Amateur Gamedev General ~ /agdg/ + /vm/

reading my post again it's pretty useless, so sorry about that. I can't help you

all i can say from my limited experience with UE4, is that if you're working in blueprints, it wont let you do anything serious with the engine, i think you should be able to look at the scripts directly to see what level streaming is doing.
Also usually in their livestreams on youtube they go into more information on certain aspects of the engine.

youtube.com/watch?v=GkDg9GPpzXE

May be the kind of thing you're after.

Fuck blueprints with a shovel. I tried being lazy for a few days, hoping that I wouldn't have to learn UE4's coding standards but I lost my cool very quickly. Some people say blueprints are meant to be a prototyping tool, but even as that they're fucking useless. A short learning period of like a couple days, and I feel like I'm fucking superhuman compared to dicking around with blueprints making me feel like a child with a fucking playset. The speed difference is easily threefold.

Thanks for the video. I'm gonna look over it, but I assume like any other official documentation it talks about level streaming in context of a linear, premade game. I'm working on an autistic voxel type of deal hence why I was asking if anyone knows of a technical explanation of how exactly levels work.

If you're doing your own autistic voxel stuff, your skills are far beyond mine at the moment.

Whats stopping you from making your chunks load and unload with your own code?

Seems pretty simple to just base it on player location in some way and load / unload in a radius, or if your chunks are huge just have them load when entering / leaving a chunk with enough distance to make it a smooth transition.

As I said, I just don't want to reinvent a bicycle. I like to use the default toolset where possible. I was simply wondering if there was something that would let me easily break the world down into cuboids and serialize them as "levels" that I can stream in using the default functionality.

The hundred autists at Epic surely can code better than I can. So if they coded something I can use instead of my own shit - I'll take it.

i think you might be retarded, i'm building literally my whole game with blueprints.

Are you building Pong? Because blueprints become much more of a hindrance than a convenience once you're doing anything more complex than that. You can't even have two dimensional arrays in BP.

It's like progress whiplash, fuck me.

This prototype was made with 99% blueprints. The recenter release of Unreal allows for Maps which is an X by 2 dimension array. Still kinda sucks that you can make an X by Y dimension array yet.

Made some good progress.
+ support multiple callbacks upon unregistering/despawning entities
+ reward player callback
+ leveling up players
+ stat recalculation

next up generating enemies more procedurally and determining rewards based on their stats, inventory, and spending stat points or something

Looking for some advice here.

I'm a designer who builds his own creatures and just in general is involved in the creative process. I'm good at world building and have a good eye for how to handle these sort of things.

I want to put together a series of articles or videos discussing various topics relating to game design. An example being how enemy types in an area can tell their own story or harm the story a location is trying to tell if used poorly.

What's my best avenue in getting these out there? I've made attempts at making a game previously but realized that my strength isn't in coding, I lack the autism on that front but I excel in other areas I think would be useful to game devs and I haven't seen discussed before.

gamasutra

Coding can be a strength if you work hard enough. Don't buy into the modern western culture meme of being only a creative or scientific person. If that were true, why was the Renaissance era filled with men who were great artists as well as architects, inventors, etc? Because they worked for it.
For the other part of your post, YouTube videos, Mark Brown, Charlie Cade and other such personalities make a living off of game design videos. Vid related is pretty good.
Use social media to get them out there. Despite all the shit we give it on this board, reddit is good for when you have no subscribers and need to get enough views for the algorithm to pick up your video and start recommending it to people.

Question: I have a background in HTML/CSS but nothing else. I will have a lot of time this year to learn to program a game however I will have very limited access to internet over the next year and the internet I will have will be 56kbps slow.

As a result I will buy an external hdd or two with tutorials and assets already downloaded.

Which engine is more efficient with regards to file size and ease of learning?

Are there any books or wiki's I can download to help with troubleshooting when I'm offline? I guess it would be C++ and Java, right?

I'm really only famliar with Unity and UE4. I've never hear of the other engines in the beginners guide. Is one better for a CRPG/Isometric style game than the other?

Just make quality content, and then whore it out on the usual social media sites. Create a blog or a webpage (host your own or use wordpress) to archive and post links to your videos, and host your vids on youtube. Youtube is going to be your only option for video hosting, and if you want your videos to be seen with any amount of visibility you're going to have to know how to game Youtube's search algorithms. Otherwise your videos will not be seen, regardless of the quality of your content. If you need ideas on how you should present your content, check out analytical vids like suggest. I'd also recommend taking a look at other analytical channels that don't talk about vidya, like Every Frame a Painting, Super Eyepatch Wolf, or even Digibro. The production quality on these videos are really good, and these youtubers generally know how to write good scripts that are engaging and have clear, understandable, writing.

Marry godot-chan for 2D you won't regret it.

Why is this? What the hell gives?

Unity comes with its entire manual + reference as a local website (and its documentation is spectacular for the most part), and if you get Visual Studio with it, you'll have most of the official C# reference locally as well. You could probably archive DotNetPearls or tutorialspoint for C# reference as well. There's also plenty of videos you can download straight off of the official tutorial site for archival.
Unreal is like 99% online support from forums/wikis from my experience. There may be more videos/references and stuff, but I'd suggest avoiding it if you have limited access to the internet while working.
Same with C++/Java, there is lots of stuff you'll want to search for that doesn't get covered in documentation that comes with whatever resources you get.
You could probably download the official Java API JavaDocs and tutorials like above.
The JDK comes with a src.zip containing all of the standard library java code (anything that is not native), but it may be difficult to read if you're new to java.
I don't know if similar documentation is available for anything in C++, but either way, you will also need to find and download documentation for whatever graphics and other libraries you want to use.

These days most engines are good for any type of game given you build all of your own game logic yourself and just use the engine for rendering/collisions/asset management/etc. I'd suggest using a few different game engines before deciding on one or trying to write your own. They all have different strengths and weaknessses

I like that picture

How's Godot for networking? I've had some success with GameMaker's, but I hear that it's shit for anything ambitious. I don't know enough to judge that on my own, though.

Kill yourself, Digibrony is fucking terrible.

Hi everyone. I'm still alive.
That's all of the Slayer's primary subweapons done, now I gotta make a few alternate subweapons for his throwing daggers.
Working on these new pickups is such a pain in the fucking ass.

Thank you friend :^)

For videos? Only YouTube, there's not much alternative. You could use something else, but even less people would see it, sadly
For articles, you have more options. It's easy to start a wordpress blog, or you can use Medium (at the expense of being in the company of insane SJWs)
There's almost no point in speaking if you aren't being heard, so you might need to shill your shit elsewhere. It helps to know a guy, or have a twitter account

Articles are easy to make, but hard to make good. I recommend starting here, just write a bunch of shit for experience. Each article should have only one point, and each paragraph should provide reasons for/against that point. Proofread by taking out everything that isn't necessary.
Videos are hard to make, and a real video analysis should have a well-written script and good audio quality. Said script is just an article, adjusted to sound good when spoken. Don't worry about buying expensive equipment, and have patience when editing or doing retakes
Whatever you do, don't ramble and go off-script, in videos or articles. Otherwise, you'll end up with a 30 minute video/read that has 10 minutes worth of content. Every sentence is deliberate, so treat them as such

And what kind of programming/game dev have you tried? There are plenty of options for all types of people, maybe you just haven't found the right language/engine/tutorial yet.

Markup languages are not like programming languages, so forget everything you know.
Engines (all of them) are hard to learn without tutorials, be it videos or troubleshooting online or reading the docs. My recommendation would be to forget about gamedev and learn a language first. Python is my go-to for babby's first, plus it has pygame for making simple games all by yourself. Python's docs can be downloaded, and they're usually helpful. I can't find a download link for pygame's docs, but I'm sure you can scrape them somehow.
As for books, look around libgen and download as much as you can.

The more complex the game or engine is, the more work you have to do, which means the more problems you will run into. Starting small is the way to go. Just imagine you're living in the 80s and you gotta make the next big CRPG

just finished this
going to learn how to rig now I guess

Never watched his brony channel, but his writing is pretty damn good. He's got an ego the size of the moon, he'll almost never admit he's wrong, and he'll shit on people for shit he's guilty of doing himself, but he knows how to put together a good script. Even if someone doesn't personally like him I feel like they could learn some do's and don't by watching a bit of his content.

Don't start with an engine if you can't program in the first place. You won't be able to be productive in an engine without real programming skills. I would recommend practicing programming until you can write a non-trivial program (over 1000 lines to do something useful).

These are some resources for learning to program in C:
tutorialspoint.com/cprogramming/c_overview.htm
learn-c.org/en/Welcome
w3resource.com/c-programming-exercises/
cs.unm.edu/~fastuous/spring14/cs241/C Programming Language - 2nd Edition.pdf
icube-icps.unistra.fr/img_auth.php/d/db/ModernC.pdf

And this is a really good C IDE that you can use:
pellesc.de/index.php?page=download&lang=en&version=8.00

Slow progress but still managing. The weapon now "shoots" with a raycast that extends from the center of the camera to where the player is aiming. I added sounds that you cannot hear because Kazam don't want to record my sounds. My next step is to identify what *exactly* did i hit: Is it an Area trigger? a door? an ennemy? all that bullshit.

Also yes, the particles and animations are awful.

Use OBS to record, my nibba

C is a really nice language, except it basically forces you to use GNU/Linux, which is also nice but usually unfriendly to beginners. Troubleshooting vague linux errors on dialup is not something I'd wish on my worst enemies.

My autism says you should use projectiles instead of hitscan

Not true. Plus C is needed for the Windows API.

C is loved in GNULand, but you can program C just as well on any OS. Pelles C makes programming C on Windows pretty enjoyable. I think you can technically still compile .c files with Visual Studio, but you would have to create the file outside of the IDE and then import them.

