[ / / / / / / / / / / / / / ] [ dir / bebjush / choroy / dempart / freeb / randamu / vg / vichan / y2k ]

/v/ - Video Games

Vidya Gaems
Winner of the 81rd Attention-Hungry Games
/y2k/ - 2000s Nostalgia

Entries for the 2019 Summer Infinity Cup are now open!
May 2019 - 8chan Transparency Report
Comment *
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
(replaces files and can be used instead)
Show oekaki applet
(replaces files and can be used instead)

Allowed file types:jpg, jpeg, gif, png, webm, mp4, swf, pdf
Max filesize is 16 MB.
Max image dimensions are 15000 x 15000.
You may upload 5 per post.

[ /agdg/ | Vidya Porn | Hentai Games | Retro Vidya | Contact ]

File: 79d78610cf8d5e2⋯.png (457.88 KB, 1920x1080, 16:9, agdg fractal.png)

File: 668bcfac6cbe6d3⋯.png (77.44 KB, 263x278, 263:278, mfw-msw.png)

60193a  No.16556407

E3 Anger Edition

Watch E3, and use the anger it gives you at the current state of the industry to make your games faster

Or better yet don't watch it and keep working, then get angry at a recap thread.




>#8/agdg/ via irc.rizon.net

>Dev resources: http://8agdg.wikidot.com/resources


>Wiki: http://8agdg.wikidot.com/

>Beginner's guide: >>>/agdg/29080

>Previous thread: >>16499138



>Please contribute to the wiki if you can

b1ec9b  No.16556455

File: b9e42cd8a4cde61⋯.png (324.64 KB, 852x1268, 213:317, pickle rage.png)



if i could channel my annoyance into focus i would get this done sooner

e00eb4  No.16556471

File: b43d0b4a84dea4e⋯.png (819.08 KB, 1088x582, 544:291, fuck.png)

Anyone with ProBuilder experience who can help me figure out what went wrong? All these faces are connected and yet the light seems to go through this part.

f066a5  No.16556507

File: 216ba867eef3918⋯.webm (6.53 MB, 1280x720, 16:9, 2019-06-08 22-26-57.webm)

Added new sounds to the flare gun together with AI running away when they're on fire.

1ef9ba  No.16556551

File: deb0d8a3c724ae8⋯.jpeg (103.22 KB, 749x789, 749:789, D8hHic3U8AAKlNE.jpeg)

> getting back to work after 6 months dev inactivity

Leopard Seals now have a chance to attack you when you kick a seal pup and the player penguin will throw sweat drops when being chased by a leopard seal.

60193a  No.16556567

File: 013e66fa04f54ca⋯.jpg (34.8 KB, 640x852, 160:213, jazz_music_starts.jpg)


>Erni's Iceberg is not dead

Already a good thread.

b1ec9b  No.16556568


I make my buildings myself, i don't trust those assets

e00eb4  No.16556591


Frick. All I want is classic blocky FPS-architecture.

f066a5  No.16556596

File: d5f9fd7a50aabff⋯.jpg (7.8 KB, 184x184, 1:1, 1414702864293.jpg)



>Erni's Iceberg

b1ec9b  No.16556612


You will probably lose less time if you make your own in blender, believe me

be2f1b  No.16556614

File: b7b3744b7c7cc40⋯.jpg (78.33 KB, 523x355, 523:355, 1464091838459.jpg)


>Erotic Ice(((berg)))

6d5a72  No.16556620


Looking good.

be2f1b  No.16556622


6d5a72  No.16556632


Is that Godot? I assume you're tiling one object: are you absolutely sure both vertices are at the same position?

b1ec9b  No.16556648


looking nice

c61ec1  No.16556650


I suggest checking https://assetstore.unity.com/packages/tools/modeling/sabrecsg-level-design-tools-47418

I am pretty sure you can create presets like windows and plop them one after another anywhere, for instance

b1ec9b  No.16556652


ProBuilder is for Unity

e00eb4  No.16556672


What >>16556652 said.


Thanks, I will try it out.

a7db09  No.16556680

>This works


local shits = {
for i,shit in ipairs(shits) do

>But not this

local shits = {
for i,shit in ipairs(shits) do


what the fuck.

444917  No.16556689



which version?

I also remember unity being kinda fucky when it comes to light calculation on lower settings. see if it still happens on the highest quality.

a7db09  No.16556715


I found the issue. It's very stupid. Some function I was calling in my sprite loader (which was in a file require()'d by a file require()'d by a file require()'d by graphics/slots.t was changing the cwd.

7fedb8  No.16557238

File: e26c929d5f8ec09⋯.gif (55.74 KB, 320x240, 4:3, qPaQsrUnZs.gif)

File: 0bb7406206cf42f⋯.png (35.9 KB, 1366x768, 683:384, unknown.png)


Radar view cones work. After I get my NPC and security camera working, it is just bug fixes and clean-up to demo day.

7fedb8  No.16557255


> shoot down rockets

> quickdraw

> sniper duel

> save the hostages

> target juggling

> headshot challenge

> panzer dragoon lock-on system

> rhythm shooter

> feed starving animals by shooting bags of food from a distance

766db7  No.16557633

File: 238800fb3c7a928⋯.jpg (79.58 KB, 427x604, 427:604, all base instincts turned ….jpg)

Some anon challenged me to make this game and haven't posted since, I'll post the game just once more to give him a chance to see it.


560de3  No.16557962

File: c894058ddfe604e⋯.jpg (122.96 KB, 769x1078, 769:1078, 00b.jpg)

>client sends sizeof of a struct of a character, void pointer, server expects a size_t

>client then sends the actual struct, server expects a struct with the size provided

>server determines what the void pointer is by the character

is this retarded?

e49ba4  No.16558589

File: fe83b624ec76f15⋯.gif (936.6 KB, 500x360, 25:18, clothing3.gif)

This is a few days old but I finished my system for saving, embedding, and loading custom clothing textures.

f066a5  No.16558671

File: aa16516664b5f1a⋯.png (64.31 KB, 1656x740, 414:185, ClipboardImage.png)



Thanks anons. Right now we're working on fog of war and starting work on our hubworld that will kind of frame the game.

9bc73d  No.16558799

File: 71e8d8193f0666c⋯.png (102.35 KB, 1137x507, 379:169, ClipboardImage.png)


Proper line of sight would be far cooler tbh.

f066a5  No.16558819


Yep, we're working on it, what you see is just the first iteration. Hopefully it will be done sometime this week.

d3189a  No.16559000


That's just how sending packets over a stream like TCP works, but there's some things to note:

>size_t is not a fixed-size type. On win32, it's 32 bits, and in win64, it's 64 bits. Use uint16_t or something similar defined in <stdint.h>.

>Any struct data fields may also be subject to the above. Though this is usually not as big of an issue.

>Any struct data that is sent over networking should also have padding disabled. Wrapping a #pragma pack(push, 1) and #pragma pack(pop) around the struct will do that for you. This isn't just important for "different architectures" like the above, but disabling padding will significantly reduce the bandwidth usage.

Serialization is a pain in the ass in every language, and so is networking.

But in the amazing world of C you can abuse flexible struct members and variable length arrays to make it much easier for yourself.

54ae77  No.16559863

File: 0840827fd5928f8⋯.webm (2.86 MB, 1592x870, 796:435, dorf_voice_acting.webm)

File: 632fb1165ac868e⋯.webm (1.08 MB, 500x500, 1:1, FEARBritbongReplicas.webm)

File: 3d0cbceb675bcab⋯.webm (2.49 MB, 800x586, 400:293, EDVoicefag.webm)

File: f15e8a62754c9b3⋯.mp4 (374.72 KB, 700x566, 350:283, RadmodVO.mp4)

Shilling my voiceshit again.

If anyone needs any voice acting in their projects, I'd be happy to assist.

26b757  No.16559931

File: 8a566de300e9280⋯.gif (1.47 MB, 355x225, 71:45, thumbsuplion.gif)


I never realized I wanted a Britbong voice mod for FEAR until now.

560de3  No.16560574


>size_t is not a fixed size

>disable padding

what the fug, well thanks for noting me that. I'm not sure if SDL's networking library does serialization by default (probably not) but I'm not too worried at of this moment to try to pack as much data as possible

d3189a  No.16560675


The thing is, if you only compile for one platform (e.g. 64-bit windows) you don't run into these issues. The aforementioned issues only occur if you compile for multiple platforms that behave differently (e.g. win32 and win64). There's more things like endianness or char not being 8 bits, but those are so rare these days you can safely ignore them.

cf662c  No.16560941

File: bbc2cb374babf5c⋯.png (5.62 KB, 430x349, 430:349, ClipboardImage.png)

I'm hoping to work on my Super Metroid randomizer hack a bit more this month.

My current difficulty is taking a number of room nodes and connections and building them into room shapes / converting a directed graph into a planar graph without intersections. For example, rooms can be rectangular or square, or they can be hallways or columns, or they can be T shaped, whatever. Lots of variety. I'm trying to figure out how to expand the nodes so that they're interesting shapes, aren't purely random, and that edges/paths are preserved (room A -> room B).

Like if I have a bunch of nodes, I might end up with pic related afterwards

e49ba4  No.16569400

YouTube embed. Click thumbnail to play.


55d030  No.16569498

File: f24604e2ab0a0d6⋯.png (240.59 KB, 700x700, 1:1, 1482185134679.png)

How the fuck do I make a UI that doesn't look terrible in Godot?

c21366  No.16569698


What kind of UI, anon?

47cc4e  No.16569970

File: c4a16d9fc1787c0⋯.jpeg (48.96 KB, 800x486, 400:243, ed010f7af16c9c1cd1f79be18….jpeg)


You just need to convince yourself that it is only acceptable to not study if you are making progress instead.

Use those procastination hours

t.Student currently in exam season


cf662c  No.16569973

File: ee2912c0fe4e195⋯.png (20.7 KB, 752x636, 188:159, ClipboardImage.png)


Okay, I think I have a satisfactory first pass for determining my room bounds. It requires knowing ahead of time how many entrances/exits there are to a room, and where they are located. Beyond that, it figures it out

<Row 1

>1: Start with a blank field

>2: Disallow any existing rooms/screens

>3: Determine the doors of any adjoining rooms

>3a: Ensure the minimum space fits the game's limitations (width/height are both <= 16, area <= 50)

>3b: Run A* to ensure there is a valid path

>4: Find the midpoint between all doors. For each door, create a drunken walker with a bias towards that door - in the case of a long, winding path, use A* as a guide

<Row 2

>5: Duplicate image, please ignore

>6: Mark any partially intersected tile as belonging to this room.

>7 (small): You now have the minimum bounds of the room. Ensure once again that it fits the game's limitations (walk could have expanded the boundaries to satisfy A*, despite heuristic)

>8: Randomly connect each screen with its neighbours, ensuring there is a valid path to every desired screen

>9: Actually set the walkable areas and screen bounds

cf662c  No.16569982


This only generates the bare skeleton of the room - it doesn't place pretty tiles, objects, enemies, or even create ingame door links. It's just to figure out the room shape and position in the game map. But after I have this, I should be able to comfortably transform my room nodes into rough shapes and go from there

7d32cc  No.16569983


Design the UI outside of godot then implement it in engine.

Read the docs.


Ask more specific questions to get more specific answers.

2a7595  No.16570220


>in Godot

Basically >>16569983

UI has nothing to do with the engine, it's just a bunch of lists and static hitboxes with sprites over them.

273568  No.16570226

File: 7ec99a794dc6ff4⋯.png (13.39 KB, 189x139, 189:139, Martian_Buddy_logo.PNG.png)


Here are some hints:

< study canvas layers

< study panels

< study buttons

< study tilemap and tilesets

< study hide/show function

< study themes and visibility

< study GDScript setget

> break your project

> cry

< steel yourself

< *Just Like Make UI*

If I can do it, you can too.

inb4 my game's UI looks like shit

c89d46  No.16570229


How soon until with AI you can press a button and BAM it makes it for you??

273568  No.16570259


AI can eat the underside of a pot left in the damp shade of an unlit crevice of 30 years, get a disease, and fuck right off. Making it yourself will always be the true path to happiness and suffering. Forgot to mention this:

< study draw() and update()

< study 2D draw order based on child position in tree/branches

< study export variables

< study groups+signals

> study UIs from other games

68f95e  No.16570630

File: 61eefc1c3bd8bd1⋯.jpg (478.78 KB, 1000x1045, 200:209, __cirno_touhou_drawn_by_ne….jpg)

How do you guys even manage to /agdg/? I can never bring myself to fully commit, because even though I'm definitely not doing it for the money and the biggest joy I could get from a game is to have even a single soul cherish my product, it always seems such a daunting task requiring artistic skills I don't have and with gameplay ideas that appeal to me first and to others second.

2a7595  No.16570708


>ideas that appeal to me first and to others second

Can't say I know the feeling of not having art, but why does this matter?

d3189a  No.16570716


The trick is to just like make game.

Also, actually enjoying the process is the key of keeping productivity up. If you don't enjoy it, you'll only end up working at a terrible pace.

I very much enjoy programming, but I've barely got any experience modeling and I hate myself when I do it because I'm bad at it. As a result, my productivity is abysmal when modeling, but amazing with programming.

>gameplay ideas that appeal to me first and to others second

Good ideas always appeal to you first and to others second, because otherwise you wouldn't bother making them.

What translates an idea into one with mass appeal is to understand why it is appealing to you, and how that appeal can be extended to a larger audience without diluting it.

e49ba4  No.16570748


There are tools outside of just using Photoshop and Blender. If you just use those you will be in a world of pain. Learn the tools of the trade to aid in your art.

985746  No.16570771


I'm just not a weak-willed faggot like you.

2e7c22  No.16570972


Where's your game?

54ae77  No.16571022

File: 29cc50e24c9a1dd⋯.jpg (24.68 KB, 229x226, 229:226, 1548778692898.jpg)


Should I do it?

985746  No.16571130


Where's yours?

2a7595  No.16571163



>nodev shitting on other people

Has the transformation to 4um/agdg/ begun?

985746  No.16571207


>doesn't post game

>calls others nodev

My game is in the thread faggot, can't say the same about yours.

2a7595  No.16571224


Could've just said "yes".

8e4256  No.16571231


>it always seems such a daunting task requiring artistic skills I don't have

Don't focus too much on art alone. You will learn it if you do it. Learn other game dev skills too, like audio editing, programming or similar. That will ensure that you won't burn out on art stuff alone. It completely changed me.

