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

Post a link to the new thread in the old thread when you make one, faggots. I'm getting tired of having to do it for you.

Looks good

Alright so there's been some sort-of progress on the shit I'm getting people together for. Based on input from you guys I've decided the following:

Game is going to be a classic Resident Evil-styled survival horror game heavily inspired by an early-80's slasher horror movie, like Friday The 13th or Texas Chainsaw Massacre. It's going to be set during the summer, and I think I'd like to have it set in the backwoods. Most of the game will take place in and around a large plantation-style mansion in an abandoned whistle stop town. The grounds of the mansion have been converted into a scrap yard at some point.

Anyone sound interested? Since I'm shit at coding I've started writing some 80's-style synth pop tunes and I'm also drafting the story. Once I get more shit figured out I'll start working on models and textures.

I wouldn't mind pitching in to help. While my ability to model humans and draw textures is total ass, I like to think I'm competent at architecture and environmental design, as well as sound design.

Professional project?

Damn son that's like a vast amount of the work

Sounds interesting. Is it gonna be campy as fuck? Id love to help, but my modeling skills are pretty subpar, and I can't program. Can you list what you need, and I'll try seeing what I can help with.

pumpkin I dont want to tyexture it

Why do I have the feeling that anyone in this thread who can program is already working on their own game.

Are there any good libraries for handling 3D physics/animation?

ultra amateur. I've mostly been looking for other folks on /agdg/ who are interested in learning how to work together in development with a team.

I wouldn't be averse to some help with stuff.

I want it to be majorly campy without sacrificing atmosphere and horror appeal. One idea I had was to put a VHS filter over the game with noise and all, so the colors and stuff look like you're watching an old tape.

As for what I need, I can't get anywhere without a competent coder. Experience with about any 3D engine is necessary. I don't care whether it's Unity, Unreal, etc. I've worked with texturing in several of them and will adapt.

I also need someone better than me with modeling. I can barely get rudimentary shapes. We don't have to make something current-gen obviously, but it'd be damn good to make sure you can tell what everything is.

I've held a job writing before, but anyone who thinks they've got ideas that'll fit and has a great grasp of english is welcome.

If you can paint, sculpt, doodle, sketch, whatever, I'll take you. Any kind of art.

If you're even a little interested, get ahold of me via Steam and we'll talk sometime. At the end of the day, this is a hobby project, and the goal isn't to make money, it's to make a game. If we make something good enough, that might change. Come into things understanding that.

steamcommunity.com/id/JustCham/

Because they are, probably. I've been helping people with textures, proofreading peoples' stories, and pitching my ideas since about 2012 on /agdg/.

As a PS, I'm about to head to bed. Don't worry if I don't get back to you immediately.

I guess I used the wrong word, is this a project you are making for the purpose of selling or is this free?

...

My ultimate goal is experience. I'd like to learn something. I don't have any intention of selling it unless when completed it's something that a majority of people would be comfortable asking money for. In that case, I'd make sure everyone gets paid.

*a majority of the people who worked on the project

holy fuck it's late. I really need to sleep.

I'll see what I can do to help, just don't be disappointed as I'm probably less than competent. You seriously might wanna find a programmer though.

Is it possible to make Unity run a 2-D game as smoothly as Gamemaker?

Good luck, and get ready to learn how to program.

run smoothly? yes
as smoothly as GM? probably not

Inside each and every one of you there is a skeleton just waiting to come out.>>10906720

How am I able to fuck up my post twice, fuck.

Do you have any idea what the fuck you're saying? In what way is 60 fps in gamemaker less smooth than 60 fps in unity?

Sound design is fun.

Figured out how to apply separate sound effects to individual sectors rather than action types (boy that's a weird sentence to type out), so I made different space door sound effects for the different types of doors–the sliding door, the blast door, the slow door, and the fast door.

I'm getting a lot of usage out of the General Series 6000 and 8000.

i think hes talking about how easy it is to make the game run at 60 fps.

even unity blows gamemaker out of the water for anything even relatively related to performance

He's a modest guy.

Oh fuck it's terminus.

If it's gonna have clunky tank controls and fixed cameras I'll help you. a bit

This

Reminder that said skeleton sees everything that you see.

Remember that next time you have your dick in your hand and think nobody is watching

Managed to fix the lightmap bug, if anyone cares.
I'll commit this later tonight.

Also got the UV preview thing to stop crashing. If you ever do some immediate GL drawing in Unity, make sure to restore the old render target after you're done.

Because they are, either alone or on a team. If they didn't have interest in it they wouldn't be hanging around here.

The bug boxingdev had with that hammer exporter? I might have to look into hammer now then.
Your slav skills never cease to amaze me, user.

Yes, that bug. Turns out Unity expects lightmap UVs to be normalized, it can't do that while packing the lightmap.
It's not Hammer, it's Radiant.
Hammer is Source (HL2), Radiant is idtech (Quake).
The import plugin is hosted here, though the repo hasn't been updated in a while.
gitgud.io/nogames/trinity

Other improvements I've made since that commit

I'll continue development on this if there's any interest.
There are some things I want to try with it, like;

I've also written some bits and snippets to make the Radiant workflow easier for myself, such as one click project sync between the two programs. Might add that later too if anyone wants. But please keep in mind that cleaning up internal code for distribution takes extra time.

Finding programmers ain't easy. And taking it upon yourself to learn to code takes longer than looking for a programmer.

the ass is evolving

If you use UE4 you can actually get pretty amazing results without a programmer.
I attended a talk from one of the Epic dev's and he showcased a game he made, completely in blueprint and it was pretty fucking impressive.

Now, sure, a competent C++ programmer would get better performance, but you can at least start with that. And once you have something to show off, maybe you'll find some interested people.

Oh and btw, learning blueprints and some general coding skills is pretty important if you want to land a job in game dev, since many designers are expected to at least be able to script some prototypes of their ideas.

But even then, you have to pay the guy, and explain your idea as clearly as possible, while also discussing what makes the game good (and where you could disagree).
And that's assuming you find a good programmer that isn't already working on their own thing.

Belive me, if I could just pay someone else to do it for my- I would.
But I'm going to have to learn code, and make it all myself.

Then I'm gonna suffer getting artists and music, or make it myself and bullshit how my shitty work is a "style".
Just like indies and "pixel art".

I'd do it for you but I'm on an old laptop.

did any of you goys managed to install Renderman for Blender on Ubuntu/Linux?

because this is what happens when I try to do that

NCRInstaller[25072]: ncr Debug: Requesting license from QUrl( "renderman.pixar.com/forum/entitlement.php" NCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve CRYPTO_num_locksNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve CRYPTO_set_id_callbackNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve CRYPTO_set_locking_callbackNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve sk_freeNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve sk_numNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve sk_pop_freeNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve sk_valueNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSL_library_initNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSL_load_error_stringsNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv2_client_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv3_client_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv23_client_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv2_server_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv3_server_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLv23_server_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve X509_STORE_CTX_get_chainNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve OPENSSL_add_all_algorithms_noconfNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve OPENSSL_add_all_algorithms_confNCRInstaller[25072]: ncr Warning: QSslSocket: cannot resolve SSLeayNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function CRYPTO_num_locksNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function CRYPTO_set_id_callbackNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function CRYPTO_set_locking_callbackNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function SSL_library_initNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function SSLv23_client_methodNCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function sk_num NCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function SSLv23_client_method NCRInstaller[25072]: ncr Warning: QSslSocket: cannot call unresolved function SSL_library_init NCRInstaller[25072]: ncr Debug: "There has been a communications error connecting to the Pixar license generation system. an unknown network-related error was detected If error persists, please notify Pixar customer support"

I'm rather new to programming, but I'd like to finally get off my ass and just, like, make game.
The game I want to make would be 2D physics-based with realistic orbital mechanics on a solar system scale. What engine should I pick for it? What tips would you give to a beginner?

I hope you have a passion for math.

Nothing is a better advertisement for artists to want to work for you than making a game that looks fun to play.

I've got no problem with that. Math is fun when you have somewhere to apply it.

Unity has an 2d mode, good built in physics, relatively easy to start with and has a shit ton of resources for free on the net. (Just make sure to use visual studio instead of monodevelop)

Besides that i heard good stuff about löve2d, but i myself have no experience with that.

