Amateur Game Development General ~ /agdg/ + /vm/

But with the blast shield down she can't even see, how is she suppose to fight? :^)

The other people are working on it, they're just post their progress in the Discord chat to everyone else in the thread rather than in these threads.

Also with your UE4 blueprints programming point, I don't know where you're hearing that you can't do that in Unreal Engine. If I'm understanding you right the Introduction to C++ Programming in UE4 tutorial walks through the basics of doing what you described.
docs.unrealengine.com/latest/INT/Programming/Introduction/

That's about what we're aiming for.

It does work like that and I use it for exactly that. Likewise I'll write my own code functions that can be called by blueprint for more advanced math which makes everything run much faster in the end. Comes down to what sort of team you have too though. Working purely in c++ is much better for my sanity, unless I'm working with UMG / UI stuff.

Oh, is there an /agdg/ Discord?

There is always people that have a thread open, and having to go to the catalog and search for agdg every time, multiple times since the new thread's not been made yet usually, is a hassle. And like I said, it's common courtesy. Hell we used to link the previous thread in the OP as well, but the lack working archive for fullchan means that's pretty useless.

Died, tried going into the first level again, and it froze up completely while the console just stopped updating.
First time I actually downloaded it too, I can see why people complain about the slippery movement. Booted up DooM as well to compare, and I can say the following things:
Also not saying you need this, but I noticed something about DooM

Some more general notes:

No, it's for their particular horror game. We do have an irc.

Ohh, is this going to be the 80's slasher thing?

I haven't heard anyone say you can't do that, just some people saying that coders and, uh, blueprinters are turning up their noses at each other, which I find to be weird. Maybe I just shouldn't listen to hearsay.
From your link
Choice of IDE's aside, that's exactly what I mean.

Yeah, me too, but I'm trying to build some skills and habits that will make me a viable team member rather than continuing the never ending cycle of trying to do everything myself plus reinvent the wheel on top.

the respawning is really buggy, you should re-start the game every time you die until its fixed

I will try adding head-bobbing, to make it feel better. Thanks for pointing that out.

The robot enemy having blind spots is intentional.

The reason that the resolution is like that is windows API does that unless you tell it otherwise which I havent gotten around to.

I didnt put in auto-reload for any reason. But, I think its important to make the player switch guns.

The first gun is pretty bad. Do you think I should have more alternatives to the first gun? I noticed that i make the player use it maybe more than I should.

This is because the movement is bound to wasd, not WASD.
(its how the binding system works, ASCII codes)

I should get more textures, I plan to do a pass where I put in more props and textures. Thanks for pointing this out.

Thanks for giving me a new "stuff to do" list, I will try and fix these issues. Some of them are rather easy to change, like view bobbing and the resolution. Others are harder to do.

Yeah good stuff. Thing is, if you can program in c++ you can do blueprint with ease. It's just a shittier way to visualise what's going on if you can read code. You can see so much more and debug so much easier through code than blueprint. At the very least though, if people are using BP they should learn to use reroute nodes to clean up spaghetti. Cleaned up blueprints aren't too awful to look at.

I'll call it an "Alloy Sprocket"

Was looking for you posting again. Is that a raycasting engine?

I'm going to make an MMORPG

they sell frameworks for that on the untiy asset store
go wild

(checked)
I like your progress man, keep up the work.

no. It uses a BSP tree and then renders polygons using openGL

it would be pretty cool if it was raycasted , though

What's with the lack of lighting?

there is lighting, I made it less pronouced in the past few versions, though.

I'll make it easier to see the lighting again.

you can see in this picture that things are darker in the distance- thats the lighting model.

I mean like light sources, it looks like it's just distance based. looking to do it easy?

yeah. when I was working on the game, I realized I would have to cut a lot of corners just to finish the engine- doing it the proper way would take at least a month, maybe more.

I don't want to spend so much time on this one project that I get bored and stop working on it- so I try to limit amount of things that the engine tries to do, so that I can finish it sooner. (I started working on this November 2015)

My goal is to implement real lighting in my next game, because I will be able to easily surpass this older engine once i start programming a new one from the ground up with all the things that I learned about engines.

Fuck off, shill your shit game somewhere else. Sage

How much to go on this one, and are you going to make another sprite based fps? your work shows promise fam.

What the uh, What the fuck.

Do the Koreans work for peanuts? How the hell do they keep creating those massive MMOs with crazy graphics for so little?

Wooden Cog?!

Mineral Mechanism?

the real question is how they even get allowed to make games
i thought their government was completely against vidya since it's run by turbo soccer moms

thanks!

this game is almost done. I spend the vast majority of time making new levels for the game now, and once all the levels are done, I just need to go over the game once last time to polish it, and then I can release it. (hopefully before new years)

All the programming goals for this engine are met- so its just a matter of making new content and fixing bugs / polishing features.

I don't know if I am making my next fps sprite based. However, I think I might need to do that, even though im not sure. It depends on how long the rest of the engine takes. I will probably use some sort of model format for this.

The challenge with making a model based FPS for me is that I want to make my own CAD software for making the models. I know this sounds really autistic.

The levels for this current engine are made using my own CAD software which crashes a lot since there were a lot of C concepts that I didn't understand fully in January when I was writing it. (the engine and map editor are completely written in C)

However, the engine doesn't have this issue.

For what purpose? There is blender already which is breddy gud.

That sounds a little too hardcore, just build an export add-on for blender or something, I can understand map tools since map making tends to be unique per game, whereas models always seemed like a different implementation of the same thing.

Are you interested in making unique or interesting gameplay concepts or just making an intricate engine built for your own needs that handles FPS gameplay?

I like programming recreationally- so its fun for me to write things like that. I have already written CAD Software before. But i do plan to use blender for my next engine just because of time constraints.

I enjoy making the engine so much, that its more fun for me than making the actual game. I think this is true for most enginedevs. Although, im probably not having as much fun because I have to use my aufwl map editor.

However, once my engines are better, I will be able to make more unique and interesting gameplay concepts. Right now my engines aren't good enough for the gameplay to really stand out.

robot girls were a mistake

Haydee only reminded me of some scrapped stuff I had lying around, in my particular case I not pretending to go after anything else except something good to look at, the original plan was to make a game engine female mannequin to mess around with.

I really should know this but how do I rig a model in blender that's made of separate pieces?

Bind two meshes to the same skeleton, dummy.

You rig each piece.

When you think about it Haydee is just like that, you test stuff here and there using that player model.

Who's that faggot that was working on a cooking game? Is it exclusively about cooking? How autistic is it going to be? What other mechanics will be present?

The body is pretty much done, just the skull left.

...

Last thread, we talked about the pros and cons about unreal engine 4.

Now, let's talk about Unity. What do you think about Unity?

>100% AUTO AIMED COMBAT

most games tend to run like crap despite no fancy graphics is going on and my computer ain't that old either.

Someone, punch me in the face.

what are you planing on doing m8y?

I don't know.
I could start with concept art but I feel really unsure about everything I make. The voice in the back of my head is never content.

Most of what's made with it is shit, but the engine's alright as Tetrachrome dev proves.

fucking duh, thanks. It's been just about half a year since I tried rigging anything so I completely forgot that shit.

forgot muh pic

The Unity team are more concerned with adding engine features than fixing long-standing bugs and performance issues. I would recommend to avoid if possible.

Delivered to my client, not perfect by any means but he liked it, Maya seems to be giving me a lot of trouble adding materials and properly rendering the results.

...

Nice fucking gig.

well hello /k/

BTW, can I see some of your work too? I am certainly not even good, but I'd like to learn.

• Good engine design overall; proper ECS, alright scenes and prefabs, things generally make sense. You can tell it's a dedicated general purpose game engine, and not an afterthought SDK to an existing game.
• Clean scripting workflow, your code is a DLL that gets recompiled and reloaded as soon as you change anything. The time it takes to implement something is as short as it gets.
• Despite the lack of source access, there's an API to extend or bypass almost every of the engine's systems.
• Relatively painless and customizable asset importer.
• Editor scripting is easy and powerful for what it is.
• Great docs. RTFM is the solution to most beginner and intermediate problems.
• Pretty lightweight on its own, if it runs like shit it's usually your fault.
• Backwards compatible, 5.1 is still compatible with shader model 2, and 4.x can run on really ancient systems.

• Telemetry in Personal. You're allowed to firewall it in the editor and runtime, but you'll have to warn your users about it.
• Bad and buggy serialization.
• Managed script language. (you can still link to your own libraries if you need to)
• Awful release cycle, too focused on catching up to UE4. Has anybody in charge of that ever tried to make a real game?
• RETARDED caching system. Shader, occlusion and GI cache can add up to hundreds of thousands of individual files. You might want to make a dedicated partition just for those.
• Bloated runtime. (about 50 MB regardless of the game) Supposedly going to be fixed with modules.
• user why are you using unity? :^)))))))))))

Contrary to the engine's reputation, programming skills are more important here than in UE4. The engine itself is just glue between a few pieces of middleware, plus a pretty good and portable renderer, a simple scene system, a shitty serializer, and an editor to handle it all. There are no noodle editors, and the defaults and examples are all barebones; you're expected to make everything by yourself, including some of the editing tools. (if you're using the default inspector for everything, you're doing it wrong)

tl;dr mediocre engine but mods will fix it

I'm not really arguing that you're shit so much as the client is, since you're obviously not completely happy with it.
But legit, good job on landing that commission stuff. Makes me wonder if I can pull that shit off. As for shit I've done for other peeps here's one I did 4 free. Never quite finished it and the project ended up dying though.