>with gameplay ideas that appeal to me first and to others second

So? Just do it man. My entire reason for devving is to make a game I would like because I hate what the modern industry has become.

68f95e  No.16571588

File: 1c3e469810bef8d⋯.png (49.58 KB, 1032x793, 1032:793, 1c3e469810bef8df3f5c664a9f….png)

Thanks for your words of encouragement.


>but why does this matter?

Because on the one hand I'd want to make something I'd play first, but on the other even if it autistically fits all my criteria I'm afraid it'll just be a pile of garbage to everyone else. I can go into detail about it but I don't wanna sound like an attention whore, mine was a genuine question since I respect independent devs like those who post here.


>If you don't enjoy it, you'll only end up working at a terrible pace.

I see. That makes sense, the whole journey, not the destination deal, right? Hope so.

>Good ideas always appeal to you first and to others second, because otherwise you wouldn't bother making them.

I'm more worried there won't be an intersection. Not on what's popular, but what's good.

>What translates an idea into one with mass appeal is to understand why it is appealing to you, and how that appeal can be extended to a larger audience without diluting it.

Whoa, that's pretty deep. I guess your'e right on that one.


I'm not really constrained to one tool in particular, in fact I'm of the opinion that if something works with some tool, there's no reason to force another to do the same - in fact I tend to edit shit with goddamn Paint.NET, except when I need hand-drawn stuff which makes me go back to cs5.


There's probably some truth to that.


>like audio editing

You mean like, sound design? Are there even courses or books on the matter? Or should I just try and use my own point of view instead of learning stuff first?

>My entire reason for devving is to make a game I would like because I hate what the modern industry has become.

Thanks for the reminder.

2a7595  No.16571651


>even if it autistically fits all my criteria I'm afraid it'll just be a pile of garbage to everyone else

If you think it's great and put all your feeling into it, then there's other people who will think so too. It might be a problem if you're interested in some obscure thing like typical hacking games which are boring as fuck seemingly by design, but otherwise I myself wouldn't be concerned about making a game that I wanted to exist.

83782f  No.16572369

File: ec32f7dd06d90a4⋯.gif (94.01 KB, 320x240, 4:3, v0NAkrBLMD.gif)


Super Mario jump implemented. An A press is an A press. You can't say its only half. Air control and turning modified to prevent moving in directions the player is facing (like how a car moves forward while turning) instead of the direction of the input (like a soldier turning at angles and moving forward).


Just Like Make Game.Either make it or don't.You are not obligated to either way.

e0422b  No.16572482

File: e059d4fccb0b896⋯.png (133.64 KB, 512x512, 1:1, 32b1673a85f0fdee59959a1cb9….png)


>My game is in the thread

We have IDs here, 4cuck

871354  No.16572825


Make a game mockup (fake screencap) with your intended ui and sprites, then implement that into your editor.

20e07d  No.16573644

File: f49247c3d1b7219⋯.png (46.36 KB, 1280x720, 16:9, HUDplanning0.png)

File: a1ba17a1df750b5⋯.gif (46.56 KB, 320x240, 4:3, d10UZxEkDY.gif)


I agree.

cf662c  No.16575717

File: 2996baa0675a347⋯.png (11.83 KB, 604x787, 604:787, ClipboardImage.png)

What's the usual method for determining if a certain pixel or map cell lies within a line? Is it Bresenham's Line Algorithm, or something else? I'm trying to work with polygonal shapes to do stuff

f927b1  No.16576528

File: f014962b77b5a08⋯.png (23.16 KB, 640x480, 4:3, 79a.png)


To play doubles advocate I often jump across multiple IPs every day because I work two jobs, both of which have their own wifi as well as my phone's cell internet which makes four IDs a day. inb4 phoneposter scumAt least a third of my game was coded on my phone

e74f05  No.16576579


What exactly are you trying to do? Bresenham's algorithm does not find every cell the line intersects, it just finds a bunch of cells that look like a line.

47cc4e  No.16576598

File: b5a9c54fcb1ef3f⋯.png (96.15 KB, 1556x532, 389:133, Dirty_dumb_phoneposting_sc….png)

47cc4e  No.16576682

File: 43ddc4bb6330036⋯.png (391.28 KB, 685x958, 685:958, collect_the_following.png)


You already have implicit knowledge of all cells in a given line,since that is the definiton of the line

Further implementation depends on what exactly you are trying to use it for.

2e7c22  No.16576977


>You already have implicit knowledge of all cells in a given line,since that is the definiton of the line

There are multiple definitions of a line, not all of which include being composed of all cells that contain a line.

60193a  No.16576992


>You already have implicit knowledge of all cells in a given line,since that is the definiton of the line

The definition of a line is two points, anon.

cf662c  No.16577086


Technically, a line segment is two points

47cc4e  No.16577175

File: db9f2c7bf35e00e⋯.png (487.89 KB, 856x541, 856:541, pixelprosciutto.png)


Two points by itself can represent any connecting structure,not necessarily a line.

>>16576977 is correct though,I fucked up.

The point still stands since you can derive the "all cells" definition from the other ones

f066a5  No.16577438

File: 2b6a2393abb7a20⋯.png (48 KB, 1034x744, 517:372, ClipboardImage.png)


Why would anyone expose their game to someone who goes

>post ur game fagget

It's just opening up to some autistic hater.

Progress for the anon requesting LoS, here's the first pass. Interpolation and blurring coming up.

e74f05  No.16577788


A line is uniquely determined by two points and each two points define a line.

55d030  No.16577927


That sounds like a good way to think about it

92b52c  No.16578014


b1ec9b  No.16578041



It's a great motivator, the more dogshit the industry becomes, the more my game will sell

976b72  No.16578070


Too late.

20e07d  No.16578588

YouTube embed. Click thumbnail to play.


More of the same but with sound and new standing sprites. Think I'm done for today, been going for about 12 hours.

17f3f9  No.16578603


>no decent games

Guess we need to pick up the slack.

4f6a0f  No.16578852

Could someone suggest a good torrent for Maya ?

be2f1b  No.16578867

4f6a0f  No.16578895


Anon could you suggest img tracing within blender if I'd tried rendering models on it ?

cf662c  No.16579435

File: 36d2ff6190ba830⋯.png (67.75 KB, 776x795, 776:795, ClipboardImage.png)

File: 024522b7dfad447⋯.png (65.77 KB, 776x795, 776:795, ClipboardImage.png)

File: c6b4e151facb4f5⋯.png (55.27 KB, 776x795, 776:795, ClipboardImage.png)

File: 8ac71703f7d1115⋯.png (59.42 KB, 776x795, 776:795, ClipboardImage.png)

As a possible approach to shape some rooms, I considered random walkers. Needs some attunement

bacf60  No.16581374

File: 8e843e2d0a024c4⋯.png (309.06 KB, 633x619, 633:619, 7c105dc411442e2268d0469918….png)


Consider cellular automata. With a bit of tweaking it can make some really cool shapes. Plus its easy to offload it to the gpu

cf662c  No.16581510


The process for any CA is that you have an initial state, a set of rules, and then a finished state, after applying all rules simultaneously?

089805  No.16581571

File: 30240e85263ee7e⋯.jpg (17.32 KB, 700x525, 4:3, aos.jpg)

What is the best way to create a voxel based game?

Any good texts on this?

cf662c  No.16581608


Is there any particular difficulties you've had with it? You should be able to implement and work with marching squares/cubes algorithms comfortably, I think

089805  No.16581682

YouTube embed. Click thumbnail to play.


Voxel based is not exactly a cubish game, this game for example is entirely voxel based, and more closely to what i want to do.

I know (in theory) how to do a minecraft clone on some engine like unity or godot, but making the same thing i would do on this minecraft clone on a game that looks like vangers looks really wrong.

bacf60  No.16581695

File: 988b1a503e71c78⋯.jpg (233.35 KB, 1801x897, 1801:897, brogue_197562_full.jpg)


yes, but the grid is usually binary and the rules are created in terms of number of adjacent positive cells. conways game of life is a neat application of it. The roguelike Brogue uses it to make its caves

650933  No.16581733

YouTube embed. Click thumbnail to play.

Blender 2.8 has some really cool stuff.

cf662c  No.16581825


It took me a minute to understand what you meant by binary-based; each cell is either 0 or 1 (I thought why not just use boolean, but then I realized it's not really number-based even though some languages interpret any non-zero value as 1/true)

bacf60  No.16581907

File: 0c107393ba58066⋯.png (629.36 KB, 916x577, 916:577, gamedevelopment.PNG)


might just be because I have a background in mathematics rather than computer science.

Heres a screenshot so this post is still agdg related

7d32cc  No.16581970



You should be able to use a sparse voxel octree.


I was literally just watching that video.

7d32cc  No.16581973


I messed up. Meant to reply to >>16581733

089805  No.16582126


>sparse voxel octree

Thank you, i was looking for some key words, and that's exactly what i wanted.

7d32cc  No.16582177


Happy to help.

47cc4e  No.16582268

File: dc59451482e3bf0⋯.pdf (2.49 MB, HighResolutionSparseVoxelD….pdf)

cf662c  No.16582335



Is it possible to perform cellular automata in-place? I feel like the answer is no - mutating the source array will alter the results, and you will always need an additional data structure to store the output

Also, found this:

https://www.please use archive.is/2015/07/28/how-do-roguelikes-generate-levels/


bacf60  No.16582706


I suppose you can store multiple levels of information through bit shifting and just access a unique level each pass.

lets say youre using a ubyte:

store starting data in the 1st bit

>1< -0- -0- -0- -0- -0- -0- -0-

on first iteration, for each cell:

read from adjacent cells 1st bit

>1< -0- -0- -0- -0- -0- -0- -0-

apply the rules of the CA and store the result in the 2nd bit

on second iteration, for each cell:

read from adjacent cells using their 2nd bit this time

-0- >1< -0- -0- -0- -0- -0- -0-

apply the rules of the CA and store the result back into the 1st bit

then repeat for however many iterations you want. this way the iterations dont have to be stored in another variable to prevent it from interfering with its source.

or you could just use a struct or something if youre a pussy

d3189a  No.16582721



Minecraft's approach of "just put everything into a block array of blocks lmao" works perfectly fine.

If you want a higher resolution of voxels, then it's not just about using an algorithm like sparse octrees to compress the data, but you also need a way to efficient traverse the data and perform collision detection. The main cost of voxel worlds isn't in memory, it's in the read/write speed.

But if you just want your world to look less blocky, there's other approaches you can take:

>Tesselation to dynamically generate details on blocks

>Marching cubes or naive surface nets to tesselate into a smoother geometry https://0fps.net/2012/07/12/smooth-voxel-terrain-part-2/

0e0b10  No.16583081


Here is the source code to VOXLAP, the engine that runs on:


d3189a  No.16583158



9bc73d  No.16583205

File: 0a8f246821793d8⋯.pdf (5.41 MB, schroeder2015.pdf)


>Marching cubes or naive surface nets

There's no reason to use either of this algorithms anymore, especially Marching Cubes. The way to go is "Flying Edges". I implemented that algorithm on the GPU using Vulkan for my bachelor's thesis and it's quite impressive.

While it is substantially more complicated than Marching Cubes, it generates an indexed mesh with no redundant, overlapping vertices, all the while drastically reducing memory consumption and running a lot faster.

The algorithm was designed to be run in parallel on the CPU, but it still did great on the GPU. What's cool is that because it scales so well across CPU threads, you can give it a thread budget (maybe based on how many cores the PC that's running your game has) and it'll just work.

For the potential Anon who tries to implement it: Go check out the implementation inside the VTK library. It should help you, if you get stuck.

e05e0a  No.16583920

File: 250353963c92dc5⋯.png (105.07 KB, 1280x720, 16:9, ClipboardImage.png)

File: d5b6659909bd697⋯.gif (231.39 KB, 320x240, 4:3, YD2MRSTVhK.gif)


Security camera has an exported and edited UV, goes with the base on the right. This should be in the next demo. Also some more tiles for the radar, these have to be manually set because I don't want to make a system that iterates through meshes and draws based on what should be where. Also some hazard space icons, which will be useful for the turn-based combat and squad movement to prevent moving units somewhere that will result in a suicide. This way I don't have to put more 3D position nodes in the map and blog up editor space.

bc38b0  No.16585076


Looks cool, what are you working on?

cf662c  No.16586301


Thats a huge micro optimization and i think doing that would cause more overhead than just shitting it into another array freely, because of how C# handles certain things.

My big hurdle is detecting the adjacent cells in a quick manner. Specifically avoiding having to nest the bounds checks inside the loop itself

4f6a0f  No.16586519

File: 9b3e19e54a4bf55⋯.jpg (145.2 KB, 1280x720, 16:9, hoji.jpg)

Does concept art help your game ?

Would buying books on coding , AI creation or Animation help ?

How good are Video tutorials in compassion to reading a book or figuring stuff out from various other sources like purchasing various stuff on creating code for the game that'll help me better understand what I'll be making ??

cf662c  No.16586539


>Concept art

I'd say insofar as you'd be basically making character sheets to keep design and style consistent. If you're more than a 1MA you probably want some, or if you're an artfag with an elaborate project.

>How good are video tutorials

Everyone learns differently. Do you learn by doing? Watching? Reading? Do the right one.

4f6a0f  No.16586605

File: 4478aa5aa14d1ee⋯.jpg (17.84 KB, 601x601, 1:1, pinzo11.jpg)


Watching is normally much better for me , though I'm concerned that I'll miss key details with how to properly do it. With if I'd get myself a couple books it would be much easier since it could teach me some stuff and I'd learn step by step. I'm normally nervous that I'll miss stuff to do with how I'm making the game. Since often I've never properly gotten a good grasp on it and I'm worried that I'll mess something up.

Maybe a tutorial will give me something through a video that a book won't in a couple videos , I'm still scared I'll mess something up and possibly end up doing something that will make it look horrible or play horribly.

d51743  No.16586646


I'd say planning out your game in general, getting everything you want down on paper is very important. Concept art is a part of that.

d3189a  No.16586759


I'm kinda skimming through this, but if I understand this right: This is essentially a whole slew of optimizations to produce a mesh similar to marching cubes, but do it significantly faster, mostly through efficient use parallelization and reducing memory access?

The reason why I named both marching cubes and naive surface nets is because they produce very different results. Naive surface nets has the nice advantage of being identical to Minecraft tesselation with the vertices moved a bit, making it trivial to mix Minecraft-esque block construction with smooth naive surface net terrain.