I've watched a bit of Digibrony's non-pony, I just refuse to acknowledge the fag's new name content and it's painfully bad. At best he tells you basic shit any amateur filmfag would notice immediately and explain better, and at worst he's a lazy idiot with shit opinions and shit taste.

he's a lazy, idiotic pothead and is a good argument on why marijuana should stay illegal. I strongly advise against anyone listening to him, and there's many other youtubers that cover similar topics he does much more competently.

10/10

...

What are you talking about? I find C easier to write on Windows. Also the IDE that I linked too only works on Windows. C being easier or even better on Linux is just a meme.

You can compile C in Visual Studio as long as you don't use features outside of their implementation. MSVC only implements C90 fully and has an incomplete implementation of C99, that means that certain programs such as #include int main(void){ char *str = "the quick brown fox jumps over the lazy dog"; size_t sl = strlen(str); char tmp[sl+1]; for(int i = 0; i < sl; i++){ tmp[i] = str[(sl - 1) - i]; } tmp[sl] = '\0'; puts(tmp); return 0;}
Will not work in MSVC because it uses C99 features outside of their implementation (tmp is a VLA), but Pelles C can compile it fine. Pelles C actually implements the full C11 standard which gcc doesn't implement, so you can use threads.h in Pelles C while on Linux glibc does not implement threads.h so you have to use Musl as your standard library to use the full C11 standard. Right now my project doesn't work in MSVC because I used some more modern type checking features in the newer C standards, so it doesn't compile.

Oh fine.

I just repeat what I've heard, not surprised that makes me wrong.
Never looked back after I got gcc working on linux. I'll check out Pelles next time I'm on Windows.

Best laugh I've had all day.

what kinda cashgrab can I make in a weekend?

a card game, but each card is a game

A FNAF clone.

A post modern game where you're a geometric figure that floats from point A to point B.

nothing. Because you have no talent, no knowledge of technical aspects, no motivation to learn or try hard, no desire to actually make a decent game nor any ability to do so.
In short you're just a shitter without any personality and a shallow person that can go fuck himself.

I want for my bullets to behave like in source engine games where they instantly travel like a laser. Not like PUBGAY, (which could be in the godot engine's case an instanciated scene that will go to where the player is aiming.) Raycasting should give me what i want.

Not giving you my cashgrab idea thats for sure.

Law of leaky abstractions m8. The purpose of an engine is to increase productivity. It cannot replace learning. This is why math teachers always make you do arithmetic by hand, even though there are calculators. That other user is 100% doing the right thing by researching how his tools actually work before attempting to use them.

A meta joke game. Like something that's only about opening lootboxes or dlc. Doesn't matter if it doesn't have gameplay. That's just part of the message.

I would disagree- the reason that you use something higher level is to learn less about things. This is why languages like Java, C# exist… Learning being replaced by a black box is what the abstractions are for, in general.

Of course I agree, that learning is the right thing to do, and if you go from knowing how to make an engine to using an engine, you would be very productive, but in practice engines are a crutch for people who can't make their own engine and don't want to learn how.

Made a minimap. Feels good to add new stuff.

Anyone who asks that question never actually makes anything.

Participate in a couple of Game Jams to find out what you are capable of.

Thanks

That's a Doom mod, isn't it?

What is it about those older early 3D games that have a certain "look" to them? I don't know if it's the geometry or the limited palette.
Is there a way to force colors to palette in a modern OpenGL thing, but without the use of shaders?

Older 3D games look a certain way because of several reasons, Doom, Quake, Quake II without hardware acceleration have a 256 color limit, so this creates the kind of limitations on how many colors you can have in a certain scene, but also you have to consider the type of texture filtering used, because this also has a big impact on the look. You can tell that the webm is probably using Nearest Neighbor filtering as well as mip-mapping, Nearest Neighbor texture filtering allows games to have a "pixelated texture" look, although mipmaps are required to show the game at high resolutions without the background just looking like a fuzzy mess of pixels.

Games running with hardware acceleration generally used API's like 3dfx Glide, or OpenGL 1.1, these are different from modern OpenGL in that it's fixed-function graphics and has no shaders, and they encourage a certain look and feel because of the limits of the API.

Also, another important part of how old 3D gets its look is through lighting, lots of games used vertex lighting which was fast and light mapping which was fancier but both have generally been replaced by newer techniques, so you don't see them that much. And the sky is probably using more basic techniques like scrolling around a looping texture, instead of something like cube mapping, which a newer game would probably use.

Modern OpenGL has compatibility profiles, so you can use older code if you want. That said, you'll want to use the lowest spec possible to make sure your game actually runs well; GZDoom is a slow piece of bloated shit that runs everything through OpenGL 3+ and the project lead wants to add Vulkan soon for some fucking reason. Keep in mind that Crysis used OpenGL 2.

Just asking, because I think a Hexen styled aesthetic from scratch would be a nifty project in the future.

Rectangular and triangular collision map functioning as intended. Also added movement redirection if you're pushing non-perpendicularily against a wall (so you slide along it). You can now also catch some fucking fish.

+server side AI
+wandering
+targeting when attacked
+attacking back
+taking damage

Fuck yes!

I extracted the Spinter Cell Chaos Theory Soundtrack and noticed something peculiar. The songs are split into short 11 second segments so that the game can smoothly switch to one song depending on the level of enemy alertness. Is a song system that can change tunes depending on certain conditions complicated to make? the only other time I noticed something similar was in MGS Rising.

Hey that's a pretty nice looking character.

oof, not you.

You. This is a nice looking character.

What's happening to Sam's face? Is he on drugs? He's been looking more and more like an old alcoholic lately.

8agdg.wikidot.com/fps-arpg-hybrid is doing a great job at that

Its a state machine which is controlling the music. If you have an object which can be set based on what's happening in a frame during the render loop like, "being seen" or "killing someone who sees you" then you can easily tool that state to decide how the music is going to sound and then quiet down as it moves its state back to normal. Further that 11 second clip shit probably was done more for console porting as to not use up a bunch of memory for sound which won't be played for many frames. On PC you would just load the entire sound file and seek to where you want when you want based on the state of the system.

Baldur's Gate 2 did this. Before the days of Youtube, I wanted to listen to the Bhaalspawn Battle from TOB, so I had to get an extractor and it had several song bits split up in this way.

A few other games have "mirrored" soundtracks, that are identical, except more instruments are active or removed, so it can fade from one to the other, depending on action

"Fuck you pay me?" its a mysterious game with hidden secrets but you have to pay real money in order to progress.
Whats at the end of the game?
Who knows. Fuck you, pay me!

A procedural zombie sandbox mmorpg dota fps with lootboxes

Nah, the better title is obviously Paid in Abyss
Basically you just pay money to go down the abyss to see what's at the bottom. They say no one's ever made it; most whales just commit suicide when they blow all their money and wind up homeless. You could be the first, goy! The game also lets you take out a loan if you don't have the cash, but the interest is determined by your level of descent and it compounds daily.

user has made a thing with Vulkan.

And the best part is he admitted to know jack shit when it comes to anything graphical related and it shows.

There's a good chance this dinkwad has given up on working GZDoom, there's a lot of rumors and signs looking like he's getting ready to just make a fork of 3DGE, another Doom sourceport.

(checked)
Are you sure about this? He did nuked his repo some time ago and later he came back to make more quality codes.
Oh boy I sure cannot wait to try out GZ-3DGE with a Vulkan GPU!

MIDI games used to just have different channels for different levels of alertness, so they can turn instruments on or off to get this same result.

Didn't Peter Molyneux already do this one?

I have been making this song for a few days, I'm almost finishing I just want to get near 4 minutes, but I'm getting really tired of working on it. What you guys think? The violin segments aren't perfect but it'll do for now.

Now I gotta catch up with real life work that I have been leaving on the side.

Those of you who do 3D animation, especially in Blender, how the fuck do you progress past the dopesheet? How did you learn? I'm finding YouTube tutorials to be inadequate.

Coming Summer 2020, RokkuMan 17: Dokuta Wairi no Fucking Vaporware.

you don't even need a model, just use rigify and go
in case you have autism, there is cadnav.com/3d-models/human/ for your autism

bullshit, the only reason i could learn to program was because godot in its limited usefulness, didn't let me get lost

Oh gosh I'd better post some progress then.

Wrapping up improved entity-entity and entity-surface collisions and have most of hitscan weapon done including calculating the height in world space of things intersected so it can work when hosting multiplayer. Works through portals and everything but I need to add checks for colliding with top/bottom wall segments and do some proper testing. Please excuse wonky vertex lighting in a few sectors and excessive draw count in a few areas.

By "cant program" I mean literally being unable to do fizzbuzz. If you went from not being able to do fizzbuzz to being able to program because of godot that's impressive.

Finished the sprites for two more objects.

The first is the classic Seeker that moves towards the player (at half speed), and the second is a new object, the Pawn. It also moves at half speed towards the player, but can't hurt them. Instead, after moving, it pauses for about 3 steps, during which time it can be pushed. If it is pushed or left alone, it will again take another step towards the player and pause once more.

...

I also made a trap object. It holds an object in place until a release is pressed.

This doesn't look too visually annoying does it?

which ones are traps?

Flappy Bird but Pepe-themed where you avoid hammer and sickles or DRUMPFT-themed where you have to avoid rape acusations. Be sure to post all over reddit

subarashii

Every single one. I just tiled it a lot to see if it looked too weird. Most levels would probably only have a handful

probably wont look as weird when there are just a few

Are you sure you're actually a Jew?

...

Here's a randomized level with a few object types

3d render of amerimutt when ?

gets the point across which is the main point, right?

switch to the graph editor and refine your motion. most of the polishing happens in the graph editor.

Might seem like a dumb question but Goolag didn't give me an adequate answer within 3 pages. Building a new PC in a few months that's gonna be my main development machine. Putting in an SSD for the OS on that and I'll be taking the HDD out of this machine into it for basic storage and backup. The question I have about this is will this HDD automatically reformat itself when I put it into my new build (I don't care about preserving files. The few I do I have a craptop I can store them on)?

