Amateur Gamedev General ~ /agdg/ + /vm/

I didn't really "choose" programming. I got into it as a kid by accident (fucking around and making games on a TI-84 with TI-Basic when I should have been doing classwork) and it's been my career for a decade now. Programming chose me.

I'll look into that. I'm just tired of my shitty programmer art. I've paid for commissions before, but would kind of like an entire game that was just my creation.

Is drawabox really that bad? As a programmer, it kind of appeals to my autistic exercise-driven nature, but I was planning on supplementing it with other material soon anyway. They cover observational drawing starting in lesson 2. I'll see if I can find Bargue's Drawing Course online for free.

Been a cellist since I was 10, so I got lucky there. You want to get steeped a bit in music theory and at least learn one instrument. Your best bet is getting a cheap keyboard from some local Guitar Center or something (they usually sell floor models for cheap if you ask; I got a decent Yamaha keyboard worth $150 for $80 because it was a floor model and had a tiny dent on one speaker grill) and learning to play a bit. Learning to write music without playing an instrument is possible, but much harder from how I've seen people do it. Guitar can work as long as you don't fall into the "tab trap", where you "learn" just by playing tabs but don't ever connect what your hands do to the music theory of it all. Piano is the very best for learning because what you do with your hands is the closest representation to the sheet music, and it makes it trivial to do things like build chords and play around on a scale. It also makes the octave structure very obvious.

So learn to play the piano a little, learn music theory a lot, and use the piano to explore and supplement your music theory learnings. After you have a solid base in music theory (at least understanding scale structure, chord structure and inversion, chord progression, and modes), you can start researching music composition and the like, but most of music composition is a pretty simple extension from music theory itself, plus some techniques. You can play with writing simple tunes while doing all the previous steps too, but probably should refrain from trying to actively research too many things at once or you're going to hurt your own retention and get burned out.
Music is hard. There's no real "fast-track" to writing music from nothing, and it's not really easy to skip all the music theory unless you're some kind of savant who dreams interesting tunes and can lay them down without losing it.

Would an apolitical, non-pixelshit indie game do well on kickstarter? the only meme-alignment it has with it is that it's kind of a board game

who am i kidding there's only one way to find out

So do you start with laying down a chord progression first before thinking of melody or what's your process? Is there a method for approaching rythm or do you just experiment aimlessly?

Just make really attractive promotional material, I think that's most important

Nothing does well on Kickstarter these days. The hype is dead, even normalfags think its a scam.

I'll say up front that I'm pretty unskilled at writing music. I've always been an instrumentalist first and foremost.
There are multiple approaches, and there is no canonical way, but the usual beginner approach is chord progression first, then melody, because the melody has to agree with your chord progression and harmony. The order you write them isn't actually important as long as they work together.
Writing music is a lot like writing a story, in that the rules of what works and doesn't work are steadfast and can't really be violated (unless you're very experienced and know exactly what you're doing), but the approaches to actually doing it vary wildly and are usually very individualized. The important part is learning those rules and being able to recognize very quickly whether what you're doing is going to work or not in the long run and how to recognize and fix mistakes.

I personally mix it up as often as I can. Sometimes I'll lay out a melody and build a progression and harmony behind it, sometimes I'll put in a solid chord progression and build a melody on top. When I do melody-first, it usually ends up sounding more interesting but is more likely to clash (my progression can end up pretty wonky). When I do chord progression first, it sounds far cleaner and neater, but usually more formulaic and boring. That might just be my own inexperience, though. My intuition is that practicing both approaches is probably best to get a good feel for one while you write the other. I've known skilled writers who could build both in their head and then write it all down at once, essentially writing the whole piece beginning to end instead of layering it on incrementally.

BEHOLD
Tic Tac Wars 2: Tacticool tic tactics

weekend projects sure are refreshing

As a guitar player I can say this is fucking true, learning music is quite difficult, specially by yourself. I recommend you pick an instrument and get some classes, as a teacher's guidance can take you quite faster than trying to figure things out by yourself. It also helps to avoid bad practices in technique and posture while playing.

I hate teachers and the concept of having to learn "theory" with a passion and would probably never do such a thing yet i pick up new instruments really fast without instructions given time and this whole "you need to do x to git gud" attitude most people have pisses me off as well. Despite this being the case and me knowing pretty much jack squat of music theory I cant even read notes I like to consider myself an okay composer. I wouldnt say learning music is difficult or hard. It just needs time and effort like anything else.

A nuuuuuu cheeki breeki iv danke

Will your game will be a mix of first person + third person to command armies? What happens if you don't give them orders and just want to stick to first person? will they be controlled by AI? Would be nice being able to choose either of the playstyles at will and still be able to win.

Well, it is an experimental project and there are many possible ways to do this.
1. The easiest way, original plan: FPS (team of players) vs RTS (archon mode) assymetrical gameplay
2. Battlezone way: can build and give orders from FPS view
3. Machines- Wired for War way: have the FPS mode be an option, join the battle to help your units but not much more

and mixing up all of the above to the extent that will prove the most fun.

Maybe have a mode with hero units that can be AI controlled, but are more efficient if a player takes over

currently It seems that this is true for every unit, regardless of its strength.

I will pay $50 (fifty) USD to whichever user makes a functional, polished, AAA clone of MMBN3/6 with netplay and modern conventions and art direction. Will pay after completion, I don't want to get burned again :^)

i made a neural network pathfinder for homework today
i've got a graph with 7 elements, each connected to each other at different costs
in the old dijkstra version all the paths were 2-3 elements long
first time it finishes the path here it's 2 elements, and after a few hundred generations it usually has 3 elements in most paths

i played with both neural networks and learning algorithms in college, but the only thing i learned is how "not to do it".

Nope, this is not how rockets work

Is it written in HolyC? :^)

Make a new weapon out of that tbh

ok negro jazz man

I implemented loot crates for my game.
I'm ready for shekels

Barrel launcher, i like it

I hope some most lootcrates have only a -33% off the purchase of the next lootcrate for the next 6hours in them.
:^)

Morning lads.

Thats some gay shit user

Good morning by the way

...

Why do you put "theory" in scare quotes? Music theory isn't some ambiguous thing. It's literally math for the very most part. You wouldn't say "I hate the concept of having to learn "math" with a passion" when talking about learning calculus. I'd recommend you learn what the word "theory" means, because it doesn't sound like you think it means what it actually means.

Of course you wouldn't, because you haven't learned music. If you are actually writing music that sounds decent without having any amount of music theory and without being able to play an instrument,
1. you're a savant and aren't comparable to more than 0.01% of the population, making your advice useless to the average learner,
2. you're bullshitting, and you actually suck
or 3. you're going to hit a ceiling where you won't be able to improve much at all without actually going back and learning the theory you've missed.

I assume that trying to play difficult Touhou songs slowly in Synthesia doesn't count as "learning an instrument"

is there a git gud guide to blueprints

just pirate the mp3 like a normal person

VLC is an ok mp3 player, the real travisty is not still using the original Winamp. it really whips the llama's ass

...

Eric Lippert's C# articles are always a good read

How should I go about selecting enemies in a turn-based combat system when they're laid out like pic related? The way I typically see it done, and the way I'm doing it right now is for all the enemies to be in an array, and it's cycled through by pressing left and right. What I want is to be able to press up, down, left or right to select enemies based on their position, and I also want the players to be selectable.

I used to have a nice WinXP+Winamp+SRS audio sandbox set up but I keep formatting computers and haven't recreated it. yet
I'm not really sure I can buy Omni Audio's OSTs, since they went bankrupt maybe 3 years ago. I don't think they were ever available So that song is 'pirated' as gamerips are the only source for NFS3 music.

Sidescrolling nun punching game is still in the thick of internal reworks, so no screenshots. The animation editor, which took a long time to finish at my slow pace is mostly done. I want to have at least one finished level in time for the next demo day, but before I work on levels I'll have to rework my collision tree so it doesn't shit itself when there's more than 10 colliders in the game and finish my planned main character animations (~15). Not to speak of designing enemies, animating enemies, internal stuff for AI & player damage handing, playtesting, maybe a boss, basic background assets… Originally I wanted a tutorial level and a real level but that might already be too much to make it in time.

Take the x/y position of every character on screen and make the arrow keys work on that instead of their IDs. The exact selection mechanism is probably preference, maybe something with distance, only looking in ie the left and not the right direction, but preferably someone that's actually closest to the left/right/up/down arrow when pointing from your current selection.

fug does releasing some kind of terrible demo count as finishing a project?

Don't be silly, I've already drenched my life in sin by having more than 16 colors. Still 640x480 though. Anyway, I got the color working properly. All I had to do was tell opengl the data was in BGR format, duh.

Why do you want to make games user?

For example, if you're selecting to the right:
Just flip the axis if you want to select up/down. I compare the center of each object to determine the distance between them. You might want to add some checks for some edge cases, like trying to select left/right when everything is in a vertical column.

If banshee wasn't a piece of shit that keeps crashing randomly it would be the best music player tbh, second best is gmusicbrowser even if the gui freezes from time to time.

I usually have one array for all the enemies alligned by y and another by x, works pretty alright

Jesus

That's because it's not aligned. It's a vector-based font and I'm doing what I can to make it look like my concept because Unreal doesn't let me use bitmap fonts.

Its called BGR not little endian although it's the same effect. Also are you checking glGetString(GL_EXTENSIONS) for GL_EXT_bgra and using BGR_EXT ? This is a reminder that not everyone has a graphics card that can load BGR textures! This txt has the extension specification: khronos.org/registry/OpenGL/extensions/EXT/EXT_bgra.txt

Remember, if it's not in OpenGL 1.1, query the extension! Also, is that a power of two texture? I bet it's not, you also have to check this extension as well: khronos.org/registry/OpenGL/extensions/ARB/ARB_texture_non_power_of_two.txt Since it wont work at all on graphics cards that don't have this, and it'll just error out and probably show up as plain white. Other than that you're thing looks pretty good user, keep up the good work!

However, if you didn't bother implementing a bitmap parser that supports non-power-of-two DIB's, you don't need to worry about that. In general you should just use power-of-two textures for everything though.

How does the biome distribution look? I personally think it looks too squiggly, but I'm not really sure how to make it look less squiggly.

