
Let me rephrase: you wouldn't try to charge money at all unless you thought that your game was at least decent. The fact that you're increasing the price to avoid it "looking bad" means you're already under some kind of impression that it's worth buying.
Replies: >>35590
(1.5MB, 1024x682)
>you wouldn't try to charge money at all unless you thought that your game was at least decent.
Why not? The goal was to release a game, not release a great game. It intentionally has an extremely limited scope.
Replies: >>36143
(2.4MB, 1920x1080)
(2.2MB, 1920x1080)
This is always going to be a problem with gaydev communities. People will recognize effort and don't want to shit on something where someone seems to have tried. This is perfectly fine for a "first completed game", and with the rare ability to actually finish something without financial compensation unlocked, he has proof he can make shit and start working on more elaborate games with more interesting mechanics and decent art.

I'd comment more on the gameplay but I don't play traditional puzzle games and couldn't figure out what it was from the trailer. It also looks like mobiletrash, especially with the iconic star counter and giant touchscreen buttons. Main thing I'd do is add a proper environmental overlay for each of the puzzles. The hexagon grid should still be visible or obvious, but it would instantly move this game from mass produced mobile game to something casual consumers would seriously look at. If orange hexagon world doesn't catch my attention, gray hexagon world probably won't either. I actually hate looking at it.
Replies: >>35594 >>35595
>Main thing I'd do is add a proper environmental overlay for each of the puzzles.
I don't really know what this means, do you mean an abstraction of the grid to overlay on top? I'm not usually a fan of this because it often turns into people only watching the abstraction like in Bamham or Diablo games but it's an interesting idea.
(481.4KB, 1280x927)
(224KB, 453x371)
thanks for asking, I happen to have ripped shit off the internet as an example. The background could be 3D, or preferably to suit my tastes, a mostly flat painted background.
Replies: >>35597
(1.3MB, 1280x720)
(373.3KB, 480x360)
(211.5KB, 320x240)
(1.5MB, 980x735)
(1.9MB, 1280x720)
point is nobody wants to look at flat colored hexagons for more than a few minutes. by the n64 era they had figured this out.
Replies: >>35597
(578.6KB, 1129x631)

Oh right, yeah I get it now. I had a huge discussion with my partner about this actually. Basically he only wanted the hexagon grid, nothing else. Just adding the nebula/gas cloud took weeks of convincing. It looked even worse without that. I would've wanted to add planets and stars in the background I think because they could've been animated as well. Just not enough time. But I completely agree with you about this. I also don't like the orange.
Replies: >>35713
>man with no games in heated debate with man with one game
Don't mind that fag anon, you are one of the best among us just on account of actually releasing something.
Replies: >>35709
(105.7KB, 958x960)
i could be inviting a lot of criticism here, but have you considered giving steam keys to jewtubers and other fags with a somewhat decent following to get some more publicity? you may be attracting a reddit-tier audience that way but if they're willing to pay for your gaym or at least spread the word then is it that bad?
Replies: >>35686
Yeah we have talked about it both here, internally in the team and with our publisher. The problem is attracting content creators with a big enough install base that would also be interested in this kind of game. I haven't had any luck yet but I'm trying still.
>with my partner
Not to put too fine a point on it lad just in case you're new here. But do you mean business partner? If so, then I'd use that phrase instead on imageboards. Given the degenerate kike-driven state of current year, my partner has very distasteful connotations around these parts.
(9.6KB, 247x260)
I guess colleague is more correct.
Replies: >>35722
There's nothing distasteful about having a cute trap boyfriend.
Replies: >>35722
heh, if you're all that experienced then you should have already known this term is a dog-whistle for filthy neolibs, anon.

go dilate
Replies: >>35739
Language barrier anon. 

