/v/ - Video Games

>>15078507

Looks like an excellent way to make generic cookie cutter models that all look the same.

>>15078563

so… anime?

>>15079019

But can you make your waifu real in Blender?

>>15079019

sure you can, but this seems faster and easier, especially for the faggots among us that can't do any art and for whom perfecting the topology and anatomy of their waifu model would be endless suffering and failure like me

>>15078563

>make generic cookie cutter models

so we'd be making aidolmaster clone

>>15078946

maybe you're tight

time for endless suffering

>>15079349

Are you gonna add a cute girl?

>>15080102

and xhe is black even, so empowering :^)

>>15079349

>almost a year later

>no Linux build for Speebot

Nigger, if you're too busy or something I could do one for you.

>>15080203

>Only shitty anime that aren't worth your time have characters that all look the same.

>proceeds to post anime character designs by Yoshiyuki "sameface" Sadamoto

>>15080342

You trying to read images files to turn them into levels or something?

>>15080755

How did you manage that?

>>15080766

no? it's a standard surface shader

i'm just trying to make it fade out if it's far enough. so far that works, but for some reason the lighting is bugged

>>15080749

Sadamoto only has sameface on young characters. When it comes to adult characters there's a great variance in his designs. Which is why I posted Evangelion because the adults all have a good variance in face shape.

>>15080773

Pitifully, using

https://web.archive.org/web/20080912171714/http://www.fortunecity.com/skyscraper/windows/364/bmpffrmt.html

and googling behind others who had my problems. Also I'm using c++

>>15080793

I know you're trying to hurt me. What would the advantage using PNG besides compression and better management of space? I'm not concerned about such things right now but I probably will one day. They're both (bmp and png) free to use commercially, right?

>>15078487

Pretty cool algorithm. Maybe resort and combine if you feel like it before building the GL texture and uploading to GPU.

Have you looked into 3D texture atlases? u,v,w tex coordinates, each page has to have the same u,v dimensions, but most gfx cards can handle quite a few w-coord pages.

>>15080877

From SFML:

>The supported image formats are bmp, png, tga, jpg, gif, psd, hdr and pic. Some format options are not supported, like progressive jpeg. If this function fails, the image is left unchanged.

Most file formats should be open these days. At the very least, .gif's copyright lapsed some time last decade, so it's been free for a while. Jpg is too lossy to worry about, but might be nice to support, and bmp, but also .png remains best for games in most cases.

As for audio, .ogg and .wav seem to be preferred. .mp3 is heavily copyrighted, still

>>15080926

Nah, I haven't had a chance to look into 3d textures yet. I used XNA for a long time, but only for 2D, and then as of about a year ago moved into SFML, which is strictly 2D, but way easier to manage resources and stuff, I've found.

>>15080997

You don't want to distribute your 4k photo textures as PNG files, JPG is the best for that purpose.

>>15081049

Personally, I think 4k textures are a dumb meme, so I never looked into them much. As they actually shipped as jpegs? I'm in love with indexed png's, you can make so much toaster stuff

>>15081204

>began to minimally considered

Just put me out of my misery

>In order to implement texture packing, I need to maintain a sorted set of coordinates that are allocated and unallocated

>This means I need a tree of some kind

>Try to implement red-black tree because it's self-balancing and has essentially the same footprint as a regular BST

>Remember there's a trick to saving a ton of pointers and data density by just packing all your nodes into an array and inferring the position

>Stare at screen for 30 minutes because I can't fucking remember how I did it before

https://www.cs.usfca.edu/~galles/visualization/Algorithms.html

Wow someone put this on the wiki

do i really need to know all this complex math shit…

I only know the basics like addition, subtraction, multiplication and division..no quaternions trigometry rocket science

>>15082320

You should at least understand trigonometry (sin/cos ratios) and linear algebra (interpolation). Understanding the logic behind matrix transformations is useful, too, though you don't have to implement it yourself usually.

>>15082320

Depends on what you want to do. Is your game 2D or 3D? Are you using an engine or building it from scratch?

Like >>15082542 said you need to understand trig at least, and I will add that you also need to have a strong regular algebra foundation. You can get away with minimal linear algebra knowledge if you are using an engine and you aren't doing anything complex.

>>15082606

i wanted to make a beat em up/hack and slash but NGMI bros because i dont know the slightest thing about any of that and yeah i'd be using unity

>>15082627

NGMI?

Is it 3d or 2d?

>>15082692

Not Gonna Make It

it's 3D

>>15082696

Alright. The higher you go in dimensions the more math is involved.

If the game is simple enough you could probably get away with only learning trig as you need it.

>>15082706

Nah I'm not even gonna bother, I've failed every single math test throughout high school and we didn't even have any advanced stuff like trigometry, i'll stick to being a motherfucking wageslave cashier

>>15082627

Just like make game

It's all hobbies for fun, anon. Any anon that finishes a game can hardly be called amateur in my opinion, and once you figure some basic shit out, you probably won't need these threads, except to show off.

>>15082729

>>15082721

>>15082721

As a cashier does money counting come easy or do you struggle and count on your fingers? Failing math tests in high school doesn't necessarily mean you don't have an aptitude for the subject. I had a parade of shitty teachers and even though I didn't grasp a lot of shit the balance and logic of math made a natural sense to me. Studying on my own it comes easily enough, especially when applying it instantly to a project, seeing the results, tweaking the values and seeing the results. Unity is a good environment for such experiments. Do as >>15082706 says and study these subjects as you need them. Don't get bogged down in what you can't do, utilize what you are capable of while improving what you aren't. Also there's hundreds of youtube fuckers doing unity tutorials and if you watch enough someone usually someone expresses the stuff in a way that breaks through your ignorance – speaking from experience.

>>15082807

idk i just use a calculator but i do have to use my fingers if i try and do it without a calculator

>>15082807

Same with me. I struggled until university where I had some excellent instructors. I actually failed the calculus entry test but the head of the mathematics department let me study and retake it. My trig really sucked. Doing projects on the side that used it really helped me get up to speed.

>>15082813

No one will come into your house and keep you from using a calculator, besides that is a lot of what programming languages do for you, calculate things. You just need to know the proper formula or which measurement to use in what case. You don't even have to memorize these, just be aware of their existence and know which book or website to reach for at the right time. The rules and logic that govern them are the most important aspects and you're likely somewhat familiar with these being that you live on earth and have maybe kicked a ball once in your life. It may not be as hard as you think. I'm speaking about 2D games and 3D games through the aid of an engine here. If you were to build a 3d game from scratch you'd need powerful math skills.

>>15082819

Sounds fortunate, anon. What's that you're working on? Looks sort of like the Mako.

>>15082926

You would want three for jumping. One for when they leave the ground, a loop for when they're in the air, and an animation for when they land.'

Have you checked the wiki? http://8agdg.wikidot.com/resources#toc14

>>15081786

You mean like 2-3-4 trees?

>transparent shader can't receive shadows even if the alpha is at 1

kill me

any other way of fading meshes in unity in/out while also having shadows?