Take my image properly this time.

i've always wanted to learn music but life has a funny way of telling me no, wanted to draw, and was a good artist in middle school. but traded it in so i can get better grades and get out of that hellhole without anyone following me (didnt work one bully followed me to high school), now i hate myself for forgetting those skills. funny thing is that i can sculpt now and at an amateur level which i am glad about. but who would hire a guy who can only do models and not draw. anyway sage for being a blog post.

but question anyone know any good reference material im kind of running blind here, tried looking up tutorials on the torso but nothing, most worry about pose more then anything. also one thing that been bugging me, and that is the clothes because how does it work as topology? does it go with the skin or on top of it, and should i sculpt it first or no? (see 2nd pic)

Thanks for the info. All things considered, I'm starting to become paranoid that I've made a horrible mistake or that I'm clinically retarded. Is it *supposed* to take a really long time to load textures?
As a test, I made an all encompassing function that parses a bitmap file, generates a texture and calls glTexImage2D, generates a mipmap and all that. I tried calling this function on every iteration of the program loop, and it ended up taking like 3 whole seconds each time, even when using a power of two texture.

I messed around with opengl a while ago and used a library to load png files, and I really don't remember it taking this long to do anything, so I must be doing something wrong.

what do the grey lines mean?

there's afew things RTS need to get right otherwise your game mechanics better be solid as hell. first off, camera needs to be simple to control and easily reset, no one wants to get all turned around while in a micro intensive game. secondly it needs to sound good and it's look need to be simple to understand just what something is from a glance if you're spend even a moment to second guess your self you've wasted precocious apm. finally you need to make sure there's a decent balance between how fast the game can be and how slow it can be, what I mean by that is no one likes super fast gook clickers but at the same time no one likes super slow snooze fests like gray goo, the only way to really balance this is by looking into your resource management aspect and making sure the player has access to alot of resources but the harvesting of it is just delayed slightly via either harvester travel time or refining those raw resources times.

also if you're going to pull a maelstrom with unit control unless you add coop but only for units like executive assault, it's a feature which will be over looked massively and even looked down on in the RTS community.

I do like what I've been seeing though

Grey lines are just a grid made by the Unity inspector. Green lines are basically a way to make it easier on the game engine when figuring out which tile you're hovering over.

I think he's talking about the grey tiles snaking all the fuck over things.
Height?

Yeah, terrain height.

Post your code, let's see what exactly you are doing. Also, be sure to post things like your specs, GPU, what kind of storage you have, if you have the latest drivers… that should provide a lot of insight into why it's taking so long.

One thing that i'm noticing: why are you calling glTexImage2D to create a mipmapped texture? The proper procedure is here:
khronos.org/registry/OpenGL-Refpages/gl2.1/xhtml/gluBuild2DMipmaps.xml

I have never played micro-heavy rts that I enjoyed.

It's a pretty old laptop that would have been top of the line in its day. 2.4 GHz Core 2 Duo, AMD Radeon Mobility HD 3650, 4 gigs RAM, mechanical hard drive, Windows 7 64-bit. I'm fairly certain I have the latest graphics drivers I can get, given that this chip hasn't been supported since 2013.

As for the code, this will no doubt be embarrassing as hell, but whatever.
GLuint LoadTextureFromBitmap(const char* filename){ FILE* bmp = fopen(filename, "rb"); unsigned int width, height; unsigned char* data; fseek(bmp, 0, SEEK_SET); if(fgetc(bmp) != 'B') { printf("Not a compatible bitmap file!\n"); return -1; } //get width fseek(bmp, 18, SEEK_SET); width = fgetc(bmp); width = width | (fgetc(bmp)

that's why I suggestold him to find a happy middle ground between gook clicker and city builder via sensible resource management

Woops, forgot to mention that "Texcount" is a global variable defined earlier in the file. It pretty much does what you'd think: counts how many textures have been created.

Ok, so I'm already seeing where this is going wrong… look at this:
data = malloc(width*height*5); fseek(bmp, 54, SEEK_SET); int counter = 0; for(int i=0; i

Woah, okay that speeds it up to being able to call the function like 400 times a second. Thanks! I had no idea that would speed it up so drastically.

So basically he's calling a top-heavy function so many times, instead of one batched call to read data?

File streams still scare me, but I suppose it's basically just like reading a byte array until you hit a terminator, in a sense, right?

Computers are optimized for large amounts of data being processed at once. It's generally faster to ask for 1000 bytes once than ask for one byte 1000 times - this is because of many things, for instance, doing that with fgetc is 1000 system calls, while fread is only one system call. System calls are a lot more time consuming than normal functions, because they incur a context switch. Any kind of file I/O requires system calls. Basically the program has to stop and save all of the stack and the registers and then go into kernel space, and then when the operation is done it has to restore that information.

You can read more about context switches here:
en.wikipedia.org/wiki/Context_switch

Here is a benchmark on standard file I/O:
left404.com/2011/03/17/disk-io-in-c-avoid-fgetcfputc/

Yeah, that's how you should be thinking about it.

Sorry, I probably flew off the rails a bit up there. Yeah youre right. Im more likely either 2 without knowing it or 3 than 1 tbh.
a.uguu.se/Yq5cDBhhxej1_Decisive.ogg
heres a thing I made recently.
Feel free to bully me to death

Working on some apartment blocks in Half Life 2. Lightmaps are so comfy, fuck realtime shadows.

I feel my soul being sucked away just looking at it

commiebloc/apartment

What additional things can you do with pre-baked maps, as opposed to dynamic lights?

Good shit good shit

So I made a dummy player object that OnIdle, responds to movement. It sets a target position, and a timer, and looks like it moves until the time elapses. Then it moves. Actually the position is updated at the halfway mark. I did it this way so objects have a fixed, concrete place in the grid, and the interpolation is just a fancy effect, so it's more deterministic.

What's happening is that the OnMoveEnd, when the timer expires, it checks the tile at the object's position, and if it's equal to "1", a green square in this case, it sets it to 0, and updates the score by 15.

The args that get passed are the current GameState object, current Actor, other Actor, and a position. Only the state is guaranteed to pass every time; other things depend on what's happening and what gets passed. But I should be able to model collisions, interactions, that sort of behavior now

It doesn't require the stencil buffer, or any kind of writing to the GPU except for the initial load, so it's faster. If you don't have to calculate the shadow in real-time (like with a static light source), it's an optimization that you should be doing.

You can calculate non-realtime lighting models with lightmaps, which can make the light look better in a scene, however the tradeoff is that it all has to be static light. The source engine however doesn't use more advanced lighting models. Even though you can calculate source engine's shadows in realtime on modern hardware, it's just stupid to do that since the lights are all static.

Are there any light casting techniques that take advantage of say, quad trees or volumetric rendering? Or is it mostly ray tracing? I don't 3D much, or at all, really

Quadtrees can be used as a culling algorithm, so for example instead of saying that, I have a light source and a list of polygons, and I will calculate the light-map by casting rays from the light source onto every polygon in my list, which is obviously very slow, since what if a polygon is behind another polygon? All of those ray-casts are then a waste of resources, since we know all of them will fail.

I can organize the list of polygons into a Quadtree and then use the Quadtree to find a list of polygons, which contains every single polygon in my original that is visible to my light source. And then it's a huge optimization since most light sources can only see a few polygons in a typical game level out of hundreds.

Of course i'm just using the Quadtree as a data structure for a culling algorithm here, and I can replace that with any other type of data structure that can be used for fast culling, such as a BSP tree, and get a similar result. The culling algorithim is not really related to the lighting model and the lighting calculations. I don't know about any lighting models that use quad trees to actually obtain the brightness values.

Volumetric rendering is a form of computer vision, where we usually take sets of 2D data and form a point cloud, and usually interpolate a polygonal model out of it, which can then have regular lighting models and graphics techniques applied to it, but you could also interpolate 3D data following other graphics paradigms out of the dataset, such as ray-traced data, or voxels. But, that's a whole other rabbit hole. Volumetric rendering is a method of collecting and interpolating data to produce useful 3D data, and not an actual rendering (as in, drawing things onto the screen) technique.

I’m glad. As stated, this is a random on-a-whim weekend project, so all you see was done in 3 days – thus it’s extremely early and subject to a lot of changes. I might keep working on it or just let it sit around until I’m done with Beelzebox. It is one of my “dream games”, tho. Tic Tac Wars 1 was actually a failed attempt at it.
Press left alt to rotate. I’ll have to add some double-alt press thing to get it to reset. The way it works now is relatively smooth, though. Kinda like CoH but not 100% copy of it. I wish I had a camera like that back when I was working on LD38.
I’m ashamed to say that sound is one of the last things I ever think of when making game, when in fact it seems to be more important than graphics.
I’ll stick to the desert (maybe I’ll call it Bumfuckistan or something). I’ll try to make some models this week before the coming Ludum Dare. We’ll see if I can achieve the simple but cool look RTSes need.
Amen to that. I was thinking of a WarCraft3- like pace, with hero units gathering experience to make things more varied. I also added a sprint for the units, akin to Universe at War. Double click to get the units to run faster for a period of time.
I’ll take your points into account. I don’t really have any ideas on how to approach this topic yet.
I never really realized that game had that. I gave examples of what inspires me here And yeah, an archon-like cooperative game with some dudes playing as FPS units and some working on the strategies is currently what I’d like to see. I’ll be experimenting with the concepts. Oh, and I really hope to make the unit-control side a bit more fun than most games that tried that.

So you could basically do an octant sweep from a light source, grab all the points in a region, do a quick cull, and form a 3D mesh from them (eg, a heightmap, but turned on its side), and then render it that way?

One thing that always confuses me for lighting is that you rarely seem to have "global" illumination. Eg, you can apply whatever tricks you want for shadows to work, but it's always done in the context of a particular viewer and has to be recalculated for every camera. Or I could be wrong

You would have 3D polygons that you can then calculate light maps for. I don't understand why you'd say "a heightmap, but turned on its side". We don't need to transform anything. Also, we would be using a polygonal mesh generated from the hightmap data.

What lightmapping is, is that it calculates the intensity of "light" at set of sample points on a given polygon, and then generates a texture that represents the lighting of the polygon, and then this texture is blended onto the polygon. The only thing this does is reject polygons that a light source cannot influence. Pic related should explain what I am talking about.