Don't have great tips, just go step by step, break your game into small pieces and try to make these and use the tutorials out there.

Despite having been responded already, I'd like some more opinions on the matter:

Ey devs, bit of a more artsy question, but still kind of general

How do you get that 3D, "Nineties Arcade machines + PSX" style of graphics and sound?

Low-res textures.
Low-poly models.
Low-complexity design.
Low-precision model placement.
Low post-processing, if any.

Low-fidelity samples with heavy reuse.
Low priority on the "sound" of the sound, more focus on the content, melody.

High color saturation and contrast.

making a group of planets orbit a sun without it collapsing into the sun after a few cycles is pretty goddamn hard, so you might have to fake those. Multiplanet gravity for smaller objects is doable, fun math. If you want objects to also apply gravity to each other you might have to write a quadtree to get nearby objects in an efficient manner, or find some other method for doing that.

Affine texture mapping.
Per triangle / per object sorting.
Low quality animation.
Paletted textures.

I haven't felt this awful in months and this has completely ruined any chances of me getting anything done today. Why can't normalfags accept that some people don't sleep well on those big, retarded mattresses everyone buys?

I was thinking about getting real orbital data from NASA and working off of that. I figure Earth isn't falling into the Sun or flying away any time soon.
How exactly would a quadtree accomplish that?

I have the most experience with C# and I've heard some good things about the 2D mode, so I guess I'll go with that.

You might want to see a doctor about your spine. Otherwise you might end up with some serious vertebral problems further down the line.

don't forget about using sprites as background objects and other stuff you don't have to animate very much, and that most players won't pay much attention to.

From what I understand talking to my master programmer friend, blueprint can handle most small projects without any performance hiccups. But if you tried to make anything big or with lots of scripted events(ie: if you tried to recreate half-life 2 in UE4 using blueprint) it can bog down performance.

Do what suggests. Make super simplified models(We're talking slender rectangle as an Assault Rifle) for placeholders and throw a fun prototype together. You can attract artists with a good idea.

(checked)
I am sure the gubmint uses some kind of unknown weapon to target people like you, unfortunately it works only on mattresses with metal/iron coils on them.

I used to have a lot of back problems but those went away after a bunch of chiropractor visits. Sleeping on hard surfaces feels more comfortable than on soft surfaces anyways and my back is in pretty decent condition now.
Are you implying that sleeping on hard surfaces is somehow going to give me back problems?
sage for non-vidya

I think stuff like hl2 would't be so bad, since it's level based. The real problem starts when you have either open world or extremely large levels (Think of witcher 3) where you have to stream content.

In a perfect world you would have programmers who make the underlying systems and expose certain interface like functions as blueprints to the designers, so they can create the game.

Sadly from my experience most designers don't touch the engine apart from changing some values on objects. And don't get me started on artists. Worked with some, who didn't even install the engine, so i was stuck importing their stuff.

NCR fags blown the fuck out legion best faction win every time

lel

I personally sleep on the floor on just two thin blankets sometimes, really comfortable for some reason.

That's exactly what quadtrees are for, finding objects in specific areas quickly.
If you mean how they work, basically you divide the entire map into parts, and divide those again etc. until you get a sane amount of objects per part, and then when you want to get all objects near an object you just look which part it's in and get the list of objects from there. That's a lot faster than checking every object on the map to see if it's close enough.

Do remember than as of UE4.13 blueprints get compiled instead of being interpreted, so they might be even faster now. I haven't really ran any tests yet though.

You could also just recompute gravity less often for distant objects, since the falloff is smaller. Cache the combined gravitational force, recompute it every N seconds or miles traveled.

folds!

how high-poly are you going to make this? Usually stuff like folds in clothing are done with the normals and color map

Assault class demon type with extra junk ammo

Thanks for making search trees click for me. For the longest time I thought they were just a novel data structure to hold objects and it "somehow" whittled the choices down to one.

I understand how they work but I assumed you just always got a single item from them, whereas you just grab a subregion to run a search on a smaller chunk.

not sure how you'd compare two nearly adjacent points in different areas though, like X= 0.1 and -0 1 … would you just grab the parent node (which could contain many many more points) to guarantee you hit it?

But to get those details on the normal map, you first have to make a high poly sculpt, and then bake onto a lowpoly retopologized model.
I'll rework the bum folds now, though, they feel a bit off.

Oh shit, is that true? Guess I'll dip my toes into blueprint again. I just hate having to constantly download updates for UE4. I haven't made much outside of small prototypes, but if the newest version has better blueprints I think I might just stick with that one

I do shit in reverse then. I paint on a texture and then bake a normal from that before scaling anything down. Just the way I learned.

There's just different workflows for different results. Sculpting first gives you complete control over the model, but is more labor intensive.

I've never made it to the normal map phase of modeling, is it difficult?

I don't really model much, but I've done a lot of texturing and baking maps for people, and here's my method.

First, I take the UV map for their model and paint a texture on using either Substance Painter or photoshop depending. Then I take that and run it through CrazyBump as a "photograph". From there, I tweak the settings of the different maps and then bake them all out. Boom, they're ready to go.

depends on the workflow,
most common usage of normal maps is pretty simple, you make a very high poly model (by sculpting, scanning something in, or just manually modelling it), then making a lower poly version of that model, UV unwrapping it and letting your 3d program bake the high poly model's normal onto the low poly one.
You can technically paint normal maps manually, but that's mostly for little adjustments and specific things.

And how rough is painting them manually? I know my way around a tablet so it can't be too awful can it?

When close to a border, you get the region(s) close by. Which if they're not too full shouldn't be a problem. If they are really full, you only get the regions closest to you.
Just taking your own parent node wouldn't necessarily work, for instance if you're in the top-left node of your parent and you're in the top-left corner of your node.

During a quick search I stumbled upon this too, seems pretty relevant. I've only worked with quadtrees for collision checking myself, so I didn't know the specifics for gravity, turns out there's specialized quadtrees for that.
mathandcode.com/2016/02/16/quadtree.html
tl;dr you calculate mass for each node on each level of the tree as if they were one object with the collective mass of all objects within, then traverse less deep into the tree for further distances. That way you can do gravity with every object in the scene instead of ignoring far away objects entirely which would be a problem for heavy objects like suns and black holes.

Fuck, guys who got ahold of me on Steam about the 80's horror game, shoot me a message when you see this. I accepted the invites late last night and can't remember what your names are.

…what horror game?! This sounds interesting

Well the difficulty is wrapping your head around which colors are which directions. You'd get the same result (but much easier), if you just copy your model, and sculpt on it with your tablet, then bake that onto the original model.

I'm looking for folks interested in teaming up on a project. The short and long is that it's an 80's slasher flick-themed classic RE-style survival horror game. I had a couple of guys who were interested but I forgot their steam IDs.

I have some experience in UE4 with C++ and Blueprints mostly blueprints as well implementing animations and AI in UE4. I'm not an expert by any stretch but I should at least be able to help get the project off the ground and I think working in a team would be a good experience. If this post gets dubs I'll put my current project on an indefinite hiatus from my current project and help you guys out. If not I'll just keep working on my current project.

You import a reference image of a spherical normalmap and just pick the colours of the direction you want. I also remember seeing a video of someone who wrote a plugin to draw normalmaps based on the angle you hold your tablet pen in, but I can't find that video. Plus you'd need one of those more expensive tablets for that, anyway.
I've only ever seen artists make hand painted normalmaps for hand panted textures though.

off by one
rip teamfags

If they can get out a design doc doesn't even have to be good or detailed, just has to be more than an idea and some concept art I will definitely get on board and contribute what UE4 knowledge I have. It's just I've been on two game design teams that fell apart the moment anyone had to do any work so I am cautious about abandoning my current project to work on something new with an unfamiliar team. I'm torn between my desire to join and work with a team that cares about the game they are making and the memory of the times I've been burned.
But if kek blessed my post I would take it as a good sign and join even without evidence that progress would be made.

I still haven't fixed this

almost done with the clothing part, gonna do the hands, feet and head now, after that - cleaning up the sculpt

3d porn when?

Are you going to bake all of those details into a low poly mesh?

There's already 2d lewds of her, I'll probably do 3d porn sooner or later, just not with this model.

yep

You can't just say that and not post them

that's one glat ass. What are you, Irish?

Sage for off topic.

com'on user, plump her up
Otherwise, it's looking great.

