e73ff5 No.14617171
Demo Day scheduled for May 5th - less than 4 weeks away!
"Devving each day keeps the depression away" edition
Resources
>>>/agdg/
>>>/vm/
>#8/agdg/ via irc.rizon.net
Links
>Wiki: http://8agdg.wikidot.com/
>Beginner's guide: >>>/agdg/29080
>Previous thread: >>14595695
>Sister thread: >>>/vg/27335
QUARTERLY DEMO DAY SCHEDULED FOR MAY 5TH
Please contribute to the wiki if you can. Even if it's your own game, every little bit helps!
0b5226 No.14617184
AGDG? More like ADHD
HAHAHAHAHAHAH!!!
e73ff5 No.14617188
>>14617184
That was uncalled for
a7d62d No.14617199
POST PROGRESS
I don't have any
ea0909 No.14617208
I came here for the Vidfeo Games
ad75cf No.14617253
>>14617199
Starting to make menus for a shop, and I hate making UI.
5364af No.14617258
>I've been programming in a fictional language for the past N days
How do I rehabilitate myself into using C or something again so I can make actual progress? Every other language is just so much worse that my morale dies as soon as I try to do it.
e1a941 No.14617285
>>14617258
Just do it, nigger.
8ee4e6 No.14617308
>>14617258
At least you haven't been busy writing lore texts for days instead of finishing models.
ad75cf No.14617320
>>14617258
The language that you're using for your game isn't what's holding you back. People can make games using assembly or lisp, and even python. I'm using Godot right now, and even though gdscript is fucking garbage I can still make progress just fine. As long as you know what you want to work on in your game, you should be able to accomplish it regardless of the laguage your using.
e73ff5 No.14617325
>>14617320
I think he's one of those autistic people that get frustrated that a toolset works in a particular way, so he abstracts it and makes his own the way he wants, and each time steps back.
I'm the same way
9b0130 No.14617329
How do I make music? I can make up tunes in y head and hum or whistle them, but when I try to recreate them they sound nothing like what I'm thinking.
1f2028 No.14617342
>>14617199
all i managed to do during easter holiday was bugfix and do unimportant homework
i finally got those pesky interfaces to properly serialize once i realized that private variables actually don't get saved just when you think you know how shit works
GOD DAMN DOG AT HOME HAD FLEAS, NOW I CAN'T STOP SCRATCHING MY LEG
>>14617325
yea, implementing features that you don't really need tends to just waste time
example being the shit i'm working on
mount and blade uses fucking text files for dialogue and works just fine, with the minor downside that the dialogue is somewhat generic as a result of this
e73ff5 No.14617351
>>14617342
I have a decent way to plot graphs and nodes, but I don't know how I'd add scripting to a dialog tree, eg using text files and checking if someVal = 10
Is this where I'd inject some Lua code or my state machine with opcodes?
0b5226 No.14617352
YouTube embed. Click thumbnail to play.
>>14617329
Get a physical, digital USB piano, practice practice practice. If you want to hear some average guy talk about music and some of the philosophy behind it, then go to the 1 hour mark, where you can see him arrange shit live.
1f2028 No.14617359
>>14617351
as i understand it, butterlord writes all the code/dialogue/items/locations/troops/pretty much everything in text files and then compiles it via python
so technically it doesn't really read text files at runtime
5364af No.14617370
>>14617320
Obviously, but that doesn't mean it's as fun to use and provides the same capabilities without making me write twice the code.
The language is meant to be as low level as C and generally work in a similar manner, but structured in a way that makes it much easier to write things and doesn't look like vomit. For example type inferrence, better native string manipulation, special array methods and dynamic arrays, structures can have default values and/or initialized however you want in a single line, cleaned up and more consistent syntax, special loops that reduce the need for nesting and weird tricks, and so on.
>>14617325
>I think he's one of those autistic people that get frustrated that a toolset works in a particular way
Yes. I'm the kind of autists who thinks everything is shit in one way or several, although I think C is shit in a lot of ways.
378496 No.14617671
To the anon who was talking about lucid dreaming and so on in the last thread, do you have any tips on handling the depersonalization? I have a pretty similar set of symptoms to yours, they just work a bit differently. The depersonalization is one of the biggest blocks between me and more efficient gamedev.
0b5226 No.14617685
ad75cf No.14617728
>>14617671
>The depersonalization is one of the biggest blocks between me and more efficient gamedev.
Do you mean to say that experiencing symptoms of depersonalization in your waking life is fucking up your gamedev work? If that's the case then you need to see a doctor or takes steps to improve your sleeping habits if LD'ing is negatively affecting you.
f44343 No.14617737
i've got a job last week and my game development stopped to a crawl, but i'm not gonna stop of course, i wish at least one of the 2nd exodus is here to see my game finished someday, although i'm considering fully migrating to another chan
>>14617184
how will we ever recover?
73df61 No.14617897
>>14617671
>avoid lucid dreaming
>masturbate less
>sleep when you feel tired
>eat more veggies
>exercise more
>go see a doctor
e73ff5 No.14617984
Oh hey cool, I learned a thing
So SFML has a handy TextEntered event that fires when the focused window receives a text input event. Normally, if you have a focused field that can accept text, you'd handle it by pushing some value from the event into the field.
Anyways, just to make sure it worked, I was mashing some random keys and I decided to try Alt+letters. Alt+A is ascii 0x01, Alt+B is ascii 0x02, etc. I was aware of alt+numpad to input any value, but I didn't think the low range of ascii was bound to alt in general. Pretty neat
63e76f No.14618005
>>14617897
>>masturbate less
Is that seriously connected to mental issues like depersonalization? Honest question.
e73ff5 No.14618021
>>14618005
When you climax, your brain releases dopamine, eg the chemical responsible for happiness and feeling good. The sleepy part afterwards is from seratonin/tryptophan. Jerking off too much can cause prolonged chemical imbalances
e33205 No.14618022
>>14618005
I dont know, but if you are honest with yourself: Are happy with the amount of time you currently spend wanking?
You dont have to turn it into your new religion and preach to everyone about the glory of nofap, but you dont really lose anything by not wanking and you may free up some time to dev instead.
0b5226 No.14618052
>>14618021
This is very true.
EVERYBODY PAY ATTENTION TO THIS POST!
It can cause ADHD-like symptoms, like mood swings, fuzziness, forgetfulness among other things.
e73ff5 No.14618071
So with C#, how do I tell SFML to use a lib/ folder when executing the game? Specifically, I can reference the lib folder fine, but SFML seems to look at the .exe's directory and ignores the root path the dll itself is in.
If everything is in the game's root path, it works, but I hate the clutter of like 30 dlls beside the exe
bba7b7 No.14618130
>>14616135
I never actually saw him post actual progress, so I don't know.
is this him?
https://brnydev.itch.io/brnygame
is this rustfag's "game"?
17857d No.14618141
>>14617671
Combating depersonalization is difficult, as there's a lot of factors involved, but with the right tools at your disposal it's rather easy to combat; as without knowing the source of the contributing factors you're in the dark without a flashlight swinging at who knows what.
The core issue here with depersonalization is that it's an unfamiliar feeling to most. Which means one cannot recognize contributing factors (as they're such small degrees of intensity), and lacking the ability to recognize the contributing factors leads to a culmination into something that's psychologically inhibitory.
In essence, the contributing factors are such small degrees of intensity for that "depersonalization feeling" that it's easy to be consciously unaware, and quite normal to be unaware of.
So, an analogy is the "feeling" of falling, which is learned to be consciously recognized through repeated exposure at varying degrees of intensity, and has a rather low-threshold for consciously recognizing that feeling as "falling" (which is an instance of a psychological phenomena that's recognized as a "feeling").
While it is of a higher-threshold, depersonalization has the same exact ability to be learned, and to be consciously recognize; as to be conscious aware of the "feeling" of depersonalization (at the varying degrees of intensity).
Though, a note here, getting the right tools to really combat it is not a path for cowardly. It is a matter of consciously choosing to expose oneself to deeper levels of depersonalization, and to learn to consciously recognize the feelings of depersonalization through exposing yourself to the "epitome" of depersonalization; e.g. ego death, or gnosis.
>How?
Some may frown upon this, as it's empirically evident that "higher powers" as it were have been building a vilifying narrative in the west, but the more intense psychotropic plants used in a therapeutic manner can help immensely with the more elusive psychological ailments.
For combating depersonalization I'd recommend salvia (salvia divinorum is the technical name), and the reason for this is it's notorious for causing an ego-death trip; in addition to the trip only lasting 15 minutes or less with resoundingly therapeutic results.
Also, side note, it is legal in some US states.
This anon is on the right track thou >>14617897 for more mundane efforts.
It should be noted that lucid dreaming is definitely a psychological tool, and can be misused/used irresponsibly; as with anything that involves rewiring the - at this point - relatively unknown realm of the mind.
4f53b4 No.14618150
>>14617342
Goofy question. I see you post alot, but your combat does not look fun.
What games are you taking inspiration from for combat? In your current iteration it looks like the rts genre.
13c436 No.14618170
What the shit!?
UE4's player controller's physics calculations seem to fall apart at low framerates. I noticed that the player takes significantly more fall damage when the framerate goes down. Since it's physics based, you'd expect some variance due to how physics engines work. It also wouldn't be surprising, if the calculations go wrong when the framerates get really low. Like if the PC completely chokes on the game.
However, that isn't what's happening. As you go from 60 to 30 FPS, there is a minor decrease in falldamage. It's within margin of error, so nothing worth worrying about. Once you go from 30 to 29 FPS, there is a notable spike in fall damage. After that it completely explodes.
Why can't things just work?
882b3b No.14618180
b4a361 No.14618183
>>14618130
I don't recognize any assets in that game that he showed off here.
b52b28 No.14618191
Can somebody give me a hand with pointers?
I'm trying to create a singleton, I have a private static pointer declared in my header
class Display {
public:
static void createDisplay();
static void remakeDisplay();
private:
Display();
~Display();
Display(Display const& copy);
Display& operator=(Display const& copy);
static Display* p_display;
};
and I'm trying to define it in my cpp file
void Display::createDisplay()
{
static Display display;
Display::p_display = & display;
}
I get an error "undefined reference to Display::p_display"
I am new to C++, should the Display class be put inside the cpp file, while I keep the static methods inside the header?
17857d No.14618198
>>14618130
i look at the attached twitter and the games UIs are different, and the twitter is too old to be a new alt by him.
he did use UE4, the genre looks the same, and their profile images look like they're from the same doujin but that's about it.
Also, he didn't get shit done, so it can't be rustfaggo.
66d6c6 No.14618201
>>14618170
everything falls apart at low framerates, anon.
>>14618130
No, although it's pretty funny that his Patreon uses a Hirame profile pic, those used to be my go-to thing for avatars.
e73ff5 No.14618210
>>14618191
As someone with no practical C++ knowledge, I thought header files were just a collection of method and class signatures to help out the compiler
882b3b No.14618216
>>14618191
p_display is a member not a type; also you need to remove the space between the ampersand if your are attempting to dereference something.
fcce75 No.14618220
>>14617329
Definetly learn to play an instrument along with taking some basic music theory lessons. I recommend learning to play piano (as >>14617352 says, get a physical digital USB Piano to practice) as is one of the most versatile and useful for learning music theory. I strongly recommend you to get a teacher for that if you can afford it (If you can't just be constant and put into practice everything you learn about music theory so you can assimilate the concepts better)
e8b1b8 No.14618223
>>14618191
Somewhere in a cpp file, you have to actually declare Display *Display::p_display;
bba7b7 No.14618229
>>14618198
>he did use UE4
>the genre looks the same
>their profile images look like they're from the same doujin
>that's about it.
honestly, only this
>he didn't get shit done
convinces me it's not him
the resemblence is uncanny
882b3b No.14618230
>>14618191
>>14618216
Here's what you were thinking:
class Display {
/*** what you essentialy were trying ***/
typedef ::Display p_display;
public:
static void createDisplay();
static void remakeDisplay();
private:
Display();
~Display();
Display(Display const& copy);
Display& operator=(Display const& copy);
static Display* p_display;
};
Save inclass typedefs for template insanity. :D
b4a361 No.14618243
>>14618198
I'm pretty sure that twitter profile image is the same one he uses for disagreement.
b52b28 No.14618247
>>14618223
It's defining the address the pointer will point to?
>>14618230
Haven't seen a typedef before. I assume it's shared between the header and cpp then?
13c436 No.14618248
>>14618201
>everything falls apart at low framerates, anon.
29 isn't low though. At least not low enough for something like that to occur. I'd expect that at <15. The problem is that 25-30 is something people who game with minimal system requirements might experience in certain places. Basically, I want to ensure that my game works at 20 FPS. Just to have enough of a margin of error so that people with shit hardware can still have a great time. Everything else would be irresponsible.
e73ff5 No.14618269
>>14618248
Floats are numerically unstable, and low framerates or extremely low/high numbers fall apart. Most gamedev projects use floats over doubles anyways.
Why is your damage tied to framerate? I assume you're adding a velocity per frame, the velocity stacks up and gets fucky with low framerate, then does impact damage?
e73ff5 No.14618271
>>14618247
Think of the header as a contract for what the class will do, and the cpp file is the implementation/fulfillment of that contract. That's accurate, right?
b52b28 No.14618280
>>14618271
I've seen it described like that. But classes in headers are making me scratch my head.
66d6c6 No.14618282
>>14618248
No, 29 isn't low enough to cause issues. I don't know what's going wrong in your case, but considering that UE4's vanilla stuff is generally pretty solid, I'd start looking through your own code, first. Make sure you're not forgetting to make shit frame-independent where applicable, and make sure you're not multiplying by delta second twice. That happened to me before. Otherwise I don't know why shit would go wrong at 29FPS. The safest bet is to write your own code, to be honest - I tried to make vanilla CMC work for me, but I ended up writing everything from scratch. It took a while but it works, and feels better than CMC.