Lighting models are global, it's always done in the context of each light source. However, we treat certain light sources a lot like cameras, in that we generally apply culling algorithms that are the same as the ones that we apply for cameras, since it's the same context of trying to reject surfaces that are invisible.

Here is some further reading from Michael Abrash on the subject: twimgs.com/ddj/abrashblackbook/gpbb68.pdf

I'm finding my self to be really excited for your project and I hope that you keep at it mostly because there was another german dude who was making an RTS that I got interested in but he went back to his home planet or something because I've seen nothing of it other then things I think are it but dunno

I would have to play some kinda build inorder to really tell about your camera but I'm guessing mouse wheel is zoom in/out? you might want to have your hotkeys configurable by the player tbh

it is more important then graphics only because as I've said graphics need to be as simple as possible and absolutely clear on what a unit/building is and is doing, just don't have things be stupid "machine" sounds and you'll be good on that

I guess it really depends on what you want to do with the game but what you could do is have it so that the players spawn each as their own hero unit and then after "pre-early" game they can go into becoming different things, like the one who's the general controlling the base and units the like, one who becomes an army of him self which can get reinforced by the general and then super epic hero unit or some shit honestly it depends on the numbers of players you want in a match

I can tell you right now that randomly generated hidden resources is fucking cancer as hell so don't ever do that nor do randomly generated maps, on that note though maps are very important and if you're abit stuck with figuring them out you can always just make a map tool and let others make them tbh. I don't suggest you follow universe at war design of having every faction their own resource even if that's cool as fuck but thinking about maybe you can have a main resource and a faction specific one with power being what converts main resource like it is in supcom

the issue is that there was really not a whole lot you could do as a controlled unit because with a RTS you need more then just one unit and also why would you just want to be one unit when you could be the whole entity? I think that you just need specific and interesting things a player controlled unit can do, you could even have it so that when a player controls a unit it gives that units abilities or spells it normally wouldn't have

do you have a general setting your game is going to have or is that all up in the air too?

well, this IS /agdg/. missing devs and dead projects are a sad part of life over here. But I am determined to at least try to release some of my games, this might be one of them.
Maybe it's time I go open source on some of my projects and just share the git so some of them might live without my input. I'd have to clean up the code before that, though

when I'll get home, I'll post it. I don't have any qualms against sharing WIP stuff anymore. Feedback is fun. Don't expect it to have control customization yet.

Noted. Tic Tacs are a bit too simple/lazy, though. It’s a standard Unity capsule asset.

I’ll try to test some of the gameplay/mechanics suggestions of yours. I was thinking of having a commander with full RTS control and teammates who would only be allowed to control units through FPS but I’m not sure if that would work. There’s also the option of “assymetrical multiplayer”, where one team is 100% class-based first person and the other 100% RTS (Natural Selection and Nuclear Something come to mind- but is that fun?). Again, a lot of area for testing. I’ll need to add the multiplayer for that soon.

Currently I consider most anything where units can go pew-pew. Maybe Sci-Fi Not-Dune, maybe some near future C&CG middle eastern warfare. I am really fond of Dark Reign 2 as well and was thinking of making some Cyberpunky cityscape setting but making assets/levels would be time consuming as hell. WW2 desert rats are likely out of the question because that shield unit is something I’d like to keep and that’s not WW2 tech.

I wouldn't go open source tbh, just out of principle tbh

I'll wait till it's in a better sate tbh it'd also be better for you as well tbh

well yeah but if you can understand what that unit is from a glance as far away as possible then it would be fine tbh

well I'll tell you that executive assault does that and it is a bad game tbh. the natural selection approach is kinda a cluster fuck though, specially if you handle it where everyone can be the commander and also if you have it literal one side is humans and the other is controlled like zombie master then you'll need to balance it correctly, it could also be like that werewolf custom game for wc3

youtube.com/watch?v=WnK2lSIY0ek

it would help in keeping things simple and easy to spot a cyberpunk cityscape design there's not many rts games with that theme although there is alot of games coming out now with it so I'm going to suggestold you make it like genesis rising with flesh machines tbh

...

You could just swap the R and B bytes instead of using an extension, it's not that hard.

7.62 high calibre
enjoy the joy of a 5 minutes gunfight for over two hours

Granted unless I'm retarded I'll only have to do it all once. But holy shit is it soulsuckingly tedious.

Does anyone knows if this would work after importing to Unity?
I want to split the textures of the body and the face so i can switch face textures to make expressions (im not going to move the face mesh) on my characters.

That's actually what I was going to do at first, but I couldn't get it to work. It either ended up giving me a segfault or the colors would be fucked up. Obviously I was overlooking some retarded thing, but I'm just so tired of looking at this code at this point. I guess I'll fix it at some point.

reminds me when my brother starting making his own UI component library in java, after months he got bored and gave up we don't even use java anymore

Almost done switching over to portal based rendering. Just a few things to fix. FPS gains I've seen vary from around 20% (240->280) to 130% (225->520) depending on area so that's something I guess. I want to move on to level editor and some gameplay stuff already, fucking hell.

I already use Git for source control but I would consider it for projects I wouldn't commit to.

I guess the video makes it look better than it actually is. It is a step in an interesting direction imo, although underwhelming.

now that looks fun. I mentioned the pacing of WC3 earlier, I guess it's still what I will try to aim for.in the current tests.

That's an interesting concept but I personally was never a fan of biomechanical themes/styles other than Giger. Carving machines into flesh - fuck yea. Living spaceships and machines of war? dunno

Gunna fire up blender soon. I'll be making some shitty models but with them I'll introduce the giblet mechanics I had in my LD39 game.

I think I saw that .webm on the wiki first :D
I really wanna see this textured properly

My hurdle with UI is instance management , general design, and rendering.

Eg its a game so polling by clients is probably better than event based updates
What about scope? Sure you can have a root form but should every instance have an ID? Should you track every label instance in a global container manager?
I made a button class that responds to input. Is a scroll bar not a combination of 3 buttons? How does rendering / size affect the clicking?

Juat little shit like that. It really bogs me down

Me too. Hopefully it won't be too long now before I can start putting together a look and feel.

I played it and it's even worse then underwhelming, it's down right offensive how bad it is

well so long as you don't hardlock your self into a straight up clone then that's probably a good idea to keep things moving

other then candle men I got nothing here's my shitty drawing tbh

I don't think that's even possible with the thing I have going

Anyway, this is my half-assed attempt at making a soldier model.
What do you guys think?

The world doesn't stand a chance.

afaik, unity supports up to two uv maps

The camera seems to jerk when you go up the stairs. Is there a way you can make it smoother?

Sure, I'd been meaning to improve step ups since the instant height adjustment was annoying.

That looks a lot better. I was going to post that I came, but you beat me to it in the video.

>the fix is on the wiki, three fucking lines of code developer.valvesoftware.com/wiki/Ambient_generic:_stop_and_toggle_fix

I added mouse input routines to the engine, but my matrices are off, hence the distortions.

WEW

thanks valve

Autumn now has incredibly crude animations for each direction, instead of a single static image

OH SHIT ITS STARTING TO LOOK LIKE A GAME NOW

...

best feel

tried github.com/fruxo/turbobadger yet?
No exceptions required, no RTTI, no STL. opengl. pretty small and fast. fully skinnable.

I wrote my own too in straight C, opengl batched and including distance field font rendering. It's a pain in the ass. It's also a solved problem so just use a decent library or MIT or BSD license codebase.

glTexImage2D creates a mutable texture object. So this means many many runtime checks. This is going to be slower. Use glTexStorage2D instead.

glTexStorage2D() to create the texture object in opengl & glTexSubImage2D() to copy the data in.

...

Wow disregard, I made SFXR export as 16-bit rather than 8-bit and it worked just fine.

Really gotta refactor some stuff again. This sort of querying/invocation is just too fucking much

we rigging today

user, you're giving me too much credit. I'm whining about creating this shit in UE4. It's much less work from a technical standpoint, but it's still soulcrushing when implementing new functionality.

help
i think i'm gonna redo the animation from scratch, i only had reference images from the side so i couldn't get arm distance correct and i can't tell when he's rotating or not

the shield guy should use a flame thrower

nah special flamer guy with crit tanks on his back

So I'm moving pretty fast with my Morrowind-inspired UI. I'm actually really fucking happy about this. Most basic functionality is there, and now all I've got to do is actually put together an interface and then I can start blocking out the gameplay I've got written on my GDD. Progress feels fucking great.

Anybody here ever make a guided missile mod for quake 3?

its a hw assignment and our class is pretty lost

Just implement these simple steps in whatever Quake's programming looks like.

Obviously just pick a target by taking all of the players positions and then choosing the nearest one, then bias the velocity by adding new velocity vectors which are created by taking the position of the missile, the enemy player, subtracting those with each other, normalizing the result of that subtraction, and then multiplying that by the magnitude of the bias you want. Then, add this velocity vector to the velocity vector of the rocket.

My one-way port, versus SubTerra and Supaplex's

Okay here's the full set

look up proportional navigation, it's a lot easier than this webm makes it out to be

Nevermind, those were garbage. Much better ones here I think. Gonna redo the metal gradient on the big pipes, though

Colors were flat and shit, made it better again. Hopefully done

I don't know, I'm considering quitting. But I don't know what else to do.

Underrated post.

It's even flatter now. And kinda pillow-shaded.

hue shift the colours to make it less gay, sine all the shades are just green, add a hint of blue to your grays but keep the gray so it's not gay

I'm on a limited palette and don't have a blue ramp with that many shades

don't get down! I think that may actually be harder.. from what I've seen of UE4 user interface stuff. Good luck and god speed.

This is the slowest I've seen the /agdg/ thread in a long time. It's been nearly three days and we're not even half way to the bump limit. Where is everyone?
before you ask, i'm working on stuff but it's a big chunk of work so it might take a couple more threads.

Black Friday, busy with work, holidays are coming. I work for the next week straight.
Less shitposting, too.

I've been spending the entire weekend researching and buying parts for my first PC build. All the parts should be in by Friday hopefully.
Also I'm a nodev so I can't really contribute.