I posted them a couple of times already and I don't want to come off as THAT big of an attention whore, but wow I guess someone saved them

I'll fight you.

Nah man, flat is best

I'm getting mixed signals.

bring it potato farmer

Gonna go on the record and throw my vote in as anti-flat

Do what ever you want.

with hips like those a flat ass just looks weird

lewds are a service to your fellow anons

how? why? You just trying to be contrarian?
Understandable for a flat chest, but a flat ass is just… no thank you. Much prefer one with at least some roundness.

this

...

I suppose it was the folds that made the ass look flatter than intended, is it a bit better after the adjustments?

what ass?

much better

indeed it's better, nice

will there be jiggle physics ?

Alright, glad you like the revised version, thanks for the input.
Here's my progress so far, compared to the ref sketches.

I'm gonna focus on other gameplay stuff first, then I'll experiment with little extras like that.

i just dont know

can anyone point me in the direction of what I'd need to learn to make homebrew games for the 3ds?

Once again, if you're one of the guys who was interested in teaming up for that horror game project, give me a shout via steam or something.

Good job user. Looks good.

I just need see some concept art/design doc then I'll be on board.

I recreated the entire thing with only the ability to move around and hit another object that is exactly the same as the green squares in that .gif, it still doesn't work. I tried turning the gravity on and it still didn't do anything. It also seems to have nothing to do with if I'm using collision shape or collision polygon.

That's gross fam.

Looks good, holy shit, add in the other parts though.

It works really well. Give CrazyBump a pirate some time.

I've watched other people do it, It always looked lazy to me compared to baking from a high detail mesh.

Oh yeah. It's not the best way, but on less detail-oriented games it'll work.

To reiterate for anyone who might be looking for a team to be a part of:

The game is an 80's slasher film throwback in the form of a survival horror game. Not like "walking sim" survival horror. Think, like, early Resident Evil and Silent Hill games.

We've got a modeler, writer, and texture fellow, but we're looking for a programmer who's experienced in Unity or Unreal or any 3D engine really.

If it sounds interesting and you're willing to commit to a hobby project, get ahold of me on Steam and I'll get you an invite to the Discord server.

steamcommunity.com/id/JustCham/

Someone who knows OpenGL wants to tell me why matrix transformations are applied in reverse order despite the fact that it's completely counter-intuitive and there's no benefit in doing so?

don't count on me, but (I think) I could make the base for a RE clone in a few hours in godot.
I don't do steam and is very very probably that I wont do it but if I actually go through I will be posting here and in the next thread

Not trying to sound mean, but even though this is a hobby project (aka no super-tight schedule, no deadlines for the mostpart) I DO still want people who will be around and who will stick with it from start to finish.

No idea about OpenGL, but there are very specific rules on how to work with matrices.
Depending on if you multiply first the rotationmatrix or the translationmatrix with a vector, you will get different results.

calling glRotate() then glTranslate() will be like translating *then* rotating. If I want a planet to orbit the sun, I have to call glRotate(timeOfYear) *then* glTranslate(distFromSun)

It's ass backwards and OpenGL authors did it because it matches the most common math notation, but fuck them. I'm asking how to get around that.

What do you fags think of having both pre-defined dialogue options for the basic stuff with talking to followers and then having manual input for just minor things like "How was your day?" or "What do you think of x?"? Waste of time and effort or a way to greatly improve immersion?

I like it, Teaching Feeling has something like that and IIRC it gives you alternative methods to activate scenes and improve your relationship.
Would it improve immersion for you? If yes, do it.

Well, if you want a way around that, you can just write your own function where you do your own matrix translation.

waste of time and effort imo, but you should do what you want.

How do you enginefags deal with cross-cutting concerns?

C# Attributes

what is the top five things that would be interesting to see in a rpg maker game?
I've seen a SHMUP made in RPG maker and a live battle system akin to Zelda, but not much else.

Using C++ here. I was thinking of making my component lists global, but I feel there's a better way. I'm not sure how the DOD people do it.

no offense taken, I will (try to) do something for myself then

some progress:
finally managed to get collisions to work
it turns out you actually need a collision shape besides the rigidbody
also did some building blocks and implemented character relative and tank controls

While this is cool, our guy is mostly experienced in Unreal. If you'd like to hop on board, you're always welcome, but I really want people who are going to stick with a project.

Speaking of which, I've got the fixed camera working. There's still a lot of work to do but it's working in it's most basic form. I'm trying to get it so that whoever does the level design can do whatever they want with camera placements and they don't have to worry too much about where they place the camera triggers.

That's FANTASTIC!

Progress!

(Checked)
I might use it as another way to directly ask a question to try and call out some of the character's bullshit instead of having a another option to add on top of that. Makes you want to take some kind of note taking but that really depends on if the player wants to keep note of that. I might have to test it out on some players whenever I get around to doing it.

Mind if you elaborate? I'm genuinely curious since it's something I've been debating over.

Pretty good.
But you won't be able to entirely avoid camera triggers, so there should be some support for those, maybe as volumes that force a specific camera to be used while the player is touching one.

Suppose I won't need to help that user since you're on it already.

Looks good, dude. I dig it.

Not the same guy but I share his opinion, it's true that the "input your own question or statements" feature is a huge immersion-boost, but 90% it's not really worth it, not as a player nor as a programmer.

If it's too simple or normal like "how was your day" or "whats your opinion on X" it might as well be a button, I as a player, am going to get tired of writing the same thing over and over and it won't feel any better than pressing a button, unless you make it so it detects your tone, so "how was your day?" and "how are your buttcheeks feeling today pussyfart" would give different outputs depending on who you ask or when, but this is a ton of work for you as a programmer, since you have to make it so it detects the tone and write and apropiate respose depending on who you are talking to, and a ton of work as a player too, since once you figure it out you have to try different tones with everyone about anything, and 90% it will not really be interesting. also, there's nothing more immersion breaking that treating somebody like shit, receiving a "fuck you asshole", and immediately talking with them again and have them act like nothing happened, so you now have to take mood in account too.

Now if it's more specific like "did you really saw Maxwell take that hammer or are you full of shit?" or "Do you really believe Dr. Robert is dead? Because I think he might be hiding somewhere" Then it really ticles your immersion hard, you really feel like your interacted with this living enviroment and not just bip boop quest completed, but this is really a nightmare both as a player and as a programmer, since you have to check for a billion different posibilities, and for the player since you might be missing important info because you didn't thought to ask, or couldn't find a combination that the game wanted to take, so you are missing content.

All in all, player input is a neat idea, and really nice when it actually works, but most of the time it will have you cursing or tipying the same thing over and over. For a more concrete example of what I'm saying, check the game Façade

I'm starting to believe OOP is just a fucking meme.

This vid has literally stated every single major problem I've faced in the past, and honestly I don't see why objects at the top of the hierarchy aren't just global namespaces because 99% of the time they contain data the needed in a different branch of the hierarchy. All the ways around this void the point of OOP and are retarded.

That topology

This is what i can tell you for graphics

github.com/dsoft20/psx_retroshader
Texture filter mode on all textures -> point
All texture resolutions somewhere between 32x32 to 128x128 max
Screen Resolution 320x240 or 640x480
Turn off shit like antialiasing or ambient lighting or any Unity extras, except Fog, make Fog black
Apply pic related standart asset unity camera effects to camera: Motion Blur, Color Correction Lookup and Color Correction Ramp. Use the pic related Ramp. Ramp should be a black to white gradient with 64x1 or 128x1 pixels. Use the ColorEnhanced3D16.png from standart assets for Color Correction Lookup. With this you should have a little bit of motion blur smearing like the original PS1 and original possible technical colors for the ps1.
Make all your models 500-2000 tris max, download some original ps1 models from models-ressource for reference.

There might be more shit to add, but this is how far i got with reproducing all the technical ps1 details.

That first models pretty great

That car makes my dick diamonds.

Good information, thanks user.

The guy is right about high-level object-oriented programming.
He's wrong about pretty much all of the other shit.

If you're not using OOP to roll your own datatypes and leave it at that? You're doing it wrong.

