/agdg/ + /vm/ ~ Amateur Gamedev General

Is there a good FOSS 3D RTS Engine that doesn't have a absolute shit tier pathfinding?

Hey. I'm too depressed, impatient and self loathing to try to code, program or draw. I'm nothing but an ADHD ridden bastard child. All I want to say is, I want to wish all of you that are making an attempt here the best of luck. Even praying a bit.

Don't be like me. Be like you, if you've already learned. Then you've done the hardest part, all that it takes is patience, self-love and a vision. That vision is gonna get you far. Don't give up.

You could always get into DIY wood stuff.

Does 0AD have good pathfinding?

Still have quite a bit of poly reduction to do with this model as 3,000 tris is a bit much for a gun attachment, but the basic shape is pretty much done.

Also I have an idea I want to run past the people here while I'm thinking about it. One of the big downsides of having a holographic sight in a game is how it can take up so much of the screen like in pic related. However in real life that problem doesn't exist as both eyes are open and you are focusing past the body of the sight. I want to emulate this and I'm thinking that I could do so by selecting only the sight using custom depth and applying a significant Gaussian blur to it. Is there a better way of doing this?

Thanks man. I appreciate you. I hope you can find a hobby that lets you make something you enjoy.

Hi me

That doesn't stop me from drawing and making game though. The only one who will try to stop you is yourself, and the only one who can decide not to stop is yourself. The decision of whether to use your time doing things you like or toss it into a ditch is ultimately on you.

alright, I fixed my vulkan issue from one of the stupidest bugs ever

so my problem was this line of code:

res = vkr_create_buffer(vkr,&vkr->stage_buf,&vkr->stage_mem, ub_size ,VK_BUFFER_USAGE_TRANSFER_SRC_BIT , VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT | VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT);

Fixed too:

res = vkr_create_buffer(vkr,&vkr->stage_buf,&vkr->stage_mem, ub_size ,VK_BUFFER_USAGE_TRANSFER_SRC_BIT | VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT);

The problem is that I mixed in the buffer usage flag VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT with the memory usage flags, meaning that it was asking for memory with invalid parameters and failing to allocate, which then made the function fail, which then made another function fail, and I didnt put a sys_warning ANYWHERE in this, and i explicitly did not handle the error code coming from the last function to fail because of this. But now I have what I was looking for: I can now apply matrices to each VBO individually, meaning that I can keep working on my program.

Now in a situation where this error comes up with a second draw call

going to figure it out tomorrow

Do that, and fade out everything except the edges. That's how your eyes would handle it.

Do you have an image so I can better visualize exactly what you mean? It seems like your saying make everything but the edges of the sight transparent. This is what it looks like if I just do the blur effect.

That blur seems a bit too extreme to me.

Covering one eye doesn't actually make things appear darker.
When you've got something in front of one eye only, everything except the edges and fine detail is filtered out. Try it yourself.
Blurring would also occur due to lack of focus, but as said, it wouldn't be that intense.

but that's wrong though

Close one of your eyes, put your finger in front of your other eye, and then focus on something in your monitor instead of the finger.

Soon, I shall begin testing around.

22 Days until i can leave my job and devote myself to my game

Oh boy, I know that feel.
I suppose it helps that I'm working on so many things at once. If I get tired of conversation mapping, which I easily do, I go try writing some music. Or sometimes go do some drawing. It kinda sucks that I have a lot of the generic coding out of the way, so I gotta do the art and conversation mapping first before I move on with that any further.

Deminds me of pic related.

If you're not already devoted to your game, X future event probably won't make you either. It's just an excuse not to do it right now, you'll have another excuse afterwards.

I can't devote to my game the 8 (+2 for transportation) hours/day i have to spend at my dayjob user
I do work on it on my freetime

Don't believe in yourself, believe in me that believes in you, do not give up!
If you have ADHD problems don't try to work on a rigid schedule, try to do a bit of everything
One day work on drawning, another on code, next on 3D modeling
Do what you feel like doing, but JUST DO IT

MoM user, With Another Video this time Covering the Water Tome Spell Book! (As well as redoing the portraits for the characters so they are sprite based!)

Questions are always welcome!

Really nice, but isn't this tome more of a "Time & Space" thing than water?
What about a water tornado that hurs enemies? and if you use it on water it becomes a tenctacle that drags enemies into the water and drown them
Or a bubble spell that reflects projectiles and bounces arround

Lloyd's Beacon is typically a water spell in RPGs such as Might and Magic, as they generally use water fountains or puddles to travel Around.

The Time Freeze aspect really solely due to the fact it has the Word Freeze in there, which Is a bit silly. Spells are suppose to act more so as support then actual direct combat however, such as buffing, sentries, and so forth.

What about making this the Ice tome instead then? Lloyd's Beacon looks like an ice crystal and all, you could replace the elementals with a frost armor or ice wall

Does your setting has only the basic elements (water,fire,earth,air)? Or are there more?

They is Dark and Light tomes, too. Also, Wouldn't Ice fall under the category of water?

In my setting Ice(cold) is a composite of water+shadow
I have 18 elements in total (6 + 12 composites)

From the short testing play I made the pathfinding doesn't seem to be absolute garbage as it is the case with Spring Engine I tested like 3-4 games or so on Spring and all those games have shit tier pathfinding such as units getting clumped, units pushing the other units away and thus breaking the formation despite the unit could move around it and much more also the formation widget it comes by default on all those new Spring engine is shit too.

Direction and cinematography as they relate to musicality, applied to all things, but especially to controls and character animations. People don't seem to understand that everything has a rhythm and dance to it, from cutscenes, to sound design, to controls, to even menus. When I touch a game, this is always the first thing I notice, and the games that have it are almost always the ones that stand the test of time.

I think lots of people in the modern western industry are effectively failed film students who are mainly interested in making cool visuals or telling stories, and they don't understand that these things are secondary to the core of video games, which is the gameplay. Visuals and stories are well and good, but there's something that they can never capture, which is the player's participation in the rhythm. There are emotions that come with it, and they vary depending on what the activity is, ranging from the joy of dance to the thrill of the kill. Well-animated character action games are exemplars of this, and it's not just because they're well-animated, but because the timing is built right into the controls, and the two synergize.

Vid somewhat related, at 4:16.

Will we ever see your boipucci?

Anyone here need a writer? I'm willing to work for exposure. Just need to really GET THOSE NEURONS A-FIRIN'.

Speedbot user, you need to translate your game into other languages, as spanish and french. Doing so will give you much more exposure.

i think this guy might need your help user

pitch me a story for an RTS game centered around a rebellion.

civil war 2 electric bungalow

federal authority is getting out of hand as a few major cities dictate how smaller rural peoples live out of touch with the day to day lives of the rural population the urban-ites pass laws that make day to day living more and more difficult. from minor things like export taxes eating into their ability to export goods competitively. to endless "anti-pollution" measures that kill the small family owned businesses replacing them with larger corporations that can afford lobbyists. larger and larger portions of the native rural population are finding it harder to find work, wile the final straw takes place as violent foreigners are dumped into once tight nit culturally and ethnically homogeneous communities.

take back district by district slowly cutting off food and materials to cities. liberating choked counties giving control back to the people that live their. interrupting manufacturing and supplying of the federal military. disrupt food, water and electricity to cause riots to disrupt day to day life, forcing a splitting of forces to controlling the cites.

you win the game when every jew is dead and the cities are burnt to the ground and a new confederacy of sovereign states are formed.