Graph editor. Dope sheet should only be used for the most basic of key frames (Like you've clearly already done) learn about the different kinds of automatic interpolation (You'll probably be using spline for most characters) or go full autism mode and do the majority of the frames by hand with minimal interpolation.

the logic part needs to be alright, otherwise you will never learn anything, but if you can get in if x do y mode the games write themselves.
I knew very little indeed. most stuff entered from one ear just to exit through the other, i could just follow any tutorial and forget everything right away, even then my first real project was 2d zelda just using the default icon as various sprites

don't you have a son now? how do you find time?

i think he means stuff like IK and constrains

But how do you prevent things like Mega Man there completely contorting inside of himself instead of flipping fluidly? Do you have to keyframe every quarter rotation? If I find that I need to go back and change something, then the keyframes later on in the animation now affect the placement of the frames before them, so I can't just delete them and redo segments. How do I handle that?
I'm slowly getting the hang of it. What's the proper use of the NLA editor?

Do you mean bezier?

I already have some experience with IK and constraints, since they were used in the rigging tutorial I did. I'm mainly trying to figure out what kind of workflow to use. How do you go from a blank canvas to a bunch of characters and objects properly interacting together, complete with camera work and audio? I knew the learning curve would be steep, but the process is different than I thought it would be.

Added some neat procedural stuff to the spawning system so I can hook up arbitrarilly complex generation for whatever data I might need, as specified in data files.
Right now it's just picking completely random colors for the mobs, and assigning them slightly random stats, but I have room for it to be expanded with functions that pick values around other values and stuff, for example, to make all mobs spawned by a spawner have similar colors and stats, but still have some variation.
I have some plans to have procedurally generated maps as well, so this system will probably become very refined in the future, with seeded generation, to make sure maps generated always end up the same, (so long as the generation algorithm and anything else stays the same)
And I'll probably use it to generate equipment and other things as well once I get that far.

I see, I was assuming that user has literally no programming experience like he said, and those links are supposed to explain to you the basics of structured programming.

He literally did. It was a cube thing where you had to chip off bits of a voxel cube with other people in a single worldwide lobby. You could pay more money to break off more pieces of the cube at once.

ShittySmileyGrid guy here.

Did a thing, though not as impressive as some of the shit you guys doin I'm happy with my progress.

I made the beginning of an animation system for my game objects.

Using SFML so basically im just storing rectangle coordinates that correspond to a sprite sheet. Every tick it checks against world delta time and increments the position in an array (or if its at the end it loops back to the start).

I then made a container for my animations
/ automated the creation based off passing it a spritesheet with how many rows cols and sprite height / width

Now to push this into game, make some quick dev art, test and then start workin on a game state system tomorrow.

gooknight bois its 3am here

thank you for that vid, I've been binging the whole channel since then

What I'm doing for my game is I made an Image class. This stores a reference to a Texture2D and a Rectangle[]. Images are stored in a global resource by an ID number. Then, any drawable object has an index that lets the draw code look up and find the Image (with the object containing information on how the texture should be drawn)

Basically, I set up all my Rectangle[] at the start of the game and that becomes either a single frame or a multiple frame animation. All you have to do is make a utility/helper function that takes a float (as percent) and transforms it against the length of the Rectangle[], then you can avoid having timers altogether, or perhaps have a few global ones and avoid explicitly tying it to the texture data

You guys have any idea how I can make controls for my climbing mechanic palatable?

The climbing is procedural and kinda like that one climbing game by Foddy, in a sense that your location is based on two "anchors" that represent your right and left hand. I figured out a way to figure out the location of the ledge and all, but now I need to figure out controls that feel good.

For now the movement scheme so far is:

Perfect solution would be to reuse one of the movement keys, but I'd like to keep space as "move up" within the anchor space and the initial ledge grab. Any ideas?

Oh right, I forgot to mention that the game is first person.

I think one of the Assassin's Creed games did this, and it worked well.
Looking at your controller, you have the ABXY buttons on the right side, Y on top, X to the left, B to the right, and A to the bottom. What they did was have the X/B correspond to the left and right hands, A correspond to the legs, and Y relate to vision.

You got me there.

Jumpscare battle royale

Flappy Breast.

It's not 100% sure, but I'd bet money on it given how he hasn't done anything for over a month and other devs haven't been able to reach him, and the only one that has basically got an angry response boiling down to "FUCK THE COMMUNITY".
I might be the one responsible for it. Oops. Oh well.

Good then, I hope it stays that way. Graf hasn't done anything meaningful to GZDoom during the 12 years, I'm sick already of his OGL 3.0 and Vulkan bullshit and then he has the nerve to talk about modern rendering yet when it comes to anything graphical related the only shit he did was a broken shadowmaps and the gimmick bloom effect. Too bad that this whole situation is not improved at all since Rachael is taking the lead now which means he is going to be Graf 2.0. Fucking hell the dynamic light has been broken and bleeding for ages yet they don't bother fixing it. But Vulkan garbage is more important for those retards.

Yeah back to square one in no time indeed.

I never understood this setup. Ever since Ghost Recon 1, I always assigned C to crouch and X do prone. Hell, it seems that C was the standard input for crouching for the longest time, until half-life at least.

also

If you have C for crouch and Space for jump, crouchjumping gets awkward.

Also fuck leaning. It supports cover faggotry.

Then your game will be shit.
jk

You should have specified you want the game to be fast-paced. cover faggotry isn't a problem when you have a slow-paced game with deadly gunplay.

i like setting it around whats needed
Zdoom doesnt NEED vulkan, new games do because of horrible optimization,
it needs more graphical improvements

It kinda is. It gives too much advantage to campers. A camper normally has to peek, and endanger himself. Leaning makes this half as dangerous.

I always thought of campers as retarded children, never saw them at an advantage in any game I played, even with lean.

Camping is a legitimate playstyle, albeit often poorly balanced. You can't just dismiss playstyles by calling the players who do so children. The game has to be fun no matter how you choose to play.

I'm not dismissing them. Just sharing my experience and opinion of'em from my online game experiences. If anything, I'd try to encourage and empower them (through lean, for example) because their playstyle seems to me to be a self-imposed handicap.

Why do browser games look so good?

This looks good to you? It reeks of Facebook plasticky trash

Make the keys rebindable

Perhaps the issue is not to limit the player, but to give him more tools to fight? Simply add leaning, but then make the movement speed high enough to kill anyone who camps. I don't see or hear any "camping" in Quake.

Can you make a game that looks as nice as that?

They already are. But I'm working on a default control scheme which has to be logical and ergonomic.

this results in power creep that leads to overcomplication.

Ever read that story about the naked emperor?

Because that's pretty much all they have, relatively competent graphics made by a C-team in a slav country that decided to create their own facebook bait next to their usual C-team jobs.

It really just results in the gunplay turning into the kind of fast arena shooter style, and then suddenly you can get wiped 30-0 quake style online by someone who knows their shit. What's wrong with power creep?

Because they can't do much else well. Usually they're limited by retarded platforms, inferior devices and/or interpreted languages, can't do twitch gameplay or use keyboard controls, etc. On top of that, they're generally expected to be super casual stuff, which implies a lot of polish (and usually a relaxed feel/atmosphere). Something that doesn't look polished is more likely to give the impression that it isn't something you can jump right into. Also static 2D elements are relatively easy to make good looking.

Yeah, but if I'm not going for gunplay being fast and solely twitch based.

It's a sign of poor balancing. You're forced to buff everything to a ridiculous degree just to make sure it keeps up with the rest of the arsenal.

Fixed

Arena FPS don't have balanced Items/Maps on purpose- and that imbalance is what adds a lot of the depth to the game. Basing the gameplay around item pickups makes the idea that every gun must be viable against every other gun a non-problem.

But, you aren't making an Arena Shooter, so whatever.

Why are there so many games where a single frag grenade can't kill someone? I don't care if it does 99 damage, I just threw a fragmentation grenade into an empty room with a guy huddled in the corner. He should be fucking dead.

So basically I have a global texture manager class that loads and keeps all my textures alive.

I have a game object container that essentially holds onto a commandcomponent, a physics component, and a graphics component.

My graphics component has a "Sprite Sheet" (essentially a texture reference with sprite width height num of colums and rows) and an AnimationContainer. (which is the array of animation arrays)

The main reason I'm using a timer is because my game is going to be delayed turned-base. IE: each turn is like ~6 seconds if you don't move or use a move / item or whatever in that 6 seconds enemies n whatnot are still gonna move / attack n shit

Isn't the main problem that most game frag grenades only do damage on distance? While that makes sense for the blast damage itself, the main purpose of a frag grenade are the fragments, so on explosion it should also shoot a bunch of raycasts into multiple directions and each fragments should do around the same damage as a bullet. Even if someone is 50 feet away the fragment could kill him if the fragment hits his head.

Because games aren't real life.

For a complex scene with lots of characters you'll need to get a good rig first. You will want to switch between IK and FK often and learn to temporarily parent bones to others. But work flow is the same. You animate each skeleton and in the same action. I've always avoided NLA editor outside of using it to delete actions but that's only because Im trying to do make my stuff for a game.

Because the gameplay can degenerate into a match of who shoots first and insta-kills the enemy. Low time to kill can remove fun and skill from a game. If you think of such a game, it becomes a match where everyone is constantly spamming frag rounds into doors / windows and moving around as quickly as possible. Now this can be fun if the movement speed is high enough, though. Half-life is an example of this kind of gameplay.

Of course if you want to prevent the game from being too deep, then you can make the frag grenade an insta-kill with low movement speed, but this is only really common in games like COD. Although in tactical FPS of course it's different just because it's not about aim battles or fast movement but positioning and strategy.

