New Jam Next Month
Helpful Links: http://alloyed.github.io/agdg-links/
New Threads: >>>/vg/agdg
> Demo Days
> Free audio
That's an issue I take with all the existing material out there (AFAIK), including the stuff I wrote a couple years ago. They just leave things at "yeah use a structure of arrays", and don't talk about ways to then minimize memory usage and improve tightness, because having lots of sparse memory isn't good for the cache.
I've posted about this before, but it seems to me like the solution is to break your entities up into partitions. Each partition would have rules about components.
>every entity must have this component
>no entity may have this component
>these components must always appear together
Then, based on these rules, you can store component types in different ways, either straight forwardly in an array or using a single level of indirection.
Learn yourself some game programming patterns: http://gameprogrammingpatterns.com
Noob with noob question here.
>making shitty game in gamemaker
>put an object in a room, object has both create and step events
>it won't work
>put code on the room and use create_instance to put it there that way
>it now magically works
those are just to import basic starter assets such as first person controller and etc.. (you dont even need to use them)
and i think it changes the default view a bit
id use a 2d map for planning the battles
and then take the battles in 3d
>try to use a spritesheet from opengameart so I don't have to use pixels
Dear fucking god what a disaster
The memory itself is contiguous, but the components that have to be processed may be sparse. Doesn't really help the cache to work with arrays if you still have to skip over things all the time. The partitions are to reduce the amount of skipping, and as a nice side effect the memory usage (even though it's not really a problem).
Damn are you making a MMBN clone?!
Oh, I see. Regardless of access patterns, the same issue will occur with an OOP system. Also you can still sort/bin the arrays based on the sparse access patterns such as MRU. It is not a total loss.
Turns out there's a SamplerState.LinearClamp I can use during the spritebatch.begin() call that removes the texture filtering. Neat!
satire post right?
Yeah it looks awful. I think it's because the scale is set so stupid zoomed in because the tileset I'm using is 8x8 that the slowdown as it gets closer to the player target is way more apparent. Tried to zoom it out a little more to see if it looks better but still not a fan.
I'll have to tinker with the placeholder tiles I'm using and the camera movement more tomorrow.
Is it possible to make a rectangle around an image in a panel using Unity?
I've been looking for hours and can't figure it out. I've tried GUI.Box, but that only works for screen coordinates, where as the image is in RectTransform coordinates or some shit.
Don't make me have to get a second image for a border bros.
any basic camera will do the gamecube one because its the n64 game
the ds ones and the recent one have that thing were it looks like its rolling on a sphere
can someone tell me what its called?
>Languages you need to understand to use ASP.NET
This is gonna be so fucking hard but it's gonna be so fucking worth it
That's not what happens.
First, you have something actually handling the collision detection and probably the physics at the same time, typically middleware. It doesn't care about entities, it just cares about shapes and rigid bodies and stuff. But there may be some extra behaviour based on the results of this. For example, something breaking apart if a force threshold is surpassed. To accomplish that, there would be a system (function) that looks at the component that holds the physics measurements as well as the component that shows how the entity breaks, and destroys the entity if the force is too high.
This isn't too different from how it would be in a more traditional architecture, where you'd have a collision handling method that can be overriden by entities or components. But it has more flexibility, as the behaviour can depend on more than just whichever data comes with the collision "event" and each individual component. That's where the traditional ways really fall flat -- they don't handle component interaction well, besides through some predefined channels with elevated privileges (like collision events).
The big thing to realize is that components are not responsible for their own behaviour with ECS. They are just data (structs), and the systems (functions) are what actually provide behaviour.
>Should a component sub system not uncapsulate the functionality?
Encapsulation is a means to an end, not an end in itself. It only creates problems if you try to apply it to ECS.
>Should you not strive to have all the data treated the same (in a batch) and keep the data opaque?
What do you mean by this exactly?
I'm sorry, it just sounds like a shitton if shitty checks and edge cases, which is the complete opposite of extensible or flexible. That each instance is just "data" is of no importance, the important part is that everybody's behaviour is inevitably tightly coupled to a main method (which then calls the particular functions), which, ironically, is really fine if your game logic is simple, but quickly gets overbearing as you start adding more and more particular behaviour.
I'm just not convinced on your words. Maybe it's something particular to your engine.
oh, that's easy
overhead camera at a slight angle
I looked into it before, I don't remember it finding a name for it but instead of making the floor a flat plane, you should make the ground into a cylindrical shape and use the same camera at a 30/45 degree angle.
players won't notice the curvature backwards because it will be behind outside of their view, but they will be able to see the ground curving out in the distance
Then, to keep the game logic in 2d you'll need to project 2d x/y coordinates onto the curve. I don't remember the maths for this part, sorry
2. Linq and sql are not necessary. Also it depends on the database you use.
3. "Database coding" is redundant, because it's understood once you know other things.
That sounds like gross extrapolation to me.
>I'm sorry, it just sounds like a shitton if shitty checks and edge cases, which is the complete opposite of extensible or flexible.
