Amateur AGDG General

Procrastinate all weekend edition
Resources:
>>>/agdg/
>>>/vm/

Previously:

Other urls found in this thread:

mrbook.org/blog/tutorials/make/
bitbucket.org/scons/scons/wiki/SconsVsOtherBuildTools
mingw.org/
metamod-p.sourceforge.net/cross-compiling.on.windows.for.linux.html
truevalhalla.com/blog/list-of-html5-game-publishers-sponsors
allfearthesentinel.net/download?file=moon_man_v1_3_1.pk3
godotengine.org/article/godots-new-renderer-progress-report-2
forum.smokin-guns.org/viewtopic.php?f=32&t=1523
pastebin.com/benU20Cd
twitter.com/SFWRedditGifs

algebra related, i asked in last thread
i've got the 2x2 matrix A = (6 11, 0 6) and i need to find A^71
as far as i understood, the usual way to do this would be to find the eigenvalues+eigenvectors, and from them find the diagonal version of A=B, and then just calculate B^71, since it's easy to calculate when it's diagonal
but for my case, i'm only getting 1 eigenvector (x, 0)
does that mean that i can't make find a diagonal version of A?
in which case, do i actually have to calculate A^71 like in the last two lines of pic related?

ehh not today, gotta make some textures and shieeet. What I learned now is that it is actually possible to make a texture overlay by using the fill layer that should really help with recreating the texture I try to match the new camo pattern to the old one as close as possible including the smaller details and the previous fuckup shouldn't happen anymore.

To all of your questions: You're absolutely correct. Save yourself some time and bring A into Jordan normal form first though & look into binomic series (if you already had that), or use a computer to calculate the partial steps. Calculating this by hand would be hell.

Yeah, you don't have enough eigenvectors for a basis.
That said, calculating A^71 should be p.easy if you scale it by 1/6 first.

Nevermind, do that instead.

i already ran it through a calculator once, and the numbers were huge so i figured i wouldn't bother with it
i'm just making sure i'm not missing out on some extremely easy way to do it

Posting this again:
Can anyone lend me some advice with 2D cameras/sprite rendering?
The way I handle things when the camera is centered on the player is by making an adjustment to the offset equal to the change in position.
It was working just fine while I had things scaled according to 1080p, but when I set up a smaller internal resolution for easily scaling it started to mess up.

At different points rounding the numbers into integers turns out unfavorably, having the change of entity position be 1 off from the change in offset. This results in the entity jittering spasticly when it should be still.
I'm really lost as to how I can fix this right now, so help is appreciated.

It's probably still doing its math according to 0.00=0 and 1.00=1080

What program is that user?

Substance Painter.
I wonder why he uses that for tiled textures, considering that's exactly what Substance Designer is there for. Still, if it works for him…

Thanks user. I'll look into both.

I shilled for Substance Live here a while ago. 19$/month and totally worth it. Very friendly license terms, too. You got to keep it after you paid them off, unlike Photoshop where you need to pay for as long as you use them.

Sounds really nice. Now I just have to git gud enough to where I actually have a reason to make textures

In 2D a camera can be represented as a box that covers part of the map's space.
Similarly, the player character can also be represented with a box.
(A box is just an x,y coordinate paired with some width/height values.)

So, assuming you represent the player's position as a point, and you draw the sprite centered on that point…to determine the camera position and where to draw the player sprite relative to camera:

pbox = new Box(); //player boxcbox = new Box(); //camera boxcbox.w = SCREEN_WIDTH;cbox.h = SCREEN_HEIGHT;pbox.w = playerSprite.w;pbox.h = playerSprite.h;cbox.x = playerPosition.x - (SCREEN_WIDTH/2);cbox.y = playerPosition.y - (SCREEN_HEIGHT/2);pbox.x = playerPosition.x - (playerPosition.w/2);pbox.y = playerPosition.y - (playerPosition.h/2);

Given this untested pseudocode doesn't have any bugs in it, you'd then just draw the player at the position specified by pbox with the width and height specified by pbox.

If the player position you store is actually the corner point of the sprite you draw, you just have to make some small adjustments to the box calculations:

cbox.x = playerPosition.x + (pbox.w/2) - (SCREEN_WIDTH/2);cbox.y = playerPosition.y + (pbox.h/2) - (SCREEN_HEIGHT/2);pbox.x = playerPosition.x;pbox.y = playerPosition.y;

Ugh I hope there's no bugs in that I'm pretty tired right now and didn't test it at all, pls forgive if I fucked up

hmm, is it possible to make a new material in Substance Designer and import them to Substance Painter?

That's the point of it.
Designer is there to create the materials you paint with in Painter. You can also bypass Painter and directly use Designer -> Engine for prodedural stuff.

Ahhh shoot I already know I fucked up.
For pbox.x pbox.y, you need to take the calculated camera x and y position into account:

pbox.x = (playerPosition.x - (playerPosition.w/2)) - cbox.x;
pbox.y = (playerPosition.y - (playerPosition.h/2)) - cbox.h;

If I made any other mistakes than that then just go ahead and gas me.

Alright sounds cool.

I got my hands on Quixel Suite and god damn it's insane how powerful this thing is.

So let's say I'm generating a mesh made of quads, 9 vertices wide by 9 vertices tall (or 8x8 quads)

Should the mesh be made so that the 9th vertice is at X=1.00, or can I use some other value (such as 9, or a different scale altogether)? I've always thought it shouldn't matter because you can use a world transform matrix to scale and position it properly, but I don't want a lot of random units and scales. The reason why I wouldn't use 0..1, is that if I jammed 100 vertices into that range, floats are somewhat inaccurate as fractions increase, right?

I was gonna check out /vm/ because I want to work on a mod, but that board seems to be one foot in the grave already.
So I'll ask here, does anybody know of some decent resources for modding minecraft? Last I checked, the most popular version was 1.7.10.

While that did point out something redundant I did, that essentially how I have it working.
But something else just occurred to me thanks to you.
I was doing this:
int PlayerDrawPosX = (int)(EntityX + 0.5) - (EntityW / 2) + (int)(TileMap->Offset->X + 0.5);

And I realized that maybe I could eliminate the quirkiness by rounding the value into an int separately instead of at the same time:
real32 TestX = EntityX - EntityW / 2 + TileMap->Offset->X;
int PlayerDrawPosX = (int)(TestX + 0.5);

And that instantly removed a lot of the jerkiness. It's not totally fixed yet though so I'm going to check what else I can adjust.
Thanks man.

How do I get over stage fright, /agdg/?
It's been several years since I last finished a game and I've been looking into/tinkering around with various freetard engines in hopes of finishing one again, but even though I finally feel ready to work on something again I keep stalling or hesitating as I always used to. I don't have the excuse of "I should know more about X before working on Y" anymore, I just want to get this shit out of my head even if I don't sell most of it.

Any good learning materials for dummies on cmake and makefiles in general?

You have too many expectations for your game. Just make something intentionally shitty.

I've done that multiple times before and people still loved the results. It's confusing and not very helpful if you want constructive criticism.

JUST DO IT
Pick something that just doesn't take an absurd amount of foreknowledge and learn while you go. Actually working on a project really helps you along when it comes to learning the related techniques.

