/v/ - Video Games★

>>16711799

>>16711799

>>16711811

Are you using godot? It always flips my skeletons in it.

>>16711957

no, unity, but the fault is definitely with mixamo because the auto rigger always gives me the same results if i give it a mesh with no bones, only the character i originally downloaded from them had the wrong bone orientations

they literally just never bothered to update their character library to have the same bone orientations, probably because it doesn't matter unless you want to know where your arm's forward is

You guys should make a thread about your game when it's done. Just saying.

>>16711996

Sure!

Got a list of lines you want me to do specifically?

>>16712117

i'm not much of a writer so whatever you can think of works too

What do we have here

Where'd you come from?!

Intruder!

You won't leave alive!

im new to these threads, just started reading them with the last one after i had a sudden urge to start taking programming seriously for the first time. my real first love my whole life has always been my fascination with computer operating systems/core programming and terry is my greatest inspiration of all time, but i also want to make a game for sure.

in the last week ive put my head to the grindstone and came up with an idea i'm really passionate about, a starcraft: brood war clone, down to the same isometric perspective and 2d sprite graphics. i originally wanted it to be a direct clone or source port with all the same units and just never release it (i'm a neet so if i ever actually pick up speed on programming i'll literally never see the light of day again like terry so who cares), but my idea grew from there into more of an "engine" or base game where players can freely create their own races and add them into the game (an idea i took from total annihilation: kingdom's ancient modding scene).

i'm planning on having a tile-based map editor with pre-created tiles so that all the tiles can support features like zerg creep or more, so all possible races supported by the engine can work on any map. (imagine playing zerg and the map didn't support creep, you'd automatically lose.) at least 4 base asymmetrical races are planned to come with the game, plus obvious stuff like 60fps and lan support. that's pretty much all i got, i think it's a solid design doc, my only real fear is that i'll get slammed with a lawsuit if i ever release it (will be released in public domain) just for being a 2d rts basically since there's really not a whole lot you can do to be different from sc:bw and be fun, and more importantly my engine will literally support creating the exact zerg/protoss/terran from brood war and importing them as a fan mod since i plan to have creep support, so i might get ass raped.

>>16712203

As long as they're distinct enough visually it doesn't matter if it mechanically rips off Starcraft. Even then the worst that will happen is you'll get something like a letter from their lawyers saying you're too close to their IP and to change some things. They don't take you directly to court with a lawsuit. If someone creates a mod that uses their assets that will be their beef to deal with and shouldn't involve you at all unless you're hosting their mods on your server or something, then maybe you'll get a DMCA.

>>16711999

Checked

>>16712230

makes sense. i honestly just realized that i have no fucking clue what to do about resource management, so it might not even have minerals/gas for it to be an issue (though, someone could just make a race scripted to have an incremental increase in minerals over time or something and basically accomplish the same thing). i'm also trying to figure out how multiplayer with download-from-host races with their own scripting would be anywhere near balanced, especially since you can just copy a random hacking race with the same visuals as a competitive one and rename it and use it competitively as a cheater. i guess i could put the race's wad md5 visible in the player lobby so you just glance at it to make sure it matches to know it's competitive, and maybe keep a list updated of "official" md5s that are at least somewhat balanced.

>>16712264

just throw everything in this post in the garbage. the hash of the wad idea is terrible because nobody will want to patch their races to bug fix since everyone will think it's a cheated race.

maybe the brood war "set races, everything else is UMS" system really is the best, i dont know; actually playing an rts modded to have more races makes the game infinitely more enjoyable even if they're trash from my experience.

>>16712264

You could have a system where race customization is point based and clients verify the total points used to ensure its within a set limit. Of course you'd need to limit customization so players couldn't script things entirely, more like custom skills built from building blocks, like: Unit has average stats (100pt base cost) with a stun ability (1.5x cost modifier) that's an AOE (2.5x cost modifier) and pierces 1 armor (+50pt base cost). It would be a lot of work balancing costs but you'd have a vague semblance of balance rather than blind trust. As an added benefit it might make race modding more approachable for those that don't know a scripting language.

>>16712516

thats a good idea, i didnt have anything like that on my radar. i'll toy around with the idea, i really liked halo 3's map building system like that where you had set money and items spawned decreased total money. maybe this could be a third option and just be fun on its own. imagine if you could mix races, like if you took all the points out of siege tanks and goliaths on terran and swapped it for reaver/shuttle.

>>16711997

So that people can accuse me of being a shill while digging into my history for drama? I'll pass and release on other platforms lol

>>16712127

I'll try to get a batch of them done tomorrow.

>>16711984

Could you say the following lines for my FPS game?

"Well, lets get this party started."

"Remember lads… Subscribe to pewdiepie!"

>>16713750

Raul's looking pretty not good, Cristian.