Why not ask around? you could have landed the commission instead of me, I have only done commissions for two people here and at least I still offer support and patching when needed.

How did you get paid for this?

Anyone here use Godot?
Is GDscript too slow to bother with? I read they're implementing (mono) C# in 3.0 but is there any reason to not just write everything in C++ from the start?

By following the client's instructions, thank you.

Is it possible to convert a model which got skeletal animation to vertex/object based animation or something? Because I would like to export them to .md3 format (I am modding for Zandronum 3.0)

Skeletal animation seems to make animation actually a tad easier to use and organizing all the action stuff.

burn

Welp, it looks a bit better now it's not perfect since I am using Atlas but at least I am not wasting textures unnecessary.

comparison

Okay so for food skills, I've got Cooking, Baking, Confectionery, Condiments, Brewing and/or Winemaking, and Dairy and/or Cheesemaking

I don't know if its worth splitting Winemaking and Cheesmaking though. Ideas?

Lookin' good

Thanks fam.

Hmm, Maybe try drawing some stuff you really like the most or what you think it would be easy to do?

A new area for the level that I am working on

What the fuck are you designing that would need a confectionery skill?

...

It is.

You could add more flashy skills that you can only unlock when your cooking level is high enough, like Molecular Gastronomy.

Also, Mise en place, which means setting shit up before cooking so it isn't a huge mess, could also be a skill.

...

Maya is very powerful user. Mastering just takes time. Something doesn't work just that way you like? Fine. Create a python script in the program that works just right.

Cooking is shit like roasts and stews
Baking is stuff like pastries and cake and cookies
Confectionery would include glazes, candied nuts, licorice, gum, any kind of sugar based treat, basically
Condiments are things like pickling and relishes, also sauces and dips and broths to compliment and prepare for other dishes. Could also include jams and jellies. Surprisingly important skill

You're right, theres probably enough content for each to merit their own stuff

Yes but you didn't answer my question
What are you making that requires so many skills with so much depth.

I don't think you'd need condiments as a separate skill, that's getting ultra tedious, making a sauce as part of a dish and making a sauce and storing it to add to subsequent dishes is no different, I've made all kinds or preserves and it was really no different to cooking, with confectionery that's a whole discipline of preparation and uses a largely special set of tools and ingredients so that I can understand.

You do have to balance these things, I'd recommend against spinning off cheesemaking or winemaking into their own brackets, there's too large a crossover between dairy and brewing

Are there any particularly good tutorials, books, or video series' that help improve ones modeling abilities? I'd really like something like a course on the subject.

My duck still only detects collisions when hitting from underneath them/with his head

On a conceptual level, I've always enjoyed the idea of a single player MMO-like game (but of course that's silly to do alone)

Anyways, I realized there isn't really an Animal Crossing game on the PC. Farming games are also really comfy, so why not combine the two? So, there'd be a focus on farming, cooking, fishing, and some crafting/production/survival elements. The idea is that you'd survive and forage alone for the first week(s), and after building more shelter, you'd attract random waifubait villagers into the area.

Mechanically, they'd offer better crafting recipes, and improving relations with them would involve quests, giving gifts (like preferred foods or items), and that sort of thing.

Not to be a spoil-sport, but in the first sentence aren't you sort of describing exactly what Stardew Valley is?

I haven't really played it, I avoided it entirely because it was associated with Chucklefish

I know it's kind of a big project to take on solo, but I just kind of want a rough idea on where to take it when I do get around to it

Get into drawing, pretty much all skills from that carry over. Anatomy, composition, proportion, form, etc.
I recommend "figure drawing for all it's worth" by Loomis, if you want a very quick start. The book might seem pretty hard at first but if you stick to it and practice all the stuff you should get really quick improvement. After that - a lifetime of building up your knowledge and visual library.

I've done that for quite a bit myself, though I've got stuff to learn there as well. So is it mostly about working from reference for a while? Is there a good goal of how many sculpts a day I should go for?

Just trying to find some direction at the moment.

That's some nice art man. You have very confident linework.
When it comes to reference, I think it's best to use it all the time, not just the beginning.
As for daily amount, don't overwork yourself, I have a very loose goal of "one drawing a day". It can be a 30 second doodle, but it has to be everyday - being a "person who draws every day" just changes your mindset about the whole thing and in the long run I end up being more productive than if I just forced myself. You can probably apply that to sculpting and pretty much any other thing.
If you don't have problems with motivation then just keep working however long you want to daily, but keep some low minimum like at least 10 minutes of fucking around in blender a day. This way if you have a shitty day, you don't break your productivity streak.

PRACTICE, mate.
go full /loomis/ on your concepts.

I want to record .webms of my games, what is the best free tool to do this? Just searched and found a lot of bullshit programs that either recorded with no sound or wanted me to actually buy it to record properly.

What? Just use OBS and then something to convert

This is exactly what I wanted, thanks user.

also nice trips

Thank you.

also, I'm pretty bad with work consistency so making it a point to do 10 minutes at the least sounds like a solid plan. Also, I have an art reference folder but don't have much for 3D models so I'll go ahead and start getting that together as well. I appreciate the advice user.

waifu material right there

bretty gud

...

Bill Clinton.

RAPE:
KILL:
MARRY:
FUCK:

...

MUH FELLOW MURRICUNS

I like Dwarf Fortress but you've just made a light and simple game pointlessly heavy. Whatever floats your boat I guess.

This tutorial was really good it covered a lot of the basics although you can ignore a lot of the early stuff thanks to dynotopo, and a lot of the later stuff has changed thanks to Cycles

CG Cookie has a lot of good content but they update less than monthly, I'm surprised anyone holds a subscription.

Also the guy who made the face tutorial has been working on a new one to replace it.

Practice makes perfect and you've got a lot of practice ahead of you.

You saying "pointlessly heavy" reminded me of Wurm. There's something like a hundred or two hundred skills, basically all of which have some small purpose and associated titles.

I gotta be careful posting heads in the future

Yeh, I figured my art knowledge would put me somewhere above cringe tier, but not quite in the mid tier, though I've yet to do a real sculpt and am more focused on doing as many as I can quickly so I can get the mileage down. I'll definitely be following along with that video tomorrow, and by any chance would you have a decent one for the figure? I'd say that's my weaker point at the moment. It seems there were bits of anatomy I could ignore or bullshit when drawing them but now that I'm working in 3D they are sticking out like a sore thumb.

I remember carrying five logs back and forth at snail speed to my charcoal pile.
that was fun

I was 8/v/'s collier for a while. having to carry several tons of dirt and several tons of coal just for a few dozen coal.

We are only memeing on you because we want to encourage you user, don't feel bad about it.

Last one, I promise

No please, pic the 3rd face and do a Bill Clinton version

just 4 u

Thanks

I've been working on a bunch of small parts of my terrain generation, and a bit of optimizations to the backend of things; so most of it I can't post about w/o a blog heh.

Here's what I've been working on that's visual though. Using noise derivatives to - in essence, procedurally create a normal map at runtime - but in reality I'm editing the per pixel normals with noise derivatives at runtime (not static, can make it 4d, alter with time, or sin(time), etc), and ranging from surface features to more complex details, and to effects.
It's a bit of a bitch for LoD, bcs if you have noise that's higher frequency than the pixel refresh rate it wonks out, but it's manageable w/some adjustments so you don't get any of that aliasing.
I'm looking into something like parallax mapping or displacement mapping at runtime… so these aren't just altering the normals, but also displacing the pixels themselves.

shiiiiiiiiiiiiit user, this looks rad as fuck

I have done nothing to work on my game in two months.

Then get to it, mate.

Last time when the Korean Final Fantasy-inspired one man game thread was created people threw a fit that he wasn't really showcasing a game since he didn't make the assets.

So what is wrong with making a game using store-bought assets?

work on it now

Maybe ask the people who threw the fit.
That thread is still up.

Literally nothing as an indie. You practically have no choice. The biggest downside is that other people can use the same assets as you have. If you've got the money to finance your own assets you should, so long as you can make a good game to go along with it. A game with custom made assets by a good lead artist will always outshine store bought assets aesthetically.

Because these threads are about developing games, people that complains about stuff like that doesn't come to these threads, you won't find that kind of mindset here.

If you keep it as place holder art nothing, if it's the final product then it's like rpg maker.

Pretty nice review, thanks.