>>15084796

Yeah, I think. I'm mostly stumbling through it and hoping it works

>assemble a team

>we're making serious progress

>game is playable online, combat mechanics nailed down

>time to ask for money

>everyone blocks, deletes me or tells me "i don't give money away over the internet"

>survive on cryptoshits trading

but i signed up for this, i'll see my game made someday

>>15086871

Why the fuck would you ask money from your own team members?

>>15086924

i didn't ask my team, i thought it was a given that i asked acquitances for money

>>15086936

I seriously hope you're autistic, because there's no way anyone can seriously believe this.

>>15086936

>>15086871

So basically your acquaintances don't believe in your passion and dream?

Just keep going I say

>>15086936

No, your post very much made it sound like you just lost your entire team.

>>15086871

>>15086936

>asking people for money

At least tell me you have a really good reason for not finding the dosh yourself.

>>15086936

Did you at least tell them you'd pay them back in full with interest?

>>15087308

>with interest?

shoo shoo, jew

Shit, I need a convention for naming and organizing private methods/fields. Previously I just used m_someVarName but it feels wonky giving that same convention to statics and methods

>>15082914

Yeah it's the mako ripped out of ME1. I'm just messing around with a planetary exploration thing and getting the driving feeling right. I'm modelling my own rover model in blender, much larger. with interior. I'd like to make a game about salvaging wrecked ships and abandoned power plants, refineries.

>>15088180

I use some_function() for private and anotherFunction() for public. don't bother with the rage. I already know it's a degenerate style.

>>15086871

You should ask for loans, not free money. Even from close relatives.

>>15088408

>Rent from sister and her husband

>Le sexual assault face, she moves out

>Nice guy but I think he just got too drunk and frustrated at money and took it out on her unfairly

>Still rent from him

>Offered to buy her a $300 bed for her kid, using my work discount, so the new place is furnished

>Her job screwed her at the same time because she was an "independent contractor" at a small local media company and the owner of it quit and decided not to pay anyone

>Had to spot her another $2,000 so she had rent money

>She filed a civil suit so she can get backpayment that is owed

Shit sucks for her. I'm borderline NEET and just saving what little bits here and there while I work part time. Sage for the blogpost. tldr, family

>>15089356

Say "Rabbit rabbit rabbit" because it's the start of a new month?

Is your game inclusive to the RGBA+ community?

>>15089378

shoo shoo evil rabbi jew

>>15089378

I'd repost immunity dog for paranoid anons but I haven't saved it.

Anyone know how to handle drums/percussion?

https://www.dropbox.com/s/d3g48yo9ed5hutn/howdoidodrums.wav?dl=0

https://www.dropbox.com/s/hnbl9uxez7c7m6i/howdoidodrums.mid?dl=0

>>15089514

Just whack them with a stick.

>>15090875

And now it's working. Asking for help really does magically fix any problem I have.

>>15089458

Funny, the version I saw didn't have the save requirement.

>>15089365

Where's she buried, Tewinigger? Also Hong Meiling is a good-for-nothing dirty whore who would gladly get fucked by bald old men.

>>15091111

Nice quads my dude, but remember this, Google is your friend.

>>15091111

Nvm I figured it out.

>>15091441

>Google is your friend.

Google wasn't helping because the explanations where convoluted. At least the ones I got.

>>15091111

The way I understand it is that the inverse matrix for the bone transform converts your vertex coordinate from world space to the bone's local coord sys so that you can apply the bone's local animation transform to it.

I'm assuming you are talking about the bone's inverse bind-pose matrix.

>>15091462

I think you have multiple misunderstandings.

>>15091510

I think your comment provides nothing of value.

>>15091513

I don't write in C so I can't advise you well, but the way you phrased your questions leads me to believe you have a weak grasp of some fundamental things.

>>15091462

The camera would either hold a pointer to a position or a copy of the position. If the position changes then it the region of memory it's in doesn't change, only the values in that region. If the region of memory it's stored in changes that can only be done by copying it to another allocated region of memory. You should not worry about performance but about learning at this stage. You're already writing a faster program than C#, etc.

>>15091523

Yeah well, unless you're going to mention anything specific at all, I'm not going to gain anything from that comment.

I know how to use C but I don't know how to design things in it because I'm used to everything being abstracted object systems that magically justwerk.

>>15091462

Your objects shouldn't be jumping around in memory tbh. Sometimes people do it to keep arrays packed for performance, but then the cost of updating pointers, or accepting that you can't refer to them directly kind of follows along (this is also why people warn about optimizing too early - optimizations involve making assumptions, which puts constraints on everything else you can do, including other optimizations). In your case you could just have the object have a reference back to the camera so it could update the camera when it's removed.

>I know allocating data randomly and referring to it through pointers is very bad for performance, but I'm having difficulty designing things in another way.

Then do it in a way that works and worry about optimizing it later.

>>15091462

Stop thinking in abstractions.

>But if that something is in an array, it's position may change frequently

It's position in the array? Why would it do that?

>how do I link a sprite(s) with an object so they can move together?

Draw the sprite at the specified position?

I think that other anon has a point when he says you have multiple misunderstandings.

>>15091462

If the things to be followed all had an ID besides just their index in the array, you could go through your array and find the right one. You'd likely only need to do this once per frame at most, so it'd be a negligible cost. Also, this sort of thing might come in handy later when you want to get any specific item. Going by array indexes alone can work if your set of items is pretty static or only grows.

Alternatively, you might make sure to always update your camera when the thing that's followed moves.

Or have the camera be just a struct of coordinates and have the followed thing update it instead. If it has a particular flag set (or just a camera address defined as something other than nullptr), whenever that thing gets processed, it should update the camera position. Why does the camera need to track something? You won't be making a new camera or moving it around in memory that often, right?

Followed thing jumps around in the array? Doesn't matter, the followed thing does the updating.

Followed thing dies? Camera just doesn't get updated anymore.

You might want to share how you're organizing your data and general game loop.

The empirical "array of things" probably shouldn't see that much unpredictable movement.

>>15091462

>But if that something is in an array, it's position may change frequently

That's not how pointers and arrays work, in practice. If you have something a camera follows, it's just about always going to be a struct of some kind, and you would make an array of those an array of pointers to structs rather than just an array of the structs themselves, especially if they're going to be changing order a lot. As such, the actual pointer to the struct is never going to change.

> like how do I link a sprite(s) with an object so they can move together? I know allocating data randomly and referring to it through pointers is very bad for performance

You're most likely going to have a lot more "objects" than sprites, so you just create an array of sprites, and instead of giving the "object" an int to the index of the sprite in the array instead of a pointer

>>15091542

Referring to your original post then: pointers are how you do references in C, and there are no objects unless you mean instances of structs. You probably don't want to tie your render and game logic together too tightly so be careful of these references between game entities and rendering entities that you mention.

>allocating data randomly

You don't do it randomly, you do it for data that needs to outlive the scope it was created in.

>>15091539

>You should not worry about performance but about learning at this stage

I've already made a fairly extensive game engine in a scripting language, and understand how to use C. I'm just not used to designing things in ways that are more "data-oriented", so I'm not sure what approach I should take for things.

