/v/ - Video Games

>>15993767

The best time to be devving. By the time you finish your game the crash will be over.

>>15993767

>Letting gloal economic factors influence your autistic passion project

Making a game is an absolute waste of time. Go write something useful like a compiler or a renderer. And for God's sake don't use some proprietary or huge ass engine and/or library.

>>15993767

The best time to invest is the bottom of a recession. Like wise, the best time to make a game would be in the dip when AAA is weak.

You got your demos ready anon?

>>15994139

>>15993767

They're not denying that a crash would occur, it's just that they acknowledge that it would be beneficial to them because other devs (especially AAA types) will be brought to a lower level. Easier to compete

>>15994139

I just think you're cuckchan cancer that needs to go back where you came from.

>>15994006

Nope!

>>15994139

>cuckchanner does not understand history

The crash, as it happened, only affected AAA. Hobbyists still remained. Hell, gaming was barely affected in Europe, since PCs were the majority of the market there, and many games were made by a single dude, who sent a floppy to a publishing firm.

Would the buyerbase be affected? If you're normalfag garbage like Undertale or a Hat in Time, sure. But if you're small time, and your work spreads from hobbyists to hobbyists on obscure indonesian clay-sculpting forums, you won't be affected, since the normalfag market affected by a crash was not going to give a shit about your game anyway.

>>15994006

I wanted a demo to be out but it's too much left to do in two weeks, especially since it's a networked game and I haven't figured out how I want to do public servers. You can connect to others via IP, but I don't think anons want to hand their IP to other anons on a Vietnamese manga social media platform.

>>15994158

A crash of indie would also be good. Chase out the cucks and trannies and replace them with people who want to make comfy vidya.

>>15993767

>>15994139

So are you reddit or cuckchan?

>>15994361

High impact games tried that. It didn't work. Now fuck off back to reddit.

>id+ filter

>2/3 of the thread disappears

Great start to the thread.

>>15994520

People need to just ignore smelly dumb wojakposter scum.

>>15994520

>>15994524

cuckchan wojaks should be a bannable offense if they aren't already

>>15994537

They usually are, but the mods tend to leave large hours of non-activity which leads to Bruce spam and Aaron posting his own butthole everywhere.

>>15994542

That's are because the mods are all mark pretending to be other people inb4 criticizing him suddenly wakes him from his slumber of ignoring cuckchan

>>15994834

>Patreon

Nobody here is going to give you money. Try cuckchan or reddit instead.

>>15995091

Nigger I am programming a game in fucking MATLAB of all languages did you really think I was seriously begging for money for this piece of crap

>>15995094

>tfw your game is so good that /v/ thinks it's too polished for patreon.

>>15994834

>Patreon

Kek. Is there actually any non-lewd videogame that even makes money on Patreon?

>>15995392

Nice doom clone брат.

Posting previous thread number for archiving purposes: >>15940593

>>15994275

>>15993547

Should make your flares unrealistic so that when they hit something they stick there and act as a light source. It'd help with seeing shit too

Any of you go to trade school to learn how to program? I’m curious.

>>15995925

Those variable names are the work of an asshole. I don't know why anyone would do this, unless their AI is programmed to function for every fucking cell in the game by hand. I'm pretty sure that's not what this guy is intending to do, I have no idea why anyone would name shit like this. I tried something like this once as a joke for a college programming class. The professor took one look at my variable names and handed it back to me. He said either I could use sane and descriptive variable names or he wouldn't bother to read it and just give me an F for the project. Good times. Copy the code, then use Find+Replace to replace ALL instances of these fucked up variable names with something more unique so it becomes easier to read. Do it with the goddamn functions and classes as well since func_50077_h is some bullshit. It's not you anon, it's this code being full of worthless variable and function names.

I can tell you right now "field_50085_a" has a lot of shit going on, and should probably be EntityFunction or someshit instead of gibberish. "field_50083_b" appears to be a part of the x,y,z navigation, but can't really tell from that snippit. "field_50082_e", "field_50079_f", and "field_50080_g" all are getting grouped together when passed, so I imagine they are most likely coordinate data. "field_50084_c" appears to be used as switch or flag to keep track if the AI should be active or not. "field_50081_d" and "field_52011_h" appear to be variables used to help determine the game state to see if the AI should continue being active. I'd need to see more of the code to understand more, but from this little bit, this is what I would guess.

>>15995947

sorta

you're going to have to sit down and learn how to like make game and you'll simply have a better understanding of the words you are reading

>>15995925

Tell me who did this disgusting shit so I can find him and call him a degenerate fag

>>15995947

The amount of programming knowledge you need depends heavily on what you want to do. So the question is, what kind of game do you want to make? Tictactoe is extremely simple to implement using just print statements in a console application, but I assume you are imagining something a bit more involved. Tictactoe is a great place to start though so don't knock the idea completely

If you want to make something that's not very graphics-intensive, like say a text-based rpg, then you probably do have enough knowledge to get started. If you want to include some simple 2d graphics, you could learn how to use a framework like SDL, or you could save yourself some headache and use a game engine such as Godot, Unity, or Unreal. Making a game in 3d is probably too involved for your first game, but not impossible. I would steer you towards something simpler, though.

If you've only had a fairly limited exposure to programming, I suggest that you use an engine like Godot or Unity to start learning. I think they support C# and some other OO languages, but I don't know all the specifics.

>>15996307

fuck unity

tell me one game made in unity you play and like

>>15996396

Cuphead, Darkwood, and Hollow Knight were all made with Unity. I won't argue that Unity isn't plagued by shovelware, but you can still make decent games with it. Although in order to get the most out of the engine, you may need to invest some time in learning how to best use its functions. I personally don't use Unity, though.

>>15995947

Clone a simple game. Tetris, Lunar Lander, Asteroids, etc.

This is also a good way to learn the very basics of OpenGL or a given engine. You can clone Tetris in both Unity and Godot to figure out which you like more.

>>15996307

If you're using Godot you're using GDScript. C# support isn't fleshed out and the documentation is lacking.

>>15996420

>Parkitect

explains why it runs so shitty

Anybody got one of these drawing of a body that I could use to 3d model, it needs to be a female body. Please.

>>15996476

Search for "ortographic reference female" in any search engine and pick one of your liking.

Optionally you can download free 3d base meshes like for example here:

http://wiki.polycount.com/wiki/BaseMesh

>>15996515

>says a minor is the perfect female form

>posts a lewd girl that looks like she could easily be an adult

pick one, CIAnigger

>>15996632

Maybe the hormones are making you a pedo instead. :^)

>>15996872

Terrible advice. Passion projects have more chance to resonate with a moderate audience than an app in the bloated mobile dumpster fire.

>>15996396

Cities Skylines

But indeed, fuck Unity.

>>15995995

>writing

>not the most oversaturated market by a mile

what? writing a book is much easier than a vidya.

>>15996965

The vast majority of agdg projects simply do not have the polish and quality you would need to reach any sort of audience, the passion does not matter at that point. Sure there are exceptions but 95% falls into this category.

>>15997085

That wasn't my point though. Better to make the game you like than mobile crap.

>>15997110

Absolutely

Hello Lads,

I started the learnopengl.com tutorials recently, so far I've been working through them. My end goal is I want to create a Dungeon Crawler in a Doom Style engine. After I have a solid grasp of the graphics part I want to get into the actual game programming paradigms (I hear object oriented is shit tier which I've kinda thought was true already), relevant maths to get a grasp on, etc?

Any of you peeps have the good reads / tuts about that kind of stuff? Devblogs you follow, good talks, stuff like that?

>>15997265

>Dungeon Crawler

Have you ever played a game called Mordor? If so, what are your thoughts about it.

>>15997265

For your game logic, use an entity component system (ECS). A simple, single header one:

https://github.com/redxdev/ECS

And a full featured, production ready one:

https://github.com/skypjack/entt

Using an ECS is really easy once you understand the basic idea.