Where? The checks I'm assuming you're thinking about are only done to say "should this system operate on this entity ID?" and the entire point is that there are no edge cases.
>the important part is that everybody's behaviour is inevitably tightly coupled to a main method (which then calls the particular functions)
There's nothing stopping you from making a sequence of systems that are run through in a loop and that can even be changed at run time.
I am not confident in my terms.
In my implementation all data is handled the same and at once. Striving to at no point do you need to know that you are dealing with particular data types as only the component subsystem itself cares about what the internal representation of the data is. I guess the way I have it designed is a component doesn't provide data, it provides functionality.
There are cases where this encapsulation is broken and data needs to be shared, but I do this through mediator subsystems too to try and minimise how much internal representation leaks.
It is working well so far, but maybe I have not come into the most complicated cases.
I'm on agdg everyday.
I don't post as much though but yeah, I lurk a lot.
I think it's important to be surrounded by people who have the same passion/goals as you even if you don't talk.
You mean kind of like how Unity works? I don't see any benefits to that over having components as pure data and having systems provide functionality. It just leads to big issues of interaction between components.
Made a new map in my love2d level editor, special for you AGDG. Hope you like it. In addition to the camera scrolling I got working yesterday, I've also added collision.
Maybe I am still applying an OOP mindset to it. I don't like the idea of systems coming in and touching data as they like so I have the component control access.
One thing I am seeing is more and more data does need to be exposed, but I think that is just how games go, maybe. It is only first try and ECS and first iteration.
gone homo sold a bajillion copys and its litterly one of the most primitive coding for games ever
like i could do that in a week end while high and farming wsg in wow
dont need to reinvent wheel just make game srs
>Where? The checks I'm assuming you're thinking about are only done to say "should this system operate on this entity ID?" and the entire point is that there are no edge cases.
The checks must be on a case by case basis. Think of it simply, you have two kinds of enemies with two different tresholds, and two different death animations, and your method checks, "oh it's this type of enemy? it then has this treshold. is it dead? then do this kind of animation" Obviously it's fine and dandy if you have two enemies. But 10? 15? What about other types of instances, do they get checks too? How many ifs do you need? How nested are they going to be? Do you really think this is fine for a very large scale project?
>sequence of systems
It's the same thing though? You're not getting rid of those ifs, except with inheritance. Or maybe you're a real smartypants and like using metaprogramming techniques like duck typing, or reflection or lisp homoiconicity, I doubt that though.
I haven't worked on my game for a week
My friend is on break from uni right now and he wants me to play assfaggots with him all day and he's my only real connection to the rest of the world
what do I do
Yeah, it can be hard to break free from the OOP Kool-aid that is given out so widely. Encapsulation can be good, obviously, but it serves a specific purpose (not letting users meddle and put an object in an invalid state) that doesn't really apply to ECS.
>Think of it simply, you have two kinds of enemies with two different tresholds, and two different death animations, and your method checks, "oh it's this type of enemy? it then has this treshold. is it dead? then do this kind of animation"
These things are data-driven. The breakage threshold is a value in a component. A particular death animation is indicated by a value in a component. There's no code that says "if this type of entity, do this", because in an ECS there are no entity types. An entity is the sum of its components. You don't seem to understand what you're criticizing here.
>Everyone has to do things the way I DO IT!!!!
>Can't decide if I want to make my game in 2D or 3D
>The checks must be on a case by case basis. Think of it simply, you have two kinds of enemies with two different tresholds, and two different death animations, and your method checks, "oh it's this type of enemy? it then has this treshold. is it dead? then do this kind of animation" Obviously it's fine and dandy if you have two enemies. But 10? 15?
I've only been trying to do ECS for a few weeks now so I'm probably way off, but the way I understand it is that you don't really set your systems up to look for "types" of enemies. Your enemy entities are just collections of component data. That component data should contain all the necessary information about what death animation is to be played, etc. Then your system just does:
And runs the animation for whatever your component data contains for the animation for entities that have the requisite component collections.
For instance in my project for animations I have an animation component that has an enum State (dying, hitting, dodging, standing), and an array of keyframe rectangles for each state. It also has a double frame per second. Then my animation system simply looks at the state data of the animation component, and runs the correct keyframe array for whatever spritesheet is presented by the Display component's data.
In that way I can have dozens of different enemy "types" and animations, but the system would only be as simple as one call with generic logic.
>The breakage threshold is a value in a component. A particular death animation is indicated by a value in a component
I get it, so you just have generic objects with a million variables to cover any kind of instance they can possibly be? But you're still not getting rid of those ifs tho? Because I very much doubt you can really have a "particular death animation" all sorted out with just a single value. Which means, either your engine is doing the match between the value and the animation for you (i.e. those fucking if/else blocks) and you're confused, or you're doing those if/else blocks yourself but not admitting it? If it's the former, stay out of the conversation? The other guy is seemingly not using your particular engine, so you're just confusing him further.