>>14618198
>>14618243
I don't use anime profile pics anymore, mainly because these days they make you look like a faggot.
inb4 "but you're a faggot"
e73ff5 No.14618289
>>14618282
>anime pics
Remember anon that made the thread a week or two ago with one and deleted it after?
66d6c6 No.14618294
>>14618289
must've missed it. Did he get bullied?
5b79f3 No.14618295
YouTube embed. Click thumbnail to play.
RPGdev,
I found something that might be valuable to anons dev'ing in Godot 3: a link to an answered question regarding how to draw values in-game, on screen for debugging or as babby's first HUD tutorial:
https://godotengine.org/qa/11780/show-numbers
Also included is a tutorial video by Gamefromscratch explaining node heirarchy. in case anyone needs it. Not gonna hooktube or mp4.
>>14617184
Oh boy, image macros. I don't know what I'll do with myself now. Get hidden, Snake.
13c436 No.14618316
>>14618269
I know about floating point rounding issues and this isn't the issue here. Look at the graph again. Look at the values. Those issues occur when you're dealing with either really small or really big numbers. Having 1400.0 fluctuate between 1350 and 1450 isn't a floating point problem.
>Why is your damage tied to framerate?
That's what I tried to find out. What I do is take the vertical component of the velocity vector at the time of touching the ground and put that into a function that evaluates a graph that maps velocity to damage taken.
All the values come from epic's character controller and I also use their OnLanded() event. I literally just take their value, negate it and put it through an evaluation function.
>>14618282
See above. The problem occurs outside of my code.
Given how stable it misbehaves, I think I'll just use a correction function that fixes the damage for framerates between 20 and 30. I don't care what happens below 20.
>>14618294
To the point where he deleted the thread. Anyone who gets upset at anime girls being posted on this website should just fucking kill himself.
66d6c6 No.14618320
>>14618316
Eh, probably some mathy fuckery going on and messing you up. Numbers can be weird. My advice is to let it rest for a while and focus on something else. Leave a //TODO on there and get back to it later, you might have an epiphany when you do.
13c436 No.14618325
>>14618320
But it already is on my todo list and the next thing will be INFINITELY more of a pain in the ass to deal with.
66d6c6 No.14618333
>>14618325
Do something fun to distract yourself. It's a good thing to be hard-working but if you give yourself a burnout the downtime will be much greater.
13c436 No.14618336
>>14618333
>01:19
Well, sleep sounds nice.
66d6c6 No.14618340
>>14618336
Fellow eurofag. Enjoy your sleep - tomorrow things will be clearer.
17857d No.14618345
>>14618170
From what I can see UE uses a variable time-step for physics; which could be why you're seeing issues here (dropped frames, lower than usual timestep so more interpolation, etc).
I'd look into the physics timestep, and if you're using the substepping feature; I also saw a forum thread for implementing a fixed timestep.
17857d No.14618354
17857d No.14618358
>>14618354
also the UE docs suck necrotic dick
sage for bantz
141999 No.14618460
HookTube embed. Click on thumbnail to play.
>>14615651
Yes, they are being rendered on both sides, I have back face culling turned off right now. Embed related is the full song.
66d6c6 No.14618469
>>14618358
Yes, and it makes me fucking mad that most tutorials are trying to teach you how to use (((blueprints))) instead of code.
b4a361 No.14618471
e73ff5 No.14618489
bec214 No.14618565
>>14618489
There really should be an easier way to do this
141999 No.14618573
9f6833 No.14618840
>>14617320
i want to start making a Digimon World 4 clone using godot but i just don't get it, also i have a year of College programming, mainly C/C++ (i'm doing C# and web programming this year)
17857d No.14618861
>>14618840
>but i just don't get it
articulate exactly, in detail, what you mean; what do u not get, and why.
this has a purpose, just do it, be as verbose as u want
9f6833 No.14618909
>>14618861
I only ever used Unity barely and it sucked hard, and it discouraged how to work in an engine, i cannot get used to working in a 2d or 3d graphic space or working with 3d models, i only have limited expierence with graphics, i only know learned how to code properly. I think my main problem is learning the interaction between all the working parts of an engine
TL; DR : Unity was shit and didn't help me to learn properly back when i tried and know that i know C/C++ i feel Discouraged
Fuck Unity
9f6833 No.14618914
9b0130 No.14618930
9f6833 No.14618954
>>14618930
i went trough some of the 2.x tutorials, but i had finals around two weeks after and i forgot a lot.
i think the problem is me
maybe i just have to start with a fresh mind. ty anon
e73ff5 No.14618966
Since this is sprite based and has absolutely no 3D to it, do you think it would look okay if I faked reflection, by drawing reflective objects to a separate layer, and doing a mask effect and making them look brighter the closer they are to the player?
e73ff5 No.14618969
>>14618966
Ah, this is probably something that I'll have to actually implement to see how it looks in motion
17857d No.14619005
>>14618954
>>14618909
Learn how either of those engines work, and go in with the mindset that you're learning; making your dream game is something you need to earn.
The first months are basically just fucking with the engine, coding prototypes, getting familiar with how an engine works, and realizing that you will always be learning. Extra stuff like 3D modeling, animation, texturing, shaders, and all that good stuff comes later.
So, get familiar with the design pattern the engine devs build the engine around, the execution ordering for loops, scour the docs when you have questions/issues/etc, and read/watch a lot of tuts/info dumps on the current topic you're getting to know like the back of your hand.
The same applies to godot, unity, UE, but YMMV bcs different engines have varying qualities for docs/tuts/info dump resources online.
>Unity was shit and didn't help me to learn properly
Unity has arguably the best docs, tutorials, and info dumps online; with info ranging from the common to extremely obtuse issues as it's so widely used.
I don't think unity is the issue here… as I found an absurd amount of information online to work through when I started years ago.
d9bd6d No.14619048
>>14617199
Started creating the game's town, and my inspiration for the city is this
Can't promise it will look as incredible as these refs, but i will try
My brother is making a custom character controller so the protag will have more freedom of movement in a map with more verticality
80bd45 No.14619218
>>14617199
I'm working on an altered design of the APS underwater rifle for animu stalker. Essentially making 'Fantasy' versions of weapons to avoid any issues with real life weapons manufacturers. As well as flex some creative muscle to bring these real world weapons in-line with the game's aesthetic (Which I have written down as "Comfy Innawoods Anime"). The idea is to exaggerate some of the features a little and not be as autistic about detail similar to classic anime but still allowing for some detail like Valkyria Chronicles.
I'll work on the final design probably tomorrow because I spent all day flying. Right now I just want to be in this nice bed after spending 14 hours sitting in airport and passenger jet seats. Seats seem deliberately designed to be uncomfortable and give you back pain.
>>14617329
>>14617352
Additionally, Chris O'Neil's video on music theory is a pretty good baby's first music lesson. https://www.youtube.com/watch?v=eBsyEpZjjyw
73df61 No.14619542
>>14618954
of all the digimon games why remake the shittiest one?
e33205 No.14620061
how do you guys handle comments, that are basically just "ids shit :DDD" in an environment that you control like your youtube channel?
Im generally opposed to deleting stuff and banning people, but if they dont offer anything constructive what should i do?
1f2028 No.14620066
>>14620061
leave them be?
i have highschool friends telling me i'm gay on 5 year old videos
eba6ab No.14620070
>>14620061
>youtube
>handling comments
These don't belong together.
Unless you're posting in a your own blog or somewhere you expect real discussion, don't try to police the comments.
b52b28 No.14620093
>2 AM
>Almost got window to load in c++ translating my old paradigm
>learned a whole lotta cpp shit
feelsgoodman.jpg
Good night lads, don't stop working on your projects.
74e098 No.14620108
>>14620061
If you can't handle shit being thrown your way, this isn't the hobby for you. But I'll tell you now, trying to silence criticism is gonna do you more harm in the long run than you can ever know.
13c436 No.14620559
>>14618354
Thanks, that seems to be it.
>>14618469
It's super easy to translate Blueprints into C++ though. If you right click on a BP function node, it even gives you a shortcut to open the C++ definition of it.
397fbf No.14620578
>>14620061
Just turn off likes, dislikes, and comments like every other pussy.
7eee83 No.14620580
>>14620061
>in an environment that you control like your youtube channel?
If you feel the need to police your own customers, then you've already made a major mistake: You actually believe the community's main reason to exist is to advertise and positively represent your game, rather than just a place for people to discuss your game and socialize.
If you actually go forward with that idea, you'll end up creating a worthless community of yes-men that will act as if every change you make is a gift of god. This is even less constructive than >ids shit :DDD, and will harm you even more in the long run.
If you want good PR in the comments of your Youtube channel, then you can just respond to the people who complain. Many shitposters actually turn around and give you actual constructive criticism once they realize they can get a better product out of it. Not only that, but Youtube highlights comments from the channel account, showing everyone who scrolls down to read comments that yes, you are actually responding to criticism!
If you're too lazy to do that then just do what >>14620578 says since nobody really cares what Youtube comments say.
091cf8 No.14620775
>doing external code gen
>generator takes few seconds
>unity takes few seconds to process
>VS takes 3-5 minutes to process external changes
>strip VS to bare bones to remove all bloat possible
>super fast booting now, but code gen… still minutes to process external changes
>try the new unity compatible rider IDE
>setup takes 45 minutes to get same as what I'm used to
>takes a few seconds to process external changes from code generator
fuck, VS is bloated garbage, I just saved myself so much time beyond what the code generator saves me.
a7d62d No.14620782
>>14620061
Ask them why they think it's shit. Be polite and try to get them to engage further.
First "proper" level coming along nicely.
1f2028 No.14620809
>>14620782
wait, that's not godot
how the fuck did you change the folder color
dc1754 No.14620812
>>14620775
I see this pic all the time, is it from a VN?
a7d62d No.14620813
>>14620809
Rainbow folders, got it for a buck on the asset store.
1f2028 No.14620814
>>14620775
>VS is bloated garbage
i've been using notepad++ instead of it for a while now, productivity has skyrocketed ever since i stopped having to wait 3-5 minutes for VS to load all of it's bloat
1f2028 No.14620820
>>14620813
>paying for faggotry
5364af No.14620822
a7d62d No.14620828
>>14620820
I think it's pretty good for organizing your folder structure further.
dc1754 No.14620839
80bd45 No.14620869
>>14620061
You just ignore it. Especially since almost every comment system will display the most upboated :D comments near the top. The comments tend to curate themselves over time. Comment sections, especially on jewtube, are a perfect example of order stemming from chaos.
000000 No.14620906
>try to ADHD do breakout tutorial on unity
>simple script to add score when bricks gone
>doesn't work and no error messages
>fuck it just copy/paste code from tutorial
>still nothing working and no errors
>which demoralizes the most not even a hint on whats wrong with that shit
fucking fuck
maybe I should Godot then
1f2028 No.14620913
>>14620906
maybe you just didn't do it right
913708 No.14620917
>>14620906
It's been awhile since I've touched the Unity official tutorials. They're old extremely outdated and I found myself having to google anything related to particles and UI. It's a fucking mess. Usually the thread for the tutorial on the unity boards have updated code by members because the mods are useless too.
cb4285 No.14620923
Do we have any deposits of game design/dev books?
Right now I want some theory or ideas of "best practices" for level design. The wiki is a bit bare http://8agdg.wikidot.com/wiki:level-design
Do we have a volafile or riot room or torrent or mega?
Or does anyone have a recommendation that I can track down?
>>14620906
>still nothing working and no errors
Did you try something more basic, like a print? Maybe it's your setup.
636f2a No.14620937
000000 No.14620973
>>14620937
I guess i'll start with this then >>14618295
e33205 No.14620997
>>14620580
I actually responded to someone that said it looks like i did not put in a lot of effort and got them to elaborate a bit.
Im talking about people that dont even write a complete sentence.
Again i dont want to heavily police comments. I want to moderate as little as possible, but i dont see any value in responding to people that obviously dont give a shit about the game, when there are people that will at least write complete sentences and tell me why they dont like it.
a7d62d No.14621000
>>14620906
>>try to ADHD do breakout tutorial on unity
Never gonna make it.
80bd45 No.14621012
>>14620906
As pointed out here >>14620917 they're all extremely outdated (Some of them are for Unity 4 for fucks sake) which is fucking laughable because you'd think with all the money they make off of having a meme engine they could afford to update their tutorials to work with the latest version of the engine.
It's why I jumped ship from Unity early on and never looked back. It's impossible to learn the engine if tutorials that are barely a year old, sometimes not even that, are broken because they keep fundamentally changing how the engine works. Go to Godot, it's the superior C# engine and is rapidly catching up to Unity.
086af8 No.14621017
>>14620782
What kinda game are you making, anon?
a7d62d No.14621039
>>14621017
Educational game for kids somewhere between ages 5-8.
70a969 No.14621165
>>14620997
You've got two options, ignore it, or blow them out of the water with your bantz.
If you can't bant, ignore them.
I only ever really dealt with one shitposter who I'm convinced was a microsoft employee and they didn't know how to bant so one well worded response shut them down.
Their youtube channel was full of videos to brag up microsoft products and call of duty gameplay.
d9bd6d No.14621184
>>14620061
shitpost back or ignore, let people have fun
ad75cf No.14621259
>getting weird behavior in Godot where the ready() function in an instanced scene is being triggered twice
>check the github issues page and it's a known bug, but they've apparently fixed it
>I check the commit and they only fixed it in 3.0 for whatever reason
I should be able to apply the fix myself since godot is open source right? The fix they applied is only like 3 lines so I doubt I could break much.
7a3eb2 No.14621272
How do I take .fbx file with animation clips in Unity and make those animation clips separate from the .fbx file? When I try to search this shit, I find advice to just select the clip and press Ctrl+D but it doesn't work. Did something in Unity change since those advice were made or am I doing something wrong?