Other people may not say this, but I would recommend you avoid cmake as much as possible.

For Makefiles, something like this ought to get you going.
mrbook.org/blog/tutorials/make/

The point here is disabling your internal quality control if it's overzealous. Make first and judge once it's done instead of judging first and being too scared to make anything.

I shouldn't have spend too much time on painting at every damn pixels, meh at least I will be hopefully faster on my next one. Damn it is really worth it through it would probably take me forever getting a similar result with Krita alone especially that damn thing has hardly any tutorials when it comes making a gud texture such as making metal looking like metal and not like concrete.

i thought cmake was a good way to achieve cross platforim compiling

Is the user here who was searching for testers? I wrote you mails since, but you disappeared, so this i just to check if it's still going on or what.

Are you looking to cross-build for other platforms, or (cross-)build on different platforms? If you're the only person who will ever compile your code, you're free to do whatever you prefer to get the job done.

CMake is OK for a user trying to build your program, but it can be a serious pain in the arse for developers/maintainers. Windows programmers tend to prefer it since it'll play nice with Visual Studio.

I personally just use a plain-old makefile for private coding projects, and autotools if I'm distributing the source code to the public.

Autotools isn't perfect either, and there has been a long debate over what build system works the best. Here's a good overview on the subject:
bitbucket.org/scons/scons/wiki/SconsVsOtherBuildTools

yes but this doesn't really become significant until numbers get really high
if you are that worried about you could also just use doubles

ok so i'm not sure what you are using cmake for but the whole thing can be pretty confusing so ill just give you some knowledge
first of all cmake isn't a compilation tool it doesn't compile libraries for you it creates Makefiles
These makefiles are then used by your compiler to actually compile the libraries
secondly most libraries will have a CMakeLists.txt file which tells CMake what to do
thirdly once you run cmake once it will create a cache file of the settings used you can fine tune the settings in the file and the next you run CMake it will use the adjusted settings
lastly there is a CMake GUI which is a little easier to use than the command line version it's made for windows i believe but there might be a linux version.

I'm sure it will be fine. I'll just have a user-defined scale factor when creating the mesh so it's not arbitrary.

Can't really use doubles, since most of XNA is written around floats

My main goal is to compile (mostly) on windows without using visual studio and its cancerous shit compiler/environment.
I have a working mingw and was hoping to find a way that works on win and linux alike.

well getting vs shit to work on linux might be hard
I can tell you i achieved windows/linux cross compatibility using mingw on windows compiled my libraries with cmake
and clang on linux again also compiled my libraries with cmake
The only library im using though is sfml
though I'm convinced i could probably add sdl to it if i needed to also might try to experiment with adding glut or something if feel like it.

The build system isn't magically make cross-building work for you without the right cross-compiler. MinGW allows the building of Windows binaries from a UNIX environment; you can't build Linux binaries using mingw, only Windows binaries, unless I'm retarded and missed something.
See the basic explanation of MinGW on this page:
mingw.org/

I don't have experience building Linux binaries on a Windows machine, but you should just be able to grab some GCC cross-compiler packages from within cygwin and build Linux binaries using that:
metamod-p.sourceforge.net/cross-compiling.on.windows.for.linux.html

I switched to Linux days after I started learning programming because I saw what a mess developing stuff on a Windows machine was going to be and wanted to learn how to directly use the compile instead of fiddling with it in an abstract manner through the GUI of an IDE.
But PC gamers tend to have trouble making that switch (I was hooked on Nintendo games at the time so that wasn't an issue).

I don't want VS shit to work on linux.

I don't want mingw to build linux binaries.

I want to build my project on windows using mingw because I've fucking had it with microsofts shit ass compiler.
As an added bonus i want to generate makefiles that can be build on linux using make or whatever native unix shit is around.

Sorry for not being clear, the main objective is to move away from VS on windows because it is utter shit.

well what libraries are you using?

r8 my first gaem

Please, tell me the final boss is a giant vagina.

Not quite. A float is basically like a number writen in scientific notation. You know, 1.2645*10^6 == 1264500. Except the base is 2 instead of 10. Some of the bits are used for the (binary) decimal number (called mantisa), some are used for the exponent. The problem arises when the number of significant digits is large. For instance, if you add a number close to zero to a number far from zero, if there are not enough bits in the mantisa to represent the range from smallest to largest the small part will get truncated. So basically, there should be no difference between numbers in the range of 0..1 and 0..1000000. You'll be able to encode roughly the same number of distinct values in the interval. As you move away from zero, it gets more iffy however, since you have a "base" value to encode along with the range you actually use. Either way, none of these concerns are relevant. Even 32-bit floats have so much precision you'll never have to worry about it in anything besides physics engines that work with spaces at least the scale of arma 3. All modern non-workstation graphics cards work exclusively with floats and don't even support doubles (afaik). Unless you need micrometer precision at 100km from world origin, floats will do.

...

I thinking of it in terms of how some fractions can't be easily represented and the loss of precision that comes with it. I remember finding some value with a float using integer multiplication (or division?) and ending up with like 12.000002 no matter what I tried. I just figured that if I have my mesh expressed as 100 points between 0..1 and then scaled it up by 1000 or whatever ended up being needed, then it wouldn't be "exact" and you'd get seams.

Though this is probably a concern for a bigger, more elaborate project.

Retooled a lot of shit and I'm almost at the point where I can "effortlessly" create meshes of quads of any NxM size. Feels good.

In terms of actually using it, probably going to stick to an 8x8 quad mesh for each map tile face. This is 81 verts, 128 tris, 384 indices each (though indices are global and only need to be calculated once). The numbers of meshes I need to define is based on the number of textures I have - one mesh per subimage, and this is then instanced and transformed by a matrix per object/face.

Well fuck my dick, wrong webm

I don't think Edmund McMillen posts in /agdg/ threads.

Supposedly mingw has mingw32-make, which is make for mingw. So if you can write a CMakeLists.txt that can generate a working makefile, you can let mingw use that to build your project for windows, on windows. The same makefile should be usable by GNU make to build on a native linux install or by the same in a cygwin environment to build for linux on a windows host, as describes.

Well, if two verticies have the same position attribute and get the same transform applied to them, they should get the same post-transform position. Either way, I've never seen this happen in practice, so you probably don't have to worry about it. You're technically correct in that this could happen, but I've never seen it.

Yeah, so ultimately it doesn't matter a bit and I just need to just like make game

Good stuff

What you're describing can be a concern, but it isn't really a problem specific to floating point arithmetic. Floats and doubles are actually the most accurate standard type for handling those cases.
The only way to retain total accuracy is to store and work with the results of division in the form of fractional numbers, which Matlab does, for example. But in terms of graphical representation on computers, all coordinate systems are ultimately represented in integers (pixels), which may be part of why no one bothers with fractional numbers in graphics fields.

9:05

...

TBH it sounds to me as if it's in you are heart, not ur head user. Try to organize, enforce some rigor on yourself and justdooeet.

Then just don't use VS or it's compiler user. Just compile w/ GCC or Clang straight on the command prompt.