...

Shadow rays!
I'm going to work on k-d trees soon, and hopefully remove a normalization in my main rendering loop.

I have to write an implementation of malloc for homework. The teacher was using a doubly linked list, but I think that those are gay so I'm making a singly linked list of binary heaps to manage memory.

That makes sense. I forgot that burgers go nuts this time of year. We have semi-sales but they pale in comparison to the upcoming pre-Christmas sales.

user, I failed an entire programming class in 10th grade, because the teacher was a cunt and i'm retarded, but 3/4 of the class failed
Don't feel bad, though.

Yeah but this is a higher level(3000) college course. And normally I don't have a problem but this class is just a huge pain in the ass. I feel shafted, I hate C++ now and I know its not the language's fault but that the class is just being taught poorly and has a giant workload.

Oh boy, what a wonderful time that was. I passed all them with a 90+ and basically flatlined when I hit college. Still graduated with a 2 year diploma after 4 years (played WoW:WOTLK and failed a semester, another time I had issues with co-op placement, because I couldn't get to (((Toronto))) or work for free, and another intake I couldn't afford a book). Debt is paid off, though.

I think at best, I managed to stumble onto OOP stuff online and copy pasted 200 lines of shit that made some particle system, but I didn't understand classes, instancing, or pointers, so I couldn't do much with it.

I did make a shitty "dodge the falling circles" game and people passed it around so that was kinda cool. I made it spam new folders in their student docs folder with random gibberish names. Only one person got that particular version and I had to stay 10 minutes after class to remove the folders. Good times.

I made some shitty maymay game about cubes chasing each other, with Funkytown playing in the background. Putin and Jesus were there, and God playing Tetris. Made a sequel later which was a platformer, and people loved it all for some reason. I know it's imbecilic, trust me, but I was basically high all the time due to antidepressants, so it seemed like a good idea.

still the most successful games in my career so far

Damn, I need to stop being a little bitch. Thanks for posting this, it was an unintentional wakeup slap. I've never failed a class before and I'm acting like a fag for just struggling in one.
Its so weird hearing that you guys had programming classes in high school though. I went to high school in a higher end southern california area(thankfully there no longer) and we never had programming classes, I just learned on my own starting in middle school with Game Maker. I would have expected the area I lived in to be cutting edge in that regard but I guess not.

When I was 10 or 12, I was dicking around with some program called RPG Toolkit, and even knew some basic HTML. I couldn't do anything fancy, but the logic wasn't entirely foreign. College was mostly useless for me, technically, and most of my professional knowledge came from dicking around after that.

For example, I had a co-op in college that used XNA. I had a miserable time contributing to their studio, and I don't even want to associate with them, because they did absolutely every piece of work on their games, not me, but that's probably why I still use it.

I didn't even know what the fuck generic types were, and my classes never told me. I had to learn about those and dictionaries by modding fucking Terraria with tConfig for fuck sakes

(checked)
fuck it, sage because blogpost
End of fall semester last year (my 3rd year of college) I had a bad depressive episode and failed every course I was enrolled in. Didn't attend classes, didn't attend finals, called the suicide hotline, was nearly suspended.
Spring semester, was kicked out of my major (Computer Engineering) and placed in a part-time probation where they give you remedial courses on academic skills and motivation, retook some courses, did really well. Head of my major says he'll accept me in on the condition that I find an internship for fall semester with a 2.1 GPA, head of the math department says they'll take me in if I don't get back to my own major.
End of the summer, right before I'm supposed to go back to school, mom pulls some strings with old co-workers who work at a small software company in a small industry which I won't describe because if I do I'll be doxxed. I talk with mom's friend, he sets up an interview, I go in and talk with my future bosses. The interview has a few dozen red flags, the largest two being that I never write so much as a line of code and they never describe any particular task they'd like me to do, just shit like (direct quote) "Assist the software engineering group with code development and testing", but they send me an offer to start next week. I'm incredibly hesitant, and consider moving into math even though it'll set my graduation back two years, but I accept.
The job was worse than I feared. I wasn't even given root access to my own machine, but was expected to install a dozen programs on the list my boss provided by myself with a temp "admin" account that didn't even have full admin permissions. Sorting that out took weeks. Testing the software was done entirely by hand with no automation whatsoever (!) using test cases written a decade ago, a third of which are so vague as to be meaningless and another third of which are incorrect due to code changes and were never edited because "it's always been that way user, everyone knows that." They have an Indian team because of course they do, and I learned at the end of October that said Indian team literally hadn't pushed any changes to the repository for months because of course they hadn't. When I learned about this, I went to my boss and told him that the reason we were missing deadlines and losing money was because we weren't getting basic shit like this right, and I asked him to push for people to do basic things like "push changes to the repository" and "check more than once a month that the codebase can generate a stable build" and he said that this wouldn't happen in the future.
And two weeks ago, after months of misery and despair, I got fired from the internship for, and I quote, "having a negative attitude." Which was frankly correct, I hated every second of that job and they hated me. The thing is that after I got fired, I could consiously feel pic related happening to me and eating my mind. I've been like this for two weeks. Thinking about this stuff is poison. Researching and buying parts for the new PC at least gave me something else to focus on, but now that I'm waiting for the parts it's starting again.
Thanksgiving was pretty bad. Not only did I have to explain to my family at Thanksgiving that I'd been fired, but my aunt came over for Thanksgiving dinner, was more of a stuck-up bitch than usual, and made my mom and imouto cry. Visiting my dad's side of the family went a lot better though (unlike my aunt they're blue-collar folk, which I'm fairly certain is the Elitist word for "human"), and I haven't been kicked out of my major yet.
The worst part about everything right now is that despite constant bullying throughout my teen years, attempting suicide in 11th grade, and getting turned down for prom 6 times (the last of which was a tentative acceptance until she got concert tickets literally the day day before bids were due) in senior year, I'm fairly certain I peaked in high school.
No matter how bad things are, just remember it could always be worse. I'm thankful that my dad has a job and we can eat, I'm thankful our mortgage is paid off, I'm thankful I have an internet connection that isn't total dogshit, but most importantly I'm thankful for my family who are willing to support me through all my worthless fuckups. I don't know what I'd do without them.

Hi Guys,

Can you guys give me some recommendations for OpenGL Tutorials? I've been usin SFML to handle my sprites and shit but I want to expand to stuff like dynamic lighting / shading and texturing, which from what I've seen gets heavy into open gl objects n shit.

Most of the shit is outdated or some video of pajeet yappin at me in his almost incomprehensible broken english.

Also if any of you boys got a good recommendation for top down zelda-esque games with good physics I can use for reference I would appreciate it.

Whats wrong with "outdated" OpenGL? Do you really want to require a modern GPU to run SFML-tier sprite graphics?

Do you really think that those graphics require anything more than OpenGL 1.1? Because they really dont. This is a game running on OpenGL 1. something, and it looks great, before you tell me that older OpenGL cant do good graphics.

And before anyone mentions shaders, I want to point out that 99% of the time you do not have a problem that has a soution (shaders), you instead have a solution (shaders) that you are then inventing a problem for so that you can justify using them.

I know that I didn't mention OpenGL tutorials, I think that:
nehe.gamedev.net/

is a good place to start. Beyond that you should be able to read official khronos documentation and read the specifications for whatever extensions you want or need.

godotengine.org/article/enjin-coin-supports-godot-and-announces-upcoming-module
archive.is/Y5UTG
what the fuck?

Just some memecoin no one will use.

Which reminds me, I really need to make some meme tools for Unity so I can passively earn a few hundred USD/mo

>>>/africa/
>>>/ebon/
>>>/scat/

I am still "trying" to learn python somehow. Man it feels more like a chore then anything else, I finished 1 book already and the next one idk I have a bit hard time focusing on it.

So instead of 4-5 512x512 textures, I think I might be able to shove everything into one 1024x1024 texture. Is there a disadvantage to doing this? I know I'd save on texture swaps when drawing my tiles, but with less than 200 visible tiles to begin with, I don't think it'd matter much even if I did it the improper way.

Is there any sort of insidious graphics shit that might catch me, like dynamic texture compression, versus pre-compiling my assets?

Wasn't mentioning it because the post is for "Modern OpenGL" fags

Yes, older graphics cards wont support this. You have to query glGetIntegerv with GL_MAX_TEXTURE_SIZE so that you know that the card supports textures that big. So, you're gaining speed that you don't actually need at the cost of compatiblity. Granted, these some statistics on the average capabilities of people's graphics cards… only 1% had a maximum texture size of 1024x1024 and nobody reported a lower score. But, I think a voodoo or something wont take it.

feedback.wildfiregames.com/report/opengl/feature/GL_MAX_TEXTURE_SIZE

XNA's Reach profile says that it supports up to 1k textures, but I guess thats the upper limit for that mode. I guess there's no harm in a few smaller textures, then. Thanks

C++ is actually that bad. Learn with C# or java instead.

C++ is poor but C# and Java are just cancer langauges, you shouldn't consider infecting your mind with them

C++ is cancer.
C#/Java are only cancer if you fall for the OOP meme.

C++ has the merit that you can ignore the bad features (until you have to read someone elses shitty cancerous code)

C# and Java don't even give you that grace. Also, what is programming without pointers and goto statements?

...

C# has traditional and smart pointers as well as goto statements. They are frequently unused (except in specific interop apps and contexts) because the language itself has syntactic sugar over many common uses of them. For example, exception handling and try/catch blocks utilize gotos in the CLI language. The expensive part of event handling is unwinding the stack but otherwise behaves as youd expect

Sure, you have those features but I don't want to type unsafe blocks over literally everything to do normal programming. Also, you still cant escape OOP cancer.

that would make him OP as hell. The very fact your units might fuck themselves up by shooting at him is enough, I think. Snipers seem especially vulnerable to that. Also, it's fun to go first person on the shielder and try to aim the ricochets back at the shooter or at his friends

I haven't done much work on the project but currently I want to animate Sam. Next, after I'm done with rocketeers/bazookoids, I'll add a flamer guy. He'll need to do hell of a lot of damage as payoff for the crit tanks.