>bosses will wipe your whole party in one or two turns unless you use the right strategies to keep them disabled
>every boss fight will probably take a bunch of tries to get the right sequence of moves down
Is this fun for most people, or just frustrating? There are no consequences for losing and no unskippable cutscenes. I personally like fights like that, but I'm not sure it would sell.
>Create convinient way to comunicate design/pseudoimplementation of a system
>Nah let's use words and not provide any details, you're a faggot
Maybe, but we have two people talking here
i want to make a morbid game
or art for something morbid
This is going to need a bit more designing then I thought.
So, I'm trying to list the possible spell variants, and so far I've got:
The damage-dealing and damage-healing ones
The condition ones : divided in the things they affect (hp, stats, speed), or just a tag for a treatment or behavior : they have a duration
The condition-lifting ones : some buffs might also work as a lift for debuffs and vice-versa
The Shaker ones: creating things (that might have effects associated) and changing terrain : the change might have a duration associated
The summoning ones: summoning monsters of the most various types, can be divided in many different spells | summon familiar, summon elemental, etc | Master Type 1
The enchanting ones: conditions into items, some permanent. Enchant arrows, enchant bows, enchant melee weapon : associating an effect to them (usually damage or condition)
Creating non-terrain items, like food and ammunition, basic conjuring
Enchanting terrain ? probably will never use
Is there any other possibilities?
So it IS an engine thing, either it does inheretance on the background or it matches the value to a particular behaviour, in any case, it's neatly encapsulated for you.
Conclusion: If you're doing a more bearbones approach, you SHOULD encapsulate.
>I get it, so you just have generic objects with a million variables to cover any kind of instance they can possibly be?
No, a given entity can pick and choose which components it has -- that's the whole point.
>Because I very much doubt you can really have a "particular death animation" all sorted out with just a single value.
Why not? Sounds perfectly reasonable to me.
Again, why are you getting this involved in a topic you clearly are not at all familiar with? Both specifically ECS, and game programming in general?
Make some sort of game where you're trying to escape from the experimentation cells of an evil corporation, like the Umbrella Corporation. Maybe you fight creepy stuff, maybe you run from creepy stuff
>Why not? Sounds perfectly reasonable to me.
Yeah, if your engine does it for you, if you actually programmed stuff, the reason why that is stupid should be self-evident.
>Again, why are you getting this involved in a topic you clearly are not at all familiar with?
I am not familiar with engines, sure. But you're giving erronous advice to a guy who is not sharing the same circumstances such as you, that much is evident.
What's reference studies? All my life it comes from the top of my head.
My dad's blood. He used to be an art teacher. Too bad he was white and had to peace out early.
I've never traced or look at another image and copy.
I'd like to do something quick based on something someone tells me to do here. Also that pic was first time coloring.
I can't program for ass.
Guy from last thread here,
Finally got around to making the stage "endless." Now to make a generation algorithm.
I made variable A keep track of the player's x position and a variable B to keep control of the spawning.
Everytime A is bigger than B there's a chance to spawn a block (and other things) and then B is increased by 16. I thought this would translate in things only appearing in multiples of 16, like a grid. But it seems sometimes shit just plain spawns on top of each other.
What do? Webm related.
I'm not sure what you mean friend. I was just responding to the idea that a system has to have many conditional statements to check for different types of enemies and decide what to do for each type, which goes against all of my understanding of how ECS works. (Which is admittedly limited.)
What are you even talking about? The particulars of the animation system are irrelevant here. You're not even saying anything to back up your argument, all I hear is "nuh uh" and lots of assumptions.
This >>128189872 but less inflammatory.
It is clear that you are at the point where you are either going to sink or swim. It entirely depends on if you do reference studies or not.
You take an image and you reproduce it is accurately as possible. This is, reproduce a photograph, not someone elses art/cartoons. Work it until you 'feel' it is accurate or until you have been working on it too long. Then do the next and so on forever.
I'd like to remind everyone of Pandora's Tower, and more importantly the fact that it shows that you can put a waifu that you have to romance and care for in literally any kind of game.
I'm talking in the context of the other conversation about the evils of inheretance. Obviously, it's not an issue you have to delve with if you're using an engine, you can have an n number of types and define particular behaviours for each without worries, it's cool.
>The particulars of the animation system are irrelevant here
It is if you're not using a fucking engine, or an engine that is not as complete as yours! Do you understand why your advice is shortsighted now? The particulars of how an animation is called are very important, that's the whole point of encapsulation!
Not to be a dick, but unless you show me your art, I am pretty sure your art is utter shit too but you say you are an artist because you are too lazy to learn how to do anything.
Well, my experience went like this.
>Implement dialogue system from scripts
>Implement battle system (units, skills, buffs, turnorder, formulas, costs, etc)
>Implement ingame animation system for skills
>Make the units
>Make the skills
>Make graphics for units/skills
>Design enemy groups
>Shit what have I been doing for the past 4 years part time?
It's just a lot of work unless your scope is incredibly small.