Actually, if you insist on sticking w/ Windows dev, then this is the proverbial shooting you'reself in the foot. VS is clearly a superior IDE, and by far the best option on Wangblows. It's debugger in particular is second to none.

Having said that, I left Wangblows behind for good over 3 years ago, have zero fucks given, and use g++ and clang basically every day.

I use CMake and the juCi++ editor/IDE for all manner of things and love them both tbh.

For those of you with 2D games that use sprite art, how do you guys handle scaling to different resolutions? And if you only have a single base resolution and scale by integers, what resolution do you pick?
I need to settle on this before I can do some design and I don't know what the best practices are.

How to achieve greatness?

You have t-

original character
DO
NOT
STEAL

Start.

And once you start, don't stop.

Still looking for amateur 3d animators and texture fellows for the as-of-now untitled horror game project. Get at me.

LEARN

love

Hard work and discipline.

Seek to understand why things are the way they are, and utilize this knowledge to gain an edge over your competitors.

Porn game

friendly reminder that if you are using an IDE you are a shit programmer and should stop "programming"

If you aren't writing in asm you are a faggot and should stop "living"

...

Lets see if i can finish the last recipes this weekend
TODO list

Are you making a cooking simulator or is this as aspect of a larger game?
Putting those GW2 icons to good use

...

He's not wrong, Vim + Makefiles are the best. I'm stuck with CodeBlocks and I hate it.

Ask your mom.

Does it really matter if your a shit programmer or not?
I thought this thread was about making games not programming.

Jews did 9/11

Hey I just came in from Reddit, how do you 8channers say it like Jews did 9/11 rite? xDD

Interesting.

Niggers.

It comes with free oven too, not pre-heated through.

that's a nice american tank destroyer

Programming is kinda like dota in a way. It turns into a giant dick measuring contest over which way is the right way to do things, even when it doesn't matter. That being said, the bantz has a place in games considering thanks to Unity a bunch of devs are shitting out half-baked products on steam.

Yes yes indeed it is. :^)

I'm thinking of using Godot for a 2d project but I've heard its performance is dog shit. Anyone used it before and know what its like?

go back

The latter, I think. But it's an important aspect apparently. Was it a HM clone? I can't remember.

Today I simplified my main draw calls, cached the index arrays when creating an NxM mesh so they're all shared and not re-created per texture subimage, and also halved the amount of data points I need to store for my texture atlas (only needed the top left and bottom right points)

What if I use both under different circumstances.

Engine recently added support for 5:4, 17:10, and 21:9 resolutions.

21:9 looks fine, but 5:4 and 17:10 have annoying cut-offs. So I have to re-do all of my fullscreen graphics now.
Guhhhhh.

Finished game will be a quite similar to harvest moon
First expansion will add alchemy and a few extra systems
Second expansion (probably final) will add a single deep dungeon and turn based combat and monster butchering

God, I hate JavaScript. How can a language be such a pain in the ass to deal with? Dynamic typing was a mistake.

Why are you using JavaScript if you are building a game? Why do you do this to yourself?

It's for the editor for a game I'm working on. I'm using the Qt framework for it, to make it platform independent. There are two UI toolkits in Qt. The classical Qt widgets, which is similar to GTK+, Swing, etc. and the "new" Qt Quick.
I'm using the latter, because declarative programming languages are a godsend for user interfaces. It's really intuitive to make UIs that way. There is a problem though. If you really want to separate the UI from the application logic, you will need some form of imperative programming language, too. Doing UI logic like sanity checks for user input or disabling some button when a certain checkbox was ticked and the like. Of course it would be great, if you could do that directly inside your UI code as opposed to adding an entire extra layer to your application. And of course the Qt developers added JavaScript for that.
I haven't checked the statistics, but I assume that my code base consists of ~80-85% C++ and QML, and ~15-20% inline JavaScript. It's probably even less JS. Those 15-20% cause me more headaches than the rest by a large margin.

Oh wow, it also added native bloom support, too.

Probably won't keep it but it's interesting to see how differently things look with it.

...

I'd experiment with it on subtle / low levels and see how it looks and maybe keep it. Could look nice if done right.

UNITY

what's the name of your game again?
not samus something something

Tentative name is the Booty Project, still deciding on the actual name. Alternative name was Not-Samus Space Pirate, or Space Pirate In Space.
Probably be something like On Cosmic Tides or On Shining Tides or On Astral Winds, something like that.

If my experience with xna has taught me anything, it may be because you're switching graphics device context and dynamic render surfaces and textures need to be recreated, but aren't.

tldr: Detect when app loses focus, when it regains it, recreate your UI state

And tone mapping too, wow his OpenGL 3.0 effort is really worth it.
it's fucking nothing.

Graf Zahl is a nigger

This takes place in the editor, not the game, so it's not my fucking problem

I decompiled it while playing just because I was curious and if I recall he was doing quite a bit of threading. Thats probably what is causing your null refs.

Render the scene to a texture, then render that texture to the screen on a quad or something in whatever resolution/aspect you like.

truevalhalla.com/blog/list-of-html5-game-publishers-sponsors

Opinions on this?

I recognize the name, that guy's a cunt that sells shovelware that helps create more shovelware (flappy bird engines) to 12 year olds. He is the mobile game's greatest cancer.

This

Basically pick an aspect ratio like 4:3, or 16:9 because its CURRENT YEAR, then use an appropriate base like 640x360 and upscale it. Or like 960x540

yeah, but he is making money and I want some money.

Shit, I'm having trouble with generating appropriate UV coordinates for some general meshes.


Right now I have a SimpleMesh object that contains the vertices and indices and other rendering data. Maybe make a "ComplexMesh" object that contains several simple meshes, then I can have a method that "paints" a texture to a Simple Mesh?

get fucked pajeet

I'm practicing modelling in low poly for my Quake 2 engine game

What you making in Quake 2?

It's going to be an FPS with rpg elements, hub levels, dialogue options and so forth. I went over some of the details in the previous thread but not a lot.

I'm planning on having a very medievil/Dungeon Keeper aesthetic and tone to the whole thing

Shabbot Shalom fellow goys!

I recently switched to SFML from game maker and while ist been a pain having to make my own sprite sheet class n whatnot its been oddly satisfying.

I'm letting my autism flow through me.

My question is, what is a good way to handle menu's and overlay n stuff? IE: is the best way really to just keep track of game state / player location and adjust accordingly?

Ex: have the overlay move with the player's camera to stay in the correct place?
When user press ESC switch render logic to draw pause menu over the game?

Please finish it, user.
I'll be waiting

Heh jury's still out on that. I'm learning C programming though. I wanna eventually sell the game if I finish it.

My experience is more in modelling and mapping. Like I've mapped in a lot of older engines. So I think creating the visual assets for the game will probably be the easiest part.

What's a good FOSS 3D engine that isn't limited by early 2000 era technology? Looking to make quake-style FPS too.

Sounds groovy as fuck. Kind of reminds me a bit of Strife.

Perhaps one of the newer quake derivatives like Warsow? or Nexuiz


Steal some code from Quake derived games fuck morph target animations.