As for rendering, it is vital to wrap all of your API code, be it OpenGL, Vulkan, DX etc. in a clean, modern C++ package, or even better, use a good existing wrapper. Going through a tutorial series like learnopengl is good to understand the concept, but you should discard all of the code you wrote while following them before beginning your engine.

>>15997327

I can't say that I have though looking at vids of it on YouTube looks pretty complex.

>>15997376

I grew up playing a lot of shareware games as a kid. That was kinda my bread and butter dungeoncrawl experience. Nobody else seems to have played it. I did attempt a simple clone of it a few years ago but I didn't know very much then

>>15997401

You gotta keep at it man, the world needs more Dungeon Crawls especially one made with love.

>>15997449

I originally was going to make it in XNA, then FNA, but I don't like the framework as much as I want, so I'm doing a simpler puzzle game in SFML right now, though it's been a few months since I made truly significant progress on it.

A pseudo 3d effect would be difficult with SFML because it doesn't do 3d texture sampling, that is textures get skewed because they only push XY vertex coordinates and forces Z to 0. It's a shitty implementation

>>15996396

Hollow Knight, Infinifactory, Cuphead, Furi, Teslagrad

>>15996396

When some "modern artist" takes a shit on a canvas and calls it a statement against homophobia, do you blame the canvas or the artist?

>>15996533

Thanks m80.

>>15996515

No, I need something realistic.

>>15994006

At best it will just be the exact same demo as I posted last time, but with Vulkan support.

>>15997265

Learn about procedural programming and DOD. Use as few abstraction layers as possible, work on your linear algebra skill since you need it to write such an engine. Design patterns are a meme and don't really mean anything. So don't waste too much time worrying about them… just write your code. The memory layout of your data is really what you should think about. Thinking about paradigms and design patterns just obfuscates what is really important. If you have a good memory layout, your code will be designed well. If you have a bad one, it will be designed badly. That's all there is too it…

Video shows my first UV mapped material on a mesh, using a generated texture. I need to learn how to export from Blender to Godot, otherwise I just get a mesh without the material (and texture).

>>15995947

if you understand variable attribution, variable read, branching and jumps, you can make a game

for c#

var myvar = 0 <-attribution

myvar = myvar + 1 <- read + attribution

if(myvar > 10) <- read + branching

myMethod() -< ""jump""

This is enough to make literally anything, it only becomes a question of how much time and willpower you have.

The actual main skills a programmer needs is abstraction and reduction

Being able to reduce a problem to the smallest operations and being able to abstract a "problem space" into a bunch of variables

I have a design in my head. I keep slacking and focusing on random unimportant shit and I've missed 3 demo days because i havent made it interactive enough.

How can i plan or schedule myself better?

>>15999111

set timers, dubtripsman. It's what I do. Set a timer (I use my phone but you can use whatever) for when you start and when you stop working, and then when those go off, you drop everything and only work on what you scheduled.

>>15999096

Well I guess I'm ready to get started then becaus I can do all of that. We're practicing abstraction and reduction in class right now so I've got those bases covered as well. Thanks anon, I know what I'm going to be doing this weekend.

>>15995947

Online tutorials are your friend. I've never taken a programming class in my life and I'm doing just fine. Think about what you want to make, break it down into simple components, look at tutorials for how other people have made similar things, and try to replicate it. Most importantly, change it as you do. Don't just copy a tutorial 1 for 1, but adapt it and personalize it. Actually think about how it works and how you want YOUR version to work.

>>15999111

Different things work for different people. I'm terrible at schedules and doing things at/in a specific time, but it works for some people apparently.

Managing lists works for me. For example I take things from my todo list and put them into what I call "quest" list, and I only work on things in the quest list, not the full todo list. There needs to be only a small and manageable amount of things in the quest list at a time or else it won't work. It also helps that when I'm done with a task, I move it to a "done" list instead of deleting it, makes me feel more accomplished for some reason. I heard of someone who does something similar, except they write tasks on paper and then cross them with a marker when done.

So anyway, for your problem instead of only taking tasks that I feel like doing, I could take a few things that I want to do and a few things I need for the sake of demo together.

>>15998481

Mesh import is pretty finicky at the moment. Right now the khronos gltf 2 exporter for blender is your best bet for materials, but the only exporter I could get animations working on was the better collada exporter.

Someone was working on assimp support for Godot, which would allow importing of a huge number of file types, including .fbx.

>>15999096

>variable attribution

Who in the world calls is that? Might as well call it "quantum number mechanics"

>>15996435

I tried to clone tetris in godot a while ago but could never figure out how to make the blocks aware they couldn't descend any further

>>15999546

I like to think I use OOP code correctly inb4 someone pipes up with >THERES NO PROPER WAY TO USE OOP CODE :^)))

I like to think of classes as packages of data and operations to use on that data, rather than as some sort of abstract class. So I'll have a Vector2 class, a Rectangle class, maybe a PlayerProfile or something like that, obvious concrete things. I'm not going to implement an IServiceProvider<ConnectionContext> or whatever meme bullshit the standard library has

>>15999552

I wasn't using an array to represent the game area either I was using nodes where the blocks were represented by further subnodes of solitary squares

>>15999552

>Two dimensional array.

>One value if empty and another if occupied.

>Each tetris piece keeps track of its own current location on the array.

>When the piece wants to check if it can move, it checks if the tiles it wants to move into are occupied or not. If so, it prevents movement.

>Piece is also prevented from moving if the place it wants to move to is outside of the bounds of the array.

Doesn't sound too hard an issue to fix.

>>15999572

>I wasn't using an array to represent the game area either I was using nodes where the blocks were represented by further subnodes of solitary squares

That really doesn't sound operationally any different than an array. You should be able to do the same sort of thing on a bunch of nodes as you would a bunch of array indices.

So for my puzzle game, since I'm wholesale stealing ideas inspired from others, I've hit a little design issue.

Each object associated with a player will all respond to that player's input. For example, if player 1 has 12 objects, then they'll all move left when he presses left, similarly for player 2 if they control any. So there's basically two styles of control going on here, which is fine. However, I don't necessarily see the value of having a "player 2", since the original game very infrequently had two-player levels, and in practice it meant controlling both sets at the same time with a keyboard.

I would like to support it though, because you can come up with some rather unique levels, but is it that bad using two WASD controls on a controller, like using the ABXY buttons to simulate arrows? I also have a replay manager that handles replays/input but I don't know whether it makes the most sense to put player input on the root object, in the replay handler, or buried on a player status object somewhere

>>15999894

You're saying you want to use, say, the D-pad for Object Set #1 and the face buttons for Object Set #2, or the left stick for Object Set #1 and the right stick for Object Set #2? I don't see anything wrong with that whatsoever.

>>15999923

Yeah that's it, plus maybe the shoulder buttons for particular actions.

I can't remember playing a game with a setup like that so I don't know how awkward it might be

>>15999927

Maybe add a modifying value that allows you to toggle between the two quickly?

>>15997945

>Everything you've just said is backwards.

How so? Do you honestly think the average agdg project has a reasonable chance of becoming a financial success? You can't be this delusional. Im not knocking agdg and im a nodev myself, but thinking you will earn money doing this is unhealthy.

>Have a Status object that contains gems, score, time, player inputs

>But also have a Replay manager that reads/writes to player input depending on whether its recording

>Can't just fire a (1) and done thing to move input to inputLast because it wouldn't provide enough of a time window for client behavior to update it

>Solution is to just remove replay handling from this part of the engine and push it down into the client code's responsibility

hello faggots.

has anyone ever made a turn based movement for an action rpg?

>>16000634

Jade Empire, Technically. I suppose you could count VATS from Fallout 3 and up as turn based.

>>16000750

aren't most SNES jrpg actually turn based movement (like FF V)?

>>16000443

Because before Steam shut down the API that let Steam Spy function. We could easily see moderately okay games made by small teams make decent enough sales for it to be considered a successful project (with the limited insight we have into these projects, of course). Maybe not successful to the point where it completely funds the next game, but still profitable .