I'm not using an engine personally, just using Monogame. But how I understand it is you don't actually define particular behaviors with ECS -- you define one behavior, and then let the data of the components drive the outcome as that behavior (the system) works on it.
>post really important stuff
>attempt to shitpost
How the animation actually gets played isn't the component's business, you're right. For the case of the death animation, the component just has a reference to a particular animation (for example, though a file name).
I added checking if the place is free, The blocks do have "solid" enabled. Apparently they don't give a shit.
Yes, I understand that much. But not every behaviour is simply given by a bunch of variables.
Obviously, if you have a simple collision system that only requests a threshold to perform calculations, then a struct that holds a threshold is going to be enough.
But anything more particular, especially animation, is going to be kind of shit, unless you're implementing such behaviour on neat little capsules.
>when u get into the deving groove
should i make a team of fellow aggy daggers?
i feel like i could lead the project even though i have no experience and mostly because 90% of the problems that occurs are because of jewery and im pretty non stingy with money
You realize the blocks are large right, like larger than a pixel?
You are only checking if a single pixel is free, this is why you get blocks that overlap each other, but funny enough, none that is completely over another.
This is the part where you feel like a dumbass
Can anyone show me some good one person dev teams that have had success at making a good looking 3D games? Doesn't have to be from this general. Just any indie devs that have successfully made, or are well into the process of making a 3D game of any kind. Preferably something that isn't overly simplistic like a first person maze or something.
>should i make a team of fellow aggy daggers?
You're gunna have a bad time. Remote dev is terrible.
Yeah that's easily possible in Godot. Or literally any 3D engine.
As far as the rest of Godot's 3D capabilities, it's not amazing, but it can definitely make gamecube tier games. You just wont be able to do any modern crazy shit, if you want modern you'll want to use Unreal Engine 4.
While you are absolutely right and I should be checking with rectangles, it doesn't help the (I'd assume logic) problem of not being a 16 pixel grid... why are things not happening every 16 pixels?
You can only take advantage of that fact if your language is truly homiconic, otherwise data and code are going to be represented in really different ways. Thanks for butting in I guess?
Well I'm not going to try and claim that there's some magic one size fits all approach to game programming, so it's more than possible that some intentions are better served using something other than ECS, but I think that most, if not all, behavioral needs of entities can be broken down into "just a bunch of variables". I'm not sure why animation would be different. At its core, at least for 2D, it's just image and timing data.
>Amount of time to make a decent quality 3D rigged and animated model
>A few weeks
>Amount of time to make a decent quality 2D animated sprite
>Three days or less depending on sheet size and quality
In b4 all my enemies are humanoid, I reuse the base mesh/skeleton and their only difference is gear and hair that is put on them via sockets.
>behavioral needs of entities can be broken down into "just a bunch of variables"
Pragmatically, no, that's not a sensible assumption.
Depends on the langauge. You could really have lists that represented game constructs with the code embedded inside, and have those evaluated at run-time, but that's only on LISP-like languages. Yeah, that approach would work there just fine, in fact, it'd be a neato exercise for anyone curious about it.
>otherwise data and code are going to be represented in really different ways
In every language I've used so far a function call is just a pointer or object which you can pass around freely, and argument lists end up just being arrays.
Sure, it may be a pain to use in some languages to the point of going against the entire paradigm.
>Tfw attempting to use member function pointers in C++ without a wrapper library
>Tfw using Java's misc.sun.unsafe features
This is what I was afraid of. Yet, you see plenty of 2D games made from single devs. Tons even. For some reason, everywhere you look though, people make it seem like a possibility. Like,
I see plenty of 3D games in dev sites like https://itch.io/jam/agdg-demo-day-6, but they never make it out of the demo stage?
I just want to be able to use my 3D models in my game. It isn't going to be first person with crazy awesome textures or anything.
just got into digital
trying shit out tonight
>he isn't using a real language to make his game
I will say as a penciler, you are good enough to not be utter shit. Sorry I said that. I encounter too many in the game dev program who talk like you do but literally are shit at everything.
>never made a game with a persistent world or saving of any kind
>never made a game with hand made puzzles
>never made a game with interactive objects such as buttons
>never made a game with AI that is more complex than walking towards or away from the player
>never made a game with an inventory
I- I don't know if I can do the Zelda jam.
In your OOP based approach, each object wouldn't have code for running its particular animation. That would be abstracted away into something that you feed the current time into and get the current frame, for example, and the object just tells it to switch between different animations.
Why would it be done any differently with ECS? You could just have a single value somewhere that says which animation the entity is currently on (which can be changed by other systems based on what the entity is doing), and when it's rendered, you look up which frame to use based on the time.
off the top of my head
the original Stanley Parable I think?
william chyr's game
Fez is 3d and practically speaking phil fish was the only stable dev
if you count it early minecraft
my 2d animation so far
Anim = stateful, has a time_left (number) and an on_complete (event, two strings + args) + frame data.