>In your case you could just have the object have a reference back to the camera so it could update the camera when it's removed.

But that means literally everything needs to have an array of pointers to things that might possibly be following it.

>>15091565

>It's position in the array? Why would it do that?

Like if you add a new sprite, you want to order them so that GL state changes need to happen less often, and maybe so that transparency can be rendered properly.

>Draw the sprite at the specified position?

At what specified position? Where do you get that position and how do you update it when the object it's related to moves? The pointer that points from the object to sprite (or vice versa) will break if one of them moves within the array they're in, unless everyone has pointers pointing to everyone else they're related to, at which point you're chasing a million pointers and might as well use something shit like a linked list.

>>15091583

>be careful of these references between game entities and rendering entities

Almost every sprite needs to be tied to something, otherwise they won't move as the object moves.

>>15091624

But you can do it without tying them in the data, instead tie them in the render logic which observes the game state or thereabouts

>>15091462

When you have references that may get invalidated, you either force to holder to check its validation anytime it wants to use it or use the observer pattern so whenever something gets deleted it notifies all objects that should be interested.

Using pointers isn't bad nor is it very bad for performance, it has its uses.

You mentioned two very different cases in your post, generally you will have one and only one camera (or at least active camera), so it's easy to update it, when you need to do so only for one or few objects, then you can refer to them directly, and you wouldn't care about performance in this case.

>>15091705

I think you need to grapple with your fundamental abstractions before you concern yourself with the performance of something you haven't written yet

>>15091767

Presumably nodes would have an annotation saying how big their largest free subarea is.

>>15091791

Oh that makes sense. I never thought about that, since I was trying to keep my nodes as structs (C#) so they would be contiguous in memory and as small as possible (eg I don't want to attach a XYWH rectangle to every node because that's silly, since intermediate nodes have no data)

>>15091767

You want a rectangles packing algorithm.

http://cgi.csc.liv.ac.uk/~epa/surveyhtml.html

https://arxiv.org/ftp/arxiv/papers/1402/1402.0557.pdf

https://www.codeproject.com/Articles/210979/Fast-optimizing-rectangle-packing-algorithm-for-bu

>>15091813

I've been following this page, but I'll give those a look

http://blackpawn.com/texts/lightmaps/default.html

>go to 4/agdg/

>everyone's using unity or some premade engine that they don't even understand

>they ask for help with fucking syntax

>go to 8/agdg/

>people are actually taking the time to understand how games function and implement every aspect of a game at the lowest level

If you ever feel unmotivated, just remember where other people are compared to where you're at. It's going to get better from here, but only if you put in the work.

>>15091977

As far as I know halfchan has a lot more projects going at the same time with more progress on them as well though.

Except they get dropped after a while anyway so it doesn't count for much.

>>15091977

I used to visit cuck/agdg/ several years ago. It was an absolute shithole, and the metafag circlejerking was batshit insane with multiple trannys and tripfags ruining every other thread. It was a terrible, terrible place that sapped all my interest in game dev.

>>15092004

You can avoid most of the poz because most of those types hate the low level stuff with a passion.

>>15092004

The whole reason I left 4chan in the first place was opening an /agdg/ thread and scrolling past the first hundred posts without seeing anything on-topic.

>>15092797

Most companies don't want to waste the time and effort and bad publicity chasing random people down; therefore, if you use said assets strictly for personal education, and ensure that you don't publish them with eg a code repository, you should be fine.

>>15092825

I think I might end up trying to sell this game. That's why I'm worried.

>>15092705

That's cool, is it for a sprite based system or will the textures be used for 3D?

>>15092797

>>15092828

Twist them to the point they can't be recognized by their creators

>>15092831

Unless I'm mistaken, it should work with either, but then I've never worked with 3D textures before, nor does SFML support them, therefore it will likely just be for sprite games.

Before, I had created a few 512x512 spritesheets for my puzzle game, but wasn't sure how to approach the file structure, because of things like animations, or not having everything fit on one image. After playing Factorio and seeing how they approached it, I decided to also do texture atlasing as well; I'll load all my small textures of random size (which happen to be powers of 2), stitch up one or more atlas pages, and remap the texture coordinates on startup.

>>15091453

If I multiplied a vertex by it's joints inverseBindMatrix, the vertex should be at 0,0,0 + its offset from the joint. Right?

>>15092828

Use them as placeholders and once your game is close to finish you can use your progress to attract artists to swap the assets with.

>>15091977

I went to cuck/agdg/ once and I was mad at how much they promoted laziness. Whenever people lamented at not being good at coding the anons there would unironically recommend the drag n drag feature of Game Maker Studio. Nothing good will ever come out of cuck/agdg/

>>15092940

Yes, that's right. offset from the bone in its default pose. I think. I really haven't figured out skinning all the way through yet.

>>15092859

you're system should work no problem for textures for 3D objects too. It's a pretty nice system you've made.

I have an atlas with textures for all the small props in my convenience store model. I placed them all by hand. was time consuming. a routine to load the texture, place it in the atlas, and change the tex coordinates would have been really nice.

>>15089640

insightful

>>15093466

In 3D, you'd use a view frustum. In 2D, a bounding box. In both cases you'd use a space partitioning like quad trees to remove the bulk of the objects first, then omit anything that doesn't intersect the camera bounds

>>15093725

>Step 1: Draw a triangle

>Step 2: Draw the rest of the game

>>15093741

Drawing objects is a snowballing process. It takes a long time to get to the triangle, but once you get there, things start falling into place.

>>15091610

Looks good. Are you using a graphics toolkit for those UIs or did you write your own?

>>15093904

I've never seen any of those youtube channels, but I'm not a big fan of content aggregaters on youtube as a general rule. It usually shows someone doing something for the money rather than out of passion.

>>15094076

Seems like you're mixing up your unsigned and signed integers, friendo.

For signed integers, -1 < 151.

For unsigned integers, -1 > 151. This is because -1 translates to the highest unsigned integer possible, which is 2^32 - 1 for 32 bit integers.

>>15094095

Then how do I force the compiler to use a signed integer? Because this is the loop:

for (signed int xpx=-weight; xpx<w+weight; xpx++)

>>15094076

Post loop.

>>15094105

By default integers are signed, so putting "signed" in front of an integer does absolutely nothing. When your variable type is "unsigned int" or just "unsigned" it is an unsigned integer.

What most likely screws you over is the fact that you are performing operations or comparisons with an unsigned integer somewhere. If you perform an operation or comparison between two different types, one of the types gets transformed depending on which types are used. If you are comparing a same-size signed integer to same-size unsigned integer, then the signed integers becomes an unsigned integer before the comparison happens.

In your code, xpx is definitely signed. But if w or weight are unsigned, then the comparison in the loop is unsigned too.

>>15094115

Width and weight were both unsigned since they'll never need to be negative. I changed them to signed and it works.

Thanks, I had no idea it would do that. Weird that I've never ran into this before.

>>15094114

Is weight itself signed?

>>15094114

>>15094120

I also want to note that you shouldn't assume printf prints things correctly. %i always prints a signed integer, so if you toss a float or unsigned in there, it'll end up writing something different than you'd expect.

