/agdg/ + /vm/ ~ Amateur Gamedev General

STOP

Well duh, look at their HP during the expedition. They're at negative millions of health.

muh_dick/cyber_10

Anyone else in the My First Game Jam jam?

I've been prepping myself for ages and never actually getting shit done so I decided now or never.

Spooky

enjoy your digiaids

Thank you, I will check that Game Jam out. It seems to have only started two days ago. I hope it goes well for you, what you have looks good so far!

Why does this feel wrong?

the only flaw is that you end up doubling your triangle count for any meshes you use, but if you're working in low/mediumpoly it's totally fine.

my mesh is 10k triangles so I figure i could cut down my duped mesh t o be about 6-7k tris on its own, and 16-17k triangles isnt too bad, but if you have a mesh that's something wild like 30k triangles then duping the mesh is probably a bad idea and you'd be better off with shaders.

anyone done crusader kings 2 mods?
i'm having trouble getting it to not crash on startup

Maybe because it took a game jam for me to get off my ass instead of my own ambition for game dev?

Thanks, man!

Maybe you should find out the error log of that game and post it here via any pastebin site.

i can't actually figure out which log i'd look into
some are empty, others have stuff in them, but the game is currently crashing while processing the flags, yet none of the logs mention the flags

mhm that is bad a log that doesn't mention what the causes of the error is hard to debug, are those flags like custom graphics or part of some "value" in a file? If they are custom graphics maybe its not in the same resolution as the other flags or it needs some additional embedded info on them.

I am just throwing some guesses here since I never modded that game myself.

he can always use a matcap, but difficulty varies from engine to engine (not sure if that is their standard name just that blender calls them that)

Performance-wise creating an outline from mesh is a really bad idea not just because of the doubling of the tri count, but because even if you set the material in blender to black, your engine will still uselessly calculate light for your outline as well, doubling the rendering cost of your already heavier mesh.

Make a lit shader in the engine you are using, in which you add a second pass with simple color rendering and culling set to front (and specify other parameters such as ZWriting to off in order to make your pass as lightweight as possible).
That way, the lighting will not be calculated on the outline and you will have much more control over it, such as it's color and thickness, even in realtime, and can add some interesting shit like border smoothing (because outlines will always have jaggy angles due to your topo) and silhouette drawing.

...

Re-read the original post:
Chances are he will be rendering out to sprites in the end.

Even in blender shaders can speed up his workflow pretty hard (same thing, customizable color and thickness, plus the fact that he won't have to duplicate, invert, scale etc each mesh). If he is making a lot of meshes this is the best way to go

It only works for smooth meshes. If the model has any split planar surfaces they'll end up sampling a single color across the whole plane.

I haven't seen that monkey for a decade.

just build your own engine lol

they should have lost the battle then
everything is fucked

Same guy here, different IP.
But do shaders work in the editor view? He specifically mentioned he doesn't want to render to see the outlines.

is this a good base model to practice sculpting, if not what can i do to make it better?

God damn it I didn't asked for a pandemic effect. Well anyway I think I fixed it now it should spawn some nigger shacks but not too much of those either, good to know that influence was the damn culprit of it all the time jesus christ did it took long it din't helped me either that Factorio takes a sweet time to load all those fucking sprites everytime I want to change some values.

Now I just need to fix these things:
-Acid ammunition, it still needs at least recolor for ayys that are burning which is not green and somehow a way so that it doesn't cause any forest fire at all. Custom graphics for acid effects will be done later I don't really feel doing it now.
-Infested horse stable which is going to use the same spawn parameters as I did with the nigger shacks and fixed graphic animation so that it doesn't spin like a madman at all. Then some minor changes here and there then I can release another version of my mod.

Sounds a tad complicated but I'll have to give it a go tonight as I wasn't even thinking about the weight it would put on performance.

This is still the big thing though. I need to be able to actively see them considering I'll be doing some traditional animation with the 3D models and I'll need to see how the lines look with the ones I'm drawing.

That minimap tho

What I've been doing that feels a lot more natural is using this guys approach to making a model
m.youtube.com/watch?list=PLvPwLecDlWRC92lV-KHLmKkoXco5tPSYD&v=pOzC_Do1w4g

The nice thing is that if you do it right you'll have different "layers" to work on considering each anitomical chunk is its own layer. I'd load up a front and side view image of a person or art with anatomy you like, sculpt a loose mesh off it, and then disable the image and do some study sculpting.

is its own object.*

the thing is i really like to mimic kazuma kaneko style which how i see it a mix of neo-romanticism and (i think) 17th century Japanese coloring styles.

you can set the shader of the outline material to ignore lighting checks though leaving it perfectly black

gonna bake my shadows and then add the texture

I always see that dumbass monkey head whenever I look up a tutorial
I don't think blender is bad

The truth is you are not even an ideaguy. I bet that instead of coming up with gameplay systems that require people to cooperate in a specific, fun and self-perpetuating way you imagine a world without boundaries that is so immersive it's a one big larping session where everything is done by the players, users imitate real-life for no particular reason with shit like people playing the role of quest givers or bandits and you live out your dream of having a connection to a breathing fantasy world. I know I used to do that when I still liked MMOs

If there isn't a team of developers hiding in your basement at your disposal, making an MMO might not work out so well in the first place.

I can gamedev though so I have some context of how to make things and how much work it would take, I think my idea is more reasonable than the typical new MMO idea you see around here. It's actually lower in scope than most MMOs, not that I'm likely able to ever do it anyway. My development "plan" is to make something very simple and then slowly add my planned content over time.

Runescape was made by 2 people from their parent's house.

Would recommend to new users having issues with Blender's curve. Not sure if I will pick it up, I'm already cozy with Blender's crap.

I'd recommend just getting used to the vanilla or a slight user modification of it, just so you get the benefit of a closer codebase and newest updates, as well as all Blender-specific plugins and resources guaranteed to work.

They are up to date right now, 2.78c. But this could change in the future of course, and yeah, some plugins may have issues.

Like I said, I'm very much used to Blender already, but it can be of use to a new user to at least get to know what stuff is hidden in there.

I'm the dumbass trying to get a GM clone of Bomberman started.

As suggested by another kind user in a previous thread, I restarted the project on 1.4 using mostly GML code, and I've been doing sort of fine. Movement is a bit stiff but I suppose that's good considering the grid-based nature of the game. I have it set up so Bomberman can move in 32px "steps", and even figured out a way to code bombs, as well as make them explode with alarms. It's all pretty cool stuff, but I'm wondering how I can make it so that the explosion range could be longer, much like how in the original games you start with a weak bomb that gets better as you pick up power-ups.

The way I'm making this work is by creating 5 different instances of my "obj_fire" object on the bomb's Destroy event, but I'm thinking this won't work well if I'm going to be expanding the bomb's range. Has anyone done anything similar in the past? Is there a simpler way to achieve this that I'm missing? Any help would be appreciated.

Looks quite neat. Although they say that "the usage is UI centered", which is a double-edged sword. Blender is hotkey centered, which makes for a steep learning curve, but once you got the most important hotkeys memorized, it's way faster than button clicking and context menus could ever hope to be.

Not quite sure what you're asking, but try making an array for each direction, and then use for events to create explosions further away from the center.

Use a for loop to create fire instances in each direction.
I don't know the syntax for the game maker language, but the logic would be something like this:

int range = 3; (that would be power of the explosion, in number of tiles reached in each direction)

for(int i = 1; i

Make it a stat that can increase and decrease (including procedures that increase or decrease the stat up to a max/minimum) you can use those later for powerups.
Bombs inherit/use the stat for their Loop with secondary stop condition when resistance/grey block is reached.

Easiest way is a stat array for each player.
That way you only have to use bomb_strength[PlayerNumber] and find a way to check who is laying the bomb.

Does /agdg/ allow fan projects
and also asking for a team to work for fun as well as learning about devving (but mostly fun)?

I'm a modeler but still an amateur.

I'm mostly looking for programmer and 3d rigger/animator.

I'm thinking about starting a project, for fun.
I wanted to do a 'competitive' 2D multiplayer versus puzzle game. Something like Puyo Puyo, but not as hardcore, and with some other additional mechanics on top of it.
I can do the programming by myself, no problem, I have finished quite a few games before, but I suck at making game assets, and at keeping myself motivated when working by myself.
Anyone willing to make some assets and help with the creative part wanna get on board and help me out on this project?

I think you can answer that yourself.
It's allowed, but usually you don't get any interested people or it doesn't work out for very long. Most /agdg/ projects are one man shows or small teams of people who already knew each other.

Probably a very controversial suggestion, but I think the game called "Bob's Game" that is currently out allows you to add and edit puzzle game modes. No idea how it works, but you could give it a look. If it works the way I think it would, it is good for prototyping puzzle games and even testing multiplayer, with premade passable assets.

would anyone here be willing to collab for a nes remake?