Well I mean there are other engines that were later released that are significantly more modern. If you want something more modern you'd have to go with something like Godot

Like Doom 3's engine is about as modern of an open source engine you can go that was originally proprietary and stand-alone. I'm not aware of any later ones.

It's coded in C++ and has things like really nice lighting and physics.

I'm actually slightly more inspired by a Quake 2 engine game called Kingpin. It's a Quake 2 engine game that has a semi non-linear progression. It has hub based levels. It also has things like money you can find to visit stores to buy weapons/upgrades for your weapons etc. That's sort of what I'm going for. I'm less interested in adding things like stats since I think it would just slow down the game.

Warsow runs on Qfusion. Which is sort of a hybrid of Quake 2 and Quake 3's engines. To the point where it isn't even able to run Quake 2 anymore because it's so different.

I know about Kingpin, it looked pretty good, but I haven't played it, I'm really hanging out for a low poly first person rpg too.

It has good aspects to it but it's unfortunately really lacking in a lot of areas.

Like it has this really good atmosphere that's dark and grungy. The style is kind of unique in that all of the characters look like 1940s mobster caricatures. The guns are stuff like tommy guns and so forth. It has relatively impressive AI for its time that follows you up ladders and in vents and stuff.

It doesn't play all that well compared to other games that used the engine though. Like Soldier of Fortune plays much better

But I have a lot of ideas on how I could expand and make a game that's like Kingpin but much better.

In the case of idtech you'll want to see what the fans have done with them. ioquake3 cleans up the Quake III codebase a lot, and there's an equivalent for RTCW, Jedi Outcast/Academy and Doom 3 but in varying states of completion.

Alright so what should I do?
ioquake3
Qfusion
darkplaces
kmquake2
quakespasm
What would have the best community support and more modern features?

Good keep us updated, I want to see more, or post your basic design docs or make a blog or something, I'd read it.

I've thought about using Quake 3's engine but the only real benefit it would give me that I would use is skeletal animation. The rest is relatively superfluous since the Quake 2 engine has a really nice aesthetic to it.

That and I'd likely also need to borrow Raven Software's scripting language from Jedi Outcast since iirc the Quake 3 engine's scripting support is fairly poor out of the box. Iirc several devs at the time like Raven and Ritual created their own proprietary scripting technology like Ubertools to get around this.

Solid option if you just want a no frills 3D engine. It's mostly just a cleaned up version of Quake 3's engine. As far as I know this would have the most support out of the ones you've listed. It runs in C++ unlike all of the other engines listed which I think all run in C.

It's like a hybrid of Quake 2's engine. I'm not sure it has any support. It's mostly used for multiplayer FPS's. It's similar to Berserk like that.

This is a popular option for fangames. I've seen a bunch crop up over the years that run in Darkplaces. The one I remember the most off the top of my head is Blood Omnicide. Which wants to remake Legacy of Kain Blood Omen in Darkplaces. They've hacked it together to make the original Quake engine look really really pretty. But it's clearly taken a lot of work for them to do.

Out of the box it comes with some hacked together support for things like skeletal animation. The main reason I don't use it is because it's mostly designed to modify the graphics a lot.

This is what I'll probably end up using. It's a fork of Quake 2 that mostly tries to extend the vanilla limits. It's like Boom for Quake 2. There's this one dev that released two indie games, one that appeared on Steam, on kmquake2. I'm not sure how much support you'd find for it outside of just general Quake 2 engine support

Quakespasm is mostly designed to play the original Quake as vanilla as possible. You'd have to contend with learning Id's frontend for Quake 1's engine, QuakeC. Which they got rid of with Quake 2.

I have written something of a design document but it's mostly all of my ideas in bullet point format right now. I plan to clean it up later.

I have a wordpress blog but it's not finished yet. I don't plan on making it public til I have some more finished assets and I've decided on a name for the project.

I'm in the early stages of making a fps puzzle game in ue4. Made progress on one of the main tools: a recoil gun that launches the player akin to rocket jumping. It can be charged to increase the impulse, but has diminishing returns the longer its held down (no cap, though).

I plan on having it used in tandem with a hookshot. UE4 has some nice support for cables now, so that helps. This is already working out a lot more better than it was when I tried doing this in Unity.

Today's animation progress: Something intended as a walk cycle but looking like a dash so I'll probably use it for both but damn it looks bad, a proto-backdash and a heel kick. Gif's all animations done so far.

this looks like a lot of work

Less than 10 hours so far.

That's not bad, but I think you should spend $2,000 to outsource the hitboxes

Any music fags in here? How did you learn and what program did you use?

I hate that I know exactly what you're getting at.

f

pretty much

why not just use EU4 it's technically not FOSS but the source code is available

So for my Quake 2 project I've been mostly assembling my tools. The modelling program I mostly use is Maya.

Quake 2 uses an old ass format called MD2. It doesn't have skeletal animation data. It's all vertex.

There is a plugin someone created on Github to export in MD2. However it isn't compiled and he didn't provide a binary, just the source code.

So I spent the last two-three hours compiling something from its source code for the first time. And it finally worked after lots of troubleshooting.

Unfortunately the plugin isn't very good compared to a similar plugin for Blender, which even includes optional headings for things like whether you want to export the animation and scale. So instead I'll have to model in Maya and export into Blender which was plan B.

So it was not exactly a very rewarding activity but at least I learned something from it.

I am done with the M10 burger tank destroyer. :^)

...

Tell the prof.

I'd argue it's probably the professor or college's fault to think those group activities aren't going to result in people slacking off, so it seems somewhat iffy telling the professor would change much beyond getting the people in your group mad at you for not letting them slack off.

the camo is so awful.

Awful in what way, does it just look bad as in low quality or as in it is inappropriate?

Why does this happen? I can work kinda fine on workdays when I really should be doing something else.

It probably has to do with the fact that game dev takes a lot of conscious effort until you reach a high level of proficiency (at which point you can put your mind on auto-pilot a lot more, and the process becomes more effortless).
If you become more skilled at whatever part of game dev you're procrastinating on right now, you'll procrastinate less in the long term.

not him , but i assume he is talking about the stretching and that it should be uniform

at least codefags produce working games even though they look like shit.

but games with gameplay don't sell anymore

Is that Reimu?

Couple new ammo pickups and an energy tank.

Maybe it's a "grass is greeener" situation, but feels like artfags have it easier.
Because graphics are the first thing people can judge by when it comes to screenshots/gameplay videos, you can make a game that has a lot of wow and visual pop to it and attract a lot of attention.
Programming isn't an insurmountable obstacle, either, it's really easy to just grab code blocks from other things and slap them together with nails, tape, and chewing gum so that they function–even if it's an abominable beast. Can't do that with art.

I don't let it stop me, though. I like making things, shitty art be damned. You can do it too, user. I believe in you!

meh I couldn't care less, I will make mods for games only when they are FOSS/Open Source and not for proprietary games. Then I don't have to worry about it's accessibility in the future.

t. Linux user

those graphics looks pretty good to me, user

Maybe my shitty crawler will have gameplay soon

Gratz user. You've entered a larger world.