I was thinking the same thing. Well, there should be a new thread for the Ludum Dare Jam so I hope there will be some progress shown.
Other than that, as said earler, this seems to be a busy week for a lot of people, including myself. Work on the terrorist model took me maybe 3 hours and I haven't touched the project since then.

Don't try this at home. Professional ideafags in a closed course.
This is the Ludum Dare group

All individual ideas are equal, i.e. worth nothing. It's how you put them together that matters. You start at the core and build up by coming up with ideas that fit the current need. You do not outline a fucking game by throwing down a bunch of random shit. I'm really fucking tired of people who think they're hot shit just because they come up with random quirky shit. You gotta make it work. A completely stupid, boring idea can be turned into a fun, challenging and rewarding game if you know how to work up from that foundation - case in point Getting Over It. And most often, you don't need a quirky idea, just a precise, uncompromising execution.

/rant

...

Nigger it's spitballing ideas for the Ludum Dare themes. They don't have time for good execution, they just have to quickly throw something together.

So decide on a single theme and then spitball ideas in a set direction

Are you fucking retarded? The theme hasn't been announced yet, those are the ones it might be.

that's not how LD works, my man.
ldjam.com/events/ludum-dare/40/theme

A week before the event, there is a voting going on. First some 3000 themes are slaughtered over the course of a few days. You are left with 48 - these get voted on in 3 rounds.

The final round is 16 themes you saw in that post. Out of the 16 you'll end up with one that will be the event's official theme.

You only learn the theme the moment the event starts.
I will be brainstorming with my codefriends then but I guess some people don't want to waste any time and will have their game ideas ready the moment the event starts.

As of the ideas themselves

most seem to be jokes more than anything else. Outside a few, I would find them hard to realize over the course of 48/72 hours.

Alright, that makes sense. I lacked the frame of reference for that picture.

Yeah sure, but that's how brainstorming works. It doesn't hurt to write down any associations and ideas that the themes bring to mind in the early stages of conceptualization. There's always a slight chance a retarded joke grows into a more reasonable idea.

that's absolutely true. Even the proposals from the picture are already brewing in my head, perhaps out of them will evolve something new and fun for the jam.

OK anons, I need some advice on which engine is best. I'm currently debating between Unreal Engine 4, CryEngine and Amazon's Lumberjack.

The game in question is a sword and sorcery FPS. CryEngine is supposedly the best for FPS games but since I won't have guns but a melee slash and bash system, I'm not so sure.

Also, which one works best with importing art assets.

There is no such thing as objectively best engine, it all depends on what you prefer. Anyways:
If you want to sell your soul to tencent chinks and can use Visual Studio without vomiting, pick this.
Pretty ok but barely tutorials or learning stuff available as far as i've seen
A meme

C# does have limited functional programming support iirc.

I have to personally vouch for how fluid UE4 is when you're doing standard shit. I've implemented a responsive, good looking, ergonomic UI in no time just by intuitively poking around in the editor. It gets a little more difficult if you're breaking the mold, but for standard stuff? Couldn't be more fucking handy.

That is, after you suffer through a month long learning curve.

gamedev fag here. Sorry for my rant, but I need this out of my system. A lot of folk, some anons on Holla Forums too, say the following

Why would I make an entire game from scratch except if I wanted to do that for the sake of it for years on end? This includes making a memory manager that probably crashes and leaks memory all the time, a scene manager that loses its references all the time and crashes and so on - and that's if I use SDL and/or OpenGL like suggested. Why shouldn't I be spending that time with the game logic with a proper debugger, memory leak cleaner (yeye always make sure it doesn't happen, but at least the world won't end) and exceptions catcher? And yes, pixelated games can be made in Unity too, no problems. I can make a game from scratch, done so many times, but WHY would I do it?

There are a lot of problems with Unity - for example, there are tools in the Asset Store that do the job better than Unity itself, and I puzzle why they just don't buy them up. This especially goes for the "better coroutines" asset. And SJW'ism is irritating as well.

Other than that, its always the little devs fault, not Unity's: for example, lighting can nowadays be easily made into Unreal quality with the free camera effects and some tweaks in the graphics/quality settings most don't even know about. Unity is so easy to get in that it is overwhelmed with idiots who use standard shaders and objects, and only opened Blender once and made an object without unwrapping it and called it a day.

Really, if you just get the Dialogue Manager and PlayMaker off the asset store and made something, you can make a 100x bigger game than you do from scratch, and you can actually focus more on gameplay, how things are told to the player and making it fun. And this is coming from someone who made his own texture-, shader- and model loaders.

God damn, does that mean there will be a micro transaction support on Godot soon?

Assuming those memecoins would catch on for some reason or option of paying with real money - why'd you care whether there's support for micro-transaction in Gotod?

Because this is Holla Forums and people think in memes and buzzwords, user. People hear the dreaded word and whatever is in association with that word is instantly evil.

You're just gonna be alienating your potential audience if your pixelshit game requires a high end gaymen PC to run at all, which will invariably be the case with bloatware like Unity.

I don't make pixelshit games, but I get your point; even then, you can easily optimize, like with the render pipeline.

because unity is free, and the assets usually are not. unity profits off the asset store by taking a cut every sale. it would be a poor business decision to do so.
they usually halt production on promised features if a popular asset comes out for that feature. they do this a lot.

Hopefully they don't fuck with development.

Posted this in the DOOM thread, but may as well also post it here too!

This is the shop area in MoM which also act as a central teleport hub, connecting to all the major areas. Each Pocket Dimension for each shop keeper is based on the element of the area, which also means they specialize in that elemental weapon and magic type!

As per usual, questions are always welcome, etc

Textured Body, now to make the face texture, i hope it exports to Unity correctly…

Why does that weapon cover 1/4th of the screen. Why not do it the Doom way and put it in middle and keeping it a decent size.

its a big gun

Why is that? You've caught my curiosity, user.

____4U____

but no, seriously, let me address this properly. I wasn't actually kidding with the first reply; 2 of the weapons tend to cover up a bit of the screen, being the minigun and magnum. They don't look correct when smaller compared to the Shotgun, even though I see what you mean. I think I've gotten use to it myself, but this does go to show I should probably release my demo soon so I can get more feed back and questions like this.

Art is learned through constant repetition and improvement. You can't learn how to draw well in a few months, for example, while I've learned CPP enough to do pretty much anything in two. It's actually a big problem for me because I lack the persistence to keep practicing art.

Basically, when you're learning programming you're teaching yourself how to do something. When you're learning how to draw, you're disciplining your brain to achieve the result you want by chiseling your skills down with raw practice. I personally respect art as a skill much more than programming for that reason; Good art is evidence of perseverance.

Right, functional programming, but not procedural programming. You can't even make normal functions in C#.

I'll point out that i'm actually doing exactly what you're talking about with "making your own engine". I have a custom engine and a custom map editor, and it's all made from scratch.

SDL and older versions of OpenGL make faster development possible - but any incursion on modern graphics programming, which is a set of seriously complicated API's designed exclusively to be used by large corporate "graphics programming" teams that can manage writing so much code and still getting something. My program's code size is several thousand lines longer than it would normally be since I have Vulkan support in my engine. My OpenGL code is only 675 lines since OpenGL 1.1 with extensions is pretty easy to write, but my Vulkan code is 2615 lines of code- and that doesn't even describe how fast I can write the code. I can write OpenGL code about twice as fast since the API is much more intuitive to use.

But if you're doing pixelshit and not any sort of complex engine OpenGL 1.1 can get some great fast results.

You could try taking a piece of music that you already know and use that as a baseline for a sound that you want. Just keep altering and changing bits of the piece, like play around with the melody or throw in a different chord progression. Music is made up of all the same fundamentals anyway; you just have to get creative for how you use them. If you want some good examples of this check out SilvaGunners vids, this autist can basically turn any song into the Flinstones theme and Love Live!, or anything else.

As far as modeling a human body, you'd be farther ahead to learn how to draw proper anatomy. If you can draw a torso and the associated muscles, it isn't much effort to 3D model it, apart from having good topology. Not sure what people do for clothes tbh.

this is a meme for a reason

cloth sims. cloth sims are fucking incredible.

What meme? I'm out of the loop on the memery.

google.com.br/search?q=cs graduate code meme&tbm=isch

I fucking love it

making an anime-girls+tanks game. whatchu niggers think?

What sort of negative influence could those companies even have on Godot ? It's not a game so they won't add loot crates or something. Monthly fee or royalties for users are also out of the question as these companies wouldn't benefit from it.

Fixed. While learning how to do good enough art is indeed way harder than learning good enough programming, you're understating how hard it is to learn the basics. Unless you had some previous experience with doing mods, scripting or something similar, I don't believe you could learn and understand what're doing in 2 months.

I rike it. Looks professional.

It's not that difficult. I just hammered the topic for 60-ish days straight - you'd have to be a fucking neanderthal not to learn programming at least moderately well in that time.

The shader is shit, but I don't hate the concept

nice

excellent. Post more

Programming isn't like art where you can simply look at it and gauge how accurately you drew something that is in front of you. Especially if its something big(ish) like writing a decently complex solo game.
You can get something to work even if the code's architecture is complete shit and everything is all over the place. I find the more complex something is the worse the implementation gets. I could for example write really small programs that are perfect in design, but once it goes beyond that its a linear path to disorganization and hacky code.

This is the most retarded hatch I've ever seen.

Post gameplay my guy, i would like to see.

You just strive to write clean and concise, and avoid repeating yourself. Bonus points for formatting the code so the relevant bits stick together and comment your autism for the future generations. Once you learn what's considered bad code, you'll be able to recognize bad code. I mean seriously, it's not fucking rocket science.

Programming is complex, not difficult. Drawing is both difficult and complex.

You're fooling yourself if you think programming isn't difficult. It depends on what you're doing. You can write an easy program just like you can draw a stick figure, but to pretend it's easy to program something no one has done or is too different such that you will have little help is both complex and difficult.

ragdoll "works"

haha double UV worked!

select all and do smooth shading, user.

Shader needs work, that black outline especially feels way too big, and the hatch is fucking retarded. Apart from that I'm interested, post more.

Lewd.

jup, pretty much. You just need to know what you're doing and you'll get far better results.

here's another pic of anime grills + tanks gaem, dont have gameplay stuff available right now, because i cant be bothered :^)

Divide to conquer is key to implementing larger scale software. It's why UML exists.