LoopAnim = stateless, just quad (4 nums) + timing (number) per frame. The actual clock for these is global. Might add an offset later but for now idgaf
this isn't counting tweening which is handled by a library/particles which are handled by the engine
My game doesn't store opaque datatypes or functions as state. Any time I need to change behavior based on state I use events/class references(literally just the class name, string) instead.
This is in a simple dynamic language, without metaprogramming, and it works just fine.
>In every language I've used so far a function call is just a pointer or object which you can pass around freely
Try and modify the behaviour of said object at run-time, man.
It's true, when it's bare metal it all looks the same, but we're talking high-level languages here.
I didn't mean the behavior itself. Just the needs. What I was trying to say is that the systems are designed generically and since what really drives the behavior is the data that generic behavior acts on, it shouldn't matter if it's simple, sparse data like collision information, or if it's more complex and numerous data like you'd need for animation. It's possible for the systems in an ECS to act on data supplied by the components without regard for special circumstances or checking for "types" of entities. That's all I'm saying -- is that I don't think (at least I've run into anyway) there's a case where an animation is going to be so complex that the data couldn't just be fed into the same generic animation system as all other entities with animation data.
Stop being pedantic. When people are comparing OOP and ECS, they are comparing entities built through inheritance vs. entities built through composition (ECS being a very specific formulation of components).
Well to be honest, some of it just is innate. You're either good enough at it or not. You just have to fine tune the skills you have been provided. Just draw shit. Reference anything and everything. Try to draw the stuff you hate drawing. I notice a lack of lower body in your images, just mostly heads.
I tried google and barely came up with anything. Which can only mean one thing.
I wonder if it's because there are devs who are thinking of breaking into 3D with the release of unity and unreal, but haven't released anything yet. Or if it's because it's too complicated to realistically happen unless the graphics are very simplistic.
Okay! thanks for that
I'm gonna whip up something right now,
but actually for the advice
I would like you to tell me what to draw
ninja turtles etc idc
I'll draw anything you ask
The thing is, releasing something without giving a fuck about the graphics is pretty risky.
Sure, some people got away with it but it's mostly because they usually got some good marketing, shilling tactics and also got lucky.
Anyone here can make your average steam, mobile game with shitty graphics.
Doesn't really matter much now, but I've updated the browser version, so that it runs much better now and with the full level.
Turns out that I was underestimating the speed of polygon booleans.
You could've just:
Raph is consumed with guilt, blaming himself and Mikey together for allowing the family tragedy to happen. He moves out of the family lair (leaving Donnie to live alone) and finds a new sewer lair where he can live and train alone and drown his sorrows in new vices of alcoholism, chain smoking and cutting himself.
Well I think one major thing is 3D takes a bit longer to get used to than 2D. And if a programmer can barely make passable 2D art they most likely wont make anything good in 3D either. This is why so many 3D games have a minimum of two members, a 3D artist and a programmer.
Level design is also harder in 3D than it is in 2D. Although it's not too much harder if you're keeping the game logic in 2D.
There's also a huge number of 2D games simply because there's always a huge number of beginners, and beginners almost always start with 2D, and just keep 2D going for a while.
But if you're already good at 3D and you're already comfortable with programming then you should be fine.
>donnie finally has time alone to do machines
What are some ridiculous/outlandish weapons for a MC?
>dying before you even have an engine
You just described one of the most iconic adventure game characters of all time
spoiler: NES Link and the fact that the master sword is made of a bunch of other swords before it.
Currently trying to re-make ramps so that there are some slopes you can't climb on until they're at a certain angle, but in the process turned them into giant tractor beams.
>Fez is 3d and practically speaking phil fish was the only stable dev
Phil was the artist. He didn't even know how to program he had to hire another guy who was the full-time programmer for a number of years so no. He doesn't count as a 1MA he's just another front man. Jon Blow qualifies more as a 1MA than Phil.
An artist at sucking cocks maybe.
Another approach from scratch.
But sadly enough, not what I expected. I expected a block every 16 pixels.
Why the fuck is this not working? getting stuck on dumb shit only makes me feel like I'll just plain give up when a real inexplicable problem shows up.
>so far 2 teams ive been in were all excited for a day or two and then dipped
Welcome to the average internet project.
Takes a lot of searching to find someone that can and will do professional quality work for free.
I'm going to be spending a lot of time on graphics. I've always been a perfectionist when it comes to graphic design. I'm more worried about coding, shaders, animations and rigging, etc that come along with 3D games. I'm still a complete beginner to game engines and coding, so I can't figure out if making the game I want to make is even a realistic en-devour or not, but it's not looking promising with the lack of actual 3D releases by one man dev teams. One thing is for sure, I'll be finding someone to help with the music, because I want someone who has good music sense and it isn't me. Everything else, I want to handle personally.
I've been making 2D art for a very long time, and while I'm not the best by any means, I am able to do simple things like character prototyping, character portraits and textures for all kinds of landscapes and such. I'm still pretty new to 3D modeling, but I have a basic grasp of how everything works at least with blender. I can model most basic things at this point, and I know the UI, which commands do what, mods, and a few other things pretty well.