Don't get me wrong, it's a cool algorithm. But I don't think voxel tessellation is a significant bottleneck as it only occurs sparringly.

2d402c  No.16587300

>Turns out my validation layers weren't even functioning for 2 months

>Switch to the fancy new unified validation layer

>Terminal output is flooded with several MB of error messages

2d402c  No.16587520


At least I've found one of the issues. I was getting a GPU hang because I was freeing some of the command buffers I was using to perform copies before they actually finished execution.

9bc73d  No.16588026


The reference implementation in VTK derives its tables from the Marching Cubes ones. It then produces a visually identical mesh. The topology looks the same, too, but is different due to it being an indexed mesh (as opposed to the "triangle soup", as it's called, that MC produces).

It's not a set of optimizations as much as a fundamentally different approach to extracting a disctetized isosurface. Whereas MC solves the problem by dividing your space into voxels, each of which are solved independently, FE looks at how the isosurface is intersected by the edges along an axis if your choice. It doesn't work on a voxel basis.

The speedups are achieved in a variety of ways, but the underlying idea is that FE prevents redundant calculations. MC calculates and stores many vertices and its normals up to four times. This doesn't just eat processing time, but also massive amounts of memory. As a result, FE can handle significantly larger datasets (while also being a lot faster).

2687b3  No.16588051

Anon shmup here


>Plan animation needed data, location, local update, update to GPU. Pretty much the whole cycle.

>Start to implement it

>Realize that I do have vertex attribute limitations


Hopefully I have some plan B, use another texture buffer for dynamic animations along with the other for pre-calculated animations that I have. Then I can specify each instance dynamic animation start with 1 attribute and send all dynamic joint in one batch.

Will be revising my plan on what I modify.

4b9a91  No.16588206

File: 31ed46076e8ac8d⋯.gif (1.65 MB, 620x620, 1:1, __hata_no_kokoro_kaenbyou_….gif)


>How the fuck do I make a UI that doesn't look terrible.

Pro webdev fag here.

do these:



You can find any number of UI mockup tools to get a good idea of where to put things and implement it later.

Aside from that:

>Focus on functionality first and take technical debt on UI. Slowly implement the UI bit by bit between more meaningful work, this will keep you sane and motivate you since UI is basically free, visible progress.

Shit's really good if you also have to meet with a customer every two weeks, since UI seems like 80% of the work to them

>Hallway usability tests(grab some random person and have them try to do something) are pretty good for getting quick feedback, just keep in mind people will become biased so you have to get fresh batches frequently

>Expert User Path VS Actual Path Taken is a handy metric

>Redundancy is great, don't be afraid to have multiple ways of doing the same thing like, an "X" button, hitting escape, clicking a "close" button, or clicking off the menu to close a dialog.

>Blue light is the Devil. Dark mode or bust.

>Good sound effects for buttons and UI actions endear people to play your game. See: every Japanese game ever

<Seriously, make something shitty first and then iterate on it over time. A shitty UI eventually can become a good one.

b1ec9b  No.16588251


You just like make game, simple

b1ec9b  No.16588262


>sah yeah sah

I like it

36b5a7  No.16588296


Get a prototype first of your idea first, then do concept art to polish things up.

746999  No.16588609


Unless you are adding in more enemies, putting in a flare gun that does an explosion on enemies faster than you that pretty much always get you in the blast is stupid.

746999  No.16588615

37246a  No.16588640


Stealth platformer with a planned RPG combat system like Xenogears meets Vandal Hearts.

2d402c  No.16588856

So after much pain, I finally have gotten textures properly uploaded to the GPU and sampled. The only issue now is that it seems to be suffering from some kind of gamma correction; not sure why. Should I be using R8G8B8A8_UNORM for my image formats?

2d402c  No.16588859

File: bec31324a6f8133⋯.png (8.04 KB, 909x996, 303:332, gamma.png)


fug, forgot to add my picture.

2d402c  No.16588868

File: 9abe552fe34ba78⋯.png (917 B, 32x32, 1:1, wood.png)


For reference, this is the sprite.

985746  No.16588936


Well we're adding more enemies and letting the player carry a second weapon as well to fight off enemies that get too close.

746999  No.16589291

f066a5  No.16590816

File: 0b522e576112c7b⋯.webm (7.43 MB, 1280x720, 16:9, minimap.webm)

New minimap in action.

b1ec9b  No.16590897

File: bab85ba4eb3e8a5⋯.png (252.37 KB, 850x479, 850:479, new inventory.png)

What are the do and don't of Inventory windows? we are going to overhaul ours moving out of programmer graphics into some sort of comfy window, also add a few new features while at it, any ideas?

997ac5  No.16590924


>I'm concerned that I'll miss key details with how to properly do it

and? your first implementations will suck, like the first time you do anything. repetition distills ability, same reason grade schoolers write the same letter over and over. just get it working, next time you know the bumps in the road and since it's not your first time doing it you already got some experience which will only get better over time.

watching or reading both have pros/cons, written tutorials give you the ability to search and find stuff easy instead of remembering times where it was talked about in a video.

imo best way is to use both and different books/videos too, because even in source materials there can be wild quality swings - which brings us back to the beginning why it's silly to worry about doing it properly the first time (my favourite is probably webdev-shit that just shills a framework without mentioning it anywhere on the cover).

be797b  No.16590969


It's not that simple rRREeeeeeeeeee

b1ec9b  No.16590985


Actually maybe i will keep the number inside the icon window, but make the icons larger and put it in the top left


It is, just do it

d3189a  No.16590988


How would you know? You've never even tried to just like make game.

53e7b5  No.16590996


Thats really great, anon.

997ac5  No.16591009

File: f8c1af69fb5961a⋯.png (151.57 KB, 640x320, 2:1, dreambig.png)

60193a  No.16591086


Depends a lot on if you want to make it controller-friendly or not. I'd personally separate things into 4/5 categories and have a menu where you pick a category, then get a list of 6-10 things listed at once, with wider icons and the ability to scroll the list up/down.

Maybe have an "everything" category as well.

Even better would be defining the menu layouts in JSON files or something, so you could provide that and something more condensed like you have currently, and players could even mod the menu to their own liking.


>using Unity for 2D


997ac5  No.16591157



completely irrelevant when learning and looking for an easy entry. concepts are engine-agnostic for the most part anyway.


fun fact, check what engine ori runs on.

0947e2  No.16591173

File: a08e933db6d9022⋯.jpg (117.13 KB, 885x495, 59:33, why_does_he_have_a_thong.jpg)


He should use Unreal Engine instead.

746999  No.16591177


Fuck yea anon, just don't make a mod that replaces all the voices. Just adds your voice for some of them.

c1199c  No.16591182



>hating unity

It's time to stop anon

5fccc1  No.16591203

File: 8c897a15f00180d⋯.png (205.77 KB, 697x768, 697:768, 8c897a15f00180d432f0d87b16….png)


>it's current_year+4, why do you still hate unity

>it's just bad game devs

>I swear

It's still a buggy clusterfuck that's only grown a little better lately because they paid some outsiders to fix their shit. People only white knight it because it's babby's first 3D engine and a step above Game Maker.

c21366  No.16591204


Have you missed Unity controversy? Don't you know about their Jewish monetisation? Don't you know it's partially owned by Tencent?

>Muh 2019

I hope, your post was ironical, otherwise you either brain dead or Jew.

5fccc1  No.16591209

b1ec9b  No.16591213


Its already in XML

746999  No.16591220


>Falling for the 2008 one weapon on the back meme

wew, if its for game balance then sure add it in. But if you want to make the game more action focused with more enemies per X, then adding in multiple weapons would be pretty cool. But again, its your game and your balance. Looks like something I would play regardless though, il look for it on demo day.

c21366  No.16591238

Finally, some fucking progress. I'm working on Godot curse of the Azure Bonds.

>Fixed data decoding

>Extracted font image

>Wasted many hours trying to display it

>Learned to write decoded data to PNG

>Read about how to dynamically create bitmap fonts

>Googled for three minutes and found ready public domain font

I feel a bit stupid, but now I know how to decode images which would be very useful very soon.

997ac5  No.16591249



like what, not wanting to get fucked over by streaming services (and was clarified a few days later)?

>Jewish monetisation

you don't even pay royalties

>partially owned by Tencent?

either you confuse it with epic or you don't know what "partnership" means, but I'm gladly proven wrong if you got a source

>otherwise you either brain dead or Jew.

if it's so bad I'm sure you can point out exactly why, especially for 2d.

f066a5  No.16591256



Unity has a lot of flaws but in my opinion it's an engine that makes it very easy to be productive. My gripes are mainly that the editor and game run on the same thread so if your game locks up you're fucked. The profiler sucks dick as well, sometimes using 30+ gigs of RAM.


>>Falling for the 2008 one weapon on the back meme

I know /v/ really hates 2 weapon limits in general but I think it can work out here. The idea is to have some set of consequence to what you bring into a dungeon so you have to think about your loadout and not just bring everything. We're gonna playtest with both the limit and without it though, so we'll see how it ends up in the end.


He's just a FOSS-fag, don't bother.

8b3b50  No.16591809


I know how to build a lost of my radar targets for moving the 2D sprites around, but it took most of a day. dictionaries

e05beb  No.16592238


>My gripes are mainly that the editor and game run on the same thread so if your game locks up you're fucked

Why don't you just, like, never make a mistake?

a9526a  No.16592296


>My gripes are mainly that the editor and game run on the same thread so if your game locks up you're fucked.

I know it's a big part of the whole edit-and-jump-into-what-you-just-made thing Unity has going, but this will piss me off to no end.

>The profiler sucks dick as well, sometimes using 30+ gigs of RAM.

>unity manual swears: Note that profiling has to instrument your code (that is; add some instructions to facilitate the check). While this has a small impact on the performance of your game, the overhead is small enough to not affect the game framerate.

How the hell is 30GBs of RAM overhead considered small enough?

2fb295  No.16592560

File: e55ba3ab7fbb512⋯.png (71.12 KB, 700x700, 1:1, CUP LOGO 2.png)

So now that the project has been announced, I don't need to be secretive and leave out details. The Infinity Cup is coming back, and I'm looking for anons well-versed in Blender to help change player models and tinker with the stadiums a bit as well. Drop me a line if you're interested in helping out. SKF @ 8chan . co

b15267  No.16592564

File: 31522135a248f4a⋯.png (639.31 KB, 1143x438, 381:146, ClipboardImage.png)

Shotgun for DT

a9526a  No.16593091


And just finally figured out Unity is currently undergoing a massive re-write that's changing fucking everything and will break compatibility with earlier versions of unity, including pre-fabs and the network stuff. I guess I better wait a year or two for things to settle before trying to use Unity for a 3D multiplayer game.

f066a5  No.16593195


That rewrite is never-ever tier, even if it does come out it will be in such a shitty state there's no point using it for years.

997ac5  No.16593399

Invidious embed. Click thumbnail to play.


gameobjects will still be supported for a while and even if they switch completely they specifically have the LTS version for that (which has 2 years of support iirc).

from what I've seen prefabs works pretty much the same just with some added functionality/usability, network was inevitable since the old one worked but had it's own issues (while still not being as shit as unreal where listen servers don't even to interpolation since like 2014, not to mention can't even do basic shit like host migration).

>trying to use Unity for a 3D multiplayer game.

if anything DOTS will probably work better than anything so far, on the the big issues was performance which seems to be fixed somewhat (not that I believe the official comparisons especially since they compare it to the heaviest/worst implementation like destruction instead of pooling etc., but it will be a step in the right direction). api is finalized for 2019.2 I think, that's when I gonna start messing with it.

I just want a contender to fucking UE which gets picked because performance/graphics while ignoring anything else to the point it's worse than quake 1 from 20 years ago - even if it's fucking unity.

997ac5  No.16593402


>How the hell is 30GBs of RAM overhead considered small enough?

if you have enough ram ;^)

9369e2  No.16593409


I'm actually considering buying a new computer to handle that shit.

b15267  No.16594414

File: 87f6b08fd92ab96⋯.png (944.08 KB, 1909x908, 1909:908, ClipboardImage.png)

Making a grain elevator

5fccc1  No.16594497


Nice sky dick, faggot.

f49b66  No.16594515


ram is cheaper than 2016 even, definitely going with 32 gig this time.

what's the average professional dev ram spec anyway?

c21366  No.16595358


Same anon.

You are right, it's not ownership, it's partnership. I could've made more research, but I don't want to.

What I tried to say isn't that it's bad. I tried to say that it's okay to hate it be skeptical about it.

Unity is obviously better than nothing for just like make a game (TM).

But I don't see why not to use Godot instead. It's virtually has no downsides unless you need:

<Occlusion culling

<Working with huge numbers of the same objects in 3D

<3rd party SDK's like AdMob

The last one render Godot unusable for many commercial projects. And it's way easier to find unity guy than to find Godot guy. So studio, probably, should pick unity/UE

But for edgy dagy Godot is almost perfect, unless you are trying to get into industry or other personal reasons.

>Muh 2019

I guess this tranny shit just triggered me too much.

Okay, about progress. I've started writing resource extractor for COAB. Images has some shitty format which I need to parse to create normal resources and use them. It's still hard for me to break down code into files. I guess, I have some mental block due to code being connected to nodes and that you have to manually load files.

b15267  No.16596354

File: d01908e4a9cfa82⋯.png (49.13 KB, 924x778, 462:389, ClipboardImage.png)

f7d415  No.16596476


Started 2 weeks ago, so far its just studying how each program works before i get into the meat of it. Always wanted to try my hands at a Guild Management Sim. Recruit people, send them out on missions and build up your guild. Nothing too crazy really. I gotta say though watching everyone work so hard on their own game is inspiring, I just hope I don't end up like YandereDev

a9526a  No.16596514


If you find yourself writing a ton of repetitive code, or coding your own math routines, something has probably gone wrong. Look at the language/engine, see if there is a simple solution already available. I once saw someone manually create an array by defining every single entry as a global variable(pages of them), and wrote a function to index the abortion.

e49ba4  No.16596953

File: 63f68443b82e726⋯.png (854.5 KB, 743x740, 743:740, 1.png)

File: ec3fe8a40a30f00⋯.png (588.07 KB, 619x606, 619:606, 2.png)

Added stockings. You can cut out a pattern if you feed it one with alpha. Maybe use it to adjust sock height.

d3761f  No.16597126


>tried to say that it's okay to hate it be skeptical about it.

the main problem is unity gets a lot of hate by clueless nodevs because shit devs shit out shit games and not being a shit engine (debatable, but a different topic). hence the >2019, since it's like memeing ps3 has no games.

