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

I just wasn't expecting to see him in that comic.

Fairly sure that's an edit.

some uk fag tried to kill trump
it was a weak attempt, but still

Someone also fucked with the wheels of his car once, like what they did to Nigel.

Getting back into the pixel art after last project fucked me up for a year. Just some loose silhouette sketching, then converting into characters. The point is to create something you'd WANT to pick as a player character, AND to make it clear just by the look what each guy can do.

1 and 4 are still WIP, for 2 and 3 I figure just some added touches will do.

Any good idea what to do with 4, specifically with the white body? I was thinking maybe a dot matrix display. Or something more mechanical?

That doesn't change the fact that that is an edit.

Good shit, man. I really like the colors, especially on #2.
I should work more with silhouettes, my proportions end up lopsided more often than not.

That happens, but working from silhouette makes you realize where stuff looks bad. Check step 2 to step 3 on #2, I needed to put the arm back a little.

Also remember to flip your image regularly, that'll really bring out any imbalances.

My process is kinda similar but I end up skipping step 1 without thinking about. Which usually isn't a problem, but right now I'm trying to make a mockup for a Wild Guns clone and the perspective gives me trouble. Doing the silhouettes after made me realize the sprites aren't as clear as they could be.
Yeah, I've been doing that ever since I found out about it.

This gun doesn't look right.

You start off too big I think.
It looks like you're aiming for very large characters but aren't really sure where to apply details, and how much. I like both sprites, but I feel like there's no particular point that draws the eye. Evil hatman somehow does this with the headscarf, but then the pants have the same kind of contrast and draw away the eye.

Yellow sombrero has his hat that kinda works as the focus, but not much going on below that (also I'd definitely restore the full yellow hat ring from #2). I'd suggest picking between hair and back, and making one detailed and the other simple. Alternatively, a nice belt may help the character balance.

Check the shading on the left leg too. I realize it's supposed to be facing away, but at least the ass, and maybe the heel, should receive some light.

Also maybe try to do the silhouette at half size, it'll help you decide proportions and accents better. Then scale it up and do the rest.

You're probably spot on with that, I'm used to drawing smaller sprites.
I didn't think of most of those things, thank you for the help. I'll try to do them over and consider this stuff.

For what it's worth, I think the side views look really goddamn fucking good.

Just two minor things: Maybe Grave Diggers's scarf could look a little more cloth-like with correct drapes and creases, and I'm not sure how the big guy is supposed to be holding his gun.

As for the rest though, that's really really impressive if you ask me.

Appreciate it. I'd fix what you suggested but that project fell through months ago.
I have a story premise for these characters in my head that I feel has a lot of potential, but right now I'm just trying new things for the sake of practice.

Finished 1 and 4. It's more or less done except maybe some outline shading.

Good stuff. Making me feel bad for crappy sprites to make it in time for Halloween.

I tried applying your suggestions and I'm already much happier with the result. Thanks again.

That's just me, but I think it looks amazing.

looks excellent
keep up the good work devvie

nice improvement, looks stellar

RoR-esque, I really like it

Need some input anons, as perceptions are a tricky thing (sight specifically here).

So I've been messing around with my shaders a bit, doing some parallax (basically, 3D'ish appearance of bump map at angles via heightmap projection of texture coordinates, in the end it's lighting effects), and experimenting with my own approach.

I'm curious, does the rock on the left of these images look parallax to you too, or in otherwords, 3D where the texture is?
Also, last image is for comparison (solely textures, nothing else).

woops, forgot texture only image for comparison

Too noisy, I can't see shit. Try it with these maps.

Thanks lads.

It's an outcome of how it's done, but I can make it less noisy if I use a central differencing method for my derivatives instead of forward differencing.

Nice trips.

I hate that cutesy hipster art.

Once again, I'm willing to voice act for free for anyone who wants it. Email me at [email protected]/* */

I have a question about 3D

when is a model considered plagiarism?

for example, I use a bunch of body parts from artists I like and mash them together in one picture/reference.
and then I make a model all on my own using that reference.

I know that if you didn't make the model or downloaded somewhere else and make money off of it, that would be illegal.

Using references is not plagiarism, it's being an artist.

Do you have any references you can post?

They keep getting pushed to some aprt in the screen, where the bottom area is sent to the middle. When I turn on "view collision shapes" in debug it says they never moved. This shouldn't be happening, I'm starting to think this is an actual bug with the engine.