1f2028 No.14621276
>>14621272
selecting the clip and ctrl+D used to work, haven't tried in ages
you could always just split the fbx outside of unity
otherwise i don't think there's any other option
why do you even need to use it outside of the fbx?
7a3eb2 No.14621285
>>14621276
I want to edit some properties of the clip inside the Unity's animation editor but I can't do it as long as it's connected to the .fbx file.
7a3eb2 No.14621316
>>14621272
Ok, I figured it out. The problem was that the animation clip's name contained the " | " sign which is illegal sign for files in Windows, therefore it couldn't be created. Renaming the clip fixed the problem.
5e524c No.14621491
>>14619542
not a remake, inspired, like a Ys game but with loot
091cf8 No.14621538
>>14620906
>have an issue you shouldn't have
>have to figure it out
welcome to gamedev
It sounds like you're doing forgetting to do something that's really simple, and could be easily missed by someone new; who doesn't quite understand how to debug yet (i.e. eliminate potential causes w/deductive methods until you reach the actual cause; which the steps to doing so is generally drawn from experience… so git gud).
ad75cf No.14621604
>>14621259
So I ended up writing a small hack to sidestep the problem; turns out actually fixing the problems was going to be more than the three line fix I thought it was going to be. Basically I just added an if statement that checks if the instanced scene has an arbitrary child node before get_node() is used. If not get_node() will also trigger in the scene it's being instanced in and throw an error because an instanced node can't find it's children or something.
091cf8 No.14621660
>>14620814
i use to use notepad++ for shader auto-complete as VS had crap support. thou luckily rider has that OOTB, which is an added bonus to convenience.
Though i'll probably be using the new shader graph feature that unity rolled out in the latest beta for general stuff from now on; as I can just go in and optimize/adjust it as needed for the final touches.
>having to wait 3-5 minutes for VS to load all of it's bloat
agreed, before I did the barebones stripping even booting VS took a solid 3 minutes (though that's for my main project, smaller ones are acceptably fast); though after I stripped it booting took only a few seconds (damn is VS bloated on initial install though, fug).
The issue I'm sure you've ran into is shit like external code changes (like adding a file in unity, and going back to VS), and VS being so terribad at handling that at anything more than a single file that it even freezes sometimes.
Especially so, if you don't have the option to add from within VS (implicitly external changes) while in the process externally adding/modifying a bunch of new scripts… it can take an excruciatingly long time while it locks the main thread to wait for import to finish.
>productivity has skyrocketed
quality, i hope other anons learn from this too
>>14621604
>indent based code scope
disgusting
>solving by smoke, mirrors, and hacks
ah yes, the old gamedev adage
da968b No.14621676
>>14620061
Consider their complaints, but don't respond… at all.
Even if you are being polite and embracing the criticism, it will give you an image of incompetence and the audience will still find some way of arguing with you. Just take their advice into consideration and don't say a word.
da968b No.14621696
>>14618909
It's not good to start off using Unity, actually.
Here's the proper order to become a competent game developer
>learn the C++ basics because the memes and books told you to
>make a console-application game that sucks huge balls
>try to make a game engine, but fail hard
>switch to Unity
>try to make a Unity project, but fail hard
>go back to trying to make your own engine, fail hard
>use Unity again, actually make something playable
>try to make your own engine again, make something okay
There you go, you should be familiarized with basic game development concepts now, and some slightly advanced structures you'll see in general programming and in games specifically.
http://gameprogrammingpatterns.com/
Also this is a pretty good resource.
If you're having trouble trying to understand how parts of the engine or even the game itself interact, then check out the pages about "Singletons" and "Service Locators"
091cf8 No.14621700
>>14621676
Yeah places like the YT comment section are a terrible platform for conversations, and it's best to completely avoid drama in such a place; further yet you can direct them to a suiting place to discuss that in your comment section or at the end of the vid (like, "discussion, criticism, and so on is welcome at X place").
I'd say to just focus on what's important, and honestly a YT comment section doesn't even make my list for putting any significant effort into (thus, I'd completely avoid potential drama, as there is absolutely no pragmatic reason to engage in drama).
As there's a time and place much more suiting for discussing that, which is more likely to result in constructive criticism that isn't completely worthless, and would be better for all parties involved.
091cf8 No.14621706
>>14621696
goddamn anon, this is spooky, that's almost to the letter for what I did when i started
8d29ab No.14621708
>>14620997
If you start censoring fucking youtube comments because you don't find them useful, you're a bigger faggot than rustfag and I look forward to seeing you fail
091cf8 No.14621717
>>14621700
>comment section
>video description
polite sage for correction
091cf8 No.14621751
>>14621700
it should be noted that you should still give ppl the freedom to comment/etc though.
thought i'd make that clear, as avoidance of drama for yt stuff doesn't mean the total avoidance of the possibility of drama (e.g. disabling comments).
as the community aspect is important, so keep the comments/etc, but the drama isn't worth putting any time into for a place like a yt comment section; is what i'm saying.
so yeah, don't bite the bait, focus on more important shit, and just like make game
20e789 No.14621752
I think I finally got the base topology right for this anime head, what do you think aside from looking creepy?
091cf8 No.14621769
>>14621752
quality edge loops in the proper places, easy to subdivide and such; nice nice
>>14621764
hue, nice
20e789 No.14621778
>>14621769
Fucking finally then, I think it is the hardest part, the rest I presume can be made with make human, anyone knows if their stuff has the right edge loops?
a7d62d No.14621786
>>14621752
Looks good to me.
>>14621764
kek'd
13c436 No.14621796
I fixed the physics situation and now face an interesting optimization problem. I have three variables:
>the physics framerate
>the display framerate below which the physics calculations start crapping themselves
>the CPU load during low framerates (note: low. The physics get more demanding as your game's framerate decreases. It's the opposite of what you would expect.)
I get to choose two. The third one is thrust upon me.
>high physics framerate so that rich Anons with high refresh rate monitors have smoother physics
>physics stable all the way down to 20 fps
<completely and utterly hammer the CPU of everyone on the low end
>stable physics at low framerates
>easy on the CPUs of toasteranons
<physics capped at 40 fps
>beautiful physics for rich Anons
>easy on the CPU of the Anon playing at the low end
<low end now means you have to run at +35 FPS or the physics go to hell
What will I do?
ad75cf No.14621801
>>14621660
>indent based code scope
Yeah, gdscript is awful. Not a big fan of dynamically typed languages either. There are C# bindings but getting them to work in linux is a hassle and it's microsoft proprietary bullshit. C++ is supported, but going by this thread (https://godotengine.org/qa/21457/how-do-i-create-a-c-script-in-godot-3-0-i-keep-getting-an-error) it's also a pain in the ass.
>>14621764
TATAKAU HEEROOOOOOO
1f2028 No.14621806
>>14621796
>The physics get more demanding as your game's framerate decreases
how?
shouldn't physics be per frame, using delta time, and work all the same regardless of how poorly things are going?
713a19 No.14621833
>>14621796
I don't get it either. If your physics don't change then producing a new frame will essentially display nothing new on the screen.
Explain your model.
a7d62d No.14621837
>>14621796
Does your game have chink or BRBRBRBR appeal? Otherwise you might want to reconsider the low end spectrum? How low end are we talking? Hell, you could even have a setting that mainly alters physics framerate and resolution.
>>14621806
Often you use a fixed framerate for physics to make it more predictable.
713a19 No.14621849
>>14621837
>fixed framerate for physics
explain how it's supposed to work
Is the player not bound by physic? Otherwise you'd get the EXACT SAME IMAGE.
1f2028 No.14621855
>>14621837
yea, but you CAN'T predict how good/bad someone's frame will be
you're either gonna have to perfectly limit the framerate, or simply do what everyone else does and use delta time
a7d62d No.14621867
>>14621849
>>14621855
I never use physics for things like player movement personally, I feel it's often very clumsy and only fits certain games.
Instead of passing deltatime to your physics calculations you just use your fixed framerate, like 1/60 for example. Then you make sure it runs at the appropriate times, i.e if you have 75fps for your entire game you will be running your physics calculation less often than the game logic and rendering.
091cf8 No.14621907
YouTube embed. Click thumbnail to play.
>>14621778
they have pretty meh edge loops (concerning deformation topo that is) and too much detail for most (as you want to have low-ish poly, then subdivide dynamically as needed, as you're already doing).
they're fine as a guide shape for making your own base body topology though. so, only as the basic shape, not using their topo, so retopo it.
For retopoing based on the guide shape just use your equivalent of blender's shrinkwrap modifier (w/a face) and project the verts to the surface for retopoing it; then use a function/hotkey to extend the edges into new faces w/mirroring for quick retopo'ing. Here's a good vid (blender though) attached for retopoing quickly.
Don't forget to consider deformation topology… it's really important; there's plenty of links on the wiki on the dev resources page for that stuff.
though, quick reminder to save a non-edited copy of your proper topo base as you'll want to re-use that for different models later (or take notes from for another base, as to get same topo w/the same "extreme ranges" offered by your deform topo setup; so animations with different bases but similar rigs look the same deforming); so always edit a copy of your proper topo base as needed for sculpting and shapekeys/etc.
>>14621801
Aw, that sounds pretty frustrating tbh.
Seems the dominance hierarchy of languages is already taken hold. For something similar that happened on the unity side of things, which may happen with godot eventually, is that JS used to be the "goto/first class" language back a couple years ago; then around a year or two ago C# became the "goto/first class" language supported.
>>14621806
appears to be variable timestep fuckery. noticeably with a reliance on multiple factors which results in a variable range of semi-fixed delta time steps.
>>14621833
>>14621849
>Delta = change
>delta time = change of time, change measured in time
>delta time used in physics
>also can be based on FPS for variable delta time steps
>has a rate of change
rate of change in the physics engine is the delta timestep (be it fixed intervals, variable, or the other approaches, see: https://gafferongames.com/post/fix_your_timestep/), they're obviously still bound to a rate of change, and that anon does have a coherent model; look at the article linked here >>14618354 to get more familiar w/UE's available approaches
>>14621867
That's one of the contingent issues due to the approach they used, and a fixed timestep like 1/60 is just a fixed delta time approach (still passing delta time… as 1/60 is your fixed time interval); which has its own issues.
Look at the article linked >>14618354 relating to UE's physics timestep. It's what they're working with.
>I never use physics for things like player movement personally
Physics can be useful intermittently for movement, but I found to be unreliable for precise movement too.
ad75cf No.14621960
>>14621604
Welp, in a not-so shocking turn of events it turns out that I'm retarded. The reason the ready function in my instanced inventory node was being called twice is because a few weeks back I experimented with using that node as a singleton and never removed it.
091cf8 No.14621984
>random tornado warning
>shit going CRAZYYYY outside
welp, living in the tropics can fuck off I wanna dev
don't go out power
713a19 No.14622023
>>14621984
I'll pray for you desu~
13c436 No.14622045
>>14621806
>>14621833
>shouldn't physics be per frame, using delta time
Very, bad idea that needs some explanation.
Physics engines are a very special part of game engines. They have their own problems which haven't been solved yet. There are various ways to implement physics engines and each and every single one of them sucks in its own way. You get to pick your poison. Before I get into this, there is one thing you need to know.
Physics engines use approximate, discrete simulations. You don't get perfect physics in realtime, because that would be too slow. What you do is you take a snapshot of the level in its current form (containing all physics objects and the relevant variables like velocity and mass) and compare it to the last one. Then you try to figure out what happened in between and fix the problems in the current one. Take a collision for example: a collision is detected when two objects overlap. You then take the previous snapshot and use its information (e.g. the last known positions and trajectories of the respective objects) to figure out how they must have collided. After that you resolve the collision, meaning you unstuck them, rotate and move them around. Since collisions are detected by looking for overlapping objects, it is possible for objects to pass through each other. That happens regularly when small, fast objects hit thin ones. During one snapshot the object is on one side of the wall, during the next it is already completely on the other, thus no collision is detected, because there was no snapshot during which they overlapped. The longer the time frame between two snapshots, the more likely that becomes. That problem occurred in Fallout 4 when you unlocked the framerate. There are videos of this on YouTube where you can see glass bottles fall through shelves. With that in mind …
>tying the physics to your rendering framerate (variable time step)
Bad idea, because you tie the accuracy of the physics simulation to the framerate of the game. The slower your game runs, the more time passes between two snapshots and the more unreliable your physics get. If you've ever played a 3d game with lots of physics, you know how finicky they are. And that's when they work properly.
Then there's also a second problem. The delta time is the time it took to render the last frame. That means how much time passed between the last frame and the one before that. It has no relation to how long the current frame is going to take. Even in well programmed games it varies from frame to frame and you don't want that given how sensible physics engines already are. You simply need a more reliable timer to eliminate that variable. Your physics engine needs to know how much time passed in between two snapshots and it needs to know that with the highest degree of accuracy possible. Delta time is good enough for game logic, but not good enough for high precision stuff like physics. This requirements means that all well programmed game engines have physics run on their own tick.
>fixed time step
You can make physics run at a fixed rate. Unity does that and Unreal can do that, too. The upside is that you get the highest degree of predictability that way. The downsides are that they are not in sync and that you get temporal aliasing. The former is self-explanatory. In Unity, physics run at 50 fps by default. Let's assume your game runs at 60 fps. Here is a timeline of how your game updates itself (P means physics, G means game): 0ms (P+G), 16ms(G), 20ms(P), 32ms(G), 40ms(P), 48ms(G), 60ms(P), 64ms(G), 80ms(P+G), …
You see what happened there? The difference between them added up. The time between your physics engine getting updated and your game rendering is variable. That difference causes a hitch in your game when the simulation catches up. And on that timeline I assumed that the game always runs at perfect 60 fps, which won't happen in real life. Fixed time step is perfect for console development, when you know exactly what hardware you're running on.