Pic related. Most Game Maker tutorials didn't have this problem. There seems to be an assumption that I'm already very experienced with C++. Also, is it even worth learning how to do blueprints first? I was under the impression that they were the same kind of trouble as Game Maker's Dragon Drop features.

The code that you posted is almost perfect in Game Maker. However, there's no need to use int, and range should be a variable on the player that's passed to the bomb upon its creation, not something that immediately precedes that for loop. The for loop should only run when the bomb's alarm goes off.

Yes, you're supposed to know C++ before getting into UE4's C++.
Common misconception. Blueprints aren't actually (much) slower anymore nowadays as they get compiled, and for many things thy can be faster to get working than typing C++ and compiling. Ideally you use a combination of the two, making basic functionality in C++ and stringing it together in actors using custom blueprints. Just use blueprints while you learn C++ for now.

Is that Agumon?

True, but only because Runescape had shit graphics and fairly limited content. If that's the kind of MMO you want, go for it, but keep in mind Runescape was made in a time where there were barely MMOs around and graphics generally shit. If you would release the same thing today, don't plan on any success.

how stupid is it to post on an unreal engine forum to ask for recruits to make the fan game. I'm assuming its legal because the game that'll come out won't be for sale/free.
I don't care about making money I want my dreams to be real.
sorry for the autism

I would say you shouldn't ask for help before you have something you can show. If you don't have a portfolio, start by working alone and then show people what you already did.
If you want others to give you the time of day, you need to sell yourself to them. And you can't do that on the basis of hopes, dreams, ideas and concepts.

I mean, you're already asking for a lot when you make a non-profit fan project. Might as well make a good first impression.

...

I posted in the /agdg/ board with this question, but decided to post here since it seems more active:

So I'm learning Python 2.7 for Roguebasin's Roguelike tutorial with Libtcod so I can try my hand on some glorious ASCII carnage.

I've also heard about Pygame, but I wanted to know your guys' opinions on it before diving too deeply into it

Link to the Python Roguelike tutorial if curious.
roguebasin.com/index.php?title=Complete_Roguelike_Tutorial,_using_python+libtcod

I just wanna slice things up like the original mgrr tech demo.

Why has god forsaken realtime dissection?

I give up on this.

Popeye wave-dashing around some weird structure?

dunno, going to make the sprites, but I've given up on coding it.

...

I'll never stop posting, despite you never noticing my posts.

When you guys say RPG, do you mean a game which has a "Story + a level up system + possibly turn based but doesn't have to be"?

Is this supposed to be MDK?

RPG as in something similar to Gothic, FF etc.

Mixamo fuse is fun. I pirated it a while back and I've been using it to shove out models for Kowloon.
I made Teagan from Uncommon Time in it. I keep deliberating as to whether or not I should add Holla Forums in-joke characters into the game.

It looks like skyrim… So you want to make skryrim?

No, I just wrote gothic as an example for a typical RPG. I don't want to make "le epic openword experience" but rather a nice third person 3D fantasy RPG.

I spent little time on progress because the last two weeks I was implementing 3D collision algorithms for work in Javascript. At least now it works as well as you could expect a collision algorithm to work, and I've learned more about the topic than I'd ever have wanted.

Do you believe Unreal Engine is too much for a noobie? I'm not really planning on creating and releasing a game that would sell, but rather learning how to use an engine well.

Why not try the engine first?

Easter eggs are cute. Memes for the sake of memes are unbearable and prematurely age your game. Tread carefully.

Who LDJAM 39 here?

not me but I look forward to your postings

I have a problem with GameMaker 1.4 and gamepads. Googling this is taking me in circles. I have a standard 360 wireless controller. I simply want to move with the D-pad. Control stick works fine. D-pad, though, I don't know how to get it to work. "gamepad_button_check" is the function, correct? When I put "gamepad_button_check(0, gp_padr)" in place of "keyboard_check(vk_right)" it does not work. But, using the "gamepad_axis_value" function does work to control with the stick.

To me at least having the option for a d-pad is good for a simple pixelly game. I can't stand playing an old platformer with a control stick.

Yeah I'll be there.
We'll see if me and my friend will join it as a team or solo.
While the latter option is tempting, I may not survive it. I think I might be too old for 48 hrs without sleep.

Unreal certainly has a steeper learning curve than say Unity. That's mostly due to there being more to learn. Unity is kind of a blank slate engine. You create a project, you've got nothing but the engine. Meanwhile Unreal has a "GamePlay Framework" that consists of quite a few classes that already do a lot that pretty much any game needs to do anyways (like handling player spawns for example). The problem with that is that you need to learn what all those classes do before you can use them.
Unreal's UI system (UMG) is more complex than Unity's UGUI. That's something I found out the hard way when porting my game's inventory from Unity to Unreal.

You can certainly use Unreal as a noobie, but expect it to be an uphill battle; Not that there's anything wrong with that. Just expect it to take some time before it clicks. Also start with blueprints rather than diving into C++ and don't sabotage yourself by starting with a complex game. Thankfully the Unreal tutorials are quite good.

I'm currently teaching a friend how to Unreal and these were the first three tutorials I expected him to do (in this order). He's studying CS with me and wants to collaborate. He knows how to program, but doesn't know game development.

>docs.unrealengine.com/latest/INT/Videos/PLZlv_N0_O1gasd4IcOe9Cx9wHoBB7rxFl/w4XlBKeE46E/index.html
>docs.unrealengine.com/latest/INT/Videos/PLZlv_N0_O1gY35ezlSQn1sWOGfh4C7ewO/EFXMW_UEDco/index.html
>docs.unrealengine.com/latest/INT/Videos/PLZlv_N0_O1gbY4FN8pZuEPVC9PzQThNn1/yS-yQfo0lc0/index.html

After that we made Pong using Blueprints and currently we're working on a Geometry Wars-esque top-down shooter. Once we're done with these, I'll do a couple Unreal C++ lectures for him and we'll translate those games from BP to C++. Once he got that down, I'll show him the hybrid BP/C++ workflow (which is the preferred way to work in Unreal, assuming you know C++) and do a little refresher for using git.
I'm taking this serious because I'm working on a serious project.

I'm sorry to bother you all again by asking the same thing but as it stands we're in real need. the space station 13 Holla Forums server needs more coders who can code byond, please consider helping out as we do get population on our server and many a good time had.

please contact either chaosyourgod or 8chanserverhost on byond.

...

I'll be your best friend

Can you give an example of what you're looking for asset wise? I might be able to help, just curious what you're going for.

HAAH
WAAW

actually we're looking at fixing combat so it's not shit

...

...

trying isn't going to hurt

Keep in mind you are talking to one of the most infamous "ideas" guy in all of SS13. All he does is sit in the thread posting terrible sprites while suggesting things that would take months of work to implement

I am curios how his sprites looks like, care to post some of his examples or something?

yeah they're shit I know

There's worse, it's not that bad.

hmm meh at least it doesn't look like nu-pixel shite which is in my opinion looks far worse.

still trash tho tbh

So user, are you upset about Microsoft abandoning MS Paint? Because clearly, that's what you used for this shit.

I used krita actually

That said they do look like shit, but I'm pretty sure all of SS13 looks like shit so it's probably just right.

You could fool me if you told me those are official assets.

Microsoft isn't abandoning MS Paint. It's just being moved to the Windows store. Similar to how they did the same thing to Minesweeper and Solitaire.

money.cnn.com/2017/07/25/technology/microsoft-paint-windows-store/index.html

Even if they didn't, there's literally hundreds of open source clones of pbrush people could use

I've actually never found a proper MS paint alternative. It starts up instantly, and makes simple cut/paste jobs easier than any other program I've ever tried.

Is it? I read that they're replacing it with Paint 3D and legacy Paint will be a .exe you can technically find, but it won't be featured.

That'd explain it. I guess they changed their minds.
I have no use for Paint, since Gimp is on 24/7 for me. Paint was only good for saving print screens and I can do that with Gimp too.

The image doesn't have an alpha channel tho.

I figured out how to make arches in gtkradiant so I made a bridge.

This game (Star Trek Voyager Elite Force) doesn't have a water shader vanilla. Instead it has this slime that hurts you. I might learn how to make shaders in Quake 3 so I can make a proper pool of water.

Why are you modding that particular game?

Because I like playing it and because I've always wanted to make my own Star Trek episode.

I also have some experience modding Quake engine games so there's that too.

Cool. I hope you succeed.

The great thing about modding Quake engine games is because they all use similar editors/technology. Tutorials are largely interchangeable.

Like I was mostly following a tutorial for Call of Duty 2 with making the above image. There's actually extremely few tutorials for Elite Force specifically. However since the developer Raven Software also made Jedi Outcast and Jedi Academy, which run off the same engine and use the same scripting language. Tutorials for those two also apply to Elite Force.