Why is my puzzle game character collecting gems in a horrible amalgamation of room themes and styles

o yeah i guess i forgot to mention the rebellion is supposed to be against an alien race that have occupied the planet. The occupation happened sometime like maybe around 2030 or something, so it's technically sci fi but most of the alien technology comes from the aliens.
The aliens are here for unknown purposes but they have been integrating humans into their force both as workers and as fighters but with heavy genetic and cybernetic modifications.
All of this is heavily inspired by the rebellion depicted in half life 2
So i was thinking more a story from a more personal angle kinda like Freeman's story in half life 2 but not focused entirely on 1 character like half life 2 but more focused on the important people in the rebellion.

Don't do this shit, it releases mustard gas.

so x-com? there are 2-3 levels here depending on where you want your story to take place the local command the individual and the global. you want a character? who the fuck is he. and why wont he just get killed right away? why are their alliums and why do they need people/ earth? why not nuke from orbit? if they need earth intact then why not genocide the natives or are they doing that? why do they need the extra manpower? are they exporting goods or are they colonizing?

...

i'm gamedeving full time and its going to take me like other two years, unless you have that much money i would advice you to not to

I'll keep that in mind, though as said you don't close an eye when you look through a holographic sight. But the effect with both eyes seems to be pretty much the same when I try it out. I can use fresnell to select out the edges in a relatively smooth way before applying a transparency and then applying a blur after the fact. It will take some messing around with the numbers to get it looking right but it should work.

maybe the people are the native people are the point of the invasion or at least part of it.
Anyway i thought i would be interesting to maybe do 1 mission with an individual then go to a local level and then near end go all the way to global. I just thought it might be fun to have some drama in there.

at least I'll be able to make progress soon I guess.

at least yandere dev made a game

Im making a fantasy world setting
WITHOUT GODS OR DIVINE BEINGS
Now, to fit my 18 elements into the universe physics…

I am severly retarded and want to move this box once a second but it doesn't move
Please bully me until i understand whats wrong with this code
#include int main(){ sf::Clock frame_clock; sf::Time frame_limit = sf::seconds(1.0f / 60.0f); sf::Time frame_sum; sf::Time frame_time; int frame_count; sf::RenderWindow window(sf::VideoMode(800, 600), "gaem", sf::Style::Close); window.setFramerateLimit(60); sf::RectangleShape box(sf::Vector2f(50, 50)); box.setPosition(10, 10); box.setFillColor(sf::Color(150, 50, 250)); box.setOutlineThickness(1); box.setOutlineColor(sf::Color(250, 150, 100)); while (window.isOpen()) { // returns time since the last restart frame_time = frame_clock.restart(); sf::Event event; while (window.pollEvent(event)) { switch (event.type) { case sf::Event::Closed: window.close(); break; default: break; } } frame_sum += frame_time; if (frame_sum.asSeconds() >= 1) { box.move(0, 1); frame_sum = sf::seconds(0.0f); } window.clear(sf::Color::Black); window.draw(box); window.display(); } return 0;}

`>>12572288
the elements are your gods

How could you mess that reply?
You better be ready for the bullying
The magical elements are tools for the human race to master

Is there no hope for man?

your mum is a tool for the human race to master

Man can create it's own hope

What does your debugger say? Is the time sum actually increasing? If not, why not? If it is, then why is the check failing?

This is a very simple problem, cmon user use your fundamentals

the frame_sum increases properly until it hits ~1 and then goes back to zero but the box never moves.

I am more interested to know if this general approach is retarded and if there is a better way to implement timers in SFML

Thats a framework written in c++. Sfml isnt a language so you cant "implement" anything in it.

Your timer seems fine though it looks like your move function is fucky

its easy to find out how long time has passed:

#include clock_t start = clock();while(1){ clock_t end = clock(); if(end - start > CLOCKS_PER_SEC){ move_box(); start = end; }}

Your first issue is that frame_sum is NOT initialized anywhere, meaning that the first value of frame_sum is undefined.

What is the percision of your clock? Because I suspect that this is what is happening:

The clock only gives time in milliseconds.
Each frame takes less than a millisecond (over 1000 FPS)
Therefore every frame takes ZERO milliseconds.
Therefore frame_sum does not increase.

I find this macro funny, not sure why.

What do you guys think of Udemy? Apparently they are having a sale right now.

I guess it depends on the specific course. What are you looking at in particular?

I use it when I have the time (which isn't a lot). I imagine it depends on what tutor/course you get, my coding course was frustrating but I'm doing other courses and they overall seem good. They also have forums which you're ideally suppose to use but I never do.
I've also noticed they do sales relatively often so don't feel like you need to decide now if you're not sure.

Pretty good but really depends on the course, though you can probably find most of their shit for free on cgpeers.

Not really sure. I guess some of the engine courses that teach Unreal and Unity. Might pick up some coding courses too since you're supposed to have lifetime access. Any recommendations?

Well shit, it's the first I've heard of it. How does it work?

It's a torrent site specifically for GFX and tech stuff. Registration is usually open on the 1st and 15th of the month.

Gimme a little more to work with and I'll see what I can do.

really? I've had an accoutn there for ages but thought it was just useless graphics shit

...

don't be a nigger

(noice)
At least get a part time job, man.
Don't starve yourself

Everything will be fine, i have savings

how many savings?

...

He's implementing a timing system using the classes of the framework, there's nothing wrong with his verbiage.

Enough for a year, if by then im getting 800$/month on a patreon account i will be earning more than my current job

I have started the project months ago, demo soon
Gotta take risks to follow your dream bro

...

Hospital, Nursery, School and Gym events halfway done

It is hard, however someone did all the work for you and made the modeling program

do you think 2D artists need to code their own software as well?

...

I didnt say that- rather, I said that the hardest problems with 3D have been solved for you.

unless you work in SISGRAPH most of the 3D big problems have been solved by now.

that is true, most of the problems were also solved for the people who wrote the modeling program, too

this is correct, the main limitation now is hardware to implement more realistic simulations.

but most of the problems are well understood now, and what is left are more features that help with the workflow or simulation details.

I made a thing a week ago.

Candlejack? No, will not be in t

Will this game include implicit or explicit lewdness? How does the face generation work?

What is this?

Implicit, it will be a farmsim with some extra fantasy mechanics, there will be plenty waifus, no need to rape the faeries

it's me messing with opengl.

Oh yeah forgot to talk about the faces
I drew several presets for different parts and made an algoritm to assemble the faces based on the faerie traits (cheerful faeries smile more, lewder faeries have full lips, etc…)

Blocking faerie panels on event throw done, now i just need to implement the events and their over time effects

oh uh sorry for accidentally replying to you

That sounds awesome, I'm looking forward to a release. The face generation sounds like something I've been interested in trying for a while. It looks like it's working great for you, I especially like the various pupil shapes

lets be friends!

Even the pupils are trait-based, it was a ton of work, but it was all worth in the end
The traits are also easy to mod, i keep them on an external .txt file, and to add a new one you only need to add a line of code and an icon

why aren't you making a 3D game?

because i have to finish this UI based demo before i jump back into making the main game

Sure, let me just start up Autodesk Handmaker, Autodesk Facemaker, Autodesk Treemaker, Autodesk Sandwichmaker…

Because it took you 20 minutes to make a shoe mesh and that's before any part of the actual game engine enters play

Are you saying you learned from your mistakes, and used a text file rather than five thousand if statements?

can you stop shitposting about your stupid fucking shoes?

I think you're being too optimistic, we havent seen the inside of that text file yet

more like three minutes if you count program start up there is no fucking reason for gimp and krita to take longer than blender

Maybe the text file contains five thousand if statements.