Unreal uses a mix they call substepping. It's explained in detail here: >>14618354
Here is the short version: You use a variable time step with an upper boundary. Think in frametime rather than framerate (1000/fps = ms). If your framerate goes down, your frametime goes up. If you want your physics to run at 60 fps (16ms), but your game drops down to 30 fps (32ms), Unreal prevents the time in between the snapshots from becoming too huge by performing extra physics simulation steps, since there is enough time. That keeps the time since the last snapshot below 16 ms, but it also means that the more time passes in between frames, the more physics simulations are performed. Or in other words: the worse your performance, the more physics simulations your CPU is has to do in order prevent them from going to shit.
And that fucking sucks, because the people who have bad performance are the last people you would like to perform extra calculations.
404549 No.14622056
>>14621700
>Not using said drama to market your games
Get on my level.
2949a3 No.14622098
>>14619048
>>14619048
I think I may have misinerpreted that anons post >>14621796
I think they’re using a fixed time step…
Fixed is reliable, and a good direction to go; just give players the option to get that FPS even with toasters.
It’s be great if that anon clarified as to their approach (look at the gaffer games article for determining).
>>14622045
The semi-fixed Timestep approach looks pretty appealing as it solves some of the big issues w/variable frame rates (w/o the big issues of a purely variable delta time, which f.e. was misused in the gamebryo engine for fo4 as they tied it to the FPS itself).
Though I’m a unity pleb so I’m stuck with fixed, which isn’t so bad, and I’ll just give users performance options to get a solid 60fps (I use a 1/60 timestep).
>>14622023
Thanks anon
had to resort to mobile posting
power is intermittently going out
2949a3 No.14622108
>>14622056
Viable approach as seen with the original GTA, and games like hatred; though again a time and place. Biting bait in one’s own YT comment section is petty, but laying bait on the other hand can have great effect at the proper time/place.
091cf8 No.14622131
>>14622098
goddamn i suck at posting on mobile… first post quotes are supposed to be: >>14621907
sage for correction, excuse my plebbishness
13c436 No.14622160
>>14622098
The point of the semi-fixed timestep is to stabilize your simulation, if your performance goes down. You basically get to say: "My physics runs at x fps and is stable all the way down to y fps". The problem is as I said: It hurts those with the worst performance the most, which is quite unfortunate.
>>14621837
I sourced a PC with used parts on ebay. I decided that those will be the minimal system requirements and will optimize my game so that this PC runs it well. If it runs on weaker PCs, all the better. But the following PC will run it at 720p with +30fps no matter what:
>AMD Athlon II X4 620 4x 2.6 GHz
>4GB RAM
>AMD Radeon HD6670 1GB or Nvidia equivalent (GTS 450?)
>Free disk space undefined (as much as necessary)
The CPU is really awful. Its outdated architecture and only 2.6 GHz clock frequency make for rather abysmal single threaded performance. This means that the overhead I just get from Unreal's systems is rather painful. Thankfully Unreal now has a forward rendering path. It's so much faster and I don't need many dynamic lights at the same time.
091cf8 No.14622228
>>14622045
Ah ok, I see what approach you're using now; thanks for the clarification. My initial interpretation was sorta correct.
I'd just stick with the semi-fixed timestep/substepping approach, as it keeps things consistent, and is worth the causal issues you'd need to mitigate.
Each has their own issues, but overall I'd say the semi-fixed is worth the pains it can give toaster users; which in itself will make them choose lower-performance options.
>It hurts those with the worst performance the most, which is quite unfortunate.
Giving them options for what the actual target "game" framerate (not physics timestep) will encourage them to tune their settings to be at least that of the minimum framerate offered (so, minimum framerate offered = where simulation is still stable and conforms to lower bounds you provided).
As long as you give them a viable range of options to tune performance, and encourage them to remain within that bounds I'd say it's fine (without explicitly telling them, as most don't want to know about that stuff, they just wanna play).
>Delta time is good enough for game logic
semantics. this term can have multiple meanings due to different contexts of it being used.
It could mean the physics engine timestep (be it fixed/etc), or it could mean the time the last frame took to complete; or many other things as it's just "rate of time change" which the exact meaning of that depends on the context the term is being used in.
>>14622160
discard that post, didn't c your ID as temp switched to mobile, nor did i read your entire post b4 posting which answered the question i posted anyways; thus why i repeated the same point we came to for the fo4 variable time steps point
75a562 No.14622248
I've now implemented different days and a basic transition between them where you see how many fish you caught that day. Needs a lot of polishing but the rough groundwork is there.
13c436 No.14622423
>>14622228
The problem is that the player doesn't get to choose the physics settings. I set them in the project settings and those settings apply for everyone. So I obviously have to find reasonable settings. This brings me to the conundrum of my initial post.
Increasing the max. physics framerate makes for smoother physics for people with high refresh rate monitors. It also increases the framerate threshold below which physics break apart, which hurts people playing on toasters.
If I stretch the window in which the physics are stable, the people with the weakest PCs pay for it.
If I'm able to hack variable physics settings into the game, I'll do that.
141999 No.14622477
>>14622045
>it is possible for objects to pass through each other. That happens regularly when small, fast objects hit thin ones. During one snapshot the object is on one side of the wall, during the next it is already completely on the other, thus no collision is detected, because there was no snapshot during which they overlapped.
This is a fault that comes from using a collision test that checks to see if points have collided with the object, and not lines. If the engine uses a collision test that actually works, then it's not possible to do this at all. For example a BSP tree does not have this problem. So, even though in quake I can bhop to an unlimited speed, I can't fool the BSP tree. Variable time-step works if your game has good collision detection. This is why quake 3 did not have issues using a variable time step for physics, except the 333 fps bug that let you jump a tiny bit higher, which was fixed rather quickly.
Fixed time steps are lazy and make the game physics move in slow motion when you cant render as fast as the fixed step. Just fix the bugs in the variable time step physics instead of trying to avoid the problem, that way you don't have to slow down the game even more by substepping.
80bd45 No.14622499
HookTube embed. Click on thumbnail to play.
>Playing EFT for sake of research And because it's fun
>Comes to light that the firing rate of weapons is tied to framerate to disastrous effect
>Low FPS will lower your weapon's RPM
>Sudden framerate drops will make your shots not even register
I feel like people have been bitching about gameplay functions being tied to framerate for the last decade yet studios are still making this mistake. Why are they still making this mistake?
091cf8 No.14622569
>>14622423
>The problem is that the player doesn't get to choose the physics settings.
Hence my suggestions on encouraging players to stick near the lower bounds of stability at minimum.
If you really really can't stand setting the bar for performance, you could always do multiple builds with different physics settings; though that in itself causes issues.
>which hurts people playing on toasters
Well, it's pretty obvious from the UE default settings it isn't designed with toasters in mind.
UE is known to run well on mid to high end pcs (within a few year range of being considered, at that time, "mid to high end"), and mid-low or low end pcs are by default fucked; unless you offer a good range of performance tuning options that you have to put effort into providing.
13c436 No.14622628
>>14622477
ika?
What you're talking about is called CCD (continuous collision detection). Modern physics engines can do that just fine. You typically get to enable it on a per object basis. The reason why they don't do that for all objects by default is that it's incredibly slow and not necessary for the most part. You would use it for small, fast moving objects like projectiles.
Quake gets away with that, because the physics in that game are laughably simple. The geometry (level geometry in particular) is super simple and the levels are mostly static on top of that.
There is a reason why modern game engines moved away from BSPs as their primary way of creating geometry. They don't scale well at all when it comes to detail.
What you could do is model a moderately detailed environment and dump the triangulated mesh (as it would be send to your graphics card for rendering) into an .obj file. You can do that in Unreal and what it'll do is show you how awful the topology of geometry generated by BSP trees is. I added a screenshot of a test level's raw BSP geometry and after I fixed it by hand. Look at the triangle count. And that level is tiny.
BSPs love lots of flat, parallel and coplanar faces. And that's not modern games. With high levels of detail you get lots of split faces which up your triangle count. If you do that enough, you get to the point where the cost of working with all those extra triangles outweighs the benefit of the BSP data structure.
575e2c No.14622635
>>14622499
and people here like to pretend eurojank doesn't exist. that's pretty bad.
34dc5b No.14622650
>>14622499
Game devs offload physics processing to the GPU for efficiency reasons. That's why physics become tied to frame rate
13c436 No.14622652
>>14622569
I've posted the hardware I'm optimizing for here >>14622160. I already prepared a test level where I have a button that spawns 10 physics objects inside a relatively small room. Tomorrow I'll just benchmark it on that PC. Just to get some real world performance numbers. I want to see how it scales.
13c436 No.14622661
>>14622650
They're using Unity, so if something is framerate dependent, it's due to the code they've written.
397fbf No.14622673
>>14622661
Physics is always frame rate dependent, anyone who says otherwise is lying.
13c436 No.14622675
>>14622673
I meant gameplay systems. If the rate of fire of your weapon is framerate dependent, that's because you programmed it that way.
e73ff5 No.14622676
Sorry for the blog, but
>Get scheduled 5 days off in a row, maybe now's a good time to work on my game again
>Take monday off, so 6 days, cool
>Today, actual day off, wake up with a scratched cornea which seems to happen every month or two for no apparent reason, also a toothache
>Work calls and asks me to pick up 3 more shifts
>Basically only really have today and tomorrow off, and today is a write off
Fucking hate this
141999 No.14622693
>>14622628
yes
You're considering BSP as a rendering algorithm, when you should be considering BSP as a collision detection algorithm. For example quake 3 uses BSP for collision detection but does NOT use BSP for any kind of rendering because of the increase in polycount. Also that comparison with the polycount isn't really good anyway, it looks like you're using a compiler that doesn't know how to deprioritize polygon splits because that is pretty excessive for a BSP compiler to generate.
I know that we don't use BSP tree's in modern engines; but it's just a trivial example of how this problem is solvable.
Anyway a modern engine doesn't use the actual, high-detail geometry of an object for collision: it uses a simplified model that is easier to work with. See:
https://developer.valvesoftware.com/wiki/Collision_mesh
https://idkudk.blogspot.com/2014/08/creating-custom-collision-for-unreal.html
So it's not like we've moved past "laughably simple" geometry since such an approximation is good enough and easier to work with. It's just that we aren't colliding things with the literal world mesh anymore.
091cf8 No.14622705
>>14622652
best of luck, and even if worse comes to worse you'll get a good test bed for what performance tuning options you'll need to offer/work up to providing.
>>14622676
damn, that actually blows.
Honestly, sometimes it's best to just say no or to at least negotiate, but this is obviously highly dependent on many factors.
>wake up with a scratched cornea which seems to happen every month or two for no apparent reason
humm, I'd figure that out, I've heard that's really painful.
do u use goose down or any other "feather included" pillows/comforters/etc?
Also, could be that you're scratching your face in your sleep, which can be caused if you're subconsciously driven to do so like a breeze on your face or something to that effect.
e73ff5 No.14622793
>>14622705
Not sure really. It happens every month or two without fail, but seems to be related to lack of sleep and/or hydration when I go to bed. If I stay up late one night, and get say 5 hours of sleep and have eye fatigue when I do so, it'll probably happen. Or say I get 6h of sleep the first night, and 8h of sleep the next two nights, if I don't recover that missing 2h, it will catch up and hit me, I think.
I've seen doctors but they don't know shit. The hospital said they can keep me for observation (this was a number of years ago) but couldn't offer a solution. I usually just use a damp facecloth and sleep it off with painkillers.
6f7199 No.14623138
Fuck everything, I'm going to make shovelware.
f2c3f4 No.14623231
6e9873 No.14623241
>>14618469
>faster than coding
>simpler than coding
<some functions are not available
I dont really see the downside since its based on C++
6f7199 No.14623242
63f11e No.14623288
http://picosong.com/weNiT/
Worked on music and some model shit again today.
Anyway, I'm posting to say that Kowloon is on hiatus now since I'm working on a collab project with a few other people.
An anon last thread was asking about our group so I can give some details but it boils down to this:
>Last years UT stream
>An anon group forms
>Want to make Teagan Quest
>End up making Die Totenmaske for a contest
>TQ gets pushed back due to personal issues of a few members
>Working on D.T 2 now, members are people like myself, !Sena/Dorfanon, a /bane/ anon, a guy who did some designs for On Cosmic Tides (Moth girl related), an anon who did some work on previous Teagan Quest attempts, a FUCKING CATHOLIC, and a leaf anon.
We had a few other members who weren't involved in D.T and they're a brilliant bunch.
Our goal as of now is to make the best game we can utilizing some of the concepts from the old one that worked whilst eschewing RPGmaker as well as all the stuff we screwed up on.
80bd45 No.14623401
>>14623288
Is that model made by your team? I can't believe a model that good would come out of an /agdg/ thread.
Nice fuggable waifu, too
f2c3f4 No.14623435
>>14623401
>spoiler
There will be waifus in DT2.0 too
63f11e No.14623453
>>14623401
!Sena is pretty fucking skilled.
As for waifus in D.T we have plans for a few so far but it's mainly concept.
The one actual female character in the rpgmaker original barely got any of her personality/storyline shown due to gameplay and engine/time limitation.
80bd45 No.14623486
>>14623453
Man, that concept art reminds me a lot of my own art. I'm not sure how I feel about that.
I forgot the eye brows on this piece, plz no bully
dc2a8f No.14623491
>>14623486
>hair coming out of gas mask
Yes I still mad.
f2c3f4 No.14623493
>>14623486
Damn, you actually managed to make a cute "hazmat suit"
63f11e No.14623498
>>14623486
>Hair coming out of mask
f2c3f4 No.14623503
80bd45 No.14623512
>>14623491
>>14623498
Good, good. Your anger fuels me.
>>14623493
Thanks
f2c3f4 No.14623524
>>14623512
>Good, good. Your anger fuels me.
Better listen to them anyway. Still, I think it's acceptable if it's only to protect from very light hazards. That should be ok.
141999 No.14623542
I've now isolated what I believe is the last major error in my compiler. I'm going to trace through its debug output and hopefully fix it in the next few hours. Then I will be able to move on from the BSP compiler itself and rework parts of the rendering engine to work better.