Then, talking about principles, if you want to just shovel out mobileshit puzzle games hoping something sticks, then maybe you should just get a regular programming job. Because then at least you're guaranteed to make money until the day your employment ends.

>>16000634

Does FFXV's wait mode count?

>>16000763

>That one time last year WoW left an API open and someone iterated it and got all the subscriber data and found they were leaking subs and most accounts (like 30% or more) were alts or bots

>>16000763

I don't have the numbers on hand but i recently watched a >GDC talk about steam sale figures in 2018 and it's grim, the market is practically dead for almost all but the fringe cases and huge publishers. One of the headlines from that talk was that the average gamedev who has his game on steam makes significantly less than minimum wage.

>if you want to just shovel out mobileshit puzzle games hoping something sticks, then maybe you should just get a regular programming job

I will never do that kind of stuff, i was just saying that it is more financially viable, going from the assumption that you want to make money from this in the first place

>>16000865

Not viable for San francisco california SJW faggots.

I can buy a new home with 30k where I live.

>>16000896

Yeah, that's what I don't get, the internet is almost everywhere, it seems like cost-to-live can be wildly different.

>>15999373

>online tutorials

NO FUCKING NO

don't listen to this dullard, online tutorials are HEAVILY segmented and almost have no material describing the functions or the reason for the implementation. It's all source code.

Read a book, unless that book is also shitty. Books are hit or miss, try to find one on amazon that has good ratings.

>>16000929

80% of my cunt makes less than 500 bucks per month.

>>16000941

You should just use multiple resources at once. Books are sometimes outdated.

It's a good thing because you spend time applying the stuff you've learned to make it work on modern systems.

>>16001056

SS13 clone when

Unity and Prefabs are pain in the butt

>have separate camera object and player object (camera follows player by default but it can toggle for example to security camera or cutscene)

>weapon models are attached to camera and they show whenever tracked player equips weapon

>whenever player picks up weapon it adds monobehaviour for that weapon that handles ammo, reload, firing behaviour, etc.

>now I want that monobehaviour to fire, for example, a rocket

>whoops! you can't magically summon a prefab from script, either put every single possible prefab reference to a player or load them from resources folder

>>16001354

scriptable objects might alleviate your suffering

Spent most of the night working on some library-tier code for my engine. this is to make working with tiles and actors easier for other client code I will write, a sort of mini API. Does it look decent or is it shit?

https://hastebin.com/jageqowive.cs

>>16001394

ok so what do I do is

>define a scriptableobject

>make a scriptableobject asset with reference to prefab

>instead of adding weapon monobehaviour to gameobject instantiate scriptableobject asset and summon rocket projectile from it

is that right?

>>16001056

Your perspective mapping looks wrong, but I'm not sure if its meant to be like that.

>>16001415

>regions

It's shit. That'll be 500 atoms. Personally I'd make those static functions and pass in a map as a variable

>>16001354

How would I know? I just give out advice, I don't know how to do it :^)

>>16001424

>instantiate scriptableobject asset and summon rocket projectile from it

If I understand correctly, when you instantiate it you lose the original asset's references. Why can't you load from the resources folder?

>>16001520

It kinda messes up my folder structure. I would have tons of explicit resources folders just for keeping prefabs I want to instantiate ingame. I'm gonna try making a scriptableobject for keeping prefab references and if it won't work I'll just deal with it.

>>16001056

Use ortho scale nigger

>>15997265

This series for vector & matrix math

http://blog.wolfire.com/2009/07/linear-algebra-for-game-developers-part-1/

There's 4 parts, but it unfortunately does not cover quaternions.

>>16001495

I had a system kinda like that but it seemed like I had to qualify two references to change any data anywhere which seemed wonky.

I need a way to search for actors by position and maintaining a jagged array seemed best, even moreso than a dictionary (due to how I index them by coordinates, I'd have to create a tuple object for every check and operation, no good). I also need to sort them by position which is why I have the single array, that's for updating.

There can only be one actor per tile, but and actor may use several tiles

ay, how many fags here bothered to make their own game engine?

>>16004454

At least 2, that one robot 3D platform guy and the Sigma engine developer.

>>16004454

I made my engine in C++ and SFML. The engine is done so I'm working on the other parts. Haven't shown the game off yet. Maybe on demo day if it's presentable enough.

>>16004150

Your video made me check discord

Last thread there was a bit of a freakout over the Unity TOS change.

They recently put out some updated info (they updated their TOS again) that I thought anons would want to see if they haven't already:

https://blogs.unity3d.com/2019/01/16/updated-terms-of-service-and-commitment-to-being-an-open-platform/

As far as I can tell they listened to the devs that suggested fixes in this thread:

https://forum.unity.com/threads/recent-tos-update-blocks-the-use-of-spatialos-to-make-games-in-unity.610447/

Rundown:

>TOS changes are per "yearly" version e.g. unity 2018, unity 2019, etc

>transparency, TOS per version on github

>use any service you want

>however, unity will not offer support for non-partnered services

>improbable/spatialOS are no longer in violation of TOS

SpatialOS drama:

>apparently they were violating their trademark, and falsely claiming to be associated w/unity and/or a "partnered service"

>apparently they were violating the old TOS (supposedly long before the insidious December TOS change)

>no longer in violating of new TOS due to easing up of restrictions, and they rectified trademark issue

New TOS, the changed part, which is section 2.4:

>Unity developers are free to use any service offered to Unity developers (each, a “Third Party Service”).

>Unity does not have any obligation to provide support for any Third Party Service provider or Third Party Service under this Agreement.

>Third Party Service providers may not, without Unity’s express written permission:

>(1) use a stylized version of any Unity name, trademark, logos, images or product icons, or other Unity-owned graphic symbols;

>(2) use a product name confusingly similar to a Unity product or that could be construed by Unity developers as being a Unity product or service; or

>(3) create or use any marketing materials that suggest an affiliation with, or endorsement by, Unity.

>All use of Unity’s trademarks must comply with Unity’s Trademark Guidelines

Take it as you will tho

>>16005278

>Indie devs having to deal with engine legalese.

Just go for Godot.

>>15999529

>Who in the world calls is that?

ur mom called it that when i was at ur house doing you

I've been thinking about making my own Visual Novel game for awhile. If anyone has ever made one, how much resources go into it moneywise? I've seen people make one for around $5,000 with donations.

>>16006865

Well, it'll depend on the rate of the artist(s). If you're the artist, you could make one for dirt cheap.

it's over for us

the whole team is wagecucked

ded gaem

>>16006900

i was the only member of the team that knew about this chan

>>16006905

What I meant was that you could at least tell us which gaem is ded now.

>>16006921

the "fire emblem but in real time" game

>>16006998

Don't think I remember. Could you post pics? Sorry.

>>16006998

>>16007073

I don't remember either.

Gonna be on vacation for February, so I'll only be able to code. Perhaps I'll make significant enough progress to show something off for once.

In the meanwhile, here are a few songs I wrote recently, the latter one having been finished minutes ago. Very different in style from one another.

https://files.catbox.moe/41j6zg.mp3

https://files.catbox.moe/2v9vx5.mp3

>3 days off

>OH BOY I CAN MAKE GOOD PROGRESS

>Play Factorio for 12h and only dev for like 3

Shit, posted this in the wrong thread

>>16007359

At least you didn't shitpost in video form for three days straight.

>>16005660

Absolute mad man

>>16007359

3 hours is good!

>>16005660

No sinewavedash, no buy

>>16007346

Your music sounds nice. Which software you use to make it?

>>16005278

sounds like they went the wrong way closing a loophole and fixed it. unity makes their money from the license and story, butthurt in this case is to be expected

>>16005278

unless you do literally everything yourself (good luck) you'll have to deal with legal shit sooner or later

>>15995947

>Will my knowledge of coding basics/fundamentals be strong enough at that point for this to be a good idea?

Your gonna have a hard time, but that's why it's worth doing.