Can we get some progress? Who will have a demo for demo day?
Replies: >>35741
(57.3KB, 204x359)
>Who will have a demo for demo day?
I'm making some hot progress on my non-videogame sideproject that I got derailed into.
(85.6KB, 1920x1080)
(104.6KB, 1920x1080)
Well to derail some of the sperging in this thread, I guess I'll post some progress on my level design.
I plan on having something out by demo day, even if it's very incomplete. I know the level's going to be unfinished and there will be only 1 enemy type, but hopefully it will be enough to show what the overall gameplay loop is going to be like.
It's fucking dark. I mean I can barely see anything other than the hud and I'm in a dark room.
Replies: >>35769
Sorry, I'll be adding some lighting to the level over time. It looks fine on my screen but it's good to know how the experience is on other peoples' displays as well.
Replies: >>35770 >>35803
To be honest, the current display is not the best with dark images, I've already experienced it. And if I turn on the brightness it becomes marginally easier to see things, but it also becomes washed out as fuck.
Replies: >>35771
JPG compression also doesn't help, I'm sure. Some details I noticed in-game were harder to spot in the screenshots. 
If you have the time and 100 MB (38 compressed) to spare, I suppose you could give it a look and see for yourself.
Oh, also keep in mind that the quit button in the menu doesn't work right now. You'll have to alt+f4 your way out. Sorry about that, didn't really intend on publishing that build prior to this conversation so it's full of bullshit that only I'm willing to tolerate for the meantime.
Eh, thanks, but I think I'll only fuck around with wine tomorrow. Or a vm on an other computer.
Replies: >>35780
I'm both baffled and excited about the number of fellow linux users on the board... Back to topic, wine shouldn't need fucking around nowadays. You either don't install anything or only install dxvk through winetricks for directx games. I heard it's a bitch to setup on ubuntu for some reason though.
Replies: >>35970
I have a very expensive monitor for graphics work and I can't see shit either without squinting and moving my face to different positions. It's hard to tell what you intend the brightness to be, but most monitors, especially IPS monitors are terrible at showing very subtle dark colors.
Replies: >>35819
I like the brutalist/commieblock architectural style but like the other anons said it's too dark, can't see shit. I think it's fine for the level to be that dark if you supply the player with their own light source.
Replies: >>35819
Well that's 3 people saying that it's too dark, so it's definitely something I'll look into. Thanks.
Replies: >>35820
why should you care what people think bootlicker?
Way too bright. Are you trying to kill my retinas anon?
Make a level without skybox, but with standard "Directional light" instead, because not only we anons, who downloaded your build can't see anything, but it will prevent you to fully build and setup the level.
And after the playable area/npc/weapons/ammo/keys/doors/secrets are done and tested, then apply the skybox atmosphere and see how it goes.
Gameplay first > atmosphere first.
Replies: >>35844
(908.1KB, 1679x915)
So I raised the intensity a little bit and went for a somewhat brighter color for the directional light. Is this any better? It's not yet baked so maybe it's not the best representative of the final product, but I want to know if other anons can see it as well.
You mean it's too dark even on the build? Interesting. Even on my cheap acer monitor I can still clearly make out most of the details. 
If I have time, I'll likely be changing the skybox as well, to something more of a dark grey and less black.
Replies: >>35847
(2.4MB, 1920x1080)
(1.5MB, 1125x950)
It reminds me of Hell's Kitchen in Deus Ex. It is similarly dark but has lots of small local (obviously fake) light sources.
Replies: >>35848
Deus Ex and the HL2 beta are strong sources of inspiration as far as things like architecture/atmosphere goes. As well as the mental imagery of Bowie's Outside album
Although my game is going to be a bit more linear than DX and likely not have any RPG elements.
(525.4KB, 122x95)
After a lot of false starts, I'm starting again. :>
This time I have a one page GDD made by this guy
It doesn't even start under wine.
(460.3KB, 955x630)
(823.3KB, 1783x638)
(908.3KB, 1584x638)
Can someone who's not a complete retard like me tell me what I'm doing wrong?

All I did was change the location of one bone in my animation, and then for some reason, both the armature and mesh decided to shift upwards in the "Scene view" in Blender, while still maintaining their origin points.
This creates a huge problem in Unity where once the idle animations are triggered for these guys, they start floating in mid-air because apparently Unity's too retarded to know to just use the mesh's origin point. 

Thought to fix this by just moving the armature+mesh back to the world's origin, but no. Pics related, it will somehow distort the model and act like I never moved the position of the mesh itself. Starting to get really pissed off, I don't know much about 3D modelling but if anyone's ever had any issue like this I'd love to hear how you got it fixed.
Replies: >>36001 >>36013
(355.5KB, 640x640)
>pic 3
Replies: >>36002 >>36006
Reminds me of what happened when I downloaded a model to blender and tried to link it into an another blend file. And I have no idea what I did to fix it, I probably gave up
Replies: >>36006
The worst part is that from my perspective it looks as if Blender is actively sabotaging ny shit. As far as I know I haven't done anything I'm not supposed to do.
Replies: >>36014
You may have created a keyframe that transforms your mesh by accident. (Not to be confused with a keyframe that transform a bone, which in turn transforms your mesh.)

When in pose mode, select your mesh, clear transform and double check whether there's any keyframe data associated with it.
Replies: >>36016
(228.9KB, 623x391)
(167.3KB, 579x306)
New update: I decided to manually fix things by moving the mesh to its desired location even after it spazzes out and then I discovered that apparently switching it to the T-Pose animation will cause the whole damn thing to shift DOWNWARDS (Pivot and everything).
How the fuck? Has anyone ever experienced anything like this? As far as I know, an animation shouldn't have anything to do with the goddamn pivot.
Replies: >>36016
Yup! I was about on the same train of thought when I posted >>36014 but then I go and check my animations and somehow there are references to my object's overall position in the keyframes. Fucking disgusting. Thanks, anon. You probably just saved my project from being moved to the Recycle Bin only to be recovered the next morning when I've finally cooled down
(167KB, 628x335)
Replies: >>36067
no bully
Replies: >>36068
(646.6KB, 833x696)
Demo day is the third right? if so I ended up doing my build way too early for this shit. But HOLY FUCK I DID IT, NUCLEAR TEAGAN 2.0 FOR UT-RETUNED IS IMPLEMENTED!! AAAAHHHHH!!!
Replies: >>36114
(2.5KB, 800x600)
Demo day is allways dubs day my dude.It's the second.
Replies: >>36135 >>37192
didn't know that, thanks man
I thought a partner was the person you link up with to take charge of a big robot.
(66.9KB, 800x318)
Your price is fine honestly. Keep in mind majority of people here don't even buy games and are cheap fuck NEETs. Even back on 8chan people would screech at you just for having your game on Steam. Shit, you even had people like rustfag who completely derailed the shit out of sigma-devs thread because he had THE AUDACITY to make a thread about his redsky game release.