A lot of absolute retards in the OOP world say to themselves Let's use a singleton! when they need to pass data around or provide a way to pass messages between different objects. Now granted message passing is a major hurdle and the reason why OOP sucks: it's used at high-level code where message passing is very important. And now J Random Programmer is rationalizing his singleton decision over using a global as it's safer even though it has the same access privileges as a global. Solution: don't use singletons unless you've very carefully read the Design Patterns section on them.

Also: In Java, they force the OOP paradigm which is never a good idea and makes high-level architectural programming hard. The guy is 100% right about that.

But there are just as many problems with writing pure procedural: You'll still have to pass the same data around but instead of doing so from an object hierarchy? You have to do it in a call-graph hierarchy. And you're doing that in an unprotected manner. Remember: in order to write good procedural code? You have to minimize state in the functions. So you'll be using re-entrant functions. So now you're manually juggling state for non-existant objects which is inherently 'less safe' according to the OOP folk but easier to deal with according to the procedural folk/video guy.

Maybe I only write pet projects and never run into these issues. But how about only using OOP for low-level stuff and rolling your own datatypes? Polymorphism is nice to have when you need to store all game enemies/mobs into a single array and update them each frame. But using polymorphism to swap out different 'strategies' for dealing with things seems a bit… overkill.

tl;dr: OOP can suck, it doesn't have to, try to minimize it and only make your own datatypes for low-level shit. It makes doing shit higher-level easier. Try not to go overboard with OOP, if it works? It works. At least try to follow some sane guidelines set by yourself?

I used singletons on my first attempt at a game engine and it fucked me over hard.

I threw out polymorphism for entities and adopted the DOD style components. Whats been killing me, even when I was using OOP for virtually everything, is basically the picture you demonstrated. I would have to send like 5 references back and forth between constructors and shit. A workaround I decided to use was to make a pipe object that basically connects everything, but even that was a bandaid on the issue. Right now I'm thinking about just using global lists for my components, and maybe encapsulating them through friend functions.

Every programming paradigm and development method is a meme. There's money in selling books and speeches to clueless managers and posers.
Just program, motherfucker.

After I got my seifuku I code 200% faster and with less than half the bugs fagster.

Is this like a mene or something?

I have read so many different articles and opinions about design patterns it is just ridiculous.

Everyone has an opinion and everyone states their opinion as fact.

THIS IS BAD FOR THIS REASON YOU SHOULD DO THIS INSTEAD

Then you go ahead and do what they said, until you run into a different fucking problem.

Software architecture and design is just a case of balancing pros and cons. Every single design decision you make is going to have a downside somewhere along the way. Once you have implemented it a certain way you will try to implement a new feature, only to realise the way you've done it makes it really fucking hard. That goes for singletons or injection. It doesn't fucking matter in the end.

What matters in the end is that you get shit done. Try to minimize the downsides as much as possible. Use whatever will get the job done to the quality you need it to be. If you are working on a module and a singleton will get the job done and on time, use the damn singleton.

Over time I've come to the conclusion that nobody anywhere really knows what the fuck they're doing. And that's not just programming, it applies to almost everything.

And you can't differentiate those people from the handful of experts who actually do know what they're talking about, so you should take everything as just one option, not as the best one. I also get the feeling that said experts don't megaphone their methods as the "best one", because they realize that nothing's ever best for everything.

this is correct

I wanna fuck his boipucci

Yeah, gonna have to agree with you on that.

What I'm saying is: don't go overboard with OOP, use it for small stuff, contained items, hide complexities behind single things like vector calculations or representing an item in a player's inventory. THINGS THAT ACTUALLY HAVE STATE.

Using OOP for anything that's not state? Yeah, just use namespaces for that.

This is what I think is the best 'middleground' and is what I personally use. This is opinion and personal experience.

Yeah, no one knows what they're doing other than this doesn't seem to kill me or anyone else and seems handy, I'll use this!. [spoiler]And then rockets start flying upside down.

Shit's fucking legit, man… /cuteboys/ programming team working on gaymes when?

See? I even forgot to close my god damn spoiler. That's how much of an experte I am.

user, that's a girl.

Jetpack powered double jumping for your favorite little robot!

Why do you need that much fps if humans can't even see past 30

Speebot can see beyond 220fps. No discriminating against Speebot allowed.

There's an fps slider for non-Speebots in the options

It's 24, idiot. That's why animation is 24fps. 30 is just a marketing trick that appeals to autists who like round numbers.

looks comfy, all you need to do is strap in some tank controls

chlino is a girl though

I am this close to giving up on UVs

Dat FPS

Reminds me a bit of Spyro, or Ratchet and Clank.
Keep it up, looks comfy.

up to 80 actually, i think. it depends on lighting conditions.

sage for taking the bait

/cuteboys/ programming? Aren't they just being sad all the time they can't become biological girls when they should just be cute boys crossdressing and live life normally.
Do not get me wrong i like the occasional trap.

Also sage for off-topic.

I can confirm /cuteboys/ are just attention whoring borderline trannies. Their BO single handidly killed all other gay boards on Holla Forums for attention if that tells you anything.

Looks comfy as hell, would actually buy instead of pirate.

It's not a bug, it's a feature unique art style.

The textures just need to scroll and flow around. Maybe add some transparency here and there, with slight strobing.

Sounds cool man I can sketch stuff if you want digitally and on paper,but in terms of game stuff like coding and modeling I know jack shit about that stuff.

What's wrong?

dying GPU aesthetics

Any musicfags in here? If so, what music program do you use and how did you learn to make music?

Too bad I'm using Unity with C#'s "fuck you, everything is objects" attitude.

Use a static class then.

Namespaces are just a way to organize classes

Static methods should be used to provide global-level and universal functions for your object type; specific object behavior should be handled with methods on the object itself, or with an additional object handler type (if you're using POD objects and are a stickler)

I'll do what I want, corporate monkey.

You can do whatever the fuck you want, especially if it works. But a language is a tool, and it's been designed in a certain way, and if you don't use it in that manner, you'll probably end up shooting yourself in the foot longterm

...

nice tranny propaganda, libreboot!
reported

Why do you need a fucking window to do rendering if it's just a specification? Isn't the point of implementing it yourself to define how things behave? I don't want source code, I don't want to compile shit. I just want to take a .dll, write code that uses it, and press F5 to test shitty game mechanics

...

I JUST WANT TO WRITE C# CODE AND MAKE IT DO THINGS WHY THE FUCK DOES SETTING UP AN ENVIRONMENT HAVE TO BE SO ARCHAIC

...

How people manage to live with Windows as a development environment is beyond me, especially when you can cross-compile to OS X and Windows from Linux but not the other way around.

Despite my posts, my code itself is really good quality; it's fast and modular and well documented. It's just that I turn into a lobotomized retard if I don't have Visual Studio and XNA spoonfeeding me all the environnment shit

How do you cross compile to OS X from linux?

Because fuck you, is why.

You don't link against OpenGL because OpenGL is a driver. You don't link against drivers. The only thing you "link" against is your OS, in which case you just use OS-specific functions. People never write tutorials for how to use OpenGL on Windows because they'd much rather have you use an abstraction layer so that you write code that uses OpenGL on ALL platforms instead.

Secondly, you need a window to create an OpenGL context because OpenGL is first and foremost an API for rendering stuff to a window. Things like saving the frames you've rendered to a file is an afterthought.

Then just link against SDL, and use SDL's SDL_GL_CreateContext to create a context, and use SDL_GL_GetProcAddress to get function pointers to all the OpenGL functions you want to use.
You can be a piece of shit and use wglCreateContext and wglGetProcAddress instead, bypassing the need to use SDL entirely, but then you get to go through the fun experience of having to write your own code to handle window creation/events with fucking Windows.
Granted, SDL isn't perfect, but if you're learning OpenGL it's perfectly fine.

Well… I can't even get my compiler working manually here, I'll just wait until I get home

github.com/tpoechtrager/osxcross

wew lads

Still no dice duck user?

wjat

Still no dice, as in, you couldn't make it work yet, I was asking

Duck user, that's you, the user that posts the duck with the green blocks

no

hang in there, there should be a tutorial for colisions of whatever engine you are using out somewhere!

you're a faggot

n-no

Will there be an option to be able to continue to move around as dialogue is being displayed, or to skip dialogue entirely?

after several days of writing code 1 total rewrite and several days of debugging
My line of sight system finally seems to be working
Now I'm gonna make you guys test it.
Heres the source code which you will need gitgud.io/greenman/rts.git
This should be playable on windows puu.sh/rC7CW/fc01818d4c.zip just place the executable and the dll in the folder from git
If you on linux the project should still work but you will need to compile it yourself.
To add more walls you go to the map folder and change the values after walls
To change the shape/size of the walls you will need to go to default.ini and change/add the wall definitions which well then allow you to add them to the map.
Any time you change the position of walls on the map you will need to turn on terrain image generation
to do this go to default.ini and change generateTerrainImage=0 to generateTerrainImage=1
Also you should be able to enter inside walls so any bug you find where the unit is inside the wall doesn't matter.