The ONLY content so far that is not in external txt are the recipe archetypes, they are too complex, i will have to develop some sort of script language for it, so i will do it later when i port the system to Unity

It will be properly documented later

did you type this out by hand user?

They really aren't.

Well, someone had to make sure it is properly balanced
What would be the point if i let the computer generate important stuff randomly?

Here your (you)

I just open notepad and write "shoe"

...

Okay, so this is my progress in the new engine.

The two renderers are almost the same, right now openGL has support for textures and so can render my font out nicely. It took me a while to get vulkan set up so that I can apply a different matrix to each VBO, but now that is done with. The cubes look different because they are spinning, but its the same cube.

Right now I have 3450 lines of code in my project, so even though it doesn't look like a lot, there is a lot going on behind the scenes that will be useful later.

Just repostan these so I can show someone my progress

Are you doing just the engine or the whole game from the ground up?

What is yours about?

...

I'm writing the engine, and then the game will use the engine. All from the ground up.

Puzzle game like Chips Challenge or Supaplex

Holy shit thats some hard work, good luck

Like what?

the lack of asymmetry makes this look really stiff, take a look at force by micheal mattesi

This is my previous work, it runs in my "Sigma I" engine: ika.itch.io/redsky

So I feel confident enough working on Sigma II with that under my belt.

Please don't tell me you are making a whole engine to make simple FPS game, that would be a huge waste.

FPS games are not simple at all… What kind of game would not be a waste to write an entire engine for?

Wow, Vulkun doesn't look any better, what a shitty meme.

well, I told it to draw the exact same thing, so of course. Later i'll probably add in some fancy effects into the vulkan renderer just so that there is some kind of point to using it.

Pretty sure he's fucking with you.

I know

...

Some of the major accomplishments in videogame programming are found in "simple FPS" engines. Maybe you've heard of idtech or unreal?

werks 4 me

I have a few opinions:
Change to:
frame_sum -= sf::seconds(1);if (frame_sum > some_threshold){cout

Anything math related. I love math, but tend to over-think everything far beyond what is likely the optimal solution.

>just spent the last 2 days relearning physics and flight dynamics
Oddly enough, I regret nothing.

Ho-ly Jesus. What is THAT? What the fuck is THAT?

That doesn't seem right.

frame_sum -= sf::seconds(1);if (frame_sum > some_threshold){cout

feature creep

Just some of the end results of a couple pages of calculations and scratch notes. I'm not sure about the first formula, but the second is not as wildly inaccurate as I initially expected.

The first is a turn rate function with respect to elevator/rudder fin angles and airplane velocity. Most of the other terms are constants derived from the plane's characteristics, so it only looks more complicated than it really is.

The latter is a function for determining the density of the gas the plane will be flying through, with respect to its altitude and whatever the gas is made of. Air density has a direct impact on an aircraft's lift and drag. I'm thinking of simplifying things further by using an array of pre-calculated densities using this formula, and only changing the assumed density every several thousand meters. My game world will also have at least two different gas layers, which will affect flight performance per the formula.

These are actually over-simplifications of the real physics formulas, but the latter only nets errors of about 5% so I think it's plenty good enough for what I intend.

...

That just means you didn't need it!

Try removing more lines, maybe your code will be more efficient.

Something that current engines cannot achieve, why develop an engine when there are so many out there? Unless you are doing this as a hobby and having FUN
All this effort could be used to make a game much faster if you worked on something that is already halfway done, unless your goal is not finishing the game, but the engine…

Just saying that you don't need to reinvent the wheel

It's hard to overstate how useful making an engine is as a learning experience.

You most certainly shouldn't do it if you want a game out soon or dislike programming, but you should definitely consider giving it a try if you want to get better at programming and have the time to spare.

Having your own engine also has it's own advantages. You know it through and through and can easily make it do anything your programming capabilities allow you, without having to google how to do X or Y in unity or why it is giving you an error or shifting through all the padding to do something simple.

Factorio, Terraria, Minecraft, and The Witness all use their own "engine" and they're all exceptional games in their own ways.

Decompile your code; that line was probably optimized away by your compiler.

Couldn't you make a "rope" by just clamping the object's location inside a sphere and attaching a rope visual from the origin to the object?

Both true for me

Can i just make a rope and kill myself?

But how will you make game if you're dead?

but if he's dead, then he won't have to be upset about NOT making a game!

I have to ask, why did you make a vulkan renderer, instead of an OpenGL 4.5 one.

kek the thumbnail to that video is a meme by itself

Reinventing the wheel is not the correct analogy, your position is more analogous to "because someone has painted the Mona Lisa, portrait painting in oil is now a worthless endeavor"

"Meme" doesn't refer to single images with funny text you kikebook using refugee scum

If you made a game about Richard Dawkins taking over the with the power of meme evolution, who would the enemies and Final Boss be?

Wasn't he already working on that?

That comparison would only hold water if this was AEDG.

That was a game using the meme of evolutionary genes, but this would be a game with pure memes, and I accidentally the whole universe:

You wanna know the saddest thing?
The game with actual memes would have met it's kikestarter goal.You know Reddit would have thrown their life savings at it faster than Bernie supporters did.

Yep.

Reddit would have to be the final boss: Its hugbox would suck in and destroy all creativity until the memepool was too stagnant to survive.

B-but muh Kekistan!! Look at our diverse, Kekistani niggers and vaginas!

That's probably true but I strongly believe that engine development is a useful and worthwhile pursuit

okay then

Slapping together a triple-triad-meets-Magic prototype in löve this afternoon.

...

A gamedev can't create games without an engine the same way a painter can't paint without paint. All painters in the past mixed their own paint. Nowadays you can buy premade boxes of different shades of paints that you can use to create a painting without mixing them, they're just expensive. And you can still mix paints yourself to get the exact shade you want. You can add thinner to give it the consistency you want. But oy vey, this is APaintingDG, not APaintDG!

I fucking knew that I'd regret not posting "analogies are by the retarded and for the retarded" the first time. Every time you point out one of the thousand ways in which an analogy is stupid, somebody comes and stretches terms and concepts until they are worthless, and for what? To keep the stupid analogy alive and nothing else, because who cares about the original point anyway.

Said point being that most great games have been made in engines written by people other than the game authors and your hobby programming project is not going to outdo commercial engines. Comparing yourself as a single person with idTech and Epic Games doesn't mean you're ambitious, it means you lost contact with reality and will burn out after a few months. Argue hypotheticals about painting all you want, it won't change this.

I don't know if I should be impressed or jealous.

The game looks like ass and shit though.

Enjoy mixing paint for years before you actually get to painting.

Your analogy sucks.

This just makes me think of Bob Ross for some reason.

You never know, sometimes the autistic lightning strikes.

Ol' Bob wasn't the best painter skill wise, but he was goddamn fast for the sake of presentation. Y'all should take note.

Most people aren't gonna be able to pull a Toady.

most people will die mediocre.

True. Unless you're incredibly stubborn and learn what makes something truly good, you're better off working as part of a team or getting a stable job.

That's true as well but in the end engine building is more complex and essential than mixing paints is to painting.

"Outdoing" commercial engines is not the point of making an engine. Many people who make games don't do so to compete with with other's games, topping "great games" is not really the point. To stretch your favorite painting analogy even further: people don't paint to somehow beat davinci and the mona lisa, they do it because they have something they want to express.

Curiously omitted from your post: The point of making an engine.

It's part of the expression of game making, some people have ideas that stretch to the bottom of their programming abstractions.