80bd45 No.14623547
>>14623524
For my current plans radiation is only a light threat in most of the game's locales. With only a few areas needing any 'high grade' protection. The primary environmental hazard is currently planned to be anomalies.
63f11e No.14623558
>>14623542
Good job lad, it's looking great.
88f5e2 No.14623779
What's so difficult about making a debugger as good as visual studio? Why isn't there a good competitor?
dde486 No.14623793
>>14623779
Microsoft's programming tools are world class, and the only reason they are still relevant in the professional fields of computing. And yes, they are pretty fucking hard to make.
141999 No.14624111
>>14623558
Thanks!
I manged to fix the bug, however there are still more bugs that need to be fixed. So, I'm still not finished.
1f7872 No.14624207
I need to figure out something with music. I can't make any myself, so I need to buy it somewhere. Do kids these days like chiptunes? I am so out of touch. I figure it would go well with my asthetic but I have no fucking idea if kids would just chuck it in the trash.
66d6c6 No.14624260
>>14623241
Blueprints are much less efficient compared to C++, and they're just unpleasant to deal with. Coding is faster when your fingies can type fast. Really. There's very little reason to ever touch Blueprints unless you are dealing with UI, or specialized shit like animation blueprints. Call me a grumpy fuck who hates innovation, but Blueprints just reek of being a special needs product for people who can't figure out coding.
66d6c6 No.14624272
>>14624207
you should not pander to what's cool, you should strive to make your shit so good it makes whatever is involved in it cool. I grew up in a shitty Russian town and playing STALKER made me feel like I was living in a cool place because it made Russia feel like a comfy, stoic, tough-love place instead of being a corrupt shithole with horrible infrastructure and a higher concentration of retarded people than the U.S. It made me feel a patriotic hum in my chest despite the fact that I really hated the reality of what Russia was. STALKER had such an effect on my perspective that I now fondly remember the things about Russia that STALKER made me love.
You can see a similar thing happening right now with oldschool anime like Hokuto no Ken and Yu Yu Hakusho becoming cool again because JoJo managed to make corny badassery cool all over again.
5364af No.14624279
>>14624207
>compromising your own vision for your game in order to pander to what you think "kids" may like
Hello we at Generic AAA Studios would like to hire you as lead game designer
b4a361 No.14624282
>>14624207
Kids these days like Ed Sheeran, Kendrick Lamar, Justin Bieber, and The Chainsmokers. My question is why do you care what they like? If you like chip-tunes and it fits the game then that's all that really matters.
fb3b6c No.14624345
first post in a long time (im a lazy fucking nigger who usually just lurks)
rate my WIP build tool with holograms n shieeet
1f7872 No.14624351
>>14624272
>>14624279
>>14624282
Well, none of you are of course wrong. If a game specifically made for kids includes music that makes kids throw it in the bin instantly then why use it? It's also a matter of objectivity, I really enjoy a good chiptune myself but that of course clouds my perception if it's an actual good fit for the game. Solodev is the ultimate echo chamber after all.
I am however thinking that more orchestral music but with midi-overtones might be perfect. It preserves the kinda "low-res" feel of chiptunes while still being a bit more modern and friendly than chiptunes.
7a3eb2 No.14624402
>>14624351
In my experience children don't like old things so you need to find out whether chiptune is "old" or "retro" to them, because they might be more receptive to the second one. My advice would be to look through some games (especially mobile shit) kids like and see if there's any chiptune or something similar. Or simply play chiptune to some kid you know and ask them for their opinion on it.
1f7872 No.14624441
>>14624402
Yeah I've got some nephews and shit around that age, I'll try and play some music for them. Thanks anon.
7eee83 No.14624461
>>14624207
Kids like everything they're told to like by their peers.
ccec82 No.14624466
>>14624345
would need to see a vid, looks ok, everything else from the unity asset store?
3ba1d9 No.14624599
Hey, I'm checking the eduke32 engine's source code and most of the time, they don't use classes. Should I use them ?
091cf8 No.14624607
>>14624402
That's only because their peers told them so, as this anon said >>14624461
It varies widely, and it's common for said peers to have shit taste (see current epidemic of mobile shit being popular, and cash-in "gamedevs" who just want money like >>14624207 >>14624351 )
On average children are taught to be content with anything they have, which is good for certain contexts, but for games and entertainment in general they have a right to be more choosy (it's the one damn thing they have control of); so it's less of "mobile shit is good enough", but "mobile shit is all they're allowed to have" in terms of their choices.
>Or simply play chiptune to some kid you know and ask them for their opinion on it
If they have any kind of respect, or subconscious drive to get your approval; they'll "like it" if you say it's good.
It's sad, but very true; if they're the average kid.
You have to have a really solid relationship with them for them to really be open with you, as, based on how they were raised they more/less sensitive to things like disapproval/approval (f.e. authoritarian parents make kids desire approval, and fear disapproval to the point of preferring to lie to not get disapproval), and some kids are just spiteful little shits because they have tyrannically authoritarian parents (so they'll always say what they feel will make you feel bad, I have a family friend's kid whos like this, I feel really sorry for him bcs his parents are terrible).
Man, so many bad parents out there, it's fuggin' ridiculous, and an epidemic.
Like seriously, why are people so goddamn dumb? Why not invest a month of your free time into doing research, and learn parenting techniques, child psychology, and explore how to be a good parent?
So many issues with society could be solved if people made the effort… before they messed up their kids life with their one-off "experimentation" with their psyche (such as that family friend's kid I mentioned); which they have to deal with for the reset of their life.
091cf8 No.14624635
>>14622793
Damn anon, I really don't know what to say other than, I hope you feel better soon.
sage bcs off topic
1f7872 No.14624663
>>14624607
>cash-in "gamedevs" who just want money
Ayy
I'm pretty well aware that making a mobile game will not earn me any money. But it does raise the question, do you think it's unreasonable to do some adjustments for your target demographic? As an example, I think it would be hilarious to add ragdolls to enemies having them fly to kingdom come when they die, even gib them. Kids probably would too, but their parents probably would not, which means my game will not be played (at least looking to the parents I know who have their kids playing educational games). Instead I'm transforming enemies into various animals and having them scamper off the map. It's fun too, but not on the same level (for me, at least), but you're pretty much saying that's wrong? Do you always only make games for yourself?
13c436 No.14624787
>>14622693
How's Sigma 2 coming along?
>You're considering BSP as a rendering algorithm, when you should be considering BSP as a collision detection algorithm.
Hmm, I wonder what PhysX uses. Maybe BSPs for static and Octrees for dynamic geometry? Maybe something different altogether. Hard to tell given that it's closed source and I couldn't find anything on it either.
>Also that comparison with the polycount isn't really good anyway, it looks like you're using a compiler that doesn't know how to deprioritize polygon splits because that is pretty excessive for a BSP compiler to generate.
The Unreal BSP compiler is supposed to only be used for blocking out the layout of the level, hence the built in feature to export it into a mesh. That doesn't stop incompetent and/or lazy developers from using those inefficient BSPs in their levels though.
I know that we use collision meshes, I have to model them after all. By simple I didn't mean the tech used, but what was done with it.
Anyway, I found a nice solution, if I manage to get dynamic physics settings into the engine, which shouldn't be that hard. I just benchmarked some settings on the toaster and tweaked two separate settings. I'll simply add a "increased precision physics" option to the game. Standard precision works for 13-75 FPS without being too demanding and increased precision is for people who play at +90 FPS, but raises the lower limit to 26 or 40 FPS. But that shouldn't be a problem for someone who owns a 120 Hz monitor and the sort of CPU and graphics card to go with it.
091cf8 No.14624806
>>14624663
>I'm pretty well aware that making a mobile game will not earn me any money.
Well, why are you making some game that you won't play? What value do you derive from it?
Do you enjoy playing educational games made for kids?
>do you think it's unreasonable to do some adjustments for your target demographic?
My philosophy on gamedev has a "demographic" as an implicit outcome, and not an explicitly defined "target"; so it's more like "the demographic" and not a "target demographic". So this is a moot point.
>I think it would be hilarious to add ragdolls to enemies having them fly to kingdom come when they die
The main concept here that parents have an issue with is "dying", and "death" being shown to their children (as you've realized, via transforming the enemies).
>but you're pretty much saying that's wrong?
I said no such things, you're attempting to strawman some words that were "put in my mouth" as it were.
>fly to kingdom come when they die, even gib them
>[is] that wrong
The design concept of exposing an adolescent child to an innocent rendition of "virtual death" is no big deal, as it's not reality, and being able to distinguish reality from fantasy is an integral stage of a child's development.
Although, there is a threshold the child has to pass - for the child's psychological development - as to when this is defined as "no big deal", i.e. when a child can cognate this. So, if they're in adolescence then it's fine as they should have the cognitive capacity as to cognate this concept (unconsciously).
However, parents do not have a conception of this idea in most cases, and any degree of death is an instant "trigger" as it were to protect their child from the actual implications of reality itself.
>Do you always only make games for yourself?
Essentially.
Simply put, it is because I enjoy the process, and want to make something that I can have fun with as a result of that process.
I also like the option of letting other people enjoy it, and to engage in something I also enjoy; which is to have fun.
As I love to see people having fun, and if that costs me nothing to do that; then that's even better.
Anything else is a byproduct.
>>14624787
Quality plan mang
a7d62d No.14624909
>>14624806
>Well, why are you making some game that you won't play? What value do you derive from it?
I enjoy working on concepts that I find interesting and good. As long as I think someone might like it and I like working on it, that's good enough for me.
>My philosophy on gamedev has a "demographic" as an implicit outcome,
Well that works perfectly fine for most games, but for an educational game it does not because you lose focus or will bloat your scope.
We seem to be pretty much in agreement to be honest, just with different philosophies on what makes gamedev fun, which is fine. The democratization of game development tanked the professional side of development but it did make the true indie a lot more fun.
091cf8 No.14624981
>>14624909
>We seem to be pretty much in agreement to be honest, just with different philosophies on what makes gamedev fun, which is fine.
Agreed
>for an educational game it does not because you lose focus or will bloat your scope.
Understandable.
>As long as I think someone might like it and I like working on it, that's good enough for me.
The issue here is, is if the only thing that "guides" you is what is interesting + if other people derive fun from it; is that it's difficult to gauge exactly what it is you want for mechanics, and scope of what your game entails even with being defined as "an educational game with w-y age range".
Reason being, is that you're inherently rely on an extrinsic element to yourself (which introduces a boat load of issues).
So, it comes down to learning, "what is fun, on average, for said target demographic"; which is difficult unless you're in that demographic so it would thus emerge intrinsically.
You see what I'm getting at here?
In essence, an approach that inherently relies on an extrinsic element to yourself, for the derivement of a highly intrinsic/subjective phenomena, is inherently flawed.
To mitigate this you need to either: learn your demographic, do heavy playtesting with your demographic, and such things to gather data of an intrinsic/subjective experience (that, "fun for other ppl").
However, the outcome will always be skewed depending on the subjects that inform your results, and very likely to be flawed.
The only way to entirely mitigate this is to make a game for the demographic you're intrinsically within.
882b3b No.14625117
>>14624345
add post processing
676de6 No.14625625
>>14617199
I've finished my shitty popeye idle in a record of six hours.
7fb313 No.14625633
>>14618191
if you're trying to implement singletons I have to inform you that you've been misled by some 14 yo pajeet on a shitty forum (gamedev.net).
Never go cargo cult.
Hint, you are trying to implement resource management routines by hiding global identifiers in a static member of a class with a singleton, you accomplish nothing other than adding boiler plate to access a global not visibly defined as a global.
Second, if a singleton is actually worth using, you should use boost.
The oldest proven way to encapsulate global state is through API programming.
Use namespaces, and stop trying to make everything an object.
stop thinking about game programming, and read bjarne's book's if serious about C++.
7fb313 No.14625689
>>14625625
nice work.
I'm seeing popping/noise around the shoulder.
141999 No.14625767
>>14624787
I've been making good progress on the compiler, pretty soon all of the bugs will be fixed. I made a thread on /agdg/ here:
>>>/agdg/31318
I bet PhysX uses some kind of collision detection that has no branches in it. Since, branches on the GPU are the slowest thing you could do. So I doubt that it uses a tree because of that.
e33205 No.14625901
Does anyone know, where i can get well rigged 3d models to practice with?
Copyright doesnt matter, since its just for practice.
1f2028 No.14625913
>>14625901
mixamo.com
i've been using their shit mainly for the free animations and retargetting
676de6 No.14626170
>>14625689
Thanks.
Making modern NES games.
a7d62d No.14626283
Someone on here told me to implement groups being killed by answering one question, turned out pretty good, thanks to whatever anon that was.
9db2b0 No.14626314
>>14626170
>Making modern NES games
No you aren't.
You're not using the NES's color palette, your sprite has too many colors, and your sprite is too large.
397fbf No.14626316
>>14626314
Popeye is made up of background tiles, I'm planning on having the game run at 12 fps to make keep up with the NES' capabilities
9db2b0 No.14626349
>>14626316
That doesn't change the palette problem, does it?
676de6 No.14626484
>>14626349
>>14626316
>>14626314
There's no need to follow 100% the NES limitations, as long as it looks good.
e73ff5 No.14626889
>this tripe shows up because epic indie dev bot retweeted it
https://twitter.com/JDWasabi/status/984183095665614848
tldr: I virtue signal for diversity, so why didn't I get noticed while being a woman??
dde486 No.14626896
>>14626484
If it can't run on the hardware, it's not an NES game. If you're going for the '80s/'90s arcade style, just say so. Arcade boards were far more capable than consoles back then.
21aa56 No.14626926
>>14626889
you don't follow bots
676de6 No.14626927
>>14626896
I'm still trying to get the working style
2949a3 No.14626951
>>14626889
kek, the mongoloid monkey looking fuck doesn't know how to be assertive, and expects everything to be handed to her.
she's the definition of entitled, and definitely has full on princess-syndrome.