Famous last words

well you shouldn't be able to enter the walls at all but i currently don't have collision activated for the walls
I'm honestly not even sure why i't would be any effort at all I guess it makes it easier to test LOS under weird angles.

Also i was just thinking I'm using a sprite sheet for visual
I wanna do 16 directions for my sprites the sprites will be 100x100 ingame and the average animation might have ten frames
That would take up 16000x16000 pixels on a spritesheet and that for just one animation
on average units will have probably around 5 animation there will probably be 20 units
that would take up 1600000x1600000 pixels and thats excluding literally everything else that needs to on there
How the fuck is this even reasonable currently my spritesheet is 8000x8000 and thats already pretty fucking large.

Well, look at a game like Diablo 2. Each character was made out of several sprites and patched together, each one being like 32x64 at the absolute most. You had a head, shoulders, body, legs, separate arms, plus a huge amount of weapons and armor, 8 frames of animation and 16 directions per character. There are literally tens of thousands of images compressed into the game's MPQ files. Monsters are much simpler because they don't wear equipment and have a simple attack/cast/hit/death animation. They wouldn't have loaded everything into graphics memory either, maybe the monsters known to be in the zone, plus an appropriate set of sprite data for the characters' and their equipment. Environments are just regular isometric tiles and props.

A few possible suggestions, though not all might be applicable to your units
Also, this ddoesn't change shit for the amount of pixels you have in memory during gameplay, but split those sheets up into one per animation because opening and editing such a large image would be extremely painful for you

Like I've said to everyone else: If you're willing to commit and stick to the project until it's completed, then sure. Before I say yes or no to anything, can I see some of your other work?

Anyways, since I can't dev proper today, I guess I could try ideaguying for a bit. There isn't really any equivalent game to Animal Crossing on the PC, is there?

What I'm thinking is a game with a number of skills for crafting, gathering, fighting, etc. You'd start with nothing and a crude campsite. After you complete a few tasks, you'd attract NPC villagers to settle a plot of land near you, like AC. They'd have favor levels, and every day would ask you to complete simple tasks for them; each would have a different set of 2 or 3 skills, so they would actually act as randomized skill trainers.

Initially they'd have crude shelters, and eventually cabins and respectable housing, as long as you keep feeding them resources. A blacksmith NPC would build a forge beside their house, for example, while a hunter might build a skinning rack. This gives you access to their crafting features without necessarily needing them yourself; once your town gets full and you have all you need, you could boot them out and get a new NPC with different skills and build your own tanning rack.

Another feature would be a daily market. After you get a few NPCs, you'd attract trader NPCs; basically from the same pool but they don't settle, they just show up for a day to trade with you and others. This is a good chance to get training in foreign skills or buy rare materials. Once you reach a very high level, you can build a guild dedicated to one group of crafting skills, which would give you access to the highest level of recipes within that skill.

Does this sound fun?

Tried this before?
adarkroom.doublespeakgames.com/

was at work all day, didn't c your response this morning

It's mostly this
My main concern would be that it's really hard to get this right (as mentioned in previous threads about proc gen, same applies here, nearly), and tbh I was thinking about doing something similar… but opted for a way simpler, and more intuitive approach that's a re-imagined method of applying the core idea.

So, what I realized was that this takes A LOT of time with w/e method you want to use for manual typing w/varied responses/etc; in addition to it becoming very generic if you use one of the easier automatas; such as state machines.
The only way to really get around that generic-ness is using a more advanced approach, which is extremely time consuming, requires a higher mathematics background (or a lot of time/dedication/effort)… and this more advanced method being natural language processing using neural networks (this is mostly understanding the logical meaning of the sentences, and not directly the answering, but it can "spice up" the answers as it can word things differently; i.e. input training set data of similar sentences, and output mostly "logically correct according to training data" but varied responses).

It is a really neat idea, but to do it right (at least, up to my standards, when I was considering this) is just so damn hard. Else, even if you do the approach of using a simpler automata w/fully manual typing it feels like a knock-off or generic, and is again very time consuming if it's a full length game.
Imo, it's better to directly limit the scope, and that's what I'm doing (I don't want to go into the "how" of my approach, but the core idea of a "manual typing" system is still there yet reimagined, but limited, yet not so limited or a broad scope that it's respectively underwhelming or overwhelming).

Well, maybe… There's not really much you can do with that, except click buttons to hoard things and wait for bars to fill up so you can get more options

sure do you want me to post it here ?

Yeah, sure.

Very nice, very nice indeed, thanks buddy

How do you get your GPS to be that high and how do you plan to keep them that way on toasters as well?

ah yes

I know that feel But I'm the reverse.

Finally felt alive enough to get back making progress. Started fleshing out the sprite sheet and putting the UI together. Added gamepad support, but need to figure out a decent way of handling assigning stats compared to keyboard just being able to use number keys. Maybe a menu focus button or something.

Server side is handling all the collision detection, resolution and pretty much everything else to avoid cheating. About ~0.2% CPU is used per client so far. Updates get pushed to clients and they interpolate locally to smooth movement out.

>not listening to yurobeat while programming for MAX WORK EFFICIENCY
Look at this tranny and laugh at him.

Thanks for the insight both of you. It's something that's been on my mind if possible since one of the key elements of what I want my game to be is really immersive and having this kind of possibility to directly type in your own options for very specific things along with the per-determined options to make talking through your men a lot quicker just for basic things that you will commonly do. It's something I'll still end up toying with but I won't have high hopes being able to implementing in the end.

The dialogue pops up without interaction, just move close to an NPC. Move away and it disappears.

My engine supports variable framerate, and the user can cap the FPS anywhere from 30 to 250 in the options menu. FPS only affects the rendering, not the logic. The logic is tied to a separate loop - 64 tick rate. There are a lot of things to tweak in the options menu, although I'm sure this will run fine on toasters anyway.

Sorry for being late I was busy with some stuff,If you want to see more work let me know

TOP TIEROPTIER

Top tier taste.

This is good work, user.

quality stuff user, now… lets see them lewds

Nice stuff. If you're interested in hopping in the discord, add me on steam and send me a message saying who you are.

steamcommunity.com/id/JustCham/

I've spent three days trying to fix the UV shit in my model importer and I'm about to kill myself.

Anyone here want to be Indian labor and help me out? I can offer $10 ;-;

I don`t think I`ve ever tried lewds yet.

I`ll send you a PM on steam

I added an option to the cameras to let them follow the player.
I also pretty much completely reworked the camera switching system, but it's so much more flexible now. The level designer can now do things like put volume triggers inside volume triggers inside volume triggers or have multiple volume triggers assigned to one camera. The previous system didn't allow for that.

That looks pretty good man,Perfect for a horror game.

9/10

Resident Evil Styled game, please.

alright guys, time for the hands, wish me luck

good work!

Man the strength of a following camera is so understated. It's easier to effectively implement than any other camera system, if you design the level right and place the camera correctly it's fucking golden.
Especially when you don't have pre-rendered backgrounds because then instead of having it snap to another perspective you can have it smoothly transition to another perspective. And do all sorts of fancy shit involving the cameras movement if you see fit.
Shame AAA has it's head too far up it's ass to see those strengths and instead "camera issues" has got to be the most common phrase I hear in game reviews/discussion.

Thanks.

That's kinda what we're planning on. We'll see where it goes.

Well, next on the list is camera-on-a-rail for those smooth movements.
Or do you mean like the transition in webm related? That was just changing two numbers and a drop down.

Also ignore the weird filter going on. I was trying to figure out how to make it look VHS-ish to give it that 80's vibe but I gave up and decided I would come back to it later when I'm willing to learn about post processing in more depth.

So did anyone actually trie this?
I understand it would take a while to download and it's probably pretty buggy
I can't really think of a better way to distribute the game though.