Stop whining and just like make game. There are many successful indie games that used a big engine, and some of the most successful indie games including the most popular computer videogame of all time uses their own engine. Meanwhile you've made jack shit and are trying to tell other people what to do, this kind of arguing is 4/agdg/-tier nodev garbage.

This couldn't be more vague blather if it tried.

Bloated proprietary engines aren't anywhere near the Mona Lisa, something like the older id Tech engines would be closer.

Just because you don't understand it doesn't make it vague. I'll rephrase it if you're having trouble: "Some of the things people want to express when making games exist at a level you can't see in the finished product, they are ideas about the lower level functioning of their games as a program"

your doing it wrong and should just stop then.

I agree, I guess those modern engines are more like the works of Jeff Koons

I can't wait to see your new hit game user, I'm sure it will make all established developers look like fools

…really, mate?
If your game is just "X GAME BUT LESS BAD", then it doesn't have much a reason to exist, as at least seventy versions of "X GAME BUT LESS BAD" already exist, all "less bad" in their own ways.

Do something original instead. Something that'd be the oranges to everyone's apples, since this thread loves analogies apparently. I'm not saying make your own genre, that's ridiculous.
I'm saying make something that could not be described using other games.

wrong.

Liberal arts major or troll?

If John Carmack did it, so can I!

Anyway, here's my (maybe incomplete?) list of reasons for making a custom game engine for Speebot and all my future games:

- No fees, my budget is $0
- No risk of shitty EULAs terminating my license
- Complete control over code: including being able to debug the engine, as well as optimize for maximum performance
- Versatility: ability to add new features as required without constraints
- Freedom to use a good FOSS programming language with fast compile times
- My opinions > convention of other engines. I find that the ability to choose and create your own workflow is very important for productivity.
- I learned graphics programming and general engine architecture instead of learning a drag-and-drop interface of any specific engine, which means that even if I change tech my knowledge will still be useful.
- Game requires custom non-cookie-cutter features, such as a built-in level editor
- I can and will re-use the engine (and evolve it) for future games
- Tons of knowledge on the internet about it
- It's fun

Have you ever programmed anything user? Sometimes established frameworks don't work the way you think they should or need them to, and in the end videogames are just multimedia programs. That said, do you actually have any arguments or have you just given up on the discussion?

he took game design in college

Oh and obviously some of those points are void if you compare to a FOSS game engine.

Honestly, for me, if RPG Maker sufficed, I'd fucking use it. But there's certain things I want to do, very specifically annoying things, like the movement of the textbox, the day to night cycle and such. And since I can fairly easily (read: I'm capable of it) implement those with GM Studio, then it suffices well enough for me.

Between having a life, and being too retarded to code, learning a new programming language and making a custom engine would not only take me too long, it'd be an unnecessary waste of my time at that.

This alone is enough for a reason
I personally hate programming and only do it because i NEED it, i love creating content, but need a working base first

Pretty rich when coming from somebody who went straight for 2deep4u after "GRAND IDEAS THAT DEFY ALL ENGINES no I can't give a single example lol" didn't work.

Don't get this wrong, you seem to be JLMG, so keep doing what works for you. The main reason I argue for engines is because I see tons of agdevs for who it doesn't work, yet they stick to making an engine from scratch for their 2D platformer because engines and Gamemaker aren't trudev and everything else will be slow.
That said, some of these things aren't really reasons. For instance, you can easily reuse a premade engine and "I don't need to debug the engine" beats "I can debug the engine" by a mile.

Its simple. Vulkan is newer (and supposedly better?) than OpenGL 4.5. Therefore, it makes more sense to learn about Vulkan than it does to learn about OpenGL 4.5. Besides, any new hardware that can use OpenGL 4.5 can probably use Vulkan.

So whats wrong with my code then? I'm not seeing why you think that isnt a correct way to do this. What value clock() starts from is not important, as long as calls to clock() give increasing values and CLOCKS_PER_SEC is correct. No clue what is because I don't use C++.

However, I didn't test the code, and I do use the os-specific apis to calculate the time myself.

It is for FUN as you say, I find engine programming to be a lot of fun. Maybe one day i'll have an engine that achieves something that other engines cant, but that's still
just an excuse.

I don't think you've understood this argument, and no one has said there's anything wrong with using existing engines. It's just you arguing it's somehow wrong to make new ones.

I don't know what JLMG means, but thanks(??)
I agree that engines are in a lot of cases the way to go, especially if there's an engine that can fulfill all your requirements out of the box - especially true with 2D games these days. I still think that programming skills are important no matter what approach you take, and that FOSS frameworks > engines, but that's just me.

Slightly related - video of a comfy finn talking about raycasting engines.

That isn't raycasting. He's projecting walls onto the screen and using portals for hidden surface elimination. That's how Build renders shit.
Doom's renderer also wasn't a raycaster, but it used BSP and some kind of segment sorting instead of portals.
Wolf 3D does use true raycasting, though.

Yeah you're right.
Here's a video of an actual raycasting engine.

You're confusing the culling algorithm with the actual rendering algorithm.

Doom uses a BSP tree to Z-Sort the walls. The BSP tree doesn't actually draw the walls, rather by walking through the BSP tree you can call the functions to draw the walls in the correct order.

Hidden surface elimination is also not the rendering. It just tells the engine what surfaces to not render and which ones too render.

You'll want to check out the demoscene then and borrow tricks from them… Just saying, that's proper Carmack style.

I support your reasons, but one is missing that really seals the deal for me:
- I'm creating a canvas for gamedev which promotes mechanics other engines lack.

That reason alone is enough for me.

The in-engine level editor could possibly be that. More mechanics is good.

Also, many nodevs thing engines are nigh on impossible. They're not. Back in the day every game had its own engine. There was no engine to license, enginedev was part of gamedev. That was before hardware acceleration / dedicated 3D hardware, and games looked more unique back in the era of software rasterizers. After the advent of hardware texture and lighting everything looked the same because the hardware only gave certain features, and all "engines" programmed to that API.

Shading languages are slowly getting us back to the golden era of engines, but they're not all the way there yet. Back in the day the video RAM was main system RAM. That meant physics could update graphics immediately. There was no "readback cost". Compute shaders are getting there, but GPU can not talk to network interface, so it's still not as free as the good ol days. Back then, any particles could affect gameplay because they were all available to the gamelogic and network code. Shared Memory Architecture is coming, but it's not there yet.

Carmack was relevant in the old days, when things were simpler and less constrained by hardware APIs. Everything after Quake was meh. Even Doom3 was less that the state of the art when released. Even Carmack's (in)"famous" stencil shadow algorithm was previously patented by someone else…

Scenefags are used to seeing enginedevs "borrow" tricks and take credit for them. Just saying, don't look up to Carmack too much. He just lucked out and had enough money to come to market first.

"Good artists borrow, Great artists steal."
i.e., you think them so great because they've inflated how much was truly theirs.

History is a struggle. a contest of wills of peoples struggling against each other to grow or to die. if you want to make it of any other theme then maybe you should make a rpg.

what?

The possible side effects of enabling this are too horrifying to even consider in the current zeitgeist.
I do agree with your point however.

I tried clamping the magnitude of the constraint's length to the initial distance between objects, and then use a cable effect between both points. It didn't work too well. It probably didn't help that I was trying to attach one end to the player pawn, which kinda works off its own physics (and which tended to cause me to fall through the floor). I know it's possible, since I've seen it done, but I just wanted a simple rope constraint similar to the one in the Source engine but for use in Unreal.

I'll come back to it at a later time.