Fixed that for him

Im here faggots, finished the first test version, i will add more info on the ingredients and more output feedback later, im quite tired right now

filedropper.com/lostcook

ideal tastes in a dish overall are
saltiness, sweetness, sourness and bitterness: balanced at 60, more than that becomes unpleasant
savoriness can go all the way to 100
most characters will appreciate spiciness and mintiness if they are below 40

Nutrients are a mere abstraction

that literally could have been done in 100 lines of code

how many did you use? 1000? 2000?

i deleted everything and farted at the computer, it worked somehow, im going with it

upload the project so far
you won't, faggot

the source, i mean

im not implying you want to steal it, but i ain't giving you fags more bully ammo so easily

I'll condense your abomination to 100 lines or your money back. seriously. I'm willing to put my autism where my mouth is.

dude, everyone knows you can put one command right after another in the same line

I'm not as much of a faggot as you are

stop avatarfagging then

...

bothers you?

post your code
you won't because you know you spent days writing thousands of lines that could be condensed to a hundred

Concept art. The best thing about it is there's virtually no difference between the concept art and the final product; assets are literally copypasted straight into the engine. The only real difference is in size and onscreen position.

i'm surprised this isn't dead

its possible, but right now im in a tight situation and can't afford rewriting
I need to finish a good demo of the cooking system and the faerie village while my brother builds the map and put a character in the world so i can have something to show in the kickstarter video
I will probably lose my job soon because the company is doing really bad right now, i escaped the first wave of layoffs, but im on the list for the next if things don't improve
Since i live in fucking Brazil and our economy is pretty much fucked getting a new job will be hell once i lose this, so instead i want to make a living with crowdfunding making my dream game, i do have a plan and a market to pander to, if i take 3 years to make it and only sell 5000 copies at 15 bucks (on steam) i will still be making more than i would make getting a programmer job arround here.

Post yours.

Muh sides.
Shit looks wicked though.

that looks neat

you need realistic expectations, street shittter. Odds are you'll sell like 20 copies at $10

I know I did

On that note, I sold another 2 copies this month!

what is the genre of your game? also how it differs from others of that genre?

soundcloud.com/sonicdahedgie

A couple imitations I tried awhile back. I have a better mic and can do better voice now but that should give you an idea. I'm honestly looking to have some small jobs specifically to practice my craft.

I'd be practically rewriting it for you to prove my autism is superior. It's not like you'd lose anything.

Nothing, really, I'm selling shit on the unity store. Turns out most nodevs would rather spend 20 hours than $10.

It's way easier to sell copies of a merely decent game on steam than of a really good asset on the Unity store.

probably

Project SocJus needs a Mode-7 gameplay sequence with Vivian on the bike

One should never turn down a free refactor.

I've changed the reload animation on my gun like 5 times, I can't just get the way I want it :/

Fuck everything.

Starting idle animation with the stuff that counts.

nice designs

Added a score, goes from 0 to 100

wew art

This is basically what doom did for many of their sprites. They just used composite images of toy/machinery/landscapes instead of other sprites.

I recreated the project and have the same issue still

holy fuck loading a bolt action rifle like a shotgun. fucking libtards.

Break action rifles do exist but they are nowhere near as popular as bolt action for obvious reasons.

A lot of large-bore hunting rifles are break action. However, the image in that comic shows the action separating effectively right over where the bridge of your hand would be. In addition, the gun also magically gains what I think is supposed to be a bolt in the next frame.

As sort of an ironic note, a huge chunk of the assassinations named in the comic were carried out by leftist and/or democrats.

If anyone wanna try out the new version

filedropper.com/cooking

Just extract the zip and click the .exe
i hope you have a good anti-virus

Alright, so I finally went ahead and remade my grid generator. It creates a rhomboid offset grid of hexes, and then it disables all of the ones I don't need, thus creating a hex-shaped grid of hexes.
For testing purposes, so I can try to understand and get the most out of that amazing hex guide (redblobgames.com/grids/hexagons/), I have my debug layer displaying the offset coordinates in the center of each hex (odd-r), and the y,x,z cube coordinates (in the same order as the example pic clockwise from the top left) in the corners of each hex.

I used the coordinate conversion equations in the guide, but the resulting coordinates look very, very wrong. The center hex should be 0,0,0 for starters.

Does anyone know where I messed up or what I'm misunderstanding about this? I get that, strictly speaking, I don't need to muck about with these coordinates and I can just use my array of hexes to do things as if they were a square grid, but I'd really like to understand what this guy is talking about and why using cube coordinates is supposedly vastly superior.

Hahaha holy fuck you're right. Dumbass liberals

It occurred to me that it might be off because I generated my grid in a rhomboid shape, which invalidated certain aspects of the rectangular offset grid in the example.
I created another script to do the grid as specified before converting to cube coords, but the cube conversion results are still not right.

I made the offset axes run from positive-negative half-width and half-height, and now after conversion, the cube's origin is exactly where it should be, but now the coordinates don't match up along the cube axes. I'm puzzled as fuck, but it feels like I'm close to a solution maybe.

my fix to my fix to my fix accidentally altered my grid's layout from odd-r to even-r

So you guys have a idea what would be a good engine for a 2d cuhrazy beat em up?
I have a pretty shit pc but can play most games fine on max settings 2002 games anyways
any ideas?

any one of them would do fine?
pygame, lwjgl or game engine if you're looking for specific examples i guess

Just something good with sprites i guess

literally any game engine handles sprites
if "something good with sprites i guess" is honestly your only criteria then just pick one that you like the name of and use it.