Really remembers me of gamemaker

Also more notes…
Later, once I feel like adding it, I can easily apply multiple types of different procedural "normal maps" to different textures, based on their blend strength in my triplanar shader, and that means rock-like features for rock textures, sand-esque features for sand textures (same as u can see the different textures in the fourth pic).
I need to setup some type of profile/properties table so I can just pass in the type of normal map I want… w/some non-gpu side code w/the settings I want for the procedural normal maps per the texture (pass in rock flag, w/detail amount flag, etc).

For now these are tech demo esque features, and not mainstay features… so I'll finish up the small details or qualms, and then move onto adding more features (maybe… finally finish implement CSG, instead of getting distracted by adding more detail features).

thanks user

Just keep practicing, you still need to be profficient with the software as well as just understanding 3d concepts.

I don't have any videos on getting the form right but try downloading someone elses model in a pose you like and sculpt over it, no-one will know you did it, or you could spend a lot of time making a nice basemesh for your own work, I always reuse my work.

Because it makes it look like they did something of value.

The first two pictures remind me of morels and false morels. Types of edible (and their inedible false version) mushroom. Around here people call them dry land fish.

Looks like it could be a god-tier horror game.

This is basically my life everyday now. Constantly putting school off because it's been relentless and then I end up wasting a ton of time. The part worst part is, even if I wasn't wasting this time I still wouldn't be able to work on my game, because I'm obligated to do other shit at the moment.

Side-note. I forgot how comfy these threads where.

I've added over the shoulder aiming. The character stops moving when aiming and can only aim in limited an area in front of them (70 degrees left or right, straight up or down). There's also a button to switch shoulders and a check box in each camera if it turns out we don't what the player to be able to aim in certain areas. In the future I'll probably let the player turn when they reach the edge so that they don't have to un-zoom, turn, then zoom again.
I've also slowed down the movement, added a run button, added a crouch button (for now it just slows speed and changes the capsule height, no actual crouch animation plays), and changed the movement to relative to camera. Also I did some work handling movement when the player switches cameras so that the character doesn't walk back and forth when switching to a camera facing the camera they came from.

Thanks. I'll do my best.

quite RE-esque

You prototyped everything I wanted to prototype and more, I must admit I feel envy.

Also, are you the guys who teamed up some threads ago? I really hope you have the courage to make those models and props happen.

This still looks sick, are you the guys making the 80s slasher thing?

Do you think that you'll be able to make the OTS aim work? Seems like you might want to go with standard non-zooming aim, since it's going to have a fixed/stationary camera if I remember correctly.

Maybe save it for using with a sniper rifle or something? I don't know.

Shooting in RE4 > Other RE games. I assume he's shooting for the best of both worlds

What are you using for modeling and animations? Blender? Cinema 4d? Maya?

Those look like the default UE assets, so those are prolly just his placeholders.

Thanks.

Yes to both of those questions.
As for props, the other guys are working on them they've shown in progress models in the Discord. Once I get the models you'll probably see them in my updates.

It's not finalized. We're going to see how it goes but we're discussing alternatives in case it doesn't work. If you have other ideas that you think would work better feel free to make a suggestion. The more options we have to pick from the better.

Hopefully that won't be the only thing we're shooting. :^)
But I agree, I like having more precise control over my aiming.And OTS is easier to implement than auto-aim.

I'm not doing the molding/animating. But the guys who are doing the modeling are using 3ds Max.

Yep.

It's cool to see something like that work out. I'm too paranoid of other people not pulling their weight/trying to take over/not taking it seriously/etc personally.

There's 4 of us including me.

i seek help
i'm trying to make my shitty a* pathfinding multithreaded, or in otherwords i just pause for a frame every 250 nodes during the search
the first 2-3 paths work fine. but after that the unit starts going wherever he wants rather than where i tell him to
and he gets errors for null nodes in his path
what am i doing wrong?
i tried reimplementing the dictionaries since i thought the searches might somehow be intersecting (even though with 1 unit that shouldn't be happening) and that only made performance worse.
the user that made a github for this kind of abandoned it
NodeElement _start, _goal, closestPossible; PriorityQueue frontier; bool searching = false; public IEnumerator MultiThreadAStar(NodeElement goal) { if (searching == false) { Debug.LogError("new search"); double tim = Time.time; searching = true; SquareGrid graph = PathfindingMap.Grid(); int count = 0; frontier = new PriorityQueue(); frontier.Enqueue(curNode, 0); _start = curNode; _goal = goal; closestPossible = _start; while (frontier.Count > 0) { NodeElement current = frontier.Dequeue(); if (current.Equals(goal)) { break; } foreach (var next in graph.Neighbors(current)) { double newCost = current.costSofar + graph.Cost(current, next); if (!next.passed || newCost < next.costSofar) { count++; if (count > 250) { yield return null; count = 0; } next.costSofar = newCost; double h = Heuristic(next, goal); double h2 = Heuristic(closestPossible, goal); double priority = newCost + Heuristic(next, goal); frontier.Enqueue(next, priority); next.cameFrom = current; if (h2 > h) { closestPossible = next; } } } } Debug.LogError("finished search in " + (Time.time - tim) + " seconds"); FindShortestPath(); searching = false; } } public void FindShortestPath() { path.Clear(); NodeElement next = _goal, start; if (next.cameFrom == null) { next = closestPossible; start = closestPossible; } else { start = _goal; } path.Add(next); for (int i = 0; i < start.costSofar; i++) { next = next.cameFrom; path.Add(next); if (next == _start) { path.Reverse(); break; } }/* for (int i = 0; i < frontier.elementstoremove.Count; i++) { NodeElement node = frontier.elementstoremove[i]; node.passed = false; node.cameFrom = null; node.costSofar = 0; }*/ }

no wait, i was just imagining things
doing this with only one unit doesn't cause any issues, unless he sees an enemy, then things go batshit
the issue is when i'm commanding multiple units. the searches definitely intersect
but using dictionaries killed performance in here
what's my alternative?

Anyone know if stealth games just use the physics box for vision detection?

Depends on the game.

you could use colliders instead of the bounds, but it's unnecessary
you could just change the size of the bounds depending on whether the character is crouching or not, for any other case using colliders would simply be excessive

I was thinking more of the objects in the world.

As in. If the player is hiding behind a prop, how does the AI know not to see through it. I was thinking physics object, but there are things without physics like foliage.

just use colliders that don't mess with the physics
unity has OnTrigger in their colliders for this

anyways, here's the alternative with dictionaries
in this case, i get no errors for null nodes, but for whatever reason, some of the nodes don't have their correct predecessor
in the function that recreates the path, it should reverse the path once it reaches the start. but for whatever reason it doesn't reach it, and some units teleport to their goal and move from it to their initial position
i have no idea why this happens, the priority ques should be stopping this
pastebin.com/fHFBJi6w
pastebin.com/VJp43Ja8

In that case yeah, depends on the engine but stuff like UE uses visibility channels and box bounds etc to prevent traces from seeing through it.

I was actually doing my own engine. Primarily because I wanted ridiculously smart AI, where they would render a very low quality frame on their side and see if they could actually see you or your shadow, but it's starting to seem like it's more trouble than it's worth on modern hardware.

You should just use displacement maps at a level this fine grained.

You can do that in any engine.

I can imagine it would be slow as shit

That's literally what I've wanted to see AI do for years, add in some way to store frames for reference so they can tell if objects have moved and I'd jizz hard.

You only have to draw a depthbuffer of the static scene, a depthbuffer the character, and then check if the character's buffer's pixels are equal or closer than the depthbuffer's pixels. No need to render the scene itself, unless you also want to figure out a method for camouflage, in which case yeah no.
Unity might actually be able to handle this to some degree as I remember their cameras having some kind of layer system, but it would need some special shaders. UE doesn't have that last time I checked (needed it for a special effect back in 4.11 I think).

That sounds excessively complicated for what you want. Couldn't you do something like pic related? Cast a few rays from the player to their shadow and then have enemies check to see if they have LOS to those points.

That isn't so hard to set up in UE4. I've already got that going sans shadow. What's difficult is writing good AI search code. So far mine is relatively simple there. AI is fast as fuck because I have a completely custom system separate from the engine to handle all of that. It's not even a stealth game but I want very reactive AI like FEAR or STALKER but better. The AI system is fast enough to control 1000 AI with less than a 2ms game thread reduction iirc. Game code doesn't run quite so fast however, but I can have over 80 AI running my current behaviours (avoid grenades or other dangers, advance towards opponents last known location if you can't see him, investigate sounds and visuals, shoot and kill opponents you can see, retreat to cover when you've lost morale). Going to expand that to make them move in more intelligent ways and eventually form squads with nearby allies and work together towards goals like how FEAR did that. Taken a bit of a break from it atm because I'm dreading having to go into the UE4 pathfinding code and making custom path finding that makes paths that enemies can't trace to (out of their vision - so in cover) cost less, so they'll run behind cover when available. Also got to add better cover seeking to it for in combat with partial cover.

It was basically intended to make decent camouflage and shadow detection, rather than just hiding behind objects.