Yes, most games just use a shitty radius sphere and use a number gradient for damage calculation.

Don't even try to make that limp dick argument. This isn't some sperg fit about "muh realism" from ARMA fags. This is a game design gripe.The point of grenades from a game design standpoint is two-fold:
To flush out players hiding in bunkers, trenches, corners, etc
To take out enemies that take little to no damage from small arms fire. Like tanks, armored enemies and otherwise
A grenade should force a player to move from their position or die. But as it stands there's a lot of shitty games on the market where grenades will not kill the player unless they're fucking stepping on it.

No it doesn't if you take the time to balance your gameplay. No one who plays Rainbow Six: Siege will tell you that there is a problem with grenade spam. The grenades in that game are fine tuned perfectly. If you know where the fucker is, you can throw a grenade at them and they will either die or be pushed out of their position where they can easily be picked off.

Would anybody be interested in a new tutorial series for NES programming?

I've been doing NES programming for fun. When I get my bearings enough, I was thinking about writing a tutorial series covering effective programming for the NES using modern toolsets (no NESASM or shit like that, just cc65 and ca65).
It would cover some base architecture, proper use of the tools (how to properly configure the linker, especially, and to use it for the correct purposes), up to building a complete simple game. It would only assume that you have some base programming experience and that you can at least grasp assembly enough to hit the ground running with it. It would cascade up from the simple (building a pong game with simple sound effects) to the more complex (building a scrolling platformer with sound and music).

There are a few that already exist, but most of them are incredibly sparse or very incomplete. The best available is Nerdy Nights ( nintendoage.com/forum/messageview.cfm?catid=22&threadid=7155 ), but Nerdy Nights has a ton of flaws:

* Majorly, it uses NESASM instead of a real assembler-linker combo. .org directives and shit make the code a disgusting mess. You have to either use NESASM when using it, or try to take the concepts and apply them to a proper toolset.
* It opts to access almost all RAM directly by address, instead of using real symbols until the very end (when it should do it immediately).
* It is written as a series of fucking forum posts instead of a real contiguous and interlinked website. A proper tutorial set should live in a git repository so that improvements, suggestions, and fixes can be made as real pull requests and if the author abandons it, it can be forked and maintained by somebody else.
* Many of the lessons assume that you have downloaded the source code and are using that, instead of properly facilitating writing your own code while reading along, and having the source code supplied as a bonus for context and help (this is really egregious; a tutorial set should be directly followable building your own work as you go); with Nerdy Nights, you can not get a working sample up without going off his source code base, because the tutorial is insufficiently written.
* The code snippets frequently don't actually match the source code download, meaning if you even try to follow along directly, the behavior will be wrong.
* Nerdy Nights does not sufficiently explain memory mapping or how the NES file is built, and never once touches on the meaning of the iNES header beyond just saying what it is and how it goes. This, and tons of other information like building sprites, is completely insufficient, forcing you to use the nesdev wiki.
* It contains copyrighted rips of sprites from Mario Bros instead of using its own samples. This is illegal, putting the tutorial files at risk of being taken down, but also it makes the programmer use other sprite files, instead of teaching them how to make their own.

My tutorial set would be

* A pong clone. A series of 7 or 8 tutorials covering the NES architecture, a bit on number systems and assembly, and getting a proper toolchain set up and configured, how configuration works and how data is mapped to the file, and then from the file to the NES memory map, up through programming a single-screen game with 5 or 6 sprites on screen (which are designed by the programmer) and spitting out sound effects when necessary.
* A platformer. More advanced tutorial set on how to do more advanced stuff like scrolling, complete backgrounds, game states (title, play, pause, etc), and writing modules of the code in C. Single-stage, but with a bit of discussion on how you could manage multiples.
* A small RPG. More advanced, but more tangential. Talk about bidirectional scrolling (which is very hard in NES programming), how to manage save data and state, and how to structure a lot of code across many different files, and how to structure things efficiently and manageably in memory and in the ROM.

Probably written with Sphinx, because it already has a shit-load of built in flexibility for indexing, inter-document linking, file management, and has source-code highlighting including specifically for ca65.

I'm going to be writing it either way, just wondering if you all give enough of a shit that you might be willing to work through the tutorials and give criticism.

Yea, the kid is a cool little dude. Relatively low maintenance, compared to his dad, for now at least

When my wife went to visit her mum for 6 hours, it was me-time, essentially. I still have maybe 2 free hours per day and I spend it on either playing vidya, coding or extra sleep.

These are supposed to look like colored easter eggs, but I can't quite get the feeling right.

The bottom is TOO wide. I know it's meant to be a bit wider than the top, but it's closer to a teardrop shape than an egg shape.

I have a pretty decent one. Assuming you mean rig as in computer, and aren't referring to character rigs in some way I don't understand.
Are there any rules to knowing when you should use IK vs FK? What is temporary parenting used for aside from things like characters holding objects? Can you use keyframes to control these?
Oh shit, I didn't realize that two armature's keyframes would be added to the same dopesheet. Should I be adding all of the bones in an armature to each keyframe, or just the ones that changed in that keyframe?

They're too dark, you need a lighter tone.

Try dithering the shading to get a matte effect if you've not got enough tones in your palette to do it. Adding patterns would help too I think. And use the shades you've got to hint around the curves to make it appear smoother.

Anybody else REEEE'ing over this?
cgpress.org/archives/unity-to-drop-support-for-monodevelop.html

wasn't this on the horizon since they added VS to their package? I look back with dread to the days of using monodevelop

I'm happy about this but I also feel like I fucking better deliver now.
Good motivation to keep working.

sculpt user here back with my hand does it look better or not, and if so what is wrong, i really just want decent quality. also one question, for retopology anyone used any of the addons for blender like retopoflow or Bsurface i just want to know if i should go for them or not and if they good/which one is better.

...

Maybe one relies more on CPU and the other on GPU, and the potato has a worse GPU compared to how much worse its CPU is?

It's most likely that, considering that the potato has one of those Intel chipsets with 32mb of video memory.

But I'm only drawing one sprite and some text in Love2d, whereas I think my GMS test had more going on graphically, which makes me wonder why there's such a huge drop.

I thought love.run had at least 1ms sleep in it so 1000 FPS/UPS was the maximum?

Getting there, the palm is the wrong shape and I think that's the crux. Hands vary from person to person, but my palm is a rounded square, with the fingers sans thumb taking up roughly the same area if I lay them over my palm.

You can change that in the settings of the love.run() command. I set it to 0.1ms, to see how fast it could get in theory, and been working with that.

Ah, that makes sense. It's been ages since I've had to muck with love.run.

I'm using Blender and Unity and I'm trying to split a model up to allow for equipment swapping. However, splitting a model creates normal seams like this. I understand why it happens, but I have no idea how to fix it. I've tried manually adjusting the normals, but they don't animate properly. I've also tried using a normal map, but that didn't work out either. Everywhere I go, I see people with similar problems that never get answered. Would anyone here happen to know how to fix this?

Add all the bones. Because if you change them in another animation and then you switch back animation it will keep the others transformations. FK and IK are different forms of precise movement. Google differences. Also parenting is not just for picking up objects. Rigs are not just for arms and legs, they should come with other bones with drivers to make certain animations easier. Go look at Shinobufags thread on the /agdg/ board. Yo should have like 100 bones minimum fir a good rig.

Tell me fellow pilgrim, is the journey from gms to love2d worthwhile? I don't like gms at all but with some months of experience I'm PRETTY comfortable with it.

The difference between us, it seems, is that I've used Game Maker since version 6, for probably 7-8 years now.
I moved over because controller support on GMS was borked and GMS 2's interface was absolutely disgusting from the time I spent with it. I've said earlier that I only stuck with GMS because "It just worked™", but it stopped working so I moved on as well.

Love2d was painful to get used to at first, but five-ish days in I have a solid grasp of what I'm doing. And something about getting basic functionality to work by your own hand rather than having it spoonfed, and figuring all the concepts out like that, it's comfy without turning into engine-devery.

Reworked how I handle animation dependent movement speeds and added the ability to add screenshake to animations. So that's a thing.

the rumble when the big guy is walking is too much, just use it on the landing.

Is it true that it's easier to learn and use older versions of OpenGL than newer OpenGL and Vulkan?

It's certainly easier to learn OpenGL than Vulkan.

Absolutely. It's also not nearly as useful, but it's arguably a better place to start learning.

It depends. Unless you want to develop for hardware that's toaster by toaster-standards, you shouldn't even think about touching OGL 1.x. After that you got OpenGL 2.1 as one sensible target. OpenGL 3.3 would be the next one after that. There has been a major rewrite of the specification between those two, so learning 2.1 will have the downside of making the jump to newer versions more work. After that, you get OpenGL 4.1 and Vulkan, which restrict you to somewhat new hardware; especially Vulkan, obviously.
Here is some data on the hardware we're talking about:
What does this mean for you?
>if you want to make something really simple, 2.1 could be a sensible choice, as it runs on just about everything

Oh, and speaking as someone who has just finished his introduction to graphics development (with Vulkan), I can tell you this: Vulkan isn't really all that hard, but it's a lot of work. It's often called a "low level API", but calling it an "explicit API" is better at bringing across what it is. The code I wrote didn't feel low level at all, but I sure had to tell the driver ever god damn detail about how I'm going to do things.
Hope this helps.

I'll try to make it more oval shaped, then.

I think they are just bright enough, but too saturated, looking at it.

I was going to have a set of pastel-only and decorated eggs, 16 in total

uhh, that is the top of the hands this the the palm/bottom of the hand

Thanks anons, I was thinking about creating something basic mostly in 2D with some 3D elements.

Then go with either 2.1 or 3.3 core profile. Personally, I'd go with the latter, since it has the programmable function pipeline and got rid of other ancient crap. It'll give you a solid foundation in modern graphics programming.