Trying not to repeat others here:

>Refactor everything you do constantly. Look up what refactoring is if you don't already know.

>Break up functionality into separable components(functions, objects, ect.).

>With every step, ask yourself "What problem am I actually trying to solve?".

>Keep a legal pad in front of you to jot down your plan before you start working and to write notes.

>Get a reference book for whatever your working in(your Uni library probably has some texts on at least what languages an engine will be using)

>A little effort everyday is more powerful then a weekend you bunout in.

>>16007642

FL Studio/Fruity Loops. Instrumentation is a mess of VSTs, default plugins, soundfonts and other shit I've found on Google. Any time I write something of a slightly different song, I spend hours getting the instrumentation just right.

>>16008632

Discovered another issue, because I flipped the white/black values in the list for my class and had to make adjustments to the sheep entity class so the world spawner uses the correct color weights, one thing that doesn't do that is spawner eggs. So everything else works except that, world spawns white/gray/black and rare brown sheep but the spawner egg always spawns a black sheep because according to vanilla black is the first color used.

Guess I'll just put that in my to-do list for now since I want to try other things first.

>>16008840

>Need something to pressure you into deving

Existential dread and the crushing weight of dreams unfulfilled is a little too much, though

>>16008857

That's overboard. You just need a small push like "I need to do laundry first shit". And that will make you often times a more self productive person because you'll end up getting shit done. If you are too lazy for that, then that's a whole other and serious problem.

>>16009456

>How the fuck do I make Windows 10 actually respect my borderless fullscreen settings?

by installing a different OS

>>15996300

looks like the work of a decompiler

>>16011142

>Linux isn't a bloated piece of shit

Just because it's not Windows doesn't mean it's not a bloated piece of shit. Windows just makes Linux look good by comparison.

>Eventually, people started to run games in something that was referred to as "borderless window" or "borderless fullscreen" or "fullscreen desktop" or whatever stupid name they felt like calling it.

This is the best way to do fullscreen in most cases. It's a nice touch when a game has that as an option. Some games involve having to use other windows while being exclusively fullscreen. I usually play them with two computers. One for playing the game and another for communicating and looking stuff up.

>>16009754

>>16011401

Also digging through SDL source code, if SDL_WINDOW_FULLSCREEN is set it tries to call ChangeDisplaySettingsEx using a display mode generated from SDL_GetWindowDisplayMode (which returns the display settings if the window would be fullscreen) together with the CDS_FULLSCREEN flag. For SDL_WINDOW_FULLSCREEN_DESKTOP, it passes a NULL display mode instead with the same CDS_FULLSCREEN flag, meaning it just resets the display mode to default. It's hard for me to tell what exactly SDL passes into display mode as digging through this code is a pain, but the behavior is logical: for borderless, it doesn't fuck with the display and it behaves like a normal window.

As for the window, SDL_WINDOW_FULLSCREEN and SDL_WINDOW_FULLSCREEN_DESKTOP behave identical when it comes to determining window style; both generate a "WS_POPUP" style with no other flags set, ignoring all other flags on the SDL_Window as well. Which seems to be the correct behavior for creating the bare minimum of a window.

>I see what you mean now. It's just an effective "fullscreen" without a real fullscreen setting. Makes sense… this is interesting, I haven't added this feature to my engine (or used windows 10 extensively) so I never ran into such an issue.

A lot of engines and games these days have this feature. It's quite common, and your game wont have that "PC mustard rice"-quality feel without it.

If you're too lazy to implement it, people will often try to emulate it by running your game in windowed mode and using a program like shiftwindow to resize and position it. This however is not always perfect. In the case of my project, the Windows start bar appears on top of it. (Although it does actually behave like proper borderless, funny enough.)

>HKEY_CURRENT_USER\System\GameConfigStore\GameDVR_FSEBehavior

I've set it and it doesn't change shit.

It's probably because the solutions you, me and everybody else finds are regarding exclusive fullscreen being forced into win10 hybrid cancer mode, whilst I'm looking for a solution of borderless fullscreen being forced into win10 hybrid cancer mode.

It's hard to google anything decent because there's no consistent terminology for the issue.

>>16011599

What a strange situation… I don't think I know how to help you anymore just because, I don't have windows 10, and any mentions of this behavior is absent from MSDN. If you could find an article about this on MSDN, it would probably be very useful…

>>16011645

I've dropped all my projects because none of them feel fun.

Ah, what the hell, I've been trying both to come up with something interesting or motivate myself but I haven't done anything all day.

Help me please, how can I make my game about trying to get through a forest while two tribes of subhumans fight it out fun.

What I have for now:

first person, I do this both to avoid having to do main character animations and to maybe support VR.

tribe 1 uses spears, throwing them if you run, dodge one and you already have a weapon, if they all throw em and you defend the area where they threw them at, they get afraid and leave.

tribe 2 uses something more advanced(?) that invalidates the first strategy, and the player can befriend them temporarily by giving them food they have a chance to find in the wild.

The most efficient strategy however is going to be making them fight each other a la DooM.

However, I'm having a lot of problems figuring out a balanced and fun way to go about it. Just randomly dumping tribes around doesn't really lead to an interesting gameplay loop, I need to think of a better goal, better prioritization from the tribes, and in general ways to make the game fun.

>>16011645

Yes. The first thing I did was make the basic controls and movement a joy to use. Not simply 'adequate' or 'controls so good you don't notice them' but 'moving around is fun.' It doesn't matter how much I manage to fuck up everything else, as long as the game runs I'll still have fun moving between points A and B.

>>16012750

stick your dick in it

>>15997265

If you're using C or C++, read through some of the Doom source code. It's pretty well built.

>>15997449

Good to hear. What do people prefer around here in their dungeon crawlers? action combat? customizable magic systems? turn-based?

I'm building one too - similar to Dungeon Master, fairly simple, centered around fighting larger creatures (larger than 1 movement square), and "event" creatures with a few stages of behavior.

>>16011645

Do a hard analysis of games you know to be fun. Play those games again, note the EXACT aspects of it that are fun (not just to you, that's autistic. What about these games is objectively fun and widely acclaimed?). Do a rigorous analysis of those elements:

>How many m/s does the player character run at?

>How many meters/character lengths/heights do you jump? Base and with items/powerups.

>What is the min/avg/max combat time per enemy encounter?

>plot skill use time/flashiness against dmg dealt/satisfaction. Is there a trend? Where is the sweet spot?

>What is the min/avg/max time between random encounters? Average number of encounters per location to location route (overworld/dungeon)

>How long does it take and how difficult is it for one to maneuver through the RPG town to upgrade gear, bank, get plot updates/quests/etc. and get back to the action?

>What are all the playable character achetypes and roles in the party?

>Weapons, skills, physics, etc. etc.

Apply to whatever you found most fun about these games, hopefully ones similar to what you want to make. Distill the "fun" out of games in your genre and replicate it perfectly. That is your cake. The frosting on top is your setting, characters, story, music, art, "comfiness", whatever that makes it your own game and not an uninspired derivative. Good luck.

This is a prototype for a mecha game I proposed to /v/ earlier this month. All the models are placeholders, I'm only trying to get the gameplay grounded atm. Gimme your crit.

>>16013514

I wouldn't go so far as to call what you've shown "gameplay" yet, as it's just the skeleton of a walking, shooting, first-person display. Before asking for critique again, I would create at least one enemy, and show how you would engage with that enemy in a to-scale blocked-out level. I believe that would be the bare minimum example of a mech game's gameplay.

I will critique what you have shown, though. I would like to see the FOV opened up more, you're looking out of a porthole at this point. Making the cockpit more of a hemisphere would give you a lot more options:

>You could see you own legs if you look down. Doubly cool if you can walk in one direction while turning and shooting in another. Triply cool if you can step on and crush cars/people/whatever.

>You could see your (presumably shoulder mounted) missile launchers. Give them the proper blinding explosion of fire and smoke like a missile launching from a silo and I'd be all over that shit.