2f4cc0 No.14627041
>>14626889
>be autist
>try to network
>fail
>blame diversity
Ok
2949a3 No.14627180
i know this is shilling, but goddamn, rider is far superior to VS; i've been wanting these IDE features for awhile.
Interface is actually usable/easily customizable, settings are sanely organized, it's stable, it's actually fast, it has amazingly useful tools (more thorough refactoring tools, defining "rules" for formatting so no more accidental messing up of your format), and it has plugins I actually want.
e73ff5 No.14627197
>>14626926
Is there a way to follow a tag so it shows up in my feed? I follow a few other /agdg/ twitters, and a bot or two, just to see a few different things
25bb5b No.14627305
>>14627180
Intellij is great for all the languages I've used it for.
24aff4 No.14627361
>>14626889
>So short she thinks people's neutral gaze is an eye roll over her head
1e2e62 No.14627667
I got kinematic movement working somewhat, but it wasn't quite doing what I wanted. But I had an epiphany while on break at work, and now I know what I was doing wrong. I feel kinda stupid, though, since the solution was obvious. I was summing up forces on an airplane, solving for velocity, and setting that as my next position, instead of solving for position (which is exactly what the "kinematic equations" boil down to essentially). Now to implement what I supposed to have done from the start. Manual labor kills brain cells, I tell you.
>>14623288
>anon group forms from UT stream
>UT stream
Those were fun times, speaking of which, when's this year's stream supposed to occur? Did I miss it, or is it not til August or some shit? I forget.
bec214 No.14627814
>>14626927
>sailor suit
>wide hips
>thin waist
>legs bare from the knee down
Damn, Popeye's looking sexy
Your proportions might be a bit off
e73ff5 No.14627827
>>14627814
>Popeye, but instead of getting bigger forearms every time he eats spinach, he gets THICCer
ad75cf No.14628023
>>14627827
>Popeye, but rule63, and everytime she eats dick she turns into a mega-slut
b4a361 No.14628056
YouTube embed. Click thumbnail to play.
Good news everyone!
Simulated robots can now learn martial arts!
ea0909 No.14628069
>>14628056
That looks like Toribash but cooler.
b4a361 No.14628093
>>14628069
It's just a tech demo, it's far from being an actual game.
c0cfed No.14628099
I feel like a nuDev faggot for using premade engines, however when I look at modern OpenGL 3/4 I am disgusted by the C++, yet OpenGL 1/2 is too unoptimized and undocumented. Maybe I'm more or less angered by modern technology rather than game dev.
9b0130 No.14628108
YouTube embed. Click thumbnail to play.
>>14628056
>not simulating muscles with a neural network
6f7199 No.14628115
141999 No.14628125
>>14628099
OpenGL 1/2 has a ton of resources, they are just older. If you target older hardware, then they are optimized for those cards, and all that you are losing is performance on new cards. But since it will run on older cards, that means that it will still be very fast, just not as fast as it could be.
c0cfed No.14628126
>>14628115
Might be nice if there was tutorials and it could work on old hardware.
b4a361 No.14628160
>>14628108
They explain in their blog why they avoided this method.
>In most RL benchmarks, simulated characters are represented using simple models that provide only a crude approximation of real world dynamics. Characters are therefore prone to exploiting idiosyncrasies of the simulation to develop unnatural behaviours that are infeasible in the real world. Incorporating more realistic biomechanical models can lead to more natural behaviours. But constructing high-fidelity models can be extremely challenging, and the resulting motions may nonetheless be unnatural.
c0cfed No.14628231
With lack of decent frameworks on C, I'm thinking of learning Java to use LWJGL. I know it's a pajeet language, but I can't find any decent shit that uses pure C.
9b0130 No.14628243
>>14628231
If you're going higher-level, why not C++?
c0cfed No.14628258
>>14628243
I've worked with C++ and its a pile of garbage with hacked on OOP and portability. It's major ass and I always hated working with it couldn't stomach beyond Hello worlds.
c0cfed No.14628267
Then again, if I'm forced to use an IDE I'm dropping Java immediately out right.
e73ff5 No.14628289
>>14628258
>>14628267
Let me shill C# to you. It had both SFML and SDL bindings, and OpenGL.
You also have Mono as an open compiler and Monogame or FNA as frameworks that supercede the Microsoft XNA framework, which was pretty robust.
Personaly I'm using VS by choice with SFML
Plus the syntax is basically the same as Java
dde486 No.14628314
>>14628267
Have you considered using the true successor to C?
https://dlang.org/
141999 No.14628317
>>14628231
Why not SDL? It's written in C and probably does about everything you want. SFML is written in C++ but has C bindings that you could use, too.
Alternatively you could shed frameworks entirely and learn how to program against the OS like so:
http://www.winprog.org/tutorial/
http://nehe.gamedev.net/tutorial/lessons_01__05/22004/
https://www.openal.org/documentation/OpenAL_Programmers_Guide.pdf
https://www.khronos.org/registry/OpenGL/specs/gl/glspec15.pdf
c0cfed No.14628327
>>14628289
C# is good, but with mono being owned by Microsoft I don't trust it. I'll look into monogame.
>>14628314
memeland also barely any bindings.
>>14628317
>winprog
>windows
Yeah no.
>nehe
Nehe tutorials only get you so far and his requiring that huge list of libraries is a no go.
b4a361 No.14628330
>>14628267
>if I'm forced to use an IDE I'm dropping Java immediately out right.
It's possible to write Java without an IDE, but you'll be about as productive as a lawyer without a library.
66d6c6 No.14628337
>>14628099
>Avoiding decades of work piled on by giants like Epic just because it's not ebin haxxor enough
cmon now
141999 No.14628338
>>14628327
You really only need to link to OpenGL and X if you're on Linux, like this:
https://www.khronos.org/opengl/wiki/Programming_OpenGL_in_Linux:_GLX_and_Xlib
Beyond that the OpenGL spec I linked should be all you need to learn more. Tutorials are really only abridged versions of the spec, anyway.
c0cfed No.14628345
>>14628337
Epic did nothing. It was all Carmack at iD.
66d6c6 No.14628352
>>14628345
Okay but still, it's an AAA-grade engine given away for pennies
80bd45 No.14628364
>Fiddling when I can with a very early first-person character controller
>Still doing a lot of research on similar games
>Discover there was actually a flash game in development with the working title "Innawoods"
>Was going to be an isometric survival turn-based tactical RPG with a nice artstyle
>Game died and the only thing that's finished of it is the character creator (Which can be found here https://innawoods.net/)
There's a lot of dead and failed projects in this genre I'm discovering. Feeling a little bit of pressure to one, not let my project die and two, to make it really good to make up for all the dead Stalkan/Innawoods games.
e73ff5 No.14628366
>>14628337
>>14628352
>>14628345
If it's not open source then you are the product being sold.
Reminder that Epic is owned by chinks who mean you I'll.
b52b28 No.14628370
Can some anons give me advice? I'm lost at what language I should build in.
I have already built about 1/2 of my base engine in Java, and although I'm quickly learning C++, translating and rebuilding the engine difficult and daunting. I'm just not sure what to do at this point.
b52b28 No.14628376
>>14628370
>rebuilding the engine looks difficult
1e2e62 No.14628401
>>14628370
Fuck the haters, do it in lisp.
1297c4 No.14628432
>>14628364
>innawoods
Ah… memories.
e73ff5 No.14628468
>>14628432
Speaking of old shit, what was the date of Catmouth Island?
e73ff5 No.14628469
>>14628468
Oops I mean fate
7a3eb2 No.14628526
>>14626889
>she fell for Open & Inclusive™ meme
She has yet to learn that autists have to swear fealty to some social justice feudal master before being allowed to network with normal people.
9b0130 No.14628533
>>14628364
I remember that game being discussed on /k/ a while back.
>Feeling a little bit of pressure to one, not let my project die and two, to make it really good to make up for all the dead Stalkan/Innawoods games.
Maybe focus on just the key mechanics and see if that is fun and then build from there?
d3dd78 No.14628618
You can now fish up your waifu
70a969 No.14628640
Took me a minute until I saw the filename and grinned.
e73ff5 No.14628729
1f7872 No.14628771
>>14628370
Making an engine is the surefire way to not finish your game, because of faggotry like this.
9db2b0 No.14628827
>>14628771
Funny how three out of the six finished projects from these threads are made with a custom engine, and one with a modified version of an open source one no one uses.
2f15b9 No.14628830
>>14628366
How much of epic do you think the chinks own?
I get not wanting to owe 5% of your raw revenue and having a license attached to it that could mean that they may try to stop you from using it (although i think the most that theyve done is ask that their logo is not used in the game or for advertisement).
09ca03 No.14628839
>>14628771
People don't normally switch languages all the time.
>>14628827
This makes me wonder, what's the ratio between games made with a self made engine vs games made with a big engine like unity/gamemaker?
9db2b0 No.14628894
>>14628839
>what's the ratio between games made with a self made engine vs games made with a big engine like unity/gamemaker?
Depends on if you count shovelware. If you do, tens if not hundreds of thousands to one. If you only count relatively good games, nowadays probably still a hundred to one since AAA has started to use Unity/UE4 more lately. At that pont it becomes kinda blurry if you should count those AAA games made before then as custom engines, though, since Unreal used to be one of those too.
1f7872 No.14628902
>>14628827
What games are that?
>>14628839
>People don't normally switch languages all the time.
It seems pretty common to be honest.
09ca03 No.14628915
>>14628902
>It seems pretty common to be honest.
Maybe among noobs who haven't made much thus haven't established their preferred way of developing things.
9db2b0 No.14628921
>>14628902
>What games are that?
Speebot, Hypnorain, and Red Sky. Hellbreaker was made in a modified version of Urho3D. The other two projects are Die Totenmaske v1 (RPGM) and Shinobu's husband's marriage simulator (UE4, not publicly released)
>It seems pretty common to be honest.
Only people who are pretty new to programming and still looking for what they actually want, plus people who never finish anything.
141999 No.14629027
>>14628902
People who start a lot of projects and never finish them can switch languages a lot; they don't have very much work to lose by doing so.
Also:
http://8agdg.wikidot.com/general:finishedprojects
>>14628827
It's probably showing some kind of skill gap: The programmers who did finish their games, are programmers who deliberately did everything in the most difficult way, even though they could have been much more productive using engines and other tools. When really, we should probably see more people in a middle ground between abandoning your project and doing a ton of extra work on top of finishing your game.
1f7872 No.14629147
>>14628915
>>14628921
>>14629027
>Maybe among noobs who haven't made much thus haven't established their preferred way of developing things.
>Only people who are pretty new to programming and still looking for what they actually want, plus people who never finish anything.
>People who start a lot of projects and never finish them can switch languages a lot; they don't have very much work to lose by doing so.
Sounds like AGDG to me.
09ca03 No.14629201
>>14629147
It's not like people who use Unity/etc are any different.
141999 No.14629332
>>14629147
This is just the nature of game development- lots of anons start projects and then abandon them shortly after, and chase the next language/engine/framework, thinking that this will be the one that gets them past the honeymoon phase of their project.
2949a3 No.14629799
>>14629332
>honeymoon phase
Honestly the main issue I ran into into the past, for bigger games/code bases, was maintainability; which admittedly was from my inexperience.
In the process of getting over the hump of “just code monekeying shit” i made a lot of smaller prototypes (2 half finish mostly mechanically complete games, a semi-finished engine, a half dozen “this would be cool” tech demos, so less than 10k lines of code per).
After that I made two larger frameworks (15k+ lines), and that maintainability issue was biting me in the ass despite being pretty feature complete for what I was going for; which as many probably guessed was due to bad practice and ill suiting design paradigms (OOP street shitting, even with hints of SOLID, mixing a lot of design patterns without clear API programming).
That is, this happened even with going to uni for CS, and were taught to me as being “best practice”.
Despite that shit, i found its best to go with an approach that’s centralized on procedural programming, API programming, and true ECS for frontend; that’s EXTREMELY consistent. Hiding any inconsistencies to my main design pattern behind a separate namespace of my API so I can just code it and forget it.
With, in my mind, various design principles being the guide to how this works for me. Just articulating those past issues, and solving them through principles like inversion of control, layers of abstraction, SOLID, and such really help to keep things consistent.
Also consistent formatting helps a lot for a huge codebase.. oh damn does it.
Through that culmination of shit going wrong, and bad practices being learned, tried, failed, and killed; I’ve finally reached the mentioned goal of maintainability. Which has made the “code this idea” extremely easy due to having a very well defined “lens” by which to visualize how to program it.
In addition to all the other things that have to occur mid-development being easy (maintaining, extending, changing, etc).
So I’d say it’s mostly inexperience, and the prevelance of bad practices being ‘’’so easy’’’ to form; in addition to those bad practices being touted as “best practice” by educational institutions + online teachers. Which leads to messy codebase that simply fall apart, and are a nightmare to maintain (which is fine for small stuff, but terrible for a dynamic codebase like a game).
2949a3 No.14629892
>>14628370
I would articulate, to yourself, why you’ve come to this point.
After that, figure out how to prevent it, and learn from the mistake (I know it sounds a bit tedious, but really articulating that stuff in detail can help u see it in a new light, and really helps to learn from it).
2949a3 No.14629900
>>14629892
>mistake
More like, “mistakes that built you up to this point”
Also, it’s generally not the language (but the big shift from auto to manual memory management may be a big consideration for u, not sure), but in the case of most of the issues it’s how the language is used.
sage bcs dbl post
d9bd6d No.14630601
YouTube embed. Click thumbnail to play.