>But I don't see why not to use Godot instead. It's virtually has no downsides unless you need:

because unity has a fuckload of learning shit available, and the post was about getting started. the linked 2d/3d kit is easy since you can do stuff without coding and the principles/how shit works are easy transferable once figured out. not everyone wants to start with an empty window and make a cube rotate etc.

nothing keeps people from learning stuff with unity, then do stuff godot. the first few projects will be small and crap anyway and after that he should now if he wants to keep going or switch to other stuff like modeling etc.

8ea0d4  No.16597320

Enginefag here

I'm using GLFW callbacks to get my window's size and framebuffer size.

I was under the impression these two were different from one another, looking at the documentation, but the size of both of these is the same?

0e0b10  No.16597469


The framebuffer is just client area of the window that is being updated by the program. When you say you want a window to be a certain size, you generally mean you want the client area to be that size. GLFW is probably just dealing with the client area size and not the true window size. So, they are the same.

8ea0d4  No.16597717


After seeing your post, I played around and discovered what GLFW is calling window size is the same as the framebuffer size, and these are both different from the window's size including the insets, at least on my OS

Thanks anon

b1ec9b  No.16598833

My gay brother will probably soon post his dog shit huge ass mobile interface for blind people. Please shame some good taste into him

d3189a  No.16598881


I can see he's your brother.

e4d567  No.16598921

File: 5df096b233d611b⋯.png (11.54 KB, 500x250, 2:1, Oekaki.png)


>open game

>loud voice blares from the speakers











d3761f  No.16598941


>not just porting coc to braille

e4d567  No.16598943


>not just porting cock to breathe

6b1fac  No.16599006

So, where can you get people to betatest your game? Steam communities if you use game making engines?

b1ec9b  No.16599025

File: 6892fad0dd851bf⋯.png (144.82 KB, 591x356, 591:356, new inventory.png)

The sort buttons are not functional yet, but how does it look?

b1ec9b  No.16599033

File: 1792d81f8abb581⋯.jpg (99.35 KB, 1280x720, 16:9, kotoko happy.jpg)


Not a feetfag, but i really like this

b1ec9b  No.16599036


>expanded icons from 38x38 to 64x64 so you can actually see what they are


8ea0d4  No.16599212

In general, how should the relationship between the client and server be handled?

2e4ac8  No.16599226


This question is too general to be answerable.

b1ec9b  No.16599232


Anything that matters is the server job

UI is for the client, but make sure the client doesn't know what he doesn't need to, otherwise you can make client-side hacks to see through walls and shit

b1ec9b  No.16599302


The client is slave to the server

The client asks for things, the server may or may not give the client

The server never lets the client do something it cannot check if it was done properly one way or another

This could actually get a chad vs virgin image to properly illustrate how much of a bitch the client have to be

f90657  No.16599314


Seems to me that a client-side wallhack would just be dependent on the client knowing where an enemy is, which seems like the kind of thing the server would need to be sending to the client.

Unless you're implying that the server should only be sending information about where the enemy is if the server detects the enemy is currently in the player's visual range.

8ea0d4  No.16599320


If you want specifically what I had in mind I was thinking of how an object that is managed by the server would render on the client's end.

8ea0d4  No.16599324


For example if an object that is managed by the server would do nothing but render on the client's end, what would be the most effective way of transferring position, angle, animations, etc

Saging for doublepost

f90657  No.16599335


Well you definitely wouldn't be sending animation data over the pipeline. In general what you would do is assign the object a state (isReloading, isJumping, etc.) and the client plays the animation based on the state. Sending entire animations over the pipeline is a complete no-go.

Location and rotation is stored on the server. Any player interactions are replicated ASAP so the server can change its end and then replicate to all other players. You'd probably just replicate location and rotation using whatever units you use in your game engine (float, if I had to guess?)

8ea0d4  No.16599358


I've seen some games use numerical ID's in place of Strings or booleans for animations and object states (ex Dark Souls). What do you think of transferring state and animation data like this?

My last question is: how would the client know which states/positions/et cetra belong to which object being managed by the server, and then pass that along to the correct models?

My initial solution was to make client copies of objs that only contain client-relevant data, passing that along to the Renderer, but I'm unsure if this is a conventional approach.

937229  No.16599420

File: 1cea27ad7d7d0fd⋯.png (139.62 KB, 641x629, 641:629, 1cea27ad7d7d0fd39407d5021d….png)

a18060  No.16599453

File: da2061252db89dc⋯.jpg (98.17 KB, 456x380, 6:5, da2061252db89dcabfc4758556….jpg)


Can you adjust dress size? Can I give my Shinobu a sexy zettai ryouiki?

1ab2ee  No.16599683


>I've seen some games use numerical ID's in place of Strings or booleans for animations and object states (ex Dark Souls). What do you think of transferring state and animation data like this?

Effectively it's the same thing. You're simply setting a state using an identifier other than booleans or enums.

>how would the client know which states/positions/et cetra belong to which object being managed by the server, and then pass that along to the correct models?

I'm not entirely sure what you're asking, but it sounds like it would depend on the framework or engine you're using. Are you coding this from scratch or are you using something like Unreal or Unity?

b15267  No.16599996

File: 1fc9a6d675c09d6⋯.png (65.99 KB, 1098x820, 549:410, ClipboardImage.png)

File: 7fa88401d3b7a49⋯.png (2.32 MB, 780x1019, 780:1019, ClipboardImage.png)

4f5046  No.16600146


Will be a boss or something like that?

e4d567  No.16600200


Believe it or not, that is actually the protagonist of Sgt Hale's Shinobu Project 2. Every object the player character can carry is being held by a separate arm on the character model as a form of precaching for performance reasons and the game swaps between them on the fly. Of course, since the entire game is played in first person you will never see your head, torso, or legs without enabling the third person camera through the console.

b1ec9b  No.16601451


Ok this is canon now

015b8f  No.16601678


White looks too bright to me, also the bottom right corner overlapping the scrollbar looks weird. If I were to click on it when the scrollbar is in the lowest position, which element would be active, window resize or scroll?

b1ec9b  No.16601919


The resizer has the priority

I will extend the window bellow so they don't overlap

And try to change all white to a light beige or something we are using xmllayout, some of the elements are hard coded

d3189a  No.16601989

YouTube embed. Click thumbnail to play.


Depends on the model.

In Dead Reckoning, every tick the server sends over the latest gamestate to the client, whilst the client sends over his inputs. The client also extrapolates the latest gamestate from the server, applying inputs it recently sent as well.

If the extrapolation is poorly implemented (e.g. you don't properly account for inputs, or the gamestate the server sends over doesn't contain important components) then the player will experience periodic "warping". I'm sure you've seen moments where you try to move in some way, only to be warped back every ~.25s, until you move backwards and that works fine. That's a case of the client extrapolation diverging from the server.

In Lockstep, the client sends their inputs to the server. Every tick, the server groups up all the received inputs and sends them to the clients. The clients can then apply these inputs to their gamestate and run a few ticks of game logic.

This makes the simulation 100% in sync, but introduces an input delay roughly equal to your ping. Your game logic also needs to be fully synchronized and deterministic, unlike Dead Reckoning.

The new kid on the block, Rollback, is essentially a better but far more computationally expensive Lockstep. Clients send over their inputs to the server, together with the ticks these inputs were pressed on, which just like Lockstep groups up these inputs and broadcasts them to all clients. However, in Rollback, clients keep a history of gamestates, and upon receiving an input that was done in the "past", it rolls back to an earlier state, applies the input, then fast-forwards. The advantage here is that clients can immediately apply their own inputs when sending them, essentially removing the delay of Lockstep on the client side, at the cost of having to compute multiple ticks in a single frame. (The resulting ticks per second is equal to "ping*tickrate²". Note the square.)

In practice, most games apply Dead Reckoning as it is the most simple. Just spamming over the gamestate from the server end, clients extrapolate the movement/animations of entities, and its often already good enough.

Sometimes Dead Reckoning is not workable, like with Strategy Games featuring too many entities. In those cases, the gamestate would become excessively large, so sending over only the inputs becomes far cheaper. The input delay is a non-issue too, as the focus is on strategy, and not split-second timing.

Rollback is the new cool kid on the block. It was popularized by GGPO a few years back. It is utilized mostly in fighting games, as Dead Reckoning is too inaccurate, whilst Lockstep has too much delay. The rollback-warp that occurs in Rollback is also a non-issue in fighting games, as attacks generally come out so fast you can't react to them anyway. Fighting games also have barely any game logic going on, so the computational cost is negligible.

Embed is a GDC talk about people implementing rollback into their engine. Most of it is just hacking their own engine to make it work efficiently enough though.

b1ec9b  No.16602369

File: dc3b423f2f7a4d6⋯.png (191.19 KB, 590x379, 590:379, new inventory.png)


What about now?

20843b  No.16602459


Are you making a life sim? If so, can we know more?

60193a  No.16602461

File: 9c73ffa1b21f0d8⋯.png (241.75 KB, 590x379, 590:379, ClipboardImage.png)


Those colours are quite oversaturated and too green in my opinion

20843b  No.16602463


Forgot I still had the :^) flag.

b1ec9b  No.16602492

File: 80d9afde2b813e9⋯.jpg (251.23 KB, 850x1207, 50:71, cooking loli.jpg)


Are you talking about the whole UI or just the text fields?



Yeah i though you were being sarcastic there

My brother and I are making the comfy farm sim we wish harvest moon was, the protagonist will be the only man in a village filled with girls, and the features will all be very deep, we want to have the best cooking, farming and fishing to ever exist in such a game.

Combat, alchemy, and a large world map to explore are secondary features we want to implement one day, at first the gameplay will be focused on farming and the town.

You can read more on the devblog, i update every friday


60193a  No.16602747


Entire UI. I'd also change the gray bar at the top, but I'm not sure what I'd change it into.

b1ec9b  No.16602774


oh the gray is actually the sky, the UI is semi-transparent

b1ec9b  No.16602796

File: 1d1c49fe599433a⋯.png (372.45 KB, 732x505, 732:505, inv with background.png)


Here with the background, but i don't see how this is too green, i will test something with lower saturation

b1ec9b  No.16602811

File: ab6203ae6f32aa8⋯.png (369.95 KB, 732x505, 732:505, inv with background 50 sat….png)

You are right this looks better, i will change the colors

650933  No.16602902


A drop-down for sorting would be better, so you could sort by name, type, value per item, value of the whole stack, count, weight, etc.

0c6e7b  No.16602903


Totenmaske is looking good.

b1ec9b  No.16603095


I see, makes more sense than buttons indeed

e05beb  No.16603566


Looks nice. You are going to be homeless.

d43a53  No.16604735



e4d567  No.16605435

File: 8816f386e410ca9⋯.png (110.9 KB, 1080x1106, 540:553, Rust wins again.png)


>When June rolls around and cis het males start complaining, I will usually dig into why with some people. Inevitably, it almost always boils down to personal distaste for “gay shit”, which is usually code for “I’m not super secure in my identity”.

>You’re fine.

>To paraphrase Chris Kluwe: watching a Pride Parade isn’t going to make you start lusting after your best same-sex friend. If anything, it might give you a better fashion sense.

>The last thing I’m going to address is the sentiment that Pride month isn’t needed anymore. Rather than waste paragraphs explaining why this is not true, here’s a challenge for you dudes that say this:

>Live as a transgender woman for a month.

>I promise you, you will understand the need for it by the end.

You heard him, Rustfags. Understanding LGBTQ oppression will make you an even better programmer than all those cis white C weenies. Now go out there and be fabulous for a month!

d3189a  No.16605498



I know you want to write your cool essay on faggotry but I would prefer it if none of this rainbow shit stained the thread at all.

11a0fb  No.16605510





Why don't you kill yourself for spamming the thread with shit nobody cares? fag month is not vidya and much less vidya making

f066a5  No.16605525

File: 421d7705293c999⋯.mp4 (2.19 MB, 1280x720, 16:9, 2019-06-16 21-13-56.mp4)

File: f967f2ca8cc77d1⋯.png (83.34 KB, 1395x549, 155:61, ClipboardImage.png)





I ain't reading that shit nigger.

We moved the shop out from the dungeons and into what's going to be the hubworld. Here the player can buy new shit (obviously), select what dungeon to play next, change their loadout in their "home" or visit the graveyard filled with corpses of your predecessors. Also implemented a progression system so instead of having a random selection of weapons on offer you progress through tiers in three different branches, pic related.

cf662c  No.16605546


>fag month is not vidya and much less vidya making

Leftist shits are trying to make videogames political. I wouldn't support a dev/engine/platform that encourages it.

e9cfd4  No.16605817

I'm still waiting for the games

47cc4e  No.16606265

File: ee363ea5349ab6b⋯.jpg (126.2 KB, 547x642, 547:642, madshitforstinger.jpg)

aed8d7  No.16606481



>He came into this thread just to get his shit slapped


Where's Memoirs of Magic though? That shit's done to, although I don't think the dev posted their release here


4824da  No.16606647



The wiki is horribly out of date since anons aren't known for taking responsibility about some other people's entries and the fact that a lot of projects just get dropped without notification.

cf662c  No.16606650



MoM anon hasn't posted here in quite some time. I think he cut ties with us because of our controversies, but I'm talking out of my ass here.

4824da  No.16606653


Wasn't the guy also some giant furry who was into diaper shit?

aed8d7  No.16606660


That's weird considering he's been here for years, maybe even the GG exodus, and never cared.


>citation needed

aed8d7  No.16606661



Nevermind. It says it right on his page, although nothing about diaper shit


47cc4e  No.16606715

File: d0223b65f1020eb⋯.png (43.34 KB, 512x268, 128:67, No Escape.png)


>I think he cut ties


He can only hold his silence.Once an anon,always an anon.

We are warped too much by the cornucopia of ideas that are imageboards to cut ourselves off.Assuming he didn't migrate to a different board he should still be here.

I know it sounds stupid,still the truth

The only exception are Yandev tier spergouts,which we would have noticed

cf662c  No.16606718

File: 3672caba6803483⋯.png (353.24 KB, 515x614, 515:614, ClipboardImage.png)

>Official art


9b143d  No.16606724


I think the thing I hate the most, and I hate all of it, is the fucked angle the spine has to be bent in just to get the ass in the shot.

