/agdg/ - AAA Game Dev General

----[ Recap ]----
Game: Seasons Past
Dev: ambientdev
Tools: gamemaker: studio, paint
Web: ambientdev.itch.io/seasons-past
+implemented buildings with shadows
+figured out how rooms should look
+basic structure of the game is there
-still need to figure out how i'm going to spawn enemies in terms of game design, not implementation
-need to work on animations

>just found out Game Maker doesn't have weighted tiles for pathfinding

Truly, this is the best time to be alive.

anime was a blessing

Made gun boxes!

Now make a game

cant do it

We'll laugh at you when our games are done, and you'll be left alone, even /agdg/ won't be there for lazy nodevs.

So I need some AI chatter for my little game. Particularly things the AI will say when

>engaging
>retreating
>bitching about being hit
>death screams
>firing
>aiming
>repositioning

Here's a video of what I have so far, with each action having roughly four randomly selected lines with a 60% probability of playing at all, with the firing/aiming chatter set to 20%. The goal is to give the player some indication of what the AI is thinking, and making the player feel less lonely when not playing online, which we've mostly finished implementing. Sort of, I mean the physics are fucky but that's part of the charm, I think.

youtube.com/watch?v=SRQ_WrvEm9s

Can any idea guys help me out here? I have a whole bunch of other lines recorded for functions the AI doesn't do yet, but that's for later.

ignore the weird audio positioning and balancing, the AI being retarded and the janky fps please, I just got this shit working like eight minutes ago. I plan to record lines from people other than myself when I get this shit more or less ironed out.

>Going back to a project you haven't touched in months

Th..Thanks

>come back to an old project
>remeber THAT bug
>drop it

>>retreating
fug, I mean repositioning, retreating isn't in yet.

Anyone here with box2d skills?

>come back to old project
>remember THAT bug
>don't be a weak willed faglord and give it another go instead of giving up like a limp dick queer
>fix it in 10 minutes because of fresh mindset
>dev and have a great time making progress for the rest of the day

I dabbled. What up?

>listening to a dev song for almost two weeks
>hoping to send him energy to continue his game

I'm trying to do sort of

I'll take things that never happened for 200, Alex

That would work as a line

>RETREATING... I mean repositioning

It happens all the fucking time. You run into some really difficult problem, and then next morning you solve it instantly because you've given your brain time to refresh.

You'd know this if you weren't a nodev.

Well, there's one. Thanks senpai.

Jesus christ what the fuck why did backspace posted a reply

Anyway, as I said, I'm trying to do soft of one way collision respons response in a sense that one body reacts and the other goes on like nothing happened to have doors similar to ones in hotline miami, where character goes through, rotating them out of the way, but doesn't slow down.
Right now I'm trying to do it by modyfying box2d's insides, but's it's a slow process since I'm a brainlet, but I was wondering if it was possible to do with mass/density without going out of Box2d's "comfortable" boundaries.

>want to make a game
>isometric sprite simulation game, in the same sort of vien as theme park and other bullfrog games
>would also be sort of like dwarf fortress
>have programmed before, mostly in java and python, know the basics and have decent knowledge of GUIs
>no engines that I've come across seem to be good for the sort of game I want, all the games I listed seem to have custom C engines anyway
>not entirely sure what kind of performance this game needs. There's lots of entities with complex decion making going on, like in DF, but unlike DF theres probably not so much going on behind the scenes
>not sure what to do because making a game largely from scratch in C seems like a nightmare

What sort of language/library would you reccomend for me and this game? Should I just bite the bullet and do it in C?

>next morning
I explicitly said
>old project

Who here /nodev/?

I'm one, I'm just here because I want to support /agdg/™ games.

It doesn't matter how old it is, it still happens.

haha us gamers right

>>would also be sort of like dwarf fortress
[sluts laughing]
> Should I just bite the bullet and do it in C?

Just make it in unity or unreal. I don't think any other available engine has the performance for the kind of compute you're looking for.

You might be able to tell I'm awful at writing lines.

ohgodsomeonepleasehelpme

You either want sensors to make your own door-getting-hit callback
iforce2d.net/b2dtut/sensors

or maybe you could do something with collision filtering
iforce2d.net/b2dtut/collision-filtering

You'll have to fiddle with it a bit either way. Doing something with the door not actually colliding with the player but being a sensor which you manually apply a force to (equal to the player's velocity) when it gets triggered should do the trick.

Try going over some RTS games, they have a lot of that AI chatter when moving units around.

Use any language in the same level as C/C++, even Java would work if you know what you're doing, then find libraries/frameworks to handle the graphics and other systems for you so you won't have to go full engine-autist mode.

Big engines are bad for that kind of games because they have a lot of bloat that isn't very flexible and sacrifice performance for ease of use, not because they use slow languages.