You should also pass -Wall -Wextra to your compiler so it vomits warnings at you. Many compilers give warnings for comparing signed to unsigned integers, since its something you should generally never do. (And in the case you do need it, an explicit typecast improves readability and makes the warning go away.)

>>15093741

>>15093767

Indeed, the first steps are the slowest.

>>15093904

I have never seen them and don't care even remotely, why do you?

>>15094127

Don't forget -Wpedantic -Werror

>>15094114

Lang/engine?

>>15094467

Those flags are useless though. -Wpendantic is only for when you want your code to be 100% language standard compliant, and -Werror is just to stop retards from ignoring compiler warnings.

>>15094518

C

>>15094115

>>15094127

Really useful and well-written responses.

Also, you can explicitly cast types. This can maybe make things a bit clearer, and are at least good experiments to help understand variable types and when the compiler is automatically casting types for you.

float y = 10.3f;
int x = 3;
int z = 5;
float p = (float) x / (float) z + y;
float r = x / z + y;
printf( "p: %f, r: %f\n", p, r );

>>15091441

>Also Hong Meiling is a good-for-nothing dirty whore who would gladly get fucked by bald old men.

*Meiling Hong

also delete yourself

>>15094597

Yes, it was a simple joke

>>15095509

Storage closet for cleaning supplies or a laundry room.

>>15095515

oohhh, true, almost forgot the laundry room

should i put the bathrooms underground then?

>>15095509

You could put the bathrooms where the pink room is. But then you run into "Well, what goes where the bathrooms were?" Maybe an attic? You could have it locked and the player finds a key somewhere and have it be the "Dev room/concept/scrapped content closet"

>>15095519

Outside, three small onsen (men, women, mixed)

>>15095509

storage and laundry. dumb waiters up to the rooms from the kitchen. laundry chutes from rooms to laundry level. A 5 story tavern? I like how you think. Maybe make the top half a conservatory or games room.

>>15095523

>>15095530

>>15095534

thanks for the feedback, will work more on it and post progress later

>>15094463

because people should read tcrf and not give jewtubers that don't even do their own work money?

>>15095626

> It would seem really (((unfair))) if he didn't.

So am i just being retarded, or can you not put constants into a struct? I'm using C.

>>15095626

no idea

>>15095993

Would that even work? C doesn't have constructors, so you can't set it. It'd be whatever the memory spot it allocated had.

>>15096025

OOP can turn into spaghetti code if abused. You shouldn't cling onto buzzwords just because people say it's the correct way. You should figure out what your goals are and pick the tool based off that. 80-90% of the time OOP works, though.

>>15096025

Regardless of if you learn it or not, you shouldn't "follow" any particular paradigm. They all have their own elements of good and bad, and you're only putting cement into your boots by subscribing to one of them and disregarding the rest.

>>15095993

I think you had to leave the constants declared but uninitialized, and then give them values during initialization.

Unless you do some memory fuckery to bypass the constness.

typedef struct {
  const int dicks;
  const unsigned fucks;
} Fuck;

Fuck my_fuck = { 2, 0 };

printf("%d dicks, %u fucks\n", my_fuck.dicks, my_fuck.fucks);

2 dicks, 0 fucks

>>15096049

>>15096121

>I think you had to leave the constants declared but uninitialized, and then give them values during initialization.

nah, doesnt't work, compiler complains about assigning to const.

How much of a difference to using const would it make to just not ever manipulate a variable after it was initialized?

>>15096204

> I want to stop being a pleb and actually learn professional practices instead of some retard who was super stubborn for many years and refused to read any theory books because I thought I was too good for them.

Books are shit watch conferences. Game devs show up to CppCon occasionally and other conventions. There's also GDC but it's 90% hardcore hard-left SJW pozz, and 10% people who actually make games

>>15096241

>How much of a difference to using const would it make to just not ever manipulate a variable after it was initialized?

Compiler might not know and not be able to micro-optimize around it and you might change it by accident. There's always const_cast but I don't know if it's in C.

>>15096263

None. I usually realize I fucked up hard somewhere in my engine and start over.

>Compiler might not know and not be able to micro-optimize around it

I guess I'll just have to deal with it, regarding structs. But I'll continue using const wherever else i can.

> you might change it by accident.

I'll try to not be a dumbass.

>There's always const_cast but I don't know if it's in C.

Appearently it's C++.

>>15095993

Yes, you can use the const type qualifier. You must initialize and declare the variable at the same time. You cannot modify a const variable at any point during runtime.

You may be trying to calculate a value during runtime and store it in a "constant" variable as part of your struct? If so, just use a standard, non-const variable to store it. Give it some name that suggests it's a constant and don't change its value.

Don't try to get tricky and declare a variable const and then change it during runtime using a pointer to the const variable. The results are undefined by the C spec I think, which probably means random crash.

>>15096280

>>15096412

If we're going there then there's like 4 people in the entirety of 8chan's /agdg/ who are allowed to give any advice.

>>15096426

Well I was crashing and burning a ton until I started watching conferences and reading articles. The OpenGL Super bible, which is the defacto book for OpenGL graphics, didn't do shit for me.

>>15096025

There's no reason not to learn it if you want to be a better software designer. You don't have to employ it in every piece of software you write, but it's good to understand how other software designers have chosen to solve problems you run up against. You should learn functional programming as well, the more design tools are at your disposal the better. Again, you don't have to believe the things that fanatics of these systems espouse, but you should be able to leverage them when you want in the ways that you want.

Is it possible to immediately exit an ienumerator without a frame advancing? Specifically, I want to go through a number of checks, then if certain conditions are met, wait a number of frames. If the conditions aren't met, I want to exit immediately. yield break exits, but it still has to advance a single frame.

pseudocode

public IEnumerator BeginningOfTurn () {
	yield return Check_A ();
	yield return Check_B ();
	yield return Check_C ();
}

public IEnumerator Check_A () {
	if (Var_A == true) {
		do a thing
		wait half a second
	} else {
		exit immediately
	}
}

public IEnumerator Check_B () {
	if (Var_B == true) {
		do a thing
		wait half a second
	} else {
		exit immediately
	}
}

public IEnumerator Check_C () {
	if (Var_C == true) {
		do a thing
		wait half a second
	} else {
		exit immediately
	}
}

>>15096426

You could try "Design Patterns" or the "Effective C++" series. If you want C, you might try using the imgui library and reading it's source, as well as anything written by stb. Or just look for some of the major C open source projects and check out how they organize and access their data.

>>15096572

Report and ignore.

>>15096572

What did I miss

>>15096695

Do you know that the IEnumerators in

>>15096505

aren't for collections right? Heavily simplifying, the compiler wraps the yield return blah statements in a state machine, and when the function is called It generates an IEnumerator that starts at the beginning of the function and picks up from the last place it yield return'd was called when MoveNext() is called.

Unity uses them to handle timing, and they can yield return an object that describes how long to wait before invoking. In Unity, they will ALWAYS wait for the next frame if they are yield return'd from, since the structure of Unity's event loop only updates Coroutines once per frame.