If only Carmack had known about Z-Buffers (which he admittedly didn't, even though they were used back in the 70's by Catmull & Clark), then the BSP system wouldn't have been needed.

1970's 3D graphics:
vimeo.com/16292363

Given that every pixel is looked up through Doom's lighting table anyway, one extra z-buffer compare per pixel strip, or even per pixel, wouldn't have been expensive and the time saved in reducing overdraw would have more than made up for the extra z-buffer compare. Z distance was already being computed to make further away things a bit darker. The Z-Buffer could have taken the place of the BSP tree in RAM. The reject table could have been enough to determine which sectors to draw depending on which sector you were in.

My point is that Carmack didn't know what the fuck he was doing when he made Doom, and that shit worked fine… except the BFG. Holy fuck, the dumbass screwed up the best weapon in the game, making in useful only among armatures or the extremely skilled player who knew its quirks. To the vast majority of players it was completely broken and unpredictable.

Think about that. Most infamous weapon in one of the biggest games… the game credited with popularizing the FPS, and the weapon was totally fucked up. Alpha versions of the gun had tons of visible particles, almost a rainbow of flying shit, but the optimized version used one big blast and made the hitscan damage rays invisible, and so the gamedevs didn't even realize what a huge mistake they had made not being able to see the damage trace lines.

The BFG blast goes off and the rays fire out of your body no matter where you are in relation to the blast, and the rays are pointed in the same direction as you faced when first firing it.

So, you could fire a BFG down a long corridor, then run the opposite way and into a room of enemies / players fleeing from the sight of it, and when the blast went off you could frag them all, with death coming out of your side or ass, even with no line of sight to the noisy BFG9k blast, even so far the opposite side of the level you were out of hearing range. You could also fire the BFG just as you were falling to hit the ground and the grunt noise would make the BFG silent – no loud charge / fire sound. Combined with the insane anti-logic of the BFG9000, that always got me called a hacker… I'd just quietly run into a room, then everything unexpectedly dies without warning.

gamers.org/docs/FAQ/bfgfaq/#3D

How it should have worked, IMO, was that the damage rays shot out radiating from the blast (not the player), and at the start of the detonation compute the reflection vector for any rays that struck a wall. Or at the very least, computer the damage rays using the players current look position.

Carmack is a false god.

Doom used two 1D zbuffers to clip the sprites, they knew what a z-buffer was but also knew it was expensive to have one

BSP tree takes less memory than a Z-Buffer, and is faster and also can be used for collision detection. Two birds in one stone. That's the reason they used it instead of Z-Buffering.

It's hard to make progress on game. My job is terrible and my wife openly hates me. Maybe everyone I know will die.

You don't necessarily need BSP for collision, they actually needed it to have convex sectors to draw the floor and ceilings sorted properly (of course the Build engine solved this without using a BSP)

Just, like, make game.

Overdraw just one wall back to front and you've already made z-buffer faster than BSP…

well, now that z-buffers are implemented in hardware, yes, its fast, but in software, BSP wins.

You need a new wife or none at all

You're still wrong. Not that it really matters.

Consider that in Doom you'll be rendering tens or hundreds of wall surfaces, computing a z-depth per wall column, then doing a light level lookup and texture map lookup per pixel, drawing those pixels… and then all that work gets covered up by many nearer walls – multiple times. Perf test it, the engine spends most of its time doing those per pixel lookups and then writing them to the display memory. Z-Buffer allows you to not do those lookups or writes. Slowest part of the engine is drawing the walls.

There was actually additional memory for display allocated for use in multi-buffering, but anything beyond double buffering in a software rasterizer is a waste. Some of that RAM would have been better used as a z-buffer.

You can bench it yourself if you want. I've tried it on a couple doom source ports, so I'm confident in my assertion. Problem is that many tricks used in old WADs rely on the old render method, such as invisible floors and leaky vis-planes, so there's no reason to improve the software rasterizer now. Iterating the sectors, rendering their walls, rejecting those not facing the player (dot product per wall), and z-buffer lookup per pixel is much been faster. You process less data per pixel when you reject via Z-Buffer.

The thing to note is that Doom already had a precomputed REJECT lump which is a table of area vs area, each cell being a single bit. This was needed to tell if a monster could possibly see the player from a different area. You could take any two nodes, look them up in the REJECT and tell if you could see between them.

If you already have such a table that you're iterating to cull hitscan visibility calculations between sectors for gameplay, you can be using it for rendering too. If BSP was so good why even have a REJECT table? Because BSP is shit. That's why Quake uses a shitty bastardized version of Portal engine technology – a transparent "visportal" brush is used to indicate which areas divide up the game. Could have just used a full on portal engine rather than a hybrid portal engine + BSP. Again, it would be faster not to use BSP for rendering.

Quadtree (for 2D) and Octree (3D) is good for culling of collision detection – where you have a sparse amount of active objects to determine if they're in the same general region. You can also use them for coarse view fustrum culling. However, iterating walls in z-sort order each render pass is just literally retarding to execution speed – it ENSURES the MAXIMUM overdraw. Gag me with a fork.

Time to take Carmack off your pedestal. He fucked up the BFG9000 – AND left huge swathes of unused memory which did absolutely nothing. The COLORMAP entry #33 is unused, but exists in the WAD as all black and is loaded into RAM needlessly anyway, a remnant of prior code (the lump data could have been shorter with no effect), and the color tables of PLAYPAL #1 and #9 are unused (due to a dumbass off-by one error in the code). It's like he didn't even test the item pickup render code. I would have placed test colors in the palette and iterated every palette at least once to verify they were even in use. Think about that. This "Rendering Engine God" didn't even get the palettes to work right.

First to market doesn't make you a genius, it means you got lucky, had more manpower, cut more corners, and had more funds than the other guys to bring visuals to the public faster (most of which were developed in the pirate underground of the demoscene). When that first mover advantage was lost, so too was Carmack's ability to "innovate".

It's laughable to try and say "BSP tree takes less memory". That's dependent on the number of surfaces it contains, but a z-buffer is a constant fixed size, regardless of input used. O(1) is faster than O(n). Sorry, it's true both theoretically and especially in practice. Besides, RAM wasn't that big of a deal or Doom wouldn't have been pissed away in so many screens worth of RAM.

Lower your expectations all around. Make the smallest game possible. Then build on that. Wifes love money. If game = shopping for her, she'll love your gamedev. Prostitutes are probably cheaper though.

I cant actually find a comparison of the speed of both, I suppose you are correct in the case of doom if you say you've tested it, although I think BSP in a normal setting should be faster since you don't normally check the z-depth of each column when you're using BSP trees.

Really I couldn't talk about doom's rendering engine and why it does what it does since I haven't even read the code.

Why don't you post your tests? Its a curious topic.

I think carmack is a good programmer, just because he did get results, if the doom engine is so badly designed, it must have been possible for someone smarter to make a much better engine in the same time… like Ken Silverman's BUILD, except he was a little late (I think because he's younger / started later than carmack?).

I have an array of colliders that is compiled automatically on start.

How do I make an OnTriggerEnter script that works with every collider in that list automatically?
I feel like I should know this but I'm drawing a blank

...

Well it's an array of Gameobjects off of which I get their colliders, same thing

Also I'm using unity, keep forgetting to mention it

Training faeries now work
It permanently increases the lowest stat, plus a random one if the job panel is working at 100% efficiency
(School -> Wis/Cng/Lck, Gym -> End,Dex,Agl)
Anyone knows how to make the WebMCam show popups? the faerie selection popup was not recorded for some reason

is this a porn game?

No, settlement management minigame

but, yes, there will be porn

What're you making this in, and how are you releasing it?