>And you will know my name is THE LORD when I lay down my vengace opon thee

Where the heck is the android sdk root folder? Just refreshed my PC. Why do I always have a hard time looking for this.

>tfw only ever played supreme commander
>tfw the tiny fov of games like warcraft and starcraft litterally made me sick

Thanks for tip though I'll look up some youtube compilations

Nice.

For firing or engaging? I'm thinking firing as it has a much lower chance of playing so it won't get overplayed.

>even a great looking slick game like monolith is struggling

its over

Where you installed it

Add a charging shot and have some monologues like these play while it's loading

I just started yesterday and have spent most of the day creating awful shitty programmer art sprites, so technically I haven't been deving yet

I plan to use pre-existing tilesets so my game doesn't look like a complete mess.

docs.unity3d.com/Manual/android-sdksetup.html

Assuming you're on unity.

It's trying too hard to be retro.

Small screens and big pixels are nauseating.

I have absolutely no programming background, but I'm knowledgeable enough with photoshop and audacity to make enough assets to get by. Where should I start?

>paying to shitpost

somethingawful is that way fag

I need to state - Collide but idk what to type pls help

public void OnCollisionEnter2D(Collision2D colInfo) {
if (colInfo.gameObject.tag == "Player")

To be fair its pretty niche and doesn't really do enough new things to be worth looking at if you've ever played shit like binding of isaac.

Forgot your link

C++ and SFML

>unlike DF theres probably not so much going on behind the scenes
There's very little going on behind the scenes in fort mode, CPU-wise. All the fps hits are from querying and pathfinding against all the shit in your fort.

>There's very little going on behind the scenes in fort mode, CPU-wise.
I'd like to shitpost and say its mostly RAM bound, hilariously enough. Did some testing myself and it seems to be CAS latency that's the real killer. Damn game only uses one core and runs entirely in RAM.

>There's very little going on behind the scenes in fort mode, CPU-wise. All the fps hits are from CPU-heavy tasks.

...

The distinction I didn't make clear enough was that what's going on in the fort isn't "behind the scenes". The fort the scene, it's going on in the scene

>not paying to run a shitposting bot to train your neural net
I operate on another level

Makes me wonder if any AI can currently recognize sarcasm or ironic shitposting.

apparently there is AI that can recognize people's mood from facebook posts and tell if someone is manic depressive etc

Fair enough, but people on facebook are usually being completely honest in how retarded they're acting.

This

>binding of Isaac
I'm not a fan of monolith but unlike isaac it's a game that I can enjoy for the graphics and music.
Binding of isaac was a piece of shit and i'm sure it was popular because of meme 2edgy le farting loading screen.

>Just make it in unity or unreal. I don't think any other available engine has the performance for the kind of compute you're looking for.

coding the game as c# scripts running on top of the bloated mess that is unity is going to have better performance than c code? what the hell are you talking about?

not him but

>i'm sure it was popular because of meme 2edgy le farting loading screen.

Sorry for the faux pas, but my gf plays it on her 3ds all the damn time. I played though it enough to see all the content and it was pretty boring in terms of design,so I agree with it being shit.

She likes it because it's simple to understand and just challenging enough to be forgiving, but has enough variety to be interesting for a normie. The fast turnover rate relative to most roguelites is pretty appealing. Monolith is more a game for yesdevs and gaymurs that have a enough braincells to rub together, which happen to be in short supply.

I'm sure if polygon or some shit writes an article about how lolhard so challenging it is things will turn around.

>what the hell are you talking about?
Pre-made engines. If you want to enginedev fucking go nuts, but shit like godot or gamemaker aint going to cut it for heavy sim games.

Neither is Unity.

It'll be better for sure. Cities skylines is in unity, remember. I do agree a totally custom engine would be worlds ahead.

In that case Godot will be even better when 3.0 comes out and they ditch their scripting language.

>when 3.0 comes out
In that case it would litterally be better to make your own tools than to wait for some other twat to make a release unless you're jammed up their ass on twitter or whatever.

...

Godot 4.0 will be out before you're done with your complex simulation game, it takes years to make games that aren't shitty arcade games or Unity memegames.

>your
Not him, I'm arguing for reason. You also shouldn't be spending years alone unpaid on niche games, that's silly unless you're doing this as a hobby and a hobby alone.

You shouldn't be making games if money is what you're after and aren't willing to work for a long time before you get anything for it. You're living in delusions and lies if you think this is a good or reliable career option.

Also my point was waiting a few months for Godot 3.0 is not exactly a bad idea.

Why are you shilling that much Godot? What's so special about it?

We have gamemaker for 2D and Unity or Unreal for 3D.

This is enginedev edition why are you discussing engines

>not making a memegame and getting cash
steamspy.com/app/34900

>Also my point was waiting a few months for Godot 3.0 is not exactly a bad idea.
Okay, this is a reasonable stance if you're willing to sit on your hands about devving for so long. I suppose assets and design docs can be made in this time, but my point is

> it would litterally be better to make your own tools than to wait for some other twat

Waiting for someone else to do something is like waiting for the sky to fall.