Right now the biggest hurdle is learning the game's scripting language to make cutscenes. Thankfully the animation sources have been released and there's a blender plugin for md3 so theoretically there aren't any technological hurdles I'd have to overcome to mod the game.

that also looks exactly like the source engine's hammer tool, you could probably just as well make HL2/TF2/L4D/Portal maps

This is in gtkradiant. It has an option for the traditional 4 way view.

There is a way to decompile .bsps from Source and recompile them into Quake 3. I actually found a tutorial on doing it.

I have made Half-Life 2 maps in the past but it doesn't interest me as much.

Just a random report, given the recent release of Splatoon2:

As I've been developing my game, I've been having my dad playtest some Nintendo games here and there. (He has barely any experience with games, is very bad at them…)
Kirby: Super Star Deluxe, Super Metroid, Super Smash Bros., Pokemon, Zelda (Ocarina of Time, Wind Waker, Breath of the Wild), Splatoon…

Of all of these games, the only one that REALLY 'clicked' and was truly fun for him was Splatoon.

As you all know, practically every game out there claims to be "beginner friendly" in hopes of fishing filthy casuals.
The truth is, 99.99% of them fail to truly be good for people who are new to video games.

Splatoon is the 0.01% that actually, truly is beginner friendly, to the extent that someone who has hardly ever touched a controller before can pick it up, play it and enjoy it within 5-10min.

For advanced players, Splatoon has some good features and some frustrating flaws, but there's some really excellent design decisions not done in other games, so it's definitely worth studying and analyzing.

I'll end this ramble with a warning: do not just go about designing every aspect of your game with "beginner friendliness" in mind.
You'll end up with a game like Smash 4 WiiU/3DS, for example, where the game ignores player input half the time for the sake of being kind to beginners. (In reality, this makes the game confusing to beginners (because inevitably their input will get ignored for some dumb reason) and frustrating for veterans.
A much better line of thought is: how can you best convert new players into experienced players?

...

Given that Source is built on GoldSrc which is built on the quake engine, I dont think its much of a leap to say that Hammer IS the quake engine's editor

Has anyone found the answer to how to stop thinking up game ideas when you're already working on one. It's like my brain is sabotaging me.

Wew.

Write them down; and I mean properly. When your brain pushes them into your conscience, it tells you that it doesn't want you to forget them.
Don't expect it to work instantaneously though. It'll take some time, but it should work.

To be technical there is no official editor for the original Quake engine. At least not one that was publicly made available.

The original Quake was created on NeXT computers. And they used a map editor there called QuakeEd. The source code to which wasn't released til a few years ago.

However at the time developers were starting to use Windows and fans were making their own editors. Like there was another editor called Qoole around the same time and Worldcraft. Which Valve hired the developer of and would later rename to "Worldcraft". They dropped support for the original Quake engine like a sack of bricks pretty fast with it too once they renamed it. A ton of Quake 1 maps were made with Worldcraft though, including the recent Machine Games episode. However there are more modern alternatives like Trenchbroom nowadays.

When Quake 2 released, Id made their own Windows editor called qeradiant. Which was the basis for gtkradiant, which people use for Quake 2/3 editing today.

Whoops. Meant to say "Hammer"

To be fair, making your game accessible to people who don't have any experience with videogames isn't an easy task. Especially in this day and age where most AAA games use control schemes that require a lot of face buttons on a controller (or keyboard) and need lots of UI and interactive menus. You can't really expect someone whose never played a videogame before to grasp all these things at once. Most modern games are accessible/more enjoyable to people who are familiar with videogames because concepts in modern gaming are based on games that came before them.

Take for example, Super Mario Bros and Breath of the Wild. People who were new to videogames and played Mario Bros had a much easier learning curve to deal with. With only a few face buttons, all they needed to learn to play the game is how to run and jump. They'd learn to jump on enemies, hit blocks, and platform using one jump mechanic. The game then becomes harder by throwing difficult enemies at you, harder platforming segments and the like, but the player will learn to deal with these because all they need to do is understand a single mechanic to beat the game (along with only needing 2 face buttons and a dpad). Now, if you were to have someone whose never touched a videogame before play Breath of the Wild you can start to see why they'd have a hard time with it. There's quite a few face buttons to learn, along with understanding how to move your character in a 3D environment while controlling the camera. There's contextual button presses, lock-on targeting for combat, and a whole lot of interactive menus and such. A new player will have to learn how to move their character, how to navigate different menus, and combat just so they can play the game. And that doesn't even prepare them for actually progressing in the game because they'd need to understand how to do the main quests and where to go/what to do and how to improve their stamina and hearts.

I don't think there is a game, including Mario, that really teaches someone not only how to play a specific game, but how to play videogames in general. It's a skill that you build playing various games over time, and you need to be aware of certain schemas to be able to play todays games.

I don't know if it's the best way, but making a game with simple mechanics that can be exploited for high-level play would help. The game Furi does a good job with this. You can shoot, dash, and attack, but you can also do charged variants of these same moves. The entire game is based around 1v1 boss fights, and the normal mode is easy enough that you can beat it with just these four moves. However, once you're used to the game and know boss patterns/timings, you can start to find ways to use the charge moves effectively and deal more damage.
Embed related.

who here /cozy/? post your devstations

Who sleeps up top?

I do

TBH last good mspaint was on XP.

Didn't that version lack a crop button?

I remember using my Windows XP computer and was surprised how Paint lacked a bunch of features.

Before anyone asks, I just made the cake today. A bit of a celebration.

Ive always gone to Katamari Damacy for introductions to games; the concept is novel and simple, the presentation is charming yet not childish or girly, and short (2-10 minute) levels allow for breaks and a sense of progression.
Games are most engaging when it's easy enough to know what you need to do. Coloring the stage in splatoon is a clever objective, since you can always see it and are never prevented from doing it. One of the biggest hurdles is the exploratory factor of progression (be it in objective or in learning the game). You cant get far in Super Metroid without something in your mind thats asking, "whats on the other side of this door?", since the next question would be, "how can i open this door?".
Figuring out and connecting two mechanics to each other, like mario being able to break bricks depending on his size, requires a similar level of questioning. I suppose its this level of curiosity and imagination thats unique to people who play games, which is pretty hard to teach, and easier to learn.

Side note: a lot of people have a hard time controlling free cameras. Since moving your head/eyes around while walking is a motion often taken for granted, it's hard to relearn how to do it, and juggling two actions at the same time can be confusing to new players. Many japanese developers (especially during 6th gen) have strayed away from controllable cameras, possibly for the same reason. Resident Evil 4 solved this perfectly in my book, since the camera is locked to movement until you start aiming (while not moving).

Yeah why the hell not

Introducing Toaster Station™.

...

How much porn you got on that thing?

Spent a week UV unwrapping a bunch of church props to texture.

That's microfiber cloth and lens cleaner my dude. Who saves porn to their hardrive anyway?

Make an IWBTG style platformer about avoiding traps and fill it with as many ambiguous statements involving traps as possible.

I uploaded a new version of my Factorio mod ARWM 0.1.9 for Factorio version 0.15.26.
Get it here: mega.nz/#!VRVQnTZb!wKOf0iPb42MI-ak30gZ7qZ-5-KDZV80xPgfSvf0JCMI
Crosslink: >>>/vm/1766
Please report all bugs related to my mod at my /vm/ thread instead since I might miss those reports if it was posted here, I haven't added much content this time apart that my custom enemies spawn now and the 2x5 warehouses types can be researched now, since I focused most of my development on balance instead.

Looking good, mate. You should post to /3d/ sometime.

I do. Mostly because sites baleet it constantly

...

sounds like the selkath

there's a million fangames out already
and only a handful of them is worth a shit
don't waste your time

Very nice.

Make a game about a trapsmith who uses disguise skills to get into places to set traps.. places like.. Women's bathhouses. The Queen's royal chambers. A tea party full of the daughters of the wealthiest nobles. Places like that.

I need some serious help with arithmetics here.

There is such a method to get the distance of a point to a line segment. however, what I want is to get the vector between a point and a line segment.

I tried adapting the point to line segment distance method, but it is not working when the point is between the segment ends. Here is the current thing I have:

float3 directionToSegment(float3 v, float3 a, float3 b) { float3 ab = b - a; float3 av = v - a; float3 bv = v - b; if (dot(av, ab) = 0.0) { return bv; } return cross(ab, av) / length(ab);}

Hey, do you think you could post those webms of your gore systems again? I wanted to link it to some people as an example.

Would you mind giving some advice to an amateur 3D artist who has no fucking idea how UV unwrapping works? I'm using blender.
Thanks.

So the vector that points from the point to the middle of the line segment?

cross(ab, av) returns a vector perpendicular to those two vectors, which if you look at the situation as a flat plane would be pointing directly in or out of your viewing plane. You could use thaat info to get a vector between the point and the line, but it wouldn't be t the right position. What you actually want is, off the top of my head:
V - (A + (AB * dot(av, ab)));
AB * the dot is a vector along AB to the position nearest to V on the line, A plus that is the actual position that vector would be in, and V minus that is the line from that point to V.