4824da  No.16606727

File: 9518b494f22643c⋯.webm (1.6 MB, 836x420, 209:105, NunFPS - Firebomb.webm)

Bomb doesn't do damage yet but will soon.

aed8d7  No.16606732

File: c80f9d67e9a59df⋯.png (827.14 KB, 649x559, 649:559, c80f9d67e9a59dfc618efd8aba….png)


Yandredev was able to cut ties because he's one of those opportunistic fucks that will sell anything for a little bit of power.


As soon as you pointed that out I noticed those retarded assless chaps

aed8d7  No.16606734


>Dynamite strapped to bottle

That's so fucking retarded I love it

4824da  No.16606740

File: 49c83f5eba35fda⋯.gif (32.46 KB, 136x228, 34:57, Cultist Dance.gif)

File: 2e7740547c00db1⋯.jpg (47.45 KB, 378x500, 189:250, molotov1.jpg)


Maranax perkele!

cf407c  No.16606745


Its on the same tier as dolphins with snorkels or birds with jetpacks

47cc4e  No.16606747

File: 262422c05f87756⋯.jpg (61.83 KB, 569x559, 569:559, SIN BIN.jpg)




The color implies that his boots go above his hips

47cc4e  No.16606750

File: 35728479b791955⋯.jpg (12.55 KB, 480x360, 4:3, I am not a clever man.jpg)


the sin bin is for the artist,in case it wasn't clear.

sage for doublepost

aed8d7  No.16606755


>Second picture

That's still not loony toons dynamite. I'd honestly keep it the way it is.

4824da  No.16606759


I'm not gonna change anything about it.

aed8d7  No.16606766

File: f3767a3ac95f644⋯.jpg (35.9 KB, 768x576, 4:3, f3767a3ac95f644cdd2d511270….jpg)

54844b  No.16607134


If you're going for cozy you're still a long way off. The window title is misaligned. The placeholder text in the fields has very poor contrast and clashes color temperature. Corner rounding is inconsistent between everything. You're mixing serif and sans-serif in what looks non-intentional. The window chrome itself is muddy and depressing compared to the bright, fresh ingredients. And I'm still bitter that you didn't even respond to the tool-tip feedback I gave.

6061f5  No.16607153

File: ab7a3bc4ad63efa⋯.png (105.76 KB, 487x274, 487:274, A separate place for every….png)

0ff4e7  No.16607726

File: 7c89751560e63b0⋯.jpg (123.16 KB, 336x422, 168:211, 56776324_p0.jpg)


Could you guys recommend any video course or other method of learning programming?

I tried to get into it once, but I burned myself out and retained basically nothing I've learned by doing way too much at once. Another thing is that while I was learning these basics, I tried to figure out how the car's engine worked when the lesson/exercise was just to turn the engine on because I didn't know why I was doing what I was told to do.

I don't even know if it's really for me, but I'd like to try it out one more time, even if I have to take things really slow instead and treat myself like a retard (because I probably am).

b8cb8e  No.16607733


Best way of learning programming is to program.

323632  No.16607763


SICP is pretty good.

6335b8  No.16607787


Not really relevant to the thread at all. Bringing it up if somebody mentioned the Amethyst engine would be appropriate, but this isn't the GamerGate thread, and random blog posts aren't needed here. This isn't a news thread for dumping general events tangentially related to game development.

b1777b  No.16607793


>Another thing is that while I was learning these basics, I tried to figure out how the car's engine worked when the lesson/exercise was just to turn the engine on because I didn't know why I was doing what I was told to do.

This is a good thing. It means you're curious and that's important to learn literally anything since you're naturally inclined to look for information on your own and fill the gaps of what's missing.

If this how you function then you don't need a video to explain it. All it's gonna happen is that you're going to see a video for an example, you'll start doing it but it will feel very constraining. You'll spend the entire time imagining what else you could do with anything they throw at you but you're following the exercise so you won't.

Best way to learn in your case is to do 2 things, possibly at the same time:

1: "Just, like, code!"

Pick an idea, any idea. Not for a videogame but for something that you can program. For instance, try to make an NPC that walks around and does several chores in a house. Program his routine, the way he recognizes and memorizes tools, the use of each tool, passive things that happen in the background (plants growing, dust accumulating). These are multiple things that you should take one at a time, but completing each one will teach you how to program, all you gotta do is find out how to do them.

For instance, the NPC will need to recognize tools for a task. How does he do so? Maybe the tasks have a tag regarding the correct type of tool necessary and every tool can have several tags that tell the game it can be used for that purpose. Maybe they have a magnitude for each tag and the NPC uses the best tool available for the job.

How does the NPC remenber the tools? Just their location? Or the container that has them? Maybe he can look around by raycasting and detecting objects with a routine to wander while he searches them?

How do you do any of this? Look information for it!

Instead of being handed a toolbox and told what each tool does, you're handed a problem and you go look for tools that help to solve it in a toolbox. You're not only seeing a practical example of everything you find, you're actually doing something on your own, can find plenty of short examples to learn, commit mistakes to learn from them and you'll find multiple solutions and tools for the same problem.

2: "Does this clown taste funny to you?"

Cannibalize code. Find someone that already made something similar to what you did, read their code until you understand what the fuck it all is and then incorporate it on your project. DO NOT COPY IT UNDER ANY CIRCUNSTANCE.

Copying teaches nothing and you'll often end up with something that works for their specific problem but isn't optimal for yours. If you truly understand how their code works, you can adapt it to your problem easily.

This has the benefit that you're gonna see someone else's style and details, learning a few tricks with them automatically as well (as long as you understand it) which can help correct a lot of problems and mistakes you'll make with self-learning. The more different coders you read, the more you can potentially learn and speed up your work later on. You'll also save a bunch of time since there's a lot of things that don't need to be reinvented. Screw the fags that tell you to code your own engine, you just need to retrofit the wheel to your specific cart and you're ready to go.

17f3f9  No.16607798


>video course

Seconding >>16607733

If all you do is watch videos then it's no wonder you forget all of it.

I'd unironically start with Javascript. Very easy to use and start experimenting without any other tools than a browser and a text editor. Then move to a compiled language when you have a general idea of what coding is like and how to make it do things.


>book that teaches a meme language used by no one and is radically different from the actually relevant languages for gamedev

Great idea.

e0422b  No.16607805


I'd recommend finding something that teaches you programming first. If you try to get into it with something like a Unity tutorial there's a good chance you'll find yourself easily frustrated. Because you're trying to learn both programming and games programming at the same time.

My personal recommendation would be those "learn programming in 24 hours" try to find one for the language you want to learn videos and try tackling them 1-2 hours at a time. Then once you're done with that, then you'll find it much easier to jump into tutorials related to programming for games.

There's a reason why colleges spend the first year of a game development degree teaching you basic programming concepts and usually in only one or two languages.


Java and Javascript are what I started with and having learned other languages since then, I can attest to them being some of the more beginner friendly languages.

323632  No.16607821


It teaches programming, not Scheme.

b8cb8e  No.16607829


It teaches functional programming.

b1777b  No.16607830


Seconding the Javascript. Just the way variables are handled (if it talks like a duck…) help newbies a lot since there's a lot of crap you won't have to deal with and with very little practice you can get to a point where javascript reads like a regular text book in english.

Once you're comfortable with it, changing to something like C# gives you more control and tools to work with while still being laid back and easy enough to use. There's probably other options here but you'll hardly need anything better than this.

323632  No.16607866


The iterative parts might as well be imperative and the entire third chapter deals with global state. Don't just repeat what you heard elsewhere.


In my experience the quirks of Javascript confuse beginners more than anything else, but C# is not a bad choice. It has a few game frameworks and the advantage that it's well supported on Windows.

36b5a7  No.16607897


Nothing wrong with video courses, you just need to apply what you've learned into personal projects. Don't copy, just figure out how to do it in a different way.

17f3f9  No.16607930


>In my experience the quirks of Javascript confuse beginners more than anything else

What experience? Are you a teacher?

I recommended JS based on personal experience. I'm not a very patient person but the simplicity and ease of use of Javascript made it very easy to get into and make simple game-like things with. The "quirks" that you need to learn out of are significantly easier to leave behind than it would be to learn a more complicated compiled language from the start, especially if you're not sure if programming is for you.

323632  No.16607989


>What experience? Are you a teacher?

Something similar.

60193a  No.16608014


I wouldn't recommend JS purely because it's not strong typed. When you're learning programming, it's essential that you have a proper idea of what you're working with at what times, and JS' "every number is a float" and ability to add strings to numbers and shit ruins that.

Simplicity and ease of use are good for when you know what you're doing, not when you're learning how to do things. That just makes the experience more painful when you do have to try the harder stuff, makes you think "ugh, why do I have to do this shit I want to go back to [easy thing]"

t. someone who couldn't program anything but GML for years until he learned C++ because he had to.

f066a5  No.16608018


Find any tutorial that is basically cloning a classic game like Pong or Breakout in a language like C#, Python, Lua or similar. Everytime you encounter something which isn't fully explained/you don't understand you google it and read up then you move on with the tutorial. This should in theory help you not get bogged down with irrelevant shit (at your level) that people who just pick up a book gets stuck on.

cf662c  No.16608047


There is exactly one (1) thing I like about weakly-typed languages: the ability to use an if statement on any object; anything that isn't null or 0 will return true. It's a wonderful shorthand for checking null

17f3f9  No.16608080


I moved from JS straight to C and didn't have any trouble learning the difference. Then again C is extremely clear about what it's doing so maybe it's an exception.

>Simplicity and ease of use are good for when you know what you're doing, not when you're learning how to do things

I think that's backwards. When you're first learning, the most important thing is to get a foothold in what you're learning and then start expanding from that. And the more things you add into the equation, the harder it is to get started and the easier it is to get discouraged and give up. When you know what you're doing, you want the ability to do precisely what you want and not have some simplified system obfuscate it away from you.

>makes you think "ugh, why do I have to do this shit I want to go back to [easy thing]"

I do agree with this though.

cf662c  No.16608101


The counterargument is that if JS lets you do things one way, and you start doing it that way, and literally everything else does it another way, you'll get confused and not know why

f066a5  No.16608149


Well chances are you are a fucking moron then who probably shouldn't be programming anyway.

cf662c  No.16608314

File: 41437570c8629ac⋯.png (6.83 KB, 671x161, 671:161, ClipboardImage.png)

I'm still trying to implement this, but not having any luck with how to organize my data:


>Takes an input data set

>Looks at each cell to generate rules such as (red, green, right)

>Also generates the frequency of each input value

>Supposed to be able to generate an XY field from it that looks similar to the input, but is procgen

I can generate the rules and frequency, but the actual output is what's messing me up. Every cell is supposed to have a list of possible values it can contain, but what I opted to do was pick values in the most "solved" cells first, eg, start in the corners and continue in that manner. In a way, it's similar to A* or other pathfinding algorithms.

>Not sure how to poll the neighbor cells without resorting to bounds checking constantly

>Not sure how to search my list of rules to prune out only a handful of possible values, without bruteforcing it multiple times every cell

>Not sure how to mutate my cells after every update (pic related), since the priority queue implementation I'm using doesn't allow direct member access

0ff4e7  No.16609179







Thanks for all the feedback and suggestions.

0f0be4  No.16609227

Has anyone played with the idea of porting in part or whole some of the open-source role playing rules into a video game? Stuff like OpenD6 or Fate or any OGL game , I've been having this idea for a month but I'm not sure if it is a too daunting of a task to do even in the most primitive form.

fc072d  No.16609233

File: 9e80d7f5e98af9e⋯.gif (217.87 KB, 320x240, 4:3, hrpbkLqxJF.gif)



FOV cones and radar icons apply to view targets.

b1ec9b  No.16609429


>The window title is misaligned

I was thinking of removing it instead, what would be better?

>The placeholder text in the fields has very poor contrast and clashes color temperature

Same thing i feel like removing it, i think it is pretty intuitive even without it

>Corner rounding is inconsistent between everything

The only non round is the background of the inventory frame

>You're mixing serif and sans-serif in what looks non-intentional.

I will unify all fonts later

>The window chrome itself is muddy and depressing compared to the bright, fresh ingredients

I tried brighter colors, but then saturation is a problem, I am not good at choosing colors

>and I'm still bitter that you didn't even respond to the tool-tip feedback I gave.

wait, where? Sorry if i missed it

084352  No.16610072

How do I git gud at modeling, preferably with blender?

Programming has been very easy to learn but I can't make any more progress since I don't have any assets to make a game with.

084352  No.16610081


also I'm still at the early point where I could make my whole game in 2d instead of 3d if that would somehow be easier, but I feel like that would be even more of a pain in the butt

985746  No.16610115



Just like practice. But honestly the complexity of making a 3D game, especially if you're doing a lot of animations, is significantly higher compared to a 2D game so if you're on the fence it might be worth reconsidering.

f35684  No.16610252


>role playing rules into a video game?

every time I read about it I always want to make a chargen/combat simulator for FATAL. the system is so autistic it should be perfect for it. also boardgames in general. tabletop simulator is nice, but when I play digitally might as well do it properly.

>too daunting of a task to do even in the most primitive form.

core gameplay logic should be pretty easy to do. I'm more wondering how to implement board interaction. either via code or some visual workaround (set a cell as holding a player/mob and target the cell or check if the camera can see it being used).

6b1fac  No.16610338

So… can I ask questions related to RPG Maker MV? Can you set up alternative attacks without plugins? or at least alter the skill ready animation of a type of attack instead of all having the magic skill ready animation?

60193a  No.16610713


>Not sure how to poll the neighbor cells without resorting to bounds checking constantly

Manually set up for loops to check the edges (no bounds checks needed because you know the bounds of the edge or corner you're working on), then use a standard double for loop on the center cells without any bounds checking.

An alternative is to expand the array by one cell on every of the four sides when loading it, with a special cell value for Out of Bounds, and only iterate over the center ones. Less code, at a cost of some extra memory usage, and the compiler might optimize it better.

>Not sure how to search my list of rules to prune out only a handful of possible values, without bruteforcing it multiple times every cell

Not 100% sure what you mean, but if you mean what I think you mean you can (probably) turn all neighbors and the current cell into a single number, then use a massive lookup table to see what the result would be. For reference, by turning cells into a single numer I mean you have the cells

>3 0 1

>0 2 1

>1 1 4

And turn it into


Or preferably keep it as small as possible with 3 bits per numer turning that into


Downsides are obvious: a large lookup table is required (I only used this technique once with only 1 bit per cell), and every lookup is likely to be a cache miss, but you'll just have to test if it's faster than brute forcing or not. Generating the number to look up is as easy as NANDing and bitshifting 9 numbers, that should be really fast.