Godot has a very developed 2D toolset that doesn't require advanced programming knowledge. Love2D is also a good engine (I've heard) that uses Lua, a rather high level language that isn't hard to get into. You sound like you're very new to gamedev, but don't let that hold you back. Everyone starts somewhere.

i think im retarded

does the loli kidnapping sim dev still post here?

...

I've only seen him make a thread on /agdg/ when it became a thing, never seen him post here.

A shitty looking game some fag had been working on for years. It gained meme status on halfchan when I still went there.

I still think it was all mockups.

some of these combinations would be neat

Cosmic horrors are the only real convincingly unstoppable enemy I've ever seen. They just wear the "There is absolutely nothing you can do to kill this thing, the best hope you have is to seal it off" so well.

...

...

Many villains can be played up correctly to actually seem convincingly unstoppable.
Dio is one of my favorite examples of this.
Time stop is an absolutely overpowered ability, and Araki takes this to the farthest end when writing part 3. In general, Araki takes seemingly mundane and overdone abilities only to play them up to their respective glory.
(cosmic reset)
(time rewind)
(complete nullification/time skipping)
(infinite force through spinning shit radically)

The one thing I haven't seen done well is a Cosmic Horror villain, besides some novel books and some media.
I hate Undertale as what it is and what surrounds it, but the Flowey fight was the only decent and cool part of the game. (tho its still faggy tbh)
It took an "Omnipotent"/horror boss and dissected and expanded upon familiar movement and battle mechanics, to both accent the fight and provide the ability to actually test one's ability.
However it fell seriously short. Flowey's fight was extremely easy, even compared to the rest of the game. Flowey had a really small arsenal, and while that did help with being able to predict his attacks, it caused his intimidating appearance to look for naught.
Overall, a omnipotent villian with a good boss battle is still upcoming, there hasn't been many games that could pull this off, and well at that.

When I meant cosmic horror, I meant something along the lines of in the pic. Basically some other wordly monster with so much sheer power that you can't even fight it directly. Stuff like stopping a cult from summoning it or it not even being in the world but still putting it's influence on it. The only examples I can think of are Half-life 2, where the combine take over the world in 7 hours and you don't even know what else there is beyond them and Yahtzee's random games.

Also shadow of the colossus does this

Well for a nigh unstoppable cosmic horror it sure as fuck controlled like shit.

Yeah, idk if they had a cannon explanation for that.

Since cooking user is too shy to post his code
Why don't you take a look at mine.
Theres a switch statement that has over a hundred cases in it
And various other shit that just plain bad.
gitgud.io/greenman/rts.git

For what purpose?

Hotkey parsing

I only found one with ~90 cases, for a bunch of control groups.
Does this mean you can only select 16 groups of units in this game?

no you can select infinite amount of units

All that stuff is just hotkeys for creating and recalling control groups
Also you're right there isn't actually one with over a hundred cases i split it up into three seperate switch cases because reasons.

it's looking real rad man, I will be looking forward to it.

Thanks man you know just because of that im gonna work on my game today

I find it a bit funny that your game is going to be Half-Life 2 like because there is a RTS game for spring named Journey War which is Half-Life 2 inspired through instead of a Rebel groups there is some sort of hippies mutant like things or w/e they are named correctly.

Theres also a straight up half life rts fan project called lambda wars
but it fucking sucks has no original ideas at all
Also I might not end up using half life 2 models at all at which point the inspiration becomes much less obvious i think

hmm is the gameplay wise of that lambda wars good at least?

I only played it briefly and I remember being pretty disappointed by it

Which graphics software allows me to quickly make level design/mockups using a tileset?

tiled

What would you like to see in a sprite-based ARPG hack'n'slash?

Dont think about limitations, just about you personally would find fun.

we could always use more kamakazi japs in this world

an army of player-controlled skeletons

An army of skeleton-controlled players.

whoops, i meant player-controlled skeletons

skeleton controlled players is better imo

Don't correct me, faggit.

huh, i didn't notice someone already corrected me here

wait fuck
i was right the first time
you confused me with your duplicate post
fuck you nigger
player-controlled skeletons or bust

Thats weird, I made a post about skeletons in my own project, but doesn't seem to be appearing in this thread.

I was browsing Holla Forums, and I think I accidentally made a post about my project on Holla Forums now

THE SKELETONS ARE FIGHTING BACK

so… nuffin?

It's an extremely close zoom in of a skeleton. I didn't fuck up and accidentally open oekaki at all.

Speaking of skeletons, I decided previous 16x16 sprites were pretty shitty so instead I'm redoing them heavy based off a CC-BY-SA set. Also made a screen to pick your class and show differences. Need to draw some class emblems or something at the top to fill them out.