No, shortest distance. This should be obvious from his code, where av and bv are correct in those if statements.

Expeditions are now working 100%
Fucking finally

I can't figure out why it did not work. I bet it's something silly

I enjoy the feeling of blindly doing things.

Your game better run vorestation code

I've worked non-stop on solving a huge game-breaking bug, only to realize two hours later that I made an additional controller object, so every action registered twice.
Shit I'm tired.

Does that affect A presses?
Specifically, do we have to resolve real vs virtual A presses?

Well, every A press is registered twice, so every half press become a full press, I suppose.

It runs like a fucking dumpster fire.

Related.

Feels bad man

Over/under how long it takes this gif to show up on Gaming and Pandas timeline?

but that's ghey

>gonna move in with fiance in six months
>he's gonna take care of me financially
>i can gamedev full time
>life is looking comfy for the first time

Sounds quite dangerous.

How about you get the fuck out of here with your blogpost gay bullshit?
Ninety percent of why people hate you faggots is because you refuse to act like normal people and have to flaunt it everywhere. People were fine even with MoM user until he did what everyone predicted he'd do and start spamming his furry fetish everywhere.

Eh?

That looks like my former mayor. RIP

Fuck off faggot

MoM user started doing shit like posting images of his fursona in the thread. He did it at least twice and the previously mixed opinions on his works swayed towards people getting tired of his shit.

Dubs confirm.

my google-fu sucks.
I have enabled some option on accident when I was rigging, and now faces are always transparent.
I have textured/solid applied for maximum draw type, but still there's transparent faces.
I think I hit some "wire mode only" hotkey and I have no idea what key disables it.

Picture of it happening, and proof that I have texture max display mode on.
Also can't just start a new project.

figured it out, it's "z", JUST "z"
fucking blender i swear

If you hit "shift + z" you get a live render of the scene. Blender keybinds are fucking terribly documented.

I want to get into game development and have an idea in mind. Would it be pretty easy to make something like a mix between Daggerfall (much smaller scale of course) and Doom coding wise? I find something really comfy about 2.5d first person and would really like to make my game with it. Is this a good first project?

A good first project looks like pic related.

Fair enough, but that's more a "Hello World" exercise type of thing, I mean like within about 1 year could I pick up enough to start making at least a 2.5d FPS engine?

once you show us your hello world you may ask where to go next

Unless you're already very skilled at programming, no it isn't.

Yeah I'm using Blender too and also an amateur. When you texture models, you have to convert your 3D model into flat 2D space so you can paint it and then save it as a png and jpg. It's like peeling an orange around and around until you have one big long strip of orange peel, then you flatten it. Since it's physically impossible to have completely perfect flat orange peel because it comes from a sphere, when you flatten the peel, some parts get distorted. So when you paint it, you have to take into account these distortions to look right in the final 3D model. That's why game face textures look really weird, because they are long and stretched to take into account the distortion from the nose and cheeks and other facial features. So to be good at UV unwrapping, you have to unwrap your model in a way that reduces these distortions and makes it as flat as possible. It's problem solving.

Here's an example from my church roof.

Does it make sense to break up "story" and "script" in a game like in movies?

I absolutely love doing UV unwrapping. It's like a a complex puzzle to solve. I learned it for 3DS Max but I plan to learn how to do it in Blender soon.

These days there are a lot of auto-unwrap buttons built into 3D modeling programs and plugins, but knowing how to do it all by hand is a good thing to have under your belt for when those auto buttons fail to handle whatever strange thing you just made. If the model is particularly boxy/angular/not very complex a lot of times I just apply a box unwrap to start the process, and then move and stitch together the peices it makes where needed. This saves a lot of time but you still are getting your hands dirty dealing with the unwrap.

I guess, especially if you keep the script in an easily parsable format so you can easily turn it into something the game itself can read.

I mean creatively
I'm stuck with the story but I want to try my hand at writing, and I have kind of a world and some characters already

The picture hanging there aways haunted me for some reason. I think I grew out of it though

Yeah, sure. I attached the main webm I use to present it. More details are in this post:
beelzebox.net/2017/04/18/heart-on-a-platter-blood-and-gore-part-2/

I haven't really advanced the gore system since the adding the intestines, the other game systems needed work But I hope to use some advice I got here to improve its performance

I don't remember this at all and im always in these threads, are you sure ur not full of shit?

I hope it will be possible in your game to collect all demn shekels, shekel sniffing when.

(1)
I'm surprised you don't remember it honestly. He pretty blatantly posted images of his fursona whilst discussing character design or some shit and got told to can that crap.
(checked)
Thanks, Beelzanon. I wanted to link your work to another user because your gore system is really impressive. I don't think I've seen another game with that sort of detail.
(checked)
Thanks for this this guide as well. I'm trying to get better at modelling so I can make some guns for Kowloon.
My modelling is pretty basic right now. I'll try and apply what you wrote tonight.

Selkath?

This is better:
wiki.blender.org/index.php/Extensions:2.6/Py/Scripts/3D_interaction/Dynamic_Spacebar_Menu

All the stuff in a single menue. I heard about it before but I thought that native spacebar thing was it, but its not.

I wish Blender was hotkey centered, its main issue is having a bit too many different types of menues and stuff you can't even hotkey (select edge loops doesn't work as a hotkey, you can't hotkey 'display->only render', etc..)

Blender had a fork for a while called Rocket 3D Clay, it had no buttons, it was wonderful.

If you don't have a really, REALLY good reason for doing this, it's not fucking worth it.

Still working on this game.

I've planned to make an overworld system with mini-games, a store and other stuff apart from the management section (.webm), which is the main part of the game, but I'm on the edge wether I feel I'm ready to feature-creep just yet or not. I've worked on the management and dialogue part for weeks, but I'm getting a little tired of working on it. Should I continue working on the main part until' it's finished before I go work on other stuff around it?

for some fucked up reason, I had to divide the whole thing by the square of the line segment length

float3 directionToSegment(float3 v, float3 a, float3 b) { float3 ab = b - a; float3 av = v - a; float3 bv = v - b; if (dot(av, ab) = 0.0) { return -bv; } return (a + (ab * dot(av, ab) / pow(length(ab), 2))) - v;}

but now it's working.

now I have a 'capsule collider' to bend grass when the 'grass bender' object moves more than half it's range in a single frame

also, I am pretty sure I can optimize it by using some clamp magic, but honestly, it can wait

maek burger game

I definitely misspoke there, I'm going to use someone's engine for sure. After looking at Daggerfall Unity I'm leaning towards Unity or maybe Source. Unity I know has plenty of tutorials and documentation out there.

If this is babbys first game project: start in 2D, make a few simple games in Game Maker, and when you've got a feel for the structure and got some experience under you belt, then move on to 3D.

...

I have no interest at all in console. Godot looks interesting, I'll give a proper look in to it. What makes unity depreciated?

Okay, maybe I'll make something closer to Wasteland 1 then.

I don't think you're listening.
Making pong isn't a "hello world" excercise, if you've never made a game before, I'd even suggest starting out even simpler.
Making video games is a humbling experience for newbies.

Fair enough, I am probably underestimating it. Still Wasteland 1 is a pretty simple game given it's mostly just stationary sprites, the world map, a few animated portraits, text on the screen and math and flags behind it. I'll just start learning what I can and go from there anyway, like you said I'll have to start with a few simple projects.

I think he's shitposting. Godot 3.0 is a major update, with big improvements to 3D.

These retarded fish people from Star Wars: Knights of the Old Republic.
If you want you can put manaan playthrough in youtube search or something similar and you're bound to find some clips of them speaking.

Hey /agdg/, I haven't been posting on here for a while so, I am back to doing work, so, this is what I have: My map editor can translate the mouse coordinates into the "game-world" coordinates, according to the veiwports, etc etc. The next thing to add is placing rectangular prisms around the place.

HNNNNNNNNG

...

Why yes, I did make a unit to do just that.

Yeah, see the only map editor that I have used is Valve Hammer Editor for goldsrc which of course is just worldcraft modified for goldsrc, so i'm copying what i'm used too. I want to do it right this time, so this editor will be enjoyable to use, this time.

Game engine that I am working on. The successor to my last game engine, "sigma 1"

beels mad fan

I am cursed to be forever an ideaguy

Write your shit down at least so you don't forget it. Get your cool ideas down and then take some time to go through them, find one that looks doable, make a spec and just like make game.

How do you start in 3D level design? Make drawings from the top and side? Just start in the editor? How about creating large interconnected metroidvania-style worlds? Any literature on this?

UE4's setting and changing pivot is absolute fucking cancer.

I used to have a diagram about the different methods of designing levels but it mostly boiled down to planning a level out first and designing it afterwards in stages.