You can restructure your logic a bit, if you actually need to run 3 different timed behaviors, you'd want to actually kick off 3 IEnumerators.

If you want to kick off one timeline that just waits and does 0-3 things when it runs, you'd want to just kick off one with one check.

I have no idea if this will get formatted correctly or not, but I hope illustrates how you should structure shit with Coroutines/IEnumerators.

// Mode 1, one enumerator, does 3 things.

// Same wait time for each.

public IEnumerator BeginningOfTurn() {

yield return new WaitForSeconds(.5f);

// If they're just booleans, just fucking check them

// if (varA == true) is fucking redundant

if (flagA) { DoThingA(); }

if (flagB) { DoThingB(); }

if (flagC) { DoThingC(); }

}

// Somewhere else:

StartCoroutine("BeginningOfTurn");

// Mode 2, different enumerators, kicked off at the same time

// This would allow you to have each have custom timing

public IEnumerator ThingA() {

yield return new WaitForSeconds(.5f);

if (flagA) { DoThingA1(); }

yield return new WaitForSeconds(.5f);

if (flagA) { DoThingA2(); }

}

public IEnumerator ThingB() {

yield return new WaitForSeconds(1.5f);

if (flagB) { DoThingB(); }

}

public IEnumerator ThingC() {

yield return new WaitForSeconds(2.5f);

if (flagC) { DoThingC(); }

}

// Somewhere else:

StartCoroutine("ThingA");

StartCoroutine("ThingB");

StartCoroutine("ThingC");

>>15096788

I didn't fully understand the question, no. I just had a long day

>>15096788

They're different things and the flags aren't bools, I just threw together the pseudocode quickly. The issue I'm running into is if the flag isn't set, I don't want there to be any delay at all. For a specific, practical example: healing and damage over time.

First, I'd check the healing over time. If it's >0, yield to a coroutine that runs some code for that and waits half a second or so. Then check damage over time. If that's >0, yield to a different coroutine that runs some other code and then waits half a second or so.

Now, with that specific example, I could check the healing/damage values in advance and only yield to the coroutines if they're >0, but in other cases the checks are more complicated and that might not be feasible.

>>15096842

If you're checking it in one place, it's feasable to check it anywhere, Scripting in Unity is entirely single threaded.

You probably want to change your scripting from being mostly Coroutines to being hardly Coroutines, if you need to check shit, make a function that returns a bool and check that

How do I into monospace? Could never figure it out. Oh well.

public struct HealOverTimeData {

public float healAmt;

public float tickRate;

public int ticks;

}

public bool CanBeHealed() {

return curHealth < maxHealth && curHealth > 0;

}

public IEnumerator HealOverTime(HealOverTimeData data) {

float healed = 0;

for (int i = 0; i < data.ticks; i++) {

curHealth += data.healAmt;

healed += data.healAmt;

if (!CanBeHealed()) {

if (curHealth > 0) { curHealth = maxHealth; }

break;

}

if (healed > data.healAmt) { break; }

yield return new WaitForSeconds(data.tickRate);

}

}

public void StartHealOverTime(float healAmt, float tickRate, int ticks) {

if (CanBeHealed()) {

HealOverTimeData data = new HealOverTimeData();

data.healAmt = healAmt;

data.tickRate = tickRate;

data.ticks = ticks;

StartCoroutine("HealOverTime", data);

}

}

>>15097348

Soon you'll be moving mountains, anon.

>>15097984

Looks a lot like zombie panic source.

>Think you know things

<don't know things

>>15098134

cute as usual, but why does her hair explode?

>>15098162

Bad collision placement.w

I did it. I packed textures into a data file and loaded them in SFML. Holy shit it's tough being a brainlet. Now it's time to learn to draw!

>>15097602

Nice, what kind of mod are you aiming to make?

>>15098141

that's been my experience in looking for a job for the last 2 years

don't worry, they never pick you even if you're actually good at what you're applying for

>>15098726

>I could get by with just SDL's internal use of OpenGL, but I don't like how it doesn't facilitate any kind of batching

Exact same thing happened to me. And now look what it made me do. It took me over 1000 lines to draw a fucking triangle all because I just wanted some efficient rendering of lots of objects. I'm probably going to have to change my career path to justify building this engine. ARE YOU HAPPY NOW, SDL?

>>15099127

That's interesting, I really enjoyed FEAR and the engine definitely deserves more love. I'm looking forward to whatever you do with it

>>15099127

Are you going to try porting the half-life levels? Some of them would be great for FEAR.

>>15099292

Firealpaca somehow manages to have an even worse text tool than Krita. Medibang is just firealpaca with shekel scrubbing schemes added to it (uninstalled as soon as it asked me to "log in" to change fonts).

>>15093725

jusus, why was it so hard

>>15099292

>GIMP is clumsy

You mean you're clumsy with GIMP? It's as straightforward as photoshop once you know where everything is

>>15099794

No it fucking isn't, not even close. One of these days I'm going to make a list of everything that's wrong with gimp so I can just shove it down the throat of anyone who tries to claim that gimp is as good or fluent as photoshop.

>>15099892

Neat.

Are those dynamic or is that a prefab?

>>15099892

Looks like potential.

>>15098444

What, you made an arbitrary RGBA array and passed it from disk into an SFML texture?

>>15099967

Does this require a "bullethole" texture on every surface?

>>15100156

So there's bullet "objects" that get instanced around a wall?

>>15100278

yes, exactly. I did that beforw with deferred decals, now I plan to use both

>>15100575

The lack of punctuation bothers me more than it should.

>>15100668

Have you seen the Rainbow Six Siege GDC talk on how they do bullet holes?

>>15099815

Well until that point, keep your baseless whining to yourself

>>15100715

I did. but I'll have to re-watch it. My approach is a bit different, that's for sure

>>15100715

tl;dr ?

>>15100354

I think what you want to do is just stretch the vertices. Flow maps might be difficult to make with procedural vertices. Basically just keep the same speed texture through out all and just move the vertices so a water face traverses greater distance and looks like water is flowing faster. It's gonna look choppy if you have that few of vertices. Use a lot more. Still confused what you want to do with your tool.

>>15100751

>baseless

Almost as baseless as the claim that GIMP can compare to PS.

How about the fact that every time you paste something, you have to click the new layer button to have it fucking go into the canvas? And after you do, the layer boundaries fixed and you can't draw outside the pasted area until you go into the menu and fix it yourself. In fact if you move any layer the same applies, you have to constantly fix the pixel area manually. And if you expand the area then the transform tool boundaries will cover the entire canvas rather than the actual pixels. And if you ever wondered why the eraser doesn't work, it's because GIMP also requires you to manually add a transparency channel to some layers you add. I haven't experienced it myself but I'd guess you have to add other channels if you paste a grayscale image or something.

