/agdg/ + /vm/ ~ Amateur Gamedev General

To be honest mate, I'm fresh out of coffee and haven't slept since yesterday.

What I'd do is I'd pretend to be one of those deaf-mutes. And by that I mean I'd re write the code without any matrices, just get the threading side of things nailed down. Then I'd adapt the properly debugged threaded code to the problem at hand (and keep a copy around in an example file in case you have to do that again later).

Your ThreadedMatrix is calling the Determinant function with false (no recursion, right?). But you can't set finished = true and also check (!d.finished) without synchronizing.

See, that's why SkyNet used the time loop so the T-1000 can lock up Sara Connor to prevent the mutation of John!

repostan from last bread since someone was helping me
if i limit it to one thread, it works but slower than before
if i do more than 1 thread, i get an IllegalMonitorStateException
i made a new class that stores the reference to the matrix, calculates it's result, and when all the instances of this class are finished they notify their parent to calculate their result
the parent in this case is a matrix, if it can make a new thread it makes a new object of the previous class, otherwise it directly calculates the result of the submatrix in that thread
after it's told all the threads to start calculating it waits until they've notified it. after that it sums up their results
so the question is why do i get an error when i'm using threads
package project;public class ThreadedMatrix implements Runnable { public Matrix matrix, parent; public boolean finished = false; public double result; public int index,x; public ThreadedMatrix(Matrix _m, Matrix _p, int _i,int _x) { matrix = _m; parent = _p; index = _i; x=_x; } @Override public void run() { // TODO Auto-generated method stub result = matrix.Determinant(false); finished=true; for (int i = 0; i < parent.threads.size(); i++) { ThreadedMatrix d = (ThreadedMatrix) parent.threads.get(i); if (!d.finished) { return; } } parent.notify(); }}
package project;import java.util.Vector;public class Matrix { public static int threadMax = 1; public static int threadCount = 1; public double[] elements; int size; public boolean separateThread = false; public void SetElement(double val, int x, int y) { elements[size * y + x] = val; } public Matrix(int _size) { size = _size; elements = new double[size * size]; } public double Get(int x, int y) { return elements[x + y * size]; } public Vector threads = new Vector(); public double Determinant(boolean quiet) { if (size == 1) {// shouldn't really happen unless someone specifically // enters a 1 sized matrix return elements[0]; } else if (size == 2) { if (separateThread) { threadCount--; } return elements[0] * elements[3] - elements[1] * elements[2]; } else { double sum = 0; for (int i = 0; i < size; i++) { Matrix sub = new Matrix(size - 1); for (int x = 0; x < size; x++) { if (x != i) { for (int y = 1; y < size; y++) { sub.SetElement(Get(x, y), (x > i ? x - 1 : x), y - 1); } } } if (threadCount < threadMax) { Thread t = new Thread(new ThreadedMatrix(sub, this, threads.size(), i)); threads.add(t); sub.separateThread = true; threadCount++; t.start(); } else sum += Math.pow((-1), i) * sub.Determinant(quiet) * Get(i, 0); } if (threadCount > 1) { try { Thread.currentThread().wait(); } catch (InterruptedException e) { // TODO Auto-generated catch block for (int i = 0; i < threads.size(); i++) { ThreadedMatrix d = (ThreadedMatrix) threads.get(i); sum += Math.pow((-1), d.x) * d.result * Get(d.x, 0); } // e.printStackTrace(); } } return sum; } }}

I meant to reply to:

…but it wasn't posted yet. Fucking time loops, how do they werk!?

Never knew Avatar had so much in the lore, was there any Sound bender in the story? I remember metal, blood and lightning only

oh nevermind then
ok so i need to lock something until something happens, but with your explanation i got completely lost
what do i lock and how do i lock it

That's definitely one thing a lot of anime/manga do well; it's generally in the background though compared to say a game.
Same with beserk, great lore for this type of stuff too.

Not sure on the sound one.
Though it's apparent the graph on the left side is highly inspired by naruto (one thing that show does well is the lore/world building imo).

There was also combustion (mercenary hunting the group), lava (again a mercenary, and in the shitty korra series had a protag character with lava bending), and spirit is pretty much like their aether.

There's several ways to do it.

See "Volatile" keyword (I might have imagined this from another dimension), Synchronized objects, a Mutex.

The pattern is called producer / consumer, and there's a pitfall called Dead Leaks? Dead Locks, something. Whatever you do remember to Reap your children lest they become Zombies… but that may be something from another Unixverse.

Also, I forgot to mention: If not modifying the Matrix, you can share the main Matrix between all the Determinators and just zone their actions to ever decreasing ranges.

If you get dejavu that means another thread had a concurrency issue – or they changed something.

Who the hell came up with this? First levels are a mix of neighboring 4 primary elements. Does that mean that next levels of mixed elements are also combinations of next level? Water + Earth = Plant so Metal + Ice = Gasoline?

Korra does not exist user, the story ended after Aang defeated the firelord.

There wasn't any. They probably didn't have any ideas for that.

alright nevermind, apparently thread.join can just wait for the thread to finish rather than pausing the main thread
so i'm avoiding errors this way
if (threadCount > 1) { try { for (Thread thread : threads2) { thread.join(); } } catch (InterruptedException e) { } for (int i = 0; i < threads.size(); i++) { ThreadedMatrix d = (ThreadedMatrix) threads.get(i); sum += Math.pow((-1), d.x) * d.result * Get(d.x, 0); } } return sum;
but the issue is still there
1 thread runs fine
multiple threads run at least 10 times slower rather than faster

Just take a bit of transparency, add a bit of noise, and voila.

It's shit. But it's a start.

I'll make a webm if I can remember how to feed video frames into ffmpeg via its standard input (and how to remake a demo record / playback feature).

Smoke simulator 2017?
It looks very neat

The problem with sound bending is that it doesn't work for a single element. In normal situations it'd be airbending as the usual medium for soundwaves is air, but the moment you go into the water or put your ear against a wall you'd be able to hear at least some sound through that. So the only person who could really effectively soundbend would be the avatar, other people would only be able to do part of it.

pls rate.

horrible uv mapping on the shoulders/waist

the model or the animation?

sure, but GIMP and blender are trash lol.

both

There's overhead for thread creation.

You have to make a kernel call at least, and perhaps map pages in to create a stack. Context switching is low. How many threads are you creating? How many cores do you run?

Probably won't see any improvement if your code spends all its time copying the same unchanging matrix around, creating more threads than you have cores and then waiting for CPU to synch results.

If it's something you'll do a lot of then you'll want to have a pool of worker threads already setup, and a manager of that pool which breaks down the big problem into small chunks, doles out the work loads, collects results, and hands out the next part of the workload to the free threads until the task is done.

Having more threads than cores running means the CPU might use hyper-threading to execute a second thread while the other is waiting for synch. That's faster than an explicit context switch at the application level.

Starting way more threads than you have cores isn't going to be helpful. The sum total of threads in the application could exceed the cores by a significant factor because you might have multiple worker pools, only some working on a specific task at a time.

Try creating 4 to 16 threads only. Give them each a significant batch of work to do before they have to be joined. Best case scenario: Start those threads up at application begin. Have a synchronized function which adds a matrix to the workload and waits for its result. If you want to do more than one thing while the matrix is being processed then create another thread.

Think about how the thread works. You have a call stack – a section of memory allocated to make function calls. Function calls are fast because the memory is already allocated and when you call / return you're just using the next lower batch of data, shuffling a pointer to the top of stack back and forth.

When you create a new thread you don't just tell the CPU I want to run two sets of code (you could but you need a stack to call functions in your language's paradigm). So, that means you create a whole new stack – big enough to call all the nested functions you'll need to call in that Thread.

If you want to just create a bazillion little threads constantly then you need to go use Erlang. The procedural+OOP paradigm is to create a big stack and treat threads like mini processes.

The colors on the model are too vibrant, also i need a 360º to judge the form
And the animation is not looping properly
or maybe its a shitty gif

i've got 2 cores and i'm trying from 1-20 threads, anything more than 1 being slower
guess i'll try and use indexes rather than referencing new matrices every single time, that's probably what slows it down so much.

same colors.

blame GIMP

It doesn't make sense to apply the color limitations of a 2d console to a 3D model. You should use the colors that best represent the character's design rather than exact colors of the sprite.

Color palette is cancer, mesh at shoulders is clipping, animation isn't fluently repeating since you can clearly see end and start

I'm remaking a nes game in 3D.

I dont have a clue over what i'm doing lol.

I was re-uploading the vertex buffers of each map chunk every frame because I forgot to turn off the update flag

...

NEAT

ded thred, is everyone actually developing? you faggots better not be procrastinating

less posts is actually a sign of more work here

i'm trying to figure out the correct usage of the GL_ARB_vertex_buffer_object extension since my VBO's aren't displaying…

do anyone here wants to make controversial games about serious topics?

once i get this platformer done i'm planning on making an arpg where a race of soul-sucking lizard merchants exploit sympathy over a genocide they faked in order to enact their plans of destroying the world

I'm a level designer employed in a reasonable large company, and the amount and specifics of the work I have to do leaves me with no motivation or desire to improve my skills or start any projects by myself. Wat do?

get gud.
exercise, eat better, fap less, go to church, watch motivational nigga videos screaming.

check

pls no

I'm procrastinating as we speak, UE4 open in a separate window.

mind == body == soul

get gud

...

...

In mine humans devastate two races, then create and enslave other two alchemical humunculus that can self-replicate, the massive guilt trips causes the human empire to degenerate and fall
All of this will be hidden in books and tales all optional to the blissful player who just want to plant potatoes and get a monstergirl waifu

...

Get savings
Drop work
Bet everything on your game
Get motivated or starve
Thats what im doing
10 days left before im finally free

...

Not happening, at least not anytime soon.

Thanks for believing, though. I believe in you too.

what

...

What game do you have in mind, user?

Touched up the animation a bit.

Once you've done a jumping attack and fallen for long enough (2-3x jump height) it turns into a plunging attack. You don't bounce back up, but it deals more damage.

yes, i am developing, but code isn't really postable.

wew laddy

...

How often should I fap?

I guess I re-did the menus, not much else since this is my first project. Other than that progress has been slow. I improved the ai a little bit, but most of my time was spent trying to fix bugs and crashes since they were starting to pile up.

An ultra-comfy farm sim with monstergirl waifus
I gonna outcomfy everything out there while having deep mechanics and rich worldbuilding

I don't know how much of a pixel artist you are, but I feel you should really make more parts of the sprite move when falling, at least the shield arm.

Also, judging by the resolution, the shockwave effect can be made even bigger easily, or it won't be noticeable.

Continue to design levels since your job won't last forever.

I actually liked the old UI, maybe i got used to it, even though it can get hard to read depending on the text color, keep the old one as an option, or just allow for custom themes in general

No shit, it goes in two year cycles more or less.

this game is my first attempt at animations ever and i have no idea how to make the shield arm move in way that wouldn't look retarded. The secondary motion on the end of the tunic is just to make it look slightly nicer than zelda 2's and ducktail's etc one frame of plunging attack

go for it user, I post my code all the time
Its interesting to see what other devs are working on, since art and levels only happen in the last 50% of solo-enginedev

Also, there's approximately no animation on the thing I was using as a reference. Oscar (or as this debug character is actually named, 'ignite') only has a very slight wobble as he rotates around his center of mass.

It was ok, my favorite verses were in psalms or Thessalonians; it's really about what speaks to you.
Overall, I prefer the gnostic scriptures, or the vedha is better in terms of it being older (you can see some of the inspiration in christian texts, as it all mingles together); in addition to it being more sound (fewer plot holes compared to the NT).
Although, my definitive favorite texts concerning this stuff are the buddhistic texts (odd names I can't remember nor care to, but they're mostly concerning abandoning materialistic needs, and the middle path).

also
Gotta get that legit world building user

How's everyone's wage slavery going? I grow tired of living

So cute, trying to hold his tunic down, but embarrassment turns to into a deadly attack.