I stopped taking advice on anything related to business on here even when 8chan was still around. It's completely driven by ideology and politics. If you listen to anyone on here you're basically playing to lose. Maybe you can make a quality game while also handicapping yourself, but it's like playing basketball with no arms. I even stopped posting my game on here because I don't want to be associated.
>wow how come people don't accept this piece of locally made garbage stop having standards
>waah why are videogames so shit these days?
Replies: >>36145
>I don't want any of our devs to shill their game, sell their game for more than pennies, use steam, use patreon, use kickstarter, use discord, or any modern gaming community that could get their name out there, hide their association with a board that is PR poison, there first game must have the same quality as my favorite game and mayyyybe I'll give it a pirate
>waah why are videogames so shit these days?
Replies: >>36149
>made a game and got asshurt because anons called it shit
Replies: >>36369
>Keep in mind majority of people here don't even buy games and are cheap fuck NEETs
How dare you?! You filthy peasant! Don't you dare besmirch we beautiful few who have avoided the snares of wagecuckery! I wish you many long, unpleasant days in your sweatshop! May you contract the divine disease and be forced to wear the cross-dressers garments of "immunity" so as not to starve on the street from unemployability! May your master be ever-angry and your work-load endless!
>I even stopped posting my game on here because I don't want to be associated.
Most of the people here are normalfags, but you're clearly worse than that. Why are you even posting?
Replies: >>36369
What I want is for our devs to have high standards and a realistic perspective about the value of shit they make, and not expect to sell everything they make and then cry when someone criticizes that decision. And I want everyone else not to suck someone's dick just because they got their game into steam.
(56.5KB, 598x944)
>I even stopped posting my game on here because I don't want to be associated.
Replies: >>36369
This retard here knows a bit of code and thinks he's the king of the world. Shut your mouth next time if you're gonna bitch, bitch. No one cares and no one is going to buy your games(if there's any lol).
Replies: >>36369
You aren't better than us dude, get off your high horse.
Replies: >>36369
(19.7KB, 380x173)
(19.2KB, 450x334)
I'm fucking losing my mind

I fixed up my old text input system and added support for UTF8, but for some reason when there's a non-ascii character after the input cursor, the first non-ascii character changes to some other character. And it only happens if the cursor has 0 length (so text isn't highlighted).

If I switch to a text render function that doesn't render the cursor, it works fine so it can't be in the text input system itself. But the only thing the new function does differently is draw the cursor when the loop is at the same place as the cursor. It doesn't modify any variables that aren't specifically related to tracking the cursor location or break the loop or anything, all it does it draw the line when it runs into it.
Replies: >>36158 >>36161
Did you accidentally put the cursor in the middle of a character?
Replies: >>36160
(11.7KB, 298x190)
That would only matter for text modifications which all work fine. The next utf8 character breaks even if the cursor is not connected to it at all, pic related, the next character breaks even though it's in a different row.
Replies: >>36161
Impossible to debug without seeing the code, for all I know one of your assumptions (all it does…) is wrong.
Replies: >>36187
Is there any good site which can be used to upload a project page there which has political incorrect content? Obviously that leaves steam and itch.io as the former requires money to get a project greenlight + requires the client in order to use it and the latter is administered by a SJW. Or is 8agdg wiki the only option left?
Game Jolt is good as far as TOS goes, but people don't use it enough for there to be evidence for or against it.
Replies: >>36178 >>36603
(29.8KB, 670x348)
(40.7KB, 680x793)
Does itch actually police the content on there?

>mfw 1 person refunded
>No Illegal or Infringing Content or Conduct. You understand and acknowledge that you are responsible for  our User Content, and you agree not to post any User Content that is unlawful, tortious, defamatory, libelous or invasive of the privacy of another person. If you post any User Content that is obscene, threatening, harassing, abusive, hateful, racist, infringing or otherwise objectionable or inappropriate (collectively,  Obscene Content"), you must correctly tag the User Content as such through the different tagging mechanisms made available by Game Jolt, so that other Users may avoid viewing that content if they choose. If you tag any User Content that contains Obscene Content incorrectly or not at all, the User Content may be removed from the Service and your Account may be terminated. You agree that you will not post any Obscene Content to the Service's public message boards or chat rooms and that you will not submit false or harassing reports or complaints to Game Jolt about other End Users. You acknowledge and agree that the decision to classify any User Content as Obscene Content is solely at Game Jolt's discretion.

It sounds like I cannot upload Moonman doom here? 

I'm not sure but I can recall one of those Itch admin twitter was giving out SJW vibes in a manner where they do police content in terms of political incorrect games/mods.
>It sounds like I cannot upload Moonman doom here? 
Well they're basically only forbidding content that breaks laws like copyright infringement and libel. I doubt you'll find any platform that's any less lenient regarding that.

>Itch admin twitter was giving out SJW vibes
They're constantly promoting games and jams that align with woke politics in the website itself and their official blog which is embedded on the front page. They're also seem to allow blatantly TOS infringing jams if they're leftist enough.
Replies: >>36183
Sounds to me like you can upload it, but have to put the nigger mark on it so they can keep a clean front, similar to the way porn is often handled.
Replies: >>36183
Sounds to me you're allowed to upload Moonman DooM, but must tag it Obscene Content and maybe explicitly tag it as racist, and can't post about it on the forums.
Replies: >>36183
(76.3KB, 746x829)
It's a little interesting how many people typically refund games. Sounds like it's normal to expect up to around 10%. Some popular games apparently received close to 50% refunds when refunds first came out.
>Well they're basically only forbidding content that breaks laws like copyright infringement and libel. I doubt you'll find any platform that's any less lenient regarding that.
Moonman doom has a lot of stuff that got ripped off from other games and mods too, does that count as copyright infringement as well?