The problem with explaining what's wrong with GIMP, ignoring the fact that it would take fucking forever and I don't have enough dedication to GIMP to do that, is that I'd have to make a video to properly show what the problem is. It has many similar features and capabilities, but they're just completely inconvenient in comparison to the way Photoshop accomplishes the same and/or gives a much worse result. You can't make a bullet point that says "x feature is worse quality" because gimpshitters will instantly start crying about it and claiming it's subjective, but there's so many of those that I'd need to either make a video or fill a wiki page with them. Furthermore many of the problems are "this action performs worse", but to properly understand why it's a problem you need to see just how laggy and shitty the performance in GIMP is when doing certain things and how it gets in the way of what you're doing.

So I finished an implementation of a max heap, which will be rolled into my engine soon. I tried to make it formal and follow the same patterns the base C# library has. Anyone feel like skimming it and seeing if I fucked anything up anywhere?

https://hastebin.com/dijelojibi.cs

>>15099398

OpenGL has a fuckton of global state, which is the cause of many issues. (lingering state & drivers have different default state)

Vulkan says "FUCK GLOBAL STATE" and requires you to create your own objects for every part of the state and fill them with the values you want.

Turns out that there's a lot more state in a graphics than people expected.

>>15102629

Well, to be fair, there isn't much global state left in GL 4.5 core profile.

>>15102205

Of course it's not as intuitive, one of them has been endlessly focus tested by a company that cares only about money and user interface. I've never had problems with anything you just tried to describe, and overcoming the idiosyncracies takes very little time. I'm not arguing it's the same user experience, but it's very easy to surpass the differences. You're really reaching for things to complain about, maybe you should try getting better?

>>15103519

Why are you getting so flustered over someone shittalking GIMP? It's a garbage tool with an even worse UI and there's no reason for it to exist while anyone can safely pirate photoshop.

>>15103528

>there's no reason for it to exist while anyone can safely pirate photoshop

Photoshop is even fatter and doesn't have a native Linux build.

>>15103543

>no linux build

I guess the fat retards in Hawaiian shirts will have to use their windows emulator.

>>15103528

I'm explaining how it easily serves the same purpose despite the UI. Why are you flustered that someone might find a tool just as capable as your precious bloated shitware?

>>15103559

Linux users can safely be ignored, as they don't actually have any reason to exist.

>>15103566

But I run windows 7 so you can just admit you have nothing to contribute to this argument

>>15103519

>endlessly focus tested by a company that cares only about money and user interface

It doesn't take "endless focus testing" to realize that you should have a tool to draw the 2 most common geometric shapes known to man. They seem to have this weird mentality that gimp is gimp and if doing something makes it more similar to photoshop then they shouldn't do it even if it's a good idea.

You can take a look at Krita for proof, as far as user experience goes it does almost everything better than GIMP. The problem with Krita however is that many of the features seem half assed or buggy somehow, and it doesn't have as many features because "muh painting program doesn't need them!".

There's something seriously wrong with most FOSS developers.

>>15102572

Out of interest is there a reason why C# doesn't have a standard implementation? Even python has a package for it.

>>15103602

There's something seriously wrong with users who expect open source software to match the usability of aging proprietary kikeware.

>>15103602

Get a plug-in, retard. There exist plug-ins for everything you want to do.

That's one of the great things about more open software.

>>15103718

I'm not expecting either of those programs to match Photoshop's performance or have the more advanced features, but there's so many glaringly obvious problems with equally obvious solutions that it makes me wonder what the hell the developers are thinking. I mean how hard can it be to make the text layer go next to the layer you had selected instead the top of the list? That's been in Krita ever since the shitty new text tool was added, and still as of the latest update a week ago was not fixed.

What makes it worse is that Photoshop isn't perfect either, I have a laundry list of shit that's wrong with Photoshop too, for instance how it throws popups into your face for trivial things ("the layer you clicked is locked so you can't move it!") and then you have to stop what you're doing and click that fucking box before you can continue. Photoshop is much less likely to fix these grievances because it's made by a kikefat corporation and the money will keep rolling regardless of if they listen to those dumb goys beneath their castle, which makes it all the more frustrating that GIMP and Krita still have so many problems that would take me a weekend to fix were they my software.

>>15103816

Your problems illegitimate.

Here's a real problem with GIMP:

If you make text italics and it overlaps the right border of the text box it gets cut off instead of moving to next line.

>>15103602

>>15103816

An image editing tool is a very large project and Krita's core developer team is both small and paid peanuts outside yearly crowdfunding events, so it shouldn't surprise you when they don't have the resources to magically improve every feature at once.

>which makes it all the more frustrating that GIMP and Krita still have so many problems that would take me a weekend to fix were they my software

The source code is open for a reason, if you want something so badly and claim you could implement it yourself there's nothing stopping you.

>>15103877

Do you have something in mind for this "unique fighting system" or are these empty placeholder buzzwords?

>>15103889

If it was implemented right, it may change how most developers approach combat systems in the future. It's that good of an idea.

>>15103883

If the GIMP team wanted a fucking triangle selection tool, they would make one. Turns out, it's utterly useless when you can accomplish the same thing in 5 different ways AND there's probably 5 popular plug-ins for geometric shapes.

It's purely idiotic to make these kinds of complaints.

>>15103906

To be fair, Gimp's dev team is notoriously slow.

>>15103842

>If you make text italics

That's related to another, more subtle thing that Krita does correctly while Gimp does not. Gimp shows all the font styles in the font list, so it shows Open Sans like 8 times (13 if you count Condensed) instead of putting the styles into a separate box.

>>15103883

I understand that it's "difficult", but most of the problems I have are things that you could fix within a day, some in less than an hour. They're also features that are at front and center of the program, so you should fix those first instead of adding some obscure animation features like the Krita devs are doing.

>The source code is open for a reason

You just finished saying how big and complicated the programs are, and now you're telling me to learn the source code so I could fix them myself. I WOULD fix them if I was a developer, but I don't want to dedicate weeks/months learning some codebase that is most likely complete trash from the standpoint of my preferences, especially when most of the problems are so simple that it's way more work for me to learn the internals than for them to just fix the problem.

>>15103906

>mods will fix it

Last time I tried to install a plugin, I ended up with an error dialog appearing every time the program starts. Needless to say the feature didn't work, and I doubt someone will pick it up and update it to be compatible with a recent version anytime soon.

I'm working on a mod and there is a constant variable (static const) in the engine I'd like to change from within the code of a game (as in game that runs on top of the engine, not the whole game). Source code for the game is available and later on has been released for the engine as well, but I'm not touching the engine. I can change its value with a memory manipulation program, like Cheatengine and it seems to work as it should, so I know it's address and name. Game is in C++.

>>15103877

>>15103899

I also worry about being too ambitious. I think the only way to know for sure is to try and risk failure.

>>15103877

You'll fail so many times that at some point you'll step back, wonder why you keep failing, and realize you don't like games anymore

>>15104052

That's the biggest problem with FOSS, I think. Any criticism is deflected with "well, if you're so smart, YOU fix it". Instead of fixing features and trying to make their software more appealing to more people, they just say "no, YOU fix it for US". It's like "mods will fix it", but imagine if Todd Howard came out and said that alongside the retarded fans.

>>15104052

>you need to focus on specific issues if you want to actually get shit done

>it's easy to understand how a small team could be overwhelmed by bugfixes and implementing shit

Yeah, that's why they're adding animation features that 12 people use instead of fixing the features that nearly everyone uses. You're not supposed to start adding new features before you've finished the first one.