Your stabber is the stabber that will pierce the heathens!

It might not be visible in those pics, but the old ui was tv static, not very fitting for a game in a medieval setting. The code is still here, only commented out so I suppose it wouldn't be too hard to add an option to switch between both styles. I'll consider it, but I can't promise anything.

Well sprites should be exaggerated compared to realistic 3d, especially in terms of movement.

It's doing fine, user. When you consider it a means to an end, it gets better.

the fuck
they're supposed to be orange

I totally would, but it's not confined to one unified class so it'd be a disconjointed mess to follow (as I'm mostly following the entity-component-system design so logic is highly decoupled).
Working on LoD, and culling atm though.

They're green actually. So many colorblind people here.

Only more 10 days for me

THEY ARE RED YOU PLEB

Pull a notch: Hang out with indies and steal one of our ideas when we give up on it, even though having done a bunch of R&D on core mechanics. It's the execution that counts anyway, right?

Also, if you can't start a project yourself – Join a small project. Indies typically have everyone doing a bit of everything. There's always folks on tugsauce looking for teammates. Find a partially complete project that you could ask if they need some help. There are some in the FLOSS community where you don't have to do a whole lot.

The journey of a thousand miles begins with a single step. The 1st step is a doosy.

He died. That's what you get for demotivating.

Good writefag Sunday. Nice work.

His secrets were divided so that no one book contains them all.

As an embriologist that third picture makes me mad

...

I'm procrastinating testing Blender's video editor
anime is flip flap

Good for you. I'm applying for a pretty good company right now. Got my test task a week ago, and work has been slow. I'm anxious they'll overlook me just because of how long it took.

Patrician taste

Good luck, i hope they don't do to you what was done to me, assigned to code monkey job on languages im not familiar with on the worst failed projects that get abandoned by other devs leaving the company.
At least i learned C# while busting my ass there

lets try again, maybe this time it will work on browsers

Nah, I'm in for level design and so far the propositions look pretty comfy, that's IF I actually get there.

...

Too much of academia's indoctrination will make you easily triggered and close minded. Lay off the sauce, there are some areas of knowledge scientists and historians are forbidden to touch in academia. You're only mad because you think this is the 1st iteration of human civilization.

None of the courses of your over priced uni education will explain where these giant ancient abandoned canals and machine craved ports came from – They're ten feet under water off the coast of North America, and not dredged in modern times. Hunter gatherers didn't build them. If you want to find out who did you have to stop getting mad when things don't match your forever flawed worldview.

Shh, resist the urge to reply. Be quiet, like this video of remnants that lost advanced civilizations left on the bottom of the ocean floor all over the world. Some start on land and continue hundreds of miles into the ocean.
vimeo.com/album/2045605/video/47697010

In Nippon many gamedevs have been redpilled on the truth of our past for a long time. They resist and evade the censorship of Western imposition by GCHQ's War Guilt Information Program propaganda campaign by putting hidden knowledge in media. That's why so many games and anime reference the occult (just means hidden knowledge), and have "lost ancient tech" as a theme… The better to woke you with, my dear.

Sage for IRL loreposting.

Are you a meme?

Don't bother screeching, I'm a
and done

God damn, and i was just going to bed

looks like crap.
lol

but is fine.

yes goy the earth is flat xdddddd
mankind are descended from ancient aliens cos the pyramids say so xddddddd
if u like and subscribe my video u can turn into a star through the power of magic xddddd
the reptilians are watching us xddddddd
toxic vapor trails xdddddd

It's just victors rewriting history to make themselves look good. No need to poison the well with nonsense.

Both of you are imbeciles.
Go make games instead of whatever this autism is.

I don't remember River City Ransom being a cripple rehab simulator, user.

Just came in this thread to let you know you're a huge faggot. Probably a normalfag. Out.

just post the function you're working on, im sure its interesting

I don't actually use classes and when I see OO code I always think "Is that guy seriously using global variables for no reason?" until I realize that its C++ or something. Global variables are at least declared with extern (mostly) so I can spot them out.

Admittedly I have been doing this…
I was making it print out "cool information" to the console on startup so that you can see what kind of things its doing. At least i implemented line wrapping to con_printf…

If you're making an avant garde horror game, then yes,

It's my first animation using IK.

Just consider it R&D and lorecrafting based on alternative history. It's how games are made.

You wouldn't make a game about the past without doing some research. For example, who really carved entire cities out of stone like those Star-Shaped Fortresses in Malta?

Behold,
HOW THE FUCK DO I USE BLENDER: THE POST: THE GAME: THE MOVIE: THE POST

Only as an occasional ideaguy thought, yeah. But I'm not exactly the type of person that could handle controversial publicity in any matter close to 'stable' or 'sustainable'. Am too emotional for that shit.

Though there is a nice concept I've been rattling around where a prophesy is laid onto a protagonist, but the protagonist either doesn't give a damn or actively combats said prophesy. Enough in-story time will progress where the world itself aggressively morphs just for an inch of fate to progress, to where the protagonist either gives up and follows through, or blows back against the world changing even harder.
I gotta be honest though, this idea would likely work way better as a small novel.
If I this isn't controversial enough, then make the prophesy something akin to demanding a sacrificial martyr. There may be sect leaders demanding you to be sacrificed to save their kids from the world (a world changing because of "chosen fate" not being fulfilled)
I gauruntee that some site will rile people up in an "not angry, i'm just loudy voicing our reality against your fantasy". Eh, I'm just rambling at this point. What the fuck am I trying to say?

...

I know right? I've been trying to do SOMETHING in 3dsmax several times, and every time left after being intimidated by the fuckin heavyweight interface.

I wish creating assets were as easy as creating wads in doombuilder, or tiltbrush in VR by grabbing vertices, faces, and lines and sculpting.

I managed to load dev textures into Thief's map editor today. Was surprisingly easier than I expected.

I really like making maps in games with dev textures since it has this clean and sterile look and makes a more boxy map look really sexy.

Can't you do the same in 3dmax? You pretty much can in zbrush and substance, no?

I'm not familiar. I only messed with a VR setup at a friend's house and it seems ideal for 3D modelling if it were just a little beefier with what it could create.

Well I don't know, user, you could do pretty much the same with a mouse input and several viewports.

implemented a temp fix to my generation order for what I did the LoD system I'm testing (for cells).
It's a temporary bandaid to properly utilize the pooling class i wrote up.
also notes everywhere so ppl know what shit does

it is to me

...

The only one of those that looks like a stump is the third one.

thats because they are massive, and the 3rd one is a little tree if you calculated its height it would be 60 miles tall. most of its "bark" have eroded off onto its base.

these tree's were made out of silicon rather than Carbon. the earth's crust is 97% silicon.

This is hard to follow because of your OO scope… I cant understand the intricacies of this system since everything is in the class scope instead of the function scope. But it looks like you're doing something pretty cool.

Is this a new meme?

28% silicon.

played a little with sound levels
since it is not gamedev this is probably my last post about video editing

...

Its earth's ancient history. it used to look a lot different before giants cut out tree's and mined out resources.

Earth's oxygen supply is currently depleting and no one give s a hsit.
Oxygen used to be at a much higher amount 7-8000 years ago compared to today. higher oxygen supply allows you to run faster, longer and heal faster.

when people are in dire need they are put into a oxygen chamber with increased levels of oxygen, this allows your body to heal your body naturally at a faster rate.

the world's oceans were not there, the world was flooded in order to drive out these giants.

Nice. An often neglected feature.

World of Goo (among others) had a nifty "console log" system. The debug strings were replaced with game-related nonsensical strings in release mode - but devs could still translate them if needed.

I stopped myself earlier today from building a sparse quad-tree system to back the text display for scrolling when line wrapping is disabled. What the fuck was I even thinking? Text rendering is not a 2Dplatformer.

I am probably going to base a vector font on (or simply make a bitmap font out of) classic CP437 OEM font that most PC's have enabled at boot.

It would be nice to have a multi-line text input box for short in-engine scripting. I just know I'm going to make this sooner than later if only for debug / built-in fallback default font, so why fight it even if it's not exactly gameplay progress?

Shit, I just found a bug in valgrind. "unhandled instruction bytes". It's because I built everything with -march=native. It's legitimate chipset opcode (I checked the manual), just that valgrind doesn't recognize (yet). Probably hasn't encountered it due to few people building everything from sources and also doing gamedev. Guess I'll have to switch to a Linux distro with binary releases for now.

That's actually not a bad idea. I've been thinking about an "easy" mode for level editing, so why not make it top-down doom/build engine style? Can't really model like that, but it would sure beat "prefabs" shit some games offer in-engine.

DO NOT TEMPT THE ENGINE DEV. This is a trap. I recognize it. Since I agree toolsets are too complex and there's a niche for a gap, Soon I'll be making a simplistic 3D modeler in the engine AND STILL NOT HAVE A GAME.

Seriously though, just learn to use a 3D tool. Even in Blender you can just subdivide a cube or sphere then "push" or "pull" with a sculpt tool almost like clay – Even has a simple mode to MS paint style on the model / texture.

There are programs out there like Google SketchUp and Wings3D – though the former will teach bad habits, you can at least make something in an easy interface.

Normalfags: Pay no attention to the Moai in your shmup. There is no hidden meaning. You are not mastering the flesh and being reborn from a space vagina.

Carbon 666 Silicon14 777

user, are you in your right mind right now? You're not projecting an image of a sane person very much.

You know, the masm32 forums actually have an entire subforum dedicated to AV and similar software giving false positives about their programs because of these kinds of things…
masm32.com/board/index.php?PHPSESSID=ebdc28b03fb67438feb6dfc786d562b6&board=23.0

Sometimes, the best solution is the easiest one… i'm still at a 30-line con_printf and I hope that it doesn't get much bigger. (It will need a third amendment when I have to implement scrolling up through the console… and also logging the console output into a file? Features that will need to be finished someday…)

I have made one level editor and I plan to continue by making a second once I reach this point in development. I think I will just gut the shit out of my engine,so im just left with the openGL-windows context, and the console, and then use it as a base for a map editor. So basically what I have right now when SIGMA2_VULKAN_ENABLED and SIGMA2_LINUX are undefined…

Anybody have resources on implementing algorithmic stuff like z-layer sorting, AI, ect.?
I try to write this shit myself but it comes out a mess, and the answers from SO are trash

I have successfully been able to use the GL_ARB_vertex_buffer_object extension to draw VBO's correctly. It turns out my code was correct but I set the per-vertex alpha to zero and not 1 which made it transparent so it looked like I did it wrong. I wont post any screens since its the same damn spinning cube as the immediate mode opengl and vulkan renderers.

Z layer sorting can be done with a BSP tree. You could also look up an actual tutorial on writing a Z buffer.

cs.utah.edu/~jsnider/SeniorProj/BSP1/default.html

Are you going to make movement in red sky non-janky and fix the diagonal clipping glitches?

I "finished" that project, right now I am trying to move past it and do new things… so the answer is no. I know that's not the best answer, but, trying to untangle Sigma 1 is just a nightmare and while I can do it if I spend the time, there is a point where you are tired of dredging up code from over a year ago, code from when I was not very competent, and want to just put it behind you. I spent months and months polishing it and trying to make the movement something that people were OK with, but its flaws run too deep. If you want to see something truly awful, just go into the code, its available to download on the itch page…

I'm hoping that my next engine feels a lot better, and i'm writing it in such a way where I think I can make something that people don't complain about endlessly.

>>>/x/
>>>/fringe/

Um what

higher oxygen levels just cause things like it becomes easier for things to catch on fire and for it to spread. It also causes health problems. If you breathe air with a much higher than normal O2 concentration, the oxygen in the lungs overwhelms the blood's ability to carry it away. The result is that free oxygen binds to the surface proteins of the lungs, interferes with the operation of the central nervous system and also attacks the retina.