>>16711811

>Bought Fuse for close to full price a few weeks before it was bought out

>I've never been able to get the fucking thing to open in adobe's launchers no matter how many times I spam click it

I'm still mad

>>16714056

if you read enough of the support forums you should eventually find this bat file

works for me so far

del /A:H "C:\Program Files (x86)\Common Files\Adobe\SLCache\Vjd7fUZ1c2UtMS4wLVdpbi1HTXt8fUFMTA==.slc"

"C:\Program Files (x86)\Adobe\Adobe Fuse CC (Beta)\Code\Build\Output\Fuse\bin\Release\Fuse.exe"

>>16712230

>>16712203

>>16712264

Visually, Starcraft basically rips off Starship Troopres, with a touch of WH40k and Aliens tossed in the mix. I'd say take Starcraft and at least one other visual inspiration.

Also, if this is your dream game, set your sights way lower: it might be more feasible to make a shitty game clone (like Pong or Space Invaders) that you can blitz through in a weekend. Build on your knowledge with each little project.

Even the veterans of this thread wouldn't touch an RTS game unless they were crazy

>>16713750

careful with those triangles.

>>16714068

thanks for the advice anon, i really do appreciate it. i dont have any illusions about it, i was basically just blurting shit out so i could look at it and see if it made any sense, and found a lot of flaws in the design already. i'm just drafting something to work on in the far future, i dont really expect it to get anywhere for at least a year, but maybe it'll inspire someone else to get a draft out and start hammering out ideas for something they'd love to do like me.

>>16714163

Dude just install Unity

More useful: If you want to program, you have to install shit to your PC so your code can get translated into an .exe file. This is something called a Compiler. Most languages have one, but some are interpreted languages (eg lua) and don't have a compiler in the traditional sense.

You'll also want to download an IDE of some kind, to manage your code. Even if you neglect doing this, Notepad++ is a wonderful utility program for miscellaneous text shit

>>16714725

>code in an image

>>16714736

It's just smaller on the screen, I don't like to bloat my posts too much with code. It's not very useful to anyone to be able to copy, any way

>>16714725

Can you add this effect to everything that goes above certain speed in-game (projectiles, falling objects, etc.)?

Logo screen has strong Unreal engine 1 vibes by the way, which is great.

>>16714632

neato

i guess i need something for when they die too, if you've got time

>I'll get you… yet…

>This… can't be…

>You… dare…

>How can…. this be…

grunt noises would be nice, although i think mine work well enough for the time being

>>16715297

fag

>>16715291

Probably, but I still wish it could be today forever so I could take it slowly

>>16715358

take a close look at how you are comparing coordinates at the edges of aabbs.

for example, if you are using <= in some places and < in others, it might be causing the problems you are seeing.

>>16712203

Whatever happened to the anon that was making that really cozy looking RTS with the steampunk aesthetic? It had pre-rendered sprites for the mouse, I haven't seen it in ages.

>>16715291

Already submitted.

>>16707265

What's the easiest thing to develop in godot? 

I'm watching a tutorial on how to make a space shooter,
but the version that the guy is using is older and
they changed some functions name and
structures. I'm watching gdquest's videos since
there's a lot and still updates, but there's not a
lot of material outside for the engine. Even the
official site has problems with downloads and
the guide.

A 2D in 3D example, something like Doom 3 computers.

Make a new scene, add a 3D mesh. Give the mesh a script, a new material, child viewport, branch the viewport with a panel and a button. Add a camera, make it current and extend the scene root. Make the mesh's viewport use 2D mode, add a camera to said viewport. Download the Godot 3D in 2D demo project, copy the script on that quad mesh, move that to your mesh's script. Make the script find your mesh and viewport if needed. The script is to draw a viewport's contents onto said spatial mat. Make a second script on a canvas layer (extend scene root!) that tracks mouse position relative to where it is in the root viewport (this should now make 2 viewports). Trace that input from the viewport to the quad, then make that input relevant to the button.

Probably not entirely easy, but there is no way the info for each one of these isn't out there somewhere.

>>16707265

2D stuff is easy because you don't have to fuck around with importing meshes, which is infinitely more finicky than it needs to be. 2D platformer is the easiest of all because you don't need to fuck around with AI. There's a billion tutorials for the stuff, just search for godot 3 instead of just godot.

If you want 3D, I'd recommend Miziziziz's tutorials. They're short and to the point. He also has stuff on 8-directional doom-style sprites and Wolfenstein levels, and that's pretty much all you need for a good, easy to make FPS.

>>16715358

remembered a few places for ya now that my brain's not all retarded. you may need to convert models to use. and you could also just slap a different texture on the cube if you want something different.

https://www.models-resource.com/

https://opengameart.org/

https://www.turbosquid.com/Search/3D-Models/free/for

>>16715232

>>16715102

What's the difference between a framebuffer and a texture you can draw to?

Let's say I have 2000 actors and I need to sort them by their position or some other metric. Their position will update frequently, and they should be sorted at all times, because their update order matters. What is the best data structure for this?

Right now, I'm using an Array, and sorting it before I begin my update logic. It works, but I feel like I'm "doing it wrong" since I don't think it will scale.

>>16717628

>>remove all traces of the thing I added

something something version control

>>16717633

It was literally no more than 6 lines of code either added or modified

>>16717628

which lang?

>>16717638

C. This usually happens when I overflow an array, but I didn't do anything that should affect that.

>>16717643

Are all your variables initialized properly? Doing x; instead of x{} has fucked me more than once.

>>16717313

Only one way to find out, add 2000 actors and see if it still works.