>They're constantly promoting games and jams that align with woke politics in the website itself and their official blog which is embedded on the front page. They're also seem to allow blatantly TOS infringing jams if they're leftist enough.
Exactly that is what I'm meaning, so then Itch io is out of the question then.

Alright I see, well then I guess I'm gonna risk one of my cock li e-mail for that then but not now because I don't feel yet motivated enough to write a detailed readme so that perhaps one or two person actually bothers reading it. Thanks for the explanation.
Replies: >>36184
Write your own webpage on neocities.org and it should generally be fine. If your project is FOSS you can use Github/Gitlab Pages (or any-Gitlab-based-website Pages). Sourceforge has a similar feature but I would discourage using it due to them mishandling code in the past.

>does that count as copyright infringement as well?
Probably. You could always not distribute the infringing parts and just write a program that takes them from their sources. See for example the SM64 rewrite, it can only compile if you have the original ROM.
Replies: >>36185
>Probably. You could always not distribute the infringing parts and just write a program that takes them from their sources. See for example the SM64 rewrite, it can only compile if you have the original ROM.
Then the mod cannot be played at all if I would have to autistically check of every assets that got added over the years where the majority of the source of it is not even known.  And my modelling/animation skill is not good enough to replace all the "copyright infringing" content with my own assets and even if I were good at it I'm not really willing to do that for essentially a shitpost mod. 

>Write your own webpage on neocities.org and it should generally be fine.
Hmm I see, then I would have to dust off my old neocities account and brush up the site so that it looks less like shit. 

> If your project is FOSS you can use Github/Gitlab Pages (or any-Gitlab-based-website Pages). Sourceforge has a similar feature but I would discourage using it due to them mishandling code in the past.
I already use Gitgud for Moonman Doom, Microtank and a couple other projects.
Replies: >>36196
I'm mostly just ranting how incomprehensible it is, but if you want to look at the code then here it is.
Replies: >>36191
(24.3KB, 312x180)
And here's what happens when I switch to my old text drawing function with no other changes. It works exactly as intended, except it doesn't draw the selection cursor.
As you said, the code regarding selections seem to be basically separate from the font drawing. Looking through it, it seems pretty sound to me. Assuming that the other surrounding code is correct.

At this point you're just gonna have to dive in and start debugging. The best place to start I think is verifying whether it actually fetches the correct characters out of the string as you parse through it. Would probably help as well to use a debugger like GDB and step through your code.
Replies: >>36195 >>36201
(249.2KB, 623x391)
I just started to match the old function to the new one bit by bit, and I made a discovery. There's only 1 image_fill_rect() function call (when the cursor doesn't have length), and if I change it to image_stroke_rect(), it magically works. The best part is that the latter just calls the former 4 times. Also, I tried changing the coordinates passed to image_fill_rect, and the character still broke but became a different one than before.

Basically the bug is some complete nonsense somewhere and possibly not related to anything that I've been working on. I'd normally assume that some of my data isn't allocated correctly, but if that was the case then the change should be permanent, not reset from moving the cursor.

I've been avoiding using a debugger because they require me to completely change how I've set up my project. Maybe I don't have a choice.
Replies: >>36198 >>36201
>Then the mod cannot be played at all if I would have to autistically check of every assets that got added over the years where the majority of the source of it is not even known.  And my modelling/animation skill is not good enough to replace all the "copyright infringing" content with my own assets and even if I were good at it I'm not really willing to do that for essentially a shitpost mod. 

What anon is saying is just distribute the mod as a patch. The patch contains only your original content (donut steel) and none of the copyrighted material from the base game. The patch can be legally distributed but is worthless unless the person playing it has the original game to apply it to. This is a common way a lot of romhacking websites get around the legalities of copyright. As for creating the patch, just have the same program that applies the patch also be the one that creates it. I don't really know enough about the technical side of it, but a while back when I played a lot of romhacks, the program they reccomended for patching the games was the same one used  by the romhackers to create their patches in the first place. It would take the hacked version of the game and the original, unhacked version, compare them, and strip out everything but the stuff created by the romhacker when creating the patch.
Replies: >>36199
>I've been avoiding using a debugger because they require me to completely change how I've set up my project. Maybe I don't have a choice.
Fuck you, set it up anyway.
(37KB, 445x402)
Such as doom.wad and doom2.wad? The mod can be played just fine with freedoom.wad. 
>The patch contains only your original content (donut steel)
The only original content I added is a shitty done spear weapon for Waffen SS/Gitler and a Chaingun 42 weapon for Gitler. So then I would have to provide a link at the readme that tells the user where to get the remaining Moonman assets and then run a python script that merges the 2 file together?  

The way (G)ZDoom or Zandronum handles mods is this, there is a commandline for iwad and another commandline for pwads, the user needs to define a iwad path and then load optional any additional pwad, so for moonman doom it would like this:
./ZDoom -iwad freedoom.wad -file moonman.pk3 mm2.wad (I'm not sure if this is the correct commands, I usually use a launcher like Rocket Launcher to start the engine) So from this command it tells ZDoom to load freedoom.wad first then load up moon man doom and finally memento mori 2 megawad. So I can just provide one of those source ports, freedoom wad and a bash/shell script that instantly starts the mod. 