Specifically if you look at famous levels, like DE_Dust2. You can see there's really only 2-3 specific areas that are just connected together. Like T spawn, bombsite A and bombsite B. The level was clearly designed with those areas in mind and the rest of the level was added onto to compensate for it.

Generally people get bogged down for instance with detail when in reality they should get a solid base of the entire level first and add detail later. Like design the entire layout of the level first, and then make it look pretty. It's similar to how you would design a house or a piece of furniture, like you wouldn't sand/paint a table before you've attached the legs.

You can see it in timelapse videos like vid related where he starts out with a very basic idea and adds detail to it gradually after it's already been fully realized. This is a very important part of the process, and if you skip it you will waste more time than you should. It's why dev textures are a thing.

Also do tutorials and do them multiple times.

You can draw it out if you really want to (as in a floorplan) but this tends to be more reserved if you're making a multiplayer map like one for CS Go or something. Most people create a basic layout of a map first and create sections or zones for things later. Or use reference images of buildings for environments. John Romero described creating Doom's maps by a process of just building and running around the area several times to get a sense of where to place things like monsters and weapons.

It's also good to involve other people in the process. Like ask someone how your map looks or ask them to play it.

Metroidvania style games tend to have specific locations that are just connected together somewhat arbitrarily. It's not actually that cleverly designed.

Thanks user, although im not creating a multiplayer map but the world for an RPG. I find it hard to for example take a Cathedral or whathaveyou and then build a level within it. I think i'll just keep trying. Starting with the rough geometry before thinking about what in the world represents those shapes is definitely a good point.

more tiddie progress… got a lot done in terms of blendshapes and general rigging, though there's still work to be done, since there are some weighting glitches, as you can see, plus the hair still needs work.

I also need to make more clothes.

Why does agdg hate Source?

Please make it so that when you cum inside of her, her pupils go heart shaped. This is my fetish.

Its very hard to do good noses with animu, but you pull it off tbh.

It's a mountain of duct tape on top of the Quake engine and you're better off just modding one of the id Tech engines instead of fighting with Source's many quirks.

what if somebody made morrowind but anime?

What if somebody made darksouls but with magical school girls?

how do you prep for dumb players

For an RPG it can be very different since you need to think bigger picture stuff like worldbuilding. Like what kind of world the game takes place in. IE: Is it low fantasy or high fantasy. Stuff like that.

None of these things are strictly Source concepts I'm just using a source engine video to illustrate my point.

So, I'm working in python to make a text tamagotchi-like program, but I'm having a problem with a timer.
Basically, I have this loop that decreases the health variable using time.sleep as a timer, but I'm not sure how I can interrupt/break out of the time.sleep function if I want to input commands immediately rather than wait for the timer.

I'm fairly new to python so, I'm not quite up to speed with some stuff, so I could be wrong about this, but do I need to learn how to thread my loops or how to create an exception? I've done some research, but I don't quite understand. Pic related screenshot.
t-thank you

the problem is that: python's input is blocking and so is python's sleep afaik. so, I dont know how because I dont write in scripting langs but just find a non-blocking version of the "raw_input" thing and , then also do not use sleep to sleep because it is blocking, just measure time passed in a loop yourself (by the way you have to call sleep for the smallest amount of time possible in this loop to limit CPU usage, dont ask me why because you are a PYTHON programmer and this is not something you would need to think about except its how you make a while(1) loop not take up the whole core (usually they dont because of blocking things like "raw_input" ok?

Another really important part of the planning process is establishing context.

For instance. What kind of place is the player going to. Is it a town? How wealthy are the people there. What architectural style do the buildings have (IE: Rustic fuedal architecture or like skyrises). What kind of jobs does everyone have. What is the player's objective and how are you going to lead them towards it?

There's very little harm in overplanning what kind of place you want the player to visit even if he's just going to run in and gun everything in sight with a shotgun. Because it establishes constant points of reference that smarter players will pick up on. And adds smaller details. Like if you have a bunch of warehouse maps you can guide the player through it via a railroad line to establish which direction supplies tend to go. You want the level to tell a story as much as the plot of the game does.

This sort of thing is something Valve is really good at. You can see it with Portal and with other games like Left 4 Dead 2 where an enormous amount of worldbuilding is limited to graffiti strewn around the level.

noice, how many types of shiny shekels do you have in-game? Also can this unit steal borrow shekels from the stupid goyim? Cuz making a Elder of Zion empire ain't cheap.

How am I supposed to structure code in C in a way that makes sense, when you don't have methods?

For example if I have 2 enemies with different attack types, I can't just do enemy1.attack() and enemy2.attack(), instead I'd have to make a special attack function just for each enemy and do something like enemy1attack(enemy1) enemy2attack(enemy2). That's obviously retarded and there's no way I could remember all the functions relating to each individual object like this. I want to move to C++ just so I could have methods.

Am I supposed program in a different way and where can I learn it? Or is C just not meant for complex programs?

Alright, so basically I need to find an input module that won't block, measure out the time in a loop by itself, and use less sleep functions to limit CPU usage.

The goyim mine gold from gold veins and return to a closeby silo to store it temporarily. Then a cart comes to take that gold to a treasury where it will be converted into currency which is then used to fund unit production, etc.

In the space between silo and treasury, the chosen sits in wait to steal the goyim's gold and bring it to his own treasury, or vault, and fund his own campaign to remember the 6 quadrillion.

Also there's jewels and stuff that lizards use, but they're less valuable and therefore less desirable.

I'd play it to be honest.

What you're calling "methods", are just pointers to functions.

The solution is easy: Just store a function pointer for the attack function and call that. That way, "enemy1attack" can be abstracted too: "enemy1.attack()" like how you want. That is what function pointers are for.

Really, its the right tool for the right job in this case, but they don't see a huge amount of action in all of your code. It sounds like you need a bit more practice on C since you're just trying to apply OOP thinking onto every single aspect of your program like a java programmer. Function pointers are only useful for small sections of your code where these cases come up. Other than that, the kind of scenario you described is really rare.

I dont know how python works so take it with a grain of salt. I'm just telling you how I would do it in C.

Understandable, thank you.

I'm retarded for not thinking of that, thanks

That's why I asked where I can learn the other way. People always whine about OOP and C++ and what have you but I've never been able to figure out what the alternative looks like in practice.

I guess a good example of how C programmers like to write things is here:

C has no "try-catch" or exceptions. You're supposed to do things like that using goto's and longjmps. I have a code file that has 30 long-jumps in it for errors. (its not a unicorn… you can use them!) Check out how linux likes to do this: web.archive.org/web/20120331202351/http://kerneltrap.org/node/553/2131

You also don't get generic types, so you can use void* and its great. I usually do some thing where: I have a void* in my struct, and then a uint8_t, and I use the bitwise operators to set certain bits on this uint8_t to tell me what struct i should expect in the void*

By the way, I rarely use enums: Instead I will just declare a uint32_t or something and start setting up bit flags in it, I find this useful because I can set multiple flags at once, which makes it better than an enum. Its like 8 _Bool's!

You don't get the STL so you have to roll your own data structures. What this does is you can usually do better than the "one size fits all" approach. On the downside it takes longer, but if you just memorize how to implement a linked list or whatever its not a big deal.

Maybe a lot of these things are stuff I do as a C programmer, and you could write C++ like this, although maybe its designed so you have an easier alternative. I also make sure that I free everything by using the kind of "spin-up - spin-down" idea, where all your code that allocates memory should follow that model.

In general its hard for me to explain not writing OOP because… its like if you have functions that take in a bunch of structs, and its not implicitly tied to one object (struct) in this case, which I think is cleaner because most of my stuff is operating on multiple objects (structs) so having a strict association seems limiting.

New versions of python has something called asyncio. It's designed to solve exactly the problems you are describing. It's difficult to learn, but well worth it. Also, it's not as difficult as threading and is basically a necessity for (good) UI programming.

Bear in mind that function pointers introduce indirection. They're good for polymorphism, but you don't want them in performance-critical code. So don't do
for (int i=0; i

Her arms seem kind of short.

...

It's a feature.

There is a converter for 2.1.x -> 3.0 projects in the works.

Anyone know any good tutorials for making clothes over a model in blender? I'm thinking of either using the shrink wrap modifier, or mask extract.

Want them to work as Apex cloth later?

It's a fucking alpha, wait till the stable release before migrating your shit.

Not really, I want to have control over everything in the animations, even the way the clothes move

In that case i don't know, sorry user

The cloth simulation modifier can bake into animation frames, which you can then further modify to your tastes.

some tests with the bobs

I plan on doing that, it'd be easy. also I want to experiment with a toggleable X-ray function

Thanks, good to know it improved. in earlier versions people were complaining the nose was weird.

yeah, it looks like they're a few inches short. I don't want to have to redo the entire rig, so if I fix the arms it will be much later, if I do a second Sena H-game.