Where's the video game related progress?
Worst thread in awhile

Indeed.
Make what you think is the most fun, and hopefully in the process of doing so you'll make video games great again.

soft body physics works for ropes rather well, or a simple solution is to use just joints with constraints.

ppl like you give that engine a bad name

Probably not, there will be marriage and stuff since the main game is a farmsim (not with the faeries though, they are pure)

Pure C# because its easy as fuck and fast to make, will port to Unity later, i will release as an .exe and put a download link here when its properly playable

When that time comes, I'm going to decompile it and finally get access to your infamous cooking code

Its ok, i will rewrite and improve it on the Unity version

Why don't you just make it in unity in the first place?

Would take much longer
I used to try building everything from the ground up or do the hardest parts of my projects first, the result is over 12 abandoned projects, i found the secret to keep myself motivated: do stuff FAST get results FASTER
i have a super autistic project i was working on years ago where i was going to make a fucking periodic table of fictional atoms and simulate physics and chemistry based on atomic bonds
I will also reuse also the complex magical lore i made before with 26 Elements and shit

is cgpeers down for anyone else?

So like, if you have a Form, how do you like … draw shit to it? Like if I have a texture, and want to draw it, but also draw things on the texture. I guess like Paint?

I've managed to it it by directly setting the byte data of the object, but that feels really wrong, and like, something like OpenGL "does it" and not wrong

downforeveryoneorjustme.com/cgpeers.com

downforeveryoneorjustme.com/http://www.downforeveryoneorjustme.com/

Im using WPF to draw the UI elements
does it still counts as pure C#?

What the fuck does pure C# even mean?

I consider it "C# without using an engine or other languages"
Not that there is anything special about it, just describing what im using right now

Falling for that trap I see. If it was straight up better in every conceivable aspect they would discontinue OpenGL, which they're not

I just realized that instead of having a timer per object, I could just have a timestamp per object and on update, compare it with a global timer; that saves an assignment or better

Thinking about starting up a side project that I could push out in a couple of months time while I finish up assets for my current thing, my NEET funds are running low.

Does anybody here know of any good SHMUP resource or program for creating bullet patterns? Setting up look-up tables for projectiles by hand would be a god damn nightmare.

I remember seeing some sort of Bullet Markup Language project, I don't know where it is on the web off hand but it should be quick to find. Not sure if it will be useful to you, and I'd be interested to hear what you settle on.

Vulkan has barely begun to enter commercial usage, so they're not gonna discontinue OpenGL anytime soon.

I imagine there's a lot of legacy soft/hardware that depend on the OpenGL paradigm.

I'm currently using Gamemaker so I don't know how useful this would be for me. It is XML based which means I could write a parser for it, but all my experiences with XML have been a pain.

just write your own algorithms dumb nigger
if you know basic trigonometry its not really that hard (just time consuming), and if you do find it hard then you probably don't know enough about programming to be attempting a complex SHMUP game

start with a simple function like BasicCircleSpray( numbullets ) that fires bullets outwards in a simple circular pattern
then extend that into AdvancedCircleSpray( numbullets, normal, spread ) that will fire an evenly spaced cluster of bullets along the normal angle using spread to fan them out

after that you could mess with creating spawn origins as well as giving the bullets special properties inside their own internal logic thread
eg. bullets that travel in a sine wave along their trajectory, or bullets that partially track the player or bullets that gradually accelerate/slow over their lifetime

once you've got enough neat algorithms made you can start combining them all together to get all kinds of neat effects

How do you generate a point that is guaranteed to fall within a polygon? I don't like generating a random number then using a while loop and testing to see if it falls inside it or not.

Asking for eventual particle/surface generator

easiest way would be writing an algorithm to place a point somewhere inside a rotated triangle, then decomposing said polygon into its component tris
then you'd loop them, pick a tri at random and generate the point somewhere inside it
you could probably extend that by weighing the tris by their surface area so there's an equal probability of the point appearing anywhere inside the polygon as a whole (instead of it being more likely to appear in areas with lots of clustered, small tris)

Turn the polygon into triangles, then you can do this.
mathworld.wolfram.com/TrianglePointPicking.html

Oh okay, so basically I grab a random triangle from the shape, treat it as a quadrilateral, and then generate a value [0..1] like normal, but wrap it to fit the triangle? Neat

Does anyone else find the wording on this confusing as fuck

...

What genre is this?

Puzzle

Do you have to explain it? How important is it that the player know about clock ticks or statistics to that degree? I'd cut it down to just a remark on how documentation seconds don't necessarily reflect wall time, and that the ratio may differ further if there are problems detecting the monitor.
If you really want to add the details of the game, have some link at the end take you to the Autist Edition where you have everything laid out in excruciating detail.

I'm okay with certain kinds of pixel-based games not doing this, particularly if they're period pieces of sorts. If it's anything like Boulderdash, It might make some collision responses fucky if there's a spike of lag at the wrong time, require separating the graphical representation from the simulation representation and produce that small unevenness in movement if the velocity and framerate don't really click. The visual thing's only really jarring if it's a static or linearly-moving camera at low resolutions.
But it's going to look bad and/or play bad on monitors with refresh rates that aren't an integer factor of whatever the standard happens to be, so that's a consideration to make.

Well, it certainly has that 90s "we're getting into making programs for Windows but have very little experience with computers and literally have no idea what we're doing because there's little precedent on how things are supposed to be done, but we're still going to do our best to make it presentable" aesthetic.

...

You don't discontinue things simply because of legacy support. Really I don't know what Vulkan and OpenGL 4.5 can do, just that Vulkan is supposed to be better in some way (performance? I dont know)

Are there any other decent private torrent sites?

Sorry if I wasn't clear but, that's not my game; it's a game from 2003 that I want to clone, which itself draws inspiration form Supaplex, Bouldadash, and Chip's Challenge

Also, despite the graphics, the guy that programmed it is pretty fucking competent. I've never had it crash, and I just recently popped open a saved level from the editor and it has checksums and hashed data so it couldn't be easily fiddled with outside of the game.

That's not competence, that's just being a shithead

I mean, look, I'm giving you the option in my game, but why the fuck would anyone want to, anyway?

It's a single player game, right? Who cares if someone makes their own levels? Only a perfect mix between jew and autist would.

The game is free, though. And it does have a robust level editor. You just can't do it outside the game itself.

then why does it matter if they edit the levels themselves?

I don't know? Again we're not talking about my game, but the one I posted above

no shit

Vulcan is designed with current gen GPUs in mind, and GPUs changed quite a bit in the past few years. Pre-compiled shaders, multithreading, and it's a lot lower level in general.

Yeah, i know that much, I Just mean that I have yet to actually use features that OpenGL 1.4 doesn't have, so how much Vulkan benefits me isn't very quantifiable until I start using these new features.

From what I've seen the multithreading aspect is a gamechanger (designed for the multi core paradigm), both for utilizing/balancing additional cores for dispatching work to the GPU, and utilizing compute/cuda cores more efficiently; so in essence better utilization of the GPU, and cores.
Soon they're integrating OpenCL with Vulkan, which will be huge too, and will make it graphics + general purpose API for utilization of the GPU.

If you want to see tangible results, unity supports OpenGL, and Vulkan; you can choose which to use for the graphics API (thus changing calls n such to equivalent), and with the heavier demo scenes n such there's a tangible performance boost in FPS.

I told you, I know about this already.

I've got no interest in unity, i'm writing my own engine.

I'm planning to add a fn-fal to my game but I can't decide which one to go with here.

should I go with the class black and wood, all wood stocks, or the israeli version?