The game is going to be 3D, but in a top down style with everything zoomed out a bit, so having super high quality texture work isn't a big deal. I'll want some cool lighting and shadows and such, but that's completely outside of my knowledge at this point.
Programming I am not exactly comfortable with. I'm decently competent at learning, even if a little slowly at times, but I think I'll be able to get a grasp on it eventually. I'm just worried it will set the bar too high by starting coding on a 3D game instead of 2D or 2.5D.
We're all in it for free at first
they all think money
but i think growth
Serious for what? I am.
I felt more motivated when I worked as part of a team, I wanted to have a couple of days where we just talk about games and things in general
but they wanted to jump straight into making games and discussing splitting money
like fuck man
It's a long process to be honest. I have had multiple stages where I thought my art was good, only to realize later that it was really pretty bad. I'm finally at a point where I can make something that is good by most standards with a little time and effort, but the process was long and took a few years of work. Some of that time was spent as a neet, doing nothing but drawing and reading books and guides. A few months or so. It helps if you have a passion to do it. If you don't have the passion or don't want to put in the time, you should try for a simple style as there are plenty that look very appealing with little work. When I say simple, I mean simple though. The advice I'd give you is spend a lot of time with people who are decent at the type of art you want to do. Learn perspective and how it works. Learn values and proportions. Do a lot of work, don't stick to a single image for too long starting out. Don't get overly attached to any starter images you make, move on. Don't be afraid to fail. Good luck.
been trying to setup SDL in xcode all day. That shit just wont work. Trying to get it to work in winxp now (it's my primary VM). Can someone suggest an IDE (that isn't VS) that is easy to set this shit up in? thank you anons
>FL Studio demo actually comes with MORE features than the 100$ version
>you have to pay 200$ for the option to save your music
Holy shit. That is the most jewish thing I have ever heard. What do you fuckers use to make your own music with, assuming any of you do?
Finally got terrain generation to work.
Added animation for shooting while running.
Nothing bad to report, just figured any progress post is a good post.
This "Game" is pretty much a decent base for a gimmick. Was experimenting with those numbered blocks but if you guys got any other thing you think could work let me know.
>torn between making pixel art and drawing my stuff in photoshop with a brush tool
>can't even tell which kind of art I'm less shitty at
If I make a level in Unity with a first person perspective, ignoring textures and model look, is the camera the only thing that I need to change to make it into a top down game, or is it more in depth than that? For instance if I want to design a game in FP view, then simply change the camera from the player PoV to a high up camera, is that an easy task?
it is that easy
you could get more advance with it and like have it change the controls for the 2nd camera
i think the way you would want to do it is having two cameras in the prefab and disable one when its not active
place_free() doesn't check a single pixel area. It moves the calling object by the x and y arguments, checks the object for collisions, then moves the calling object back to its original position.
When you reference player.x, you should instead use something like floor(player.x/16) * 16. There's probably a more elegant way to get it on the grid but I can't think of one right now.
I'm just afraid my dumb ass questions do not warrant a post. Too basic level. This is the collision handling part of my jump code
Yeah that's exactly the way I ended up fixing my problem.
Is the point of those two nested loops to lock the character to the collision block so that there isn't a gap? If so, you might want to do y += sign(1) rather than y += sign(vsp).
Actually, now that I think about it, there are a couple reasons why this code doesn't make sense. The while() loop directly contradicts the if() loop of which it is nested inside. There's no way the code inside the while() loop could possibly run. Unless I'm misunderstanding the sign() function.
I did take it from spalding's videos. It's not that I didn't look for alternatives, but that seemed like the shortest approach at the time.
I would be lying if I said I knew exactly about the nesting, but its vsp rather than 1 to account for the speed build-up of gravity. I don't think it has to do with the rectangle itself.
I wouldn't mind just still having a rectangle while ignoring the arms for collisions. But I can't just go and add an integer to the x position on the meeting place since the same line has to be used for both directions.
attached gif is just what I did while awaiting responses, crouching and shooting, running and shooting.
i think you can add a jumping sprite and also a next to wall sprite which will have smaller collision boxes. you don't want an arm just overlapping the wall next to it.
as a temp thing just copy the standing sprite but change the collision mask on them
>But I can't just go and add an integer to the x position on the meeting place since the same line has to be used for both directions.
Don't you have a -1,0,1 ranged variable somewhere in there for handling direction? You multiply it by that.
>banned for a day because of bantz
Truly, these mods are worse than Hitler.
Guys I want to make an autismgame, any ideaguys got anything for me?
There are some generals where I get banned for the slightest hint of rudeness to the point that I've just stopped posting in them to avoid the constant 3-day bans and other generals where I've relentlessly shitposted for years and never even been warned
>Does the while loop method solve that issue?
Yes, because you are checking in intervals the same size or smaller than your smallest possible colliders (in this case 1 pixel). If you can say "I am never going to use colliders smaller than 16 pixels" then you can step up to 16 pixels per iteration. Doing it pixel by pixel is somewhat handier because you don't have to calculate the exact intersection point afterwards. The drawback of this method is it is more performance intensive, particularly if you are moving something very large distances a single frame (1000 pixels becomes 1000 iterations!). Most of the time this will be fine though unless you have tons of things moving each frame.