You should probably post a few screens, at least.

I'll post an mp4

man that flikkering is real fucked up obs needs to get it's shit together.

it's not obs
your gpu is literally dying

only the recording looks like that though
The game looked fine while i was recording this

this was fairly frustrating, the base for the sculpt got really fucked up at one point

You should attempt to do some scripting mods yourself, make a doom mod, or write a very small C & SDL game prototype.

When you or someone on your team goes full feature-creep in every conversation you will know enough to shut them down and not all dogpile the programmer as lazy and a negative-nancy for shitting all over your ideas.

seriously though, "idea guys" trying to lead projects should just kill themselves. Get at least enough skill and experience to understand the game making process or GTFO. Go flip burgers.

If you wanna go for PS1 era, remember that models, while having a skeleton, do not transform besides rotation; and in the end, they are made of distinct separate parts that touch each other at the joints.

The trick is very visible on Meryl (notice her shoulders, her elbows and head are thesame) and older fightan games.

I will never understand this, I find sculpting with flat shading makes the object easier to read.

Your shit's all fucked up and you haven't even finished the clothes yet, if you want to avoid what I guess is the bug that makes you move faces on the far side when sculpting one side you just zoom in a whole lot.

There are also a bunch of games from that same era that didn't do this, it's entirely optional.

What you working on user?

I wonder how those hands would feel.

Seperated body parts always looked pretty bad though. Quake's models still look fantastic and I think SH1 didn't do it either.

Actually Somewhat incorrect. The Reason why some games back then had segmented body parts is due to the fact that with the lower poly count, they couldn't get them position properly (This was also a thing in the dreamcast era).

In terms of fighting games, this mostly affected grappler characters; in the Rival Schools sequal, the models were smoother, albeit for the grapplers who needed segmented parts.

What are you having problems with? I can probably help.

Matrix transforms are applied in whatever order you decide to apply them in. Modern OpenGL doesn't have a matrix stack.

If your issue is objects are not rotating about the point you think they should be: that's an issue with your understanding of linear algebra and not an issue with OpenGL. You should do the math by hand a few times to gain a deeper understanding before blaming your tools.

Nostalgia? I mean I love quake but the only thing cool about the enemies is their innate design, the models are absolute shit, Thief tier too few gons models.

I mostly sculpt with flat, I've just been trying to experiment a bit lately. One thing I'm wondering about though is - what should I use for the final bake. I mean either option looks ok, but i'd like some happy compromise where areas with detail have much sharper jumps in shade, whereas large planes of nothing are smooth.
Some ideas on how to get that with blender would be very appreciated.

There's a boner joke to be made somewhere here.

Enemy model for an fps game, right now I'm sculpting all the details and later I'll bake them to a low poly model.

With enemy designs like that, I am getting Time Splitter Vibes

I didn't play Quake until very recently, actually. But I preferred playing it with the classic graphics settings.
Devil Daggers is an example that emulated the style very well. To me the appeal is the same as pixel art, I guess.

Baking with smooth is the way to go, it should look as sharp as you need it to. if it doesn't you could try creasing in non dynotopo or baking twice and splicing the bakes together, but straight smooth should look best.

I played Quake for the first time last year and I think they are ugly, now the texture work is really nice but the models don't take advantage of the excellent texture work, in many cases the UV maps are fucking retarded and overlap, the Quad Damage texture was drawn by hand in perspeective from 2 different angles, although I don't have that texture on me, the Shambler texture shown is how all enemies were done, look how inefficient it is, I mean to an extent it's understandable that they drew the textures like this thanks to the really limited visualization of 3d software in 1995/6.

Devil Daggers gets it right though, with just enough gons for a nice silhouette and low res textures to give it that pixel art style.

That's not the problem. Using triangles and/or dyntopo while sculpting seems like a great thing at first, it may give more freedom to form the sculpt, but it will sometimes bite you in the ass later when you bake on your lowpoly. If your sculpt is not clean, there can and often will be artifacts that seem near impossible to get rid off, especially on organic stuff.

Here is some pic related, i hope you can see what i mean.

Protip: Use shrinkwrap and other modifiers on your Basemesh to adjust your lowpoly to your highpoly. Also, creating or downloading simple starter basemeshes is even easier and faster than retopoing again and again, from my experience. For example make a low-midpoly game character with good topology for animation, and always use that one as a basemesh to start human sculpting.

On-rails or the camera smoothly transitioning from one fixed location to another. If you do it right, fixed cameras is one of if not THE best camera system

I am using Blender, what is a good file format to export to for my code to render?

obj is a nice and simple format that covers the most basic needs. If you have animations and rigs and stuff, fbx is the way to go.

sauerbraten.org/iqm/

I can code in C++ but have no experience with DirectX. Is it worth learning?

Do you already know Vulcan OpenGL? If not it may be better to learn how to use something that is available on all systems then learn something that is Windows specific.
But in general yeah it's worth knowing.

I plan on staying on windows for now. It seems easier at the moment to just use visual studio.

a.pomf.cat/wzuqhc.webm

New weapon for space piracy, the BioLogical Entity Neutralizer & DNA Randomizer, also called the B.L.E.N.D.R.
It's a pulp sci-fi raygun you'll pick up from Planet X. Functionally, it's a pretty typical railgun–accurate, fairly slow rate of fire, hits hard. But it also has a unique death sequence for enemies killed by it. Or, well. Not "killed".

Placeholder graphics, etc, etc, but I'm still REALLY not happy with the graphical effects.
The rings don't seem to fit, the hit/morph effects don't seem to have any real impact, and the particles flying everywhere just seem random rather than actual energy dots off a raygun.
Took some inspiration from pic related and TF2's Righteous Bison, but it looks like I'm gonna need to do a lot more.

Some reference material that may be helpful.

What's the human limit of 1MA?
What's the most a single developer could make?

That's a question that doesn't have much of an answer, considering we've seen many incredibly impressive things that were made by either one person or were 95% one person.

I suppose the only real answer could be "the limit of one-man devving is how much you're willing to put into devving".

yeah, but what's the most impressive thing a 1MA has ever archieved?

I now have a cameras on rails.
Basically there are four points, a "rail camera", and a trigger volume. Two points keep track of where the player is between them and two points are references for the camera to move between. The position of the camera between it's two points depends on where the player is between their two points while inside the trigger volume. You can also make it move from point A to point B in a round about way, or have the camera rotate in a certain way by modifying some curves, just to give it that extra level of flexibility.

I also set up source control but that isn't as cool looking.

Cooking system:
Pan recipes:

pancakes (not done)
dumplings (not done)
[main ingredient] jam (next, need to think the formula for perfect fruit:sugar ratio based on fruit sweetness)
battered [main ingredient] (not done)
deepfried [main ingredient] (not done)
[main ingredient] toast (not done)

Then Oven, Knife (raw) and Fermenter recipes

WHY IS THIS HAPPENING

is this actually as shitty as it looks?

Anybody here use game maker studio and /or watch uheartbeast videos?

Cave Story
the only things he DIDN'T do was beta testing and ports to other consoles (which he only supervised)
pixel is one of my all time favorite game devs, up there with carmack

...

slowly but surely things start to come along. Today a menu system, tomorrow a fully fledged release title i wish at least

Quite a few threads ago there was an user posting progress on his monogame project, was a 2D topdown diablo/zelda mix(by his admission)
I had a couple questions, but he stopped posting by the time I had them.

it all depends on the one man, really

Things are going smoothly. There are some animations I still need to add, and the animation system itself needs an overhaul. I'm also not sure if I'll add in the Chao talking to the player. Otherwise, saving and loading is good, fruit's good, stats are good, eggs are good. Need to move on to minigames soon, which might need some thought.

The only game I actually felt was engaging across the different versions of TCG was Chao Ball, in Sonic Pinball Party. The others are kind of blah. In addition, the memory-matching game in every version is easily cheated. I feel I should have at least 2 different games, though… On that note, a game that's actually affected by your Chao's skill levels would be nice, but I'd rather wait until I have the 'basic' version that I have here finished.

Hey agdg, I got a new version of my game, so I wanted to make a post about it.

My carpal tunnel is getting better, so I was able to get some work done in the past few days even though I couldnt work for most of the week.