That was one thing I was thinking of for the player model itself, but I haven't thought of doing it for the actual shadow. I scrapped the shadow idea, because I get worried about doing anything that is too unreliable will be more frustrating than intense/fun.

Yeah that's understandable. Sometimes it's a fine line between TECHNOLOGY and being frustratingly realistic.

alright, i've nearly got the pathfinding working how i want it to
currently, when i tell a unit to go to a destination, it still checks the surrounding nodes of that destination in case it's occupied (otherwise, it will look through the entire map every time i tell someone to attack an enemy, if i do that i'd have to drastically lower the max searched nodes per frame), but the pathfinding itself is also using multithreading, usually takes 0.7 seconds to find a path to an unreachable target, with 250 searched nodes being the max per frame
now my only issue is that for some reason some units tend to get stuck, and there's some very rare moments when framerate drops a bit

Thats pretty interesting. I'm thinking of scrapping the idea I had though in favor of focusing more on a thief-like game, where the AI is dumb but it's set up well enough so that it's still very challenging to deal with.

It would/could work for the shadow, but not the camouflage as you'd need full rendering for that, plus some kind of edge detection looking for human shapes. Basically what the kinect does except multiple times. Well, that might be "doable" as well but I'd expect it to be really fucking slow. Plus I don't think there'd be much of a benefit for that over what MGS3 did.

If you make a thief-like game that'd be really cool. There's a decent niche there. I'd certainly give it a shot.

Why? UE4 and Unity both let you do low level rendering calls. The only slow thing here is the technique itself, but it's not that bad of an idea if implemented correctly and optimized.

There's a number of advantages to doing straight up rendering. First of all, raycasts usually use physics data. Second, raycasts see no difference between an opaque material and an alpha one, and will not go through "holes" in a texture. Third, above a certain number of raycasts, using the GPU is just more efficient.

No, you'd just encode logical info in the output color. You don't need to (and shouldn't) render the AI view with the same shaders as the player.
I can imagine this technique would be very useful for bots in multiplayer games, actually making them aim and pay attention to the environment.

You can only optimize it so much before it starts getting janky and unreliable.

Use seperate dictionaries?
I'll be honest though I don't really understand what the problem is.

Oh yeah, it didn't seem mean spirited at all.

I was just worried I was derailing the thread.

especially considering I'm doing this more for animation than game development, but I figured you guys would know this sort of stuff best.

Yeah, I heard people build asset libraries out of their work. That's one of the things that drew me to this medium. Now I just want to get decent as soon as possible so I can see how the animation tools compare to traditional. As I said, I'll watch that video you posted today as well as maybe download a pic of a skeleton and draw in all the muscles so I can figure out just which ones I really don't know.

I want to create my own game, and am relatively proficient in low-level coding and have enough time on my hands to learn other things so I think it's possible but I have one major problem:
I can't model to save my life
Is there a repository for 3d models which are free (preferably GPL/Public Domain licensed) where I can troll for 3d models to use?

Did I mention I'm poor as hell? Yeah, I'm poor as hell.

Oh, also I need a free game engine with which to told my hand while I make it. But I could settle for Unity if I had to.

opengameart.org/art-search-advanced?field_art_type_tid[]=10

yea, i thought dictionaries were the issue, the actual issue was that each unit was checking the entire map to reach his enemy rather than the nearest reachable node

Many thanks user!

how easy is to get people interested if you make a prototype?

Like gathering a team based on your prototype?

I'd be worried about getting the right people interested. Not just dumbfucks who are like "OH COOL GAMES" and then proceed to not do anything.

I think I'd be seeking some publisher, you know, professionals.

...

I rather get a basic deal and get some publisher to help me make a quality product than wasting years making some shit that will look like garbage.

What do you mean by this? I thought the publisher just handles distribution, not even marketing for indie games.

If you want a real world example of how to deal with publishers, Frictional Games (Penumbra and Amnesia). They went through Paraodx Interactive to publish the Penumbra series, although I'm not sure whether or not this was AFTER they where already selling it online. They must have gotten fucked over, because later they sold Amnesia on steam and through their own private servers. Eventually they decided to make a boxed version after the game was already wildly successful, and published it through THQ. Instead of using the image they had planned out (Pic 2), THQ published it with Pic 3 and it has widely been hated by the fans and considered a laughing stock.

And they never got machine for pigs published, although that was made by SJW hacks so naturally it bombed.

I dunno the word, I don't mean marketing and publishing but getting a budget and getting a team to make it.

You mean setting up a studio. If you look again at Frictional games they got a loan to make Amnesia, and this was after they already have made Penumbra. For Penumbra they where just a bunch of guys who made a game in their free time, not too differnet from the anons in this thread who teamed up to make a horror game. I don't know how they met though. Just have to keep your eyes open.

dunno, I gave up on my dream of making a full professional game by myself.

I rather not give up on my dream of being a game developer and just focus on making prototypes.

It was pretty much just an user asking here for people to team up.
That can definitely work, but there are loads of potential problems with teaming up with strangers.

You'll be better off searching for a studio and applying for a job if you want to be a game dev.
It's possible to found a studio with some people, but hard to get money if you don't have already something to show.

I have a prototype pretty much set up already, with VS mode and co-op mode, playable online, some people might find it fun to play, and the game's mechanics are deep enough to make you want to play another match or two.

I will keep advancing even if it feels like i'm going nowhere.

REDDIT FORMAGGIO XD

(checked)
What game are you working on? Do you have any screenshots.

I'll record some gameplay later, alternatively i could post the steam workshop page.

But that would be:
So i'll post a webm later.

So it's a mod? Are you prototyping in Tabletop Simulator or something?

nope, gamemaker, webm related: my biggest inspiration

hmm does anybody got a Idea what I should change this vehicle into?

Previously I used as a arch-vile replacement which as its ability to resurrect corpses but different flame related weapons I can't remember what these were.
But I removed it since it was too small and much more annoying to deal with.

I plan later to make some few "APC" like vehicles which carries troopers or something like that in limited manner i.e. they won't flood the map like pain elemental does.

GZDoom still has no support for such shading system so I couldn't care less, unless somebody managed to add these somehow.

oh wow… it really does look like Clinton. Now I feel dirty.

Hover Tank

Hmm so a Hovertank with flamethrower/ball/whatever weapons? Well alright I will try to modify that one into it soon enough.

Actually, you forgot to turn on mouseaim and you didn't turn off mouse movement.

Well if it has flamethrower/ball/whatever weapons, make it a hover catapult

recorded something

hmm well I am going to stick with the Incendiary based weapons for now, then later I will try to modify it's weapons for a catapult one.

The game runs at 60 FPS, it's just my machine can't record at that speed.

Hey, if i had money the game would already be made

forgot webm

So if I see it right it will be a medieval themed RPG game?

Not really, it's a mix between micro arena (SC/WC3 custom map) and magicka.

huh hmm, well good luck with that I guess.

I take it you're the same user who uploaded his code one or two threads ago. I'm the user who told you he'll rewrite your code and upload it to GitGud. I got distracted by reading life obligations and playing way too much Elite Dangerous, but I'm still going to do it. Just work on your version and I'll work on mine. I'll post the results once I'm done. I want to try a couple of things out, like offloading the work to another thread and putting the map data in a Hilbert curve.
That isn't multithreading, but time slicing or preemptive multitasking. Multithreading means letting multiple "parts" of your application work independently and in parallel. For your A* implementation that means moving the algorithm to another CPU core and let it work until it gets the result. It then notifies the game's main loop once it has said result regardless of how long it took. They work at the same time rather than taking turns.
The only worthwhile improvement one can do to the default C# dictionaries in terms of performance is implementing a data-oriented dictionary. That means no key/value-based but two lists (one for keys and one for values) instead. I'd fix all the existing problems before doing that though.

Am I correct when I assume that you don't want units to intersect? That's difficult and is not something you solve in A*. That makes the map dynamic (whether a tile is passable sectors changes while the algorithm runs) and A* cannot handle dynamic maps. There are pathfinding algorithms for dynamic environments, but they're slower and way more complicated than A*. I'd solve that problem outside of pathfinding, while your units move. A* gives you the paths (which may or may not intersect), but a unit only moves, if the next tile is free.

(checked)
Thanks, you probably remember me from my other project, which has been stopped because i want to make it 3D.

right now, when a unit reaches the next node in his path, he checks if the next one is occupied by someone else, and recalculates his path accordingly

A* absolutely can work with dynamic maps. The drawback is that the path must be recalculated potentially every time the map changes, but infrequent changes and having different resolution areas/maps for it to traverse can make the impact not as bad.

You dont want to calculate 1000 paths every frame at 60 fps and redo it every time the player does something, thats just stupid

Good work guys.

mhm finished another one.

But still my question still stands, since md3 can only handle vertex animation and it seems to support per object animation as well, what's the best way to deal with it when I want to use skeletal animation instead?