Ditch those display panels and use translucent HUDs projected onto the cockpit glass. Much cooler, better visibility.

Add missile lock-on/fly-by-wire. The missiles you've shown are simply bigger, slower bullets.

That's about it.

>>16013514

You know, you can just post a video instead of exposing us to your faggy youtube/twitter account among other things I won't share

>>16011748

I found the information that describes the fuckyness.

It's a really shit quality video on youtube. No joke.

<4:55:

>The way direct flip works is that the compositor will detect that your window covers the entire screen and will actually decide not to compose. It takes itself out of the picture and puts the asked content directly on the screen.

Apparently this exists starting from Windows 8.

There's some newer versions too (iFlip, iFlip immediate, and google also shows people talking about "iFlip true immediate" because of course immediate doesn't actually work as immediate), but it all boils down to not treating the borderless fullscreen window as a window.

There's some DXGI articles related to video such as this one: https://docs.microsoft.com/en-us/windows/desktop/direct3ddxgi/for-best-performance--use-dxgi-flip-model

But this is all DirectX stuff that I don't touch at all. Nor is there anything mentioned about disabling it.

At least I found a "fix" for my own project.

If I make a window with SDL_BORDERLESS, it'll behave like a normal window.

If its size & position equals that of a display, it starts to behave like exclusive fullscreen for that display.

If I make it 1px taller or wider than that, it behaves like a normal window again.

I'm not sure if I want to write such a disgusting hack around my project. But I don't think I have a choice since slow alt+tabs agitate me.

>>16013554

Extra effects like the missile smoke thing would have to be added in post but I'll keep it in mind.

The cockpit is OBVIOUSLY not final but I'll redesign a better placeholder using your notes.

>The missiles you've shown are simply bigger, slower bullets.

And that's not what missiles are?

Really what I'm looking for is crit regarding the movement and the feel of the mech, that has been the main focus for the past few days for me. I also want to know if anons are okay with the weapon system on display which involves switching between each individual arm and shoulder or controlling both arms or shoulders simultaneously.

>>16013611

Then post a demo, you stupid faggot

>>16013611

OK. Thanks for clarifying what you wanted critiqued. I agree with >>16013641 though. We have no idea how the movement and shit feels just by looking at it. There is also nothing for us to reference scale or give a sense of speed. That said, I'd say don't worry about what we think yet, just keep making progress. It looks like you can move around and shoot, so keep working towards a Minimally Viable Product. Show that off then.

The "feel" of your mechs will come much later, when you've added acceleration to the thrusters, shocks to the legs for when you land, sound effects, etc.

>>16011645

I often get stuck playing my game.

The only project I didn't have fun with was a city building idle game of sorts, because after I had made the basic engine features I realized that I didn't have a clear plan of what to actually put into the game.

>>16013957

I don't use SDL_Renderer, and I use OpenGL.

Digging through more SDL source, SDL doesn't actually do a whole lot extra with the window if OpenGL is involved. It just sets the pixel format (SetPixelFormat()) and creates a context (wglCreateContextAttribsARB) with WGL_DRAW_TO_WINDOW_ARB and WGL_DOUBLE_BUFFER_ARB, but nothing that would actually configure the swap chain to specifically use the modern flip features.

I really don't think the issue originates from SDL.

>>16014039

What a shame… that's really a poor situation. I'm going to have to do the same thing in my codebase eventually, maybe.

>>16011599

>you, me and everybody else finds are regarding exclusive fullscreen being forced into win10 hybrid cancer mode,

Wait, what is going on here? I'm on win7 still . what is the hybrid mode win10 forces on dedicated, fullscreen resolution modes?

I'm using GLFW to manage windows, GL contexts, and monitors (multi-displays). It has a borderless desktop window mode, a standard windowed mode with all of the win7 widgets, and a real, fullscreen dedicated video mode.

I always assumed people used borderless desktop windows because they had shit video drivers that did not properly switch video modes and GL contexts (AMD shitware). The problem with borderless windows is you will most likely have 0 control over refresh rate, color grading effects. It's a poor substitute for a real fullscreen mode that people opt for as a halfassed fixed for bad video drivers.

>>16014452

I usually use borderless windowed mode for strategy or turn based games where I want it to take up the whole screen but also want to switch to other programs easily. Never had a driver problem with exclusive/real fullscreen, but I'm using Windows 7 with a recent nvidia gpu.

>>16014452

>what is the hybrid mode win10 forces on dedicated, fullscreen resolution modes?

See >>16013581

It occurs on Windows 8 and newer.

>I always assumed people used borderless desktop windows because they had shit video drivers that did not properly switch video modes and GL contexts (AMD shitware).

I talked about some of the differences here >>16011142

I feel like I keep programming in circles, so I take a break to play a game, then I want to clone that game

>>16013032

>as a /tg/ nerd, there are plenty of games I would play even as a 1:1 adaption just to skip the setup and to save space.

Things like Magic, Battletech, and 40k are already pretty well covered. I know what you mean though. I liked Warmachine but never found good digital version and couldn't afford to buy and paint more than 2 or so units.

Any interest in the HARP RPG system here (Rolemaster)? I really like it's individual weapon tables and crit tables, and alternate charts for medium->large creatures.

http://ironcrown.com/harp/

https://en.wikipedia.org/wiki/Rolemaster

I have my basic combat mechanics system implemented. It took way more work than I initially thought. It has 5 different status effect types, 6 different damages types, and 4 different shielding types. Also area of effect, timed damage with decay, and timed secondary effects. I can add a new attack/defense/ability just by filling in a new instance of a Lua table including the descriptive text and necessary scalars. It's not a component system. All the available fields are hardcoded as well as the logic in C. Although new combat logic can be added with a custom Lua function definition added to the attack/defense/ability table.

>>16011142

Excellent writeup. Thank you. I wasn't aware of this windows8+ fuckery. If the 1px window expansion works I'll use that hack I think and just use some offset defines for my opengl viewport() calls.

>>16013712

It's probably God trying to prevent you from wasting your time on a useless project and go start a family.

>>16012750

>on starting with Unity?

>unity

Go pay for the premium version because the free version literally say that they own everything you do in it. Then get back to me.

>>16015678

>that's not how intellectual property rights work.wav

Also

>4.2 Your Content

>As between you and Unity, you own all right, title and interest (including, all intellectual property rights) in and to the content you create using the Software, Developer Services and/or any content you post to the Site or in the Communities (collectively, “Your Content”) (other than any components of the Software contained therein or used in connection therewith).

>>16013008

Thank you, this post cleared my mind a little.

Initially I wanted to avoid designing or researching anything and just go with the flow, but that didn't work out so well.

>>16015984

Look for youtube tutorials for the basics then mess with stuff and look up things when you can't make it work with enough digging. Ignore the shitposter, just like make game.

>>16009027

I thought I almost had the side checks for the burning log working, I got it to check if there was an adjacent fire on any of the 4 sides, and set a boolean to true, then run a loop if the boolean is true to check if all other sides are solid, but there's a serious problem with how I'm iterating over the list of sides.

I thought I could iterate over the list, remove the element from the list that matches the fire block from the list, break the loop, and iterate again over all other sides, but it looks like the iterator is destroyed after it's been looped through.

Tested this externally and it's true, if I make an integer list with 4 elements, run a while loop on iterator.hasNext() to print out each element, then run the same while loop again, I get nothing.

When I try to remove the second element from the iterator, it also removes the 4th element.

What the actual fuck.

>>16016339

i'm not that anon >>16012750 but yeah what you're saying is true

>>16016434

Why do people use iterators instead of for loops with a length property?

>>16016462

It probably would have been easier doing that than trying to figure out which iterator does what.

I think what I'll try next is setting a color modifier of the block based on how long it's been ticking. Maybe in stages, so first tick it turns red, second tick it turns orange, third tick it turns black, and on the 4th tick it drops.