Why are my compiler errors always really small stupid shit like missing a semicolon or missing a return type. Happens all the time when I focus on one thing for like 3 hours and write a bunch

I found a solution to this problem. Create face overlap using invisible faces. See pic related- I have difficulty describing in words

That's because of the subdivision modifier. You will have to make sure they overlap 1 loop more and then when you are ready to export it, apply the subdivision and trim off that extra loop. Subdivision at level 1 affects vertices 1 neighbor away so if you cut off the foot, the remaining edge and its 1 neighboring vertices will be affected.

So you duplicate the first loop of edges from each side of the seam. When the foot in this example is placed against the ankle the ankle should have one loop across the seam taken the ankle, and the ankle should have one loop past the dream taken from the foot.

So I wanted to see which approach was fastest to set flags for my tiles in C#, basically just modifying a value (occupied flag)
>bool[], byte[], int[], custom struct with a flags enum, HashSet

The arrays were more or less the same speed, being 86, 87, 95, 99 ms respectively, but the HashSet was SLOW for insertion, 154 ms. Probably going to use byte[]
hastebin.com/iqufezitav.cs

Got this up and running in five days in Love2D, having never used Lua in my life.
It's a basic player character with animations, movement, font and camera, as well as other things like splitting up the game into multiple .lua files and timers for both the animations and reading control inputs.

It's not much, but considering how overwhelming Lua was at first, I'm pretty happy.

I'm sold.

It worked! Thank you so much, I can't tell you how much I've been fretting over this.

There was no subdivision modifier, it's just how normals and smooth shading work with split seams.

I can't drawn and never did any 3D modeling but I just downloaded Blender, how many hours of practice before I can make cute anime girls

10000 at least. You should stop on over at >>>/loomis/ and pick up some art supplies. If you have some potential you might be able to start making some cute anime girls in a month or two; really depends on how often you practice and how well you understand the elements of a composition. I've been learning to draw recently myself, and it's actually kind of a nice break from coding. Tonight I'm working on structure.

Should I learn to draw before I start with 3D modeling or can I go straight to 3D?

If you're going into character modeling learning how to draw in 2D can help a lot, especially with anatomy. It's not a requirement, but 2D and 3D both benefit from each other.

I think it would be necessary to know how to draw before 3D modeling. I'm sure with enough trial and error you could get decent looking 3D models, but it would be pretty frustrating. When you create 3D objects via modeling, you're still just essentially working with design principles you'd use with 2D drawings anyways. Like says, knowing how to draw will help immensely with characters models. Actually, you would have a hard time modeling any organic object without a solid drawing background.

It's the other way around. Modelling helps you draw better. As long as you do these things more or less simultaneously.
You start understanding the forms and then bring them to the canvas.
The modelling process itself has very little to do with drawing.

Post specs please

Yes, OpenGL 1.1 with optional extensions is a really great target and lets you program much faster

Here's another picture of my project that I put on the wiki, I am working on the BSP compiler right now.

I did it… Metroid Prime doors. And i though i would struggle. To the user last thread who my swaying reminded of Metroid Prime, you got my inspirations completely right.

I've always wondered the posibilities about converting s BSP to an OBJ. Is it really feasable?

That's some slick presentation you got going on there user, nice work with the door.
How's working with godot?

Thanks user! It's an absolute breeze since they made QOL improvement on GDscript. All it needs is JIT optimizations.

i haven't gotten anything done since xenoblade2 came out please kill me

Kay.

ain't gonna spoonfeed you, faggot

Glad I could help! I haven't seen anyone else describe this trick online so feel free to share it with others.

Silly user, this isn't the AGDQ thread.

Unreal 4 has that functionality built into the editor. It is super useful for the standard workflow for level design ("blocking out" the rough layout with BSPs, exporting them to your 3d suite and remodeling them one by one). And you do have to remodel them, because the topology generated by BSP trees is godawful.
There is a reason for why BSP based level editors always restrict you to a fairly low resolution grid (Valve's Hammer editor uses 1" for example), and there's a reason for why modern game engines don't rely primarily on them anymore.
If you look at how they work, it becomes clear. The meshes they generate are far from optimal, because they produce a lot of unnecessary edges. And the more detail you add to your levels, the worse it gets; and it gets worse fast. Simply put, BPS trees don't scale for shit.
They work great for old games like Half-Life, because those game's level's consist mostly of perfectly flat surfaces, many of which are parallel. On top of that, there isn't much stuff in them to begin with.

The only upside that BSPs have nowadays, is that texturing is super fast. No modifier and no trick I know of in Blender is anywhere near as fast as texturing, say, a room in Hammer. I lost so much time on the UV mapping of the retopologized meshes, I'll have to write a Blender plugin to ape Hammer's functionality.

I'm very interested

I think you intended to write to

Anyway, while I won't be making a NES title, seeing a cool tutorial like that on the wiki would be awesome

Perhaps the shake when the creature walks should be toned down a bit, instead of it being removed, as suggested. You'll need to make tweaks but it looks hella good, as all your stuff, my man.

I'd definitely be into in that type of series. It's very interesting to read up on how even relatively simple things we take for granted took much more effort on such restrictive hardware.

A BSP is really a few types of data structures that let you prove things about a bunch of polygons really fast. So, while the BSP does not describe any specific 3D geometry itself, you can just take the polygons it's referring to and convert it into a .obj. rather easily.

Although it doesn't really make sense to convert 3d data that's been organized into a BSP into an .obj because the BSP conversion increases the polycount by quite a bit. You're better off taking the data before conversion into a BSP and using that.

The benefits of BSP is that you can do things like collision detection and line-of-sight testing very, very quickly, as well as depth sorting which means you don't need a Z-buffer for the world, which made it a better choice for software rendered games like quake and half-life. Less old BSP-based games like HL2 would use a BSP for the low-poly geometry and then stick a bunch of models into the scene to compensate. Quake 3 would use a BSP tree for collision detection but would not actually use the mesh with the split polygons, instead they would just render with the original mesh because they didn't want to increase the polycount. This is because by Quake 3 Z-buffers are hardware accelerated, and thus BSP's main selling point is gone since it's fast enough to just use that.

Does topology matter for static objects such as furniture and buildings?

Of course, you don't want unwanted crooked things, do you? Games should be like real life, get it?

Man you must be a blast at a party.

the bones/fingers are good, the flesh looks bloated and way more work is needed on the finger webs.
the thum's "armpit" is completely wrong (it should continue the line made by index finger), it is too long (tip of thumb should be near the pinkies first joint after the knuckle)
thums tendons could be visible
the index and middle finger's tendons should form a "v" shape closer to the centre of the wrist.
the pinkie knuckle is too low.

the nesdev wiki is sufficient for technical reference, assume your reader knows how the machine works.
usually tutorials are useless because they assume an innapropriate level of prerequisite knowledge.

a lot can be learned from the smb dissasembly, a whole series could be based off it's code.
I would have a chapters on debugging tools, macro assembly, fixed point arithmetic, and concurrent logic (jump tables).
a good article would show how to do game design with specific constraints in mind.

I think there is more interest in rom hacking than homebrew, and few people know how to inject their own assembly into a rom.

Yes, I was still measuring for that. Either the fingers seem small or the thumb seems big, the pinky seems like it's further into the hand than it should be. Again I'm basing this on my own hand though, at the very least the thumb seems too big.

spent a few after-work hours making a couple particle effects, this shit turned out to actually be a lot easier than i feared it would be at first.

this is a first quick pass for some effects, mostly just for the sake of having some stuff to work with and be able to refine further down the line; placeholder if you will.

apologies for lack of webm

All that foliage is crying out to get blown back from the shot and explosion.

Hot damn. Looking good, user.
That blowing spiral looks a bit late and a bit high. Here's some source material for you.

projectile connection is also way late, should be instantaneous at that range

Depending on the kind of game he's making, I've seen hit-delays like that in many turn based combat games, so it might not be unintentional.

Is this the new Valkyria Chronicles?

yay! love to see you still here, user.
I'll agree with the two other dudes: the grass and shrubs should sway on the gun firing and the projectile should hit instantly. Other than that, looking great.

I kind of realized now: you showed us the swimsuit cutie in the last thread, right? really promising stuff

Just because other games do it wrong doesn't make it ok.

I'm planning on doing shallow coverage of much of that. Enough that you could get by without referring to the wiki. I'll point to the wiki for in-depth technical stuff, and only expand when the wiki's content either feels inaccessible or incomplete. Certain things like the sprite format will have to be covered in a bit more depth in order to properly explain how the user can go about building their own tools for building pattern tables (a simple POC will be included showing obvious methods for converting between an indexed PNG and a pattern table file; it can be done in only a couple dozen lines of Python). I'll also refer to more complete tools available on the internet, because you can only get by on a POC shim for so long.
Anything that is better off being referred to an external source (especially the nesdev wiki) will be done so.

I feel that most tutorials are useless because they either run in an inappropriate order (the Nerdy Nights one, for instance, doesn't introduce subroutine calling or defining variables as labels until the very end, which is fucking stupid. If your reader is being expected to either already know assembly or to learn it, they can manage subroutines and memory labels, especially because they've already done branching at the beginning), or they leave wide holes in the middle, jumping from disparate concept to disparate concept without any connection (any tutorial set that has you dump all your code and start from scratch for the next lesson is flawed; it impedes the ability of the reader to follow it and maintain consistency in their head), or they cover the high-level concepts without getting specific at all, expecting the user to glean all the particulars from the source code without explanation.

That's planned, but much of it will be out-of-band (the tutorial set will refer to a set of "appendix lessons" which cover that content where appropriate) to avoid diluting the main content. Once the learner has learned macro programming, they should be able to go back through the tutorials again without having that part getting in the way. Each page, including all the appendixes, will have its own well-commented compilable source code reference. Anything that can't be better explained by me than another source (like the nesdev wiki) will be directed there instead (though I'll provide a working commented code sample).