Which sorting algo are you using? Your use case seems to have a partially sorted array, which some algorithms benefit from (I don't remember which ones). You could also skip the sort operation, and move the new element to the correct position as the metric changes, instead of in a separate step.

Array sounds right, but I'd have to know more about the data you're storing/accessing. It also depends on how the language works, since certain languages handle the data differently.

>>16717643

>>16717628

Use the following compiler options:

-Wall (tells you every problem with the program)

-Werror (doesn't compile until you solve them)

-O0 (doesn't optimize out your stupid code)

If you still have problems, use -g and learn to GDB.

I found it

I was iterating a different array pointer twice on each loop, so it rolled into unrelated memory and was modifying that during the loop

It was something I must have pasted accidentally when I added the new thing

>>16717633 was right all along

What a great way to spend an hour

>>16717741

The AddressSanitizer would have caught that and you wouldn't have to guess.

>>16717746

I don't like using debugging tools, they're such a pain to set up for each project. I can usually find the bug just by printing a few values that I suspect to be related anyway. Don't recall ever had a weird and misleading bug like this before though.

>>16717774

Linting your way through can only do so much, it's not feasible to use it when you're doing a really big fucking loop. Knowing the stack trace is a great boon to have when you're trying to squash a bug, I highly recommend learning the most basic debugger.

>>16716391

The overview of the specification for framebuffer objects goes into detail about how it's different: https://www.khronos.org/registry/OpenGL/extensions/EXT/EXT_framebuffer_object.txt . It's the same concept but there are differences in it's implementation and features.

>>16717313

Maybe you could use heapsort to do this, at O(nlogn) it's better than quicksort in theory but is still slow. What are you doing exactly? Its hard to know what the best thing to do is out of context. Maybe you can make additional assumptions about the scenario?

>>16717774

If you're going to use C, learn to be effective with the debugging tools. Become very good friends with them. Speaking from my own experience it has been the key step to being effective with C in moderately larger projects (OpenGL rendering system, scene graph, portal occlusion system, auto-dungeon map builder, etc.). Printf debugging is never necessary. It's slower, it's less effective.

Build any tree or other data iterator that uses pointers into a function. Don't copy/paste these types of routines. Get them right and re-use. Maybe you're already doing this. I hope so.

Reorganize your thinking to not send single data items to a function that processes per frame. Send the entire array and operate on it all at once, or in blocks. Set yourself up for SSE wherever possible as early as possible.

If these three things seem unnecessary or absurd to you, seriously revisit your decision not to use GameMaker, Godot, or Unity.

>>16717791

It's not that I can't use a "proper" debugger, I just don't know how to set it up easily. The best free one I know of is in code::blocks, but for that I need to create a new project in a special folder and import all files manually which aren't in a very obvious hierarchy and I normally use a batch script for compiling so I have to port the correct settings from within that too.

If there's a better way to debug with a GUI and without paying 5000 dollars and my privacy to Microjew then I'd love to know.

>>16717809

Alright, I'll spoonfeed you: ddd. It's just gdb but with a gui.

>>16717814

How about one for the OS that videogames are actually relevant on.

>>16717809

This is a good debugger:

https://www.pellesc.de/index.php?page=download&lang=en

>>16717313

I remember doing some tests about this exact same situation. My end result is: You can have different tables holding tight packed data about your actors, in your example, the tight data would contain only relevant information, which is the position/metric, and the actor ID, that's it. Even if you update the actors and the data table every frame, you'd still get a considerable boost due cache hits. CPU just loves cache hits. I used this for my sweep and prune collision function and the results were very satisfying. But I don't have the numbers anymore. Of course, you only get the speed up when you have an important number of actors.

Also avoid heap memory and memory allocations in general, you can use the stack and allocate tables that will cover all your needs beforehand. In my case, I know that I'll never have more than 10000 objects, so I had tables with that size. That's how Quake did it iirc.

>>16717804

>Reorganize your thinking to not send single data items to a function that processes per frame. Send the entire array and operate on it all at once, or in blocks.

I believe this is a really good habit. But I'm really rusty and I have to ask, by sending an entire array you have to pass the pointer to the function right? if so, wouldn't dereferencing the table each time you want to access an item add overhead per access?

In my case, each function already knew which table to use, you didn't have to pass the collision data to the collision detection function because it only cares and will always care about the one table it was made for.

On a side note

>follow a nice enginedev devlog

>great progress

>interesting techniques

>check the rest of the videos/posts

>it either completely dies or switches to unity

This saddens me, a lot of people break before finishing a game or even an engine. I suffered from this too, part of it is that GUI programming sucks and making good editors is hard.

>>16717809

I use vscodium (vscode without botnet) and plugins for c/c++/embedded, works like a treat after setup and installation of plugins

>>16717693

>>16717793

>>16717899

My game has a hard map size of 256x256 (byte x byte position), logically reduced to 128x128 (anything larger is too sparse for gameplay), and an actor limit of 2,048 (arbitrary amount). The only constraint I have is that each tile may contain 0 or 1 actors, and each actor can occupy 1 or more tiles. This is so that updates happen predictably and I can easily look up what actor might be adjacent to another (eg polling GetActor(x,y) and if null, then I can move into that space)

Right now I sort using C#'s Array.Sort method. Using their provided method for this seems to be rather quick, though I'm sure faster methods exist. It just seems cumbersome to manage several thousand objects and start removing and reinserting them as they move.