I have a new bad habbit of buying asset packs to make prototypes for them

I guess it's not that wasteful though, really.. I'd rather spend 20$ on an asset pack and have many hours of dev time learning and have fun, than spend 80 canadian shekels on a modern AAA game that lasts 4 hours and sucks shit

Alright, lets say if I wanted to make a tank game, would darkplaces or ioquake3 let me be able to create terrain?

As an artfag, our work should come second. We can make a game look appealing and give it context that will spur the players' imagination, but we can't make it fun (on a base level, at least).
I want to see more interesting prototypes with programmer art so I can make art for them.


On my own, all I can make are 'what ifs'. Like, a fake screenshot of what a game may look like if it were real, which is kind of difficult for me since I have no concrete idea of how the game would or should work.
Essentially, I have no real goal to work towards. I find it a bit easier if I have a 'mold' I can fill.

Artfag here, we just need to start with simple games like sokoban or breakout but with artistic twists.

(checked)
But I want to do art for games where you shoot or stab niggas. Otherwise I can't get excited about my work.

no, you artfags need to partner with me and maek game together

I don't like devving alone, but I appear to be a rarity here

Make it a sokoban where you have to shoot or stab niggas.

There is moon man mod for Skullfuck 2.0+, what about that?

allfearthesentinel.net/download?file=moon_man_v1_3_1.pk3

>I don't like devving alone, but I appear to be a rarity here
Well, you never know how reliable other people are. Also, their taste might be shit.

I also don't like devving alone, but working in a team is another challenge. Keeping communication and momentum up is fucking hard, especially with strangers you've never worked with before.
Also what said.


I didn't mean literal niggas.

What's a good open source 3D engine to make a driving game in?

i'd say godot, but from what i've seen the physics there are iffy
unity isn't open source
not sure if UE4 is
and i have no idea in what state cryengine/lumberyard are right now

Did you poked around with the Grit engine?

UE4 is open source but not free software.
Cryengine is pretty much the same but unlike UE4 you don't have to pay past a certain amount.

That image reminds me, a veterinary clinic that I know the owners of got broken into and the security camera didn't catch the guys face because windows update forced the computer to restart, so the security camera wasn't recording all night long.

I've heard good things about Phaser.

So much for cushy OS eh? I wonder if he installed gahnoo+linux after that.

I am not sure how modern website looks like but have that in return, hue.

This is actually not true. Chinese developers do it all the time, they take sprites or maps from other games and photoshop them a bit until it suits their purposes.

This project is cool. This guy is making a Halo themed Quake I mod.. for the 3DS.

NEW VERSION RELEASED
>>>/agdg/28230
That was a busy week and very productive weekend

That's not necessarily a bad thing.

Bare in mind that if you code something and the project was never finished, it doesn't exactly say very much about your skill as a programmer

Whereas if you make art and models for something and the project was never finished. You can still reuse those assets someplace else. And you can put them on your portfolio for prospective employers.

I think it would be monumentally easier than creating a new game for the 3DS or trying to port the original Halo to it.

Since the original Quake has not only been ported to the original DS, it's been ported to the 3DS too. And from what I've heard they already had some hacked together multiplayer for it.

It's not a bad idea if you're almost incapable of producing visual assets

Both engines make terrain by creating a static block and manipulating the vertexes. You can see a dude do it in this video with Trenchbroom.

The end result looks a little jagged though. If you want smoother terrain like what they have in Source with displacements you'd either have to program a lot of it yourself or model a lot of it separately and import it.

I would dev with others it's more so that I can never trust people to not just leave the project after a few setbacks or completely try to take the project in a new direction.

I worked on a game dev project some time ago and I remember I gave very explicit instructions to the concept artist about what I wanted the main character to look like. And they barely followed it at all. Like I wanted the character to wear a full faced helmet and she drew him with a face.

At this point I'd rather do most of the work on my own and get other people to fill in the gaps for me later.

Not surprising.

Serious question though.
What IDE should I use if I'm an artfag trying to make a 2D action platformer?
I was experimenting with Godot but then I read this
I haven't encountered any issues but I wonder if I'm going to run into unforeseen problems someone here knows about.

I'd just say to use GameMaker™

Need input on a mechanic for my dungeon crawler. Combat can't be paused, will be ATB-based, and mostly automated.

Armor acts as an effective pool of extra hp. You start with 15 hp with 300 hp being "a lot". I like the idea of having item effects like "+25% increased armor", but with values this low, you don't see much of a difference, which led me to work on this system.

Basically, all armor items add to your armor rating, which is a value that does nothing by itself, but is converted to armor based on your class. For example, Plate Mail might add 120 armor rating, while Leather Armor adds 20 armor rating. A warrior might have an armor factor of 20, meaning he'll get +6 armor points (6 effective HP) from wearing the plate mail; wearing the leather armor only gives him +1, which is better than nothing. A mage can't wear plate armor, so is limited to leather armor. However, their armor factor is like 40, which is means they get +0.50 hp, so basically nothing.

Some armor can have innate increases to armor points, though. For example, maybe Plate Mail actually adds 120 armor rating and +10 armor points, so everyone gets at least some benefit from it. Actual damage reduction (either flat or %-based) is going to be rather hard to come by and quite valuable. Shields will be the primary source of it, so make you decide between damage and defense.

Finally, all worn items will have a weight. This affects the charge time of your ATB bar, so caster classes will inherently be faster than warrior classes.

r8 pause screen
F U L L R E S

Its performance is perfectly fine. The biggest flaws of Godot right now are its unfinished 3D tools/renderer which are still very usable.

I'd need to see how it works in action to see if it would be any good. It's one of those things that could be really awkward and make the game lean towards one type of character over others.


This is an example. Since if caster classes are legit just better players will feel like the game is unfair.

A good example is how in Dragon Age Origins mage characters are comically overpowered compared to literally everyone else.

This does tie into the game's lore though and makes you naturally prioritize targets like mages as being the most dangerous. Which is how it logically should be.

A whole new 3D renderer is coming out in 3.0 using OpenGL 3.3. The update is scheduled for early 2017, but you can download the alpha from the glse3 branch on github.
godotengine.org/article/godots-new-renderer-progress-report-2
Godot will look more like a modern engine hopefully with this update.

Sorry, the branch is gles3 and it's far along in its development apparently, more of a beta.

ioquake3 is fantastic - but you're going to be hunting all over the internet to gather enough information on game logic and the content creation pipeline.

ioquake3 is also C.

ImGui is a neat library.

Alright I got a pause screen and chat with text to speech like Moonbase Alpha. Can't record because the software doesn't pick up the voice.

...

It's taken me longer to do the timelapse videos that it did to make the pixels

I just remembered. There is a tool to make terrain for Quake engines

forum.smokin-guns.org/viewtopic.php?f=32&t=1523

This one is a tutorial for how to do it for the game Smokin' Guns. Which is an ioquake3 game.

Whats some good tutorials for someone learning Unity for a 3D game?

7/10 it needs more a e s t h e t i c tbh fam.

If you use OSI's definition of open source, it's not open source either

Hitpoints are awarded when you reach the a new highest level among all classes; the amount you get is much higher with non-caster classes and because of that, it's better to level up with a fighter class first if you want to take a hit. While you can do casters first, you won't have a good foundation of HP, or things to do when your magic runs out, but you'll have really good killing power.