>>14621752
>anime based head
>don't look like an alien as much as it should
d9bd6d No.14630659
>>14628056
love the bullying
141999 No.14630681
>>14629799
Writing bad code is inevitable when learning to write good code. But finishing projects is all about being dedicated to writing your unmaintainable mess, instead of stopping before you cross the finish line. Having your code turn into an awful, hairy mess that you don't like anymore is just the end of the honeymoon phase for a programming project. Eventually you will stop feeling good about the code you have written, but to finish a project, you have to continue despite that.
d9bd6d No.14630701
>>14630681
>Writing bad code is inevitable when learning to write good code
you can remove that conditional
your 6 months old code will always look hideous to even yourself, unless you stop improving
7fb313 No.14630733
>>14628401
modern lisp has better gc latency than java.
When racket adopts the chez compiler, its gonna be a serious option.
naughty dog uses racket.
141999 No.14630737
>>14630701
It's not like most code can continuously be written better- after a certain point you will be able to write most simple routines "well enough". At that point programming skill has more to do with understanding complicated data structures/ academic papers and being able to apply them to your code.
6abeea No.14630744
>>14630681
> tfw not really programmer and feel consistently great that my pajeetcode even works
d9bd6d No.14630759
>>14630737
if not the structures and procedures, you will still hate the approach you decided to take
fbb6ea No.14630777
Aloha fuckheads!
It's your resident waifu simulator faggot dev here.
Here's more hardcore (not rly hardcore) wonderful wolfgirl action!!
Spent some time creating somethign that's supposed to be this anime 'super rage glow' aura stuff, tried another approach first, using a mesh skinned to the character with some flamey looking material on it, but we decided to go with a post process effect instead.
I need everyone to subscribe to my goddamn website too, because it makes me feel like i matter or something. go to www.shimapanzer.com and give me your email! >:(
Shit's coming together nicely, going to have to take a pass as a proper decent looking UI design soon, which wil be fun, and i'll actually have something INTERESTING to post here, instead of shilling waifu after waifu after waifu.
141999 No.14630781
>>14630759
This is sadly true. I have to spend some time destroying the OOP-styled code in my project rather soon.
1f2028 No.14630784
>>14630777
shit, how the fuck do you get the flamey postprocessing?
i want that shit, i only have generic glowing
fbb6ea No.14630807
>>14630784
here's a bit more exagerated flamey effect, you get the idea.
It's just a screenspace effect added to the 'outline mask'
it's a few panning masks overlaid/multiplied to get the flamey effect, then added to the outline essentially.
HOPE THAT HELPS!
ec793f No.14630822
Son of a bitch, I need some help from anons with a fresh perspective. I'm trying to write the Field side of my RPG, but I cannot for the life of me figure out what to name components related to a character on the field. CharacterData refers to raw data in my code, so I don't want to reuse that for something else. The only thing I can think of is "Toon", but that feels repulsive and disgusting and I don't want to use that term.
Fuck, man. I got Lua integration into my battle system and things are coming along great, but naming the all-encompassing world character classes is extremely difficult.
f2c3f4 No.14630828
>>14630777
Nice dubs but you write like a goddamn redditor
1f2028 No.14630830
fbb6ea No.14630834
>>14630828
Nigga, i make fucking waifus and tanks, you think the faggotry of this might be getting to me and influencing my writing??
ec793f No.14630839
>>14630830
Pawn is a good one, Actor sounds pretty all-encompassing for characters. Should've been obvious, so thanks for the direction.
2949a3 No.14631022
>>14630681
Eh, I've finished tech demo projects and made my framework projects feature complete before moving on, but your argument is relying on the sunk cost fallacy; which is in itself also a pitfall of the project cycle that one needs to be aware of.
It's good to have experience in working through that, but it's definitely not an if… but a when.
However, this inevitability can be mitigated to the point of it becoming trivial. Hence, my point on maintainability, and proper usage of design principles; such as inversion of control. E.g. want to rewrite all your logic? Ok, go on ahead, it won't break anything if IoC is done correctly.
>>14630737
>At that point programming skill has more to do with understanding complicated data structures/ academic papers
This is one of the most important skills, but it's not the superset of "programming skill"; it is a subset of programming skill.
Bout to get hyper autistic
What that is a set of, is the ability to cognate different levels of abstraction. That is to say, being able to abstractly visualize algorithms while having the ability to coherently articulate them into code for your own usage case.
Without going into too much detail, what this culminates into is the ability to visualize the flow of one's logic into a consciously constrained, and orderly manner at each of these levels in the same contextual thought process; i.e. to metacognate, or to see "the whole is greater than the sum of its parts".
Although, to do so in an effective manner is a whole 'nother ball game. Noticeably, this is an everyday skill, but generally a skill most willing lack a coherent conception of.
i.e. they use it in a conscious manner, but lack the self-awareness of the act of doing so (the act of metacognating metacognition).
Now, to refine this skill is directly related to programming, and includes your point; although the point here is to refine not just a single set of this skill but the entire superset that is the root of the skill.
Simply put, to build a proverbial toolset for programming with design principles, and design patterns; is to refine said skill.
Note that I'm leaving out the ability to articulate your toolset, i.e. to code it using your language of choice. This is important too, but is generally built through experience, and the process of trying, failing, learning, and rebuilding.
ad75cf No.14631052
>>14630777
Obligatory edit. Also nice trips.
2949a3 No.14631064
>>14630807
>>14630777
>meme spacing
>cringe inducing wording
not sure if this done ironically or not…
Nice models though, looking quality as always.
>>14630822
PC (player character), agent, avatar, entity, unit, being…
I'd go find a thesaurus and look at synonyms.
ad75cf No.14631098
I really need to start working on custom UI themes, godo't defualt is gross.
>>14631052
Whoops, uh fair warning I didn't mix the audio at all so rip headphone users.
397fbf No.14631130
>>14630834
You do write like a redditer, but we ignore it because you're actually making a game
086a50 No.14631161
Occasionally, I'll accidentally solve a bug and not know how I did it. When this happens, should I just accept it and move on for progress's sake, or try and recreate the problem for knowledge's sake? I feel like either one would be negligent.
aa0f1e No.14631216
>>14631052
was it a Japanese version of 'what does the fox say'?
141999 No.14631235
>>14631022
If you finished your projects it doesn't really apply to you, since you completed everything, but I think that you're using the sunk cost fallacy as a way to rationalize a lack of commitment to finishing projects.
I don't really follow the second part of your post
ad75cf No.14631289
>>14631216
Nah, it's a song from the Dragon Ball Super OST.
b8ecfe No.14631295
So i have been dreaming of making 2D games ever since i was a kid and had a fondness for games i could only describe as 'top down like pokemon'. Things like Links Awakening, Legend of the River King and Dragon Warrior all conformed to what i would later learn is a tileset based spriteart graphical style.
Now i'm an adult and work mostly in video because i found it a lot easier to make my ideas come to life when its tangible but i still have that childhood dream of making my own games like those. A dream recently revitalised by playing games like Corpse Party and Yomawari Night Alone.
Question is how do you start? as a kid i fucked around in an early 2000's version of RPG Maker but that seems pretty limited. I've seen some excellent stuff about making tilesets and gifs and importing them into unity to achieve something good. But then i heard about people using C+ and custom stuff from scratch.
How do you get started? i've fucked around in RPGMaker enough to make maps, towns, transports between screens, dialogue windows and options and interactions but i'm really feeling inspired lately to make the sort of thing i want to play but doesn't exist if that makes sense.
I know theres beginners guides but personal opinions from people working in 2D are what i'm after: for someone just starting in 2D what do you prefer?
aa0f1e No.14631334
YouTube embed. Click thumbnail to play.
>>14631295
Unity 2D, there should be everything you need besides events, those you code yourself in C#
Well, database is also a thing missing in Unity that was present in RPGMaker, you would have to code each thing manually, battle system too…
Actually, you will have to code everything in Unity from scratch besides the 'draw stuff to screen' part, but it's still easier than starting from scratch scratch, because they actually do have a tilemap thingy now
aa0f1e No.14631344
YouTube embed. Click thumbnail to play.
>>14631334
and this crap was literally the first link I've found, ignore it, i caught cancer in less than 2 minutes of watching, this one seems better
b8ecfe No.14631406
>>14631334
>>14631334
Both are interesting. I'm really looking for things to learn while experimenting. RPGMaker is nice and easy. At its hardest events are like making a html web page on the backend but its been nice for some proof of concept stuff. I think a lot of the modern indie crowd and their hipster politics bullshit was so annoying i didn't think about the 'fi they can pull it off so can you angle' but others like 'kitty horrorshow' who dont care about anything but satisfying their own autistic expressions are a good source of inspiration in a way.
That and corpse party. I would love to see a documentary about that game since its basically looking at a program meant to make dragon quest clones and going 'okay lets rip out all the combat and random battles and make a horror game' and i can appreciate that.
2949a3 No.14631481
>>14631235
>If you finished your projects
Certainly not all of them. The instances of finished projects are the frameworks + tech demos which were later in my history of programming.
Admittedly, when I first started I was abandoning projects half way or part way through; which ended when I learned of the "project cycle" as visualized here >>14621538
>I think that you're using the sunk cost fallacy as a way to rationalize a lack of commitment to finishing projects.
Understandable that you see it that way, as you have an outsider perspective to the chronological order of my mentioned finished/abandoned projects, and are misinterpreting my point.
To reiterate, as mentioned above, the unfinished projects were when i was beginning; which in retrospect were valuable lessons.
The whole idea of the sunked cost fallacy, and it's application in the context of a codebase; is to just be aware of it.
The experience of it, and thus hopefully the awareness of it; leads to an imperative to implement things like design patterns that mitigate or eliminate that possibility from arising (as I mentioned above, using IoC to give maintainability).
Therefor, if done properly, you utilize the patterns that prevent a project from passing that threshold of abandonability in the first place; as it is maintainable, it is extensible, etc etc.
>don't really follow the second part of your post
Basically, it boils down to: "how to consciously refine your thought processes".
2949a3 No.14631550
>>14631161
In the end it depends on which decision's implications are more valuable to you, but if it's not a "big decision" then it's not really too important either way.
141999 No.14631644
>>14631481
I'm sure you're right about how i'm seeing things, and how I don't really know you. Reading your posts again, I think you're right: just because someone is stuck in that trap, it doesn't say that he isn't committed, but is just not experienced/competent enough to complete the project yet.
086a50 No.14631746
>>14631550
>but if it's not a "big decision" then it's not really too important either way.
It's the most important part of the most interesting game mechanic, so I guess I'm going to.
2949a3 No.14631825
>>14631644
>just because someone is stuck in that trap, it doesn't say that he isn't committed, but is just not experienced/competent enough to complete the project yet.
Essentially, yes.
I'd like to expand on that actually.
An analogy is of a beginner snowboarder trying to do a triple diamond run after gaining that initial confidence (see first mount of attached pic); while still having much further to go, while lacking that "threshold" of experience to even gauge their skill in the first place.
This applies to anyone who has pursued crafts with similar skill ceilings, "thresholds" of experience, that must be broken through.
Though, as u know, that's why the most common advice is to make an early gen clone; like ping pong, pac man, and the like as it's just simple enough to provide the necessary challenge for a beginner.
So… it comes down to finding how to create just enough challenge as to be engaging enough, but not enough to be entirely overwhelming; as to straddle that line between order and chaos.
Which is really where human beings learn the most, in the most efficient manner, and is beneficial for other aspects of being a conscious being too.
Lastly, that's one of the reasons why programming is so great, as you're always learning, and there's more frontiers one could explore, or create; in a countless number of lifetimes.
9db2b0 No.14631857
>>14631161
You upload your fixed version to your version control,then roll back to the broken version and try to figure it out. There's always a chance that it's a bug that doesn't trigger at all times and you're only assuming you fixed it, though.
>>14631295
I think newer versions of RPGMaker have different scripting (Javascript I think?) which would make it a lot easier for someone with little to no experience in game making to work with it. Doing everything yourself is relatively viable for a project as simple as a 2D topdown game, but you having absolutely no programming knowledge means you'd lose more time to learning how to make a working base than it would take to just make a full game.
a7d62d No.14631858
I just finished my first ever animation, please rate and hate. I'm worried that there's too much motion, I was going for something similar to the Crash Bandicoot intros, going to add springy sounds as well.
b8ecfe No.14631865
>>14631857
I fuck around in rpgmaker for experimentation on themes. Like can you make a horror game? a different take on survival games? or just different ideas behind stats other than just hp, mp and so on.
da968b No.14631886
>>14630681
That was beautiful, anon. Thank you. Love for your projects in unconditional.
b4a361 No.14632159
I love it when my graphics research papers teach me how to commit arson.
aa0f1e No.14632296
>>14632159
you never know when you might need to know these things
397fbf No.14632464
void Tween::_add_pending_command(StringName p_key, const Variant &p_arg1, const Variant &p_arg2, const Variant &p_arg3, const Variant &p_arg4, const Variant &p_arg5, const Variant &p_arg6, const Variant &p_arg7, const Variant &p_arg8, const Variant &p_arg9, const Variant &p_arg10) {
pending_commands.push_back(PendingCommand());
PendingCommand &cmd = pending_commands.back()->get();
cmd.key = p_key;
int &count = cmd.args;
if (p_arg10.get_type() != Variant::NIL)
count = 10;
else if (p_arg9.get_type() != Variant::NIL)
count = 9;
else if (p_arg8.get_type() != Variant::NIL)
count = 8;
else if (p_arg7.get_type() != Variant::NIL)
count = 7;
else if (p_arg6.get_type() != Variant::NIL)
count = 6;
else if (p_arg5.get_type() != Variant::NIL)
count = 5;
else if (p_arg4.get_type() != Variant::NIL)
count = 4;
else if (p_arg3.get_type() != Variant::NIL)
count = 3;
else if (p_arg2.get_type() != Variant::NIL)
count = 2;
else if (p_arg1.get_type() != Variant::NIL)
count = 1;
if (count > 0)
cmd.arg[0] = p_arg1;
if (count > 1)
cmd.arg[1] = p_arg2;
if (count > 2)
cmd.arg[2] = p_arg3;
if (count > 3)
cmd.arg[3] = p_arg4;
if (count > 4)
cmd.arg[4] = p_arg5;
if (count > 5)
cmd.arg[5] = p_arg6;
if (count > 6)
cmd.arg[6] = p_arg7;
if (count > 7)
cmd.arg[7] = p_arg8;
if (count > 8)
cmd.arg[8] = p_arg9;
if (count > 9)
cmd.arg[9] = p_arg10;
}
ad75cf No.14632551
>>14632464
>that method signature
I know you're baiting for (you)'s, but how can yo write all that when loops exist?