A little food for thought for you people. Sorry for the inevitable blogposting.

Just now I was talking with a friend of mine who is dedicating some of his time in collecting his all-time favorite vidya in physical form, after he told me that, it made me think.

Imagine your dream game, user, I don't know if that's the one you are working now or you are planning to work later, but put him on your mind now, and answer to yourself: Do you think anyone would want to have it in a disc form, sheltered in a case? Or maybe even a cartridge or flash drive? I'm just commenting on this because this whole digital download phenomenon made things pretty disposable, making shitty/meme games is getting almost indistinguishable from "normal" games to the majority of the public, but think about it, that only happens because of how the transaction model got so easy and invisible, giving 1 or 10 dollars to some faceless dev and not even seeing your money go away, it's so numbing that lots of people don't even play the game sometimes, the whole "fun" was the payment after all.

But in physical form, that's another story. It's not that cheap and even if you don't buy it in a store and see your money going away with your own eyes, the delivery time will already kill the instant gratification of clicking that purchase button on a whim. The product has more value now, it'll make you considerate your choices, you wouldn't buy a trendy indie on a case or some cash-grabbing "always online" e-sport thing. You would buy something you really like, something that transcends time, a great singleplayer experience, if it has multiplayer it has bots and/or a split-screen, something that you could handle to your grandson and he would enjoy too, something to beat at least one time in year or two.

Most of what's being created now will be forgotten, but that already happens to most of things in entertainment, the novelty here is that lots of them doesn't even belong to the users who bought them, their futures dependent of the digital distributions services that maintain them on their clouds, even modern games with physical editions would fall short to older legends for their dependencies on day one updates and such.

So tell me user, is your dream game worth being carefully stored at someone's shelf or will it be tossed to the sea of digital pastimes, more populated with things that you never know than things that you love?

Tl;DR: One fine way to know if your game is worth it, is trying to imagine if anyone would pay for it in a physical form and maybe preserve it to generations to come.

A lot of cartridged stored games weren't even worth the plastic, let alone the silicon.

I imagine people didn't want to stay with those for long. But today those could become meme games quite easily, just a matter of causality.

original idea please do not steal

kongregate.com/games/i_smell/no-time-to-explain

anyway from what I see on youtube that game still allows walking and it's a somewhat different gun/game. thanks for the heads-up though.
I wouldn't be surprised if there's another game somewhere out there even more similar to mine

I'd have more to post if the world cut me some slack this week.

Looks like it could be fun.

That looks way better at that distance. Is it that the black border is always the same size? Maybe you should have it scale depending on distance a bit. In this image it looks good, while in the closeup shot it looks way too thick.

i actually LITERALLY JUST NOW made the outline based on camera distance to object, so it's a more-or-less consistent outline, relative to screenspace, but stops scaling at some point as you get further away so the whole object doesnt get covered in black shite. as a result, the outline is also a lot smaller when you're very close to the object.

image related, compare outline of tank + terrain in the background to the church outline.

I enjoy the macabre

hello fellow polebro, jak sie mash?

dobrze, thank you. I think there might be more of us here

Releasing my games on a CD or something is an idea that I like to think about, I hope one day I do get to do it.

I think about half of the people on /agdg/ are polish, or at least a third.

Yeah that looks good, what's the gameplay like anyway? Do you control the tank itself?

Trying to make him walk on 90 degree walls and having some trouble.
At the moment I'm rotating his collider depending on the normal of a raycast that is going straight down. As that happens, gravity is also changed to be applied always straight down even when he's rotated. (red vector pointing down)
The movement (the red vector pointing forwards) is rotated depending on the normal.

This seems to work for very steep diagonal planes, however as soon as he has to go 90 degrees on the wall he just spazzes out.

Is there a more simple way of doing this that I'm missing? I can't think of any more elegant solutions, but it's pretty clear rotating the collider itself seems to be a bad idea

Does anyone know how hard is it to get your game published in physical form? I would love my game to get published through something like IndieBox.

What now?
Should I give up on metatiles and just work with 16x16 tiles?
Is there a map editor that can do metatiles that would be more suitable for my project?
Or am I going to have to spend tons of time writing my own map editor?
Or should I just create maps by hand with a hex editor?
Or do I write a tool to generate metatiles automatically from a tileset by looking for matching 8x8 blocks of pixels?

I think OpenGL 4 Shading Language Cookbook is a pretty good place to start. Using modern OpenGL means writing shaders or at least knowing how to use shaders. The book also comes along with some pretty decent boiler plate code. SFML is fine too. GLFW is also a good library for handling OpenGL windows and contexts.

i quit game dev because I couldn't into path finding in Godot. When I asked they gave me "tutorials" that I didn't even see were tutorials and there weren't any alternatives. Should I try again or am I too retarded for this?

What are metatiles? I've never touched Tiled

You are not a good programmer in C or C++. You don't know the magnitude of what you do not know yet. You aren't able to accurately gauge your experience or competence because you are not skilled enough. Unlike art, it is easy to misjudge your ability. People unskilled in art can look at it and reliably evaluate it. Not the same for coding.

Yeah, you can learn enough in 2 months to get by. You won't find your missteps until later. It will be workman-like, cluddgy, slow, won't scale, and be absolutely filled with bugs you are not even equipped to find yet. but it work. on your machine. most of the time.

You can get that far with drawing and painting in 60 days too.

Metatiles are when you take a single tile in a tilemap and actually build that tile using a set number of smaller graphics chunks.
This allowed them to save lots and lots of memory, see pic related.

what's wrong with just applying A* to godot?

Is that literally not just a tile map or byte[]? This seems easily fixed

the Pokecrystal Disassembly project has a visual map editor, dunno if you can use it for other things though

It's hard to say what's wrong without actually seeing the code, but that's pretty much the same way I do it in my game, so your solution should work in theory.

To rotate my character I'm actually crossing the normal with my forward vector, which I think is what is giving me problems.

Is there any other way to do this? I tried some stuff with Quaternion.eulerangles(normal) but it just doesn't seem to work at all

Essentially you just want to set his up direction to the normal and his forward direction to the movement vector. I'm not too familiar with unity but I think it would be something like

transform.forward = movemet_Vector
transform.up = normal

what

I didn't even know you could do that, up until now I had been changing the rotation using quaternions

However for some reason I can't seem to apply both a rotation to my forward and to my up at the same time

Happy to see you're still progressing, MoM user. Keep up the good work!

Made some infinitely expanding walls
I like the first vertical one the most, but if I have an off-center light source, it makes it really hard to figure out the lighting on the 4-way on

Here's how a sample level might look once expanded

I decided to check the unity documentation, and I think this might work.

transform.rotation = Quaternion.LookRotation(MovementVector, hit.normal);

Yeah that's what I was doing before, only problem is that it only rotates if I'm walking up or down the slope.
If I turn sideways on the slope it just goes back to zero rotation.

Dunno I'll just go ask on the forums tomorrow I guess

Here's my expanding walls, plus the old and new one-way ports. I made them visually connect, even though they cannot combine in any way or interact mechanically.

Is it possible the movement vector is being set to zero? Or maybe the raytrace isn't hitting and returning zero. Other than that I'm out of ideas.

Yeah I already have a temporary movement vector where I store it to do the rotation.

Don't worry user, thanks for the help either way

Here's all my non-animated things so far.
Pls rate

The following items will be redone: Yoshi egg, ice cube, the 3 gems, the ? mark should be a haste item, keys need to be bigger, heart and static block are placeholders for a "match 3" style of pushable block, top doors need minor adjustment, bottom row of buttons needs to be redone as well.

I've been working on scaling back my shitty gradients into NES colors (I took the 64 color palette, shrunk it to 48 non greyscale colors, then adding a bunch of ramps to arrive at my final palette). Now, I've been trying to use that NES root exclusively, except for gradients in strategic spots (eg the white striped floor or the pipes)

I remember i had to go through the same shit when i made my F-Zero clone, i don't remember what exactly the problem was but it has something to do with how unity rotations goes back to 0 after you hit a threshold or something which makes it go retard. You can fix that with Vector3.Cross i think.

>pastebin.com/8G2JEvRd
Anyways, here is my full "rayhover.cs" it's not clean but it does the job. Make a cube or something, put the script on it and put it somewhere above the ground. You should be able to drive around in your map with WASD. Copy the needed stuff from it to your own script once you understand how it works.

Why is the outline based on camera distance? Shouldn't it be based on perpendicular to camera instead?

Is the threshold perhaps 6.2850? That's 2pi, or a full rotation in radians

Fuck, there goes my whole argument. Fuck this gay world.

This seems almost exactly what I was doing before.
I'll give it a shot and see if it works for me

thank you

Backface-extrusion outlines look best when the extrusion distance is scaled with distance so distant outlines are smaller, not larger.
It basically looks like you took the default unity toon shader and slapped it on everything.

t. wrote my own toon shader

Did you do the models yourself? They look good, my dude.

I just found free assets floating around the web and on the unity asset store, and slapped my shaders on top of them, I'm a programmer not an artist.

Holy fuck I'm an absolute retard
You can use Quaternion.Lookrotation(forward, upwards) to set both rotations at once

How did I not notice this

Also pic related seems to be giving me problems

I'm taking my movement vector and crossing it with the normal of the slope under the character to rotate the vector itself depending on the slope.

Is this the right way to do it?

New subweapon, fire bombs have a smaller splash radius but these hit harder than any other, detonate on impact, and fly straight infinitely.

That shit needs some fire particles

someone let me idea fag for them tbh

you can do it for free

I'll do you for free

I've been thinking about doing something for NES, but I'm not much of a programmer. I can either learn the actual assembly language (which is apparently better for making bigger games – my ultimate goal is to create an RPG a bit like Ultima IV), or I can learn C and then put all the shit into a compiler. My issue is I'm not really sure if I'm starting in the right place. I have never programmed anything other than very simple text-based .bat file action/adventure RPGs (multiple projects with different mechanics, all unfinished but only ever because I'm very visually-oriented and lose interest easily). To get to the point quickly without rambling, I dunno if I should get intimate with NES programming through experiments on my own, or if I should just try to make a ROMhack of Ultima IV and learn by example.