Alright awesome! This program that I'm working on was originally a Discord bot and that featured asyncio, but being so new I didn't understand the API, so I moved on. I will check it out for sure. I may have found an easy work around through threading though.

So if I'm using C#, how difficult is it to set up a networked game?

I'd want to code a server with up to 100 people (realistically 20 max players). Stuff gets handled client side, but commands would be sent to the server for validation. If the client command fails validation, the correct game state is then sent back to the player

That sounds like a convoluted way of setting up an authoritative server system. There are two different approaches to multiplayer

Thinking about how a SS13 clone would work. Don't worry, I know it's a meme project nobody ever starts or finishes.

But wouldn't the gamestate be too large to constantly send over? The map, for example is 256x256 and has probably over a thousand objects on it at a time

The second approach does not eliminate hacking.

What benefit would they be getting from? Hits that should have killed them will not be registered for them at all?

Usually you're just sending what has changed instead of re-sending the entire set of data. So, object A has moved to (X/Y), object B has been destroyed, object C is now doing Z and so on. And i think even re-sending everything wouldn't be a problem bandwith-wise, think about how much bandwith 4k streaming consumes, all you're sending is maybe 2kb of data per tick.

Yes it does, a properly done authoritative server that only sends your client what your allowed to know and only accepts input that is possible makes hacking impossible.

Say someone is lagging and everyone around him has stopped moving for a second because his downstream is fucked. He shoots everyone in the head. A second later, connection frees up and the messages are transferred. He sends "headshot player 1, 2 and 3", and only now does he receive "get killed", when in reality he would've been dead a second ago. Outcome is, he was invulnerable for the period of the lag and killed three players.

...

It will prevent some cheats, but not all of them. In a lot of cases only relevant information is required for hacks to work, i.e. the information about enemies within the player's vicinity is enough for aimbots and wallhacks…
Some cheats don't need even that data to operate. My mouse has spray patterns for different weapons in CSGO built into the driver.

I can fap to this.

It's weird, I think is hot but looks pretty damn bad. Might just be the terrible faces and the poses being unrefined, or maybe I have an undiscovered fetish for wireframe ladies.

I suppose for shooters it's a different thing, but this approach does eliminate hacking in the classical sense of the word in RTS, MMORPGS, ASSFAGGOTS and so forth. Scripting is a different thing, but with this system you won't have people flying around or instakilling you with a ton of damage or being invulnerable and so on.

Does making all picked up chemicals (including household stuff like bleach, toilet cleaner, ect. and actual lab supplies which add a lot to your count.) turn into a single "chemicals" item in the inventory sound like a reasonable idea? My idea is that it would save a lot of inventory management and simplify crafting.

Yeah it for sure helps and is imo the only correct way to handle networking, but it does not completely eliminate hacking/cheating, that was my point.

Depends on how involved the crafting is otherwise and how varied the chemical items typically picked are. If the player picks up ten cans of bleach and can make both advanced narcotics and napalm out of it, then it will quickly becomes jarring.

I dig this, how did you achieve this effect? The bartender is my favorite.

You could broaden it into more categories of chemicals, bases/acids for example. But again, that depends on what you want to do with your crafting system.

Also realize that your crafting is basically A+B->C

This makes you have a huge supply of [A]. What you could do is have different qualities of crafting ingredients, but keeping the actual types small. So you'd have "household chemicals" actually be "simple chemicals" and lab chemicals be a more refined version of that.

Then, your basic recipe could require ANY chemical, but a firebomb might need advanced chems

...

that segues nicely into bringing up my current model

I've been struggling for literally like 3 days on the hair and I feel like shit about how long this model has taken me to complete, so many design shop setbacks and the hair was simply infuriating to tackle and find a good design for

hopefully the next model I do will go by doubly quick, i still need to UV and texture this beast

were there any tutorials you found to help you do clothes like that?

no that's just my 3 years of experience modelling nothing but TF2 cosmetics LOL

So why'd you stop? Didn't you make a good sack of money off of it? You could've also hopped on the DotA2 cosmetic bandwagon, and also Unity asset store

who said i stopped?

Why is Unreal's BSP editor a pile of steaming shit? I have years worth of experience with Hammer Editor and related Quake map editors and always hear how Unreal Editor has the best map editor in the entire industry. So then I've been fumbling with it the past week and it just has the most retarded key shortcuts (Alt+V+Scroll wheel button, Ctrl+End, etc. fucking really???) and slow and annoying vertex manipulation. Also the pivot is retarded garbage never being in a place you want it at and the scale in the options can royally fuck up having everything neatly on the goddam grid which has the worst options for size and 4 separate options for translation, rotation, perspective, and fucking mongoloids.

Am I alone because I'm an amateur at the editor or is UE4's map editor the result of a good engine and tons of editor development bloat?

Look at this fucking shit. Snapping points in vertex mode is fucking pointless as shit! It just snaps the whole body instead of the individual vertices that I highlighted.

Im not sure but i think you have to switch to scaling / resizing mode to do what you want to do, you're in transform / move mode right now, it's the three buttons to the right of the show icon.

Nope it still does it. I don't even know how that can happen. The location is all nice whole integers and the scale is 1,1,1 so how in the fuck is it 0.1 longer in one direction.

Fug u.

Here, a quick guide. What sort of game are you working on? I could offer some advice on audio work and effects.

No, it's just complete dogshit to anyone who has used a Carmack or non-Unity engine. It's so shit that this guy is making bank
nte.itch.io/hammuer
Me and my team mate considered buying it until I realized it would be better to just go along with Source for our project.

Is there even a way to get past it's awful style of scaling where you can't scale just one side?

And before anyone asks no that is not actually Russian at all I'm just a leaf that enjoys STALKER too much.

I don't understand your post. Also I gave up and just went ahead and built it using Blender. Then I'll just export it all as separate rooms and what not and put in the props in UE4.

I see how you'd be in for a bad time, if you worked with it as if it were a complete level editor (which it is not). You're not supposed to ship your game with any BSPs in it, hence the built in feature to convert them to static meshes. The BSP feature is merely there for prototyping or "blocking out". The workflow you're expected to use to create, say a bridge, is as follows:

You don't use the Unreal editor for creating the geometry itself. You create the objects (which includes your level geometry) in your 3d modeling application and use the Unreal editor for placing the objects in your world, scripting, lighting, etc. And that's a very good thing, because 3d modeling is quite the complex subject and you don't just add it to your editor like it's not a big deal. Even the simplest modeling tool I ever used (SketchUp quite a while ago) was worlds better than Hammer. I think it didn't have smoothing groups and texturing was a bit unintuitive, but when it comes to creating the actual geometry, you can't even compare these two.

Since I can't afford any artists, I learned how to use Blender. And now that I am familiar with it, I'd rather cut my own nuts off than go back to Hammer. Its shallow learning curve is literally the only thing it has going for itself. And it only has this shallow learning curve because it's basically a 3d modeling application for cavemen with how primitive it is.

took a break from UVing to experiment with shaders and stuff in blender as a concept for what the lighting and shaders could look like in-engine

how would something like this look? 3-tone lighting with ambient and adjustable specular per-material

is there any way to change the ambient light color per-material in unity, btw? say I want to have an orange material with red-shifted shadows, or a teal material with blue-shifted shadows. is that possible?

Iirc Epic has stated that you're not supposed to use bsps. Most people in the industry even agree and see using bsps as outdated.

You do realize if you wanna sell your project you'd need to shell out 20 grand to pay the Havok license.

Yeah but even if you do use them it's a piece of shit and an insult to its spiritual predecessor. Even if you are just doing prop placement it can be a pain in the ass.

It's just plain and simply bad.

That look really good, actually.

Yes, as they're slow and inflexible.

Releasing the source code and selling it are two different things. Carmack didn't make Doom freeware after he released the source code

I mean sure if you just wanna release an indie game and don't wanna make an actual polished game.

If you licensed the source engine I doubt this would fly with Valve. They'd most likely hide something in the code to prove it was you that leaked it. Leaking the source code in general tends to only be done by disgruntled employees that hate the company they're working for.

And if they can prove it was you that leaked it you're looking at a breach of contract/nda which can get you fined a shitton of money.

Along this line of thought, I wonder if it would be possible to make a profit of a game while not only open-sourcing the game's code, but also allowing the game's assets to be re-distributed with a (somewhat) more permissive license…

It seems like it ought to be possible if you were a well-known, reputable game developer…
You could get your funding from fans through Patreon (or better yet, through a non-profit organization of your own creation so Patreon can't fuck you over), and just release the games you made for free.

The source engine source code is open to anyone, man. You just have to release it on Valves store. I'm certainly not going to release exclusively on steam.
Also one thing you need to remember is that I an an untalented hack who just wants to make game.

* for free on Valve's store.
I should add that the plan was always to open an anonymous patreon using our "company name" and release whatever we made for free.

...

...