I honestly don't give a shit about romhacking. I like programming, and it's what I do for a living; I'm not going to write tutorials on something that doesn't interest me at all and that I have no experience in.

I mostly want to do this because there aren't any good, modern, comprehensive tutorials out there for getting a beginner with some programming experience up and running into NES homebrew development, from setting up a development toolchain to running a full simple game, and absolutely none of the ones out there are in a format that allow people to raise issues and make pull requests (there are the NESDEV programming guides, but that's more a pile of articles than anything like a single comprehensive tutorial).

Either way, most suggestions and criticisms you have could be made in the form of issues or pull requests, as this will all be in a git repository (probably on github, so I can just host out of github pages. It sucks, but gitgud.io doesn't have pages enabled). It's also probably a month or two out before I start working on it, as I'd like to have a more comprehensive understanding of some things than I have now (particularly sound and some more obscure tricks).

Yeah, I was just testing out the tech.

I spent today rewriting my camera, and also added camera pan triggers.

why tho

Because ~^*`dynamic typing`*^~ (and it prevents a few weird undefined behaviors in C)

My biggest gripe with Lua is that its 1 based indexing, all numbers are floats, and Love2D has 31k memory out of the box

Finally finished my basic graphic component system with my Animations tied in.

#ButNoProgressWithoutAWebM

How do I learn to like programming?

How much do you actually know?
I find a great motivation is knowing how to apply what you've learned in a useful way

I know how to implement basic data structures (stack, queue, linked list, tree), but I don't know when.

Think about how you'll be using the data. In terms of sheer performance, an array will absolutely have the fastest read/write for random access if you know the index of the elements you're using.

If you need items sorted, a tree is more appropriate than say, a stack. In most cases, you can get away with a simple linked list if all you need is an unordered collection of objects you need to iterate through. Stacks and sorting go hand-in-hand.

Source tiles on left, mine on right; I just redid my keys. I need to adjust the doors and gates slightly to have the new symbols, also guy says they're too similar, but whatever.

I'm old as fuck and borderline retarded but learned a decent amount about rigging in one sitting, it's really easy once you get into the swing of things.

I went with movement, collision and camera. Also splitted, it was easier than I thought.

Yeah, that got me by surprise

By the way, the glitch at the end of the video wasn't intentional.

And music, almost forgot.

Want to try to recreate Magnus the Red from 40k. Left two are my wip and the right two are some reference. Only a couple hours in but I'd like some broad critiques about the overall look and what needs improvement before I move forward.
Would also like to know if anyone has any tips and tricks to creating hair and wings in Zbrush.

I just don't have the natural knack to plan that stuff out properly. And then I go and look for tutorials but then I feel like I'm just copying some guy's code, and not really learning. Like I can't even think of how to properly set up a tile-based platformer. I can get a sprite on screen, I can move it, I can give it a bounding box, but I can't think of how to properly store a map. Maybe create an easy system that let's me build one and then read that file. Like, how would you handle collision? I was originally thinking each platform block has its own bounding box, and I'd just build the map up like that, but I know that's probably inefficient as fuck and would probably cause problems with spaces between each block. But I can't think of a proper system that would create one large bounding box around a collection of platform blocks. I'm just retarded, I guess, and that's why I can't enjoy programming.

What I would do is store a position and velocity for your player, and then on the update method, find the updated position. Then, you'd consider a rectangle between the origin and destination, and look at each tile and object within that bounding box, since it "could" intersect with the player. You'd basically draw a line between the origin and target, and test it against each of the potential objects. If there is a collision along that line, then you set the target position as close as possible so it doesn't intersect, and repeat this process until it's clear. At worst, you set your velocity to 0 because you're on a floor or wall.

Optimizations would be to limit the number of objects checked for excessive velocity (eg it gets set to 10,000, you wouldn't check everything in that area), as well as a better storage of tiles in eg a BSP tree or something to cull them faster.

There's better or specialized ways to handle it, but that would be a naive approach

Is there a design book or something that I can read to help send me on my way with this kind of stuff? I always wonder how much of this kind of thinking should come naturally, or if everyone's learning from the same 'masters'?

This site has been recommended a few times
gameprogrammingpatterns.com/contents.html

Mostly I just pick up bits and pieces as I go

Why not have a hierarchy of bounding boxes? Thus allowing you to test the player against all bounding boxes in O(logn) time. The idea is that on the lowest level of the hierarchy, every platform has its own bounding box. As you go up the hierarchy, you get bounding boxes that contain multiple smaller bounding boxes. And so, if you aren't inside one of those larger boxes, it won't bother checking if you are inside of any boxes inside that box. Consider pic related: a platformer game where the black boxes are the level geometry. I only need to check against three bounding boxes instead of eight! This scales very well, because I can have 2^n bounding boxes for every bounding box check. So, I can check 1024 bounding boxes in just 10 actual checks. Of course I need to store 2048 bounding boxes to do this, so it takes more memory.

What would it look like if the black box intersected some of the middle?

Also you could have made that into Loss

well the black boxes are the actual level geometry, the algorithm would just pick different parent boxes for everything. All it has to do to build the tree is find two boxes on the current level that fit into one box without intersecting any other boxes, then repeat for each level until the root box is created. So the black boxes can never intersect the other boxes, because the other boxes are derived from the black boxes. Pic related is another example…

shit, you're right…

What about doing everything tile based? Using a two dimensional array where every bit of environmental collision is made up of pieces that fit into uniform square tiles. It's how a lot of old games did it.

If you do everything tile based in a 2d array then culling is a little bit simpler to understand, just because you can immediately derive what tiles your player could possibly be touching by deriving the array coordinates from the actual coordinates rather easily. Then, you just check the player against all of the tiles around the player, and that's it.

Of course doing everything in a 2d array is really wasteful, so you will want to implement some basic compression of your level, just because, It's a huge waste of space, to use a brute force solution like that: a level that is 4096x512 at 8 bits per tile is 2MB in memory, so using a 2d array is just an absolutely ridiculous waste of space. If you implement the most basic compression scheme possible, like for example: if this 4096x512 map is taking place outside, then maybe the top layer will all be air tiles. Instead of having 4096 places in the array saying [0,0,0,0…], it can just say: [0,256,0,256,0,256…], Now, increasing the bits per pixel to 16 bits might actually save more space in large empty areas, but the point is that where I was taking 4096 bytes (4kb!!!) I am now only taking 32 bytes for the same information, and since tile based games love to repeat the same tile over and over, it saves a ton of space. Now this messes up our simple to understand collision detection but that was basically the only advantage to using a 2d array, I haven't thought of how to do collision detection on such a map yet and this post is getting too long, so, I will be going to bed now and maybe tomorrow i'll think of something.

...

yup, that's me.
any game without waifus is trash.

Someone tell me if this is a good or bad idea.
Problem
When shooting the pistol there's a lock preventing firing that lasts for 0.16 seconds. This can make shooting feel bad/unresponsive because if you click while the gun is firing it wont register the click. If you double click to double tap an enemy only the first shot will be fired.
Solution
To get around this, I made a "firing queue". If you fire the pistol while the lock is in place a shot is put in the queue. When the lock is removed the shot is automatically fired. Now when you click while the gun is firing it will still register the click and fire a bullet, only the shot will be a bit delayed. Double clicking double taps now as well, just with the second shot coming out later than you clicked. Only one shot can be in the queue.
It feels better than it was before, but I'm not sure if there's a better way.

honestly i think that the delay on the 2nd shit would be worse than it not firing at all.

I timed it. Between the second click and the second shot it's usually a .05 second delay and at most a .08 second delay unless you are deliberately trying to fire as fast as possible. It's hard to explain but for me it really does feel better. When I'm spamming shots I hardly notice when it's there, but it's obvious when it's not because it breaks the rhythm. Maybe I'll put it in the options menu so that during the demo day people can try it out both ways and report back what feels better.
It'll be less noticeable as well later on when I decrease the amount of time the lock lasts. 0.16 seconds is a worst case scenario.

Making simultaneously-swinging doors was surprisingly complicated.

Feels good to start a new day with no bugs hanging over my head.

At least with LuaJIT you can work around those first two by using C arrays and types, which can save a bunch of memory too.

thanks

I hope so, I know the general gist of it but I never really touched it completely.

looking good so far
here's some things I think might help, pinterest in general has alot of good images and guides for 3d modeling, especially zbrush

Ledge detection and custom physics constraint code for my player actor. It's all still janky and nowhere near finished but I'd be lying if I said I wasn't proud of the ledge detection.

It's a fine idea, but cap it to two or three max. If the user hammers the button a 10 times in a quarter of a second, it would be stupid to have the gun firing shots on its own for a full second and a half.

These guys are good enough I think.

I think I'm getting limited in how I'm defining object behavior, it's becoming a mess of subclassing even if I use function pointers. I also need to switch to an open language (instead of XNA/C#) so I was looking at Love2D.

I haven't used it much, but I remember having trouble figuring out metatables and simulating OOP stuff. In particular, variable scope and making objects.

Your main character is a priest ?

Knife's base is very simple: github.com/airstruck/knife or Classic if you want something more fleshed out: github.com/rxi/classic/

For Base, in main.lua:
Base = require("base")MyClass = require("myclass")local classInstance = MyClass()

In myclass.lua:
local ClassThing = Base:extend()function ClassThing:constructor(...) -- : is syntax sugar for being able -- to use self and refer to the instance -- you're calling the function on. Like: self.x = 0 self.y = 0endfunction ClassThing:multiply() return self.x * self.yendreturn Class

Hope this helps and I'm not repeating stuff you figured out.

No, I'm working with other anons just using only Love2D, I just mean with Lua in general, not with an extension

One thing I'm having trouble with is say, love2d.org/wiki/love.graphics this page. It has all these defined types, but I can't seem to access them in any way. The only way is to create them through eg graphics.newQuad() or the like.