Also holy shit, I looked at over 184 pages of the #fangame tage out of sheer curiosity and it looks like massive garbage, it is nothing but tumblrtale clones one after another, followed with a few pokemon clones, fucking baldi basic, sonic and couple other low effort trash. How the hell is one supposed to dig through all the games there that is not shit?
Put the mega links on some random wordpress site. Or put it on gitgud if it's opensource.
It just pastes the utf8 bytes on top of the uninitialized character data without setting it to 0 first. This is one of those "how in the motherfuck has it been working fine all this time?" moments. I was so convinced that the issue wasn't there.

What a waste of time. One of these days I'll learn to always initialize everything to zero.
Replies: >>36204 >>36224
Wew lad, the wonders of C. For what it's worth, I can explain why it was "working" all this time: The stack tends to be full of zero bytes. If it was passed through registers, these also tend to get cleared to zero often.

You seem pretty deep into reinvention territory and do some wonky things with mixed integer and floating point math that suggest low experience, so I can't help but ask: Are you deliberately going super low tech as a challenge/learning experience? If not, you might wanna consider something more high level, because you will have to deal with issues like these a lot.
Replies: >>36205
I just wish it was backwards, so everything is initialized to 0 by default unless you specifically tell it not to. Similar to how fallthrough/break in switch cases is backwards.

>The stack tends to be full of zero bytes. If it was passed through registers, these also tend to get cleared to zero often.
I can imagine that. I was thrown off by how consistent it was and how distance to cursor had no effect.

>deep into reinvention territory
That's because I fucking hate absolutely everything. I don't want to change everything about the way I build my project and install python and cmake and 16 other dependencies and tear my hair out deciphering shitty documentation for weeks and then insert 20 megabytes of bloat that may or may not do what I want it to do correctly into my project, just so I can do something that would actually be fairly straightforward and interesting to implement. This way I can make it do exactly what I want in the exact way that I want and can easily think of how to do it.

I have to deal with javascript and PHP and web software and work and I just want to kill myself the whole time.

>mixed integer and floating point math
I just assume floats auto-cast to an integer unless specifically used with floats, I hate manually casting things because it makes things unreadable. I'm more specific when I feel like it matters, for example I cast ints to floats if I'm expecting a float operation.
>and work
*at work
>I have to deal with javascript and PHP and web software at work
My condolences. I'm not going to get between a man and his therapy session.
>I just assume floats auto-cast to an integer unless specifically used with floats
It's exactly the other way around.
Replies: >>36209
(280.1KB, 128x128)
>It's exactly the other way around
Please no. Tell me no, even if you have to lie.
(260.9KB, 654x473)
>I just assume floats auto-cast to an integer unless specifically used with floats
Actually that can't be true, because the reason I started automatically casting ints to floats is because I kept having issues with ints flattening floats.
Replies: >>36213 >>36218
(174.4KB, 520x900)
They only get converted at the end when stored in an integral variable.
Replies: >>36217
That's fucking bullshit though. I ran into that problem numerous times where my floats were being flattened to either 1 or 0 where I was trying to get a value between 0-1.
Replies: >>36222
Don't use floats if you don't understand them or the language you're using.
Replies: >>36219
I'll keep using them and there's nothing you can do about it. Also the things people are telling me now are directly contradictory with my personal experiences.
Replies: >>36222 >>36309
(5.8KB, 200x200)
Can't tell if you're trolling or retarded right now.
Replies: >>36223
Are you? I don't know what else to tell you.
Replies: >>36228
Don't initialize everything to 0, your values are still not initialized if you do that.
Memory is initialized when you write meaningful values to it, if you just write zeroes that you don't want, you're only hiding bugs from tools like valgrind and UBsan.
This is why languages where all variables are 0-initialized for this purpose are dumb, you still have a bug if you think you have something you don't in the memory you read.
Speaking of those tools, https://clang-analyzer.llvm.org/ would have caught your problem unless it crossed over to an unincluded file.
Replies: >>36225
Why would I get a variable without putting something into it though? I can't recall anywhere I'd have done that. If I take a struct and then write into some part of it soon after, I'm going to run into a bug later if I didn't initialize the struct to 0 first, such as what happened with the utf8 thing. I can immediately think of a bunch of places that would be fucked for similar reasons if I didn't remember to zero initialize values, and I don't remember uninitialized data ever being easy to debug because the effect it has is completely unpredictable.
Replies: >>36233
>I don't remember uninitialized data ever being easy to debug because the effect it has is completely unpredictable.
Tools like valgrind, debuggers and sometimes even your compiler can detect uninitialized memory accesses. They are easier to debug if you actually use tools like you should. But harder if you're a printf spamming brainlet.
Replies: >>36227 >>36228
And if I don't want my entire development setup and file organization to revolve around debuggers, a zero value is much easier to detect than a random value. I've never had a problem with zeroed values before, but this is at least the second time that uninitialized data has had me chasing problems from the wrong place.
(38.8KB, 375x375)
All those tools can generally detect 0-initialized memory as well.

You're a fucking retard. Any arithmetic with floating point values is converted to floats then converted again for your final return value. If you can't even figure that out because muh personal experience I doubt you even code, just copy paste shit from the web.
Replies: >>36229
I can't remember what the context for it happening was exactly, but the return value was obviously not an 0-1 integer. It was either 0-1 float being used to create a X-Y integer, or an integer used in the process of creating a 0-1 float. Either way the resulting value was fucked and casting everything to float fixed it, and none of them were "copypasted from the web".