What's this, some kind of deformation?>>13113637

I've seen some of this guy's previous webms, it's grass (for a golf game?).

did you ever improve your pixel art user?

use this technology in a complex lawn mowing simulator

How is godot for 2d? I've looked into love2d which seems pretty good, but c++ is more of my thing.
Is it full of bloat or shitty in any other ways? Anyone got 2d godot experience that can chip in?

Godot's focus is supposed to be on 2D.

Speaking of engines or frameworks, I'm still using XNA for C# shit. I don't want to hook into OpenGL myself and I don't want to use Unity. What's something that's more modern and not Monogame?

It's node/scene setup is a little different, but I like it quite a bit.

I just downloaded it and I'm working through the tutorials. Seems light and robust, I like it so far.

I don't quite understand Scenes in general. Is it basically just an arrangement of nodes to form a template of sorts? Saying that they make a scene for a UI, for example, feels weird to me

Breath of the Wild better look, there's a new kid on the block

A scene is just a collection of nodes, so it can be as complex as a level or as simple as a button.
docs.godotengine.org/en/stable/learning/step_by_step/scenes_and_nodes.html

Pretty cool when you get your compute shader algorithm to work perfectly.
Gj m8
Be careful with compute shaders, once u learn more about them you fall down a path of rewriting (many times) to optimize with the new "tricks"/"quirks" you've found in mind, just get it good enough, and move on.

Didn't lifeweb achieve this?

github.com/godotengine/godot/issues/9390

W E W

...

Jesus christ where does someone start?
I'm having trouble finding tutorials between "look you can make the premade man run around the premade area" shit that teaches you almost nothing and the ungodly dense C++ stuff.

Then just learn Unity, it's designed for people like you

Literally why

Even the portability thing is a shit argument because we now have the raspberry pi.

Starting with blueprinting is pretty good to just learn the general engine flow, the API and what you can do with it. If you want to learn it with just c++ just think of a small project you could do and play around with it.

I started off making the mechanics for a shooter and just went from there.

Ask any questions you have here and we'll be able to help you out.

This is why Unity is more popular

...

Unity is garbage for people that want to make games, user. I want to create something good with my limited yet Jack-Of-All-Trade-sy skillset.
Source allows me to shit out good FPS design simply by following the rules laid down in Doom, Quake, etc.
I am but a humble maker of vidya and fan of cyberpunk who aims not to compete but merely to give his wares away to those who want them.
Also to gain sweet patreon brouzouf without being Jewish about it.

The Ouya was a literal scam designed to make as much money as possible whilst investing as little of it as possible. At least we got a good -tan.
I'm a retard.

AYO HOL UP
*smacks lips*
SO YOU BE SAYIN'

The map editor can now "click-and-drag" flawlessly to make boxes. I don't have "snap-to-grid" yet but I will soon. This right now has a very smooth feel so I like it.

I fell out of my seat and onto the floor, and then, as I was lying there, I have this desk that has a little door compartment on it, so I was opening the door onto my elbow, and I did that for a while as I lay on the floor.

...

SJWberry Pi already existed when the Ouya was anniunced.

that's nice to know dear

i'm going to go to sleep and forget that I wrote that

I grab the edge of the desk and move the chair in and out using only my arms

Don't worry I will remind you tomorrow.

dumb algebra question
i've got XA=B, where A and B are 3x3 matrices
in order to find X, you usually multiply both sides by A^-1, you get X=BA^-1
but what happens when A isn't invertable (A^-1 doesn't exist)
is there no answer, or is there another method of solving it?

if you can somehow stomach slavspeak, here's a bunch of lectures specifically talking about how game engines (UE4 in particular) work, from a game engine course i took and didn't bother to finish
nikoladimitroff.github.io/Game-Engine-Architecture/

I'm not a math man, but isn't the rule stated at 4:10 in this video contrary to that idea unless A is 0?
cdn.virtualnerd.com/videos/Alg1_6_2_8.mp4

What do you want to do? I searched for tutorials as well and then just said fuck it and jumped right in, googling everything i wanted to do as i did it.

I like it

...

...

Right now I've scaled myself back from C++ anything and am trying to wrap my head around blueprints.

This seems to be a recurring thing. See and

I figured I might be able to get into the basics of it after having worked with C# for quite a while.
I was wrong.

Syntax is similar, it's really not that hard. Being familiar with OOP from c# will help and having the engine do most of the low level stuff like memory allocation also makes it easy.

You'll learn it with time, but for now stick with blueprints to learn the engine flow and API. Skills from that carry over to c++ in ue4. Ask questions if you have any hitches.

I like your elephant mug. Very comfy.

I had a miserable time trying to implement A* and it wasn't even A*, it was the more basic kind of weighted path without a heuristic. Best of luck

...

This is what my desktop looks like

So here's another test with the jiggle, as well as using the nCloth jiggle function on the hair.

meant for

Is there somewhere a gud site to learn how to make a okayish looking particle effect with blender or something? I tried to make a acid splash effect but idk man it could probably look much better somehow, also I used blender internal only so far for it.

I believe in you, user. You are not retarded and you didn't waste time, put that suicidal gif away and see the truth. You are going through a hard time now but you still managed to work on your game. Try to appreciate what you did despite obstacles that make it hard for you to focus or find motivation. Whether you created something that works or learned what doesn't is not as important as a fact that you did something when you could have done nothing, you made progress.

Making it is a continuous journey not a static, one time goal and you are on the right path.

All looks gucci to me, man. Will there be physics for the hair, clothes and boobs?

Good stuff, thanks. I'm currently working on Speebot, but this may come in handy for my next game.

Here's a fresh clip of Speebot with a couple extra cheat modes enabled (unlocked after completing the game).

Mixamo is distracting me from game dev by way of me making waifus. I need to get some real modelling skill under my belt as well.

(checked)
Oh fug, you're speebot user? What's your next game about?
I could offer some advice on/offer some robotic sound effects.

I don't know what exactly it's going to be about yet, I've been mostly thinking about gameplay and music. It's going to be an action RPG.
I was thinking about using robotic voice effects in a song (embed related).

Don't start with C++ UE4 if you don't know C or C++ already, and Blueprints are worth getting into, they are easily set up, compile MUCH faster than C++, and help you discover engine features and how things work with all of the suggestion features, much better than Visual Studio autocomplete.

Whenever i do something i usually start off in a BP and then eventually migrate to C++, but only if the logic is complicated. There are disadvantages as well of course, they are slower and making complicated shit in them is annoying and can lead to a gigantic mess. Also i spend way to much time aligning the nodes nicely and arranging shit instead of working

Excellent fucking taste, friend.
Also, those voice effects may actually be synthesized. I have a Gaia Sh-01 synth and one of its defaults is something similar to the "vocals" in that song.
You'd need to bring forth some serious deepest lore if you want to style it like Neir or Drakengard. I recommend going for an usual fantasy setting.
I've a "deepest lore" setting I keep in my mind in case I ever make an action RPG.

I think he's more screaming inside because he wants to dev, but can't, because of circumstances outside his contorl

Okay so lemme see if I get this Godot shit right

So does that mean because scenes are instanced, the root node could be like a Ball, and then I have my main scene and I put the Ball Scene instances in?

Yep. There's a physics demo that does just that on the github. Make sure you have the 2.1 branch selected, godot 3 examples work differently.

I dicked around with the tutorials and got up to the pong one and started working on my own shit at that point. I am very confused

"p" is just a Vector2, you want:
get_node("Junk").set_pos( Vector2…)

Also I think setting p.x and p.y manually would have no effect on Junk. Not sure though, I'm away from my computer to test it.

Right, I want to move it downward by 5, for example.
The only I could see that was to do something like

… which is fucking cumbersome. That's why I set a local variable and updated Junk with it. But it doesn't seem to work.

You are right about p being a vector2, but it just feels a little clunky and weird the way I have my nodes set up

This should work.
get_node("Junk").set_pos(get_node("Junk").get_pos() + Vector2(0, 5))
I think it's much more readable. Using get_node() all the time is kind of annoying; I know there's a way to set a variable to refer to a node but I don't remember how. You can also put this movement code directly in a script attached to the Particle node.

var a = get_node("Junk")

iirc only numbers, strings and arrays get copied on assign, everything else is passed down as reference instead

Looks like Tesseract hates GCC7 and the only compiler warnings I get are for some crypto library, probably unrelated to my rendering issues which don't show up when built with earlier GCC releases.
This is going to be great fun.

FYI, LD39 started. The theme is "Running Out of Power"

You're a tranny, POC, or other visible minority. The goal of the game is to receive money by bumping into other NPCs and begging for it.

However, your begging ability becomes weaker as the game progresses

Write a pretentious indie game that starts off at 60fps and decreases the framerate as your hitpoints dwindle, tanking into the single digits when you're near death. Journalists, hipsters, and industry figures will line up to lick your asshole and your next game will have guaranteed publicity no matter how shit it is.