Can I bake the animation action somehow so it becomes per object based?

what's with the stretch man

man there is nuffin much I can do, the only "possible fix" I could think of is by using a large texture scale or so.

;u;

kys

...

...

If you want to give more weight to the punches, add some more frames when it hits to simulate inertia

but user, I've been doing that for five years, and it sucks.

Holy fuck nigger these look amazing

Thanks famalan.

though I appreciate the advice, I doubt I'll touch those again. Part of the reason why I'm moving to 3D is that traditional animation just burns me out. By the time I've done roughs, cleaned up roughs, clean line work (with consistent line weight), and color I feel like I've done an entire project, when really I've only done a couple of seconds of animation.

Of course, that advice should carry over to 3D so I'll remember it.

lmao smh tbh fam

Has anyone here written their own scripting language for their game engine?

i made a node-based editor for unity, like unreal's blueprints, but a lot more shitty
closest thing i've done

It's been like 3 threads already just fix it user for god's sake

i cant I think godot is borked

I haven't even tried godot, so I can't really tell, but if you haven't advanced much on it you can always switch to something else like Gamemaker or LÖVE, which are both reliable, easy to use and good for anything 2D. you could also try construct 2 if you suck at programming

no

Sounds unlikely. What was your exact problem again? As in what parts are in each scene?

ok

I can only detect collisions when things are hitting other things with their upper areas

That is not an answer to what I asked.

wat

I asked how you set up your scenes. Hell, if you really wanted help you would upload your project somewhere. The only thing you've been saying these past few threads is "collisions don't work, pls help ;A;" instead of doing that.

the green blocks are spawned by code but apparently taht doesn't matter since I set up a separate scene where it wasn't spawned like that

the greenblocks/money are a separate scene, and the player has code in it to detect when it's hitting them, which is the following:

if get_collider().is_in_group("money"):
get_collider().queue_free()

the player just have movement code which is the following for every direction/key pressed:

if Input.is_action_pressed("player_up"):
player_pos += Vector2(0,-10)

the player's collision box is a collision polygon 2d, but I switched it and tried collision shape 2d as well and it doesn't change anything. so now i just switched it back, same issue. The same goes for the green blocks, only they are currently collision shapes. It doesn't seem to matter.

The money is a rigidbody and the player is a kinematic body, I messed with their properties a bit but nothing seems to get it to detect collisions normally.

Publishers found a project, markets it and distributes it. It's basically your studio leadership.

Warning, publishers often want your IP and about 20-40% for publishing your game.

Meaning after steam you'll be left with about 1/3 of the profits. and you can't earn money on merch… or sequels.

user… I'm sorry…

Games have an avg yearly salary of about $40-60k in my country, while AI programmers for other software is around $60-100k.

Friendly reminder to license your shit properly or sleazy fucks like will try to hijack and claim the rights to your WIP vidya.

it's not like A&C is ever gonna be finished

but if you peddle your own games you're not a slave to the system rite guise? rite? (and I sure hope you're reporting your taxes goy!)

This turkish motherfucker…

get out of our thread you animal fucker

I'm not that Turk, faggot.

So how do you prove that you made/own your assets? What license is preferred?

a new level for my game; the next version will come later though, since there needs to be more improvements to justify a new release

As long as you make sure to break even you're not paying tax, and when you have employees you actually can get about 20% of the costs refunded or as a tax cut.
Stop shilling for publishers (((user)))

Only two more implementations and i will finish all pan recipes…

hen oven, fermenter, and raw prepared foods
Im so fucking tired

...

its C# dipshit
And fuck you Java is just fine

There is no way that's the most efficient way of doing that.

each recipe is slightly different, if i make one huge generic method with tons of parameters it won't be efficient

cant even tell the difference with that shit syntax

why on earth are there different functions for each dish if the parameters are all the same?
even if the recipe is slightly different you don't manually bake it all into your program like that
just load it off a text file

recipe r[] = {
#include "recipies.txt"
};

jesus

each one changes the ingredient nutritional value, taste, quality, how fast it cooks and the overall "deliciousness" in different ways at set temperatures, usual goes like this

damage the food (loss of nutrients)
—- high limit—–
safe to cook
—- ideal temp —-
safe to cook
—- lower limit —–
damage the food (loss of taste)

it differs in several ways for each type of food being cooked

It also only detects collision if the player is moving upwards

I kbow i asked for advice about how you were handling recipes but uh… I'll figure it out on my own

I missed when you asked about the recipes, but its fine if you don't want help

I'll admit I'm completely clueless about how licenses work, aside from that I want my game to be FOSS on release.

that's shit you store in a text file and load, not shit you compile into the game. What the fuck is wrong with you? You're doing a ton more work so you can code like a retard.

The recipes are not fixed (tomato+onion = ketchup), you can add anything into the pot and the game will try to figure out what you want to make and modify each ingredient slowly over time until you turn the fire off

Thats why they are methods, not simple strings or arrays of item IDs

.

im trying to go for a realistic cooking system, not the usual combine A+B to get C

that's fucking shameful

alright, I'll give you the benefit of the doubt here
can you upload that script to bastebin or something so I can tell you how to rewrite it using 1/50th the code

I… what?

also body enter/exit signals only work when I have gravity turned on, and even with that, they still have the same issue.

This is how i make soup
pastebin.com/e5tuHpf5

...

its here bro

Why only soup? Your point was that there's no way to have one class that reads all the data from a file.

How are we supposed to know if that's true if we only have one class?

because thats enough to prove it is not just a fixed recipe you can get from a simple text file
unless you want me to keep all my code in text files the game has to load and compile every single time

Born too early to solve quantum annealing
Born too late to get shekels from minecraft or bitcoins
Born just in time to cook dank food in java

...

...

Silly user, C# isn't a food.

It really isn't. Everything you listed could be be represented in a simple text file. Now, whether that could apply to your other classes remains to be seen because you only linked one fucking class.

The data in text doesn't have to be compiled and believe it or not, compiling everything to classes doesn't mean they magically don't have to be 'loaded'.

You're a shit programmer who single handedly gives java bad name.

you can't escape the bully now user

I'm not a pajava so I'm not sure if this is correct approach but couldn't you have a McFastFoodBeanFactory that just spawns food items one by one with some predefined fields? Jesus christ let's stop laughing at the poor chef and help him out here.

low effort billy

did you forget its C# already?
If you are so good why don't you prove it?

are you complaining about what exactly? the console line? im still testing this shit nigga

we're offering to save you like 6000 lines of code that'll be impossible to debug but you're arrogant enough to think that these developmentally delayed programming techniques are the absolute best way to do shit

I already gamedev for a living, nigger. you're the asshole who needs to feel good about yourself

It seems like i just hit the spot

How do you not get a gamedev job and not get fucked by the shit industry and shit work conditions

it beats working a minimum wage job while I get my degree

Nah Java is shit and C# has all sorts of unique features like LINQ and anonymous methods and code inlining.

Java also allows every class and function to be overridden by default which means more work for you to make your code secure and predictable. You could introduce a massive security hole in your game or network because you didnt declare something as sealed/final. C# is smart without being restrictive and lets you focus on the goddamn code itself.

still, doesn't make it as bad as this old meme makes it
but still, im working with C#
mostly for unity compatibility
>inb4 OMG UNITY

You can't possibly be talking about the
"100f / cycleTest" part showing up 3 times right?
that shit is insignificant

jokes on you i ALREADY have my degree, and my job is not minimum wage (its still hell and i want to quit)

>public Dish ProcessMeal(Ingredient[], DelegateMethodInstance)

Where ingredient is an object type that contains all the per-instance data such as nutrition, temperature, etc. You giveit a reference to an existing method which blindly processes/transforms a given set of ingredients based on settings

At least you chose C# over Javascript for Unity

literally curry-tier coding

its really good with rice

OMG UNITY

I think you learned to NOT post code here unless you have a problem you can't work out, if it works then we don't need to see it, programmers here can be pretty autistic about efficiency

Daily reminder that OOP is a meme

Don't hang shit on this user I got what I wanted and it isn't a final product anyway. Textures aren't relevant, just the model. If you guys wanna criticise it how about being productive and saying what could be improved or what specifically is bad?

I don't really mind, triggering autism is fun
I know i still have to improve a lot, programming is my worse skill by far, im really good at creating content, but i need a base built before i can get to that

Oh, lord.
General reminder to all that while yes, OOP is arguably overused and has become somewhat of a meme, preventing this is why it exists. When properly used at least.

But does he want help?

Also, one answer is to make a dish parent class with all the common data and a cook method, then overload the cook method in child classes as need. Then dishname.cook() or cook(this) can be used as a general call even if the methods differ.

Maybe I missed something but I don't see a class, just a pile of singletons. The soup code can't be a class, it contains a foreach loop and returns a a pile of stuff, so presumably he's calling each dish as an individual function and initializing the variables every call.

I thought about this too but he claims every dish has a wildly different cooking method, if that's really true (I kind of doubt it), you'd wind up having to write nearly as much spaghetti code since the existing method would have to have a completely different case for each set of ingredients.
Your way is still what I would likely use myself.