– Or maybe I'm better off learning C on its own, doing more text-based stuff and then getting to NES/6502 assembly later?

Many are free user. For example:
assetstore.unity3d.com/en/#!/content/54975

assetstore.unity3d.com/en/#!/content/84

assetstore.unity3d.com/en/#!/content/27676

If you're going through the trouble of learning programming, you should really develop skills that can be used for financial gain in the current marketplace.

If you're going through the trouble of learning programming, you should really develop skills that can be used for financial gain in the current marketplace.

Learn C and maybe do DOS programming, this will prepare you for more difficult systems like the NES.

Once you know C a little bit, you should try making some games for DOS, this is better as a stepping stone to learning NES programming since DOS requires you to be somewhat comfortable with assembly language. However anything on an IBM PC is not going to be like developing for a console, since PC's use a framebuffer and the NES uses a PPU (which is like a 2d GPU, but the term GPU didn't exist) but, it's still more applicable to NES programming than modern windows/linux programming.

Sounds good. I have some books on DOS that just sort of came into my possession. I also have a book on 8085 assembly language, though I'm not sure just how much it would differ to 6502. If it isn't too different, I might be able to at least gain some general knowledge on how assembly language works.

I learned to paint miniatures and I don't sell my services there, either. Also

6502 is pretty different from x86. But the idea is to become comfortable with programming, comfortable with the ideas of assembly language, and comfortable with working on the level of abstraction that an NES is on, by working in DOS. Since, in both the NES and DOS, you don't call a library, you just ask the CPU directly , or you write to a location in memory, etc, etc to do things, so it's a more comparable mindset.

I thought I had to be able to make shaders in order to be a 3D modelfag

I'd sure like it if the artists I've worked with could do shader code.
Visual material editors are not programming, and are serverly limited in what they can do efficiently.

If I was able to code i wouldn't be bothering with 3D models to begin with tbh

I would probably be doing modeling/animation if I could, I can't do that stuff, but I can code pretty well.

Code in is part of what's used to draw the bars in the attached image. Going with a shader to draw the bars highly simplifies the needed UI code; all I need to do is pipe in a "_Fill" and "_Delta" attribute to tell it how much to be filled, and the shader code handles everything else.

I'm glad it was of use

Artfags can't into shadermagic for the same reason programfags art is horrible. People who are equally good in both art and science are rare, or rather people who have the discipline or fun in doing so.

wew Update. Here are some legless sprites, I hate working on the running animations but they are almost to testing. After that I will work on adding arms then drawing jumping sprites, which will be easy as shit because now I have a ton of sprite legs (arms soon) to rip and move around.

Nice model, would like to see shaded.

I dig it, keep it up.

No offense, but I've been seeing you trace these for months now. If it's too much work, you should probably go for a simpler aesthetic.

Oh no, I just do them off and on. And I am getting better at drawing, or at least I think so. These don't look that bad, do they? They're just a template. I also haven't really posted anything in about a month or two.

Not the same user but
Theres not exactly much to them. They look so early in development that the best anybody can say is "You're probably doing good so far".
The motion certainly looks fluid enough, but thats probably not saying much to your ability since you're just rotoscoping after all.

Essentially what said. At the pace you've been going you haven't been nor are going to make any considerable progress. I don't notice much of a change compared to a few months ago.
If you're starting out I'd recommend a simpler style, such as NES limitations. Fewer colors to worry about (like 4) and fewer frames required to imply movement, which both result in content being made faster. You also learn the fundamentals of sprite art better this way.

I have awful handwriting.
Despite writing every day and trying to be neat it never gets better. Repettition doesnt seem to help me.

I cant draw. If I followed your advice and drew every day I would see no improvement just like my handwriting. And it would just be a massive waste of time.

Programming on the other hand if you invest time you WILL get better.

Its weird how way too much blood comes spurting out of the severed arm but none comes out of the torso (where the heart would be pumping blood).

You're not consciously trying to improve your handwriting. Most people don't. Even fewer get into calligraphy, but that doesn't mean you can't practice it.
Patently false, as others manage fine. Sticking to what you know is fine, but please don't discourage others from picking up a new skill.

I could turn that argument around and call bullshit too, since I can draw well but couldn't get into programming. But I know that's due to me not sticking with it because I didn't apply myself enough.

I bloody am. Mine is hardly readable by myself let alone other people.

Exciting…

Suspicious.

Adam will always have his undies, but im still deciding if the girls will get detailed genitals and stuff probably not because its not a porn game

that's cuz it's not scrypted yet

and won't be in a while. 9 hrs and we'll be jamming

whats the best way to manage different screens (e.g. menuscreen, gamescreen etc.) in SFML?

drawing will not help you improve your handwriting

are you on the discord or you have a separate team?

you mean that official channel thing? I joined but I'm gonna use our own channel to work with friends

Did you look at other people's handwriting and tutorials? Did you try different methods and utensils? How many blank sheets of paper have you filled out with just writing? Have you analyzed your sheets and tried to replicate the good examples? Have you tried making variations of those examples by playing with the letters or gap sizes?
This is essentially the same way I go about trying to pick up a new aspect of drawing. It boils down to using references, analysis, duplication and experimentation. You're simply not going to get it right on your first try, so you shit out as many possible variations as you can think of or find examples elsewhere you can loosely copy. I don't believe there's a single skill you can simply pick up and be good at unless you're some sort of savant. Almost all of them need you to make a lot of bad products as a stepping stone to make a good product, as well as looking at people who have succeeded and failed and becoming aware of why that is. But you have to do it consciously.
By the way, that also goes for things people like to chalk up to the talent meme, such as creativity. My designs from 5 years ago were very unimaginative and derivative, but I kept at it and come up with more effective and "creative" ideas now, simply because I was aware of my designs being shit and trying to work on it.

If you aren't trying all possible methods of learning something then you're not really trying. And if you are, then I guess you're simply impervious to learning or in the wrong state of mind. I don't know the solution to the former but the latter you can work on.

Okay so I'm a fucking moron.
My problem came from the fact that after a certain amount of rotation, instead of spherecasting to the ground, the spherecast stopped to a collider I'm using to ground my character.

Can someone explain to me how to use Layermasks for Raycasts and Spherecasts? I can't seem to understand them

bitshift is annoying but when you use the rules given here, you'll be okay.
answers.unity.com/questions/416919/making-raycast-ignore-multiple-layers.html

HE'S STILL WORKING ON IT HOLY SHIT YES

Is this the INFINITE DETAIL guy that didn't sell his send to medical imaging firms?

New thread when? :^)

page 13

No, this is a fairly low profile voxel fag who's been working on this rendering engine over the years. He actually works on pretty similar technologies to the "INFINITE DETAIL" guys such as point clouds, though. I'm just very excited to see something after three years of complete radio silence by the guy on youtube.

I'm trying to do a similar thing but I'm retarded and it's not working. Instead of using two textures per mesh I figured I'd use one series of textures, all being identical except for the expressions. Then by setting textures to separate materials I could swap between them programmatically. My textures are psx small so this wouldn't be so bad, I believe. But there's so little discussion on this I can't find any helpful information on the process. There is this
from answers.unity.com/questions/1344642/change-material-in-runtime.html
but, outside of code formatting and without declarations, this is just confusing to my undeveloped coder's mind.

The programming part of it will be made by my brother who is much more experienced with Unity, im not sure how he will do it.

LD40 started.

What can be done to diagnose a problem like this? It seems to randomly happen every second or so game launch.

did you consider changing the heads instead?, afaik the overhead is smaller that changing materials.

Basically, changing material or sharedMaterial only changes the first material applied to the renderer.
For mutli-sub meshes (meshes with multiple, separately rendered submeshes that take different materials), the materials must be provided in an array. The materials property of the Renderer type doesn't give you a reference to the Material[] that the Renderer is using, but rather a copy. Assigning to that property applies the Material[] to the one that is being used internally.

The guy that answered that question isn't that good at articulating it, and he's using Unityscript instead of C#…
Say you have your mesh set up so that there's two subs, a body material in index 0, and a face material in index 1. To just change that one material, you'd have to do something like the following:

// Defined in your component, so you can assign them in the inspector.
public Material neutralFaceMaterial;
public Material happyFaceMaterial;

// In some function in the component where you want to change the materials:
// Get the renderer on the component
rend = GetComponent();

// Get a copy of the Material[] that is being used
mats = rend.materials;

// Change the material at index 1 to a different one
mats[1] = happyFaceMaterial;

// And then apply the changed Material[] back on the Renderer
rend.materials = mats;

It's a lot of work and fairly verbose. For your case though, I'd actually recommend changing just the textures, since that is a lot easier, and less code overall.

// Defined in your component, so you can assign them in the inspector.
public Texture2D neutralFace;
public Texture2D happyFace;
// Also defined in the component, but not assigned through the inspector.
private Material faceMaterial;

// On initialization… (in Awake or Start)
faceMaterial = GetComponent().materials[1];

// Wherever you want to change the face…
faceMaterial.mainTexture = happyFace;

Does that make more sense?

Changing materials is pretty cheap, and many characters are done with muti sub meshes. Changing the whole head mesh out would be more expensive. Incidentally, changing just the textures or materials doesn't make that much of a difference.

You could also make a texture atlas like pic related and then simply move x,y, positions of your texture. Or only the face texture instead of the entire material.

Top or bottom?

Top

Top but the darker middle is a step in the right direction.

I opted for swapping textures as you recommended and it works great. I needed to change the faceMaterial initialization from the second to first element in the materials array to get the intended swap effect. Otherwise the code worked fine with no errors, but there was no change to the mesh texture. It does make a sort of kind of sense now. I appreciate you taking the time to explain and providing examples. It allows to step through a working process carefully and deepen my understanding. You helped me out a lot. Thank you!

No problem man. I've also been a tutor since college, so I have experience figuring out where confusion is and clarifying it.

Is also a good solution for your problem, but still requires a multi-sub mesh, and modifying the texture coordinates of one of the materials instead of just swapping the texture reference (modifying material.mainTextureOffset).