Of course the Israeli version. What are you sum dum goy?

How do I instance nodes in godot? I only know how to instance scenes. I just want to create some buttons through code.

I know u knew part of it already, but I was telling u what I've seen as far as actual games go.
Also, how to see for yourself, and if it would be worth it or not; in a game setting.
Also, get off your high horse.

...

I like the first one.

You can programmatically make new nodes like this:
godot.readthedocs.io/en/stable/learning/step_by_step/scripting_continued.html#creating-nodes
You can't instance nodes, though, you need to use scenes. There is a plugin that lets you extend nodes with scripts, but I haven't tried it before: godotengine.org/asset-library/asset/7
You can install it on the AssetLib tab, then make sure you enable it in the Project Settings.

I really dont know what you're trying to say at this point

I just made my first game, it's pretty shit, check it out scirra.com/arcade/action-games/first-game-17907

all I can say is that this should not have momentum

it feels fun

the momentum is fine it just needs to build up faster

This is basically how I'm handling my objects moving around … can you guys see any potential problems I might have later, in particular with the Position, Map, and State?

hastebin.com/azubodetol.cs

Basically the Map will be loaded and put into State, which contains a copy of the map and all the objects and stuff. Entities are mostly plain data, with the map providing utility functions

public bool IsMoveTimerElapsed
{
get
{
return (MoveTimerMax != 0)
? (MoveTimer == MoveTimerMax)
: false;
}
}

The fuck is going on there

glad I pirated this

It checks to see if the timer has elapsed.

If max time is 0, the timer is not active so it cannot have elapsed. Otherwise it checks if the current time is equal to max time (which is true if elapsed). Otherwise it returns false

dunno, it just seemed way over complicated for what it was

return (MoveTimerMax != 0 && MoveTimer==MoveTimeraMax)

SOMEONE HALP

What kind of game is it? RTS? TBS? 4X?
What does the player's turn look like?
How do you get resources?
How do you spend them? Why? Which resources are there? Which resources are the most valuable? How do you win?

Stupidly ambitious combination between 4x and Action RPG. The idea is that you play through a dynasty of characters sort of like CK2 or Rogue Legacy. I know how I want character progression to work, I know how I want dungeons to work. The RPG part is at least vaguely figured out.

The only thing I've got set in stone for the empire management part, though, is that I want technological improvements to primarily come from dungeon delving. There is no "tech tree", you have to go to a dungeon, find a technology, then you spend effort to research it.

Player's turn is spent either running around the capital city for stuff, or looking at an empire map giving orders to armies/workers out in the field. I'm thinking of straight up ripping off Civ V's "each tile has X amount of resources on it and you gain them each turn" system. What I don't know is what those resources are, or what they're used for.

I know for a fact that I want Gold and Mana, but do I also want Lumber or Stone? Maybe Iron? Maybe "greater" resources used in just about everything and "lesser" resources only used in specific things?

Civ V had a few types of resources; you had regular tiles that generated food, research, production, etc., but then you had luxury resources (copper, gold, gems, crabs) that generated happiness, as well as strategic resources which gave things like iron and horses for your units to use, and then finally bonus resources like deer or bison or fish that gave a boost to the basic resources above

Thanks to the user who reccomended that I try The Guild 2, while it was fun it wasn't quite what I was thinking of making. Think more Recettear but you slowly buy up more businesses.
Honestly, it's probably best I just stick to alchemy only, that was my biggest idea which I have the most plans for.

I know for a fact I don't want the luxury resource system from Civ V, that was pretty bland.
Also, I've realized that I need character stats that do more that just combat. Fuck. Guess I'm going to end up breaking from tradition and separating INT from MAG.

Found a fuckton of glitches, fixed all and added
Incredible all the trouble a single misplaced line of code can make
Damage/Death animation added, also minor stuff like hospital and nursery passively improving stats of newborn faeries

Will they have idle animations too? Even something as simple as eyes, blinking, mouth changing, and having ears droop as HP drops could add a ton of juice

What do you anons think of game maker for a person new to creating games, no coding experience at all, experienced in asset creation and animation? It looks promising to me because of the dedicated community that seems very helpful.

not idle (will probably do it in the final unity version)
but they have five states/faces
Very happy (happiness = 1, health = 1)
Content (happiness < 1, health < 1)
Sad (happiness < sadnesslimit)
Tired (health < weaknesslimit)
Sick (sick = true)

It's serviceable enough if you've never programmed, but eventually you'll outgrow it

Will using game maker ingrain bad coding habits ill have to shake later?

Go with the classic.

yes, but at least you'll have a game

I think I made a mistake in my calculations

I think you made a god

5 hours of progress
bug fixes
damage/death animations
faerie names are now unique, faerie is rerolled if a clone appear
minor details (new faerie stat bonus, zero amenities penality)

Add some fast loud music and it is perfect

healing and mana potions are in, and come in three different tiers each. Next step is probably going to be status effects since I'll need those for additional potions and for the spell system.

just call it a bullet hell and everything will be fine

Darken that red a bit, my eyes bleed when i try to read it

Were you trying to imitate Alpha Core from Omega Boost?

this multiplayer tutorial is fucking garbage

nice Deus Ex: Human Revolution Typhoon implant clone

Even i can do better than that

It runs that every single time something even touches the player, even if it's a wall or the ground. It catches an exception each time, every frame.

the best part is this garbage costs $240/year to access

where did you find this gem?

ITS SUBSCRIPTION BASED?

CGpeers, and I think they got it from lynda

Such a huge waste of time. Red flags throughout, and I still don't understand anything about networking.

Holy christ
Unwinding the call stack is the only expensive part of try/catch exception handling, by the way

now I remember why I removed colored names in the first place.

wew

forgot pic

The Unity 4 tuts I tried on there were all terrible, but that's when I first started Unity a few years ago, so not sure on their Unity 5 tuts.
Honestly, the tuts on youtube are better, or try the udemy ones.
In the end the quality of the lessons really depends on the teacher. Again, the best unity teachers I found were on youtube, f.e. the official unity live stream tutorials are sometimes brilliant.
The only tuts worth their salt on lynda are for things like blender, photoshop, or pure programming topics (as these are more traditional, "professional" teachers, so quality shoots way up).
Luckily I get my sub for free

Can't you just change your font to have an outline?

outlines look like shit on pixel, use shadows instead.

infernal mace looks alright because of the background

these would probably fix it
or you could get fancy and make it so your text darkens when the background is brighter

Anyone in particular you liked on youtube?

Definitely the official unity livestreams, Sebastian Lague has some great stuff (unity, and blender), and also Jonathan Weinberger is a fantastic teacher.

Thanks. You know if the udemy course is any good, by any chance?

sounds like a lot of trouble for something fairly minor. I think I'll revert to white text for now.

the problem with that pic was more with the steel dagger, which is barely visible

Just copy and paste the text in a 3x3 area on your spritesheet or whatever. Or overdraw it 3x3 times.

of course. Unfortunately, I'm not sure on the udemy front, but I've seen some courses consistently reviewed very high for unity related stuff (I would find those courses, pirate it, see if they're a good teacher w/a 10minute watch, etc).