397fbf No.14632574
>>14632551
I'm sure there's a good reason for it
73df61 No.14632732
>>14630777
>waifu simulator faggot dev here
that is shinobu anon and you know it. good work still
2949a3 No.14632740
>>14632464
>>14632574
they may have just the right excuse
also obligatory edit
e73ff5 No.14632899
470b5c No.14633019
So, what is the future of rendering pipelines, clustered forward or some variant of deferred? Is clustered forward just rising because gen 8 consoles don't have the bandwidth for huge G-Buffers at the mandatory 90fps for VR? Or is it actually better than all deferred techniques?
086a50 No.14633099
>>14631857
>Implying I have a version control to upload to.
>Implying I don't just have shittons of backups and written notes.
What do you take me for, a professional?
>There's always a chance that it's a bug that doesn't trigger at all times and you're only assuming you fixed it, though.
That's happened before, and this time, but I did finally figure it out conclusively. I was passing a wrong variable.
ad75cf No.14633296
Finally got a rudimentary shop system in place, the buying and selling of bingas' now works like a charm.
e73ff5 No.14633429
This isn't really related to anything but I'm probably going to phase away from the /v/ threads and instead shift my attention to the slower threads on /vg/ or even /agdg/ itself
Mostly since I can't stand giving traffic to this board with the way Mark runs it, and I also don't have time to keep up with the faster threads, so maybe I'll be able to focus better
ad75cf No.14633544
>>14633429
Does /vg/ get any traffic here? I know that people mainly just use /agdg/ threads to keep progress logs. There's also the irc, it's pretty slow and people usually just idle on it, but it could use more love.
e73ff5 No.14633588
>>14633544
I don't have a problem with the irc but I feel like it's just be a namefag hangout more than anything. Like it wouldn't be strictly formal and dilute from it's main purpose. Still better than disagreement, though.
Also /vg/ got new owners a few weeks ago. They try to heavily moderate things so there is little shitposting.
4d18b5 No.14633698
>>14629799
What kind of API are we talking about here? Do you break down each part of a game, things like try to group all rendering, controls, logic, ai etc? Or is it something else?
Also any books you'd recommend?
397fbf No.14633752
>>14633724
Alright, it's bad enough that you type like a redditer, but don't cap your reddit posts and share them here.
66d6c6 No.14633771
>>14633752
I'm not the guy.
80bd45 No.14633894
>>14633724
I mean. Good for them for running a good social media campaign (Looking at their Twitter stats it's fairly successful too). But if the guy wants to post here, he should really lurk moar and learn to not format his posts like a retarded redditor who thinks he doesn't need to proper utilize the unique formatting of other websites.
397fbf No.14633900
>>14633894
he's treating us like twitter goys, he only talks about actual development when he's forced to and the rest of his posts are just marketing
66d6c6 No.14634047
>>14633894
Oh, I'm not saying he's doing anything wrong. I'm just stirring the pot
73df61 No.14634261
>>14631858
nothing to complain about it. Its pretty good.
1f7872 No.14634311
>>14634261
Thanks anon, I appreciate it. I've added a small idle animation as well so it doesn't go entirely static by the end. Made the splash animation a bit shorter as well but still working on the timings and stuff.
329d32 No.14634496
>>14630777
>>14630834
If you are gonna post about your game dev stuff do so normally. But don't come in and talk like a fucking redditor or think you have special permission to do so. Fuck off and stay there instead.
2949a3 No.14635182
>>14635140
CGpeers my anon
21d180 No.14635269
>>14633019
Difficult to say with realtime raytracing being pushed by all major players right now. I don't know whether it's using its own thing or whether it is used to augment existing rendering techniques. If the latter applies, I don't know whether it's used in a deferred pipeline, a forward one or whether it works with either.
Realtime raytracing aside, both have their use cases since they have unique up- and downsides. Deferred sucks at transparency, refraction and can't do MSAA, but exceeds at dynamic lighting. Forward sucks at dynamic lighting, but can do transparency, (fake) refraction (need raytracing for proper refraction) and MSAA. Clustered forward rendering makes forward rendering suck less at doing dynamic lighting, but it's still nowhere near as good as deferred.
Deferred isn't only bandwidth intensive, it also has a substantially higher upfront cost. I'm using Unreal 4 with forward shading, because I don't use excessive dynamic lighting. Just switching to forward rendering gave me an insane performance uplift on the machine I target as min. system requirement.
If you want to make lots of use of dynamic lighting, there is no way around deferred. If you don't, forward will serve you well and perform a whole lot better while doing it.
2949a3 No.14635303
>>14633724
I had a feeling, but didn’t want to have to endure reddit to confirm it.
>>14633894
Agreed that’s it’s good that they’re running a good marketing campaign, but still, this >>14633900
>>14633698
Ill get back to u on this soon, when I’m at the PC, as I’m on mobile atm
a7d62d No.14635369
>>14635182
I'll be honest, pirating assets like that wouldn't feel right at all.
2949a3 No.14635384
>>14635269
I’d say it’s pretty much what scales the best for the ongoing paradigm shift in graphics API that is starting to become widely available (due to compute capable GPUs now being cheap, available for multiple years, and that the gaming audience are by a wide margin more likely than not to have a compute capable GPU).
This is due to that compute API being relatively new in the graphics world, and the algorithms utilizing it are slowly emerging faster and faster (paredo distribution is forming); as it’s the “bleeding edge” atm.
It’s also worth extrapolating that the “future” goto rendering approach has more than likely not been discovered yet; as at this point the compute functionality offered by GPUs is laughably underutilized.
So, from my perspective it’s the switch from fixed function pipeline, to a more general compute SIMD pipeline in terms of what “is the future”; which offers a lot broader range of approaches, traditional but parallelized data structures, and so on.
So, w/e approach that scales the best there, and can have the brunt of its work offloaded to the GPU in a massively parallel general computing (GPGPU) manner; will rise to the top.
The only approaches that I’ve researched that do so are ray tracing, or various SVO rendering algorithms; which due to their laughably parallel nature, scale pretty well. Although those are still early, and overall aren’t production ready; while offering a boat load of issues that still need to be solved.
So it’s really “up in the air” as it were.
9db2b0 No.14635395
>>14635269
The idea of raytracing as "the future" is on its own, not on top of other techniques. Although they really mean pathtracing when they say that.
And just to give a similar analysis for it as you gave about forward/deferred, it has pretty much free normal transparency, doesn't require any AA at all (it has it "built in" because traces back hits to the camera rarely go through the center of a single pixel), and doesn't give a shit if lights are dynamic or static for efficiency, but can't have too many lights (as lights are what determine the amount of rays) and dynamic ones (much like anything animated) would invalidate the buffer each frame. Which wouldn't be a problem with those temporal guessing algorithms they want to use now since they only require a single frame instead of convergence over time.
2949a3 No.14635442
>>14635369
Kek, the only ones that look worth the time to even download are the free ones.
2949a3 No.14635468
>>14635384
Minor correction
So, from my perspective it’s the switch from fixed function pipeline, to shader pipeline, and then to a more general compute SIMD pipeline
a7d62d No.14635493
>>14635442
Maybe, if was actually able to create any kind of art.
676de6 No.14635879
Can we have a proper discussion on how to replicate psx graphics using moder tools like blender and Unity?
So far I got that the psx used integer textures, like the psx couldn't do float division for their textures so they're forced to use affine projection rather than perspective, also they used a limited pallethe and low resolution, also they had to use some basic shader like pong with global illumination by default (no lighting like point or other) and used soft shading of their planes, and subsurf division set to smooth the edges.
I'm sure this could be done on blender today and Unity, and I've seen some unity psx shaders.
I want now to make a RE3 background to test my ideas.
fb726b No.14635915
>>14621538
>psx graphixs
>using unity engine
>with modern hardware requirements
>can't use OG PSX to play dat unity game
nibba why
fbfc16 No.14635920
>>14635915
Because a lot of PS1 games have far better art directions than the overproduced modern versions. It's best noticeable in the Spyro and Crash remake trilogies.
6f7199 No.14635926
>>14635879
>hey guys I want to replicate PS1 graphics in this bloated meme engine
This whole notion of "lol I'm going to make retro games in this big fat engine for retarded numales" is fucking retarded and an insult to both oldfag devs and newer devs who still give a shit about optimization. If you want to do it at all, go with some old freetard engine or go the Devil Daggers route and make your own lightweight 3D engine.
676de6 No.14635932
>>14635915
>>14635926
the average pc gamer has pc more powerfull than PS4.
hell, phones have better specs than modern day consoles.
a7d62d No.14635967
>>14635879
What's there to discuss? The specs are out there, there's tons of shaders that do it for you. Nigger just fucking do it.
6f7199 No.14635981
>>14635932
>modern hardware means I can be a sloppy dumbass
This notion is everything wrong with modern game development. If your game looks like a PS1 title and isn't doing anything really computationally expensive it better fucking run on toasters (there's still loads of them around, a good chunk of /agdg/ anons are using them and the world has loads of poorfags willing to buy toaster-friendly titles if they're priced reasonably), otherwise we get shitheaps like Dusk which has system requirements higher than fucking Crysis with nothing to show for it.
2949a3 No.14635990
>>14633698
>What kind of API are we talking about here?
In general for API programming one should separate their program into "levels of abstraction"; this applies very cleanly to game programming.
>Levels of abstraction
Distinct layers of functionality; this is more for overall design, and organizational purposes.
Granulation <- Low level, high level -> Abstraction
Separating "low level" functionality and "high level functionality" via namespaces (this is also borrowing from the "separation of concerns" principle for organizing your levels of abstraction).
So, we're separating functionality into distinct levels, from the a high level to a low level:
High level is like the front-end, so "the game". I use ECS design pattern at this level.
Mid level is like the back-end, so the functionality, e.g. "audio API, rendering API, etc"; separated into their own sections of my namespace.
Low level is: collections, custom types, math library, noise library, and other such things (hiding w/e design pattern I use behind an easy to use generic API, with heavy usage of interfaces); again separated into sections of my namespace.
Utilize the principles of SOLID for all these layers (good vid: https://www.youtube.com/watch?v=eIf3-aDTOOA).
Wiki article is pretty succinct, and gives u the basic idea.
For example, for handling how do layers of abstraction communicate? Use the "dependency inversion principle" from SOLID.
>Do you break down each part of a game, things like try to group all rendering, controls, logic, ai etc?
In essence, yes.
What those all have in common, is that each of those "concerns" (audio = a concern, rendering = a concern) exists on multiple levels of abstraction.
Each level of abstraction for a single "concern" has it's own functionality separated into its own namespace, and uses the principles of SOLID for consistency between levels of abstraction.
As mentioned I also constrain myself to an ECS design pattern for the high level; which in itself offers quite a few benefits.
>Also any books you'd recommend?
A lot of the resources I use are on the dev resources page on the wiki.
Most of the books I've read concerning programming, and game stuff in general are very specific (this is where inspiration for my tech-demo projects came from); like GPU gems, and artificial intelligence for games.
Most of the time I read SIGGRAPH papers/presentations, and watch the plethora of GDC videos available though.
2949a3 No.14635993
>>14635926
>thinks you can't optimize unity
get on my lvl scrub
>>14635932
If you want2 go down the true route of PSX graphics, you can build a custom rendering pipeline in unity with the new API.
Look at the latest unity blogs, or GDC talks on the scriptable rendering pipeline; or the lightweight rendering pipeline.
676de6 No.14635996
>>14635981
I can't wait to get a non toaster so I can use better software than the garbage I have (opengl 2.1)
6f7199 No.14636042
>>14635993
You can optimize it, but there's still a lot of overhead for a game with PS1-era graphics.
>>14635996
Reminds me, I should get my hands on an OpenGL 2.1-era toaster sometime to test Godot's performance when the new OpenGL ES 2 render is finished. Oldest GPUs I have on hand are a GTS 450 and a Radeon HD 6630M further held back by Apple's abysmal OpenGL drivers.
086a50 No.14636060
>>14635990
>>14635993
>gives u
>want2
Anon, please.
676de6 No.14636074
>>14636042
I can't even run newer software, fuck.
7c850e No.14636133
>Not smart enough to get a programming job
>too socially inept to get an easy call center job
I'll be stuck in retail till I'm replaced by robots, won't i
2949a3 No.14636135
>>14636074
hey at least u can shitpost i suppose
5364af No.14636204
>>14636133
Better make that game so you can sell it before that happens.
7a3eb2 No.14636229
>>14636133
>too socially inept for call center
If you're brave enough to buy your own groceries, then you're socially adept enough for them to train you into a call center worker. You should apply if you actually want that job. Call center job did wonders to cure my friend's autism
1f2028 No.14636348
>>14636229
i've tried multiple call centers, any place that has "friendly and social person" as a requirement don't work out
676de6 No.14636679
Is a better idea to simply give up on gamedev and focus on building a single skill.
I'm wondering why someone should give a fuck about small games made by a single guy that are mediocre in everything when they can play games made by hundreds of more talented programmers, artists, musicians than anything I could ever make.
Better focus on a single skill like illustration and make a living from it, maybe I could even become more well know for it than making shitty mediocre games.
d9bd6d No.14636714
5364af No.14636735
>>14636679
>small games made by a single guy that are mediocre in everything
Yeah, why would anyone play something like that…