2spooky4me

When I draw sprites I start with the outline, just like you'd normally sketch something with a pencil. Is it normal to actually draw the shadow/silhouette first without knowing the exact placement of every body part?

You should know where all the parts go in your head

It's hard to imagine for me without drawing the actual shapes instead one by one, instead of a whole silhouette.

I've been working on a 2D game for about 2 years with another artist. We have a very solid work and friend relationship so we have never run into any drama or issues. The game is coded in C++ and OpenGL using my own engine and map editor. Here is a video from half a year ago. There's an indiedb page I have but only the other dude writes and blogs about it. Which sometimes leads to cringy posts but I don't care enough about the page yet anyways.

There's no music or sounds yet so i usually dub videos with random pandora music.

www.indiedb.com/games/armoin/videos/first-cinematic

Thoughts?

the sword direction should be controlled by the mouse, not animation. like in arcane the armor collector
also the fog looks weird

That can certainly work, but what I mean is that as you draw out a silhouette your brain should fill in all the details of where stuff goes. Because it's just a silhouette it's very easy to change and fix- even easier than redrawing lines for each part- before you add any amount of detail.

Personally I always use a single outline or silhouette for pixel art, but with traditional drawing I always just draw in the pieces.

also, is there any realtime terrain deformation, like in worms games? technically that would be a thing for wizard units, but it would be neat

That's not fog that was a video showing a bug. Yeah I haven't figured out the controls for the game yet. There's a lot more types of attacks not shown in that video.

You aren't the first one that has suggested something like that but for the type of game that we are going for, sort of an RPG adventure game, it's not required and thus I won't bother implementing it.

come on, at least do it for a boss or something

I think if you focus on gesture and the pose instead of anatomical details it's not that hard. You'd only need the main masses (head, torso, pelvis, tits) at first; the limbs can even just be some curves at the beginning, only for you to add details later on.

No. It would change the gameplay too much at this point and it would be a lot of work. There are navigation meshes that are generated with maps right now so it would be a nightmare to make a real time random poly nav tracer.

Realtime terrain deformation is hardly a trivial feature to add. Look at something like OpenClonk; they use a bitmap using what I think is a marching squares algorithm to make it look nice, and even then there are a lot of edge cases like 1px floating pieces of terrain that act very screwy when the player tries to move over them.

Alright, I suppose if it works for you it might work for me, so I'll try it. I draw in pieces on paper obviously, but maybe this silhouette thing will work with pixel art.

Lesson learned, I will focus on drawing the tits first.

oh, alright then, i already had to change my pathfinding from navmesh to A* in order to include resources that block unit paths, so i'd imagine it would be an issue

This especially. Creating AI that can effectively maneuver a platforer environment isn't easy, but making one that can move around a changing 2d platform environment is especially difficult.

Navmesh isn't an algorithm. You can use A* on a navmesh. Navmesh is just fancy name for a graph used by an AI.

i mean i'm not using unity's navmesh, but using a grid-based A*, since i can easily set/get what's passable or not there

...

A web dev nu-male's life is tough.

...

the issue is that it usually mentions that there's SOMETHING wrong in the line, but not what in particular
what happens if you have like 5 variables in the line that could each be null? you have to make 5 different checks prior said line to see if each variable is null or not
which would have been easier for the program to simply tell you in the first place

You know what's even better? Taking care of bugs before they occur. :^)

what am i, a fucking prophet? i can't magically find every single bug before even testing once, and they're bound to occur because nobody's perfect. not even you

Compiler errors usually have columns as well as line numbers. I think the C# compiler might.

...

...

But god made lots of mistakes. Me being alive, for a start.

that just proves that it's impossible to not make mistakes, eventually

Click Debug and hover your fucking mouse over them.

Is waking up on a beach too cliche to start a topdown rpg game with?

And niggers, don't forget about niggers. They were a mistake.

Yeah

Still working on my Gynoid, how does it look?

why does it have nipples if it's a robot?

Why not? Clang's everywhere my friend.

Very smooth. You might want to add some more mechanical parts, unless you're going got an "everything is encased in silicon/plastic" kind of look.

I will use joints, not everything will be encased in rubber, only the "needed" stuff considering the excuse of a backstory this would have.

Well, why do humans have nipples?

because we've got babies to feed
do robots?

They don't have to feed only babies.

Power and reset buttons.

It's already bad enough that you need a human shape with tits, perfectly smooth and round hard plastic or bust. Your clang is weak.

Anyone know the easiest/idiomatic way to pass info from scene to scene in godot? LIke say I had a title screen and wanted to pass the selected difficulty as a variable before loading my dungeon scene?

mhm

...

...

...

Use a debugger, you dumdum.

was it on purpose, Holla Forums?

I've been using Godot for a while but haven't gotten to switching between scenes yet. Apparently you make a global script class which you dump the variables into that you need transferred.

godotengine.org/qa/1883/transfering-a-variable-over-to-another-scene

If that's a piglamist then where is his designated semen dump goat?