These are both things I once used in an A* implementation for a project that got cancelled early, so I never got to test it extensively. And originally being thought up for the needs of that particular A* implementation they're likely to apply less to your problem (especially the second idea), but I thought I'd tell you about them anyway.

60193a  No.16610751


And nevermind the binary thing and bitshifting, for non-boolean cells it's probably better to do it with multiplication by factors of 10. That only worked well in my A* thing because it was always 8 bools, so it fit in an 8 bit number and could be SIMD'd. You need to use the center cell as well, so it doesn't scale nearly as well with the bitshifting.

6b1fac  No.16611470

Do you guys get burn out?

I've been starting to work in one game myself and only creating the assets are taking me forever. And I didn't even finish up the protagonist's sprites.

I'm just feeling stressed because I had to redo some work for a few times and now I find other mistakes I can't fix by editing and just feel discouraged. It's my first game and I shouldn't even care, but I want it to at least have some care. How do you get your morale up while working in a project where you are the only one working on it?

3de771  No.16611485

Anyone know anything about GBA development? Couldn't find anything on the wiki (only original GB/GBC) and cursory research brings up fairly old resources. If anyone knows a good place to start, or good tools to use, it'd save me a lot of time. I haven't done console dev before but I'm experienced in C and ASM so I'm not too scared of the challenge (read: very scared)


Yes. The key is to keep the scope small enough to push through before you get burnt out. A first game should be a pong/asteroids/tetris clone, not a full thing. Don't be afraid to put it to the side and work on something simpler, get some experience.

b1ec9b  No.16611508


>I'm just feeling stressed because I had to redo some work

Get used

6b1fac  No.16611517



The thing is that it's just some shitty RPG Maker game, but the little template sheets I found for the assets I'm working on, just because at least I should work in the sprites if I'm going to be lazy enouth to do an RPG Maker game, And just drawing is taking me hours.

60193a  No.16611518





Why would old resources matter when your target platform isold? I highly doubt there's active tool development going on for the GBA.

b1ec9b  No.16611530


>And just drawing is taking me hours

You will get faster as you get better at it

The problem is your expectations that "It's just a RPG Maker game" are totally wrong. There is no "It's just". Everything worth doing takes a lot of effort to do, no matter how "simple" it looks to be.

f066a5  No.16611552

File: c3e432079dbfc0c⋯.png (45.99 KB, 605x639, 605:639, ClipboardImage.png)


I've tried and failed to make probably 20-30 games and engines by now. Yeah I get burnt out, hard. Gamedev is hard work and anyone who says otherwise is lying.

We just implemented randomly generated acid pools for some variety in the visuals. Some floor variations, dressing and props on top and maybe we got something good looking. Also made the walls 2 tiles high so we can do things like wall mounted traps or whatever.

b15267  No.16611560

File: 7bd569b34b6b15d⋯.jpg (263.57 KB, 600x600, 1:1, 64fce50dee2c6b859dc20326ee….jpg)



Really, get used to it is the best advice we can give. As your skills increase so will your drive and your ability.

Things will become less monotonous and much faster the better you guy.

Also, I just had to redo my games entire repo and file structure. I wish I only had to redo a character sprite, kek.

0fbf45  No.16611601


>target platform is so old

But computers and emulators are new. There's always new tools to things, and sometimes the newest thing is the worst, or a newer thing is better. I was asking here before doing research to see if anyone experienced knows where I should look first. I'd rather not waste a day trying out 12 different things only to find they don't work or aren't what I want. If you want me to do the research myself, just say that instead of posting 3 barely useful links I already found on my first web query.

6b1fac  No.16611626


I guess so. I just lowered my goal from a 100 or so characters RPG with each having personal skills and stories with night and day cycles with just a multiending lovecraftian RPG travelling through dimensions on the dream world.

Still, just drawing one character is much harder than I thought.


Fuck, I've even been thinking about doing alternative clothes, but after finishing one I just thinkg "FUCK THAT!".


I guess so. I'm ust pushing through it just not to add this game to the pile of abandoned projects which I'm just waiting to have some actual skill to translate into their respective mediums.

f066a5  No.16611704

File: 877146b573bf4cf⋯.jpg (43.48 KB, 500x500, 1:1, Remote_dog.jpg)


>100 or so characters RPG with each having personal skills and stories with night and day cycles with just a multiending lovecraftian RPG travelling through dimensions on the dream world.


6b1fac  No.16611728


Fuck, I got an eunerism. I meant that I wanted to do something big and then I decided to do something easier (in theory).

Something involving the same town, the same characters and different dungeons accessed from the same spot. Maybe get some multiple ending by playing with variables and Javascript.

Also RPG Maker has almost all relevant plugins included in some hidden DLC.

b1ec9b  No.16611734


>I guess so. I just lowered my goal from a 100 or so characters RPG with each having personal skills and stories with night and day cycles with just a multiending lovecraftian RPG travelling through dimensions on the dream world.

<Just a small RPG Maker game

6b1fac  No.16611750


I just got too ambitious at first, but decided to do something humble and still is hard as fuck.

ec9851  No.16611808

YouTube embed. Click thumbnail to play.


Audio and new map test, getting an idea for how to plan the radar. Still working on radar functionality, need some more time on that. For demo day, I want to have it so if they camera sees you, you explode with a firey explosion sprite.

8ea0d4  No.16611833

File: e12e9c915b3de9f⋯.mp4 (1.81 MB, 640x360, 16:9, 2019-06-18 13-46-58.mp4)

Got some progress. Camera can target objects to lock into their location.

Don't why the character is upside down, I'll look into that.

To do:

>smooth camera transition when locking onto an obj


>replace float arrays with vbos

>testing the physics implementation

1% done bois, 99% to go…

51c5df  No.16611864


RPG Maker isn't exactly recommended for large projects. The simple way of using it hides the fact that the work builds up in volume as compensation. It doesn't do a good job teaching you how to code much either. Animation viewer is pretty good though, lots of practice with it gives you practical experience I find.

b1ec9b  No.16612100

File: dc8c7ebb7916bb0⋯.webm (1.72 MB, 1280x720, 16:9, 2019-06-18 18-56-06.webm)

Rate our UI

Be gentle

b1ec9b  No.16612116



I remember all those RPG Maker projects i started and dropped, I am glad i finally got something i can keep creating without burning out

f066a5  No.16612407


Looks functional but ugly.

d51743  No.16612658

>>16612100 (checked)

It looks okay, but I would change a few things. First off, I would make the borders of the inventory bigger, to give off a less clinical feel, which is what you have going on at the moment. This includes changing your minimize and close icons to something more stylized. Secondly, I would change the font of your inventory around to something less basic, and resize parts of it as well - on a similar note, the white lettering is difficult to read and strains the eyes. I would recommend making the window bar at the top less transparent, or just totally opaque. An obvious change is better icons, but you're probably using placeholders there so I won't judge that. And lastly, change those popups. Way too much information popping out all at once. I would make two states - your basic pop-up, with a name and some very basic info, and then an advanced one that appears when you press shift or hover for several seconds or whatever. Likely, just keeping that top little bit of info you have (the name and freshness) will do.

Overall, I'd call it workable but I wouldn't write home about it either. At the very least, the functionality, with the search tabs and sorting, is something that doesn't need an immediate rework and should hold up well. Unless you're planning on gamepad support, in which case. Well, I think you can tell.

cf662c  No.16612667


>Factors of 10

Or you know, it'd be representing each number as a flag, which means I'm still limited, eg an Int32 would only allow for 32 possible values. Which works for simple objects, but what if I'm trying to work with Color values?

aff4ca  No.16612825


Sort them out instead of adding filters and drop down lists. Even if its bigger, if it is broken up into pre-set categories then the user will have an easier time sifting through it.

b1ec9b  No.16612839

File: 1d9cfa3f62d8fcc⋯.jpg (293.44 KB, 1000x800, 5:4, tooltip test.jpg)




>First off, I would make the borders of the inventory bigger

Bigger borders with some details drawn would 20 px be big enough?

>font of your inventory around to something less basic

For readability sake i am not changing the font

>white lettering is difficult to read and strains the eyes

Are you talking about the "sorting, name, tag" parts right? i can change those

>An obvious change is better icons

In the future i will make 3d icons, but these will have to do for now

>And lastly, change those popups

The tooltip will have a hide section button, but right now we want to focus on finishing the cooking system before anything else, we don't want to repeat the failure of the last demoday


That is what the sorting dropdown will do

Also if you guys have any samples of UIs you would rate as excellent i would love to study them

a9526a  No.16612848


I've never done GBA dev before, just fucked around with GB, NES, and Genesis. That being said, you're dealing with an ARM processor. That means RISC. If you have any previous experience in CISC ASM, it's gonna be only halfway useful, since you need to manually do some stuff you used to use a single opcode for. Thankfully the GBA processor is fast enough that you can do C/C++ without slowing to a crawl. Unless you plan on trying to push the console to the absolute limits, just use C/C++.

In all fairness to that anon, what tools may work well for one dev might be a nightmare for another. I prefer using ASM with a generic 68000 compiler on the Genesis, but I wouldn't recommend it to another person looking to get started on the console. Likewise, I wouldn't be able to deal with the abstraction of C/C++ on the Genesis, it would drive me crazy. Barring there being only one tool(PSX dev) or one source of information(actual official design documentation), you really do need to do your own research for tools and guides. If a place like gbadev has it listed under documentation, then it probably has some worthwhile info inside of it. It's never a guarantee you'll get all the info you need without some serious digging either. It was a pain in the ass to actually find info on every Genesis VDP register, many places had incomplete info.

271528  No.16612987

Shmup anon here.

I think I did most of the dynamic animation code, and test code to move a turret.

Except there's something fucky somewhere and the shader just show me 1 frame before making models disappear.


So currently at that point, hopefully it'll work soon and I "should" be good to make levels and boss. (beside missing sound effects and music, and visual effects…)

dd795a  No.16613092

File: e70a81a2e0a9a1c⋯.gif (377.21 KB, 320x240, 4:3, JouGF0LifJ.gif)

519f59  No.16613146



Don't be a fucking gay.

You were trying to make it functional right?

dcf72c  No.16613186

how to start programming

e21e7c  No.16613285


just like

60193a  No.16613318

File: 3b3e24304d0dd44⋯.png (457.42 KB, 834x648, 139:108, cpp in one video.png)

44f511  No.16613335


Search C for absolute beginners

19dcf2  No.16613354


Either you start when you were 12 or give up hope of ever making it.

985746  No.16613432



Alright so here goes


>searches should apply as you type (looks like you have to press enter or something)

>popup covers the inventory so you can't see what you would want to look at next

>can't pin a popup so you can compare to different items


>window is too transparent

>colors are unappealing to look at

>need more margins for inventory text and top right buttons

>left margin is bigger than right margin

>font sizes are off, inventory is (at least perceived) as smaller than the text on icons

>font is boring and overall too small

>text placement in icons feels off, try the lower corners instead

I think that's it for now.

9f79f9  No.16613558


>watch a coding tutorial

>do this because i say so, i'm not going to explain what this does

>have no idea why i'm doing things

>no logic behind what i just did

>don't understand a thing

>give up

I wish I could have done programming in high school

985746  No.16613561


>thinking academia can teach you good programming

oh boy

9f79f9  No.16613576


They could have taught me at least programming, I read a book, I do some code, I look at someone else's code or decide to write something for myself, I get lost. nothing makes sense to me, well basic ideas do, looking at someone else's code is the worse

I wanted to modify Yamagi Quake to make a game of my own out of it, really just interested in implementing a class system, a player progression system, the menus to use these and a few other things like forcing players to gather at an exit point to move onto the next level, couldn't really figure things out, even after reading a book on C that was actually quite well written, I can really only do simple tasks, it doesn't make sense to me, and other peoples code just confuses me even more.

Just going to stick to art, that's something I can do definitely.

985746  No.16613590


>really just interested in implementing a class system, a player progression system, the menus to use these and a few other things like forcing players to gather at an exit point to move onto the next level

>I've never held a pencil but I saw the Mona Lisa once so I'm going to replicate that

What did you expect? I did 5 years of CS specialized on game making at uni. Our first projects where in order:

>whack a mole

>svälta räv (swedish card game, very simple)



>zuma clone

That's about 3 months to do all that simple shit and that's just getting the basic gamelogic and rendering done. It took 3 years to be able to write a networked 3D game engine from scratch and make a game with it and even then that iteration was complete trash. It took probably 3-4 game engine iterations to be able to produce something that was actually decent or useable. All university did was provide a direction and free time, most had to be taught on the side.

ef6a7f  No.16613600

YouTube embed. Click thumbnail to play.


I have a friend who plays drums and is excellent at writing music, who has offered to make my music for my game. Can I get some opinions on this song in particular? I would like something close to this and MGS1's soundtrack.

b1777b  No.16613630


Takes a while to get there, but it's pretty cool so it might suit a slower-paced game with some tense moments.

aee18f  No.16613842




and before someone spergs about functional programming, it's the ideas that you learn that matter.

b1ec9b  No.16613905


I want to make it the best it can be… without taking too long because we have other stuff to be made


Thanks, notes taken

94256a  No.16614027


Install Unity Soylent™, the universal engine replacement. Like its namesake, it is so useful that you will never need anything else and if it isn't working for you, it's your own fault for not chugging it harder.

e43bdc  No.16614049

File: a061f975fe2075d⋯.jpg (439.9 KB, 2048x1578, 1024:789, a061f975fe2075dc0911c06d90….jpg)

I know I'm missing something obvious here, but I can't figure it out.

OpenGL refuses to draw GL_QUADS and GL_TRIANGLES unless I specify texture coordinates. I just want to draw a single-color triangle and it does not work. Used to, though. I traced it back and it's caused by glTexImage2D, as removing it will break textures, but then the triangle will show up. Unbinding textures and disabling GL_TEXTURE_2D didn't work either.

What am I missing here?

d3189a  No.16614058


>Used to

Most likely it's some lingering state from some other code that you were too lazy to write cleanup for.