>>16718847

fuck me, it was in a different repo

https://github.com/artellblender/collada-exporter-2.8

disregard

>>16718847

Why go with collada over gLTF-2?

>>16718869

Never heard of gLTF2, but I looked it up and it seems pretty good. Are there any good guides on it?

>>16717899

>In my case, each function already knew which table to use, you didn't have to pass the collision data to the collision detection function because it only cares and will always care about the one table it was made for.

excellent way of doing it - if you know the information before hand, bake it into the system. Really good tips and info in general in your post.

>I believe this is a really good habit. But I'm really rusty and I have to ask, by sending an entire array you have to pass the pointer to the function right? if so, wouldn't dereferencing the table each time you want to access an item add overhead per access?

Right, send a pointer to the first element of the array, not a copy of the entire array itself(!)

The compiler will move the address of the array into an address register and increment it throughout the loop. Each item of the array will not have to be de-referenced. Also, operating on arrays of data will allow you to use SSE functions and perform many, many operations in parallel (vector math, and more). If your function is math-heavy this can give pretty big gains. Or just use the Sony SSE math library. It's pretty good.

>>16719178

Of course, just cranking out something fast is what I usually do, and then go back and re-think my data structures, re-arrange the data, and look at the top 3-5 routines that are taking the most per-frame time. This usually leads to re-writing a whole lot.

Your idea of per-arranging just the necessary data for your sweep and prune algorithms is actually a really good illustration. Even if it duplicates the data and you end up updating the data in two places, it can still be the best way to do it.

>>16719263

Yes, I'm making a spiritual successor to a game that itself was a spiritual successor to Supaplex/Chip's Challenge

>>16719285

awesome. I'd play that. Hell, I'd even buy it.

>>16719307

Since you're enthusiastic about it, what sort of things do you like/dislike about similar games?

>>16719312

Shit, I only have vague memories of Supaplex as a kid. Game was way too hard for me back then. I always liked the somewhat freedom you had on how to solve a level. Each map felt like it's own sandbox with various puzzle areas that you could tackle at any order. I also liked how the scissor bastards forced you to be careful on what green squares you ate, and luring them into a trap was cool. Making explosions was always fun. One thing I hated was when there was some predetermined falling mass of balls that you had to position and time carefully for in order to end up with a pile that didn't clog any necessary resources or exit: those felt like they required too much trial and error instead of cleverness.

>>16719290

>blender's gltf exporter doesn't work with Godot 3.1

>the collada exporter for blender 2.8 doesn't work

>no other format supports animations

Am I completely fucked? Is there any other way to animate a 3D model in this engine?

>>16719399

>you can see your own legs

What genre of game are you making?

>>16719394

I know absolutely nothing about this, but can't you, like, open the model/animation in blender 2.7 and then use the collada exporter?

>>16719511

Downgrading isn't a good idea on Linux

>post modeling job on blender forum

>instantly get flooded with replies from pajeets

Is there no escaping from these streetshitters?

>>16719394

It works if you directly export a .dae file from mixamo, but yeah, Godot's 3d modeling support is garbage.

>>16719976

Consider hiring the lowest bidder, just so we can laugh at the results.

>>16719587

Can you post a shot directly from the front or rear? I think her legs are too short, but after putting some thought into it, I think it might just be the view angle.

>>16718898

go to >>>/agdg/. It's currently the third thread in the catalog.

>>16720001

Figured. I think I'll just ripoff Squeenix and go for a "HD2D" style.

>>16722191

aMAIZEing

Anon shmup here.

Beside IRL stuff that depressed and disheartened me, tending the garden a bit changed my mind and feeling a bit better (for now at least)

Had to arrange script management a bit. When actually working with "real" cases I noticed I couldn't easily have multiple move part commands. After some changes and breakage… It turns out I missed my target:

- I have a general array with all script blocks that units can switch around for behavior

- Each blocks can have several commands (move, move part, shoot, …)

- Each command can have several step (move to A, move to B, …)

- Time was checked inside commands against the current step inside a command

- To address those 3 states, I have 3 variables + unit time.

The problem is that I currently check time for the current step inside a command but multiple commands aren't necessarily at the same step, so I would need multiple variables which I don't want.

Current trail: have time checked inside the script block which would make the most sense.

And the hot stuff for later is have a structure pointer and have a data array interpret commands, but I don't know all the use cases yet as I'm already stuck.

That said, when things get moving and bullets flying, it's getting nice.

Still looks like shit though.

>recursive function

>variables suddenly become 0

I don't get it

Don't know if this is the right place to ask, but is Jira really useful to know to become a QA/game tester? If not what would you suggest?

>>16723527

No, it isn't. It's just an issue tracking system. You must have basic knowledge, but it's obtainable in 30 minutes when you are hired. And even though Jira is the most popular, many companies have their own systems.

I guess, you need to have discipline to test everything. And understanding of what's going on under the hood so you could test better. And read some QA books. Preferably those that already employed ones has read. I'm not a QA, so I don't know what am I taking about.