I can tell you that I have and probably most of the other people telling you that there is a better way to do it have probably done something kind of like this and learned the hard way that you can structure your code better.

It seems fine at first but once you have to come back to it and work with your base for several months while you make your game you will slowly grow to hate it, and you'll be stuck with it.

how does OOP help with this? You just need to structure your code so that a few generic functions can deal with all the recopies.

you deserved to be bullied

Interestingly, if user changed his code slightly and converted the arguments into a RecipeArg class, he could leverage generics and even enable his functions as LINQ extension methods

I fixed this by giving the collision box Area2d as a parent, then connecting a body_enter() signal to delete the money when it enters the area

it just works

I thought I answered that pretty well in the rest of my post. Again, I'm going on the assumption that the methods for each recipe are as different as he claims, in which case a generic function that deals with every recipe will balloon just as badly. I don't really believe that's the case, but if it hypothetically were, it would give him a simple way of keeping separate methods for separate recipes to tinker with debug as needed while still being able to call them with a single method name in a main loop.
Personally I've been focused on C for the last two years and have mostly move away from OOP and for a while heavily preached and practiced DOD, but I've been forced to acknowledge that OOP keeps certain people from shooting themselves in the feet and/or ass.

I'll have to trust you on that one, I never got into LINQ much, unfortunately.

you don't need to bloat a giant function, if there are specific things you can just pass in another function as an argument for the non-generic parts of the process. (like qsort in C)

I was asking because I program in C and dont very much understand OOP.

It's useful when you're dealing with a large-ish list of objects in a collection that need to be interacted with or processed. Eg, what user is doing. It works by using SQL-like syntax to choose object instances to invoke methods on, but it's truly raw C#.

Imagine doing stuff like

Or whatever the syntax happens to be. You can inline it, you can pop that on a single line, you can do all sorts of fun stuff.

so you're literally the college graduate java programmer sterotype? please tell me you aren't Indian too

Stereotypes exist for a reason.

R8?

It looks orthographic in that first screenshot for some reason

That would be because it is in orthographic view.

I really like it

Do the Terminator one next (Winchester 1887)

need higher poly count, I can see the surface normals

Well what the fuck bro, why you be askin what it looks like when it's in a perspective type that makes it look all crazy?

goy/10

next time you try to post bait don't make it so obvious
make your polygon density more consisten

That's the thing, you're still sticking yourself with the issue of coding the conditionals into your general cook() function. Not only can this lead to Yandev code, but you're hardcoding those function names, so any time you want to add one you have to continue to add it to cook(), and if you want to remove one or change the function it refers to, you have to track it down and alter it, each time.
In OO languages, classes inherit methods and variables from parent classes. But they can override those methods with a method of the same name. When it's called the compiler will use the more local method (in the child class), if present, or default to the parent's method if no overridden method is present in the child class.
This basically functions how you describe with the key difference that the calling code doesn't have to know the difference at all between the cook() methods of the different recipes so there is never an issue of saying "if this call this recipe if that call that recipe."
The calling code calls dish.cook() and the dishes contain the differences in cook() methods. And the calling code doesn't have to be altered to accommodate new recipes.

breddy cool

There.

Where is the density fucked? I assume it's the butt
What's wrong with paint?

have you tried making textures in paint?

or if we back up, have you tried UV unwrapping? :^)

working on a new level, with some new textures

I don't make textures in paint. I just didn't feel like booting up gimp to take a screenshot.

I want to make a game based on an 8-way airdash which can be upgraded mid-level to give you multiple dashes.

The player character's only motivation, story-wise, will be fun and challenge (why wall-jump and airdash up this floating mountain? Because it's there!), so I want a character design that screams "adventurous" and "thrill-seeking". I was thinking I would go full shounen protagonist with it, but I kind of want to try and think of something a bit less cliche. Any ideas?

oh, and I'm not sure on the method of dashing yet; I'm thinking rocket boots, but if the character design warrants it, I can just say "he/she can dash just because he/she can dash, don't question it"

bretty gud

ur mom lol

See htat mountain, you can climb it

So I need to know something. I'm working on a recreation of the Tiny Chao Garden, and one of the big, important things I wanted it to do was maintain compatibility with the real thing. That is, if you raised a Chao in Sonic Adventure or Sonic Adventure 2 and popped it out using Fusion's Chao Editor(chao.tehfusion.co.uk/chao-editor/), you can load it into my game, export it later, and bring it back to SA/SA2, and vice versa. Chao data is represented in a binary file, and while Game Maker has binary functions, is it enough to make this shit work? Did I shoot myself in the foot from the get-go since GM only uses strings and real numbers?

Maybe I'm just overthinking this. I'll fuck around and see what happens.

Well first of all, do you understand the chao data? If you don't know what you're reading in, you can't do anything intelligent with that information.

As for whether or not you can work with the data that you read in from a file from GM, as long as it reads to some kind of concrete data type (char, int, etc) then you should be fine. Even reading in from a string is fine as long as you can work with individual chars

What are the best places to find an indie artist to comission for some preliminary work? Sorry but can't be from here.

Any starbucks in california

Just a small update for today. There was a lot of brainstorming and discussion among the group but not much coding. I was hoping to get the other two inspection screens finished but instead I only finished one. It's for objects that wouldn't make a lot of sense to have the player pick up and rotate. Things like wall mounted plaques, keypads, pictures, whatever.

FUCKING LOL

I was thinking about that, and I've been investigating it.
It's on the to-do list… as a non-essential feature though, so later I'll probably add it.

Although, I've made the fine-grained details only fade in at closer distances (lerping the per-pixel normal), and at a strength relative to the distance to the pixel position for how "strong" those fine-grain details appear.
So I'm avoiding aliasing, and also having the nice detail effects only when they're noticeably visible.

Hey, forgot to give you feedback for your game. Honestly, just played Radio Tower;
The character moves WAY too slow, unless there is a run button and I didn't figured it out, it's just too slow to be enjoyable, it feels like you are crawling. Jumping is faster so I found myself bunnyhopping most of the time.
Also, you slide way too much too, it feels like you are on ice.

At first I thought "oh man the guns fire waaay too slow", but then I realized that mashing the button is much faster than just holding it, is this intended? Because once I realized this, holding the button had no purpose at all.
The first weapon is a tad too weak for my taste, but I guess that's fine since you are supposed to use it only at the begining and when you run out of ammo for all the other guns
The Laser Rifle barrel is misaligned with the spot where the bullets actually come out, if you pay attention the bullets actually come from a little bit for the right of the barrel, it's incredibly annoying.
Also I don't have any idea of what you tried to make with the shotgun, but it doesn't feel right AT ALL, it feels kind of like the S gun from contra, but with ultra tiny pellets and a shitload of damage falloff. I didn't enjoyed it at all, I would like it if you make it more like Doom's Shotgun, or if you insist with the "huge spread" idea, tweak it a bit so it doesn't suck.
In general, the guns all feel like the lack strenght for the lack of recoil, you should add at least a simple tipe of recoil so they feel like they actually pack a punch.

Sometimes if you shot at the green enemies (the first ones), they teleport to anywhere, this happens mostly if you use the zoomed right-click mode, this is incredibly annoying.

The death animation takes forever, you should speed it up at least twice as much.

I haven't played it that much, but that's all for now, I'll try to play the other levels later tonight so I can give you feedback on the other guns and such.

hehehe.. hahah.. HAHAHAHHAHAHAH
HAHAHAHAHAHAHAHAHAHHAHAHAHAHAHAHAHAHAHAHAHAHAH

oh man

...

Nice.

So I'm trying to find a way to make a rotating object with a maximum rotational acceleration smoothly accelerate and decelerate such that it smoothly and optimally points at a target.
Getting it to point at the target is easy enough, and getting it to point at it smoothly if I ignore its acceleration cap is easy, because I can just reduce it's rotation by a percentage of it's distance to the target each second.
I think to do this right, though, I need some kind of physics formula.

The closest I can find is a set of kinematics formulas:
v = v0 + at
x - x0 = ((v - v0) / 2) * t
x - x0 = v0 * t + 1/2 * at^2
v^2 = v0^2 + 2a(x - x0)

Clearly I'm retarded, but I can't figure out how to actually use any of these for my specific problem. I can work these formulas for acceleration/deceleration in a straight line, but on a rotating object, things seem to fall apart.
For instance: (x-x0) will just keep wildly flipping between -180 and +180, and even frequently equal 0 as it blazes past the target direction.

Is there a better formula for what I'm trying to do?

I just spent 4 hours working on this problem before posting, to see if I could solve it. I found formulas for angular velocity and angular acceleration, no matter how I plug things in, the results are varying degrees of nonsense.
I supposedly need it to be in radians, and I have everything already measured in degrees, so I just converted all my degrees to radians my multiplying by 0.0174532925, but nothing works.

Half of my screen is covered in debug variables spitting numbers just so I can try to understand what the shit is going on. But I can't understand. I don't have a single clue what the fuck. I'm losing my goddamn mind.