It took me 2 hours to figure out that enums don't actually exist, despite the "enum" category, and to actually supply it as an argument into something like love.graphics.setDefaultFilter('nearest', 'nearest'), I had to put it in quotes, instead of FilterMode.Nearest, despite what the fucking enum said

I have no idea where things are in scope and its frustrating

Continued testing my command input.

Now to implement collisions.

fuck off you stupid mexican

:D

What do you mean by extension user? Require is just importing the file contents to run. The base example just returns a table with some functions and a metatable set up for you.

How do you mean access them? They're object types Löve exposes so you can only create them or use object functions on them.

Yeah enums is a misnomer since you need to provide strings Lua-side. To be fair the error does say it expects strings.

Variables are global by default which are accessible across the environment (really they're just in the _G table which gets checked if you try and use a variable not available in your current scope), local variables have the same sort of scope you'd expect in other languages.

do local i = 1 -- local to this doendi = i + 1 -- error: i isn't definedlocal a = 1do a = a + 1 -- using a declared in parent scopeend

local t = 1
do
local t = 5
print(t) – 5 because of local t overrides
end
print(t) – 1[/code]

Is she going to have a hat too?

The tittyloli continues to grow, I see.

This is going to be the base model for most females, i will use shape keys to change everything from beast size to height

I think you are confusing me with some other dev

Anyone know any good tutorial for getting into 3D modelling with blender? I have ZERO artistic talent btw.

welcome to the club
i don't have any specific guide, i search "modeling on blender for unity" on youtube and basically watched everything i found, you will find a ton of conflicting information and horrible guides, you will need experience to learn what actually works and what doesn't, specially since mostly are for older versions of blender

Art is of secondary concern to the Gameplay. Your art can be replaced later. Gameplay and programming is ofvthe utmost importance at this stage of your development.

That said, if you keep hitting walls in your programming, you can make decent assets so you have "permanent" work completed

Fuck, that's exactly what I was trying to avoid having to go through. Oh well, guess I'll just have to suffer through it.

Already a programmer first and foremost. Wanting to learn modelling for pretty much this reason

Nah, just was joking about the general small bodytype combined with big tiddies.

first thing first

uninstall blender and pirate maya, it's much easier to use for first timers and allows you into 3d studio max much better which after uninstalling maya you should pirate next

It's not wrong just because you don't like it, as well. They do it as a stylistic choice and I'll tell you why - it provides clarity between when the attack was made and when the enemy was hit. Almost all JRPGs are like this dude. So are literally all turn based games. Civ games? many rpgs?

A few notes:
Other than that, it's pretty impressive, keep up the good work.

Really depends what kind of gameplay you're going for. If you want a slower-paced tactical shooter like R6: Siege, you'll probably be better off ignoring inputs. If you want something faster-paced like Doom, an input queue would be better.

He took the blackpill.

No thanks.

get a plug'in for blander which turns it into 3d studio max then

Is there an acceptable way to use colored vertex points in lieu of a texture in 3d? I assume lots of people use it as a crutch with low effort indieshit that it has a stigma to it.

Looks like horror-inspired Enter the Gungeon

In Love2d, I wrote a function that asks for four variables, specifically coordinates of a rectangle. I intend to store these in a table like so.
Is there a way to store them in such a way to where i could easily feed all the values to the function? For example, just specify solid_data[i] and just have it read all four values from that.

Your shape is good aside from titties. Your vertex distribution is meh. Especially head its too dense with vertices. Use subdivision .You wouldn't imagine how many people I've told here to use subdivision. The less vertices = the more easy shape tweaks, smoothness, deformations, and easier topology.

Give those titties some gravity and simplify the vertex dense chest abochest and head.

FUCK my phone. I meant chest area above the tits.

Are you sure you don't just want a table of tables, like solid_data[i] = { 1, 2, 3, 4 }? If it has to be flat just pass the table reference and index to start from and use index to index+3 in your function.

It's capped at 1.

That's the kind of game I'm making, so that's good to hear.

I hadn't checked on my program's compatibility with my performance target, it is doing very well in this category. It worked the first time, and with no recompiles needed, pretty good since I had not attempted running this for three months. It also runs rather well, although I think it was maybe struggling at the highest resolution, it is still a great machine, it even has support for VBO's!

So that's cool, I guess

is this any better, and yeah i have a problem with my hand, im double jointed on my index and middle finger, so theirs that. as of detail i just want to go for PS2 era, nothing modern, so i think it looks fine.

user here what i use, but FYI do you want to jump into topology or do sculpting then retopology, because from experience, doing topology without a reference is nearly impossible. as for artistic talent you'll gain as you learn i mean look at me (pic 3 from recent pic 4 from about 8-10 months ago) in short i can get good if you just take the time to practice.

youtube.com/watch?v=I6K7x1tX9rk

shit didnt post images my bad forgot about the embed thing

pic 1 and 2

pic 3 and 4

The index finger is longer than the ring finger. The shorter the index finger is compared to the ring finger, the more testosterone the person has.

Question for devs working with frameworks: What's some sensible ways to organize your code? I've only written small games in C++, and when I think about writing larger games I can only imagine the horror of initializing 20-30 classes and passing various shit to "managers" in some horrifically large update loops.
I've looked into Entity Component Systems, and they seem like an okay solution, but I don't want to create unnecessary abstractions or delve too deep into "engine dev". There really isn't any good books on 2D engine architecture that I can find, I've read game programming patterns, and countless dev blogs looking for concrete examples but I always feel like I come away having learned nothing.

I'd really rather not use Unity or gamemaker anymore, I like programming in C++ but this has been beating me over the head for so long and I'm at a loss

Looking cool, is sigma II going to be open source?
I'd make maps n shit

Glad to see you back user, the camera pan trigger looks solid.

Not FOSS but under some kind of NC license. I don't believe in letting people use my code in a commercial product without paying me, that's my only objection to FOSS. It doesn't acknowledge the code itself as a commercial product, yet somehow art and everything else is different. So, maybe in the games I make with it, I will distribute the source code under MS-RCL and then you can buy a commercial license if you want to sell stuff you make with it, I don't really know.

If you wanna check out how it is right now, here's a download of the engine and the map editor:
a.uguu.se/q1b4Uxab1Hg0_sigma2_rel11118.zip

Also you can write code that is rather effective and clean looking, even in massive projects. Since I am a C fag I don't use OOP or whatever but I do have to initialize quite a lot of stuff to actually make my engine work, and it all has to be done in a specific order, however my actual game loop is relatively small. Since you are making a 2d game 20-30 classes is a little bit much, I mean I only have about 10 or 12 (i forget) to make my game, and usually I don't even have to think about the parts that I've finished already.

The problem you are having, I think, is that you are worrying about writing bad code, instead of writing code, maybe. I think you should just start programming, it doesn't even matter if you write bad code, writing shitty code is a great way to learn about how large systems work, because you can look back at it and identify how it went wrong and how you could do better, it's a lot easier to understand how to write clean code after you write something that is a total mess first.

I'm not worried about writing bad code per say, I've done that plenty in previous projects. My most recent one I posted around a bit, it was a small RPG game. I personally can't stand to look at the codebase for it now, it's so poorly written. I think you're 100% right about learning to write good big systems from shitty smaller ones, but I'm more worried about how long that will take.
Thanks for the link, and the advice, it's pretty motivational. I think I'll try to clean up that codebase, starting with reworking the tilemaps into 2 dimensional arrays or something like that

Speaking of bad code, does file reader/parsing usually look like this in other peoples' projects?
hastebin.com/tayariciqo.cs

I have a tab delimited text file, because for the effort of making it just a flat file of bytes, it's a lot easier to edit shit the way it is now (and I'd still have to use ByteConverter or something, anyways)

CC 3.0 is probably what you're looking for: creativecommons.org/licenses/by-nc/3.0/us/

per se*

Great! Thank you guys for the feedback. I'll post again soon with another update.

Will you go inside, y/n

Oh, nevermind. I accidentally the entire fucking room.

the shinny floor doesn't really fit tbh

That's just the power of GZDoom.

can it do shadows?

Just wait for the Vulkan version of GZDoom it will fix everything.
t. Graf Zahl

I finished the second song, just added another part. Also I finished a third song, all in a single day and 2 hours. Considering it took me about a week to make the first 1.8 songs, it's amazing what you can do when you are inspired.

Shit, forgot again to change the resolution of the mp4s

the first song is very expected and fairly bland because of it, I forgot I was listening to it that's how bland it is. and so is the other song

Well my mom liked it

ur mom's a poop

It looks normal for a text-based format like that, I am doing similar enough stuff with mine.

For the kraut devs that hang out here, this might interest you.

www.deutscher-computerspielpreis.de/preis/faq

They're handing out awards for german made games, which includes prize money. There's multiple categories, but what's interesting is there's one category for students that lets you submit only a concept. So even if you're a nodev, you can potentially win.
I don't know if there's a hook to all this, but I'll probably send something in. Deadline is the 19th, so that's enough time to whip up a .pdf with some assets and ideafagging.

...

I don't hold it against him, we only had a soft agreement and in his situation I would probably take the guaranteed safe money too. But damn. At least I'm getting fairly close to the end in terms of stage assets and design.

Does anyone know the best places to find musicians? I lucked out when I found the last guy on Youtube, he is so good and does that SNES style music I wanted but now I gotta find a new guy.

Butt
can't find a good topology guide for animu butts

I haven't read into what they expect, but I figure if your idea isn't solid and has no concept art, you're not likely to win. You'll also have to state what you use the Preisgeld for and prove you're using it to fund your game, or they'll revoke your award.