>>16723527

i have a friend who does outsourced testing for microshit in poland and he mentioned that they use wikis a lot. i think if you really want something to put on your resume it would be wiki editing abilities like wiki templates. wiki templates are effectively a programming language (just with a loop limit and no dynamic user input), in my experience they're extremely fun to learn and have a really awesome syntax where you can change function/variable names as the result of other variables, so i'd say it's worth learning anyways for fun. i'm a neet and i don't do testing though so take what i say with a bucket of salt, i am an expert on wiki templates though.

>>16723607

I spent the last 4 hours attempting to write a simple "find the next object in this array", but with a few complications thrown into the function header

>>16723636

Array[++index]

*dabs

>>16723656

What's the point of dereferencing dabs?

>>16723702

dabs is a void pointer, I'm dabbing in the void

I'm familiar with programming and I want to start my own game dev projects. I see Unity mentioned and suggested often but almost never any mention of Unreal.

Is Unreal that bad? Should I avoid it like the plague? I'm new to game dev.

>>16723607

Hard at work.

>>16723780

Speaking as someone who has a lot of experience in both, no, Unreal isn't bad by any stretch of the imagination.

They both have strengths and weaknesses. (From a technological perspective, that is.) Whenever I'm working with Unreal, some of its weaknesses make me wish I was working with Unity instead. Whenever I work with Unity, some of its weaknesses make me wish I worked with Unreal instead.

You can make a point of avoiding Unreal, because Epic is very hostile and shady right now, but that's not a technological shortcoming of Unreal and it's your choice whether you care or not.

What do you want to make?

>>16723787

I don't really have any good ideas at the moment so I'm just thinking of making a single player survival game somewhat similar to project Zomboid but in first person just to learn the ropes.

I was thinking Unreal because I'm really keen on becoming more familiar with C++, but as you said Epic is being shady.

>>16723780

I dont have experience with Unity, but Unreal has been fine so far and i can "code" way faster with blueprints than actual code but thats just because im somewhat biased towards visual stuff instead of just plain text.

Importing assets like 3d models from blender or sound files is pretty simple and worked fine so far.

Only problems i had were a few bugs that mostly appear in newer engine systems and get fixed eventually. Kinda sucks if you need that exact system to work right now, but usually avoidable by waiting a bit before you go for the mayor updates.

Documentation suited my needs so far, although some official guides are several versions behind the current one. Same thing goes for community guides.

Also every month epic releases some assets on the store for free, although the usefulness of those varies a lot.

Sometimes its a cool "liquid in a container" shader, some other times its some half broken "my-first-fps-rpg" style blueprints/premade projects that are not exactly useful once you have started making your own project.

Although i would recommend just looking at some demo projects for both engines and comparing the monetization models. After that you probably have a favorite. Asset production workflow may also be interesting if you plan on doing that yourself.

>ALSO EVERYONE BACK-UP YOUR PROJECT. I ALMOST LOST HALF A YEAR OF WORK BECAUSE I DIDNT!

>>16723839

I'll go ahead and look at some demo projects for Unreal and maybe Unity after for comparison.

Thanks for the info and advice, anon.

>>16723866

>You should start with something really trivial. Pong is a great way to familiarize yourself with an engine, because as simple as it is, it'll still force you to touch:

I do that too when checking out a new engine. I use Lunar Lander/gravity game though. So far I've built it in Blender Game Engine (now dead), C + SDL, C++ & raw OpenGL, Godot. I think it took me maybe 1 hour in Godot, as I knew exactly what needed to be done, I just needed to know how to do it in Godot. So, very much recommend this method. Soon implementing a game loop/movement/UI will become second nature and just learning the tools will be the only missing piece.

My goal this year is to dump out a few twin stick shooters that are actually fun and concentrate on improving my Blender modelling skills to make decent looking characters and enemies with animations.

>>16723866

>>16723874

Some good advice I hadn't thought of. I think I'll actually try doing something really basic like Pong in C++ itself as a programming challenge then move over to unreal and do it there.

>>16723866

>using git

>>16728258

Is there any version control you would recommend over git?

>>16728293

Copy and paste to a new file, then zip the file and put it on a USB stick

>>16728350

And when that USB stick gets damaged or lost? There's no reason to not have multiple copies both online and offline. The number of anons who are willing to spend years working on something but won't spend ten minutes setting up a VCS astounds me.

>>16728388

The real trick is to mail the USB stick to friends and relatives

>>16728518

Okay so what you need to do is is use Pi encoding to obfuscate your code. Since Pi is infinite,that means it contains every digit and sequence of digits within, including itself. This means you just need to find the Nth digit, which can be interpreted as your bytecode

>>16728565

Git ≠ github.

>>16728607

>Pi is infinite so it contains every possible sequence of digits

>By distributing pi you're also distributing every piece of closed source software too

Better get yourself a good lawyer.

>>16728754

>Git ≠ github

Well you never defined where to host it and that's the default assumption.

>>16728768

I wasn't even the one that suggested Git specifically, I just said a VCS. I know some devs prefer SVN over Git because it reportedly handles binary files better than Git.