Haven't been spending much time on gamedev lately, I've instead focused on drawing. so I can make a living off furry porn in 2017
The retopo is done though, I'm gonna properly unwrap the UVs and bake the normals right next.
Also experimented with more sculpting workflows and I can get some pretty clean results like pic related.

You gotta do what you gotta do.

...

what's your FA user?

FA just did another witch hunt for cub artists, leaving the popular artists like Oob alone, yet again.

Every time I think about doing dev, I get overflowed with this feeling of dread that I can't program anything worthwhile. I'm not concerned about the art side of things, but I'm completely lost when it comes to programming, I feel like I'm yanderedev-tier shitter without a big engine to make it look like I have half an idea what I'm doing.

For example I need to do pathfinding of some kind, but I don't even know where to begin. I can't read C++ from tutorials, it's incomprehensible to me compared to something like javascript. I also tried to learn to do some kind of vector collision test (think; shooting a laser that stops at walls and enemies) but I cannot find or figure out anything that could do it. Basically it seems like, ignoring performance or prettiness, programming is relatively trivial until I need to implement some kind of algorithm, at which point I hit an impassable brick wall that I can't move past no matter how long I try.

Should I just give up and start doing furry porn commissions until I die of regret? I don't know what to do.

make some ground textures for me then :3

Do you sell bricks and bricks accessorizes?


Reminds me of Sudden Strike 1

I got them from google, I think they're from Doom.

Are you using an engine? Is your game 2d or 3d? I'd program for you senpai.


That's because I stole borrowed the art from sudden strike 2.

Try harder. It won't get easier if you keep ducking.

JUST DO IT, NIGGER.

Try to overcome the feeling of dread. Everyone starts out on the bottom tier and work their way up. Right now, you're just starting so don't get bogged down by all the details.
Trying something and failing is much better than not trying at all, no?

A dev I like said that "What you tried to do is just as important as what you will accomplish so keep trying."

Keep trying, user. Don't give up.

tbh fam I like the aesthetic of Sudden Strike 1 and 2. Well that settles it I guess, what tools did you used for to rip them? Also what are you planing for with your game?

t. Sudden Strike player

Good to know my life isn't that shit. That and not living as a pajeet or as a nigger

That's probably because you're overthinking it. A lot.


You gotta just keep trying.

My advice is to try and program something a little simpler than a game. Like a text editor or an image editor or something. Then go back and see if it's a little easier.

This sort of shit is hard and it's hard work but you gotta keep your chin up and not be too concerned about the final result. The journey becomes the destination or so they say.

There's pretty simple games to be programmed as well. I think there isn't really any advanced programming going on in a grid based dungeon crawling JRPG for example.

I actually could not find the images themselves in the game files so I nigger-rigged it and took screenshots from the editor still better than doing the art myself

The idea is to have a probably turn based game where you control a squad of SOG operatives in the jungles of vietnam/laos/cambodia. Operators will gain increased stats by finishing missions but once they die they are gone for good.
It is like a mix between rainbow six and xcom.

Personally this is a little why I prefer working with premade engines rather than trying to code my own. Since I spend more time tweaking existing code to suit my needs than creating it from scratch. It becomes more fun to test that way.

Y-you too

hmm sounds interesting, I am not much found in turn based games but anyway good luck with your stuff mayne.

Checked out xnormal and it's great, got rid of a lot of annoying artifacts in my bake, but there's still this really harsh change in lighting where there's a seam. Any ideas on how to fix it?


Don't have one actually, is it something really needed, or can I just operate from tumblr and/or twitter, but share my stuff on e621?

I have the opposite problem, whenever I think about gamedev I overwhelm my mind about all the cool shit I could work on and I end up spending a lot of time not doing anything cause I can't decide. I guess just set some realistic goal like 30 minutes of game dev a day and only ever skip a day when you absolutely physically can't work. I did that with "one drawing a day, can even be a 5 second doodle" and it helped me tremendously with discipline over the past year.

honestly yandev tier programming is fine he has made a working game after all. The important thing is that you learn from your mistakes and keep improving also I wouldn't bother with optimization at all worrying about performance is just not useful if you are a beginner

You'll want an FA because people will be more likely to commission you on it. It also has the added benefit of people on there reposting your stuff after they've commissioned you. Which is free advertising.

Do you happen to have a Steam or anything like that user?

I've already encountered both as a few of my many roadblocks

Grid based. Your party's just a position in the grid and you press arrow keys to move it. Every grid cell is passable or impassable, and collision is just saying "can't move there". No algorithms.

Just reveal all grid cells in a radius of X cells, no problem. Doesn't break much. No algorithms.

Make a simpler game if you're hitting complexity roadblocks.

you still need pathfinding for the enemies

That's what I thought, but then I started wondering, what's the point? If I can't do anything interesting at all, why would I be doing it in the first place? A game so simplified that it doesn't need any fancy algorithms is difficult to make interesting, and it's not going to teach me to solve any of the fancy algorithms either. I wouldn't be moving anywhere, just avoiding the problem like I am currently.

I've already been able to do basic game mechanics and stuff for a long time so I don't need to learn it anymore, but I can't come up with any interesting game ideas that don't end up needing path finding or ray tracing or something like that.

The point is to git gud, you nigger.
I, as a drawfag, can't draw masterpiece paintings that I want to draw.
That's why I work on my fundamentals and slowly work my way up until I reach that point.
The same applies to programming.

Experiment. Learn the language and how to use it. You won't be able to make a cool-as-fuck game right now but in half a year, you would've learned enough to build more complex shit. Even if what you make is 'boring', it'll help in the long run.

Random encounters or really fucking stupid AI.

You can't learn to build a house by nailing the same plank over and over. That was my point.

Probably different ID

I guess, though that's where the "experimenting" part comes in.
If that doesn't work, you can always ask others ot give you a helping hand and help you out or something.

I honestly think it is smarter to split the work. Why learn how to do everything? Just learn one thing and have someone else do the other.
The age of polymaths is over.

I can't really give advice on learning algorithms because I'm not sure where your problems with them are. Lots of it is about a mathematical mindset but I can't just tell you work through two semesters worth of university-level math. How did you fail when trying the vector collision test?

That involves having a team, user.
And having a team means knowing people.
And knowing people means you need to be socially competent.

Well you don't have terminal autism do you user?

What part of social interaction gives you trouble?

Good work user

I don't know, there's something about reading other people's code that's incredibly difficult for me. One of the only algorithms made by someone else that I managed to "successfully" implement was perlin-like noise generation, but even in that case part of it made no sense to me so I wrote some of it myself (probably ruining the performance but whatever).

Finding the tutorial.

As far as I understand it ray tracing doesn't solve collision with rectangles and moving objects, as it seems to be inherently based on a grid. But that's all I can find.


I sometimes want to do that. Just focus everything on art and try to find a likeminded programmer to make a game with.

The part where I have to rely on others to get shit done. I'd rather be a 'jack-of-all-trades' and ask assistance from others every once and a while than be part of a team. The only time I can think of "relying on others to do work for you" is when you're part of a studio.