This is precisely where my assumption that floats automatically cast to ints came from.
Replies: >>36230
(31.5KB, 1505x274)
Post code or stop. https://cppinsights.io/
Replies: >>36231
I'd have posted it from the start if I had it, but I haven't ran into it in a long time because I learned not to do it.
I'll accept the possibility that I divided an integer by and integer or something similar, and it was so long ago I don't remember it anymore but the assumption about getting an int where I expected a float remained.
(116.3KB, 960x720)
>Why would I get a variable without putting something into it though? 
You wouldn't, you'd write meaningful values to it, these could be zeroes but not necessarily.
>I can't recall anywhere I'd have done that. If I take a struct and then write into some part of it soon after, I'm going to run into a bug later if I didn't initialize the struct to 0 first, such as what happened with the utf8 thing
If 0 is your "meaningful" or "default" value, sure. The takeaway is to write meaningful values. 
Writing zeroes by policy hides bugs instead of preventing them, the bug is still there and now debug tools think the zeroed memory is initialized, because in the C standard sense it is. 
If you 0-initialize a C string you want to write some text to but forget to write the text, your memory is initialized and you have a valid string but you still have a bug.
>I don't remember uninitialized data ever being easy to debug
Clang's memory sanitizer can do it at runtime, it should catch all uninitialized reads your program runs into during execution. There's also the previously mentioned static analyzer.
A good understanding of the implementation of C programs and the environment they run on also helps, experienced programs can tell the source of bugs from the patterns they create. A random unexpected character probably means uninitialized memory and/or a null terminator in the wrong place.
Replies: >>36236
>You wouldn't, you'd write meaningful values to it, these could be zeroes but not necessarily.
That's what I did, there's a whole process writing stuff to the struct, but some part of it was left uninitialized.

>If you 0-initialize a C string you want to write some text to but forget to write the text, your memory is initialized and you have a valid string but you still have a bug.
How does not initializing the text and letting it be something completely unpredictable (including possibly/most likely 0s) make it any better or easier to detect? It's also much easier to have and detect an anomaly in a long string than it is to detect an anomaly in a single value.

>A random unexpected character probably means uninitialized memory and/or a null terminator in the wrong place.
No it doesn't, that could mean absolutely anything because potentially a lot of things interact with the data, that's why I spent a lot of time looking for what's changing the value. It was changing to the same character (relative to something else) every time, it only happened under specific conditions to only one character out of all the characters and it had never happened before anywhere else I used the same function, there was a pattern and consistency and precision to it despite it being "random".

If the data is 0 then you pretty much know it's uninitialized unless 0 is an expectable value. It's much easier to handle data if you assume it's zeroed by default, and you won't run into problems if you decide to add a new value to a struct later. I won't argue whether a debugger is better at detecting which, but as a human it's much easier to zero initialize things.
Why would it otherwise? Float is the wider type.
(290.2KB, 636x437)
>129 wishlists
it's $10, for fuck's sake just buy it
Replies: >>36256 >>36264
*$12, whatever the principle still stands
Replies: >>36261 >>36264
If you buy any game for more than $5, you're stupid.
Replies: >>36264
Maybe make it 9,99 for that classic "it's totally below 10 dollars goy".
If you buy any game, you're stupid
Replies: >>36303
If you buy anything, you're stupid
Replies: >>36306
>having money at all
Have fun being a puppet to the kike scheme.
Replies: >>36313
Your wilful ignorance makes me happy about the fact that your games will work - very subtly - different to how you think they will and will give you all sorts of maddeningly hard-to-track-down bugs.
Replies: >>36327
(42.9KB, 417x640)
>living in a ZOG state
lmao you will never be truly free
You seem to operate under the impression that people can't learn or change at all. When I've confirmed through personal experience that something is a certain way, then I'm not going to listen to anyone on the internet who tells me otherwise unless they or I can first prove that my previous experience might have been somehow mistaken. But if it turns out that it was mistaken, then I appreciate learning that, just like I appreciate having learned this thing about floats.
(851.1KB, 1920x1080)
(887.1KB, 1920x1080)
I've got some of the basic functionalities of the class switcher script to work now. The main difference between the HPIc menu and the class switcher script is that the class switcher script handles the keyboard input much better than the HPIc script does as I've managed to delegate the cursor mover function to a separate function instead of doing cursor input at the NetworkProcess function itself so which now calls the function and passes 3 arguments to prevent array out of bound error. Though writing this script is slightly easier as I don't have to deal with grids as I'm not planning to add that many tanks maybe in total 30 different tank types to chosse from so a simple list should be sufficient for that.