Why don't you attach an OpenGL debugger and verify that all of the calls you're doing are the right ones?

e43bdc  No.16614124


Going through it, I found that if I remove glBlendFunc, it draws it but in black instead of the color I have set, red. It just got erased by the blend function since it's all black. Considering I set the color right before drawing the triangle, I wonder why that's the case.

e9cfd4  No.16614260

File: 846192ab1b690a4⋯.jpg (15.29 KB, 479x163, 479:163, mom dev.jpg)

from MoM Dev

60193a  No.16614281


Is he 'strayan/kiwi?

c23460  No.16614286



He should. I did. I don't trust the Australian government any more. I just want to read and shit post about videogames and now they're treating me like a criminal. "Oh, it's just a temporary block until the video isn't being shared", and now months later, still blocked.

9f79f9  No.16614505



Thanks I kind of wish I had this advice in the first place, I'm just going to focus on art until I finish what I'm doing, then spend money on hiring someone to code for me, and then get mad at wasting my money on that.


I live in the country where that happened and I can post here.

0877f1  No.16614540

File: 3f7ce6c6b06e8cc⋯.webm (2.89 MB, 1600x900, 16:9, PONG RPG 65.webm)

File: 9d96f0bef2e1e78⋯.webm (2.87 MB, 1280x720, 16:9, PONG RPG 66.webm)

I "done" with my game, all that it lacks now is the story mode but i can aways patch it in later.

bb29cd  No.16614809



What version of OpenGL are you writing for? It sounds like fixed-function (1.0) as it doesn't sound like you are using shaders.

Managing GL state is a bitch. Use a GL debugger such as XLcode. verify the state is what you think it should be when you start drawing triangles. You might have an empty texture bound to the state but are not giving texture coordinates with your vertex coordinates.

b1ec9b  No.16614836

File: 13e7e35279602d0⋯.png (532.17 KB, 800x600, 4:3, inv.png)



What about now? Still remaking the top part, but everything bellow had been changed

The white text need to remain now that the background is much darker

77162e  No.16615061

File: 00a0b5f319dee0a⋯.jpg (1.62 MB, 3000x2000, 3:2, Nanner run chars.jpg)

Some character art for a game I've had ideas for that have been swirling around in my brain for a bit. Dunno if I want to post the premise of the game, nothing's set in stone.

0e41e2  No.16615071


.webms is borked

e43bdc  No.16615133


Aiming for as low as possible, yes. From hardware tests, it runs on at least 1.2, maybe earlier, though the texture functions I have implemented are from 1.1, I believe.

Either way, I did eventually figure out the issue. It was just a matter of disabling GL_TEXTURE_D before drawing colored polygons, but there was a mess of bullshit that I had to clear through that was conflicting with each other that I hadn't run into issues with until now. That's what I get for trying to MacGyver functionality from dozens of sources without fully understanding what I'm doing.

the best part is that this was an issue I ran into while trying to figure out a subset of a subset of a subset of what I'm trying to implement. I was initially working on loading in and rendering .obj files.

f066a5  No.16615318


Looks significantly better anon.

bacf60  No.16615464

just lost countless hours of work because I imported a package that overwrote several very important scripts

serves me right for not making backups but holy shit thats infuriating


8chan was blocked for me for about a week after the christchurch shooting, I guess its just a matter of how cucked your isp is. That or I'm now on a list.


lookin good, though if you're still taking criticism, the font of the item count clashes with the rest of the style

00e966  No.16615515


>making backups

Use git or some sort of version control please. It's more than just backups, it helps to see a diff of files from the last time it was working.

4de363  No.16615674


>start coding at all

>without a version control system

That is really stupid. Stop whatever you are doing and use git or something.

df149e  No.16615803

File: acaf9cb0109f9fb⋯.png (123.53 KB, 800x800, 1:1, untitled.png)

how do I make this look not crap?

5f4080  No.16615818

File: 3a91e84c028f433⋯.jpg (60.44 KB, 847x406, 121:58, assface.jpg)


How do i make this not look like crap?

6d5a72  No.16615909

File: 234c0b6e250e1a9⋯.jpg (Spoiler Image, 4.89 KB, 225x225, 1:1, poopy.jpg)


how do I make this look not crap?

4ebb6e  No.16616107


If you want to model something resembling a human, learn anatomy and topology and sculpting. Read the shinobro tips on his thread on /agdg/ and the threads on /3d/.halfchan's /3/ will give you some directions, too.

68f95e  No.16616516

File: d002c7119bf31da⋯.png (276.17 KB, 504x900, 14:25, d002c7119bf31dabbd435ee3a7….png)

I'm in a bit of a pinch, lads. Do you have any manual, book, video tutorial or resources to learn how to make a competent 3D model starting from a bunch of reference pictures? Bonus points if it's low-poly too and if it helps with some compression shit with the textures, can't do that either.

Also, do you have any good manuals on data/image compression for gaming? I know that the main gist of those hypercompressed textures is all about using DoG techniques, but I've never implemented in practice.


Sure, we'll try.

9f79f9  No.16616527


Practice makes perfect.


You don't make any sense.

d3189a  No.16616548

File: 8adf0cbfe96e5ac⋯.jpg (23.03 KB, 474x208, 237:104, bcaea94924cd1db774482a93eb….jpg)


The usual process is to first produce a few reference images that directly face the orthogonal axis. (Pic related.)

After that you use a feature like this https://en.wikibooks.org/wiki/Blender_3D:_Noob_to_Pro/Background_Images to use the 2D images as overlay or backgrounds, and then just trace the thing with vertices on each axis.

If you were expecting a pleasing-on-the-eyes-image like the one you posted to be easily transferable to 3D, you're gonna run into a lot of problems.

Things that look good in 2D do not necessarily look good in 3D, as the 2D>3DPD is not actually just a meme but also an objective truth.

68f95e  No.16616930

File: 0a709b3bd016447⋯.jpg (40.89 KB, 521x516, 521:516, 0a709b3bd01644755a1a9715b0….jpg)


>You don't make any sense.

If you meant this about my spiel on compression, I was trying to say that I've read some theory behind it and from what I could gather it's a linear transformation of an image into its anti-aliased, smaller version and its remnant details, but for the life of me I can't figure out how you could possibly save space with that and was wondering if there were any practical resources to compress textures and data.


I see, visually it helps out quite a lot, thank you.

>If you were expecting a pleasing-on-the-eyes-image like the one you posted to be easily transferable to 3D, you're gonna run into a lot of problems.

I really wasn't, but now you're making me wanna. I hypothesize that a similar, not identical, looking model could be achieved by projecting that picture to be as close as possible to what you posted with some projections, beats me how though,

>Things that look good in 2D do not necessarily look good in 3D, as the 2D>3DPD is not actually just a meme but also an objective truth.

True enough, though I've heard of some developers adding more tris than necessary to some areas of their models so that when they get to animate them they can squish and enlarge features to mimic animation smears and make all actions feel hand drawn or something along those lines.

050cfa  No.16616953


If I remember correctly, you would normally add extra polygons around areas that would be most affected by rotation and movement. So, elbows, knees, around the mouth and eyes, etc.

d3189a  No.16617009

File: 1a9728dcdbe41f4⋯.png (34.66 KB, 865x384, 865:384, distance-field-fonts[1].png)


>If you meant this about my spiel on compression, I was trying to say that I've read some theory behind it and from what I could gather it's a linear transformation of an image into its anti-aliased, smaller version and its remnant details, but for the life of me I can't figure out how you could possibly save space with that and was wondering if there were any practical resources to compress textures and data.

Compression is usually achieved by putting restrictions elsewhere.

One technique that does what you describe is Signed Distance Fields, which is utilized in some games (e.g. source engine games) to render text.

A signed distance field texture is first constructed from a vector shape (e.g. a font character) or a very high resolution image (featuring only pure black and white) by storing the distance to the shape/nearest black pixel in each pixel. This value is signed, with negative values being inside the shape.

This signed distance field texture can then be sampled in rendering to determine the distance to the edge and whether the texel is inside or outside the shape. This can be used for a variation of effects; like font anti-aliasing and outlines.

In essence, the signed distance field texture contains nothing more than a very good estimation of a shape, compressed to a very low resolution. Raw vectorized data is often still more memory-efficient and accurate, but you can't texture sample that.

Signed distance fields do not support color, nor do they even support sharp corners! But it is possible to use multiple signed distance fields to produce these effects. (A signed distance field is single-channel, so its easy to put 4 in a single texture.)

OpenGL / D3D also have a handful of built-in compressed texture formats that you can use. These usually come in the form of sharing a few bits between adjacent pixels.


bb29cd  No.16617055


Yeah, that's OpenGL for you. I had much the same experience learning it. It does get better. Try using the push and pop functions for GL state, so you can restore to whatever GL state was at after drawing things like UI, text, etc.

Also some state changes take significant time. You can use XLcode to figure out which these are and avoid changing them per-frame if you can.

b1ec9b  No.16617332


But the numbers even use the same font as the other text instances

b1ec9b  No.16617342

File: 0fb7fa76fcc51c3⋯.png (433.65 KB, 1280x640, 2:1, topology reference.png)


Study references

You can download the models from recent pokemon games and open them in Unity to study the topology and proportions

36b5a7  No.16617791



1. Sculpt your model

2. Retopo

3. Fix all issues with your model (n-gons, double verts, non-manifold edges, etc) before you create UVs.

4. Smooth your low poly model

5. Create UVs

6. Duplicate the low poly model and subdivide it to match the poly count of the sculpt, label it "high poly".

7. Project details from your sculpt onto your highpoly

8. Bake normals from your highpoly onto your lowpoly

9. Texture

10. Rig

11. Weight paint

Create basemeshes like hands or feet, most people won't notice and those parts are the most time consuming to do.

647807  No.16618116

File: 2fb5db435a3edfb⋯.png (502.58 KB, 833x787, 833:787, blender_2019-06-21_00-49-3….png)

File: 676e5e90a023280⋯.png (477.6 KB, 827x755, 827:755, blender_2019-06-21_00-49-4….png)

New model WIP

348452  No.16618145


Looks nice but consider using less polygons

647807  No.16618165


I'm using subsurf. Is that a bad idea when making organic models? I wanted to give it a try after only making hard surface models for so long.

7d32cc  No.16618199


No. Using subsurf is a great idea when making organic models, especially because topology problems are made even more apparent when you apply subsurf. Unless you're targeting older systems or trying low poly modeling as an exercise you don't need to worry too much about polygons. What you want to avoid is wasted detail. If you have a 20k polygon model, but a 15k model looks and deforms basically the same, then something is wrong with that 20k model.

How many polygons is this guy?

dfc1e8  No.16618240


It's definitely not finished. I haven't baked and turned the thing low-poly yet but as of right now this thing is around 18k faces and 36k tris if you count the high poly mask. Without it it's only 8k faces and 17k tris. When people ask for polycount which of those two do they refer to anyway?

In any case my plan is to continue making the rougher shapes and details and then sculpting finer details onto the clothes.

cf662c  No.16618319

File: 0e44d4b99d85125⋯.png (22.51 KB, 1294x780, 647:390, ClipboardImage.png)

>Make a random walker

>Generate a polygon

>Convert the polygon vertices into map cell coordinates

>Any cell that partially intersects the polygon is marked as used

Still trying to figure out some room shapes… What technique is this? Polygon rasterization?

cf662c  No.16618325


It looks like I can solve this mostly with Bresenham's line algorithm, but how would I determine the fill shape?

7d32cc  No.16618328

File: a451cc05e83775b⋯.png (812.87 KB, 1024x576, 16:9, sas-model-update.png)

File: a53153bafe05cd7⋯.png (145.29 KB, 307x504, 307:504, CSGO-SAS-Polygons.png)


>When people ask for polycount which of those two do they refer to anyway?


17k tris is on the low end of current gen games, though high for the amount of detail that model has. Pic related is the CS:GO SAS model, which, not counting the guns, has 16,471 tris.

cf662c  No.16618333


Don't 32-bit systems have a limit of 65536 vertices per draw call? Or am I mistaking it for something else

dfc1e8  No.16618388


I will probably be aiming for something similar to that or lower. My favourite games were from around the 2004-2008 era and their look still left an impression on me.

Do you know where I can get more reference like that picture?

7d32cc  No.16618475


This if the first time I've heard of anything like that. Looking around, this is the closest thing I can find that resembles what you're talking about.


Personally 65536 vertices seems waaaaay too low for a 32 bit system.


I got that wireframe image from a 3kliksphilip video.


Which he probably got from placing the model in the hammer editor and taking a screenshot.

I got the first image by searching "CS:GO SAS 3D model".

cf662c  No.16618507


This was a limit with XNA's DrawPrimitives calls. Buffer size was limited for some reason.

dfc1e8  No.16618536


Thanks, I guess it's time to install Source SDK again

7d32cc  No.16618547


Seems like a software limit, not a hardware limit.


Happy to help.

12a6a5  No.16618692

YouTube embed. Click thumbnail to play.


Hopefully it works now

cf662c  No.16618739

File: b3fef6139843447⋯.png (6.17 KB, 247x751, 247:751, ClipboardImage.png)

>Try to create non-rectangular rooms for the sake of being interesting

>Look into random points, concave polygens, cellular automata, everything seems too involved or not quite what I'm looking for

>The solution seems to be just sticking random rectangular shapes together via composite

Fucking gamedev

bacf60  No.16619474

File: f9f788ac82be291⋯.png (132.87 KB, 895x806, 895:806, colors.PNG)

heres a visual debugger for my implementation of dijkstras algorithm

f066a5  No.16619501


That's pretty much how I do it. I also punch down walls randomly to connect shapes further. That said I'm probably going to use cellular automata for caves later on. It's not that hard to implement but you need a secondary step to make sure all rooms generated are connected.

f066a5  No.16619606

Free game conference going live in a few hours. It's organized by (((Rami Ismail))) and most of it is multiculti shit but there's some stuff like "Creating Characters in Low Resolution" and "Making LA-MULANA's 2D Graphics in 3D" that I think agdg might find interesting.

f066a5  No.16619607


Forgot the link because I'm a massive faggot. https://gamedev.world/en/

d3189a  No.16619646



Time to wait for 3 days and then just watch all the good talks so I can skip through the irrelevant shit.

I also find the whole gamedev conference thing to be kinda weird. Gamedev is for most developers (everyone speaking at the conference) a capitalistic venue. Aside from bragging about the cool shit you made and marketing your company, why would you speak there?

I get the feeling its going to be filled with people that fall under the category of "those who can't do, teach". If you know what you're doing, why not establish a business and directly train people to get results?

That's not to say it's not worth watching and can't be educational, just that adequate amounts of salt should be provided.