If you're just a hobbyist developer or some faggot with no connections, you're unlikely to find yourself a colleague that shares your same vision and can work with for extended periods of time.

Besides, if you DO know how to make art assets and code, it's a valuable combination if you do eventually want to work for some studio.

Nope. They don't care, you get to pick one job and that's it. And you better be specialized in something too, or you're not interesting.
Unless you're talking about shitty indie companies, then learn to do as much as you can.

This.
I worked summer jobs as software engineer for a big corp and they don't care if you can do more than one thing. You are paid to do x so you better do x.

That doesn't really have anything to do with social competence though.
It's just reality.

Whats my best angle if I really enjoy making voxel models but despise any sort of coding?

KYS?

You gonna burn alright. I fixed the 37mm flamethrower which is also upgradeable by buying the Incendiary shell upgrade, the 2nd upgrade gives the flamethrower a longer range. The 50mm Cannon has now it's Incendiary shell ammo back.

joj

Game status: 90% COMPLETE
Waiting for the last 7 sprites to be done. Then comes all the fucking balancing of stats and spells.

Can you show gameplay?

/agdg/, I need your help. I want to git gud at making vidya, but right now I don't know much other than the basics of object-oriented and structured programming, though I generally have no trouble grasping concepts, and I'm serviceable at best at artfaggotry. I want to start a project in GameMaker as soon as test season is over, but I feel like that won't help me really git gud. Am I being a faggot and should focus entirely on that project, or learn from somewhere else? If so, where do I even start?

It's shit. But it's my shit

Keep it up user. I still have nightmares about your code though.

Don't worry user, the code got much worse.

...

W-E-W
Still, at least you're close to completion which is not something a lot of Anons around here can say about their projects.

You can't even imagine how many conditionals I can nest within loops inside looping conditionals!
I was just lazy and wanted it to Just Werk

I'm won't judge you. At the moment I seem to be too retarded to calculate the AABB of a set of points. It's easy, but for some reason it doesn't werk and I don't know why.

Show code so we can laugh help you

Now I feel strangely smug.


Did you make sure that the width of your boxes is not negative? Took me about a week to figure that out

Fixed it right after posting. I accidentally initialized the xMin variable with negative infinity which meant that no value would ever be smaller.

I remember how much I hate lua now

This is amazing.
No one can claim lack of programming skills are a barrier to making games now.

No shit. I never even used Lua or anything that wasn't Fortran or Perl for the matter before I started making that abomination, I just opened Lua and LOVE reference sheets and started writing shit on gut feeling. It ended up working.

You can change that if FPS to something like
FPS == FPS == 15 and 60 or FPS * 0.5
and be even smugger. You may want to consider stepping through Interface.draworder backwards too (for j = #Interface.draworder, 1, -1) rather than the j = j - 1 workaround depending on how that table is set up.


Honestly I'd much rather people got something out then worry about their code being perfect. It's all a learning experience to carry on to the next project.

Whoops, I meant
FPS = FPS

The best thing about lua is that it is not 0 indexed right?

–Yeah, I love how it does everything slightly differently just for the sake of being special

I enjoyed working with tables in Lua very much. So much I rewrote the table ordering code several times. It's still shit.

Do it. Make the shittiest fucking shmup there ever was. When that's done make a slightly less shitty shmup which you should be able to do because you're not a nigger and learn from mistakes. Repeat until gud.
just like make game

Wait! Are you serious?

Yes.

...

This is probably the ugliest chunk of code in my project so far… Not sure how else to make the enum values cycle through the different values. I can't use +=1, for example, because there are also None, Up, and Down in the enum type and those don't make sense here.

You don't have to make fun of me user.

can i help?

With what?

you can make extension enum functions out of those ifs (they should be switch/case though) so instead of all of that you could just call functions like facing.Turn90Left()

public static Direction Turn90Left(this Direction dir){ switch(dir) { case North: return Direction.West;//etc. etc.

I'm not making fun of you, I just thought it was on topic. My code is organized to the point where it's almost detrimental, since I've redone it so many times, and have a way of doing it now. Even in terms of a playable game you're miles ahead of me. (see )

Yeah that's a good idea.

Feels kind of clunky, since there's not a case for every combination of direction/orientation, but this will let me move around the dungeon without too much mess.

IF YOUR GAME WORKS IT DOESN'T MATTER IF YOUR CODE IS SHIT

dude no don't add an orientation value, do 3 separate functions for 90 left, 90 right, 180

it only matters for yourself and how much time it actually takes to program stuff and fix bugs

Actually Orientation previously existed for making the camera look around, and was also planned for a few spell effects.

you're only making it more complicated but whatever

"if your game works" implies there aren't bugs to fix :^)

You need to fix them first, which takes a lot longer if your code is shit.
And as the old saying goes, one man's finished is another man's beta.

This is Nintendo's way of doing things, really.
They make up for their shitty code with the largest quality-assurance armies in the industry.

But if you want to beat the AAAs, it would be worthwhile to learn how to write non-shitty code. Technology can allow you to do things 1000x better or worse, depending on how well you use it.

What Nintendo games have shitty code? I've never experienced a game-breaking bug on any of their first party games. Hell, with the Wii, you had to pay like a $10,000 licensing fee and if their unit tests crashed the game, you had to do it again.

There's a difference between comfy smug confident in your code quality & readability and xkcd-tier hipster smugness that wastes everyone's time so you can pretend to be smart. If I ever feel like rewriting the FPS changing hack I'd probably use a table of values and cycle through it so it's properly extendable. The joke in my screenshot is that more than half of the lines are still hacks.


tfw I've worked with lua long enough for 1-indexing to feel completely natural

So what?
Maybe it's because I'm a shitty programmer myself, but I consider whether something works and functions a much higher priority to whether something is coded well.

A player is never going to look under the hood, and a player is never going to care as to whether you could have done your code up in a prettier way. What matters is whether your code makes a fun game.

Don't use this as an excuse to skip out on optimization and self-improvement, though. Shitty code can still bog things down quite a lot. But you can always learn more, improve it, and optimize it.

user, I…

That's my point, user.
You can't tell their code is shit unless you were to actually have it to look at.
But the end-product isn't shit.

They test any games, whether made in-house or by 3rd-party, that touch their game consoles to hell and back.
They don't really care how the code is written much, they just check for bugs and meeting certain guidelines for a decent user experience.

But just because some code got some bugfixes, doesn't mean the code(base) isn't shit anymore.
To the player, though, a shitty codebase can be perfectly enjoyable- all that matters is that the end-product 'works' with minimal bugs.
So that's all that Nintendo concerns itself.

What you should take from this, what I'm trying to get at, is that Nintendo and other AAAs waste tons of resources, armies of programmers and testers, to get their games in shape, where you could get the exact same job done with no compromises, with only 1/100th or 1/1000th the size of their armies if you don't write shitty code (at least speaking strictly in terms of programming).

Technology is powerful stuff- in the right hands, you can be extremely efficient, but without that technology, or if you use that tech poorly, you can be insanely inefficient.