The triggerhappy use of 'shilling' bothers me. He isn't shilling. He's discussing the merits of using your own engine or using a more suitable one for the task at hand.

Unity is a bloated mess, but its the "best" premade solution in terms of pre-made engines for the task the original guy was trying to achieve. Unreal might work as well, but I'm not familiar enough with it to make that deliberation.

I like your aesthetic man! Shadows look nice!

>at least half a million dollars gross
>roughly 300000 in revenue, counting tax and steam's cut roughly
>for fucking bad rats

if it's just 2d sprites and tiles then coding the graphics would be fairly simple with opengl+some wrapper for it like sfml or sdl. you could possibly get away with just using the built-in sdl functionality for drawing sprites etc and never touch opengl itself. it wouldn't be like writing it "from scratch" because you never have to touch the windows api or anything, you just have a bunch of simple functions to make a window, check the inputs, draw a picture.

in terms of the gameplay you're going to have to code this shit one way or another. if you have basic programming experience then i honestly think just coding it yourself in c is going to be more straightforward when you have a very specific thing you want to make. just iterating over an array of structs (or an std::vector of objects if you prefer c++) and updating them is more pleasant to me than dealing with some elaborate component entity system that unity has to have to support every single game under the sun. check inputs, update the game state, draw to screen, repeat. you know everything that's happening because you wrote it.

your biggest hurdle coming from java/python would be having to deal with memory yourself but if you envision this game getting complicated and computation-heavy then it's better to be aware of how things are stored in memory from the start.

Working in the games industry is not fun and games. Poor pay, chronic overwork and unpaid overtime. Don't forget crunch time either.

Better off developing small indie games and hope they go gold instead of becoming a mindless drone in a large corporation or business.

what happens if i pirate gamemaker

Are you shadows dynamic, baked or part of the sprite, user?

you can't because it's still new and there's no torrent

Then you should pirate Fusion instead.

I fully agree, user, my point was that you shouldn't wait for a product to update to start work, and enginedev is really god damn tedious work.

i meant the old version not gm2

kcracks.com/2017/01/10/game-maker-studio-2/

Not sure if this is it but I didn't care enough to look further.

the building creates a separate object for the shadow, and the shadow sprite is pre-made

they're not dynamic right now, like i could make them scale based on game time but i don't think it's worth it

>Pre-made engines.
there are libraries and engines for c/c++. there is a lot of ground in between pure "enginedeving" and "just use unity" and it's that area that the guy was asking about. if your knowledge is limited to unity vs godot shitposting then don't reply to questions about c libraries.

Making your own engine is so much more work that you'll probably spend more time tinkering with that than waiting for Godot.

That's not my point anyway, if you think you can make your own engine then more power to you. My point as I recall was if you're not going to enginedev that then waiting a little for Godot is probably better than using Unity if you want to make a performance dependent simulation game.

no, user.

What do you mean by bloat? As long as he doesn't use all the features it should be fine, right?

I know this is a retarded question I'm just asking to learn why engines have bloat. What is the benefit of creating a new one?

The thing that gets me most about AGDG is that there are so many people here working on their own little projects, 90% of which will never see the light of day because individuals just don't have all the skills to make it happen.

If all the disparate artists, programmers, designers and modelers actually got together in teams we would see much better games and a monetary return on effort.

Enginedevs are the worst offenders. An exercise in futility. Their programming skills would be far better spent assisting other devs making games, rather then just programming problem solving.

>Enginedevs are the worst offenders. An exercise in futility. Their programming skills would be far better spent assisting other devs making games, rather then just programming problem solving.

Have you considered they don't have the skill to program functioning systems, so they turn to desperately hashing something out that justwerks, over making a fun game?

Unity is designed to be comfortable to use general purpose engine, the more you generalize something the more you sacrifice performance. High performance is usually obtained by specializing your code so it does one task well and does not try to accommodate other behavior with it or be safe so that retards won't break everything with their drag and drop thumbs.

Also, since they're general purpose, they typically employ very advanced 3D/physics mechanics that are slower but allow for something much fancier looking and flexible. They engine probably includes all those fancy 3D effects and systems into your game even if you make a 2D pixel platformer, that's why you end up with 50 megabytes even when you make a small and shitty game that an engine dev could make in 200 kilobytes.

If you want high performance with Unity, you'll probably have to rewrite so much functionality from scratch that you'd be better off just making your own engine with smaller libraries or frameworks.

how is using sdl or whatever to draw some sprites to the screen "enginedeving"? that's just making a video game. you get sdl, you put a few abstractions of your own on top of sdl over the course of a few days and here you go, you're making your sprite game now. what are these months of "enginedeving" labor supposed to be? i swear to god everyone who uses this retarded term is totally clueless.