>patience

I checked for you and the Krita text layer problem, the one that would take literally 5-10 minutes to fix unless you have a brain tumor or the program was made by yanderedev or something, has existed for 3 and a half months now. Unless I'm mistaken, the original shitty text tool was replaced with an equally shitty text tool at 4.0 which was released back in March, they released the tool before it was finished and haven't updated it at all since. It still baffles me why they decided to put the text tool into a popup window instead of a docker like everything else, but that's besides the point.

>spend ages gathering the knowledge to write your own graphics editor

I made one before I knew shit about programming properly, I've even used it for some sprite work since there's nothing else that has some of the features it has. Admittedly it was on a very high level language and only reached complexity akin to paint.net (from what I know of that program anyway), but it really isn't as hard to make one as you think. The hard parts would be some of the filters, performance of pixel operations, and the math for things like warps/transforms. But both Gimp and Krita already have those parts mostly under control, almost none of the problems I have with Gimp/Krita are related to those.

>>15103602

Open Source doesn't have to listen to their users because they are not customers. There is no incentive to give a fuck what the users think. So of course they are essentially shit. No polish. No consideration of real-world workflows of their users. Gimp is shit because it lacks polish and has no consistent workflow or even a thought to typical use cases.

Krita came along and it was by default better. Why? Because it is based on Qt and not the Dumbest GUI Ever Written: GTK. It also has paid users. So it will rapidly improve in areas that matter, unlike every other open source project ever. Incentive = progress.

>>15104113

>Any criticism is deflected with "well, if you're so smart, YOU fix it"

In this case, it's a valid point because anon is bitching about a tool that's intended for replacement soon-ish last I checked. If he really wants a quick fix for something like this, he should either file a feature request through https://krita.org/en/get-involved/suggestions/ or look into doing it himself.

>>15103945

variables with the const type modifier cannot be modified at runtime.

You found it's memory address and modified it directly. You can do this in code _although the behaviour is actually undefined_ by using a pointer to the variable and casting it appropriately. This may cause random crashes.

>>15103424

Regions are for my own sake to organize code. Years ago I used to have a region per method signature and it was awful. By personal convention, I put m_ before private fields and s_ before static fields. I want the public facing stuff to be nice and organized, with a minimum amount of clutter on the backend.

The spacing is enforced by VS

>>15103694

Not sure. Theres SortedList and SortedSet but they need unique keys. Theres a bag class somewhere but it doesnt order things around. I could probably get by with an IComparer class per thing I want to sort but thats OOP cancer imo. A general implementation I can reuse is much sexier

>>15104224

>The spacing is enforced by VS

No it's not.

>>15104113

>"no, YOU fix it for US"

It's more like "YOU fix it for YOURSELF and maybe everyone else if they have he same issue as you".

>"It's like 'mods will fix it', but imagine if Todd Howard came out and said that alongside the retarded fans"

If the game was completely free and it's source code was available, that phrase would actually make sense,

sage for off-topic

>>15104325

>innovative

I wouldn't say so. When taking a step back and looking from further away, really, all I'm trying to do is take a genre that's been disappointing and has only regressed into gimmicks since its inception, and it's best execution is only "good". It's a shame. Often times, I'll think I've stumbled upon a creative idea and then find that it's been done before.

Basically, rather than being original, since I seem to be too stupid for that, I'm just trying to do it better than anyone else. and fail, at least in the public's eye, since they seem to worship the mediocrity that's available.

before you ask for any specifics, I do post actual progress when I have it, which is quite seldom. give me a couple years. I work on this stuff every day, but that doesn't mean I know what I'm doing, or how I should be doing it.

>>15104325

The reason why we use the same set of mechanics every time is because they are often the most simplest solution to a problem we want to solve. Changing mechanics for the sake of it, without understanding the actual problem those particular mechanics were meant to solve to begin with, is not going to work.

>>15104352

>before you ask for any specifics

Please explain your innovative combat system so I can steal it, become successful and turn you into the new Tetradev.

>>15104174

You mean something like

const int x = 3;
int *i = (int*)( &x );
*i=1;

or

const int x = 3;
const int *i = &x;
int *n = const_cast<int*>(i);
*n=1;	

Both of those work, however when I use the variable that I need to change I get an exception right away.

Currently reading about dll injection. It could work, right?

>>15104415

Yeah, that's what I meant. If you aren't going to modify the source code of the library you are using, DLL injection is your best bet.

>>15104408

>innovative combat system

<You can jump on enemies to kill them.

<Some enemies require one than more jump. I call them "bosses".

<There's little actual focus on combat because combat in platformers is silly and it's most bosses in platformers are shit.

How's that?

>>15104363

>Imagine if the tabs in Firefox were broken so you have to go to some menu to adjust the tabs every time you open a new one, and this problem persisted for months or years with no fix in sight. Then you go to /tech/ and tell people how annoying it is, and they tell you "it's ok because Firefox is open source, stop complaining and fix it yourself!".

quit being a lazy bum if you want to solve your problems.

>>15104442

Sorry, I confused you with another anon.

I haven't considered implementing enemy hopping for my game though. Maybe I'll do just that!

>>15104441

Got it, thanks.

>>15104552

Having to defend your stockpile of cigars, Skittles and Arizona Fruit Juice Cocktail (Watermelon) from personae non gratae sounds pretty comfy.

>>15104224

Putting directives for your IDE into your source code is cancer and another example of Visual Studio sucking donkey dick

>>15105795

Then how can i collapse arbitrary chunks of otherwise random code?

>>15104113

>he doesn't patch FOSS projects he cares about.

You're a pathetic leech. FOSS is driven by generosity yet you aren't willing to give. People like you go on IRC or the forums and request features or bugfixes, but you should know damn well that you haven't done anything for the project to deserve having your specific problem fixed by someone else. Fix it yourself.

>>15107442

Tard

GLTF is an abomination.

>>15107473

Aussie cunt

>>15107442

Or, I could pirate a far more functional paid program instead and be happy. I don't owe FOSS shit, it's the lead developers' job to entice me to use their program. It's like a landlord trying to sell me an apartment that's run down, then telling me that I should fix it on my own and appreciate the generosity of him offering it to me.

>>15107523

>It's like a landlord trying to sell me an apartment that's run down, then telling me that I should fix it on my own and appreciate the generosity of him offering it to me.

Surely you can spot the several massive gaping holes in this analogy.

>>15107523

>Sell

You mean provide to you and others for free so that you may hopefully participate in a community project to build said house. It's fine though, there are lots of intellectually inferior people who can't figure out how to program or even use plug-ins.

>>15107520

That's pronounced "OZ-EE", not "ORSS-SEE"

>>15107535

>You mean provide to you and others for free so that you may hopefully participate in a community project to build said house.

Why would I waste my time with that, again? To keep this analogy going, I could get a nice apartment with everything working right out, just set up and move in, or I could move in into a dingy shithole where I have to do my own plumbing and heating while some other faggot walks in once a month to wallpaper a single wall, and then demand I praise him for it. And they cost the same, which in this case is free.