I'm still working on drafting this concept I've had for awhile now that's essentially an "MMO API"

Here's how it would work;
1.) There is a client that connects to the server
2.) The client itself hosts the graphics locally
3.) The server hosts the maps collision data, and tracks players progress
4.) The client connects to the server through a standard set of APIs. Incompatible APIs are not permitted to connect

The idea is that you can have an MMO essentially run on anything as long as it can connect to a network. It can be 2D or 3D (it would have to do up to the client developer to figure out how to represent height in a 2D space, however since its an MMO its not expected to have that high a degree of verticality anyways). And all you would need to port to the target platform is the asset data itself and the API for the client to use. I can host a reference client to aid developers in creating their own assets, for example. And the API could be portable, meaning you can have consoles, smartphones, and old shitbox PCs all connect to the same servers. This would in theory help greatly with the games adoption rate.

If objects are processed from left to right, and don't overwrite eachother's behavior, how can the rock on the right end up with priority? I don't think it uses a stack. Trying to emulate this behavior

wait, is that what's happening or what you want to happen?

I can only assume that the rock on the right would end up with priority only if it was specifically programmed to be that way (eg. before allowing a rock to roll down a right leaning slope, add a check to see if there's a rock 2 spaces to the right and a left leaning slope 2 spaces right and one space down. If true, don't let the rock fall)

It's the expected behavior that I am trying to emulate. I'm just trying to figure out how to do so. Both rocks want to occupy the same space, and intuitively, it should be the left rock, but it ends up being the right rock.

There might be a priority to "roll left" versus "roll right" under the hood… I'll see if I can contact the original creator and get more information on how he did it.

Also this feels incredibly fragile, but it works so whatever

I think it's just executing the checks for all rocks in sequential order, i.e. check left first for all rocks, do actions if possible, next, check right for all rocks, etc.
The result in your picture then makes sense.
The design I'd assume is something like a list, or array of rock elements; looping over each w/the left check, and then next loop do the right check.
Inefficient, but it's a simple solution to reproducing the behavior you want.

That can't be it because there are also balloon objects that go up, and the same tile conflict can happen, just in a different direction. There's too many object combinations for that to be feasible

humm, well the info you posted here is all I had to go by.
I'm not sure as to exactly what the conditions of the conflicts are, but it sounds like they're just layering the same simple logic; however, they're processing the set of elements + sequential checks in a specific order to produce said behavior.

Just, like, make game, user. I'm using it. It's simple, it suffices, so why not?

...

That looks cool in the thumbnail, actually.

This is a good feel.

streamable.com/ge4jv
Making Kowloon and devving whilst no longer a NEET sucks.

I feel you, Kownloonanon. It's hard enough just working enough to survive, let alone to come home to a second job that you're not getting paid for. Knowing that you fags are out there somewhere is one of the few things that keeps me going.

Nice, you're making progress. If you're trying to mimic the fast paced gameplay of old school FPS, consider removing the recoil.

recoil looks too harsh on those guns

Maybe in user's game you're playing as a tiny loli?

Knowing you dysfunctional fucks can make a game helps propel me to put in work myself. AGDG also works by shame because it feels bad to start posting progress and never follow up.
Good luck, user.

To be honest I am not entirely sure what I'm doing only that it's fun so far. Thanks for the advice though.>>12592732

It's dramatized for the sake of feeling powerful, loli does not sound bad though.

But the loli has to have smugly adorable voice acting. Like Duke Nukem, but innocent.

Well, I got an answer

This is why I hate programmers

Just give the enemy a shrine maiden's outfit and it'll be perfect.

T-that looks fun.

okay i'll include the classic and call it the fn-fal and i'll include the israeli model and call it the (((fn-fal)))

I gotta say though, the m16 was a lot easier to draw since it used such muted colors and smooth angles. This is actually proving to be a challenge.

What, why?

Their mannerisms. Saying shit like 'Precisely. So this shows that your assumption is incorrect." That, on top of their le xd nerd culture.

There's too many shithead autists already, programmers just have an absurdly high concentration.

Reminder to keep deving.

You can do it.

Alright, this is starting to make more sense now. The rocks actually have different priorities depending on what happens with the middle block.

This also works with any other object moving out of the hole, such as the player or removing dirt/gems, etc; this would give the left object priority.

As for my game, I just need to figure out how to best store/sort objects by position, since there's about 2000 objects an 65000 tiles at most. I mean I could hold them in an array, but I don't know how to sort them intelligently as it goes

How exactly do you want to sort them, and why? It sounds a bit like you want a priority queue but I'm not entirely sure.

There's nothing intrinsically wrong with that sentence.
Le xd nerd culture isn't a natural phenomenon, nor is it inherent to, caused by, or ubiquitous among programmers. Marketing people did that.
You're definitely right that there are more than normal, but I think you have a skewed view of programmers as a whole because of a few autists with god complexes.

What.

mods deleted some other shit thread

Tiles are static and just kind of sit there. Anything with any sort of behavior is considered an Object/Entity. Every tick, each object updates itself and perhaps does things like falling, destroying other objects, etc. Having them in a position-based order means I can have a predictable update cycle, and also predict behavior from screenshots/paused gameplay.

The thing is that I need to update every object about 30 times per second, and if I max out at 2000 objects, I don't want to iterate all the tiles to build the list of objects, sort them, then apply their Tick() method.

Priority queues sound good on paper, but most implementations use a lazy sort method which means that I only get sorted objects on pop/dequeue, which means I have to add all the objects and then remove them every time. Some sort of array feels the best approach here, since I can just iterate from [n..0], but I don't know how to guarantee they'd be sorted (insert methods would be slow, if an object moved in a certain way, it'd have to push a large chunk of items around)

"I'm gonna kick ass and chew bubblegum, and I'm all out of ass." - Loli, 2672
Probably no playable loli, possibly adorable NPC with a short quest, likely not.

The priority queues I've worked with in C# let you enumerate through them without having to duplicate it.

...

The base C# language does not have a priority queue implementation, though, does it? I had to make my own based on a min heap.

Though if I recall, the objects were not guaranteed to be sorted until they were removed

I was too lazy to make my own, and probably couldn't do it as well as these guys anyway.

github.com/BlueRaja/High-Speed-Priority-Queue-for-C-Sharp

Oh shit, that's a good pattern

Also none of those seem to be IEnumerable, so they don't seem to support foreach / iteration

Whoops. You might just need to modify that yourself, then.

Any of their implementations use a CascadeUp/CascadeDown method, which means it still uses lazy sorting; even if I did implement it myself, I could only guarantee that one branch of the object tree is sorted, not the whole thing.

NEW BREAD?

We're still on page 12.

I think I have the explanation for this. College programming.

Most kids going to them don't give a shit about programming and just want a degree, which causes them to struggle through it. The people who do it for autistic fun are head and shoulders above their peers who think they're progressing because they took a class on fucking cobalt, where as the autists actually did research and practiced. 4 years of this causes the autists to think they're a fucking genius even though they'll most likely end up in the same place as their peers.

5 hours of progress on a weekday, good stuff
Faerie Job buildings now deteriorate, but you can repair them

I hate how tutorials finish with "Now, you would never do this in a real game, but these were chosen specifically to show the concept for…"

Nobody ever covers how you should do it in a production environment, so then you have a bunch of jackasses who don't know how to do something right because every tutorial ends there.

also

Isn't the purpose of unity to have a low barrier of entry?

I don't even know what their plan is. A few years back they tried making their own vocaloid for advertising.

Yeah but it associates the engine with poor performance and bad games. I remember seeing people cite Unity as one of the problems with Yuuka-Laylee

I realize now that I didn't understand your post… you were criticizing unity for marketing itself in such a way, and not wondering why Unity has so many bad games made with it.

We are one thread from page 13, shall we do it?

don't be a faggot, respect the rules

Lit girl is a miracle of the universe

The new thread would be the one thread

Its time!

NEW BREAD

The guy is incompetent.