When implementing A*, should the pathfinding algorithm run from some object that controls the grid itself, or should it run from each individual object that will be running the algorithm?

I guess what I'm asking is, is it better to have many objects running the same algorithm, or a single object running the same algorithm many times simultaneously?
Or does it not matter at all?

How much Chromatic aberration is too much?

I hope it has some dank vaporwave music

Anything stronger than off.

Is this better?

Transcendence of the self
GOTY 2016

very good

It's like a metronome

I used the sin of time so it makes sense that it would oscillate like a metronome.

Hotline Miami 3D

Like Streets of Rage? Then use OpenBOR

Thanks for the link, lots of helpful info. Not sure I like the idea of using a singleton for this myself, but the last answer in particular looks like it might be worthwhile, letting you open a scene with arguments.
I like godot a lot so far, but coming from a 100% programming background some things definitely seem convoluted for the sake of "muh designers and le artists working with muh engine"

Recent Progress

To-Do List

After That?

I put it on the script/object that controls the grid itself, but it's really a matter of preference. That said, the pathfinding changes I mentioned above were scrapping the entirety of the A* code in favor of a variant of Dijkstra.

Pic related is what I'm working with at the moment. A* is great for real time movement and games where units can move anywhere. With what I'm doing though (turn based on a grid system), I have to calculate the distance to each square in range to find out what moves are valid anyways, so I was already doing like 90% of the work Dijkstra's Algorithm does and using A* from there was pretty redundant.

No clue what you're doing, so it's possible that A* is the better choice for you, but it's something worth thinking about now potentially save you some time and effort later.

Nice artwork. Keep it up

It's all placeholder artwork stolen from other games. I'm just codemonkeying together a prototype at the moment. The relevant bit in the screenshot is shaded squares.

Two actions per turn. Cyan squares can be reached using one action. Red can be reached using both. The character in the screenshot can move 3 spaces per action. Diagonal moves cost half an action more. The picture is the area that character can move to assuming all squares cost one point of movement (no difficult terrain) and there are no obstacles in the way.

>Diagonal moves cost half an action a point of movement more

sub 80 IQ mongoloid here, does anyone have any experience with darkbasic or gameguru?

are they ok for retards like me to make shitty meme games?

I'm more interested in how you found those particular engines and why you think they'd be better than the mainstream options.

If something advertises itself as EZ PZ NO WORK LOL U HAVE GAME LOL EZ! then I'd avoid it like the plague.

Hey, anyone here know how companies like Autodesk know if models used in games were made from a student version of 3dsMax? Do they have any idea if it was made from something like a pirated version, too? Would I be able to just export the model as an .obj, then export it to blender, and save it from there?

For storing a bunch of coordinates, what would be the advantages of one over the other of these two options?:

Would there be any performance or memory difference, assuming you're going to be storing a gajillion gorillion godzillion coordinates and accessing it very rapidly?

interested on this too, i want to trick the 3D jew

Depends. Do you need to access say, x3 in constant time? Yes? Put it in a hash table. If it's just a linear sequence that goes from start to finish, then you might not need O(1) lookup.

Not only are they not going to check your game but they can't know, if you're that paranoid recompile it in another program.

In this case it'll just loop from start to finish, but the array is expanded, rebuilt and looped though a lot in realtime.

The vectors would be fewer index accesses, as long as you usually need both x and y.
Memory wise it's the same, as long as you need both.

Not sure about cache misses and all that stuff… you'd have to wait for the DOD user to answer questions related to that.

For what purpose?

Well, I'd seriously look into restructuring your array into a hash table, especially if you're going to be rapidly restructuring elements say in the middle of the array or worse. If you keep everything in an array you'll have to shift everything resulting in O(n) or worse performance. If you put it in a hash table you should be able to get O(1) for every single entry in your table. Generally, stay away from arrays unless they're under 100 items or so for performance reasons. Hash tables are really king for performance for any large data set.

(Someone please correct me if I'm blatantly wrong, btw)

also is me ,fucking vpn changing ids on me all the damn time. if you have any questions i'll try and answer

Butterflies.

Experimenting with actors and entities that aren't allied to the player but also don't particularly give a shit to the player.
Which is basically a lot of words to basically say "factionless NPCs". Since there's basically only two types of NPCs (+FRIENDLY and MONSTER), this should allow for populating maps with harmless wildlife.

I won't have to change anything in the middle, by "rebuilt" I meant "replaced", new coordinates are only added to the end. If it loops to the end, it'll just stop until the array is replaced. If it's replaced, it restarts.

Something similar to a flood fill.

Fuck me I'm tired.

Only if you have to manipulate the data a lot. If you want to just iterate over the whole set anyway, array's are fine. As well, as when you load a model or something that you pass to the graphics card. For stuff like this, raw arrays are king.

Tru, tru.

erm, that user is iterating through the values, and only needs a simple collection.
No hashing is necessary, which is unnecessary overhead for that user's purposes; or at least from what I can tell.