I just want thing to spin good.

but there are no conditionals- you are passing a function pointer into your generic cook() function and then calling it inside that function

so, cook(other vars, &soup_function);

this way the specific function for the thing you are cooking can be part of the players state. It doesn't seem to be any more complicated than using other classes, since in each scenario, it would work like this:

cook(&soup_function);

//OOP way?
soup.cook_function();

in each scenario you have to be aware on some level of the thing you are cooking- so I don't understand the complexity of that

to be clear, the cook() function is accepting a pointer to the soup_function function as an argument and then calling it internally. but the idea would be that you could replace it with other functions for other dishes. That's why I referenced the qsort function because it accepts a function pointer as an argument

No you don't, you are purposely overcomplicating it and you are drowning yourself in complicated physics that aren't really necesary, 90% of the time the simplest solution is the best.

Also, I suck at math and physics, so even if it was the right way to go about it I wouldn't be able to help you, sorry

If there's a simple way to do this, I'm going to jump out of a window. I've been working on this single problem for a total of 18 hours today. I haven't even eaten.

Quaternions. Depending on your engine they may already have systems built in to handle them. Here's an OpenGL tutorial about them, but googling for your specific case should help you.

opengl-tutorial.org/intermediate-tutorials/tutorial-17-quaternions/

Well I mean Schrödinger did adapt Quaternians for quantum mechanics…

slerp (spherical lerp), maybe?
I've done that before, but it was awhile ago; I used a mixture of slerp and something else… can't remember exactly.

keithmaggio.wordpress.com/2011/02/15/math-magician-lerp-slerp-and-nlerp/

This is just a 2D object.
I'll look through the quaternions tutorial now, but I'm seeing lots of z all over the place. Don't know how useful it'll be.

I've thought of using that to smooth the deceleration for sure, but I need to crack the first part of the puzzle first.

Object is spinning at 588 degrees per second. It can decelerate at 18 degrees per second. It needs to be pointing at a heading of 245 degrees when its velocity hits zero, with no over-shooting or corrections.
It sounds simple, but the numbers always come out wrong, probably because I'm using linear equations and not considering the roundness of it all.

Either way though, thanks anons. Whether or not this stuff helps me out, I'll at least learn something new.

This might help.
euclideanspace.com/maths/algebra/realNormedAlgebra/complex/transforms/

I've read through this and a couple follow-up pages on rotations and circles and the like.
Good lord, this stuff is borderline impenetrable. At leas the physics calculations don't come out of the gate talking about "imaginary angles" and "complex numbers".

I've gotta degauss my brain. I'll re-read this stuff and see what I can make of it in the morning.
Thanks, user.

If my game is free to play and nobody plays it, i'm doing something wrong

...

now that I think of it, look at tweening/easing algorithms.

Ignoring priorities to work on game.

But bro, working on game is ALWAYS top priority

If you wanna use acceleration you need to solve a second degree differential equation
I would suggest you don't try and just set the speed and ignore acceleration it's not really worth the trouble

It's generally worth it to put as many variables in text file as you can
Though it probably will not reduce the amount code in your game as some people are suggesting.
It's nice because you don't need to recompile your game every time you wanna change a variable and it helps a lot with modding and such.
Also it's just nice to have alot of variables in the same place means you don't have to go searching through code just find and change some variable.

So do you understand complex numbers? They're pretty damn useful. Normal numbers exist on the 1D number line but complex numbers exist on a 2D plane. Think of a complex number as a way to express an x and y coordinates on that plane but with a single number. Because we can express coordinates on a plane with one number we can do a lot of really neat math with it. For example mixing together complex numbers and trigonometry we can get smooth rotations.

So a complex number takes the form a+bi, where i is an imaginary number. a is the "real part" and b is the "imaginary part". Remember to think about it more like x and y coordinates. a is the x and b is the y. If we draw a line from any given point on the complex plane back to the origin that line represents a rotation. Lets make a little cheat sheet.
0+0i is (0,0) so basically the origin.
1+0i is (1,0) or 1 unit out to the right of the origin, basically 0 degrees
0+1i is (0,1) or 1 unit above the origin, basically 90 degrees
What about things between? I hope you paid attention in trig! It's time to break out your unit circle and trig functions, cause this is the magical equation.
cos(n)+sin(n)i is the point in space exactly 1 unit away from the origin at any given angle 'n'.
Have some wacky angle like 72 degrees? That's around 0.309+0.951i. Want some value above 360? No problem! 743 degrees is at around 0.921+0.391i.
So how is this stuff useful? Well in the link I provided (euclideanspace.com/maths/algebra/realNormedAlgebra/complex/transforms/) we see that if we want to rotate we can use complex number multiplication.
CR=C1*C2
Where C1 is the original, C2 is how much you are rotating by, and CR is the result. All of these are complex numbers.
So lets say we have something at 0 degrees, or 1+0i and we want to rotate it 90 degrees, 0+1i. Well 1 * 1i = 1i, or 90 degrees. so there we go.
What about something a bit more complex? we start at 10 degrees, rotate -32 degrees. So to start we have cos(10)+sin(10)i an we are rotating it cos(-32)+sin(-32)i. Plugging that into wolfram alpha we get 0.9272-0.3746i we can plug that into the other magical formula
arctan(b/a)
to get what that is in degrees. So remember a complex number is in the form a+bi.
arctan(-0.3746/0.9272) = -22
and there we go, we started at 10, removed 32 degrees, and now we are at -22 degrees.

But you just want a formula.
Initial rotation x
rotation to apply n
rotation in complex form = (cos(x)+sin(x)i)*(cos(n)+sin(n)i)
arctan((imaginary part)/(real part))=final rotation
I sure hope your programming language lets you multiply complex numbers and use trig functions on them. I know C++ lets you do it if you with cplusplus.com/reference/complex/
Also there's a good chance you will get your answer back in radians so if you need it in degrees don't forget to convert.

Ok so i understand your problem correctly
You wanna go from Angle A to angle B using only rotational acceleration
If thats the case don't listen to this guy non of what he said is really relevant
You need to do what i said here however if you won't take no for answer lets try to break it down a little bit
So we have a starting angle a1 and the angle we wanna reach a2
now you can write down equations like these
but these only work with constant acceleration so they pretty much don't work
To calculate the angle at any given point in time you need to use an integral
It's kinda hard to write down an integral without the proper symbols so ill just explain it
you need the integral over the speed from the point you start moving the current time to get the current angle
And to get the speed at any point you need the integral over the acceleration from start up to that point
Now if you don't how to solve integrals that's kinda not fine but you don't really need to either way.
You need to solve the second degree differential equation heres a good link if you really care en.wikipedia.org/wiki/Equations_of_motion
now this is for motion but the problem is essentially the same just substitute position for angle.
Now here the fun part even if you solve this equation your not there yet.
This equation allows you to calculate to angle at a certain time given a certain acceleration at each point in time.
So now you need to plug in your max acceleration into the equation and reverse it
Note that if you reach the halfway point you can just reverse the acceleration and you will be at a standstill at the end.

Now if all of this makes perfect sense to you go ahead and do it this way if however it doesn't
Just use constant acceleration at which point you can use
angle = anglestart + 1/2 * at^2
Say you wanna go from angle1 to angle2
angle2 = angle1 + 1/2 * at^2
you wanna start decelerating at the halfway point which is angle3 = (angle2 - angle1) / 2
(2 * (angle3 - angle1) / a)^0.5 = t
There equation solved for the time
though really in code you could just test the angle every loop till it's at the halfway point then reverse it also this does assume that the deceleration and acceleration happen at the same rate
You might get inaccuracies though depending on your framerate if you want to i can try to explain how you can fix that aswel

is this a fallout 5 exclusive feature
;^)

A game all about mobility and airdashes with non-cliche characters? Make it about a group of janitors :^)

...

Cooking user is now a meme

I hope you manage to find some way to keep that abomination in.

That camera is too dynamic for a horror game.

alright, units don't get stuck anymore
pathfinding officially works
silky smooth 60 fps, lowest so far has been 55
only way it can get better is if the other user gets multithreading to work, so we won't have a 0.7 second delay when pathfinding an unreachable target
i can finally start working on building buildings

Did any of you have any luck with making money of half asses android games? I really, really, really need to make money for one simple reason. To support myself while working on my actual dream game that I love.

I'd be willing to spend a two or three months making one or two android games to get ad money. But I see that there are waaaaay to many people already doing this. Any of you had any success making money in this way?

working on mobile is not worth the trouble
it's a pain to get anything working there that's fun and profitable

I don't actually wanna make anything serious. I look at all those retard three man teams on reddit gamedev who quit their jobs and in two years, TWO YEARS, they made… games worse than free flash games. I can do better than that in two months by myself since phone users are braindead and don't expect quality.

I just want to make any sort of money, even $1000 per game would be good enough since it wouldn't take much to make it. But can you even make this kind of money or even this is hard?

don't you have any dignity? any self respect?