The advantage of that is you could keep all of your faces in one texture and save some memory, the disadvantage is it's a little more work setting up the material, keeping track of what parts of the texture are what face, and piping the right coordinates into the material. Also if you decide you want more faces, it's more work to add them, even if you make the texture bigger, because texture coordinates are always normalized (wrapped into [0…1) )

So, any of you guys doing LD? Shodanon?
I'm currently working on a quick, relatively simple idea to see if it's feasible or not. If it is I'll post some progress here.

If making a sphere-line collision (where the sphere contains all of the mesh vertexes) before doing ray casting for each and every single triangle.
What's the common sense test before making a line and cylinder? Or it is not complex enough to justify a pre-test?

I was tempted to whip something up based on current project but didn't have any free time today and not sure if I should even bother after losing out on a day.

I am not sure if this counts as agdg since we're a funded small slav studio, but I am preparing the Steam store page for our first game as we speak. It's been an absolute blast and a labor of love. We're doing early access for only a short while to get some feedback on boss / player balance while getting the netcode finalized for online multiplayer. Excitements abound!

What kinda skills can be used to make games?
To me 3d modeling, programming, music, writing, art comes to mind

programming >>>>>> art >>>>> music >>>> writing

figuring out an actually FUN gameplay comes before all of these.

No matter how good you code or how well you draw, you can't make an unfun game fun.

You can't make a game period without programming.

You forgot ideaguying man. Gameplay is after all the most important aspect, if you're not an AAA studio.

programming >>>>>> art >> writing >>>> music.

Music matters the least. If your game is shit but has good music people will just listen to the OST on jewtube.

You can't even make a game without programming like said so its the most important.

No one needs idea fags, everyone is capable of coming up with ideas.

I think a team of one artist and one programmer is good enough to make something worthwhile.
No one's gonna play your game if its outright horrible looking. No matter how good the gameplay is. So i think art can be pretty important too.
Low-poly is for hipsters and having basic shapes for everything can look cute but it becomes boring.

I would say that hipsters combine brand new rendering with outdated rendering, which really annoys me. It's like when you have 64x64 pixelshit and then the pixels get shaded with some OpenGL shader when the game should'nt even have hardware accel. Low poly can be cool if you commit to using 90's graphics techniques.

Basically what said, but it hugely depends on the genre of game. I'm doing something more akin to an atmosphere-heavy RPG, so once I finished textboxes, choices in conversations, collision and movement, I had most of the "mechanics" done.
Once that was done, the combination of music and art is important to set the right atmosphere, and writing is important for obvious reasons.

Once again, it depends on what type of game you make. No matter what, though, you'll need a mix of all of it. Even something like Cuphead had a simplistic, yet fun enough story to give you a reason for your actions.

What I don't get, though, is all the bitching about how X is more important than Y and Z, as if they need validation that their favorite skill is the most important. Maybe it's different for people working in a team, but who really cares what's MOST important or LEAST important, if you're working alone, you still have to do EVERYTHING. What's most important to me is what I'm doing at the moment. Right now it's world design, in two weeks it will probably be more writing, and i'll have to code small things in the meanwhile too.

Let's say someone has an idea and a foundation for a game they wanna make. They actively look for people that can help them make the game and thus create a team of people.

Is this type of person still considered an ideafag?

Does this guy have the ability to do anything besides bring people together and make them work on his idea? If no, then he's still an ideafag.

if that person is also going to contribute to the team in a tangible way with code, art, music, etc, i wouldn't consider that ideafagging. otherwise yeah.

indie projects don't need an RH department with ideas

HR department*

If this person is contributing actual content like programming/art/music/writing they are a content creator like the people they got to help them.
If this guy is paying them they're an employer.
If this person spends all their time delegating tasks, knows every little bit of the game that has been created, is being created, and needs to be created, and they set deadlines and makes sure people keep them, then they are a manager.
If they're just saying what they want to see, they are an idea guy. If this person has a design document they are a glorified idea guy.

programming > gameplay design = sound effects >>>> theme >>>> visual art = music = writing >>>> some other shit
Programming is The Father, gameplay is The Son, sound effects are The Holy Spirit. Other things are secondary. Many people can ignore poor visuals, turn off shitty music and put up with bad writing but almost nobody can tolerate fucking up The Holy Trinity. This is objective and can be proven.

Where does "making a new thread" fit in?

Finally done with the Steam Store page setup. Shit that thing takes ages.

In another 7 other new threads, or older ones being bumped.

Is this still possible without being rich?

This sounds more like a producer honestly.

It would be difficult. Moral is very important when making a game for zero pay, it's up there with discipline and momentum especially in a group. If this person is telling people what to do, doesn't accept much or any feedback, and doesn't get into the trenches to make the game with the artists or programmers, these people are going to have shitty moral. On the other side of the spectrum if people view this person as a manager and they don't do any management, (they don't tell people what needs to be done and who is doing it, they don't set deadlines, etc.), the project is going to fall apart due to lack of focus. You really have to be a top notch manager to make it work.
Paying people gets around this. People will work for a poor manager if the pay is good enough.

Producers are usually higher level and they deal more with funding the game and large scale oversight. They'll say "you need to reach Alpha by the end of this quarter" not "finish the reload animations for the P90 and get them in game by Wednesday".
Basically a producer makes sure the game ships on time, and a manager makes sure the game gets made on time. The difference is small but important.
However in a small studio a producer will be the manager, and more often than they really should producers will try to have oversight on the actual game-play. Like with EA telling developers their games need to have an equivalent to FIFA style trading cards or else they aren't funding it.

Which skill is the easiest and fastest to learn anyway? i guess how fast i'll learn depends on my IQ though, but generally speaking, how long would it take for an avarage normalfag to learn any of these?
(Programming, art, modelling)

...

Depends on what you're naturally better at. The big problem I have when hiring is that I don't need a programmer, I need a developer. A programmer is someone who's real good at doing code and reading about different sorting algorithms on Stack Exchange. But he has no clue what a vertex normal is, or what texture UVs are etc. He wouldn't know how to edit an animation to save his life. A developer knows a bit about everything and has a good idea how the big picture of game making looks like. He doesn't need to be good at modeling, but he does have to know what skin weights are and how bones drive vertices in an animation. He has to know what attenuation and reverb are etc. It's not easy but if you want to be good at it, it's the only way to go.

This applies to not just programming by the way. A modeler has to be aware on how his model will be used in game, make sure to optimize triangle counts, understand UV sets (you might need a second UV set for some special effects or separate UV islands for light baking etc.). In its true form, game development is unforgivable. It's very easy to fire up Unity and make something that works more or less, but seeing a project to its completion is for most people insurmountable.

ded tred, where is my new bread?

...

Why would you learn all of those in their entirety, rather than learning the parts that you need as you need them? I'm not saying that learning them in their entirety is a bad idea, but it's not going to get you devving very quickly if you force yourself to finish all of those first.

Also, I have almost never had to do anything more advanced than Algebra I to make a 2D game. Why do you think you need to?

How to learn:

Read loomis, watch glenn vilppu DVDs.
Practice drawing from imagination trying to recreate pictures you see from google or your anime shit, but trying to copy them from imagination (like you watch the picture then draw it, then if you forget about a part, watch the picture again).
Try to construct these pictures as if you were doing it from imagination.
Start with the movement lines, then do perspective boxes, then proportions, then anatomy, then clothes and hair, then clean up and ink said piece.

Practice perspective, anatomy, values and planes.
Learn how to render a realistic picture by copying some masterwork.

Read "The story" from robert mckee.
Read TV tropes articles on writing, read tv tropes.
Download the TTC courses about shakespeare, poetrhy and the one called writing better sentences and writing non fiction.
Also learn grammar, and seek the ode less written (poethry book).
Read the bible, shakespeare and the most important works of literature in history.
Learn to write a sonnet.
Watch at least the 100 best movies of all time in hollywood.

Read SICP and watch it's course.
Learn to code pong and breakout using SDL on C#.
Make some games with libgdx.
Read about data structures and algorithms.
Read about optimization of algorithsm.
Read the basics of computer architecture.
Learn the C programming language.
Learn algebra, trigonometry, calculus, linear algebra and physics.
Learn about OpenGL and do some graphics demoes.
Learn to do space partioning like quadtrees.
Learn the basics of game engines architectures, physics engines architectures.
Read the basics of AI.

Google "The complete musician" PDF which is a college book on music theory and analisis.
Read the basics of music theory (music theory for dummies).
Realize music is nothing more than a melody line with one or more voices at the same time playing basically the same notes of your chord progression.
Google some tutorial on the basics of harmony, chord progressions and counterpoint.
Buy a cheap keyboard.

look up some tutorial on youtube called "from seconds to hours/mintutes of gameplay"
read a book on fun and game design.
read the book of chris crawford of game design.

read on topology, learn to draw your own references, mostly if you do the 2D step, 3D is easy as fuck.
get a pen tablet and learn about sculpting on clay to master zbrush.
look up the blenderella course, the blender rigify rigging DVD, the chaos and evolution DVD.

download the animator survival kit.
Master the 12/10 animation principles of old disney guys.
Get some exercises of animation.
Master drawing.

Read the bible, the basics of history, get some hobbies that aren't games like gardening, fitness, martial arts, music, reading literature, economics, history, political theory.

Read about sexual transmutation.
Do nofap.
Get a dog.
Learn diferent languages.
Socialize.
Read about investment.

Read how things get viral, a book on viral marketing.
Read about the blue ocean strategy in bussiness.

Start doing exercise, do calisthenics or join a gym.

Cook your own food, don't eat garbage, learn about nutrition.

if you try to make anything most complex than the most basic 2D shit on SDL and OpenGL you'll need it.

trigo is fundamental for touhou shit and projectiles.

linear algebra is fundamental for shit like Mode 07.

calculus is fundamental for physics.

I forgot to mention this will take at least 10 years.

t. oldfag.

10 years for all of it or 10 years for each

This, pretty much. I needed trigonometry for several things (shadows come to mind), but it was just simple stuff that i got from a google search (still nice to understand). Aside from that, is was all simple math but it might just be because I actually liked math and was pretty good at it.

10 years for all that shit.

I've been doing it since 2010 and still lack two or three years before making a stupid touhou on C++

How will being more horny motivate me to learn programming/art? I don't get it

It's the secret of motivation.

everytime you fap you're wasting your sexual energy instead of using it to reach your goals.

new thread