The issue I would assume is expanding the array, or instead a better solution is having a static sized array with an upper size bounds (as long as the size is predicable, this is a great solution); once you reach that use another static sized array (pooled array) for the excess values or since that user is rebuilding the array just place it in the "next size up" static sized array that's pooled.

huh?
Arrays are great.
As long as you're not resizing them frequently, or doing what it's not meant for; i.e. doing that which other collections are purpose built to do, and the array is not.

Hash sets are purpose built for specific problems, as are all collections.
They're not the "one size fits all collection" user, no collection out there is "the best"; it's a matter of the right tool for the job.

neat user, it's a nice touch basically :^)

I'm still very excited to be working on this project, and very happy that the guy likes my stuff, but I kind of wish I could quit my job and just work on music full-time. How do fellow working anons manage their time between work, leisure, and work for a project?

Insomnia helps.

nice, any chance you could drop one of your "weird samples" tracks? I'm curious now.

There's your problem, right there, it's not just work/composition/leisure… it's work/composition/leisure/life/other shit.

This too thou and not sleeping for a few days in a row; drink lots of caffeine.
Once you get past day 2 or 3, you get pretty awake, and super energized + creative.

sage for clarification: "not sleeping for a few days in a row", i.e. working on your project for a few days in a row during your days off, make sure they're sequentially ordered days off for best effect

nice to see some classcucks working themselves to death

soundcloud.com/zawazawazawachameleon/i-am-a-nice-fellow

and the most recent

soundcloud.com/zawazawazawachameleon/loud-proud-and-rowdy

On top of a pile of unfinished traditional songs that I was trying to make for somebody's roguelike (sorry if I haven't been on steam lately, I'm at work most of the time on nights, if you're in this thread) And to make matters worse, I can't drink nearly as much caffeine as I used to, due to a recent string of health issues that I just now have under control (Colon troubles)

though, ain't complaining, really excited to step out of my comfort zone with this project. But I'd thought I'd ask the devs here how they deal though, lord knows, the amount of mental anguish is 10x what I'm dealing with, really respect the work ethic.

But A* is just dijkstra with a heuristic

You kinda are
Arrays are great for sequential acces
As long as you are not removing and adding elements in random positions an array is faster.
It takes a little more effort but you have O(1) to any element in an array too you just need to keep track of your indices
With that said hash tables are only slightly slower and quite a bit more userfriendly in my opinion
And definitely faster if you are removing/adding a lot of elements
Note that if you are rebuilding the array/hashtable arrays are faster

kek, nice.
I like the boss theme, and dashing rouge though; they're quality.

oh damn, sounds ruff. Glad you got it under control user.

That's what I read too, when I was reading about A* ages ago

install blendoo

A* only gives a shortest path to one tile, Dijkstra can be used to get paths to all tiles within a certain distance. It's one of the only things about Dijkstra that's better.

Nice stuff as usual Zawa, glad to see you've got a paying gig doing what you like. New song sounds pretty LISA-like, yet unique. I like it.
Is there a full source to ? Google gives me nothing.

There's multiple methods supposedly used, most of which I don't know about, but one of them is hiding information in the least significant bits of your vertex data, and at various other positions in the file. Just re-exporting with Blender isn't going to get rid of that. But like said, there's a decent chance they'll never know, and buying an autodesk license as soon as your game becomes big enough to make them consider checking your stuff should stop them from caring anyway.

That's true actually i didn't think of that

Hey guys i LOVE Tetrachroma can WE sell it to a publisher?? Since nobody owns the assets i can take them and just keep the money. I'd be giving you guys a real completed game to do this. I just want to note that I'm not asking permission, this is happening. I'm going to claim ownership to all Tetrachroma soon and sell them. THANKS

Is this a meme now?

I know this is just a parody of that sandnigger who wants to do that to A&CP, but it doesn't make sense in this case since we don't even have access to any of Tetrachrome's assets and therefore no porn of the life like textures ;_;.

I think the dubs says yes.

is /agdg/ down?

(czeched)
Yup, it's giving the same 404 page threads do when they die. Still exists on the boardist, though.

same with /art/, codedonkey is breaking things again

probably just a hiccup

Digging your tunes. That reminded me that G4F Records, who produced Furi's soundtrack never got back to me. Jerks.

does anyone have a sprites of a fat guy i can steal

Oh my god.

Your shit is weird but cool, reminds me a lot of LISA's soundtrack.

What if I used a pirated, unregistered version?

we have the horse head.

That thing could make tetradev some dosh on the asset store now that I think about it

I wanted to do that for do that for my own project which hasn't progressed beyond reading engine documentation, so all power to you in the meantime.
I could probably start at any time, I'm just getting cold feet because it would be my first 3D vidya.

What are your plans for your 3D Vidya/Engine?