I'd explain I was referring to something else but I realized with a lot of modern gamers playing popular games that's actually correct and now I feel bad.

Hold me, user.

Level editor is almost done. Feels so fucking good.

That sounds kinda gay user, just ask your gf to hold you instead.

I'll bop your noodle user.

progress always feels good

That's pretty sweet, user.
Are you using any library for the UI?

Ternary is good, but not the right tool for this job. It'd be not as easy to extend with more FPS options as it should.

If you know about them, that's cool. I've just seen a lot of people not know Lua has them(ish) and do lots of little if/else blocks where they help be tidy.

What're you working on?

see . ID changed. Didn't replace placeholder sprites yet so no ingame footage right now.


It's a nun, ya git.

It looks awful

How bothered about this sort of thing should I be?

Qt (Qt Quick to be precise). I made sure that it also works on Linux.

How much smaller than normal QT is it?
I tried installing QT once and it was like 15gb. Not even once.

It's part of the normal Qt package.
On Windows my entire Qt directory clocks in at 3.5 GB. I just checked. I installed two versions of Qt (5.6 and 5.7), although I only installed the MSVC prebuild binaries (so I don't have to compile it myself), since I don't need all the addons. Besides that I only installed the Qt Creator IDE (which would be the best C/C++ IDE, if it weren't for that shitty debugger UI).
You can select what you want and what you don't during the setup and it comes with a maintenance tool that let's you customize your installation after the fact.

How good do the graphics look?

If you plan on selling it you really need to optimize it.

that's not true and you know it

My mantra is:
Practice (many approaches), learn (as much as you can), experiment (don't be afraid to fail or innovate, but it's one or the other), rinse + repeat.


You sound like you're talking about AABB (axis-aligned bounding boxes), which has to do with intersection tests (between bounding boxes w/various cases for types of collisions), and testing against possible collision cases within the subdivided axis aligned space (usually by an octree for 3D space).
It's usually a matter of finding the correct search keywords, and then you can find tutorials/references a plenty.

Most bits of code implementing a complex algorithm fall under the category of being: "more than the sum of it's parts" (as they have something beyond those segments of code holding it all together, which are fundamental concepts of the algorithm, i.e. problems/ideas(approaches)/solutions, which were used by the programmer in the first place to implement said ideas into code).
Thus, I generally approach any complex algorithm via first doing some background research into the topic, understand it on paper, and being able to think out each step algorithmically (or in otherwords, "what does each step do, why, and how does it get me closer to the goal" - engineer; not: "how to put this into code" - codemonkey).
The code part comes next, and is a breeze if you understand the how, why, and what of the algorithm/concepts itself (in addition to having that fundamental programming knowledge).

>(not meant for games)
well… what'ya know, I mean, who would'a thunk that that would've happened, eh?
kek

If you want your gaijin scum of a game to catch on in glorious Nippon, you must fold your code over 1,000 times and make it run on Windows XP machines from 2007.

Well, fuck. I have no idea how I'm going to correctly "edge" the water between tiles that have it and those that don't.

Since the dungeon is static, maybe precaculate the edge tiles and mess with their geometry?

i don't get how microsoft can make such horrible installers
and i still don't get why i need to deal with license bullshit every few months considering i'm using the free version. i need a fucking account for a compiler.

i don't understand, what's the issue?

Neither of these look correct.

The first one ends abruptly with the water floating in midair, and the second one just kind of exists.

The water is one tile ahead

so just put a box with inverted normals under the water tile, and with missing faces if there's more than 1 water tile

Yeah but the water is transparent

so?

So each box would end up being various amounts of transparency, when they're overlaid on top of each other

Well, this almost works.

Alright, so I got a really stupid idea. What if I added an item that played a fictional tv station that was basically just a bunch of webm memes?

There, I'll just do this.

So does anybody here do VR dev or is it dead in the water?

Your right, that does sound stupid I would find it funny, but then again I don't have a very refined sense of humor.


It is supposed to look so shallow?


I don't have the money to buy a VR headset, and if I did there are other things I would spend it on.

If I want to add more depth to it, I'd have to basically create a "bucket" out of the ground below it, and merge it with other tiles that also have water. No thanks.

Help I accidentally made something beautiful

Yeah I'll just do this, realism be damned.

But lads, move() worked before with the camera in this one project I had, why would I need to use set_pos instead? I don't understand. How would I even use set_pos? As before, I still can't get that working. I'd much rather just use move(), pls upload a project that is much like mine but working:

...

It's like a high point carbine made by even more mentally challenged people.

Just make it murky.

Alright, pajeets, got a simple problem here. I've figured out a very basic solution, of just checking against a bunch of generic cases where it'd fail and returning true if they all pass, but I'd take a lot of work and I want to do something more elegant

pastebin.com/benU20Cd

This is just all around terrible.
Your biggest problem is most likely poor overall design.

Here's your (you)

Yeah okay faggot, enjoy your shitty mess.

calm down, sport

No.
I'm sick and tired of selftaught inbreds thinking they are gonna be the next notch while somehow being even less capable than him.
Do you honestly believe your question was formulated well? Try posting it on SO and see what happens.

:^)

Clearly, something triggered your autism. Do you also get flashbacks?

I do but I'd rather stay anonymous on Holla Forums while I can, since it feels like people give more honest input. If you really wanna find me, then I'll be uploading the horse thing on e621 when I'm done with another drawing in the set.
I'll see how things go in the next few months and maybe make an FA account then. Probably art station, too.

You may be able to just add more layers to your ingredients. Instead of "FoodTypes.CheeseCheddar" Have a food type called Cheese that is a child of Food, and then extend Cheese into Cheddar and Moz, then you just check if a given ingredient is a child of Cheese. However objects may be better for this than enums, because as far as I am aware enum inheritance doesn't work in c#. Anyway, read up on objects, object casting, and object inheritance.
That's the best I can give you with such limited code. If anyone else has anything better, or if this actually makes things worse, let me know. C# isn't my language of choice and I don't have a lot of experience with it. I'm mainly going off my c++ knowledge and hoping they are similar enough I don't fuck everything up.

I've seen you post before and what you described didn't sound like an error, but exactly like what you expect to happen.
Are you sure move() doesn't work and you just don't have anything static in the scene to notice it?
You didn't even describe what you expect to see when it's working.

...

show an example.

Using framerate based timers often causes a lot of problems in such cases for example

(Checked)
And Notch sold out his shit so that Macrohard can grasp their claws on his blocky ass game, not that I care much for his shit, I can't stand games where everything is made out of faux voxels and looks blocky.

Is this the feeling of disappointment fathers have when they see their sons never learning from their mistakes?

Having made money drawing furshit, I don't recommend it unless you're really into it yourself or are desperate and need the money. It can be pretty stressful.
I can give you business tips though if you'd like.

please tell stories of autism as well

None, really. Most trouble I've had is they can't decide on what they want or their character designs are super bad.

...

I'll bully you in the morning, user

Is there a new thread yet? Can't access catalog right now.

JUST COME IN, DON'T STAY THERE

V-Vladimir Putin is that you?

What's that cone-shaped thing? salmiak crepe?

Sunflower seeds.