The Apollo 1 caught on fire entirely because they used a pure oxygen atmosphere inside and it spread extremely fast. They used pure oxygen and nobody anticipated that it would cause problems like that.

Oh, you.
pick one.

Does pics related gameplay look like the product of an ordinary mind to you?

Kek. Shitty, but made me chuckle.

Best is to just pick an open source project and look at how they implemented what you want to do. Sometimes they'll be using libs that are free to use.

My favorite z-order rendering technique is to take an axis aligned octree and collide it with a model of the view Frustum. Find the node collides with the camera / near clip plane, render nodes from it front to back, walking into adjacent nodes away from the camera (or towards the camera from the far clip pane for depth first render). Just stay within the view volume, and render as you go. If you transform the vectors (1,1,1) & (-1,-1,-1) by the camera rotation matrix, then you can examine those results and determine which pattern to walk nodes to go from back to front (or front to back) in relation to the camera. Basically, everything in an octree is z-sorted (and x-sorted, and y-sorted) already, you just need to know how to walk it.

I prefer sparse octree for large scale collision detections and muh voxels anyway, and there's some tricks to vastly improve octree search over root or bottom up… Since the placement of the separating planes is implicit rather than explicit it makes many operations faster than a binary tree. A BSP tree can be made to represent an octree by using axis aligned planes to divide space in a repeating fashion. Octrees could be considered technically a special case of BSP with node structure optimized for certain forms of searches.

When it comes to enemy / NPC AI, there are so many ways to do it you should just pick a game with AI you like and see if there are any GDC talks on it (or the games it was inspired by) – or, again, check released source code.

My code for it is going to confuse the hell out of people too, since i'm using legacy GL with the extensions for "modern" features, so it looks like a program written in the 90's, with wglGetProcAdress and "ARB" suffixed function calls and all.

Thanks for the babbling, why don't you go make an /x/ board and stay there.

Do we have any infographs on how to get into gamedev for utter newfags? I see the question asked a lot and it would be nice to have. I actually started on one once but realized I don't know enough about the available engines outside the three I have used.

ie: Learning basics of programming, relevant maths, picking an engine, outlines on how to program typical game stuff like collisions, input, state machines, script effect objects, etc. Good programming practices.

When I think back to how I got into this field, I can trace everything back to trying to make a mod for a game I liked. I recommend trying to make something for an existing game and learning the process for that. The skills and know-how transfer to game dev proper, and the mod route is more defined than the more open-ended realm of game dev.

its better to start programming raw, no engine, make a simple 2D UI based game, make your own assets, learn to animate them
it doesn't even need to be a functional game, just search for interesting stuff like making a ball that runs away from the mouse, don't try to jump on a serious project without any knowledge.
I suggest starting with C#, works on Unity and its REALLY easy to learn

Adding onto this, try to optimize your game to be fast.

You'll learn a ton by making your own engine and optimizing it, because you can't do that if you program like a bargain bin pajeet.

One by one find things that are slow and then figure out how to make it less so. You'll learn to structure things the proper way and that knowledge will translate to everything you program.

...

...

post it fagget

Nah, I'm kidding, user. Glad that you're feeling motivated.
Honestly, looking at my old work just makes me hate my past self for being such an autistic faggot.

ca3old.tumblr.com/
remember to use the tags per year.

I will use next year to start a proper training in realistic digital painting.
Need to be serious now.

Autistic faggotry can be recycled user, most of the stuff i got on my elemental system dates from over 6 years ago when i tried making a TCG in java, which was inspired by an old RPG maker project i started and dropped 10 years ago
Also my cooking system is a super simplified version of the physics/chemistry system i was trying to create with a periodic table and subatomic physics with magic and shit i might revisit some of that when i make the alchemy system for the game

Current stuff is better than expected
I set the bar very low when clicking a tumblr link

not enough, but I feel motivation to start practicing realism.

Go for it user, don't waste motivation, it is a rare commodity

Mate, no fucking bully please I was like 11
It was a Mario/FF crossover I thought up because of Culex in Mario RPG. It starred Ludwig Von Koopa, Geno and Rydia, and the plot was just random autism for no fucking reason. Something about the seven Star Spirits.
That, and I stole all of the assets. Didn't matter that the art styles and music styles were clashing, I just wanted to, like, make game.
That, and it was all based on the Hello engine on Game Maker, which is shit.

I'm reusing SOME things from my last attempt at a project, a few decent songs that I reworked a bit to be less repetitive, and some attempts at a plot thread.
But when I say autism, I do mean autism.

Pic related is pretty fucking good, mate.

thanks.

remaking old projects is a good way to see the progress and feel motivated.

I need some help calling a function from another script attached to the same GameObject in unity.

I don't know why but nothing I try works, even though I've done it multiple times before.

First Script:
public CharacterControllerUtility utility; void Start () { utility = GetComponent (); utility.ClampToGround(); }

Second Script
public class CharacterControllerUtility : MonoBehaviour { public void ClampToGround() { Vector3 up = Vector3.up; transform.position -= up * 5; }}

Shouldn't this work?
Sorry if this is fucking retarded but I don't know what to do.

are you sure it compiled? if you're not getting any errors it looks like it should work

That picture is doing things to my brain.

It compiles and then when I try starting the game it tells me pic related

that means that utility is null
either they're not on the same gameObject, or you didn't put a CharacterControllerUtility on at all

Oh, the script just disappeared from the editor
I'm sorry I thought it wasn't finding it from the code, I didn't even think the editor didn't have it

So if both are attached as components to the same Entity (GameObject) you should be able to do the "go.GetComponent();" from the first script.
The second script will only be called once, as you have that in the MonoBehaviour Start() function which is only called upon starting the component (only EVER once for this instance).
You want to call the "ClampToGround()" in the first script, within a function called every frame (like update, etc), or when you need it (self set bool: !Grounded).

much better.
(not really, lol).

better than the animations I made tbh

Change the textures into rotting flesh and you got a Silent Hill monster design for you right there.

oh shit that's a good idea

One user was making a crash course in gamedev, teaching just the bare minimum to get started along the way. I don't know what became of it, but I've often thought of making something similar. I'd have to agree with pic related that just starting out it doesn't matter what platform or engine you choose. Those things you can decide better after you have experience. I always stop myself making these kinds of tutorial things because, IMO, it won't really help newfags. They'll still just ask for silver bullet solutions and go pick up an engine / framework without knowing WTF they're doing. They don't want to learn how to make games they want to make a specific idea come true. They have to learn along the way themselves.

There's so many online tutorials for intro programming I feel like if one were just focused on gamedev rather than teaching every little thing about a specific language it would solve the college "game design course" crap.

Some game courses aren't shit, but most are, and they're a huge waste of money. I once met a guy who had taken a gamedev course. He showed me an incredibly detailed design document, and stats tables, etc. He didn't even have Tetris or a game mod under his belt after that whole damned course. If only they had taught him a little bit of everything instead of focusing on "design" and character creation he could have just like made games.

I tried to start making a game with him since he seemed motivated – but he wasn't really motivated. When I said we had to make something less ambitious the first time, so either trim down his design (he REEE'd at the suggestion), or make a new smaller one (which he agreed to do but lost interest). I told him this was part of making that big dream come true, but if he couldn't get back to me in less than 3 days when all I needed was him to run the new build and comment on mechanics he designed, then the project was trashed.

I've been burned so many times trying to help absolute beginner nodevs that I'll let other people waste their time trying to teach them.

Thinking back, that's not how I learned anything, so why would I expect it to work for others? When I started making shitty little games all I had was a PC, and a few tips on ASM and C from some local (pirate) BBS. I mowed a hundred lawns to buy Borland's C compiler. Parents even confiscated my hardware because I was doing gamedev / programming rather than homewerk. I moved my stuff to a friend's garage to make games after school and started a software business distributing my wares on Compuserve.

Point being: Nothing could prevent me from making games, and all my teachers and relatives actively tried. Nowadays we have free OS's and compilers and shit loads of documentation about everything. There's free source code of successful games out there to look at. If someone isn't making games today, IMO, it's because they just don't want to.

When absolute newfags (not the ones that are actually making a game) ask about how to get into gamedev I usually just ignore it because it becomes engine wars and language wars. The only correct response is: Nigger, Just Like Make Game. You can search on any problem you run into along the way.

Do you somehow not see the blinking shoulder problem? If you do, why are you making animations for something broken?

I'll fix later the texture.

I was also going to make a crash course tutorial of my own using javascript since you can run it so easily without installing anything.

For example, you can save this as "game.html" to get a game, and then edit the code for easy results.
body{margin:0}window.onload = function() { //listen to keypresses, and link them to keyDown/keyUp functions document.body.addEventListener('keydown', keyDown, false); document.body.addEventListener('keyup', keyUp, false); //container for keypresses key = {}; //create&setup a canvas and add it to the page can = {}; can.html = document.body.appendChild(document.createElement("canvas")); can.ctx = can.html.getContext("2d"); can.html.width = 600; can.html.height = 600; //set up player and enemy player = { x:400, y:470, w:20, h:30, hp:2, reload:400, spd:4, bulletspd:4, bulletsize:15, color:"#0ff", vel:0, dead:false, shottime:0 }; enemy = { x:200, y:100, w:20, h:30, hp:10, reload:350, spd:2, bulletspd:4, bulletsize:15, color:"#ff0", vel:0, dead:false, shottime:0, wander:2.5 }; //text that's shown on top left gamestate = "Game on!"; //container for all bullets bullets = []; //start game update();}function collision(a, b) { //returns 'false' if a and b are NOT colliding, or a originated from b if (b.dead || (a.origin === b) || ((a.y + a.h) = (b.y + b.h)) || ((a.x + a.w) = (b.x + b.w))) { return false; } else{ return true; }}function drawMe(target) { //draws character and their HP can.ctx.fillStyle = target.color; can.ctx.fillRect(target.x, target.y, target.w, target.h); can.ctx.fillText(target.hp, target.x+target.w+5, target.y);}function update() { //reset and clear canvas can.ctx.font = "14px Arial"; can.ctx.fillStyle = "#111"; can.ctx.fillRect(0, 0, 600, 600); //use javascript default function to get current time (in milliseconds (look up "Unix time")) var time = Date.now(); //handle player if (!player.dead) { //movement if (key.A) {player.vel = -player.spd;} else if (key.D) {player.vel = player.spd;} else {player.vel = 0;} player.x = Math.min(570, Math.max(10, player.x+player.vel)); //shooting if (key.space && time > player.shottime+player.reload) { bullets.push({ x:player.x+player.w/2-player.bulletsize/2, y:player.y, w:player.bulletsize, h:player.bulletsize, vel:-player.bulletspd, origin:player }); player.shottime = time; } //draw drawMe(player); } //handle enemy if (!enemy.dead) { //movement enemy.vel = Math.min(enemy.wander, Math.max(-enemy.wander, enemy.vel+(Math.random()-0.5))); enemy.x += Math.min(1, Math.max(-1, enemy.vel))*enemy.spd; if (enemy.x < 10) {enemy.vel = 2;} if (enemy.x > 570) {enemy.vel = -2;} //shooting if (time > enemy.shottime+enemy.reload) { bullets.push({ x:enemy.x+enemy.w/enemy.bulletsize/2, y:enemy.y+enemy.h, w:enemy.bulletsize, h:enemy.bulletsize, vel:enemy.bulletspd, origin:enemy }); enemy.shottime = time; } //draw drawMe(enemy); } //handle bullets //loop through all bullets for (var b=0; b

...

user that's not how you punch at all
go watch hajime no ippo, everything is explained there

Can you post these with the a floor visible?

Holy shit mate.
My fucking point from still stands, this is seriously Silent Hill-tier.

kek, I'll try to make a new animation.

OR
Make it a terror game

Still working on my game, though giant bugs keep slowing me down.

Thanks to the user in a past thread who pointed out in the pseudo-C64 minigame that Cruller could basically stay in the air infinitely by resetting her jump! I agree that it's more fun that way, and actually added that into the main game. (Specials and air dodges now reset her double jump, and attacking successfully keeps her suspended in midair.)