>>16728388

I just keep five local copies, three on separate hard drives and two on flash sticks, as well as a few online backups. VCSes are far too much a pain in the dick for something that should be so simple.

>>16728768

>someone mentions git

>you assume this must mean hosting on github

Sorry anon, you're just clinically retarded.

Hey, if I'm looking to make an engine like the build engine is this where I should be looking at?

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

>>16729380

What a nodev thing to feel so smug about. Github is what almost everyone uses for online backups, and local version control is irrelevant because the topic was about losing your physical backup.

>>16729562

>Gitlab

>Gitgud

>just using your own server where you host your portfolio site

<hurr durr y u so smug

Imagine being this retarded

>>16729660

Well you can't read so I can't be more retarded than you.

>>16729665

Feel free to point out where I said anything that could be seen as a "gotcha" regarding anything you said. I could use a laugh.

Best of luck to you anons. It's been a hard year. I hope you stay safe and remain motivated though things burn around us.

>>16730234

Don't forget the panties.

>>16731044

There's a whole bunch of tutorials linked on agdg wiki.

>>16731314

The documentation is helpful if you just want to learn the functions of the program. The 2.8 version is mostly up to date if you're using that.

Most books about blender are not very good or are outdated.

How does one add tutorials and cutscenes cleanly?

I'm obsessed with simplicity and performance, and both those things drive me insane. You start with a nice clean code, but then you suddenly need to be able to disable and pause everything including the UI and injecting event system code into everything so you can monitor certain variables. It bloats the code too much.

I'm also curious how to handle compression. Let's say you have a timeline for an animation system, a certain value is the same for multiple frames, how do you minimize repeated data? and better yet, which data structure can be used to hold the timeline?

>>16732491

What I'm doing for my replay system is just noting the players input at every frame, and if it changes from the previous value, then I note the input and timestamp.

Not sure if that would help in your case

>>16732942

Learn how to do one little thing, and then either learn to do it better or learn a new thing. Repeat until game is made

>>16732942

Just build your own dev tools

>>16732942

>without overloading my computer with a bunch of bloated dev tools

Install a software distro & building platform like msys2, and then use that to install the necessary compilers and libraries. Use a lightweight text editor like notepad++ to write code, and compile it through the command line environment msys2 provides.

>that I'll ultimately get overwhelmed by and never learn how to use?

Believe in yourself!™

bought some really cheap animals and props and got my npc creator script to work with as little work as possible on my part. only really have to assign animations to the override controller, everything else gets created/assigned automatically

>>16733154

i fucking love stock bear sounds

i got random idles working

you would not believe how stupid mecanim is again

can't interrupt a transition unless i've specifically made those transitions in the controller and in the correct order

so i can start transitioning from idle to a random idle, but the next moment i might have to start moving

and since i'm still transitioning, i can't go to the locomotion state - end up moving while the random idle animations are playing

my solution right now is to check if the blend time for the current animation has passed, and then crossfade into the locomotion state. but the whole thing would have been a lot easier if i could just be allowed to interrupt the current transition

>>16733898

Scroll through this. It doesn't have to be exact but get closer

https://imgur.com/a/PIA08

>>16733928

You’re the guy making an Eva game right? Why not just use reference pics from the manga/anime

How would an artistically challenged person get decent at drawing? Only very simple sprites for a 2D platformer or something since I won't be capable of anything more anyway.

>>16733962

I'm currently following the old "copy over pictures I like" strategy.

Take the picture you want and put it in one layer, then draw over it in the next layer, in whatever drawing software you prefer. A drawing tablet is not optional, but it can be a cheap one. I like Aseprite because it isn't overloaded with useless options.

>>16733928

Go on pinterest and browse anatomy drawings. Some people make "cartoon" anatomy guides that use more shape based templates that make it look better without being too much.

>>16733962

Also, if someone wants to do collaborate gamedev, reply to me. I'm serious about this and my motivation will be off the charts if working for a team and not only myself. But I can only do programming (or I guess game design decisions etc).

>>16733962

Practice every day for an hour.

Figure out exactly what you want to do and break it up into steps.

>>16733962

Do gesture drawings, build your fundamentals, and copy anatomy books:

https://www.alexhays.com/loomis/

https://drawabox.com/

>>16733991

Tracing will only make you good at tracing.

>>16732942

>How do I get into gamedev without overloading my computer with a bunch of bloated dev tools

Just like

>that I'll ultimately get overwhelmed by and never learn how to use?

You actually get around this by not being a brainlet and slowly learn the stuff you have to to follow your dreams

>>16733850

put all your idles in a subtree, that way the transition between idle and moving is between a tree and a moving animation, therefore it should not matter if it's transitioning between idle animations or not.

In theory

>>16734195

it's always a transition between animations

even so, whether i'm transitioning between states or between animations, i'm still transitioning and one way or another i can't interrupt it

>>16733578

Guys there has to be a basic way to take the UE4 rig and attach it to your own mesh to use the UE4 standard animations.

Why is every tutorial on this outdated and from 2014

>>16734235

>mecanim can't be trusted to work the way you expect it to

Do you fully understand what the, for instance, "exit time" part of the transition means?