When the class switcher script is done I will rewrite the HPIc menu so that it shows different items so that for example a Panzer III cannot buy the 105mm Cannon as its chassis is too weak for that.
Replies: >>36475 >>36621
You're all playing to lose.
Replies: >>36459
(35.2KB, 680x722)
It could be argued that all truly creative endeavours include "playing to lose" at some level since creating something good has different implications and conditions compared to 'winning',though in the end that would fall within the purview of your conundrums of philosophy.
Is this really doom?
Replies: >>36481
(1.2MB, 1920x1080)
It's GZDoom 4.5.0, using freedoom2.wad as a iwad.
>Game Jolt
Isn't that like we well only do business with you if you happen to live in one of the 5 countries we care about but we only tell you the list of those countries after you register and fill out all your possible personal data and connect your paypal account?
(1.1MB, 1920x1080)
(1.2MB, 1920x1080)
(1.1MB, 1920x1080)
(712.6KB, 1920x1080)
And here is the main functionality being shown now. Right now the changes are visual only, I cannot even add any code that handles the weapon switches as I have only 2 weapon (105mm Cannon, Multi Purpose Device). Yes I know the model texturing is inconsistent but I cannot be bothered to fix it now.
Replies: >>36678
> I cannot even add any code that handles the weapon switches as I have only 2 weapon
Why ? Engine/script limitations or just don't need them ?
Replies: >>36685
Its because I want the weapon code written in a way I have to update it only from one file/place instead of having to check the same code several times over whose difference is only different firing rate and different projectiles that it can fire. So for example the I have several classes one is "MT_Cannon" and the other is "MT_105mmCannon", "MT_75mmCannon". The class "MT_Cannon" should do all functionalities needed for the weapon to work, any class that is inheriting from "MT_Cannon" should only define properties and nothing else such as firing rate and the ammunition types it can fire. 

I would like to try doing something similar for the projectiles because right now to make the damage upgrade to work I have to define the same actor again just to change the variable which looks stupid too me. The worst case is the APHE/U-APFSDS projectile which has 16 actors just for the damage upgrades and to make the projectile dropping work because FastProjectile actors does not support gravity whose reason is unknown to me, so I had to deploy this ugly hack. 

Then I would also have to figure out how I want the Panzer III/Panther I tank to work because both are classified as medium tank yet the Panther 1 has a 75mm cannon which is getting close to heavy tank classification, so for the "Medium" tank category I have would have caliber upto 75mm and the "Heavy" tank category calibers upto 90mm, the "MBT" category is limited to 105mm calibers. The Panzer IV is a medium tank too with a 75mm cannon but it does not feature sloped armor, so theoretically the Panther 1 tank would have more armor (=hitpoints).
Replies: >>36720
Don't hardcode the tanks. It's more tedious, but you can save yourself a lot of headaches (and open up your mod to others who want to put in a Ratte or something).
Replies: >>36723 >>37056
There isn't for me any way to not make it hardcoded as GZDoom doesn't support virtual filesystem as in a code that writes/reads custom binary/text files, besides even if I would use just the standard ZDoom class system the buy menu needs still to account of what player class the player is so that it shows class appropriate items, for example when the player selected the Panzer III Ausf. J then he cannot buy items that are made for the heavy/MBT classes.
Replies: >>36729
I was thinking you could specify compatible parts by just listing strings in the tank file, but I didn't know the engine was gimped that hard.
Replies: >>37056
(344.1KB, 566x705)
(2MB, 1600x900)
What does /agdg/ think of FOVO? Basically, instead of doing the standard perspective projection, they use a nonlinear projection derived from empirical studies. Despite sounding like bullshit (linear perspective projection is correct for a screen, AFAICT) it looks pretty good and apparently aids in judging distances. I'm too retarded to figure out how to implement it from what little they've published, though. Using their wireframe as a guide, you can get an idea of how to project the image in that particular scenario - but it's unclear how this is parameterized over FOV. Does anyone here have any experience or ideas? I figure the best way to go about this is to write a general system for warped perspectives via a grid, fit their wireframe to it, and then interpolate between that and the linear perspective. However, as mentioned in the gamasutra article, the projection appears to involve some vertex shader fuckery so there's obviously something missing from this plan.

Replies: >>36978 >>36982
>make a system that will probably be re-used in basically all of my projects
>it's nice, simple, straightforward, but extremely useful
>add the final piece
>it turns into a hideously convoluted monstrosity
>it's essential so I have to have it one way or another
Why does it have to be this way. Either I have to make a separate version of the library for each project, or write some annoying roundabout abstraction layer path for it so I can inject that last piece from the program code into the library.

I think it's a meme in the same sense that lens flares and such are. People go "OOOOOOOOO" because it mimics film/photo optics even though it's not necessarily how our brains/eyes actually perceive things, perspective distortions like this can even cause motion sickness ingame. I can imagine it being good for VR though.
Replies: >>36980 >>36985
(644.4KB, 1591x523)
It's the opposite, instead of drawing things the way it would be on a photograph, where the single lens causes an unrealistic perspective warp, this does what your brain does using both your eyes.
Humans have about 200-220 degree horizontal vision, if you count the parts that are only seen by one eye. If you only count the parts that are seen by both eyes, and combined into one image by your brain, that's about 120 degrees. Now think back to what 120 degree FOV looks like in a video game. Quake pro retardozoom. In fact, the wrongness of how games handle camera projection is exactly why some people get sick even from normal FPS games, even without it being VRshit. And it's why people tend to look very different in photographs, unless you use particular lenses or do some fuckery that smarthphones use to try and remedy it.
Now I only know a bit about that Blinky thing the other anon linked to, because I looked into it before. That just renders more, instead of a single projection plane it uses multiple to make a 3D shpe, and then distorts that to make a final image with a higher FOV without it looking (as) retarded. I think this FOVO stuff is supposed to be easier on the rendering than just drawing multiple screens and distoring the result, so it should be superior.