Here's the latest test of the main game engine - curious how well it runs on everyone's machines (and more importantly, if it's fun to play). There are almost no enemies in the level at this point, it's basically a physics test. But hopefully it's fun to move around and explore.

Controls: Arrow keys move, Z jump, X attack
Space to pause/switch characters
Caramel can fly by pressing Jump when in midair. Hold down the button to fly faster.
All three characters have various extra moves you can do by holding in a direction while pressing jump or attack, or by doing Street Fighter style controller motions.

Extra debug features:
Kill yourself by pressing C
0, 1, 2, 3, 4, 5, 6, 7 - quick character switch (also lets you bring in Player 2, though for this test their controls are mapped to be the same as P1)
Press SHIFT to skip ahead to the next room in the dungeon array.
(I don't recommend doing this, as it may spawn you in a stupid place.)

Apologies, it's still very incomplete, but I hope it's fun to play around with. Any comments/complaints/suggestions/bug reports/calling me a faggot welcomed as always.

This.
Revive PT, user. Go cash in on the VR craze, you'll make quadrillions.

lol.

I agree that they are unsettling in a surreal way, but it would seem like a good idea for him to learn how to do proper animations especially if you want to create a contrast between the monsters and any non-monster/mundane elements.

Can you make it so it shows how many FPS the game is running?

The slowdown that occurs when you hit the floor feels very disappointing. If you can, consider either removing it or making it happen only when you fall from very high.

Also these blocks do not look like you can walk on them, they look more like wall decorations.

I know it's very unfinished but this feels so completely directionless that I can't be bothered to test any more. It seems I can just run and pick random paths along the way and find more of nothing. That said I can see it being a nice game.

yeah, 3ds max can be a real bitch to wrap your head around at first but once you do its an amazing tool for game modelling
just persist and try to dig up good tutorials that aren't some fuckwit pajeet mumbling incomprehensibly into a 2 dollar microphone
mfw i finally got my head around skinning mesh to a skeleton

yeah, use the edit poly modifier

redone.

oops.

MoM user here again, with the Light Tome Spells this time!

Feel free to ask me any questions!

Not-Shantae is CUTE

I have autism now

HAHA i finally figured a design for the complete 26 Element Lotus
6 commons
12 composites (mix 2 commons)
8 celestials (mix 3 composites)
will make the celestial element icons and place them all in order when im back home

That looks pretty funny.

Has anyone tried the Xenko Framework?

Looks neat, by the nips so will work better than Unity, in C# so equivalent to babby's first game engine and integrates with Visual Studio because just fuck my registry up, fam.

But I digress. It looks good. Anyone used it?

Good idea. I'll see what I can do.

I actually have a debug window that I removed from this demo (as it's ugly and the information it displays is pretty much useless to anyone who isn't me), but FPS wasn't actually one of the things it tracked. I should fix that.

Good point about the blocks, thanks.
I hope that the areas will be more interesting once there are enemies in them, but I see what you mean about the directionlessness. The dungeon's actually fairly straightforward with just a couple of branches, but I'm familiar with that kind of very demotivating "what now?" feeling. Any suggestions on how to make it better?

Thanks. Being compared to Shantae in any way is very motivating.

What game are you remaking, again? It seems familiar.

Will your game follow a plot or is it completely focused on game play?

The game has a plot which I think I mentioned here before; They is currently a crisis in the world of magic involving a bunch of monsters made of stone conquering the lands, so the King of the Realm of Magic has requested the aid of heroes to come and resolve this conquest, with the people embarking on the quest being the 6 playable main characters.

Have some shitty oc.
Fucking gimp ate my layers

lovely.
kek

I think the main problem was that the map never seemed to end. It's very frustrating when the map just keeps going and going in these pathways and it's in the back of your mind that you left another path un-checked.

Usually in games it's fairly obvious what the "proper" way to go is and what's not as important, but it's a bit hard to describe what the exact difference is. I think usually in games you run into some kind of jumping puzzle or a tricky spot or important looking object that makes you want to check the previous path before you continue on the current one. Or maybe the sidepath just doesn't look as important (e.g. random hole in a wall as opposed to a proper pathway).

It almost looks like your map has multiple paths to get into the same destination, which doesn't suit this kind of game IMO because multiple paths make you want to explore all of them. So this kind of level design makes you go in circles looking for the remaining corridors you haven't explored, and the buffer of "paths you've left unchecked" in your mind grows very uncomfortable the longer you keep going.

The map should either be more open (i.e. paths connect to each other in many locations and you can build a mental map of the world), or more closed so that each branched path either ends or loops back to the main path in a similar location as it started. You also shouldn't split the map right at the start like that unless one path is clearly blocked, it leaves you with that uneasy feeling very early.

teleports behind you
unsheathes thumb
thumbs up at you
Good stuff kiddo!

I obtained gold or something, then accidentally pressed shift and wound up invisible forever.

z and x are a bit weird on my keyboard. Custom keys when? Doesn't have to be fancy with the names of keys displayed, just a button / menu option that take you through a series of "Press the button to use for: Jump", "… for: Attack", etc. record those key values. Could bind it to something like "Press [F1] to define keys". Not that it matters, just most indie games are hardcoded to things that are hard/weird on different language layouts.

I agree with:

Things have to stand out from the background in games. Try to practice having more contrast between background graphics and the interactive elements. Even secrets in most games should stand out more than those blocks. pics related.

Otherwise, Keep up the good work.

Yep. Things like that are perfect.

Just add audio & music and all the essentials are covered.

Some examples like that should go in a pastebin that we could link to from a resources thread/sticky on >>>/agdg/
Maybe even stick it on JS fiddle so don't even need to save files to disk. Then we can just point newdevfags to the resource thread (where they'd check in the first place if they read the OP).

The user I was replying to has a valid point though about selecting a toolset and whatnot. I'm sure we can find an infographic comparison, but it should be aimed at yesdevs who've at least made a simple game already and will understand the infographic. If they don't know the difference between entity component model and traditional OOP then it won't matter if we list those engine bullet points.

In some ways I think the decline of Flash is to blame for much of the newfaggotry. It was a turn-key solution, but nothing really fills the void left by it.

——

I'm only 1/2 procrastinating, designing the engine's debugging interface. Deciding whether to make in-engine console + logfile or network based command / logging first. Most pro engines in debug build expose a port/socket for logging/cmd io – That lets you run full screen while debugging on another machine so as not to influence render performance too much with multiple screens and windows (and also because console games are debugged this way). I'm not shooting for a pro tier engine, but it sure is a handy feature. On *nix I can just pipe STDIN / STDOUT or run it via SSH, but I can't expect users to SSH / telnet into the game just to kick some shithead from the server, and I'll need to provide a graphical interface for noobs rather than text only.

Depends on whether I want to start nailing down netcode at this stage, and whether to use raw TCP/IP or a special client to debug with that can into UDP netcode protocol (and itself would then be pipe-able). This engine is somewhat more network focused than most, but the full design isn't fleshed out yet, so if I go off and make a network attached console command interface I may end up wasting a bunch of time having to scrap and remake it. So that means I'd need to make the netcode protocols first.

animations are done, time to import them into unity.

heh

This is what the engine outputs on Linux. It turns out that programming it so VBO's are not required was already helping…

Thanks, that's very helpful advice.
You're absolutely correct about the map design: it IS designed around having multiple paths that lead to the same destination, because the three characters have different skill sets and you might want to choose a different route depending on who you were playing as. (And yes, possibly double back to get something you missed.)

Hadn't considered that frustrating aspect about the lack of a "proper" way to go. It's something that simply didn't occur to me because I KNOW that the paths all loop around to the same place. But you're absolutely right, the actual effect on a player who DOESN'T know the whole map is to become quickly overwhelmed and "lost".

I should probably make the "branching" rooms lead to clear dead ends initially. The all-branches-lead-forward design might be better used in very lategame dungeons where an overwhelming "this place is too big" feeling is more appropriate.. Also, I definitely need to implement some kind of mapping system.

Thanks for the wise advice.

Custom keys are absolutely on the to-do list. Sorry for making the default control set an annoying one. (I assume you're a Bernd?)

I'll be sure to make them configurable in-game in the next build.

Does someone need art?

what kind? my artist is in vacation for a month, i need realistic style food icons or anime style 3D models and background art

2D drawings.
What kind of BG?

comfy as fuck fields/sunsets, like a big painting
if you want to negotiate send me an email
[email protected]/* */

What does your art look like?

I am not looking for an artist, but I always am curious at seeing peoples art!

Like this.
I haven't got any realistically colored ones in this battlestation.

Nice. I think that layout will be easier to understand than the prior design (which had paths going through others). The ancients would commend such dedication to geometric autism.

Since the offset of the circles is constant you can use it as a fractal. Perhaps make expansions of it, then do geometric analysis and see what higher order shapes fall out, then use various patterns as art which references the lotus if you want or maybe make more lore based on those derived shapes.

Using the offset of one of the center two elements onto could produce some interesting results in a fractal expansion.

sage non-progress posts

On the scale of bad to worse how are these looking?

I like the style of pic 2, reminds me of some of cactus' games.
adultswim.com/games/web/hot-throttle

Looks really slick

This is terrible, that one thing moves totally unrealistically.
t. /loomis/

neat

really good, but i think the jump needs acceleration, right now it moves up and down in constant speed, looks weird

First one needs a touch more foot work,
and you could throw that other hand up in the air for counter balance or just "flair".

Descent animations so far. Could use a bit more slashy moves in that stabby attack, i.e, little swishes not just stabs.

also, can you throw your arm at opponents after obtaining merely a flesh-wound?

Smooth but with little force behind it, and somewhat awkward poses. Look more into how the human body moves as a whole when doing something and give your stuff a better sense of gravity and weight.

If he primary slashes with his right hand why does he stand with his left side forward?

It still will in cases like
Ash Thunder
Instead i could make individual lines for each, like
Fire -> Ash
Fire -> Metal
Fire -> Ray
Fire -> Thunder
But will probably become a fucking clusterfuck
I will still try it, but right now im at work (did that on lunchtime)

That was a fun watch.

Well, maybe going through them is OK if there's some kind of rhyme or reason to it, or the display indicates multiple paths.

Perhaps slightly thicker lines in the graphic, with the main color of both elements? Just throwing ideas out there. I like the prior design too, didn't take me long to figure out how to read it, so it's not that big of a deal.

God damn that's really nice.

LOL

I used gamemaker a lot back in the day. It wasn't just incredibly slow, but all the 'good' coding practices people posted everywhere were godawful.

One example being the only way people would move up slopes in 2D games; They'd use pixel-perfect collision (Built in, used by default) and check the pixel to the right of the player. And then right and slightly up. And right and slightly more up. Essentially, doing this once for every 'pixel' high they should be able to climb on a slope.

pic related was going to be their logo before massive fan backlash.

They used to have an anti-piracy measure that would destroy that assets of people who pirated it. This was before GM offered source control, and even then no one used Github. The best part was that this triggered for legitimate users. Pic related.

I EBIN TROLLED the forum on and off a few years back. The CEO was a devout christ-fag who ended up selling GMS to a gambling company. One of the admins who was heavily invested in the forum had a mental breakdown because they left their account logged in at a library computer, and then next person to sit down wrecked havoc. He blamed the community for abusing her trust and basically called everyone heartless assholes, even though odds are almost 100% that the person who did fucked with him had never even heard of gamemaker.

Speaking of the forums, there's a married, 50 year old lefty who all the children look up to. He's been grooming these kids for a decade.

I have a lot more stories about GM but I probably already sound like a huge fag

i gonna try a few designs later, i really want this to be fucking pretty because this is the "magic circle" of all elemental magic, so it will pop up a lot in future games (which will be more combat oriented)

Godot will save us from the tyranny of nonfree game engines
r-right?

I'd rather bend over and let unity fuck me rather than use godot

Thanks Anons.