>>16016462

Industry standard :^)

>>16016462

While loops are the simplest, and for loops compile down to a while loop with some extra syntax (increment and initializer), at least in C#. Iterators are retarded and involve a reference object be created with a collection. There's a TON of overhead in comparison, and you should basically use for loops anywhere you can.

>>16016590

I expected to be told that iterators were optimized or something.

>>16016462

I can't speak for Java, but in general it's an abstraction over, well, iteration. Consider going through the items of a binary tree; with generalized iteration you can write code that works for lists and trees and leave the details up to the implementation. There's definitely a point where you start to overdo it though, Haskell's Foldable and Traversable bullshit comes to mind.

>>16016563

Are you implying that Odin wants you to make vidya?

>>16016664

Every layer of abstraction and indirection will slow it down. Barely – but surely – if implemented well, and that's the best case

>>16016888

>laughs_in_c++.tga

Anybody have resources on project organization?

>>16017528

It's not that, stuff like fullscreen isn't part of X11, nor is mode switching, so it actually has several extensions that duplicate this functionality, none of which are fully documented, so i'm still not sure how to use any of the API's that are provided to do this. Eventually I will have to get the code sorted out…

>>16016874

Oh man, I revised some of the code for this so I only do one iteration, and return an int between 0-2 depending on which conditions are met

>fire && solid neighbors

<progress and drop charcoal

>fire && !solid neighbors

<progress and drop nothing

>anything else

<revert to normal log

and I've started getting an array index out of bounds error for multiplying the color of the burning block, there's only 3 colors it goes through so the list of colors only takes 0-2 as an index

for some reason I occasionally get a number way way WAY higher, now I'm getting the color based on the metadata of the burning block so it should never exceed 2 but sometimes I'm getting 7, or 11, or 15 as the metadata and I have no idea why

>>16017408

>muh zero cost meme

Template shit is expensive in its own way.

>>16018230

You'll change that stance once the program becomes large. But it's not just compile time, code size blows up as well.

>>16011645

>Would you want to play the game you're making?

Yes, that's the entire point of my game dev attempts. I am my own target audience, everyone else is just a free bug tester.

Unity 2d best option to make an ogre battle clone? Not tactics. World map, capture/liberate cities, auto battle version of Ogre Battle. I've worked with Unity before with A*, general pathfinding, etc and hexagon based maps for 4X style work. Nothing finished. More learning than anything. Just curious if anyone knows of a different engine that would be easier/better suited to the project.

>>16018270

There are pretty good compilers nowadays like zapcc, my stuff at home compiles in less than 2s on average. In a corporate scenario you have distributed compilation anyways.

>>16017477

That Argentinian dude seems pretty nice, I hope he gets to keep doing his job developing Godot.

>>16018342

godot

>>16018230

>Trading compile time for runtime performance is a deal i'll take every time

That's not the trade you will be making, or not the full extent of it.

Go try to compile the xTensor library sometime. It's header-only. Heavily uses templates. Good luck.

>>16020052

i believe it has improved but that's up to interpretation

>>16019738

That's godawful. You should wageslave until you can afford to pay an actual artist to design a more interesting character than an angry teenager in jeans.

>>16020079

its rough animation m8 just trying to get the idea across

>>16020086

oh shit thanks!

>>16020087

I was criticizing the design, not the animation which is also terrible

>>16020094

design is basic now yes, once i have his animations finished up i'm adding more

>>16020110

Is he the guy who was making that evangelion thing a while back? Or was that a different guy who does the exact same thing?

>>16020110

i've reworked it because i found the timeline shit kinda stupid.

>>16020136

the aregntiean guy was the one who bitched about yak taro stealing his idea right?

>>16020148

them polygons tho

>>16020146

>Is Tetradev from Argentina

>Can't even spell "Yoko" right

You're the argentinean nigger, you dumb shit.

Your new music is just as bad, by the way.

>>16020164

Don't be too hard on him: he's also a talented voice actor.

>>16020195

FUCKIN SAVED

Well this thread certainly got derailed.

>>16019320

Well, guess it's just me then. I got it to compile but building their basic sample code from their intro tutorial produced template errors and would not compile. Using C++14 as required by xtensor and their recommended compiler (they have work-arounds in their instructions for most major compilers). That's kind of the issue with templates though. They quickly break on anything but the dev's build system.

>>16020232

1: that looks like shit.

2: You really shouldn't be complaining about someone else "bitching". The last time you showed you reee'd about /pol/ and had a mental breakdown in the thread.

>>16020243

shit

>>16020244

i ree'd about pol?

and i know i'm gonna change it up

>>16020247

>win8.1

>i've got to clean up my desktop

>so i can export vids using photoshop cs4

>new desktop the one i had had a nasty fuckin crack

>thats gzdoom and something i used to patch melee for dolphin

>helping my friend make a garbage meme

>have to clean up the desktop

>its the gif for quick access

i wont lie i am a bum

>>16019594

check one for the music archives for tracks with an open license etc, there should be plenty.

>>16013957

>It's generally easier to work with Windows than Linux, despite the reputation…

depends what you do, it also is important what are your options to fix issues. I much rather dig through some linux autists blog or worst case check the source than have to pray microsoft has an answer on their mess of a site, and if it has it's not some mcp retard dumping his "sfc /scannow" textmacro to increase his postcount - and that's only possible if you even know wtf is actually the issue and not "something happened ;^)"; my last issue literally told me "there is an issue but we don't know that it is while giving me links to the knowledge base"

>>16014593

main problem is you'd have to come up with your own ai, or only make it 1vs1. also might be easier to keep it simple with a boardgame first, there are quite a few dungeon crawlers (some even feel like a paperversion of a digital game with all the clutter and dozen stacks of cards).

>>16020793

ok, I will compile it to linux and mac I guess

>>16020799

I would appreciate this too.

>>16020793

>>16020806

https://phantomcorps.itch.io/spaceshootergame

Ok guys, I made some new versions, but I can't test them here, I would be really grateful if you could check them out.

>>16020772

I'm having issues with performance - I've got 12k fps at very low graphical settings, and bullets/enemies move in slow motion, but the power ups move at what I'm guessing should be normal speed.

I'm not sure how shmups work because I suck at them and so avoid them like the plague, but the other thing I noticed was that in this particular (random/procedurally generated?) playthrough there were often big gaps where no enemies were approaching, and I had more incentive to dodge 80% of the time than actually kill enemies. Obviously if I were dedicated to the game from the get go I'd be pressed to level up as much as possible before the boss, but if not, I might be frustrated that Im' totally unprepared for the boss when I get there.

The other thing that you might want to consider is a limit to the focus time, something like stamina .

>>16020823

thanks for checking it out!

>The movement is really floaty and sticky. >Everything feels sort of like wading through

I will check out how to mess with the physics a little better, using the unity rigidbody thing is making stuff come out weird

>>16020835

Thanks alot my dude, I will check out ways to optimize the game, I also think It might be related with the same rigid body issue, also maybe I can use the charge to control the focus time, so I will try that

>>16020849

You shouldn't be using physics/rigidbodies for a bullet hell, they need to be tight and fast while physics are slow and boring.

>>16020823

>he's been doing this for a while

Because you keep responding to him

>>16020822

it's incidental but I like how tooltips and shit make it in the final executable

[Serializable]

public class Shooting

{

[Range(0f, 100f)]

[Tooltip("probability with which the ship of this wave will make a shot")]

public int shotChance;

[Tooltip("min and max time from the beginning of the path when the enemy can make a shot")]

public float shotTimeMin;

[Tooltip("min and max time from the beginning of the path when the enemy can make a shot")]

public float shotTimeMax;

}

>>16020788

MSDN documentation is much more clear and complete than any of the linux man pages. On X11 I have to guess what an API does because the documentation for it literally doesn't exist, while Windows DWM fully documents those features… and man pages are written in a kind of free-form way that makes it more bothersome to read. MSDN formats documentation more like a table, I can quickly understand what every parameter of some function is supposed to do, etc. Linux isn't at all as through and the style is more like notes about what it's supposed to do instead of explicit, short descriptions of what each part of it does.