That plays a big part on making mecanim work

by setting it to 0 on every single transition

Because I too don't know what it means, but setting it to zero means all transitions start can instantly. it probably means something in the lines of "do not interrupt this animation before X time", so by setting it to zero, you can have the transition start at any time. But who knows what it really means

>>16734276

there's a big difference between "it takes 0.25 seconds to transition between these two states" and "don't interrupt the transition between these two states for 0.25 seconds"

i want a transition to take 0.25 seconds but to still be possible to interrupt it, hence why it's easier to just ignore what mecanim can offer

>>16734283

Doesn't setting the interruption source to "next state" accomplish that?

>>16734352

yea, i guess i just don't want to have to specifically make every single transition, my current solution is simpler when it comes to making multiple controllers that are nearly the same

>>16734377

i don't actually have the transitions in the controller, i'm calling Animator.Crossfade from code, so i can't really define what can and can't interrupt them. My whole problem is that by default i can't interrupt them, unless i've specifically made that transition in the controller

what would you say is the easiest to use 3d modeling software that works with Unity? like even easier than blender and used to make roblox tier stuff very quickly. And I mean VERY quickly

>>16734460

at that level you're better off just using probuilder inside unity

>>16734468

I was not aware that existed, thanks

>>16734231

You can either just export the standard dude mesh from the engine and import the fbx into blender. It will have the full skeleton(its kinda messy) that you can then adapt to your mesh.

That being said, you can alternatively also just retarget the animations, which is basically just telling the engine "this bone on skel1 is this other bone on skel2" once. I prefer this since epics skeleton has a ton more bones than i need.Also i think it auto-fills in if you use the same bone-naming-scheme for both skeletons so stick with epics naming if make your own skeleton.

>>16733962

Make really tiny pixel art to limit the necessary detail and animation like most hack indie devs. That or just recolor Mega Man sprites if you're a mega hack.

>>16733993

I'd be interested in helping you with the art aspect of your project, anon. I'm not experienced making pixel art assets but this sounds like a nice opportunity to learn.

>Move event hooks into a private method called e_Idle() or whatever

>Tighten up the logic on more stuff in the actor update loop

>Somehow make it so that pushing an object breaks the pusher, it basically gets shunted out of the map despite still being present

>Made it so actors can only unoccupy their own tiles, no matter what

Starting to feel pretty tight, but doing the tile logic and implemented general behavior handlers is going to suck

i've got basic faction relations working, and behaviour graphs/nodes to get the ai to talk to the player

example - player has not talked with a pirate

if he's within 10 meters, the pirate will walk towards the player and start a dialogue

if he fails to start the dialogue, or he's already talked to the player, he defaults back to his wandering behaviour

if the player insults the pirate in the dialogue, the player faction's relation with the pirate faction gets lowered by 10, which by default was 0, so now it's -10

and since the player faction is an enemy to the pirate faction, the wander behaviour now treats the player as an enemy, and as soon as the dialogue ends the pirate will attack the player

I posted a new video update.

>>16734849

E-mail me at gamedev@riseup.net then we can find a better way of communication.

>>16736374

Probably an attempt to curtail his funnyjunk admirers.

>>16736374

Don't know why YouTube turned it off by default nice catch. I just enabled embedding.

>>16736378

Embedding traffic actually brings in more from shitty websites.

If Touhou debuted today, would it become even remotely popular?

>>16736918

yes

deer kept running into the water when the bear chases them, so i had to get swimming working since my pathfinding can only stop you from running into a wall and i don't want to have to wall off the entire coast manually

also got that stats panel

>had to reinvent the wheel to implement another feature (might have to do it again for faster optimization)

>LISPfags were fucking right, recursion is fucking based but getting walled by the OS/kernel for having a huge stack recursion is not cool

<framerate slowly dips whenever i do something

>>16737781

You can't do billions of function calls that calls itself unless you split it off (like mergesort) so you don't hit whatever number of functions the kernel think is bad. How does LISP and its dialects fucks it up?

So I'm basically between two choices, either implement (or try to integrate) a 2d skeletal animation system like Spine in my simple engine, or just prerender the animation and load it like a normal sprite.

My question is, can you run out of video memory when you load a lot of images? I see games with a lot of frames in their animations, hand drawn animation like Cuphead and the like, I'd like if there's any resource about how they handle it.

Obviously the best practice for sprites is to use texture atlas, but there's only so much you can put in there before needing multiple ones. I'd like to know if I need to plan to do magic behind the scenes to load assets as needed, which will suck regardless.

>>16730234

These look really good, but then again I am the biggest fan of early century PS2/nip 3D graphics.

>>16715282

>>16737114

Sorry for the wait, here you go dude!

https://files.catbox.moe/1lafug.rar

What do enginedevs to for networking? I've tried to learn to use Windows' winsock2 but I can't find any explanation of how the code actually works, just a rough overview of the theory behind it and then "copypaste this 500 lines of code to see it work!".

Is there a cross-platform library that everyone uses or something?

>>16739087

You'll only be killed professionally, anon. There's no risk except towards your livelihood

>>16739093

girls dont move their jacket out of the way so you can see their tits better anon.

The jacket should go over the breasts but moved slightly off to the side to look more natural.