Got any good references or tips in adding force to moments I'm kinda' a fucking idiot and has every right to bully me

Because I'm a cunt

Why's that?

Foss usually equals to shit

...

Blender and Krita are fine for their respective fields.

Watch HEMA videos and/or do HEMA.

And Blender is pretty fucking bad compared to other modeling and animation programs

...

also, the the most in-depth tutorials are from kids with shitty mics and 200 views. I am living in the worst timeline.

Give us some reasons why it's bad in comparison, faggot.
If you pull out the "muh industry standard" shit, you're a confirmed retard

The interface is a complete mess and most of the most useful functions are either hidden away in submenus while they should have a more central position.

It's full of single lines of text that are supposed to describe what each function does, when an icon would occupy much less space and be more descriptive.

The actual program itself isn't bad, but it's clear that it was created by a bunch of programmers that knew nothing about how to make a functional interface.

That's my two cents, feel free to disagree.

They STILL haven't implemented working wheels yet, and point people to asset store solutions.
I rememeber that += sometimes didn't work in one version too. Good times.

Oh yeah, back to gamemaker, my friend was just begging to be sued. They even sent him multiple formal takedown requests. Essentially, he wrote something to perfectly convert any Gamemaker executable into the original source code. The interpreted executable preserved everything, including function and variable names.

That part was great. I used it on every fucking game just to learn how the popular games were developed.

Their entire multiplayer service is total cancer and moneyjewing scheme. I think i will settle for Steamworks matchmaking and have players host their own servers like in many shooters.

take that back

Well, all that's left is ladder climbing, and then I'm out of excuses.

So, /agdg/, since I barely know how to do any high-level coding, what engine do I develop my platformer with a huge continuous world in? Gamemaker?

How does unity's matchmaking service fee even work? Do they do both the matchmaking service and host the games? Or do players still have to act as the host and all the data passes through Unity servers for no reason?

...

That doesn't really answer my question.

You mean no transitions or loading screens?

I don't know how that would work in gamemaker because AFAIK it has a set room size, but I've seen Unity games that dynamically load the map.

yeah, go for it

If you want to put in a bit more effort, LOVE2D is a pretty good 2d framework.
If you want something higher level with built in editors and such, unity is probably your best bet.

I suppose that depends on how you plan to implement it. Game Maker uses a "room" system, which might be a limitation in this respect. If you plan to structure it like a Metroidvania, it should be easy, and excellently suited to it. If you want it completely seamless, with no transitions, you might want to consider using something else.

Current progress…

Refining the way my combat system code executes. Battles are turn based on a grid. Each action a character can take in combat is in its own script.

Previously, each character had a list of actions they knew how to perform. Each entry on the list was an instance of the action's script and the script stored all relevant variables. I'm in the process of changing it so the actions are all stored in a single database and the characters' list of known actions contains a class with the ID of the action and the relevant variables there.

The entire thing is a pain in the ass, but I'm hoping it should optimize the code a bit… If nothing else, it seems like a bit more of a "professional" way to handle actions.

Does anyone know how I'd use my own relay server in Unity? Or replace it entirely? No one is going to pay $10,000 a month to do what costs actually costs $100 to host yourself

Kek.
Anyone who falls for this deserves it.

...

It's still kikish, but you really should already know what this basic acronym means.

Welp, officially run out of excuses.

Fuck it, then, I guess I'll do CV1-style transitions between areas. No loading screens would have been ideal but there's no way in hell I could reasonably get something done in LOVE2d.

Concurrent is one word not two

Also, crunching the numbers, that bandwidth usage is insane. Of course, AWS has significantly better rates at 0.09$ a gb out vs 0.49$ a gb total.

Illustration purposes user.

WAIT SHIT NO I HAVE ONE MORE EXCUSE

WTF DO I CALL THE GAME

I was going to go with "The Legend of the Dark Soul" but that's a mouthful and a half.

Just use that as the project name and carry on.

Might it be possible to trick Game Maker into allowing an infinite size area by looping the background within a room?

I've done it before in Flash, which only allows a max object size of 2880x2880. I created a much larger level by dividing it into segments (pic related). Basically, the background consisted of two objects, and loaded small segments of the level into them. Whenever one of the objects moved far enough offscreen, it was moved to the other side, and the next part of the level was loaded into it.

The result was a room that seemed to be HUGE, but from the engine's perspective was always less than 2880 pixels wide.

Just a thought. I'm not sure it's possible to duplicate the process with Game Maker, but someone more knowledgable about the engine might find a way.

Godot. Use scenes to divide the level into chunks and load them as needed.

Type A or Type B?
The change is in the outer the petals

We NES now.

A, the straight lines from the middle-ring to the outer-most ring lacks the noise the same, curved lines have in image B.

A - Looks more like you want to focus on the centre, like cross hairs. Looks more scientific and rigid.

B - Looks like you want to focus on it as a whole and flows naturally entirely. Looks more fantasy and fancy.

I guess it depends on what you're going for.

How huge are we talking?

I've seen seamless scrolling levels with 65536 x 65536 tiles in Javascript. That's big enough to fit an entire game into one level.

Any decent 2D engine should have such capability these days. That's how I would make my toolkit selection, were I in your shoes. Point being: You shouldn't have to compromise on that feature if you want it.

That said, never used a 2D engine I didn't make, so I'm not sure which to go with. Try a few out and see. That's your new job. Shopping for a toolkit and trying on engines.

B
for reasons…

Here's a slightly improved version of my level from earlier, with some of the suggestions implemented (FPS counter, custom keys via a clickable button in the top-left).

Thanks, that was really interesting.
Nice to know I was following in the footsteps of giants, even accidentally. Part of the fun of programming is figuring out how to break the system's limits.

user…

...

You can use black people, but they aren't fucking caramel. They're burnt wood,

i REcOgNiZe YOu
hIIIIIIIIIIIIIIIIIIIiiiiiiiiiiiiiiiiii DEvFaG!!!!!!!!!2111!!@[email protected]/* */!1!!!!!11111111

sO ThIS iS wHaT yOU'rE wOrKiNG oN NOwaDAySSSSSSS? fAnTAstIC!@$%!!!!!!11!!!!

iT'S tOo bAd yOu NEvER FInISHeD tHAt ViViAN GAmE, A lOT oF aNoNS WeRE lOOkInG fORwaRD tO iT!!#%^&**!!!!!!!!!!!!!111!!1!!!!!!11l

there's not a single black race retard.

...

A then

Duuuuuuuude

Ok, now to draw the last 8 elements

what the fuck

"Caramel" is actually her name. There's an unsubtle naming scheme going on with all the characters (the other two girls are named Angel Food Cake and Honey Cruller.)

Officially, I admit nothing.
But I have a funny feeling we might not have seen the last of that game.

At least he wasn't using unicode to make his text melt.

I love some chocolate elves
But isn't her chocolate skin a bit TOO dark? She is quite toasted
Also how the fuck i open this flash file? opening on new window causes it to download, and playing the downloaded file downloads it again

This tbh.
This one's skin is way cuter.

It ain't my game but I'd suggest cropping their names to "Angel" and "Cruller" because those sound more like names a person would have, right?

Open it with something meant to play flash games, not your browser.

A is OK, but when you overlay B, you can make the curved segment follow the circle curve, and then they'd start to tile more seamlessly. The curve could hint how the shape can be tiled. Not sure if you ever need to do that, but good for making fancy patterns on backgrounds, shrines and what not. You can even take out a few lines and follow patterns of overlay to make designs for "spell" graphics, etc.

I'd go with the curves if it were me.

However, the straight lines are "different", so they'll draw attention to them, and they do point at the very center where there is no element. Could perhaps represent your "unity" of elements or somesuch. The straight lines tile too, they just can't match the circles segments is all.

Yeah, I may have to lighten her up a little bit for the sake of visibility. She looks alright on my monitor, but seeing the game running on some other systems she does seem a bit too dark. (Her current skin tone is one of Flash's default swatches.)

As for what the flash file is doing, that's a mystery to me - sounds like your browser is playing a cruel joke on you. Perhaps try the standalone player if you have it?

Yes, that's how they're usually named. Their full names will probably only be mentioned once, if at all.

Also reposting the pseudo 8-bit minigame since I guess a lot of people didn't see them last time. (Unlike the "real" game engine test posted above, these games are actually possible to win.) Controls are the same as the main game, but you can also use C and V to Jump/Attack and WASD to move.

This symbol im creating represents all elemental interactions, its not supposed to repeat and form patherns or anything it would be repeating our periodic table of elements over and over for some nice effect if it had a cool shape

anyone else using godot here? I just started a little blat city clone project, let's see how far I can go with it. Currently trying to implement collision detection

Looks neat. You should be targeting 60FPS, though. 30FPS gives off a "cheap flash game" vibe; 60FPS will play better and leave a better impression. Also the wind up animations on the attacks are probably too long, though I'm sure you will realize that on your own when you start implementing enemies to fight.

When pressing the shift and character keys, it acted as if I was holding them down instead of pressing it, meaning it sends me to several rooms down instead of just one.

The symbols are the elements that make up your world, but if it tiles it implies there are many worlds, which is nice symbolism. You don't even have to show it actually tiling for the human brain to pick up on that.

Also is that music original, or did you get it from somewhere else? It sound really good!

...

I get it that they're elements. The shape itself could be said to represent the world or the magic of the world, etc. Making patterns by tiling over it, then finding shapes within that can just those shapes more "meaning".

So, if there were a "Love" spell or charm potion, for instance (this is just an example). When you overlay the shape with the two central elements made into one, etc. The shape it makes has hearts in it. So, a love spell could feature the hearts in that shape - without the tiled image. Or, a potion / shrine would have such insignias. Later you might even reveal how the two-joined heart symbol was made in the game, perhaps some sorcerer discovered shapes and led to the creation of romance magic. It's just lorecrafting.

Another example: Magic stuff tends to have arcane symbolism. So, finding different sub sections within the tiled / overlaid pattern could be how you find the glyphs for your magic alphabet.

The player doesn't have to know where they come from, but it kind of makes it easy on you and at the same time gives you something to "mind = blown" a player when it's revealed the shapes they've been looking at thinking it's just arbitrary magical marks actually have a meaning that's been in their face the whole time.

Not an essential thing, but if you've got a nifty symbol that lends itself to such things then it's something easy to do to come up with designs and such.

Fighting games. zweifuss.ca/

Any effect not listed on the blog will not be attributed to elemental magic, but to other types like alchemy and arcane
The Elemental Lotus is just one of the magic symbols i will make, there will be plenty more to be made before i think where the symbolism of everything will be, this is not the final form of it.

By the way
PRAISE THE SUN

Good point, I'll set it up for a custom Pause key too.

I shot for 60 initially, but it dropped far too often. If I'm able to port it to another engine with less overhead than Flash, I'd definitely consider upping the framerate.

As for the windup time, it varies with the character - I presume you're using Angel? Cruller's attacks are almost instant. Angel is slower, but her attacks have much more power. (Also, it's possible to offset Angel's slow startup time by cancelling into her attacks.)
There are actually a couple of enemies in the demo, though they're very incomplete - you'll run into them if you take the lower route at the first fork. There are also some boxes you can beat up in the starting room.

It's licensed music - it was composed by Inu-goya.
www7b.biglobe.ne.jp/~inugoya_x/
And yes, it's great!

Can't play it atm but just wanted to say I'm still hyped for your gaem.

well, collisions sort of work now, but the tank stops a lot farther than it should, I know that the problem is the following
world.intersect_point(get_node("tank").get_pos() + Vector2(0, -tank_speed))

this is used to check whether or not there's something above the tank, the problem is that as tank speed is 80 it is surely moving 80 pixels above it, whereas the actual movement is carried as speed*delta(delta being how long a key is pressed, so it should take 1 second to move those 80 pixels), and changing Vector2(0, -tank_speed) to Vector2(0, -tank_speed*delta) breaks the movement so that I can't actually move

Damn gradient effects are magic

Nice meme

All I need to do is learn how to make music and I can become a 100% independent developer. I can come up with really unique and catchy melodies in my head put can never put them down. I'm going to be doing MOD tracker music. Any suggestions?