Here is the changelog, its not much, but ive got 7 levels in the game now! I just want to make 10-12 levels, I suppose, and that's really all that I need to be satisfied with it as a completed game, of course I still have to do all sorts of polishing and adding better music and sound effects and doing passes on the levels to add new things or fine-tune it, like adding new props and making real level transitions and fixing bugs. So I am getting close to my goal so far.

changelog:

New "ELITE_GAURD" enemy added- can be found in the "railroad bridge" level (incomplete implementation)
Finished "tower" level
Changed music for "radio tower" level

If anyone has the time to try it out, this is the download link:

tell me what you think of it , how I could make it better and all that.

My website for this project is ika.neocities.org/

This game was amazing as a kid because I always expected it to go somewhere. It never did.

Elite 2, I think.

What did you use to learn directx 11?

What are your opinions on unreal engine 4?

Call me a shill or anything, but I spent the last couple of days looking tutorial after a tutorial of this shit, it really looks good and I'm hyped Unhappily my pc is a Frankenstein, it has a i7 but the graphics card is below toaster level, so the engine literally doesn't work above 10 fps
Anyway, here's some advantages:

The lack of love this gets is crippling suspicious, what I'm not seeing?

Dwarf Fortress?

it should have a villainess to romance

Unreal 4 gets a lot of love, though right now rarely a real game looks like the tech demos

The most impressive usage of UE4 right now is Guilty Gear Xrd series

It's open source but not GNU free software. If you want to play around with the engine source code to improve efficiency for your game they let you do it.
The biggest drawback is the royalty fee, 5% for every dollar past your initial $3000.

And then there's Cryengine. It's an awesome engine that is completely free with no royalty fee, the source code is open, but with worse documentation compared to Unreal and Unity nobody in this thread uses it. Also the fact that it's strictly C++ may have something to do with it.

Only if you're a fag that isn't gonna make a real game. They're good for placeholders or if your game is generic enough to use that shit as final assets but for something specific like I'm making, you still need artists.

Yeah it's open source. You can make your own forks of the engine but if you do - good luck updating it to new versions. Without forking it you can still see all of the engines source code to understand what's going on under the hood, which is a huge advantage over unity.

A lot of the steams and tutorials are garbage. They'll go into only BP and rarely into engine internals that'd actually be more beneficial to learn. Going further from that, most of the documentation is done around blueprint for idiots with very little information about internals or the best way to do things. This means that it's harder for an inexperienced programmer to get his bearings but it's no problem for more advanced people.

Overall the engine is very good. Quite a few people here are using it, including myself. I just don't show off most of what I've got because it's still WIP, I don't like to show off what I've got until I'm very happy with it and I don't get a huge amount of time to work on it.

i'm having a hard time picking between game maker or clickteam fusion i got from the bundles. they aren't as powerful as writing the game completely by myself, but making my own engine or grabbing shit like unity would be an overkill

Really liking your camera stuff.

I never used fusion, but Gamemaker has acceptable documentation and a criminal quantity of tutorials on YouTube. It's also free if you only export to PA I think. Really don't know the details about the EULA.

Thanks man.

Fix'd. Autocorrect literally drunk.

Watching Data-Oriented Design videos, and it turns out Ubisoft uses global singleton hybrids, which basically solves the issue I've ran into earlier. Pretty much all of the videos I've seen so far have some butthurt faggot in the comments going "THIS IS TERRIBLE BECAUSE IT VIOLATES WHAT I'VE BEEN SPOONFED"

What are you looking to make? [Obligatory Löve shilling goes here.]

a strategy game with units moving around the level divided by small squares

It's funny about that isn't it? Things are generally only terrible if you misuse them. If you made it, you'd bloody hope you weren't going to misuse it. Or if it is just stupid and either doesn't work or causes a big inefficient mess.

People are masturbating furiously over the blueprint system tho, but I think you're right because being able to live edit code in near-real-time is hot as fuck and at least as good as BP.

Who would have thought spamming objects for everything, even major things only used once, was a stupid idea?

This is the most beautiful primitive I have ever seen in Blender.

5% Royalties though.

Shaders are basically AAA only, which is too bloated for indie games with simjple graphics. If you want AAA graphics though, creating high quality assets takes time, as an Indie or 1MA you would have to be a giant autist to be able to create an entire cluster of AAA models for a complete game by yourself.

Literally spaghetti.

True.

True.

They have good basic stuff, but not far enough for an entire game. Even less if you want to have assets that don't bite each other and make your game look like Dr. Frankensteins Monster.

Yes, but not Free. Royalties need to be paid if you use their code.

Tutorials are 95% for Design and Spaghetti Script. Basically nothing on how to use the API and respective C++. Being a Beginner or Mediocre in C++ is crippling if you only have the Doc to rely on.

While Unity community is full of beginners, Unreal community is full of skilled pros. Sadly though, many of them are insufferable faggots, who will literally swarm you for criticizing anything or not being part of the hugbox.

Mostly cleanup.

I'll make sure to experiment with multi res modifier after this project. What sort of artifacts do you mean, though? Things like ripples in the normal map?

Well to be fair the blueprint system is really bloody good. What isn't is how many people use it as a substitute to actually fucking learning how to program. Instead they're more content to hack shit together that 'werks' but performs horribly or won't work when they try to expand on it. Their tutorial streams often are full of the unreal devs bumbling around like idiots, making simple mistakes they shouldn't, and trying to fix their errors instead of a nicely planned out tutorial.

Guys, help me with a resolution problem. Since it has pixel art it needs to scale well for the most common resolutions. I don't want to go as low as 640x360 or as high as 1280x720. So far I thought about 768x432 since it would fill the entire screen at 1920x1080 (1080p), so this is what I want, to fill the entire screen and to not be enormous, but to be big enough so I don't draw pixel limbs.

Is 768x432 a good one? It would scale well to 1080p but not for 720p, but apparently 1080p is the most common resolution now. But then again since it's a pixel game I expect people with older PCs to play it so they would probably have a 720p monitor. Fuck, what do you guys use and why?

Wouldn't it be better to go with 540p? It's exactly a quarter of full hd, and if you wanna go even lower, you could do 270p, like the PSP's screen.
You really don't need that much res for pixel art, consoles used to be 240p way up to the PS1 era. (I mean, the video may have been interlaced and maybe even in a different resolution depending on the region, but the actual rendering was 240 in most cases.)

Do you mean 960x540? That's what my current game resolution is and a bunch of people complained that I had a weird aspect ratio.

Rail camera, huh?
So how long until you say fuck it and make an action game?

640x360 isn't low, you'll still struggle making pixel art for that.
960x540 is the next stop for integer 16:9, it's half of 1080

I forgot to respond to this question. The reason for why I don't wanna go as low as 240p is because making good looking pixel art at lower resolutions requires a ton of skill, knowing how exactly each pixel will affect the art. If I go higher I can be a bit less strict about that since I am a beginner in drawing. But not very high since that would require a lot of time

You couldn't be more wrong.
Individual pixels will still be as important, but now you'll have x4 of them to keep track of.

How so? it's 16:9
Also, you don't really need to make the game one resolution, you could cut off/add more "field of view" at the sides. I know it's not called fov in 2d, but I'm not sure what would be a better term.

Then it's not really pixel art, the point of pixel art is to use every pixel meaningfully. And if you're still caring about each pixel, then more resolution will just be more work.

I tried drawing very low resolution sprites and I either didn't have enough space for the details I wanted or it looked like crap because every pixel counted. Just deciding how I should place the eyes on a character was hard.

I know, it was sort of in-between. Drawing HD animations is very time consuming, drawing good looking pixel art is something only good artists can do. I tried to choose something between that, not drawing in HD but not very pixelated either.

I suppose 640x360 is not that bad for a pixel art game then? It should perfectly fill the screen if you got a 1080p monitor I think.

I don't think just half assing it would help then, I think it'd be better if you focused on some simpler cartoony style while you develop your drawing skills. I recommend Andrew Loomis' "Figure drawing for all it's worth" if you wanna learn quickly.

But that's the thing, high definition animation isn't as hard as high definition pixelart, because you're not focusing on every pixel, just the lines and shapes.

Alternatively, make some simple stylized 3d models and render them as sprites with some cartoon shaders. It won't look amazing but it shouldn't be that labour intensive, since you dont redraw every frame.