What kind of issue did you have? It should give you an error code, and the error code should explain everything… Windows has much more detailed and complete error messages, they are even in multiple languages.

Although, i'm only describing C programming, both UNIX and Windows have a C interface, I couldn't talk about it in terms of anything else. If you're using a high level language, that's not really "Windows development" or "UNIX development", you're just talking about the abstraction layer that the language provides, and not the actual interface and behavior of the operating system that Microsoft created.

>>16020851

Makes sense, thanks dude!

>>16020876

yeah! I didn't know that happens.

>>16020772

I'll reiterate what >>16020823 said. The moment is floaty, enemies going off screen is a problem. It's even worse on my machine because the screen size in smaller than normal but the grey boxes have a set pixel width or something and so more of the screen is covered.

The hitbox is pretty large for the number of bullets on screen. If there are some bullets between me and a pickup I have to tank them, I can't weave through the bullets to get to the pickup.

Remove any and all physics from the enemy drops. The drops should move down the screen at a constant rate, not shoot up out of the enemies, then slowly accelerate in the direction they should have been traveling in the first place.

>>16020884

That's because the manpages aren't the full documentation, you can read the standards for that.

>>16020884

it was more meant in general, documentation is nice but doesn't help when basic stuff shits itself. before that issue I literally got "error: null" when trying to enable ICS at some point (and even if you what's causing it it's not consistent, so who the fuck knows that the problem was), if you even get an error. finding out stuff suddenly doesn't work anymore because reasons is always fun.

networking especially is huge clusterfuck if you want to do anything more than have a single interface and even then you have to constantly wrangle with the OS. simply takes time and nerves I rather spend on something more productive.

>>16020905

Thanks for checking It out, I will mess with the UI soon, I will take out most of the physics that might even help with performance

>>16020772

Yep this plays very close to a really bad clone of the EARLY pc-98 games of touhou (not touhou 1), but with even worse shot patterns, enemies, etc. You know some good music and sound is not just there to make it pretty, it allows the player to not have to look at the enemies blow up and just rely on sound.

It feels like you were originally making a horizontal shmup but then tried to transform it to a vertical.

You do not want to make an euroshmup, health bars, the whole charge/ammo thing, the inertia with the movement is a big sign of bad design of shmup in general. Also joystick/controller support can go a long way.

If you're serious about developing shmups, I highly recommend reading the following:

https://shmups.system11.org/viewtopic.php?f=9&t=36231&start=0

https://shmups.system11.org/viewtopic.php?f=9&t=39679

>>16020955

>the manpages aren't the documentation

That's literally what they're there for. You read them to find out how they work… you shouldn't have to ask any more questions after reading them. Also, not everything is standardized, and Linux's implementation isn't the exact same as the POSIX specification. So it's not that simple.

>>16020982

>networking especially is huge clusterfuck if you want to do anything more than have a single interface

Network programming in Windows works the exact same way that it works in any other UNIX with an API that is basically the same. All modern OSes use the same style of Sockets API that BSD created. It's literally the same as anywhere else…

>>16020148

Christ. I could make a better Evangelion game in 3 months and I'm just starting on this stuff. Y'know what? Fuck it, I'll make a better Evangelion game for the 5/5 demo day. Not that it's a high bar to surpass but I'm gonna do it.

>>16021036

Ok, I ve read those threads and I have to say there is much to take from there, thanks alot,

>>16021149

Sin and Punishment was my favorite not-Evangelion game.

>>16020676

Thanks, I'll check it out.

>>16020727

I appreciate the offer, but it's not worth it for this project. The game is basically Commander Keen without a pogo stick, with more emphasis on combat and less on hunting down keys/finding your way around the map. It's a very basic game that I'm making as practice. I made something similar in the past and decided to see if I could make something noticeably better.

Is there a good C++ wrapper for OpenAL Soft? Or should i just write one?

>>16020238

That is true. I was using MSVC for this one. And of course, template code compilation time scales exactly with what you're doing with it, im not sure if their benchmarks are all that complicated.

>>16021849

>Is there a good C++ wrapper for OpenAL Soft? Or should i just write one?

OpenAL has barely any objects to speak of. It's trivial to write one.

>>16021999 (TRIPS)

Her surprise is off in the first gif, make her do the jazz hands when the camera snaps, and make her wait a little before picking up the picture.

>>16022111

I'm confused. The jazz hands is when she finds an object not when she reacts to having a picture taken of. I could make her pose if that's what you are suggesting. I like the japanese peace sign. I think the timing for picking the object up is fine. I don't see why it would need to be changed.

>>16022150

Ah, I see. The pose could be a nice addition, but I indeed thought the surprise came from the taking of the picture. In that case it's fine.

>>16020267

>discord

explain that one and then kys

>>16021999

I don't really lurk these threads much but damn anon, that looks really nice.

Also, checked.

>>16016462

A iterator is a ultra convenient thing that can iterate over mostly anything, because most things made to be iterable implement Iterable and thus become iterable regardless of it's structure.

With that, you can abstract whatever something is a tree, a list, an array, a sparse array, a hashmap, a hash set, or whatever else kind of collection. You can use all of them the same way. Also, it's trivial to concatenate iterators

It's pretty convenient, but you can't optimize code using it for shit. So whenever it's worth using it or not, it's the coder discretion.

>>16022150

>Japanese peace sign

I thought in nip culture it's supposed to be making a V with your fingers as in "Victory"

>>16024805

Is that one of those IBMs?

>>16021039

>It's literally the same as anywhere else…

again, it was meant in general. API is fine and dandy but doesn't help you when shit simply doesn't work. hf figuring out why your game doesn't connect even when you did everything right because windows wants to be retarded again.

>>16025002

Yes, it's the IBM 701c but thinner, with modern ports, and a slightly different keyboard layout.

>>16020086

Damn nice!

Saved. You are an american hero anon

>>16025008

You're getting confusing replies because you're trying to change the topic from Windows programming to Windows in general, when the topic was clearly about programming for Windows. What's the point of this? I wrote a post specially about programming for Windows, and you're disagreeing because you moved the goalposts and changed the topic.

If it's about things "just not working" that's hardly anything specific to Windows or Linux. Things on your system will break, especially when you don't know how to use your system.

>>16024805

>butterfly keyboard

i never felt the keyboards on the thinkpads were any good, just had a sane layout but that action when opening is pretty killer

>>16025438

it will never make a comeback because the keyboard action was designed for a certain screen size that is too small for most consumers in this age

>>16025490

I don't think you know what modern languages do with iterators.

>>16025438

>>16025628

Its cost is impractical and the moving parts beg for failure, but it's so sexy.

>>16025438

Thanks. I want the laptop in-game to be memorable, and what laptop is more memorable than "that one with the folding keyboard". I'm happy it turned out.

>>16025458

>>16025628

Thanks anons. Glad you liked it.

Shmup anon here.

I think I have the rough explosion formula effect, but I will need to dig on how to implement it, fragment shader side.

The problem is that since I'll be doing a per pixel evaluation, the initial blast being violent can span multiple pixels. There is already some function to fill a zone but are based on distance evaluation along with uniforms: if the current pixel is within the boundary, it will be set at a certain color. While I only know the distance needed when I hit a specific pixel with high value.

I guess I can try to for loop when needed, but then I'll have to read about circle rasterization and make some shader implementation, also I'll have to think about performance, but I almost fell into the pre-emptive optimization trap.

I'll dig a bit more if there's some function that really do what I want, and if there's not, check on doing it with loops.

>>16025774

To be completely fair it was too small for consumers of the late 90's as well. The laptop came out in 1995 just as the sub-notebook craze was coming to an end. It was a best-seller the year it came out, but soon after the market shifted to a preference for 11+ inch screens.

>>16017310

That tower in the distance looks really cool

>>16019082

There's a pretty good chance they've been uploaded to gamebanana or one of gmod's content websites.