Why should I choose the shithole, again?

>>15107566

That's okay. Intellectually inferior people tend not to be able to make appropriate analogies either.

You're not moving into a shithole and having to do your own plumbing. You're moving in to a home that has everything you need and more but you just don't like certain things and feel like additions could be made to make it a better home for you because you used to pay for that nice apartment that you're now used to. In reality, you don't realised that this home actually has 3 different ways to easily do what you want inherent in the program. You also refuse to use the additions other people have made and are offering to give to you for free and demand that they be in everyone's version of their house despite the fact that most members of the community think it's redundant or don't like it.

>>15107566

>To keep this analogy going

That analogy is so fucking shit though.

>>15107535

It's pronounced ow-see or oh-see.

Why is GMS2's room editor such a pain in the balls to use? I can place tiles by left-clicking and delete them by right-clicking. But if I want to add an object, I have to hold down alt and then left-click, and if I want to delete objects I have to find a way to select the ones I want gone and then press the del key. Years ago I used GM 5 and I'm pretty sure that was easier to use.

>>15107722

>GM5-6

>Place an object in the room editor

>It fucking stacks them constantly if you don't pay attention

>Okay shit delete them and try again

>Completely delete all rooms from my game and redo them

>The object ID counter is still at whatever it was before

>This means that for very long-lived games, there is going to be permanent asset bloat under the hood as development times increase

>Told on the forums that it's a "direct x issue"

>>15107745

Ouch. Actually I don't really miss GM5 because I'm pretty sure whenever music looped it froze the game for a split second (I used MIDI files). I thought maybe it was the computer we had at the time but it still happens today when I run my old games on my current PC. I just figured the room editing process would have greatly improved by now.

>>15107830

>this is why I moved from hell to purgatory

>>15107863

And then I moved from XNA to SFML, so nonexistence

>>15107874

Why do people use SFML instead of SDL? And why not just use Godot?

>>15107942

Well I had a period of about 4 years of nodev learning experience with XNA, and while I enjoyed the way the engine handled things, I never liked the way it handled preloading assets and it became cumbersome to do certain things in a lightweight way. MonoGame needs a more recent version of C# than I have, which means I have to upgrade from VS2010, which actually requires me to update to a later version of Windows to get the more recent ones, so fuck that. And while FNA is a faithful recreation of XNA, it still has the same fundamental issues I had with XNA to begin with.

I like SFML because it's very lightweight and I can load assets without too much thought, but otherwise like learning how the underlying stuff works myself.

I tried Godot, but it felt clunky to use the node/scene system. I'm sure it works fine, but I don't want to fight against a workflow

Has anyone here watched TheBennyBox's 3D engine tutorials?

I want to blow my fucking brains out when listening to him. I have never in my life heard someone sound so fucking infuriating.

>>15108050

I don't know who that is or what you're talking about, but exclusively from your description I know you're talking about a subhuman with an incredibly fake high pitched voice.

>Make a general Game object type that serves as the main heartbeat of the game

>Have a separate project to test things, with a bunch of clutter in a Game subclass because I was testing my partitioning, but don't want to delete it

>Realize I can just make another subclass of Game that's basically a completely separate sandbox that I can blow away as soon as I'm done with it

Oh shit this is great

>>15108160

If it's invalid and at the bottom of the map, then how are you viewing it? Somethings weird

>>15107942

SFML is better organized and integrated. SDL is OK though too. I used it for a few years.

>>15108189

>130lb tatted up women mouthing off and acting aggressive in a streetfight.

It's a dangerous game. One day someone is going to add the final stamp to that pussy pass. Hope she gets lucky and lives through the punch to learn a valuable lesson.

>>15108313

Are you using OpenGL? It's a bitch like that, you might have to shift things by like 1 or 0.5 pixels.

>>15108189

BTW, that first webm, he probably talks like that because he was trained to sound non-threatening and non-masculine around a male authority figure. That's why he's speaking through his nose and not using his diaphragm. Essentially groomed by the adults in his young life into an adult beta-male. I fucking see it everywhere and it's disgusting. It's not always the kid's fault though. They may have tried to fight it but been mentally and/or physically crushed for it. I see men and women do it to their kids. Disgusting shit.

>>15108363

SFML, which is using OpenGL

But it's weird because quads and stuff work just fine. Lines and points get fucky. But that's fine, if it's a known thing I won't worry about it, since it's on their implementation, and nothing to do with my engine's fault. Plus it's a thing I won't be taking advantage of in production anyways.

>>15108367

Yeah, but the moist mouth noises, the random pauses, the nasally AND breathy voice, the spit swallowing, the FUCKING RANDOM PAUSES, the non-confrontational fake nice guy voice….

Programming OpenGL truly is the worst. You never know if your buffer index is slightly off, your matrix is wrong, or some other shit.

>>15106838

Ideally your IDE should recognize where to collapse based on syntax, if it doesn't you're using a shit one. And any IDE that would encourage you change your source for it is a shit one

>>15108710

And all it cost was the integrity of your code

>>15108715

>Organized, readable code that is easy to find things

>Bad

>>15108733

The formatting it atrocious, man, I don't know how you put in so much work into making it less readable. try throwing some lambdas in there to top it off.

>>15108397

yeah, you're right. he has a lot of other things going on that make it unlistenable.

>>15108421

Is the coordinate thing something SFML does? OpenGL coordinates are lower-left = 0,0 for textures. Also I think GL clip space (the screen with identity set as ortho projection matrix) is [-1.0,1.0] with 0,0 at the center.

If SFML is doing some rendering-to-texture for 2D lines, boxes, sprites, it may have the sampling setup weird on the texture (not using GL_NEAREST). I'm not sure. I haven't dug that deep into SFML.

Yeah OpenGL is a pain in the ass. Adding a layer on top of it helps some things, hurts in other ways though.

>>15078478

Has anyone here tried the Godot 3 tutorial serieses (???) on Udemy? Are either worth buying?

>>15108993

> Are either worth buying?

Nothing is worth buying when you have cgpeers.

>>15104552

>convenience store simulator

That could be fun.

https://1d4chan.org/wiki/Night_Shift

>>15109006

cgpeers?

>>15109036

Thanks, lots of good ideas in those thread archives. I wasn't really considering going the surreal or supernatural route. I might though. Maybe 2-3 nights with a unique event each could make an OK minigame to put out.

>>15109093

yes, cgpeers, try putting it into a different box (preferably one that leads to a search engine)

>>15109010

>>15109036

Reminds me of an idea an anon came up for a horror exploration game set in an abandoned mall.

>>15109094

Having the supernatural events be rare would be better I think.

>>15109494

I agree. Otherwise it probably loses its impact. I'm not making a full game though, just 1-3 events in a single location. This all started as a stress-test for my rendering system and physics system, and as a way for me to get better at Blender. I'll probably release it for free and ask users to help me by providing runtime statistics and crash dumps.

>>15109273

That would be an interesting setting. I like the abandoned mall videos. I'd have to use PS2-era detail though or I'd never finish the modelling and texturing.