Sun is rainbow plus leaf and moon is the skulls ones right?

OneyNG posted a few FruityLoops tutorials on his Oneyplays channel.
I think that should be enough to get you started

someone recommended me those sites in a previous thread, I hope it helps:
umlautllama.com/Audio/theory/theory.html
wiki.openmpt.org/Tutorial:_Getting_Started

nope, sun is the top 3 (thunder, ray, lightning)
moon is the bottom 3 (ice, gravity, flora)
Since the moon is always on the opposite side of the world it cannot reflect the sun's light like in real life, the moonlight in my universe origins from flowing silver flowers that grow on one of the moon sides, the other side is pure ice, and gravity core of it causes the tides in the world
Sun or Sun2?
I think the first one was too big…

Glowing, not flowing*
Hum.. i think this sun is even better

Why are lightning and thunder separate elements when they are the same fucking thing minus the difference between the speed of light and the speed of sound?

...

Lightning is the electricity
Thunder is the explosion caused by air turning into plasma due to lightning

I actually get [triggered] when i see "Thunder" near a lightning icon

Thunder is the sound of lightning.
Its like making the sound of someone getting slapped and the slap itself as separate things.

isn't:
Thunder = sound
lighting = light
ray = electricity

???

Why separate them?

...

Danke.

Lookin' good. I was able to explore around the whole place without getting stuck or disappearing. The glowing runes were a nice touch.

thunder = sound waves thunder clap, or side effect of lightning; ionizing the air particles making a sorta plasma soup
lightning = electrons + thunder effect of ionizing particles, e.g. electricity/plasma
ray = photons, e.g. light

I'd presume the sun represent electromagnetism as the binding/origin force.

well they're not synonymous in a lore sense to that user, even though in english thunder/lightning are generally confused (thus synonymous with eachother), but they're not the same thing.

ray is radiation, heat, infrared

forgot I was working on this one, finally got around to finishing it.

If you guys are curious you can read about the elements here
fablesoflaetus.com/

What im making right now are the last 8 elements which i barely touched, the Celestial elements

is there an about page or something? I can't figure out what this game is supposed to be

NO
PEOPLE TOLD ME THIS SHIT WAS TO GO AWAY

fablesoflaetus.com/everyday/
The first game will be a simple farmsim, but im going to use this same universe for several games, so im working hard at worldbuilding, so everything stays consistent in the future when these details become relevant.

Oh yeah the part about the combinations to create the Celestial elements is not clear
The common element combinations to create them are wrong because this is how the mages of the setting theorize these elements are formed, when they are actually formed by the composites of the three common elements

Example, instead of
Light, Fire, Air = Sun
The real combination is
Lightining + Ray + Thunder = Sun

Have you ever played Lost Magic?

Nope

so we're farming elves to rape? I don't mind a lewd monster rancher, I just wish someone had told me sooner.

was just reminded of it
it's a pretty solid game where you cast a few hundred spells by combining elements,

Don't use shitty paid Unity courses, they're always terrible, and attract the worst teachers who do it for a few quick bucks.
Use the official unity tutorials (live streams, all the teachers I've seen enjoy teaching), or other youtube tutorials (sebastian lague, etc) with teachers that are teaching for fun (huge difference in quality, easier to get into, etc etc).

Ah ok, makes more sense now.

I pirated them thinking I would get decent production quality for a $200 course.

Also, the official documentation is complete and utter shit when it comes to more advanced topics

Almost, Elves and Orcs lore comming this wednesday

Gonna check it when i start planning my next game… for "inspiration"

unforunately for unity related stuff that's not the case. as it's rather young, and less formal compared to say photoshop.

Like? Also what language do u use?
I've always found docs to be above, and beyond normal standards; they document everything necessary.
I use a lot of their more "advanced" features too, like the draw procedural stuff, compute shaders, and command buffers.

C#, of course. I could spend hours complaining about it, but I'll just be smug and say you've never tried to make an online game with Unity's online documentation.

For certain topics pages often link to random blog, youtube, forum posts, and occasionally the asset store.

I like a lot how it looks, but i somewhat dislike the outline being so bright, I understand if you want to keep a somewhat purplish palette, but imo it has a pastel palette feel to it that I don't think fits with a bloody weapon

I think I might have made it too dark, as well, it's just something to consider if you don't think of this as the final revision

Oh, yeah I have no interest in online games.
If i were to do it I'd roll my own P2P framework, or use an open source one that's on github as the base; then build ontop of it.
Also, don't use Unity's multiplayer framework, from what I've seen (over the years of browsing the unity forums), it's not maintained nearly at all; due to way better frameworks for multiplayer being available on the asset store (or open source ones on github).

Makes sense.
They don't want to drop support for the multiplayer, and some employees are honest enough to point you towards the information I pointed out above; don't fucking use Unity's provided multiplayer API.

Meteor and Comet elements

Next i wanna make Austra and Borea
These are the elements that will link the border elements in the middle
Austra represents creation (Flora + Prism + Glass)
Borea represents destruction (Miasma + Ash + Thunder)
Since these are very important i gotta look for inspiration before making them, anyone has any ideas?

I was thinking of having Austra being some sort of rainbow flower and Borea some extra evil skull with the top cracked (to represent thunder damage) and the mouth releasing a dark cloud.

Why make games when you can sell shitty tutorials? for free engines, wtf.

should.. should I make a paid tutorial?

that wasn't supposed to be blood, it was supposed to be rust. I might need to make it a little more orange.

I mostly use the more purple outline since it adds some contrast but I'll try experimenting with some colors.

...

First you eat the sugar.

Then you sell the tutorial.

Then you make the Pong.

sounds like a pain in the ass, how about i just compromise on features and actually finish a product instead of pretending i have any idea how to code

If you just want to make stuff with GameMaker no one will stop you.

But just about every engine except GM has infinite scrolling… Even homebrew ones for actual NES.
github.com/sikthehedgehog/dragon

game maker gets way more shit than it deserves honestly, I think for people that just want to make a game it works pretty well. For the people that are setting out to show the world how it's done, it's better to make your own engine.

Me personally, i'm a dumbass so drag and drop+light scripting is plenty.

GM isn't the only engine that has those features now. Might behoove you to look at Godot and other engines which offer similar feature-set since it's hard to find a scripting language worse than GM's.

I've looked at godot and all I really remember about it was feeling like it was a half-finished product with inadequate documentation.

I looked at GM and came away with the same feeling. A language half implemented without basic things like data structures for grouping vars. No free scrolling, just a "room" based level.

What are your specific qualms about Godot, since it at least has these features GM lacks?

If I knew what my specific qualms were I would have listed them out in a previous post. The only specific thing I remember is needing wonky hacks to make the game physics resemble NES platforming and no tutorial on how to do sprite animations.

I remember thinking its sprite sheet system was simple. I added an image with the frames on it and told it how big the cells were. Pretty basic stuff, TBH.

Had vid related bookmarked from 2014.

Not sure what you could possibly mean by "game physics resemble NES platforming" since NES games had shit loads of vastly different platforming physics. Even within the same series, consider Mario2 vs Mario3. Probably would take "hacks" (code) to make any logic system work like some other logic system. It's not like GM has a "make things work like game X on NES" selector.

Doesn't matter to me, really, what you want to use, but it just seems disingenuous / fanboy like to categorize GM as the go-to engine for "just wanting to make a game" and "works pretty well", esp. considering further up we have some commentary about GM not working well at all for some.

I remember helping a dev using GML figure out how to walk up diagonals (apparently a common problem). Every standard solution I came up with was thwarted by GM's lack of a feature. They never did figure it out and wound up using a kludge. Had to limit the run speed or else the set of sensors wouldn't detect the next block, and get stuck in the level (apparently that's a common problem).

I would mainly discourage use of GM by those who are just starting out and want to develop transferable knowledge beyond GM. When I was younger the GOTO language was QBASIC. Like GM, it taught me how to do things the worst way. Took 2 months to learn QB, but 3 years to unlearn the mindrot of "the BASIC way" when I started using a real language. People with a history of GML remind me of that when I see them on codingforums.

That said, if Godot was perfect I wouldn't be making my own engine. At the very least it seemed like it was in the right direction, and usable. GM is usable, but dedicated to the wrong direction, IMO.

Back again at the X11 "documentation"… so, I have been implementing full-screen mode. This is done using the XF86VidMode extension. However, the documentation for this is almost nonexistent. The best thing that I could find was literally just a list of function headers and some random comments about what three of the headers did, which only made it vaguely more helpful than reading a .h file and guessing what they did based on the names. Then, when I get it working by reading something online which tells you how, by trial and error, they found out that if you call these functions in this order it actually works and gives you fullscreen mode… but then… you cant set the resolution back! So, instead of waiting for the stars to align I decided to use xrandr to change the resolution back, which should be simple… but then I get this absolute bullshit description that would only ever help you if you knew what any of it meant in context, so its basically unusable.

I don't know why im STILL trying to tame a system who's creators didn't even document.

Almost no application developers do that anymore: graphical applications use some toolkit like GTK+ or Qt and games usually get SDL to take care of the X11 boilerplate, especially since they get Wayland support thrown in for free.
Use GLFW, faggot
glfw.org/

I have no interest in admitting defeat

I took the easy way out, although I still want to know how to do it correctly sometime.

system("xrandr -s 0");

Have fun implementing Wayland support, libinput support, ALSA and PulseAudio support along with OSS and JACK if you really want to get into it after you're done.
Really, X11 is a gigantic clusterfuck nowadays and it's being replaced for a reason.

I'm done with X11 (for now).

I plan on using OpenAL just because I am not touching any retarded sound system myself. Maybe one year… but not this year, and probably not the next. I don't actually have a huge interest in audio. (mostly because its a feature that you just tack on after the major problems are solved…)

...

I wanted to do that too, but then I read the changelog.
glfw.org/changelog.html
I have a higher standard of quality than this project currently provides.
i.e. I don't make rookie mistakes in CMake files, nor fuck up an event system to the point it segfaults.

My code and its libraries must not emit even a single compiler warning and have them all enabled to max (a standard even EA meets, so it can't be that high).

I ran Valgrind on a library example and it puked so many errors that it literally claimed
valgrind: the 'impossible' happened: Killed by fatal signal
This only happens when you manage to luck out and piss over the memory guards and into memcheck's internal structures – which memcheck tries incredibly hard to prevent, but if you play the error lottery a million times, then you win.

I'll check back in a year or so, by then I'll already have made a game and engine. For now it's not mature, IMO, and it's maintainers reek of noob. Good intent but accept patches I wouldn't touch with a 10ft pole.

Granted, GLFW is fine for the quality of most games – but I'm wanting to make at least the rendering and input system something to the standard that I can at least license in the medical sector (there's a niche for realtime 3D voxel rendering if you can speak the language of ultra-sound machines, and can write error free code).

I wouldn't trust GLFW to run on a system that shouldn't be rebooted often.

Busy right now, but I have full screen supported in raw C + X11. If you haven't nailed it down by then, remind me later and I'll try to help.

Opinion disreguarded. Write to the X11 protocol you get Wayland for free since Wayland provides an X11 server. This is what I mean when I say GLFW reeks of noob.

The major problem is having a 3-D world that you can walk around in. Audio is a few days with OpenAL in comparison. Sure, I could spend a month on it, and I want to do that sometime, but I don't want to get caught up like that now.

yeah, if you dont mind showing me what the fuck you did to make it sing i'd really like to see it
I dont think that I am going to have any more luck punching a brick wall than I would have reading X11 documentation

Didn't even know this, i'm glad that I wont have to think about it.