That said, for 99% of people used to videogames, at normal FOV it won't matter. It might help the few people who do get sick from FPS games. It will certainly help people who play at high FOV to gain an edge. 
And for VR, where you need to render at a high FOV it'll probably help, since what they were currently doing IIRC was rendering at a way higher resolution and distorting it with post processing. The high resolution is needed to get enough data to blow up the center while the edges get squished. So if this technique is faster than rendering at a higher resolution, it should make VR games more performant.

But I'm not reading a fucking Gaymasutra article, because fuck those SJW-infested cunts, and the forum link doesn't provide much info so all I can say is "I dunno lol."
I found an HTML5 demo of something similar years ago. Instead of doing post-processing on each frame, the raycast angles for each pixel were adjusted according to frustrum res, desired FoV and a curve slider. It was a bit difficult to find the right balance, but it could be done with some patience.
Replies: >>36985
(1.2MB, 1600x900)
(16.7KB, 480x480)
(15KB, 480x480)
Even if it's a meme, I think there's a lot of value in using nonlinear projections to put emphasis on certain parts of the scene; a game that tries to communicate extreme differences in scale might greatly benefit from this.
The papers I attached have all the relevant detail, what little there is. I linked the gamasutra article mostly because it serves as a nice overview from a gamedev perspective, with the useful observation that the GIF shows the use of vertex shaders in their implementation. Actually, after reading your post, I realized the reason for the vertex shader stuff is that FOVO is meant to map binocular vision to a flat plane - the distortion is 3D instead of 2D. Vertices towards the middle of the screen are pulled in because they're between the eyes and therefore get seen from both sides, for instance. So, working from that direction, I imagine that FOVO is implemented entirely as a vertex shader. This would explain why it's so fast compared to blinky, although it raises the question of how its made to play nice with other effects such as shadows. Now to get off my ass and actually program this.

Also, just going to call attention to the use of vignetting to hide stuff outside the window, and how FOVO resembles something between barrel and pincushion distortion.
It's a shame there's not more exploration of this topic. It always bothered me how renders and photographs, despite being more correct than drawings, tend to look subtly wrong when there's significant foreshortening or large objects close to the camera.
You're neglecting 2 important factors.

Firstly 2 different people will have a different sized monitor at a completely different position and distance from their face, so the perspective distortion would have to be calibrated for each user individually. The reason AAA niggers often use a too low FOV by default is because they assume people will be using consoleshit and looking at a TV from their couch, and at that distance a low FOV is more natural because the screen occupies a narrower space in your vision.

Secondly your "eyes" are always looking at the dead center of the "camera" because your eyes are the camera, whereas in a videogame your eyes are looking all over the screen. If you turn your eyes and look to your side, a straight object won't look all curved and weird, but if you distorted the perspective in a videogame like this then things at the sides will look curved, the distortion would have to be fixed according to where your eyes are looking. The camera in a videogame is more equivalent to the orientation of your head than of your eyes.
(833.2KB, 1920x1088, 00:01)
(10.6MB, 1920x1080, 00:15)
(1.2MB, 864x464, 00:17)
>try to refactor the cannon weapons
>make a first attempt at a ZScript only ammo switcher system
>the first method works fine and I could select 2 differrent ammo system
>try to define a static array to define the ammunitions in a way where inherted weapons gets different array struct definition
>ERROR: cannot redefine array
>fine I'll just use a dynamic array then
>ERROR: cannot cast pointer <class<Cannon_Ammunition>> to CA
>VM ABORT:  (I forget its error code)
Fuck this shit, god fucking damn it. This is fucking retarded, why the fuck I am not allowed to have dynamic array of structs? I fucking need this shit.
Replies: >>37043
(2.9MB, 200x234)
Use the power of absolutely bullshit data structuring,anon.Believe in the programmer within you.Below I have provided an absolute shit tier solution if you can't think of anything.
Define every single possible struct in a static array and then keep an array of bools/ints of the same size that tracks which ones are actually available.:^)
Replies: >>37049
(1.2MB, 234x221)
Alright then I will try with this niggerlicious method. Of course Graf Zahl has to find new means to fuck up me again.
It seems I stand corrected, I just looked at this documentation (https://github.com/marrub--/zdoom-doc/blob/master/api/files/Wads.md) and there seems to be a method to read if a specific wad file is available and reading its lump content which is read as a string so I would have to write a function that reads a specific file format. 

I will look into this later and see if I can make my stuff less hardcoded, for now I just want shit to get working in the first place before I start another code refactor. Man the few weeks I spent fucking around with Godot was less painful then fiddling around with ZScript, what a horrible mess it is.
wait when's demo day again?
Replies: >>37192
On 3/3, see >>36114. More importantly, who's going to make the new thread? We're about to hit the limit.
Replies: >>37197 >>37203
(54.4KB, 421x408)
>next demo is on 3/3
Oh shit, I don't think I'm fast enough to write 3 functional scripts in such a short time.
(38.2KB, 595x842)
Wasn't it supposed to be 2/2? Like in pic here >>34239 or pic related
Always keep in mind that I may be retarded. Did we lower the frequency because we're so slow?
Replies: >>37233 >>37234
(210.3KB, 1920x1080)
It's always been 2.2.
Nobody cared because there's only like 2 and a half people here working on games.
It happened on the agdg board, lads.
Replies: >>37254
There's an agdg board? I mean outside markchan, I'm not touching that shit

450 replies | 184 files | 133 UIDs
Show Post Actions



- news - rules - faq - source code -