I have dignity, this is why I'm poor. Otherwise I would had memed out a game like undertale or yandere sim, but I'm working on what I like, and honestly no one gives a shit about what I like until it's actually a finished product. And I'm okay with that.

But until then I have absolutely 0 problems making mobile games, I despise mobile gaming and if I could get ad money out of that to make my actual game then I'm good.

...

For a few months to make some money. I don't know why it's hard to understand, my dream game is a completely separate thing and is not related to some mobile games. I just wanted to see if there are people here who made some mobile games and made something out of it.

I made one for cutscenes. TBH hardest part was dealing with operator precedence. You can cook one up in a weekend if you're not too concerned about performance.

Oh god this is starting to look like wurm.

...

You're a fucking idiot.

well if you want to metaphorically suck dick, be my guest

I'll explain why you are an idiot and if you still can't get it then that's too bad. You don't know what dignity is, working on something completely unrelated just so you can avoid sacrificing your main project that you care about is dignity, if your main project makes no money. Dignity is not half-assing your dream game so you can get money faster. Do you get that?

Coding some flash-like games for android for a month or two is not "sucking dick", I've heard enough of these "sucking dick" and "eating shit" comparisons on cuck chan, stop trying too hard. You probably have more mobile games than me, I played a few hours of mobile games in all my life and I don't like it. So if you want to get into a dick waving contest about who has more "taste" in their platform you should had picked another person. Do you honestly think that I want to make mobile games, that I enjoy it? I hate it, and I just want to have money so I can work on my actual game. I'm not a corporate fuck so I don't have a career or a great job, so even ad money would be helpful.

It's been ages now, but I don't think I had a proper trigonometry class. I had a class called "geometry" and then a class called "pre-calculus". In both classes we mostly just watched "edutainment" videos while the teacher read a book. Then when I got to uni and tried the basic introductory calculus course, my professor was an autistic mute who just wrote obscure shit on the board without explanation. The entire course was graded on a 65 point curve. All of that, combined with me being naturally shit at math, means that I basically have zero maths background beyond basic algebra despite 5 years of post-algebra maths classes.
Blog aside, I think I get what you're saying here. How exactly does that end up working, though, assuming a random starting rotation speed and a constant acceleration/deceleration rate? C1 is my current angle, C2 is the total angle i need to traverse given my current rotation speed and my fixed deceleration rate, and CR is the angle to wait before I'm properly aligned and can start my constant deceleration?
Or does this not take motion into consideration at all? Like all higher math, this seems really simple when I look at it, and then locks my brain when I actually think about how to apply it.
I'm also fairly sure my programming language can't do jack diddly shit with complex numbers, but I'm stubborn enough that I could probably find a way to cram it in once my brain properly parses it.

My acceleration is constant. At least, I think it is. If the object can only decelerate at 2 degrees per second per second, that's constant, right?

A better explanation of the scenario is as follows:
A ship is in space, and gets hit by a large mass at an odd angle. This gives it an uncontrolled spin of rate 'x' degrees per second, as well as a drift in direction 'd' at 's' meters per second. Its onboard maneuvering thrusters can only output a thrust rate of 2 deg/s/s, and it can't correct its drift until it first corrects the spin. If the drift is fast enough that it needs its main engine to slow it down, it has to arrest its spin pointing opposite its current direction of drift (if the drift isn't terrible, then it can be ignored and weak lateral thrusters can correct it). I'm basically trying to create an "emergency stop" script for a ship in null gravity. I could cheat and disrespect the ship's thruster power for this purpose, but that feels really lazy and hacky.

I remember differentials and integrals vaguely, and I remember them not being too hard to work, though I was never given an explanation as to why you do them, what they're for, what they represent or anything. It was just "here's some voodoo you can do to an equation when you're explicitly told to". In fact, when I went into the professor's office to ask him to explain the "why" behind all this shit, he kicked me out of the class.
Now we're in physics territory again, which I've always been more compatible with, because physics is "math"+"here's why math". When math starts using abstraction, my brain starts having seizures.
I've been using pretty much exactly that equation, though, and keep getting results that are off by dozens if not hundreds of degrees. Do I absolutely have to use radians for these calculations? I have absolutely no idea what a radian is, and when I convert to radians, my results get much less accurate.
I've started using delta time to calculate time since this problem first cropped up. For now, I'm running my calculations in both delta time and frame time, separately, and the results are always consistent between one another once I covert frames to seconds or vice versa.
I'll probably have to refactor the whole game to run on delta time instead of frame time soon enough, but that's a nightmare for another day. If you have suggestions or better solutions there, I'd be very happy to listen as well.

...

You won't earn anything per game. All money made by mobile devs is made by the big mobile publishers, indies only get plays if their game is free.

I remember reading something about the few popular apps making 90% of everything and then then the rest of the 10%, which is the majority of apps, dozens of them, make very little. This is why it's hard justifying spending two months of my life just trying my luck on it, since it's essentially a luck game because mobile games are complete crap.

But yes, I was talking about making them free, so you can just implement an ad or two when you start the game. I wonder how many people get less than $100 per game that's free.

Congratulations, you just went from $0 ever to $2 per month if you're lucky.

games are not easy money in any way
honestly youd be better of making a meme game for pc and just pushing it heavily on retards/youtubers

This. Getting into games for the money is like getting into any other form of artistic medium or entertainment for the money.

You're better off getting a real job if you need cash.

Pre-calculus is the faggot name for a trig class. From what it sounds like it wasn't taught properly though.
Differentials/derivatives are used to calculate the rate of change in a function at a given single point. You know how in middle school math you learned how to get the slope of a graph by taking two points and dividing the rise by the run? The derivative of a function basically lets you do the same thing, but with a single reference point (which is impossible to calculate with the middle school method since the rise and run with a single point are both zero). It's especially useful when you're dealing with quadratic equations, since those tend to not be straight lines (so each point has a slightly different "slope").
Integrals are just reverse differentials, also called antiderivatives. To a certain degree, they allow you to calculate the original equation from a derived equation. There are two kinds of integrals, definite and indefinite. Indefinite integrals can only reliably determine the quadratic and linear components of the original equation (a(x^n) + bx), but not the constant component (since the derivative of a constant is zero). Hence why they're indefinite - integrating (x^2) + x, you wind up with (x^3)/3 + (x^2)/2 + C because ANY constant added to that equation will still result in a derivative of (x^2) + x.
Definite integrals are calculated with an interval, and require a few additional steps that I don't really remember at the moment, but the short story is that you plug both endpoints of the interval into the integrated function, subtract, and that gets you a constant. I'm rusty there though so that's something we both should brush up on.

This actually has nothing to do with trig at all it's mostly calculus
Yes
Ok assuming a random starting rotation speed makes things quite a bit more complicated actually
Now you will need to do some calculations
angle = v0 * t + 0.5 * a * t^2 where angle is the difference in starting angle and ending angle v0 is the starting rotation speed a is the acceleration and t is the time
we want to stop the rotation when we reach the correct angle
To find the speed at t we use
v = v0 + a* t so enter 0 for v and solve for t
0 = v0 + a * t
t = - v0 / a
so this is the amount of time it would take to stop rotating.
Now you put this t into
angle = v0 * t + 0.5 * a * t^2
lets call this anglestop
You will pretty much always undershoot it unless it will land exactly on the right angle so you will want accelerate for a short amount of time first
You can calculate this by using this equation again
angle = v0 * t + 0.5 * a * t^2
0 = v0 * t + 0.5 * a * t^2 - angle
but this time angle = (anglestop % 360) / 2
now to solve for t we can use the general second degree polynomial solution.
t = ( - v0 +- (v0^2 - 4 * 0.5 * - angle)^0.5) / 2 * 0.5
and that's pretty much it i think
again you will have issues with frame inaccuracies i don't really know what you mean by delta but you will need to do some kind of rounding to get this to work consistently

nu bread

I think I just might be able to pull this off now. Thanks, anons. I'll report back in a future thread once I've cracked it.

Har Har.

I do plan on taking some inspiration from Dustforce but only a little. The big similarities are freedom of movement (in fact I want to do MORE freedom of movement than Dustforce; DF tended to be mostly linear and occasionally went into "corridor lined with instakill spikes" territory, which is the first thing on the list of things that my game is NOT going to do), the goal of "reach the end as fast as possible", and general platform design. The differences will be a bit more marked. For what I have planned, you don't get any attacks; enemies only exist to be avoided (and since I'm not doing spike corridors, dodging enemies is going to be the main method of challenging the player). Also, no double jump, just an 8-way dash which refreshes when you land or cling to a wall. Additionally, each level will have single-level powerups. You start each level with one dash per jump, and can acquire more, but reset back to one every level; some levels will be designed to be completed without the maximum number of dashes.

don't want to stop being a neet, but being a neet is fucking shit.

Fair warning, software engineering focuses on management and project organization. It's more akin to a business course with a CS focus. On the plus side, people like the term Systems Analyst.