I'm here fuckers , but I'm kinda chilling because the autism squad couldn't chill their tits when I posted yesterday
Now that I'm going back to my collision code I've actually got a more long winded mess of a workaround that includes acceleration and booleans, which is probably why I never ran into the limitations of my technique.
I think I was thinking that having a while loop would make the sprite appear to teleport, but thinking about it I'm wrong, aren't I? Because this situation only comes up when the gap is smaller than the regular speed anyway.
Just want to clear up what autismgame means. I mean a simulation or strategy game like dwarf fortress or civ that involves autistic people trying to optimize, solve or just put a lot of energy into having a lot of fun. (While also being somewhat accessible to "gamers" too). Currently thinking of making some sort of spaceship management game, but I'm still open to ideas if something more interesting comes up.
>I think I was thinking that having a while loop would make the sprite appear to teleport
No the sprite should only draw after the whole process is done and it should only end up making a difference in situations where they would have jumped past the obstacle they will instead butt up against it.
PROTIP: most tutorials out there for acceleration in game dev is wrong. To do it correctly you take the instantaneous velocity of this frame and the last frame, average them, and use that average. This only works with linear acceleration, though. After linear it gets really complicated so I would just recommend using a physics library at that point.
>PROTIP: most tutorials out there for acceleration in game dev is wrong. To do it correctly you take the instantaneous velocity of this frame and the last frame, average them, and use that average. This only works with linear acceleration, though. After linear it gets really complicated so I would just recommend using a physics library at that point.
thats useful info, thx anon
bet you're some fuckhead who's constantly shitting up the thread with low quality "I got you to reply at all" tier trolling, with no attempt at even being slightly funny.
I'm not even slightly mad, though, just sad to see the threads in such a sorry state.
Add keymapping you fuck, it's not even that hard.
I was quite butthurt when everyone cried, forcing me to add controller/mouse support to my game. In the end it was much better for it.
retarded gml question: im messing with 8 way movement and i normalize the player's horizontal/vertical speed when they move diagonally. for some reason i can move diagonally into the corner of my walls. im using The Patented Shaun Spaulding method of collision detection. is this happening because normalizing the player's speed while moving diagonally produces subpixel positions and if so, should i just use round() or floor() or whatever on the player's position every frame?
last one of the night
im so happy
i found a programmer yaaaay
>he didn't even show me his shit
>but its all about growing together and learning as we go
>and not giving up
If he hasn't finished anything before, and nor have you, the chances of you guys giving up is extremely close to 100%
I noticed /v/ had a gamedev thread earlier, but it hit the limit and I couldn't find a new one so I just came back to this thread. It was a very interesting thread, and had more stuff posted than this thread. I guess I'll check back there again.
What the hell did you make? Just what the hell did you do? Think about what you're doing before you bring ungodly things into this world that you don't even understand. You must repent, before it is too late.
List of AGDG steam games:
Oh that's not what you meant. Posting anyhow!
PS can you even animate? Because that shit looks hard to animate, but would be pretty decent if you could animate it well despite looking a little janky.
I know I'm late, but I feel the autistic need to reply anyways.
I don't got free time until tuesday night, and I have none at all on friday or thursday and I don't know unity's shader language, only glsl.
Well, if you're going to be the artist, get better at animating. It's not really -that- hard.
Delicious animations can easily make your art amazing, even with a kind of mediocre base.
>tfw you realize you've been wasting hours on AGDG, but it's actually fine since you already dropped a solid 7 hours of work on your game today
realistically speaking, how hard is it to make a job out of gamedev? if someone makes a game, it gets greenlit, and it sells 100 copies on steam for 10 bucks each, how much of that $1,000 do you actually ever get? i think steam takes 30%, then you have to pay the engine fee if you use unity or UE and make a certain amount, right? then taxes? i see a lot of games on steam that have very involved mechanics and unique art styles with very positive reviews that barely sold any copies. is the dream of being an indie dev who makes games as their actual job a reality, or will it most likely always be a hobby even if you make something awesome?
well, this game made $300,000 in steam sales, minus the 30% from steam, and then whatever taxes or fees came out to be, didn't it? it does look pretty cool, but that sounds like enough to live off a couple years for dev'ing a new game right? or am i missing something
In the game, I think what you're actually going to be *doing*, is finding a number of prism shards scattered through the world, and returning them to the mother prism. Returning them all and making the mother prism complete is how you win.
But, while you're holding each shard, it floats around you and follows you, making you more powerful. When an enemy hits you, it makes you drop the shard. Certain kinds of enemies can pick them up themselves, making them more powerful. In fact, most of the shards will start off in the hands of enemies. When you die, the shard lies on the ground where you dropped it.
What are your thoughts on this?