Hello all,

Been lurking here for a long while, decided on making my own Half-Life mod which provides an entertaining experience that is about 2-4 hours long. Been learning about the Half Life SDK and playing with the Hammer editor, is there any anons here that have any good information/tips regarding modding with the Goldsrc engine? Also, is it acceptable to send early design builds to people for playtesting outside of the demo days?

Video unrelated

>>16025971

I worked on a mod for goldsource, but I never finished it.

https://www.youtube.com/watch?v=2vyblEffMYQ

https://www.youtube.com/watch?v=_o8bQk8zF68

https://www.youtube.com/watch?v=NwlY7mJYiJg

I was really bad at programming at this time, so my modifications were pretty basic. I added a few enemies and new weapons but they were simple modifications of existing enemies and weapons.

You'll want to look around at the tutorials on moddb and twhl, as well as look around at gamebanna for models and stuff. Milkshape3D is the modeling program that you probably want to use, since it supports all of the formats that goldsrc uses, but there is probably a way to set blender up to export those formats. You also probably want stuff like hlmv (half life model viewer), etc, there are a lot of cool utilities. There's a lot of information out there…

>>16024918

Thanks anon. I actually came to the same conclusion earlier because I have been feeling like shit for the last few days. We have had a cold wave over here so the windows are closed most of the time.

It's more important than people think.

I had a dumb idea for a survival horror but I'm not sure if it would work and I dont know if I should actually try to make it. Any anons mind telling me if the idea would work?

>>16027289

It won't work

but seriously post it, although personally I think horror is 80% execution.

>>16027289

What's the idea?

>>16027309

Basically the general idea is to eliminate the sense of power that a lot of them have without removing weapons. So I came up with the idea of each enemy being unkillable in the normal sense but above and around them theres a sigil with unique designs. Each sigil has a weakspot that damages and kills the enemy. Because the sigils are invisible you have to search around for notes left by people who have figured out what the sigil looks like and use it to figure out where the weak spot is. In order to make it easier I thought the enemies would twitch in the general location of where the invisible weak point would be. you can also carry an infrared scope with you that exposes it but takes a lot of inventory space however this will eventually be phased out. Later on the enemies would evolve and prevent you from using previous strategies to take them down. I have the world concept in my head but I just wanted to know what kind of gameplay it should have.

TLDR theres an invisible 3D entity above the normal enemy.

>>16027341

Sounds pretty clunky, otherwise reminds me of Alan Wakes mechanic. But like >>16027309 said, im sure it could work somehow

>>16027347

Yeah that's what I thought too when I was done thinking of it

>>16027341

Sounds clunky as hell, honestly. If you want to create a feeling of helplessness by removing the usual sense of power, then do just that. Make the enemies essentially invulnerable. Design the game so any fighting you do will at most stagger or temporarily disable the enemies rather than actually destroying them. Design it around fighting being overall ineffective. Limit ammo, make hiding and avoiding enemies a more reliable way to keep yourself safe, and above all else remember that what the player can't see is scarier than what they can. Keep them on edge with sounds around the corner, even when there's nothing there. Avoid jump scares - they're a cheap way to get a moment of shock, but they don't do anything to build the sort of dread that the unknown can.

>>16027341

I'm having trouble imagining good implementation of this in a typical horror game with mostly-linear progression. Unless it's something like Clock Tower, which requires a lot of backtracking, you would need to relegate this to boss arenas of sorts, and that would require trial and error, getting to the boss, learning that you don't have enough info to figure out how to kill him, then leaving. And really, it just doesn't sound that fun.

For something with more backtracking, rather than sigils, a somewhat similar idea could be to have various elements of the environment act as easy ways to kill an enemy. How you'd implement that, I'm not sure. Could tie it in with some story shit, like, the monster was a person who suffered horrible burns, and you could assemble several items to cause a fire, making him die instantly. You should still leave the option to kill them through brute force, though.

This also incorporates somewhat your idea of reading people's notes about where the sigils are, as this would require you to read writings about the monsters, what happened to them and so on.

>>16027387

>>16027407

Thanks for the input I thought of the sigil thing as a way to discourage players from wasting ammunition and killing everything they see as well as a possible solution to third person survival horror games since the enemies arent conventional. I also wanted to add a morality system where killing more enemies equates to a more hellish world and new areas being opened similar to how demons souls had light and dark soul tendency.

>>16027432

> I also wanted to add a morality system where killing more enemies equates to a more hellish world

That's not a bad idea. A really easy way to get something like that across would be by making the enemies children, whether ghost children or whatever, it's a pretty self-explanatory way to pull off the morality thing, even normalfags are offended by killing children. The question is, how much of an effect would it have? If it's just aesthetic, it would be a letdown.

>>16027475

I was thinking of a different ending new bosses/ enemy types and newer interactions with characters. I have a bunch of stuff written in a notepad but I dont really want to bore anyone with something that probably wont happen.

>>16027492

Word of advice, learn to crawl before you try to sprint.

Keep your ideas written down, but it's a bad idea to become dead set on game development with your best concept as first priority, because when you start learning how to make a game it's likely you will get stuck on one minute detail that gives you trouble and the frustration of not getting it to work will discourage you.

Start simple, start making basic shit, learn what you can do, and after making a few clones you'll have a better understanding of the basics and how to make your dream game a reality.

>>16027630

but the indie shit on steam is playing it safe, they're all the same low quality VN, roguelike, JRPG, low poly or pixelshit and they all share similar themes usually ripping off popular games but doing everything worse

i've got an upcoming gamejam and i'm planning to clone f-zero x

i have a list of nodes for a path and my shit linearly traverses it

all fine and dandy, i know the exact distance that they need to have moved in order to get to the next node

now what do i do if i add the possibility to move left/right along the path? the distance to the next node is suddenly wrong

>>16027630

>>16027553

>>16027645

I think it's a mixture of both approaches, but without either extremes.

As in, not total chaos, and not total order. Or, not going way beyond your bounds, and not "playing it safe".

Have one foot in chaos, and one foot in order; straddling that line. Or, in otherwords, always be taking risks but make them reasonable risks.

You should have instances of failure and success; they're all necessary learning experiences. The difference a reasonable risk retains is, is that you can pick yourself up because you were expecting that you may fail, and you can learn a lot in the process. Or you can succeed, and in the future you'll be able to raise the bar for what you consider a reasonable risk.

One such instance is the exploration of concepts that are a bit out of your reach (knowledge/skill wise), but having a solid plan (design doc/todo list, or just organized execution); while along the way assessing if what you're pursuing is reasonable or not.

So, it'd be unreasonable to pursue an MMO for your first game, but it'd be less unreasonable to create basic networking capabilities for your game w/a library.

So you're always learning while progressing towards a clear goal, and can measure your progress with how close you are to this reasonable goal.

However, this is relative to your own skill, knowledge, and know-how; so the actual heuristic determination of falling under the category of "reasonable risk" is your responsibility and will always be evolving.

There's also the point that if you lack the knowledge/experience of what your pursuing,you cannot properly assess the required time/knowledge/skill, and are in the dark as to whether or not it's a reasonable risk.

This is where you see people saying making a game is easy, or that everyone should just make an engine from scratch "just because it's better"; this is due to them lacking the knowledge/experience to properly assess the time/knowledge/skill required to perform said task (be it for making an actual game after your hand-made engine is a minimum viable product, or be it for taking an engine from zero to hero).

It's also advantageous to avoid the sunken cost fallacy (I have to stick to this, because if I changed Y feature I'd have to spend xxx hours to change everything else to fit), and to prevent a lack of dedication to a project (being a "small demo" project that you don't care about); so it's a constant balancing act.

Though, the last point is a whole post in itself, but it can be summed up as: "make a game that is fun for you, and that you want to play".

This is my own experience, and I'm not sure if it'd work for everyone.

>>16027700

well nevermind, i guess at a high enough speed it's not really noticable

>>16027734

At least this guy gets the idea.

>>16028010

>>16020823

>>16021003

Yeah, the UI should be less obtrusive