>>16736946

grassroots, can't you do whatever you want with it (to an extend)?

>>16739087

>But I am afraid I can get killed if I do this poorly.

You will never get shit done if you're such a neurotic bitch you think you'll get killed for making a video game.

>>16739195

I had something similar. While I wasn't planning anything for Demoday, I was making a ton of progress really quick and just hit a brick wall, needing to go back and rework some basics. No idea how easy or hard that will be, but I'm dreading it.

>>16739251

>Serious games for serious gamers

Then you should be so serious that you should work on it RIGHT NOW

>gamedev account

a what?

>>16739228

Unfortunately this is what almost all ideaguys are like, even in /v/ and /agdg/. They have these grandiose ideas about the world and story and what's happening, but the actual game and gameplay is just an afterthought, some vague idea of what you'll be doing, with at most some vague surface-level idea or two that should be included in it.

>>16739275

>leftist revisionism and polygendered garbage

>art

Art is not that.

>>16739272

It's not a matter of left versus right. I'm right wing too but I don't want a game to propagandize to me, regardless of which side is doing it. You could make a fun game about being a communist hipster in a liberal arts college, drinking starbucks and typing opinion pieces on his Macbook if you had fun gameplay in it and the setting served some purpose. So one more time I'll say this, come back when you have an idea, regardless of how small, for how to make your game fun. The story is easy, the gameplay is hard.

>>16739272

if you're making all that shit then why worry about your redpill game now? I have a few projects I want to do, but i know if i want any of them to get finished i have to focus on one then move to the next, just plan out what shit you're going to make in advance and when you get to the redpill game, come back for advice on how to not get killed by the jews for making an unkosher game

>>16738848

neat

>>16738020

there was this neat trick that would split each sprite of an atlas in many small pieces, and try to reuse the pieces that are the same to save up memory. Not sure if it's worth the effort, especially with pre-rendered, where the chances of repeated sprite parts are very slim

>>16739482

Not many people cry about being killed by the jews over an RPGmaker game that says "i think Hitler is a pretty cool guy, eh kills kikes and doesn't afraid of anything"

Is it okay to not make a company? Let's say I'm putting my own engine together, using resources that are 100% under CC, and occasionally making shit of my own, I'll use my own credit card and put a game on itch.io, then when I get enough money I'll put it on steam. In what way I can get fucked over doing it like this?

Having a lawyer and an accountant is simply too expensive, I don't want to deal with company and taxes, I just want to be a freelancer and dev when I can while avoiding all the ugly bullshit related to business. What's your insight for this subject?

>>16739707

Does anyone on /agdg/ even have a company?

>>16739707

There's an anon that frequents /agdg/ that's talked about him being being fined or liable for a sizable chunk of change because he didn't make an LLC to separate himself from his work or something to that effect. It may depend on where you are but I don't think it costs much to register as an LLC to have some legal protection. Otherwise if something goes very wrong everything you own is liable, rather than things that just the 'business' owns.

I wouldn't worry about any of it until you're 99% done and know you're going to deliver something, how much and where.

>>16739707

Start a company. Depending on where you are it ranges from cheap to completely free and the benefits are so good you'd be doing things the hardest way without them.

>Bankruptcy: if your company doesn't do anything meaningful or folds, you can file for corporate bankruptcy and liquidate anything the company owns in order to pay off any debts. In most cases this would satisfy any debtors who were on you for unpaid this or that.

>Income: let's say it's the best case scenario and your game goes gangbusters. Where is that money going to go? You can't just get thousands or tens of thousands into a personal bank account in a short amount of time or they'll freeze your assets and start looking into where the cash came from, assuming that you're a terrorist or a criminal or something.

>Taxes/Expenses: Let's say you need to buy a game engine that's only a couple hundred bucks. If you have a company you could write off a part of the entirety of that purchase as a business expense. This again depends on where you are.

>Grants: Depending on your location you might qualify for a grant, which in most cases is basically free money for the government. This is what is keeping Ubisoft afloat but I can tell you for free that a lot of my friends have made shitty student films, started stupid businesses or whatever and paid nearly nothing out of pocket thanks to governmental small business grants (and in every case, when their shit folded they were basically bulletproof because the government protected them against that as well).

>>16739707

How would you even sell a game as just 1 guy making a game?

>>16739707

If you're gonna make money then start an LLC or whatever your country's equivalent is.

In almost every country, it barely costs anything, and it makes you no longer liable for most fuckups. Don't be retarded and just fucking do it.

>>16739731

>a fucking nigger

>>16739777

Yes, niggers are a problem, trips man.

>>16739731

>nigger and white woman duo

>background is a huge ass grid with solid color

>screenshake for every action

>every action creates a weird delay

>CRT + Scaline memfilter

>game is set in LA

>female companion doesnt take any damage

>raul

How come you made a 3D model of raul earlier yet now its some sort of 2D game?

>>16739777

>wants to make a redpilled game

<one of his games featurs a nigger MC with white bitch sidekick

>>16739731

Do you have any idea why the thread for your game was deleted earlier?

>>16739811

It couldn't be because you turned the thread into race-baiting politics the same way you're doing now, could it?