17f3f9  No.16619667


>all these women trying to tell people how to make games or run a company

>immediately a title with "discrimination" and "women" in the same sentence

>"how to (((globalize))) games and avoid hurting feefees"

>"how to ban wrongthink in online games"

>"what it's like to be a nigress artist from africa"

<top donors: google, facebook, intel

What a fucking pozz load. It's a shame some good people usually get lost and end up in these.

b1777b  No.16619680


I would like to think that it's because some Devs treat this as an art or craft. They want more and better games made and share all they know to increase the amount of monkeys at the typewriters. They don't want to be the only ones that create something since, even if it made a profit, it'd be a single game instead of the many that could be made instead.

I would really like to think that.

There's probably some that think like this and even some that actually act like this too, but they're bound to be a minority of a minority of a minority.

d3189a  No.16619697


The issue with that is speaking at such a conference, filled with talks about sexism and toxicity, is a gigantic waste of time. You could just dump your findings online. That way people can just fucking read instead of wasting an hour listening to your talk, and you can directly link to things like shadertoy and youtube for examples or sources.

After that, places like agdg wiki can compile these articles together to create a knowledge base.

There's plenty of people who post stuff online to help out gamedev as a whole. I post plenty in these threads for that reason. But if I ever reach the point that I have my own studio I would divert all those efforts into developing my employees instead of random strangers on the internet, which has mostly the same effect but would directly benefit me instead.

6b1fac  No.16619704

Shitty question about RPG Maker MV.

In the notes on items, actors, monsters and variables what progamation language does it use? javascript or HTML5? I just want to do some weapon that needs reloading but I don't want to sacrifice the MP to put the bullets.

b1777b  No.16619722


>if I ever reach the point that I have my own studio I would divert all those efforts into developing my employees instead of random strangers on the internet, which has mostly the same effect but would directly benefit me instead.

That's understandable, but in the end you'd have a single highly trained company making games. If your profit matters that much, it's the best thing to do but if you want a better market out there because you don't think you can meet all the demand, training other people would be a better option as well. As long as it doesn't harm your company, of course.

Personally, I'd rather sell whatever project I finish to support development of other projects but keep the source code entirely open to the public, providing as much help as possible to anyone that wants to read or change it.

Not only is that a great way to help the modding community, it's a good example for upcoming gamedevs that you already had done anyway.

8297c1  No.16619921

File: 555d4e99456e61f⋯.png (2.07 MB, 1280x854, 640:427, chechen war_akihabara.png)

Alright boys I've finally gotten to the point where I feel like my halfassed attempt at cobbling together a game will have more technical depth than a AAA title. Time to stop being ideasman and just make game.

weeaboo fukushima stalker setting idea, oc please dont steal

60193a  No.16619960


I can already taste the nu-lowpoly hipstergarbage that that "Creating Characters in Low Resolution" is probably going to be. It tastes like plastic and shitty latte-frappechiatto with soymilk.


Frankly I don't want to improve the people in the current industry. I want the current industry to burn to the ground. I'd rather help the people who rise out of its ashes because they like videogames, when all the fuckers who were in it for the money and the SJW infiltrators are gone.

f066a5  No.16620226


Well as most have noted many at these talks have jewish agendas loading out some pozz to gulp. Other than that it is usually used as a marketing opportuniy for the companies.

cf662c  No.16620756


For Super Metroid however, that generated room I posted is way too big to encounter regularly. Most larger rooms are rectangular, with sizes like 4x3 or 3x2. They have a horizontal bias too.

Do you track intersections for room pieces? My method was to just take a small shape and move it until it is adjacent to the base room shape. If multiple addon shapes overlap, its fine.

My next challenge is figuring out the size and position of the rooms, now that I've (potentially) solved the shape. Hallways will run between major rooms in an HVH or VHV manner and it should be fine.

f066a5  No.16620773


Yeah I track intersections to know where I can place doors and avoid spawning things such as pits and traps. It's probably the most used metric when I populate maps.

271528  No.16620936


I fixed it.

Not sure if there was some specs change in shaders.

Before I was using:

in ivec4 boneID;
in vec4 weights;
int some_bone_to_eval = ... boneID[0];

And it worked, but when using the same definition with dynamic animation, it wouldn't work.

Setting all chained variables from the engine to shader for fixed values I concluded that this variable was the problem and rewrote it as

in int boneID[4];
in float weights[4];

While being unsure if it would still take me 2 vertex attribute or 8, which would put me over the limit of 16, but this work and I think it's how you define it with the glVertexAttrib*Pointer, since I said it would have 4 values for this attribute, then I only use 1 attribute.

Then it's minor problem with copy/pasting while being sleepy or something since I half-edited the pre-calculated animation part and half-edited the dynamic part, which resulted in nothing working correctly.

Now the test turret rotate, but rotate around the model's origin instead of its own. This should be part of the dynamic calculation I guess and I would need to evaluate this against the bone structure.

24cf5d  No.16621150

File: a7387c4f903b2eb⋯.webm (2.64 MB, 320x240, 4:3, combat.webm)

Doing box on box detection on a Z80 CPU without making a mess of jumps is quite fun.

7d32cc  No.16621224


Looks like a pretty nice mega man clone.

24cf5d  No.16621235


That's pretty much the idea.

This and shilling for the Screen 4 video mode, because everyone that does a thing to MSX 2, use that awful slow Screen 5

901711  No.16621357

What's you guys' schedules like?

I'm kinda interesting in learning, but programming seems like it'll take years to learn and I don't really wanna do it for 8 hours a day

24cf5d  No.16621386


You need to benchmark yourself, but try treating it like playing dorf.

Just get a simple language, an very primitive and easy project like doing a pong and have fun trying it out.

6b1fac  No.16621391


I've been trying to work in my game when I have a few hours free until I feel burnt out. Thankfully I'm using a game creation engine.

e43bdc  No.16621398


For programming, I used a few tutorials to get going, and from there, it was a process of logic. I know what I need to do, I logic out HOW to do it and look up functions that should do what I need. If that doesn't work, I try another approach. I have a general grasp of what I'm doing by now, at least, though I was blindly stumbling with no understanding what half of what I wrote did for at least a month.

It's not as hard as it seems, though starting is a nightmare.

As for schedule, it's whatever I can muster, really. I don't force myself to do it, since I'd probably resent it if I did. I take a bit of time off nowadays, at least half an hour, to play a bit of vidya myself, since it helps with inspiration, as well as helping relax. Naps are helpful too, it's often good to come back to a coding problem after you're refreshed a bit, helps thinking.

f066a5  No.16621657


I work 40 hour weeks and have a dog so dev time is a bit limited but I try to squeeze in 4 hours a day at least. Even days I don't feel like doing anything I try to at least do something on the game, scribble down some design, polish a map or fix a bug. Just something to keep me going forward. Weekends are 12-16 hours of working though.

36b5a7  No.16622264


Here's how my projects go

>1. Work on project idea

<2. Fuck up entirely

<3. Try to fix it

<4. Fail

<5. Abandon it because of demotivation.

>6. Start from scratch out of the blue and everything works properly.

7d32cc  No.16622403


>Start on a project

>It's going great.

>I'm making progress

>Leave the project for a year for no reason

<Oh yeah I was working on this

>Start working again.

Just recently I nuked a project from August of 2018 and I'm starting over.

cf662c  No.16622846

I started mucking around with some of the rom-level stuff in Super Metroid again. This time, with ASM code. Since I'm randomizing the rom, I wanted to also make a few tweaks that made sense. For example, rather than enforcing the gravity suit coming after varia suit, sometimes it might be interesting to play a game where you get freedom of movement beforehand, but still have to avoid heat rooms.

I made it so that gravity suit is independent of varia so that if you only have gravity, you'll still take heat damage.

17f3f9  No.16623835

File: 0c13a491fdc465c⋯.png (176.97 KB, 438x393, 146:131, 957da7df8e7c114894ea61b7bc….png)

>refreshing 8chan for days making no progress

>nothing interesting in 8chan so close the browser out of boredom

>immediately make progress

>mfw I realize I'm unironically addicted to 8chan

dc8934  No.16624072

File: 59d20d53fb66ee5⋯.png (1.01 KB, 201x60, 67:20, 8ch-ips.png)


I realized the same thing when I wrote something to track only the threads that are interesting and ignore everything else. Turned out there were only very few things. Writing such a thing would probably be more procrastination for you at this point, so I recommend something like temporarily fucking with your host file or something. Pic related are 8chan's IPs (there is some dumb spamfilter thing that prevents posting them directly), set them to and you won't be able to access the site anymore.

901711  No.16624562

Is it better to start learning coding by making stuff in visual studio?

I'm asking because there are shortcuts in Unity, like for instance Pong, you can just use Unity's Bounce physics to make the ball bounce without having to code anything.

24cf5d  No.16624610


The main lessons you need to learn are "how to make my brain create the routine i need" and "how to run the code on my head well enough to know where i fucked up".

Copypasta will only take you so far.

f066a5  No.16624631



Adding to what this anon said, most of the premade Unity stuff is absolute trash such as their character controller. These components are also the source of people hating Unity because all games that use them feel exactly the same. Avoid at all cost.

cf662c  No.16624854


If you have to reimplement everything yourself, then what's the point of using Unity? Sort of a rhetorical question, but what advantages does using it even offer?

f066a5  No.16624889

File: 3db7182e75207aa⋯.jpg (26.99 KB, 184x184, 1:1, 1422461296814.jpg)


I'm not talking about the entirety of Unity, just the components that collect a bunch of functionality like the character controller. The rest of the engine and tool chain is fine.

7d32cc  No.16625080


If you want to learn how to write a physics and rendering engine then yes, using visual studio would be better than Unity because Unity comes with a physics and rendering engine.

If you really want to really learn a programming language inside and out you start from the ground up. For example if you want to learn C or C++ you use a text editor like Vim, Emacs, or Notepad++, compile from the terminal, write your makefiles/cmake files from scratch, and so on. Once you have a good basis on how all of that works, you can decide if you want to use Visual Studio to work on the larger projects. Writing the code using a text editor rather than Visual Studio will help you understand what Visual Studio is doing and help you with debugging later on.

However most people here want to just like, make game, so learning a programming language in the context of an engine rather than in a text editor first makes sense.

100c40  No.16625190

File: 8b16e04872a410b⋯.jpg (131.91 KB, 960x778, 480:389, 8b16e04872a410b8bc6b408702….jpg)

Taking the plunge on game developing with a good friend.

Have no experience in coding, going to be getting UE4 and there user tutorials. As someone who has no coding experience, and will be tasked with writing AI and, well, virtually everything that isn't sound design and modeling, how fucked am I and where should I start.

17f3f9  No.16625197


>game developing with a good friend

>tasked with writing AI

>and virtually everything

<no coding experience

Just stop.

f066a5  No.16625207


Read the thread then start over making Pong or whatever.

7d32cc  No.16625328


Don't start with UE4. Start with smaller projects in Unity or Godot. You aren't going to make your dream game as your first game, or your second game, or your third game. Make some Atari clones first to understand your tools then go ahead with something a bit more complex. You'll learn as you go along what exactly your getting yourself into, because I think at the moment you don't really understand that. You want to challenge yourself in order to improve, but you also need to understand the challenge before jumping in.

If you want to make a game in UE4 you'll want to learn C++ first and then later go into UE4. UE4 is a great engine but a terrible C++ teaching tool.

6d5a72  No.16625379

File: 56eba8f9f118ce0⋯.png (109.86 KB, 250x300, 5:6, naof.png)


Remember suffering makes you stronger.

feaa5e  No.16625407



is struggle that makes you stronger, suffering only makes you weaker, it makes you doubt your true self, things you could esily do can't get done now.

6d5a72  No.16625570


You know what I meant.

901711  No.16625822

File: 60ad27ca1bc4fbd⋯.jpg (78.72 KB, 884x903, 884:903, 1560876859789.jpg)


>However most people here want to just like, make game, so learning a programming language in the context of an engine rather than in a text editor first makes sense.

Yeah, I'm more interested in making games than programming on its own. Not saying that I hate it or anything, I actually think I would enjoy it once I learn properly. I made the mistake of trying to make a game that was too complex for my skill level when I last tried, and gave up.


>Make some Atari clones first to understand your tools then go ahead with something a bit more complex.

After making Atari clones, would it be feasible to make a 2D platformer? Something like Mario or Mega Man?

6d5a72  No.16625932


Go with RPG maker or something like that.

7d32cc  No.16625949


>After making Atari clones, would it be feasible to make a 2D platformer? Something like Mario or Mega Man?

Yeah, just go in knowing that a platformer requires a lot of fine tuning to get the movement just right. Do some research into the Super Mario Brother's movement and you'll see things like how Mario's downward acceleration increases when he's moving down. There's also a degree of forgiveness in some platformers like the recent Rayman games where you can press the button a fraction of a second after you walk off the edge of a platform and still jump.

6d5a72  No.16625977





7d32cc  No.16626022


It's not really auto-ledge as it is ledge-assist, like console FPS aim assist. Auto-Ledge would be like in Ocarina of Time.

6d5a72  No.16626068

af7dc2  No.16626784


I need a general idea on how to do something. I need some advice on how to get a 2D field of view. More specifically I have two nodes, one that looks and one is a target. The one that looks needs to know when a node moves in front of it. There are these elements to this:

< A register of nodes in a space

< Locations of both nodes

< Rotation of the looking node

< The angle of the FOV

I know how to setup the FOV (already done), but how do I determine if a node is in the angle of the FOV of the looking node? Will post current ideas when I get home. I know WHAT to do, but not the math functions I need to do it. Anyone can answer, not engine specific.

f066a5  No.16627310


Not sure if it will be helpful but we have fov in our game that somewhat works like what you specify (I didn't write it but I think so). You can find the source at https://github.com/jaoel/LD44, it's in scripts/map/fogofwar.

0e0b10  No.16628452


Extract the clipping planes from the matrix and then use them to find out what is inside of the frustum.


In two dimensions this is easier. If you have an angle just use polar coordinates to find the normal of each clipping line. Add pi/2 rads to each angle for the start and end clip.

cf662c  No.16629807

0ae1de  No.16635870

Is gdevelop worth trying out?

[Return][Go to top][Catalog][Nerve Center][Cancer][Post a Reply]
Delete Post [ ]
[ / / / / / / / / / / / / / ] [ dir / bebjush / choroy / dempart / freeb / randamu / vg / vichan / y2k ]