I've heard mixed things on GLFW's code quality before so yeah, in that case you're probably better off rolling your own solution. The Wayland issue is another story: XWayland adds a little performance overhead (only a couple frames, but it's still something) and from what I've seen Wayland isn't nearly as messy as X11. Saying people should only support the legacy solution because the alternative has a compatibility layer is similar to people claiming we don't need native Linux builds thanks to Wine, except adding Wayland support to a Unix application is much easier than porting a Windows game to Linux.

OpenAL is pretty good, it's just surprising to see you using someone else's library for once.

Understatement of the century.

Using something general is definitely the best plan. Don't even bother with OS-specific audio frameworks.

I have always used OpenAL… its just that using other people's code is not that interesting to post about

besides, X11 is someone else's library… in fact, its all libraries… but my goal is to use as little of it as I can while still making a more impressive engine each time

so, audio isnt the question I want to answer first

I think you can get around using OpenAL, where as X11 you're basically stuck with it since the majority of linux machines use it. I'd think using SDL or a window library like that would be closer to using something like OpenAL

Also X11 is fucking disgusting I learned C by trying to use that shit and it was a mistake. I copied a ton of their bad habits

library's a library, the only way to get around it is by writing your own bootable program in ring zero like the guy a few threads back is doing. Maybe X11 would be tolerable if it was documented? At least in MSDN I know what everything does which makes it a walk in the park compared to this.

I'm just going off the context I think the other user is going in. Assuming stuff like OpenGL, which you can't really get around if you want to use your GPU, are givens.

Also it is documented. Like absolute garbage. It feels like it's meant to handle the display of +100 computers to.

yeah, i know what you mean, you have to use X11, openGL, and all that to do the things that you have to do.

I supposed what im trying to say is "documented well". Microsoft has this down really nicely and I didn't appreciate how good it was until I saw the man pages on linux.die and all the other sites for those.

msdn.microsoft.com/en-us/library/windows/desktop/ms632680(v=vs.85).aspx

MS has a detailed explanation of what every single parameter for all of its functions do, and all sorts of helpful links… I'm just preaching to the choir though, but maybe you haven't seen it.

The people currently maintaining X11 know it better than anyone else and even they can't stand working with it.
You could always look at the source code of the few Linux games that don't use SDL or even SDL's source itself if you're absolutely stumped.

This is the problem. If every function in X11 was as well documented as MS documents it I bet X11 would seem to be much, much better on the simple merit that people actually know what the functions DO so they don't write buggy code by mashing functions together and hoping it works.

What's a good engine for 2D projects? I've tried using Godot but my lack of experience makes it a pain in the ass. I'm a bit reluctant to use game maker and the project head refuses to use Unity.

Away from the dev machine right now, I can post code later. Here's the pertinent spec.
specifications.freedesktop.org/wm-spec/wm-spec-1.3.html#idm140130317598336

// Get a window state atom.Atom xstate = XInternAtom( display, "_NET_WM_STATE", true );// Get the fullscreen atom.Atom xfull = XInternAtom( display, "_NET_WM_STATE_FULLSCREEN", true );// Change the XChangeProperty( display, window, xstate, XA_ATOM, 32, PropModeReplace, (unsigned char *)&xfull, 1 );

The thing to remember is that X is the protocol for interacting with a window (potentially as a workstation connected to a mainframe or supercomputer). If you want to talk to the window manager then that's Gnome or Unity or KDE, etc. Those projects standardize on Freedesktop .org. Unlike in Windows, you can actually replace the window manager on Linux/BSD. This capability does come at an increase in complexity.

A practice that's frowned upon, but can work in certain circumstance, is to make a fullscreen application that bypasses the window manager at window creation:
XSetWindowAttributes xattrib;xattrib.override_redirect = True;XCreateWindow( /* etc. */, &xattrib );
But you'll need to also wrestle away input events too, and this doesn't play nice with window managers, hence the freedesktop standard instead.

Beware of Win32 documentation. Worse than no docs are bad docs. If you use MSDN enough you'll learn to scan the comment sections where a kind user, having battled with the disparity between API and docs will detail how MSDN did not describe how the function actually [mis]behaves (some really nice ones provide work a around for such really broken parts of Win32). If you're unfortunate you may find yourself in that role, being the first to report the issue. At least X11 has source code available to read, rather than having to poke and prod at a black box to determine it's function.

Love 2d

I managed to make some shitty games in that before, really shitty, but it's a neet little 2d engine, and you script it in Lua, although I used it fucking ages ago so maybe it's different now.

love2d.org/

I know about love, and I'm not sure about using Lua. It's a rather ambitious project, at least conceptually, so I don't think love 2D can quite cut it. that said, we are trying to keep it within the realm of possibility.

You could make the two icons so that they fit together ying-yang style or in some other way when overlaid.This could then lead to symbolism about of how something must be destructed before something else can be created in its place etc.
Though depending on how you want to handle differing magic practices and their philosophical and physical background you could just keep that stuff out of elemental magic and work it into Alchemy (eg Analysis > Understanding > Synthesis)

My only problem with Love2D is Lua… which basically has a roll-your-own OOP system, and while it has to be embedded within C it's not designed to be a good embedded language with tight language integration.

I've been giving this some thought, since my Font system (being 2D) has sprites and animations. It wouldn't be too difficult to turn my font creator / renderer into a 2D engine by itself.

There seems to be a niche that's left unfilled currently.

Perhaps we should put together a list of features that the "Killer 2D" engine would have (ignoring deployment platforms for the moment – cross platform is simple). All the bells and whistles a 2D engine could want would likely be fairly simple to provide, but the editor tools and scripting etc. to make the engine user friendly are the harder part. Devs who can make a 2D engine usually don't need all that to make the rest of their game happen. And dedicated 2D engines like GM and RPG Maker seem to be made by devs who can't into actual enginedev and language dev (those devs get gobbled up by the 3D industry). I think the problem is that no one is really doing market research, i.e., look at what the demands are. There is still a huge demand for 2D games, and thus a 2D engine, but all we have are good 2D frameworks and shit 2D engines. The sweet spot in the middle is vacant.

Below is a rambling rationale which I may use to convince myself and perhaps a few peers to work on an exclusively 2D engine rather than 3D engine. saged.

Flash is great, but getting fucked by its own publisher / developer, and the baggage it carried as a web plugin. Its future is uncertain esp. with the encroachment of mobile devices. Adobe / Macromedia seems to want to just sell the Flash Studio as a tool to make content which compile to HTML5/JS rather than have a "plugin" architecture. But browsers are major unstable and slow shit as platforms go. (WebAssembly will help, someday).

Godot is, like so many engines today, trying to be everything to everyone and just getting more complex. The learning curve is steeper than GM but GM is whack. I would have liked it if Godot had split into two separate engines rather than spanning the 2D / 3D divide. I looked into its codebase to see if I could build my own GUI to replace it, but their "editor is the engine" philosophy is literally retarding. If the engine was separate from the editor then I could have just built a user friendly front end for Godot.

When you make something for "everyone" you've made it for no one in particular; This problem plagues most engines since they gather more feature creep rather than becoming (or spawning) "the FPS engine" or "The RTS engine" or "The platformer engine", etc. Most actually successful engines have historically been made for a specific game and then host gameplay which lends itself to the features the engine provides. That's even more precise: Rather than "The FPS Engine" it was "The Doom Engine" or "The Unreal Engine", etc. Aimed at meeting the specific needs of an individual game.

Most of the games that you'd want to use the engine of, perhaps an idTech or Unreal that's open sourced, the licenses are shit and overpriced; In some cases you can't even buy a license to the old engine for releasing as closed source.

It seems like the industry has mistakenly departed from their senses in the quest to make "The Everything Engine" – It's a never ending quest like trying to make the Philosopher's Stone or find the Holy Grail. Gone is "do one thing and do it well", and now popular engines are made to appeal to the largest user base rather than provide a rock solid feature-set for a specific genre of games. Such pre-made "game cores" need not limit the possible game genres so long as source code can be modified of that core (or the hooks for scripting are generous).

Imagine if instead of using a 2D framework to try and make a metroidvania, there was a pre-made metroidvania engine which had a pre-made basic metroidvania game that you then edited. You could implement a whole new game but just start with swapping out assets for your own, making new enemies / boss logic, etc. You wouldn't have to start from scratch. The same would be true for 4X, FPS, RPG, Shmup, etc. genres. To reuse code, a common cross platform framework would support the collection of special purpose engines. It would be a middle-ground between "modding an existing game" and "developing a new game from scratch".

The question is: What would those core genres be, and what is the killer featureset for making games of that core genre. One core be a 2D platformer, of course…

What kind of 2D project specifically? An isomorphic engine is very different from a platformer, for example.

…And what are the killer features a dev like yourself wants in such a 2D engine? What are the requirements that your ambitious concept demands?

well, i didnt mean to toot my own horn when i called it ambitious, i simply meant that it demands more than something as simple as love/lua. As for what exactly it is, in terms of gameplay, it is meant to be a game in the same vein as older castlevania titles. I'm considering taking a look at the inner workings of SotN to help with the actual design of the game itself. I'm a bit uneasy about using a pre-existing game, as that could lead to issues with the developers. It would be easier, sure, but I would much rather avoid legal issues wherever I can. Mind you, this is meant to be an official game.

Furthermore, we still have some things left unresolved conceptually, such as moving from one part of the game to the next, all the different kinda of weapons and equipment, etc. Getting our ideas out and keeping everything within possibility is more difficult than it seems.
I thought to look here because I believe that a good bit of Holla Forums knows good game design. And as such, some of them might be able to help me.

orx-project.org/

Im really tempted to use ying-yang because it fits so well in spherical shape, but i want to make
something more original

looking for help with my uni Matrix determinant calculator again
i reworked it so now the threads actually behave like threads
i create as many ThreadedMatrix as i'm allowed when the program starts
each of them has a que of matrices that need to be calculated
matrices with a size of 3 or more create submatrices and push them into the thread que (and specifically pushing them into the thread with the smallest que)
matrices with a size of 2 (or 1 if the original matrix was like that) return the result of their determinant
each thread adds the results of their que submatrices into the total result
once they all run out of elements in their ques the program finishes and we get a result

my issue is that i'm still not getting much of a speed up
at least it's not going 10 times slower with threads like the last version, but on average i'm getting 33000000 ns to 57000000 ns, regardless of how many threads i have running. is it just a toaster limitation at this point?

main class - pastebin.com/tibwBerT
matrix class - pastebin.com/bcJ9du5V
thread class - pastebin.com/9cw3mkSM

I get it. I didn't mean to sound condescending in the least. I think 2D projects can be as ambitious as 3D. One dev I know is working on a 2D game the size of a planet with collaborative online gamelpay.

You're probably right about Lua if you have a ton of different types of items / enemies. I've used Lua before and it became a mess when the project scale got large enough. I used it to do scripting for a GUI toolkit (signaling 9000v+ motor controls). I made a bad choice and I regret that decision still today. It's good for smaller projects, but the language doesn't lend itself to large complex domains.

So, it would be nice if there were an existing 2D metroidvania engine that you could freely use to make games without worrying about licenses or copyright issues? Not that I have a solution at present, I'm considering floating the idea of making a 2D engine you could use to make and sell games with no strings attached. The example game would use CC0 licensed art assets so there's no conflict with commercial use if you kept some in, such as particle effects, etc.

Thanks! This framework looks neat… but no scripting. And why the fuck isn't their logo a pair of Orcs?
orx-project.org/wiki/guides/beginners/spritesheets_and_animation
Using a config file format to specify sprites is drek. There does seem to be a plugin for GIMP to create sprite sheets, so that's more than I can say for most. No level editor though.

IMO, scripting is an essential part of 2D game engines for the JLMG crowd. If users could write logic in C++ then they wouldn't need a game engine since they could just use any of the 2D frameworks available to make their game with, SDL, SFML, etc. Their partial engine looks great for a C/C++ dev to extend into a full engine, but not something someone who would use GM would pick up.