I'm not trying to half ass it, I just am self aware that I need some years of experience to draw decent very low resolution pixel art. As for high definition animation, I always thought that is harder to do. I mean I could do HD animation, but don't you think pixel art looks way better in games? And wouldn't you see all the imperfections of a beginner artist in HD too well?

That actually is a nice idea, I've seen it in action and being used in some games. But I feel that if I'd do that I might as well make a 3D game.

Yes, it will be scaled up exactly 3x.

Another thing I forgot to mention; with a higher resolution you're gonna have problems regarding the scale of things. Most 2D games made between 1985 and 1995 were built around a common '16x16 blocks on a 240p screen' grid, and by coincidence it worked pretty well. If you increase the resolution but don't work around the tile size problem, you'll fall into the same trap many indies have; slow moving tiny objects on a big ass screen. (or worse, fast moving tiny objects)

Might as well just render them in realtime then, unless you're not targeting PC. A prerendered spritesheet is hundreds of times bigger in memory than the model and texture used to generate it. I think the last thread had some good points about that.

I might as well go with 640x360 then, it would have the best support for people who use 720p and 1080p resolutions. Gotta prepare for it very well though, every pixel counts. I don't know what Castlevania SOTN used, 320x240? But the people who have drawn that aren't human, so detailed and such a tiny resolution. Just… damn.

Imperfections will be visible regardless of medium. To be honest I'd say you should go with whatever you'd like the end result to be, it's likely gonna be meh, but the only way to actually learn is to keep fucking up. Plus, I'd imagine it'd still look better than medium resolution pixel art.

I don't think it would be that big of a memory hog if it was low res sprites with a fairly low framerate, but i guess it wouldn't hurt to just render it real time unless you're going for something specific.

I'm rolling with 640x360 personally. Are you intending for people to actually play at 1x? You should take that into consideration, especially considering things like fonts and such.

I am assuming most people will play the game scaled at 3x since 1080p is the most common resolution, or 2x for people with cheaper monitors. But I doubt many will still run 720p in 2 years, so 3x it is. And there is a lot of text in the game, another reason why a lower resolution would not work well.

You could alternatively letterbox. I think someone else suggested FOV changes, but letterboxing lets you keep the same AR regardless of resolution, as long as you're willing to deal with black bars.

I'm not, I like seeing the game fill the screen.

You'll have to letterbox at some point anyway, for instance people with a 16:10 monitor.

That's true, but apparently most people now have a 16:9 monitor, so it makes sense to develop for that ratio.

What do you mean, exactly?

...

1. Never give up.
2. Try reading Loomis.

Thank you. Downloading his books now.

I did everything there was to do in that shitty little minigame, expecting some sort of breakthrough. Even hatching an egg would be nice. Nope, egg just sits there mocking me.

might as well just pay for/steal code

Ah. Yeah, not much you can do if you don't have SADX/SA2B and the GBA connector.

You can hatch that egg if you neglect your Chao and wait for it to run away.

Anyone here who has a brain that isn't immune to mathematics? I'm trying to use a lerp (linear interpolation) function to make a sliding scale. Description of the function is "With this function you can find the value that equates to the position between two other values for a given percentage. lerp(a, b, amt)"

I count all of this on GUI pixel location, so the a and b values are 904 and 1045, so when the slider's y position is at ~974, the lerp function should spit out ~0.5 or ~50%, but when I plug (904, 1045, y_position) into the function, it spits out something like 99500.

Am I reading this wrong? Is this function not supposed to tell you what percentage between 'a' and 'b' 'amt' is?

you give it the percentage and it returns the position of the value
It doesn't return a percentage it returns a value and you give it a percentage

Oh, holy shit. So my shit's all completely backwards and inside-out.
Thanks, user, I'll work this out another way.

please make the camera look down when you approach an edge

You're telling it to lerp to 9740%.

If all you want is the midpoint between two values, you can just compute the average, i.e. (a+b)/2. This even works with integer division.

this what?

this what?

Thanks for setting me straight, anons. It's working now.
I had to round it and invert it because the engine I'm working with uses a scale that's positive as it goes down for some reason, but it's working as expected now.
I did: floor(abs(100 - (100*(y_position-a) / (b-a))))

IDs

You weren't the first to point that like a disadvantage, as a citizen of an almost commie country I find that more than fair.

That's actually true and a painful truth I didn't consider, I was going to use Little Nightmares as a counter example but I just realized the team has 40 people, that's not indie shit.

What does this mean? I'm not baiting, I really don't know

True. But I believe it's possible to use them in a minimalistic way

Yes, that's actually strange, I noticed an certain disdain to coders in the start of the blueprint tutorial, he somewhat demonizes code as something too complicated and used only by nerds and "we" as artists would like to "visualize our creations more easily"

I commented this with a friend of mine he said that they could be discouraging programmers because they are control freaks and they would complain more about details than the average "blueprint user", leaving code monopoly to the devs and, as you said, veterans aka the perfect hugbox.

hmm is there a way to optimized it a bit? Roughly speaking since I don't program at all (3D modelling)

That's quite idiotic from them since the only way Blueprint was possible is by programming them in the first place.

Who will improve the Blueprint function if there is no more programers do to them?

Dumb fuck niggers.

Can someone give me a good resource to begin learning level design in UE4? I have NO IDEA what I'm doing.

pls rate.

I was just pointing out that royalties are unequal to free. Royalties being bad or not is for yourself to decide of course. Though, i am more of a hardprice type of guy, because in my opinion royalties or interest or any kind of percentage based trade is mostly leeching on the work of others.

Pic related.

Well, they are trying to market their spaghetti, and in order to sell it they try to appeal it to a certain audience, which are mostly artfags. A lot of artfags get a boner on hating coders, probably because most of them ar too stupid to code or think it is obsolete, and being able to create logic/gameplay without the need of coding through visual scripting is a wet dream for them. Same way that many coders get a boner from creating art through code, thinking that Artists will become unnecessary one day because of technology. Artfags and Coders have a kind of love/hate relationship.

not anonymous enough :^)

Looks like an asian man.

mah spaghetti nigga

Practice makes perfect user.

From the GML:
template GLM_FUNC_QUALIFIER tquat lerp(tquat const& x, tquat const& y, T a) { // Lerp is only defined in [0, 1] assert(a >= static_cast(0)); assert(a

...

Does this shit still freeze and recompile every time you add anything?

Not for me. I'm running a pirated version though so i'm not sure if it's a newer thing.

(dam quads)
read , specifically the third part.

Topkek.

Amazing analysis, you have some sharp senses.

Check this playlist:

youtube.com/watch?v=cl_eoVfNDKU&list=PLZlv_N0_O1gak1_FoAJVrEGiLIploeF3F

mhm yes indeed it does sound very plausible. I can only hope that no serious "clash" is going to occur because of that.

I've made good progress on the system to inspect items. I still need to refine some parts of it but at least I've got a good base that I understand and can work with.

Aww yeah finished implementing the Jam algorithm
Triggered when using cooking in the pan when fruits and sugar are at least 90% of the ingredients
First It brings the ideal cooking temperature of all ingredients close to 105ºC (ideal temp for making jam)
As usual for all cooked recipes it cooks every ingredient over time (time and temperature are input by the player), checking if anything is burning or not cooking properly due to low temperature
Then it calculates the ideal fruit:sugar ratio, that variates between 2:1 to 1:1 depending on how sweet is the fruit (or fruits) used
If the player gets the ratio right he can get a slightly quality increase, but a big divergence drops the quality greatly
At least the game compares acid and sweetness, acid juices added to the mix brings sweetness down from the high unpleasant levels to a more acceptable level. (all tastes go from 0 to 1, anything over 0.66 is unpleasant)

Took longer than expected

ded tred

rate

What are wrong with her legs?

I don't know if you are aware, amigo, but it looks like your chica is a negra.

I'd correct that, and fast, if I were you.

...

c'mon muchacho

but I like niggers, more than I hate whites.
:/

...

Anyone have a good source for comfy music? Similar to the types you'd hear in a VN. Just soft, simple, easy listening background noise, that sort of thing.

And I like other kind of animals too user, that doesn't make them people.

Have you checked Newgrounds audio portal? There's a lot of comfy relaxing music there, but if you are planning on selling your game you have to contact the author and all that jazz.

newgrounds.com/audio/

Thanks m8.

anyway, going to sleep.
bye pol.

New thread