Short game jams are always hard for me, I come up with a few strong ideas that I'd like to work on and then I can't decide which to choose.

Embed related is all the stuff i've done on my engine lately. I know its dumb to embed, but I don't think I can fit me talking for 16mins in a webm. It sounds better at 2x speed.

Wait, you're the guy who did red sky? Jesus, man. Wasn't that a wholly original engine as well?
Also, what model file types are you going to support for Sigma? I actually kind of want to play around with your engine now that you have an editor.

Have you considered what game you are making on your new engine yet?

I'd like to interject for a moment, that's an interface for an editor which does absolutely nothing aside from trick you into thinking its a real working piece of software. Now, as for model types, I guess .obj? I don't know a single thing about models. I know you sound excited, but you wont have a lot of fun playing around with it for a few months at least. I don't have the "rendering engine" implemented, so that means that it doesn't do much yet. So, i'll hand you an actual zip of what it does now. comfy.moe/cubmje.zip

I have an idea of what it will be, but I don't devote any time to thinking about this yet because it's so far off. So, the idea is that I want it to have enemy animations like goldeneye. The gameplay will be different though, it will be fast like quake because I love playing quake. (CPMA) now, the enemies will have "squad AI" which I daydream about sometimes.

Also, the answer to the first two questions is yes, I forgot to answer them.

I must be fucking retarded, this has got to be the easiest shit in existence:
Two floats, if they differ from each other, one should linearly approach the other per tick, until they are equal again. All i can think of is

>if A < B, A += 0.1f * deltatime
if A < B, A += (B - A) * deltatime

That sounds nice, Quake's gameplay is the best of Id. I don't get what you mean about Goldeneye animations, even though I know what they look like.

Why .obj, isn't it really only good for static objects, since it doesn't support vertex weighting, and isn't the format not compiled to binary, isn't it just stored as readable text?

I don't know anything about model formats, I just know obj is suppose to be simple, that's all. Now that you've told me these things, I wont use it after all.

In reality what model format to use is very far out-of-mind for me as I don't expect to be making this decision until i finish the part where you can walk around a 3-d level, which I cant do until i'm done implementing the culling algorithm, which I cant do until I finish the map editor so I have maps to test the algorithm on, and now we're back here again… That's why I give it no thought.

I think its odd that, almost every single time people hear about my engine, I get asked what model format i'll use. I guess that making and importing models into the engine is just how most people interact with a game engine, so all of the stuff that is on my mind isn't something that someone using an engine would think about, because you don't interact with those parts, even though they're more important and a prerequisite to doing anything with models. Loading models into my engine is a far off though for me.

By goldeneye animations I mean that, you can shoot someone in the arm and they will react to it, you can shoot someone in any body part and they will play a unique animation for that. Its only limited by the fact that I can't do anything with 3d models and have no knowledge of the topic, so I wont be able to make the models and animations for the game probably. I don't know how i'll get assets. Maybe i'll have to learn?

:^)
if A < B: A += 0.1f * deltatimeelse: A = B

Thank you user, turns out i really am retarded

My experience with model formats is just as an artist, nothing technical except reading over the formats documentation when available, I don't understand most of it but the few things I do get are useful. I get that the model format isn't important to you right now, but I think it's that people just want something painless, since painful tends to be the standard, although I guess painful is also related to a lot of old engines, newer engines are much easier at least, only a few issues, things like scale and axis not being standardized.

I thought you could 3d, Didn't you make all your weapons for Red Sky?

Now I get what you mean by Goldeneye animations, sounds cool, I loved that feature in Goldeneye, with mouse aiming that would be infinitely more fun.

I'm guessing your maps will use the additive brush method seen in Quake, I much prefer additive Quake mapping to subtractive Unreal mapping, but I think a lot of that is just how painful Unreal 1s map editor is to use compared to something like Trenchbroom which is really easy to use and very powerful itself.

I think that the models wont be hard to import into the engine… the hard part will be the fact that you will have to know C in order to make things happen in the engine! Since, I don't believe in the idea of scripting languages or bytecode interpreters in my engine, just because its bloat and I don't want to write them. I'll make it so you have to compile a dll that contains the "game-specific" code.

No, in the credits I point out that, all of those are made by: nogamesdev.tumblr.com/

I know what you mean by this, the answer will be yes. I honestly do not have a lot of experience using map editors, so feedback is important. This isn't the first map editor that I have built, if you want to see something that is truly awful you can check out this:
mega.nz/#!NAohHZIT!JTEsQr11m92y-dXaJTw3FOdjmki8Gs3YBCCcMzG1-vA

Just try and use it and you'll see what I mean, here are some tips:

That's fine, Sounds like Quake 2, but I could be wrong

I guess I remembered incorrectly about the 3d stuff, well I'm sure you could find someone to help you.

Well at least that's behind you now.

You can actually use that tool to make all of the maps from Red Sky, every single map in red sky was made using it. The way the buttons work is:

New file (this does absolutely nothing)
Open "S3DM" file (uncompiled map)
Save "S3DM" file
Compile "S3BSP" (sigma-3d BSP map file, what is used by the actual game)
New block mode
Select Block mode
Delete block? I forget
Texture cache editor, you can load in WEBP files and if you click "apply texture" while selecting a texture from the list, all of the blocks selected will now have that texture
Entity Settings, this allows you to pick what entity "create entity" does.
Select entity, this allows you to move around entities
Delete Entity, this will delete stuff of course

Now, its important to note that you cant actually know what the entities actually are if you dont remember what you had selected when you created the entity, or if you look into the files and remember which numbers map to which entities. Embed related is me making one of the maps. By the way there are some jumps in the recording because it crashed about 4 times while recording that video (I think it does it less in later versions though)

Did some map design in Blender instead of garbage UE4. I made sure to make each wall group a separate mesh so I can just easily plop it into UE4 at the very end.

I know that this is nothing, but I am still proud of it.

This is a good model format that is similar to obj but supports bones and animations and stuff. sauerbraten.org/iqm/
My engine YUME uses a format called moe that I based off iqe/iqm.

So I'm finally able to make good game assets, but I'm running out of money and need to make some now. What shop do I sell them at? Unity's monthly fee throws me off, if nobody buys anything I'll lose the little bit of cash that I do have and get booted onto the street that much faster.

God, where's that second life 2.0 bullshit I've been promised by Summer? I'd sell a couple dozen cupboards and footstools there and pay rent.

I'm not completely sure why this is, but I'm positive it has something to do with why the industry is fucked.

It's something, user. You can take the time to be proud as long as it doesn't delay your making the next something.

...

I was thinking about how funny that sounded for a while, after I watched my video again. It's like the most retarded way possible to talk about Vulkan.

I chuckled several times during the video, keep it up.

Ha, I just found out about that freeware a couple of days ago. Nice!

Found you on Steam, Senpai! Can we be friends?

Man this is fucking retarded, it would be more sane to have muzzeflash appearing from a weapon instead doing it on a ammo basis as they are doing it right now. I had this long standing issues where the muzzeflash is appearing from the turret instead at the barrel end.

Yes of course, either the MG muzzeflash looks like it is coming from the cupola or at the top of the gun mantlet, thats fucking stupid. This piece of shit ""documentation"" doesn't even gives protips how to calculate it properly. So that it doesn't look too off either, and the blender sprite creation tutorial on their stupid forums was pretty barebones too so I don't know if there is something to keep in mind out when making a armed vehicle or a turret for that matter.
wiki.factorio.com/Types/AttackParameters#projectile_center

God damn maybe I should rather stop modding at all and play vidya games all day long instead, I am sick of such shit which seems to impossible to correct it.

Take a deep breath, step back from the project for a while, do something fun (get the endorphins flowing), and come back to it later with extreme googlefu.

We all hit walls, just gotta come back when you're cooled down and can feel like you can think.

I wish I know what makes fun for me nowadays, a few months ago or so I was pretty busy playing with Call of Chernobyl for a while and even then it didn't really "hooked" me on it, I have some few other games to play but idk man I am not so excited for it anymore as I was when I got a pirated copy of C&C Generals Zero Hour which is several years ago by now.

Yeah you are right, sometimes a small break is good, meh maybe I will just play some video games for a while till I calmed down I bit I guess. It still sucks thou that their modding documentation is poorly keept. Anyway thanks for the heads up.

What I find is the most interesting part about Pelles C, is that it's the only compiler to implement the entire C11 standard out of the box, which means that somehow, it has some kind of technical merit over all the other compilers.

Okay.

by the way, you haven't sent me a request yet, its steamcommunity.com/id/ika_ , in case you sent a request to the wrong one

New thread soon.

I sent you one too just now. Holy shit there's almost 5k search results with your nickname on Steam.

NEW THREAD

SJWpie is shit.

BeagleBone Black. Much better. They actually care about the hardware, not pushing social agenda fads.

beagleboard.org/black