Orx, as well as any 2D framework, could perhaps be extended with a set of killer features for simple 2D gamedev. IMO a GM killer would need:
* No licensing headaches - 100% free.
* Game Logic / Event Scripting.
* Serialization for Save / Load game.
* Sprite & Animation Editor.
* Transparency support and various blend modes.
* Shape-based and pixel-based collision models.
* Level Editor - not just tile layers, triggers and behaviors too.
* Customizable Menuing System (menu editor)
* Streaming level data.
* Regional parallax backgrounds / foregrounds (can have different BG in different areas of a level without having to write your own script).
* Cutscene maker.
* Gameplay record and playback (esp. for speed runners and "attract mode": en.wiktionary.org/wiki/attract_mode )
* Render out recorded Gameplay demo as images and audio stream (dev feature, for making videos necessary for marketing), not 100% necessary, but much better than screen recording.
* Netplay Support (delta serialization & client side prediction at least)

Can anyone think of something I'm missing? (need a solid list to make a pitch).

I'd like to put together a definitive killer feature list. If there is any engine offering such then there's our answer for "what's the best for 2D gamedev"? Otherwise we can try and cobble together features from things like Orx and Tiled, pick a scripting lang, etc. Then implement the remaining parts if that's possible. Perhaps even put a bounty on features needed and get kikestarter to fund development.

IMO, an actual game engine (not a partial engine or framework for making a game engine) would have at least the things I've mentioned above. Artists and level designers should be able to sit down and make a game without touching any code except the scripts to modify/add logic if desired. Consider Rayman Legends enigne, vid related. It's custom made for a game like Rayman.

An engine designed for 2D platformer would be designed for a 2D platformer and have a much different "logic/physics" and level editor than an isometric game or RPG, but they could reuse basically all of the other engine features.

This is my point. Things like Orx don't replace GM. Even Godot is distancing itself from the niche it once filled by over complicating it's UI There is a plethora of cross platform frameworks but few actual engines.

If we don't ever nail down what it is we want in a 2D engine solution no one will ever make it. Problem is, some gamedevs don't know what enginedev features are possible – such as online collaboration server for devs w/ resource hosting, online leaderboards w/ speed runner demo files to watch, etc. Not saying stuff like that is necessary for every game, but gamedevs have to make their requirements known in order for enginefags to make it happen. Otherwise, wait fornever for generic engines to cater to specific 2D genres.

That's done better by external tools, but an animation editor is still a solid idea.
Completely unnecessary and wouldn't give devs the flexibility from setting up their own cutscene system.
Godot already fills most of your requirements but I'm not a huge fan of its UI or hierarchal scene trees.

I'll try running it for you, I have 4 cores and 4 virtual cores. Just give me a moment to install the Java SDK and get it running.

oh yeah, i forgot to mention, if you're doing -i (reading from a text file) the text file has to be in this format
n
a11 a12 … a1n
a21…
…
an1…
n will be the size of the matrix, the rest are it's elements
example
71 3 4 8 8 5 36 2 6 3 8 4 25 3 8 5 3 7 91 5 3 9 1 4 61 4 7 3 8 2 53 1 4 6 8 6 41 6 8 9 2 3 5

For that file, no matter what number I type after
(I tried between 1 and 32) -t I get between around 11000000 and 8400000 total, and repeatedly doing it with the same number of threads results in about the same range of ns. So I can confirm it's not a hardware issue on your end.

Then I tried running it without -q, and 2 threads. Got the following output:
And then it just never finished. I tried again with 1, 3, and then 2 threads again: pastebin.com/dJ56neYt

No matter, what number I type after -t, not sure why that turned into a linebreak.

Interesting, I have not been doing this and I have an enormous function using XF86VidMode extensions to achieve this.

But, when using your method… it doesn't work. So I'll continue using my old method.

Good point, I don't doubt you on this one. I've seen it on some comment sections but I haven't hit it yet myself since I don't dive that deep into win32.

nice.

...

I like you.
Everyone's poking fun at how shit your animations are, and yet you keep making games, not giving a shit.
I wish I could do that.

frankly, if i described the concept of my game, most people would think it'll be boring. and that's precisely why i'm working to make it fun.

lol

you stop giving a fuck after a while.

that's bizarre, i'm getting it too and only when it's not quiet
and i still can't understand why it doesn't run any faster with more threads. i swear i'm doing it correctly

alright, as far as i understand, the issue is that at one point none of the threads have any tasks in them, but they're about to get one pushed into them
before it gets pushed though, both threads think they're done so they stop working
then even if one of the threads has tasks left, it's already finished so it won't continue
i guess i can restart the threads if they have leftover tasks, but i'm supposed to know exactly when they started and stopped which becomes impossible if i restart them

pls rate.

change the punch animation, it's wrong

yeah, I'll try.

Actually I like the strong punch animation, the quick jab has to go though.

Both of the punches are really slow, actually. Too slow for fun gameplay.
That's not taking into account that you're trying to emulate an unfitting artstyle with an extra dimension and it just ends up looking like shit. Develop a style, user. Good texture work will get you everywhere.

the character looks drunk when walking backwards.

Still needs a lot of work though.

You don't say, user.
Keep at it though.

Changed Nova/Quasar into Gamma/Gaett
The names doesn't really mean anything, just made up for the reference
Nova/Quasar never made sense anyways, i used them because i was out of ideas

Only two elements left

How do I distribute points on a triangle at a given density?
I need this to make the grass so that my brother stops bothering me
public override void processDetail(Mesh m2) { for (int i = 0; i < m2.triangles.Length / 3; i++) { Vector3[] triangle = new Vector3[3] { m2.vertices[m2.triangles[i * 3]], m2.vertices[m2.triangles[i * 3 + 1]], m2.vertices[m2.triangles[i * 3 + 2]] }; } }

Send help.

Practice user, this is the before and after edition, save your current sprite and post again when you improve.

i'm gonna fucking kill myself over this thing
either i end up locking something and i get a correct result but slower with more threads, or i get a fast result with threads that's wrong 99% of the time because the threads ended up stopping before they should have, even though i can even predict exactly how many processes the whole algorithm is supposed to do
or it inexplicably gets stuck in an infinite loop
multithreading was a mistake

MY FELLOW NODEVS

i would once again like to say that i did not have sexual relations with that woman
i did however go to ledoux.itch.io/bitsy where they offer a free tile-based tile engine, perfect for those with a lot to write and not much else
visit him today at ledoux.itch.io/bitsy , that's L-E-D-O-U-X (dot) I-T-C-H (dot) I-O (slash) B-I-T-S-Y

Its nice if you made it, but I don't think anyone serious would use it.
Maybe it would be better if it were a native tool instead of web-development ?

Actually, the more I scroll down the page, the more I understand, that it is for lazy itch """game devs"""" who like to spam out shitty things they made onto the site, which will be clicked on for the moment they are on the top of the "new submissions" page, only to fall off, never to be seen again…. Its interesting to look at.

not me really, just saw it on the itch.io frontpage & a game jam for it might make one since i've got some time to waste tbh
hence for the "a lot to write and not much else"
and i also just wanted to reference that shitty bill clinton ad lol

Currently writing Elves and Orcs

Please rate my loredump
fablesoflaetus.com/lore4/

Have you tried to produce a consumer and consume their produce?
en.wikipedia.org/wiki/Producer–consumer_problem

First just get the code working correctly single threaded - remove all thread code, make sure it works. You might already have that.

Now, in a single thread you want to make sure you can break your problem up properly. Try making a function that processes a chunk of the matrix at a time…, creating a partial result pass, then a pass that processes the other part, and finally combine the result.

If you can break the problem down into passes that don't depend on eachother while they compute their own results then you can use multi-threading. It has to be a good sized chunk, not just a few multiplications at a time.

If your problem relies on the completion of the prior steps to begin processing the next step, and can't be "decomposed" then it is not paralizable and will not benefit from multi-threading, as you'll just have each thread waiting on the other and you'd be faster just single core crunching it.

If you can split up into sizable chunks then then use a thread pool to break the problem down into processable chunks:
en.wikipedia.org/wiki/Thread_pool

If the problem can only be broken 2 ways or 4 ways, then only create that many threads. It turns out that there are some fancy formulations for your problem. I'm sure there's a paper published somewhere with an example. A smarter algorithm can beat more threads.

Your posts just make it look like you're viraling this thing here

Next loredumps
I promise i will waste less time on these and go back to programming soon

single thread works fine
everything should be evenly distributed among the threads
i've been debugging this for hours, the issue is that one thread dies right before it was supposed to get new elements in it's que because for a brief moment every thread had empty ques
pic related is the example
two threads
first one thinks it's finished at one point, and right after that the second one gave it new tasks. it never finishes those tasks and the second thread is stuck trying to finish the program
and calling thread.start() after it's already finished just gives me an error
i could just keep all the threads alive until there's absolutely no more tasks left, but i have no way of knowing that for sure

How long have you been writing about this world? When did you conceive of it?

Worked a little more on my game scirra.com/arcade/action-games/first-game-18025?action=updated

Fragments of it have been popping into my mind for over 10 years
They just gathered into a more coherent mess recently

Talk about scarce ammo, you could start with a few rockets
Also what will be the next obstacle after the Doritos? Mt. Dew?
Just kidding, keep working on it user, i believe in you

thanks, the placeholder art will be gone soon

explain yourself

Your only contribution to this thread is plugging this shitty product, which makes it look like you are advertising it.

point taken

That's cool, I enjoy seeing it come together from afar

...

Are you synchronizing on your queue? The way it works with a work queue is that there are threads started up and just idle, not doing anything.

They have a workload var that's NULL or "do not work". When notified a worker will wake up, check the workload var, and if there's work to be done, process it then mark the workload as complete.

A manager thread keeps references to the threads in the pool. It's the one that gets the workload and starts chopping it up. The manager knows when there's no more work to be done. If the manager creates the thread pool on construction, then it should have a secondary function called to .doWork( matrix ); This function enters a loop that doesn't return until all the work is done. It constantly locks the queue, adds work, then releases the queue and locks the result queue, gets any results and releases the result queue. If no results it can wait to be notified, then go back to getting results, otherwise skip the wait and repeat.

Well, if you don't know when a job will be done then you've got a union worker pool.

I just realized that I made my fucking font system while working fore a company that technically owns all my work while I worked for them, even hobby shit. Congradulations $DYNCORP you are the proud owner of a text renderer that has animated sprites for characters, and I've got to rewrite a chunk of my fucking font engine.

That was one of the times I didn't get away with drawing a line through that clause of the employment contract paragraph and initialing it. Usually when they have such clauses I have them voided. If HR gives me shit I say,
At most places they keep a different contract in a drawer that allows you to do hobby stuff with a non-compete clause, but you have to ask. Sometimes they won't budge, and wind up owning shit that's useless.

Still a net-positive day, as I got my cross platform multithreading code working nicely and started on the netcode.

Maybe I'll make a plain text, and remake the sprites to have bones for stretching like that Rayman Legends video shows up there somewhere.

We page 13 now.

that only happens when they're waitin for new elements
its like this
original matrix goes in thread 1
submatrix 1 goes in thread 2, submatrix 2 goes in thread 1, submatrix 3 goes in thread 2 etc.
the que is only empty at the very beginning and end of the algorithm
i do that in the matrix class, although that's happening in different threads, is it an issue?
i originally meant that i had no way of knowing if there's any work left to be done that would actually be accurate. i fixed that by also keeping track of when i'm making new matrices, but it still doesn't fix my shit enough
i've tried making threads wait and then notify each other to start up again, but i was getting an illegal state monitor error
i've been working on this shit all day. it's 5 am already and i can't sleep until this somehow starts working correctly

NEW BREAD

Damn son. I wish i had a continuing lore, instead i have a folder that exists since 2003 and has lived through 4 different hard drives, including 43 stories for books i wanted to write, 107 folders of game ideaguying and even a few selfmade warcraft 3 campaign maps, the first game i have ever modded for. And still not a single finished game.