To test an experimental, distinctly videogamey storytelling technique I haven't seen anyone else try yet at least in a shootan game, implement Radiance Hints and design the mechanics/level architecture around using realtime GI to its fullest.
Theoretically it should be easier to make than a more standard shootan game if I have the right mindset and don't overextend myself with certain things, but all that is up in the air until I make the damn thing.

hmm, maybe try to get a few basic things going first that are "easy" to do?

That's my plan.

(checked)
Well good luck.

Basically what said.

With A* I need to know the starting point and the ending point and then it will calculate the best path from there, But if you look pic related, at this point, the end point could be any of the highlighted grid spaces. Until the player chooses one, there's no way of knowing which it will be.

I already have to calculate the distance to each of those spaces to determine whether the move is within the character's movement range. Because of that, I'm already doing like 90% of the work Dijkstra's Algorithm does. Given that, it makes A* redundant. All I had to do was add a line or two of code to keep track of which space led to which other one while finding the movement range and I was able to cut out the A* code entirely.

Important game dev question:
How do you learn to drink water?

I haven't drunk or eaten anything today after work because I've been so focused on dev. It keeps happening.

Alarm clock.

1. Have water nearby
2. Drink from it

Once you cross the threshold of no longer requiring food or water, you're golden.

Go with 's advice. Have a glass or water bottle nearby and make drinking from it a reflexive action you take while doing stuff at your desk.

This model needs to be animated then I need to do the 2 other models (light attack spider) then I am done for now and then… I dont know.

VHS and Pause Effect.

more of the new level I am working on

reminds me of this

neato

...

What a shame m8.

That stuff's really cool. I can see it going alongside stuff in JSR. Keep it up m8.

The pause effect looks fucking sweet but the "VHS" effect doesn't quite do it for me, unless I'm misunderstanding which effect that's supposed to be. I think it's because the shifting is so constant and regular.
Have you considered connecting it to the camera movement or angle in some way, so that it only shifts as the camera turns? I think if implemented right that would make it more subtle, but more effective. Just a thought.

I'll look into that. This is just the first draft of the VHS effect and over time I'll be trying to make it look better. At the moment there's no random information used to make the effect, it's just some sin waves put though a bunch of other math and overlapped over each other. That's probably why it seems so regular.

I figured it was a sine wav, of time, right?
So I'm thinking try the sine of the camera angle instead, dividing or multiplying the angle first if the spin is too fast or slow.

Sort of. Ultimately but not directly. There's a lot of stuff going on in the effect. If you are interested I can just post it.
But the only reason the distortion lines are moving at all is because they are attached to a value that is constantly changing. Are you saying that if the camera isn't moving the distortion lines shouldn't move? For a camera that doesn't rotate it would look indistinguishable from this screenshot.

Not that they necessarily shouldn't, just that it would be a little more subtle and maybe more interesting if it was was attached to something. I didn't mean it to sound definitive. I think what I'm picturing in my head is something like the feed from a cheap pre-digital security cam as it pans, which may or may not be something you'd want.

It's fine. I'm not saying it doesn't need work, it does, but tying the effect to camera movement isn't exactly what we are going for. I do plan on making it much more subtle as I go on.

Just make it lower frequency, and it'll be subtle.
The higher the frequency, the more aliasing-esque effect is going on.

People other than AnCP developers have access to their assets?

Unless they've uploaded it to a public repository, with an explicit "public domain" type of legal jargon nobody can legally obtain the rights to their assets.
In this case, if it's public domain, I believe one could claim w/e was made by them after the fact, as derivative works… it depends though.
Such as, editing a model - that's in the public domain - then claiming ownership over said edited model as derivative works, however, it has to be substantially different from the original to claim ownership… look at the Theseus' paradox for a good analogy as to the issues on this.
Note that if they didn't say anything relating to a license, and it was uploaded to a public repository, it's automatically copyrighted.
So, if someone comes along and tries to claim ownership… the original owners (those who hold the "implicit" copyright) can assert they own the assets, i.e. if they were never explicitly licensed (we're talking about if they're the original proprietors of said property, be it code, models, etc); then those original owners can claim ownership of said assets making that outside person's claims of ownership null and void.

Although, something like the actual "name" or "brand" of AnCP; I'm not too sure about…

Somebody correct me here if I'm wrong, but I think that's right.

Said turkroach says he has a way to get their assets, his second post in the thread:

It doesn't matter how he gets the assests like said
As long as there is some proof you made those assets like an online repository you own it.
There is no way he can just take just because it's not licensed.
The brand he might be able to take but even that can be contested.
Laws are really not so ridiculous to allow for something like this turkroach is full of shit pretty sure he is just baiting.
And if he isn't he is completely delusional.

Wasn't there some guy working on a warfare mod for Factiorio here? Because Factorio is getting an official PvP mode now.

You summoned me?
I'll be blunt, I am more of a PvE player type fam.

Well user doesn't really matter as long as we get more shooty things, if they can kill biters they can kill players.