Im afraid of simplifying the upper chest area because my male model got fucked up shoulders because i couldn't weight paint the vertices in the area properly from lack of options
I will probably simplify the head later since im using textures as faces anyways

Yeah, I just read the part about how you have to have material that proves you're gonna earnestly work towards finishing this.
Well, kinda curious about who's gonna win with that.

I honestly think I have a decent shot at it, since I've been concepting for a while and have a few art tests and mockups. My plan was to throw a demo on Kickstarter later anyway, but pitching the concept to this thing can't hurt.

Go for it dude, I would try for sure. There can't be that much competition.

Definitely. There's really barely any german devs, especially small ones, that show any promise, despite the interest and market being there.

I'll be rooting for ya then, Hans.

Thanks, brah.

>

Yeah, for some reason on my monocolor placeholder texture floors it ends up looking like a fucking floodlight. When it's normal textures it looks significantly more natural.
I'm assuming that the lights are actually being rendered in additive translucency, and so the brighter color a texture is the more it pops out.

It can…soon.
There's a current system being developed by dpJudas in the vid that I'm pretty excited for.
youtube.com/watch?v=OmcvIcSP4Cc

well it's looking nice tbh

Holy fucking shit, after over a decade they are developing a cone lightning system. Now this is some blazing fast development right here.

I work full time, but I'm working on a project that I really like and want to see to completion, and I want to devote myself to it full time, i.e. make it a career instead of something I can spend 4 hours a week working on. Does /agdg/ have any sweet grants or ways to generate money so I can pay the bills while developing? Has anyone had a successful kickstarter or community funding here? Does anyone have any failure stories?

I'm not looking for millions of dollars or big crazy bucks. I just want to make it my fun day job that I legitimately enjoy doing, but I have no family to assist me.

if you enjoy it don't make it your job

Pretty much.
dpJudas is the savior that GZDoom has needed for a long, long, long time but is only just now getting.

I'm still alive, just more focused on programming questions which aren't as nice and easy to post.

It's your job to advertise yourself and get people familiar with your game/project/product. Until you gain traction, you want to post on social media, make a YT channel, make a FB group page, post on Twitter, Tumblr, even fucking Reddit or Imgur if you absolutely want to shill yourself. Their circlejerky nature and feel good attitude means you will get some attention regardless.

Set up a Patreon page or a Ko-Fi account to accept random donations, too. Create a "personality" and associate it with your game, but still be professional.

Fucking Bioware somehow manages to fleece the Canadian government out of 25% of its production costs on a few games, it can't be that hard.

Guys, i fucking suck at maths and need a little help. Fear not, my question remains very simple:

I have 2 points in 3D space: A and B, and i want to calculate the distance between the two so i can scale a mesh to create bullet tracers. Can anyone bully me the way to proceed?

This is elementary enough that your engine should be able to do it already. Subtract A from B and then take the length of that difference. The length is sqrt(x^2+y^2+z^2).

godot.readthedocs.io/en/2.1/learning/features/math/vector_math.html

i've seen lenght() around somewhere in godot but have yet to know how (or more when) to use it. Thanks for the hint. I'll re-read the docs.

mite sound like boomer advice, but go to a music shop and talk to anyone using workstation keyboards (employee or customer).
you want someone skilled with midi and samplers possibly used to working with old gear like the Korg M1 or Roland SoundCanvas.

its easier to gauge someone's style and skill when they do it live as you can test their ability to work within specific constraints.
musicians are often professional time wasters and I can't really help you with finding them as it requires social skills and every musician I know makes bedroom beats.

If I was going to put SNES style music in a game, I would link against fluidsynth (LGPLv2) and use a custom soundfont to get small music files and easy sound effects.

Given vectors A and B:

Vector FromAToB = A - B;
float LengthFromAToB = FromAToB.Length();

actually no, that'd be From B to A.

You subtract origin from destination.

B -> A is B-A
A

I'm literally retarded. I'm sorry, I just woke up.

Here we go again, last try:

>B A

By making retarded mistakes like this you make me feel slightly better about asking my pre-school tier question and let this thread do my homeworks. I ain't better.
Although i am currently in the midst of vector maths experimentations in 2D in the hope to implement it back in 3D, i ain't doing my homeworks but at least i am studying hard Thanks for the help user!

Welcome back.
Looking forward to more progress, I really like punchnun.

Same guy. user use subdivision. You will save LOADS of time down the road and it will come out actually smooth and even. Fix it now rather than later. It will fix your shoulders. I bet you were pulling and pushing individual vertices to get dat flat ass smoother. Do it faget.

Good news, the distance from A to B is the same as B to A. It doesn't actually matter which one you use, as long as you're consistent throughout your code (but the universal convention is generally B-A)

All you need to know about programming is the syntax the programming language you want to do and basic logic. That's all!

answer these questions:
what are you trying to achieve?
how are you doing it?
what is your program doing instead?

In hindsight, I feel like I should've just used pure white images and then changed the draw color over time

Just do that, changing the blend color of a draw isn't hard at all, and it would give you more colors overall and take up less texture data.

So I'm trying to use Data Grid on Gamemaker studio, and there's this in the manual.

What did they mean by this? I'm making a TES clone and am using DS grids for inventory system. When should I destroy my DS grid?

Also, should I use DS grid for the NPC inventory or will that cause data leak?

I assume that GM internally manages all its objects and resources like sprites and object instances. Based on the wording I assume that it creates unmanaged code to create the data structure. Basically dont create on every frame, be smart about using it, and when youre done with it, destroy it so its memory can be freed

Yes it does, because the vector you're getting from A to B and B to A has a direction.

You're welcome, my guy.

I only skimmed the post, I thought they only needed the length (which is identical)

Round two, made some changes to what I think people were pointing out. Started adding the feathers and attempted hair, will need to work on it. I think I made the legs seem double jointed? It's hard to tell since his armor covers a lot. Would love more critiques if you guys got 'em.

the legs are far too straight still and are abit too big on the ankle side of things

Anyone else unhealthily addicted to a game or know how to break one. I am and it's really fucking my progress. I have been trying quit fucking dota for years. I just have a nasty habbit of queueing up for a game even though I don't want to play. I always relapse the same fucking way. Overstressed from life shit with no escape. Other games don't see to really give me the escapism I need, unless it turns out to be an all time favorite like dungeon keeper, stalker, thief, or deus ex. And even then, once I beat them, I don't replay for years so I always end up gravitating back to fucking dota.

Time to start working on the Doomslayer's daggers. All I've done so far is get sprites made for his default, but I got 4 other dagger types and resources necessary to make sprites.

I got sick of people always getting angry at dota it just became unfun at that point tbh and I got pretty good at the game too

I wouldn't mind some slomo ability going along side those daggers

Sorry man, no slow mo here. I don't want to make things too easy, and even the default infinite dagger does okay damage, gives 10 HP on hit, and only has a 6 second cooldown

Which game engine gives you the most development freedom, /agdg/? I've got a few ideas I want to play around with but I don't want to pick an engine like godot or construct 2 and find out down the line what I'm trying to do won't work because of the chosen engine's limitations.

There are no limitations in any open source engine, since you can just add whatever you want to it yourself if need be. So just don't use a closed source one, I guess.

It's not fun for me either mate. I afk in half my games now because I don't have the will to even play them. I just get so burnt out by life or my game that I cave. Like I said. Most games don't give me that escape or don't last long enough.

Are you aiming to make a game you'd wanna play?

it's less about making things easier and more opening things up for hard

warframe helps me alittle but I've found that a really good tabletop pnp does the job better

I got special plans for the nightmare difficulty, it won't happen in this update but definitely in the future.
Plus I'm not big on slowmo mechanics unless they involve some seriously exaggerated physics to go along with it, and I don't even want to think about trying to pull something like that off in GZdoom, AND make sure it's compatible in Zandronum 3.0

Yeah. It's the light at the end of the tunnel for me. but I get burnt out a lot. Usually because there's life shit I procrastinate on and I don't want to work on my game until that's sorted, but I'm already strung out from grinding away at stuff for weeks on end so I just take a dota break.

I know that feeling, user. I know that feeling all too well, but you gotta keep working at it. Even if that means straightening out your life first.

I just need a solution for being addicted to dota

I haven't tried PnP. The closest I came was an RP server on NWN. It was fun because it was essentially tabletop RP with a whole server of people. Too bad it was too cliquey

Walking through the logic really helped. In this case, when a fish was reeled in and came off the lure, the lure would erroneously maintain the pointer to the fish rather than correctly reverting to nil (resulting in bizarre behavior of course). It was a simple enough fix in the end.

pic unrelated

...

are you sure it is not denial?
reddit.com/r/AskReddit/comments/7pny5m/formerly_lazy_people_what_things_did_you_do_to/

My solution came after the realisation that making bitter people from dota lose their games on purpose was more worth it than actually playing this game for real. I got sick by this shitty community for saying "4v5 doesn't work techies u fuckin faggot", thinking they know absolutely everything in a very complex and situational game such as dota, while i randomed the taliban son of a bitch.
I too have have real life psychological problems as well and they can cometimes be crippling since they are problems that lingered troughout all my life. I don't know if that's your case, but Gamedev helped me alot to not feel like shit by taking my mind off the fact that i tried to KMS some years ago and overall be better by making sure i have a composed routine, and most importantly: goals. Once you have a set, solid idea playing in your mind, you'll love it so much that you'll do anything to pull it out of your mind and materialize it into a real thing.Also a little Jesus really doesn't hurt. Pray for intelligence if that's your thing as well. Last but not least, if you want to deal with your shit and make a game at the same time, why not make a game about and/or inspired by your shit? You know they say some great artists were depressed and melancholic as fuck when they became famous for their work. And if you think that will hurt your reputation, who the fuck cares? If you really believe you cannot sink lower than that, then this means anything worthy you'll do WILL be worthy for you. For me, that gives me back hope and a little sanity.

Good luck user, and godspeed.

...

Never do that again

BAKER WHERE IS THE BREAD

NEW