Messing about just trying to find an art style for my game
I quite like this but im worried the palette will get too limiting
Maybe if different worlds have different pallets or something?
I gotta start fixing some bugs and stop doing the fun stuff soon
haha i forgot to post gif
then i made it too big
oh gawd what is happening this morning
this doesnt show anything
why did i make another gif
why did i post it
Mobile online crusader kings where every player is on the same map
When you start playing, you're given a random role depending on what's available: If you were the first player, you'll become the king, other titles will be handed out, etc. But most people will start off as ordinary citizens.
As with CK2, you have to continue your legacy and aim to become king. The gameplay would be fairly similar, except now you will most likely start far lower on the food chain. While the king manages the kingdom and the counts and dukes manage their titles, player citizens produce resources for the kingdom and make up the armies being sent to fight. It would take far too long to explain how you'd go from citizen to king, it's just all about making the right connections and marrying into the right families.
If you die, you start over with a new life.
Unlikely that it'd work real-time, you'd probably have limited actions per day (not just to jew the player by getting them to pay for extra actions, but because I can't think of how else it'd work).
You'd need a lot of players for this, but you can probably make some smart AI that gets replaced over time as more players join.
~~~* RECAP MONDAY *~~~
I tried to stop being a retard temporarily
Downloaded some software to make webms
>shoulderon gaiden off hiatus
Holy shit I was doing music for that and slowly stopped hearing news of it I'M STILL HERE SENPAI NOTICE ME
+Trying out some post effects
+Started putting walls up for houses,
+Swapped some textures around again
+Started work on a simple animation pipeline
senpai dont fall for those retards, you don't have to convert everything to webm everytime, the gif was perfectly fine and converting to webm is irrelevant if you don't record especifically for that
if someone cries about filesize laugh at their datacap
im 100% adamant that what will make or break you is PR. Basically in order of importance
1. Word of mouth so people know your games exist
2.Good /unique artstyle that reels them in
3. Solid gameplay that keeps them hooked
lets post some art a programmer could probably do eh?
sometimes I feel bad for frustrated programmers ;-;
No a programmer could absolutely 100% not do that. That shit has excellent color work and shapework.
Just because something looks simple doesn't mean it was simple to be able to make
+setting up networking
~will still make assets in exchange of code and a bit of guidance on how to apply said code
feel free to use this as recap image again, it's depressingly accurate
Programmers could do a damn lot more art if they tried to use a technical illustration approach and technique, but they go to /ic/ or some other lair of concept artist wannabes and get put on the road to full blown fine art faggotry instead.
It's a grand conspiracy to ensure that established artists aren't displaced once everyone else figures out that artists are just as obsolete as musicguys.
You want to add a layer of challenge and difficulty to the game that will exclusively affect bad players. Think about this for a minute and maybe go back to the drawing board.
I'll throw out an idea, though, instead if you could voluntarily throw your crystals shards at enemies to send them into some kind of persistent buffed out frenzy state, but you get a much bigger reward for killing them in this state in addition to having your shard refunded? This could make for some really cool displays of skill where you're sharding a bunch of enemies and watching them rush at you, potentially in a state where they can kill you quickly, only to handily dispatch them and be showered in delicious sparkles.
If you're trying to take notes from the game it looks like you're inspired by, then you ought to know that those kinds of players love it when you give them options to organically adjust the difficulty for themselves.
any unity game devs want to give me links or names for some great starting tutorials? like something that will walk me through making a simple level in a series kind of thing geared for total beginners? i got the catlikecoding link earlier, but it doesn't seem to have an order for tutorials and i was wondering if there is anything else i should check out. i don't mind if it's a udemy course that costs a little, just as long as it's good and comprehensive.
tell me why then. the textures ones are all complex and squigly while i can just model it and it will be a rectangle shaped. with the map it will draw a ton of verts because the texture isnt just a straight square
im not trying to argue im honestly wondering how its not more. also i cant find any textures of a good sidewalk
So when you die to a monster(meaning that the monster is already pretty hard, or it wouldn't kill you) you grow weaker and the monster can easily kill you again, and the monster grow stronger so it can trivially kill you again.
Do you see the problem here?
thank you and tl;dr a sort of Prop Hunt only with considerably deeper mechanics
and eventually procedural placement of items at the start of each round if I ever get past player controllers and networking; it's looked really similar for the past month because I'm struggling my ass off with the coding aspect also job took most of my free time during the holidays
because fake normals need good shading to popout, if you could manually raise the bump by controlling raycast areas or paint them yourself,but its so much harder on flat surface where light only bounce once/twice
>that's an orangutan
Which is an ape.
It saw a human using the tools so it decided to do the same. I mean clearly if a smartass like a human uses the tool there's a great reason to do it too.
dev: Ace Combat Clone Dev
tools: UE4, Blender
+Better flight physics
+Overall code structure more cohesive
-Weapons haven't been transferred
-UI haven't been transferred
dev: Meneer Guus
progress: - Bonus room added
- Updated graphics
- Fixed lotsa bugs
- More juice
Thanks for playtesting agdg