Well alright I was just a bit worried about that because the T-55 can be loaded with a OP rocket ammo which I did that so that it can face on the Behemoth bugs, through it should be replaced with a biological type of ammo so that it won't be able to damage structures much.

If somebody wants to continue on that mod well that's fine by me, I might even provide assistance. My focus is now on my Micro Tank mod.

I had some few difficulty with the 0.13 release because I only got it from the Factorio.su site which is why I don't maintain it anymore, I made plenty of post explaining why that is the case.

why live

Does it say what version it is?

blender says the file is 6100, only supports 7100 and up

and now i have to reinstal maya, since for some reason it's trying to make me reactivate it, but just opens up an html file in my browser, with no options for anything to do
and this is after i've activated it a year or so ago

hmm, maybe try using a few older Blender version then if it imported successfully save as a blend file and try to open it with a newer Blender version. Check the date of your model it should be written somewhere and then use a approximate Blender version for that.

Dubs decides my next addition to the game.

(checked)
A explosive ball enemy

use an animated character instead of a ball

pickles

might do that later, i've already found that unity has a prefab explosion

off by one

a dung beetle that's rolling the ball

something that shoots fire

a dungbettle character that shoots fire

fucking make exploding balls now

dungbettle character that shots fire and rolls explosive balls

okay but where can i get a dungbettle asset? the rest i can figure out how2make

JUST DO IT

Download the Sintel model and put her into this position

My past few hours
Fuck. It's great when you can just compile and play, but debugging is a horrible experience.

Pic related

can't do that in an atheist game, user!

Why not? She's wearing pants.

The shit is fucked

rate and subscribe

Fuck, how do you design "puzzles"?

I'm trying to design a level with puzzles that are along the lines of finding the right items to use in the right places in the right order, and finding extra items that help you understand what to do and get some "lore" about your surroundings.

But it's turning into a huge clusterfuck when I try to write down lines of what goes where and what each point of interest is.

That's not a puzzle, that's tedious padding to make it longer. Going from A to B to unlock C is lame. It has nothing to do with interactions of game objects

I don't know how else to describe it. A "quest"?

You can't just walk from one item to the next, you'll have to read notes and listen to conversations to figure out how to get certain things or bypass obstacles.

Chances are you can simplify it down to nodes and lines connecting those nodes. Once you do that you can make use of graph theory.
en.wikipedia.org/wiki/Graph_theory
Removing all of the bells and whistles means that you get right down to the core of the puzzle. If you can't make moving from node to node interesting then chances are when you try to put it into the actual game you are going to have to rely on the games bells and whistles to hide that fact.

To go about doing this, take any space that the player can be in without going through a door and give it a number. Then draw these rooms as numbered nodes. When one room connects to another room via a door draw a line connecting two nodes to represent this. If it's a one way door you can make it an arrow, and if you want to distinguish doors you can use different styled lines.
If you want to divide a room up into multiple distinct sections then just draw multiple nodes and connect them based on how the player can move to those parts of the room from the other parts of the room.
This is just for basic flow, however, eventually you are going to have to test the paths in the actual level.

A lot of Zelda dungeons can be described as "going from A to B to unlock C" and they manage to keep it interesting. So long as there's more to it than that there shouldn't be a problem. Graphing the rooms and connections is suppose to reveal how much is actually going on.

I'd imagine it's hard to visualize the process unless I'm drawing on top of the map, but there's quite a bit of back and forth so it becomes difficult. Photoshop's notes would be very useful for this, but PS isn't that good for designing the level itself.

I'll try to see if I can make use of graph theory concepts though, thanks.

But those aren't puzzles. Zelda works by having you shoot a fucking eye in the Forest Temple to twist the hallway as the puzzle. The reward is the key so you can open another door.

Fair enough.

What's your opinion on innovation being hard to do in some genres, do you believe in such a thing? In what genre?

how do you get gud at modeling?

do you make alot of models and analyze what you've done wrong?

pic related for example, there's alot of things wrong with it (I 'finished' it because I wanted to test things out on my own)
I assume that you should start off by getting good at modeling and meshing or increase observational skills. Also follow tutorials.

it's terrible but at least it's better than the very first model I made

You practice and you will just become better over time. Tutorials held and it's always good to check the user manual for reference, in case there's a faster way to do something.

Is anyone else here scared by 3D stuff?

I don't mean "scared" in the metaphorical sense, as if to imply it's frustrating or daunting to get into. I mean literally scared as in fear. It's a bit hard to describe. But basically, imagine the feeling you get when you clip out of bounds, or if you turn off the global lighting in a game, or see characters in their stock poses. That shit makes my skin crawl, and I don't know if it's an uncanny valley thing, but it makes me unable to get into level editors and stuff like that because I feel so uncomfortable.

Am I alone in this?

I hope so.

Don't worry user, I'll find some way of triggering you without forcing you into a level editor.

Toast in bread