/agdg/ + /vm/ bread

already here, it's called employees.

AI that will write code for you for free when?

Pick 2

I take:
real within 500 years
good and free

What's both good and free anyway?

Allergies will just fuck my shit up if I go outside, nothing in there is good.

why

You choose "learn how to code for 10 years"

Quality and being free as in free beer(freedom optional)

Neato!

alright, i got my Wait function working (it's confusing because i have to tie it in with the LateUpdate, rather than just using a Coroutine)
someone tell me if i'm wrong, or why i'm retarded
pastebin.com/2EPTXmqN

yes
do you irc

with an example of course

Not really, mostly steam.

[spoiler]id/name?

Shoot me a mail first.

...

user, spoilers are *.*like this*.*

I stupidly assumed (hoped) it wouldn't need the closing tag, like on you-know-which chan

you mean 4chan, they are the ones getting banned for mentioning us, not the opposite

user, this is honestly starting to look really good.

I know but it's a shit chan I don't want to wash my keyboard after typing out loud their name :^)

mhm for chemical weapon shit the availability stage will be this:
Poison -> Chemical -> Biological

Also I need to think of a few standard caliber hmm..

maybe something like this:
Autocannon:
20mm, 40mm
Cannon:
90mm, 105mm, 155mm
Rockets:
dunno
Machine Gun:
7.62x39mm, 14.5mm
Grenade Launcher:
40mm

I prefer alot of ammo variety instead caliber variations, less work for balancing.

I added alse some funny sound to Blokowisto and Autocannon tonk, too bad that muh OBS doesn't like ALSA/OSS because reason, so I can't record them.

the SRBM Bavly will be removed and replaced with a stationary variant, manually controlled obviously. Unless you like that Turret's wreck shit because they are dumb like niggers. Like that totally not useless cannon turret mod some nigger made.

That better make it into the final version.

Whoops, forgot to unsage.

i'm still wondering if i should put gachimuchi sounds in here
right now i want to find a decent squeaky sneaker sound effect WITHOUT FUCKING BASKETBALLS for when they dodge

Use the locker slam sound effect for selecting a menu option or something.

wouldn't a regular punch be better?

Maybe for a heavy hit, or a KO punch. It'd get a bit tiring to hear on every punch, don't you think?

awesome job boxing dev, it's improving every post

I remember a few months ago where it was just a hitler boxing thing.
They grow up so fast~

i really want to start recreating the "boss of this gym" scene
but AI needs work
the struggle is real
i've even got my nifty folder of gachimuchi sound effects ready for when the time comes

>have to find another job soon, due to taking a break from uni for a term, and rent/etc is expensive

using stolen assets is nice

When you write one.

i feel you user
if it wasn't for uni and worrying about failing, i'd be doing this 24/7

I probably would've hated text games when I was 7 too, don't know why I thought it'd be good

Are you a Mass Effect developer?

Untextured repair tool thing and ammo box

you making a battlefield 1 game?

WE WUZ

/mu/sician here, anybody have any knowledge about making vidya music? all of my own music is either potato quality pop riffs or sloppy 14 minute long guitar + drum machine jams.

just like maek song

nah just anachronistic stuff, it's all just hobby work at the end of the day

I've heard it's just 30 second loops usually, but I might be wrong

I think all you need to do is just learn how to compose.

Text games are rad. just beat your nephew until he likes it.

I wish Starship Troopers was real, and I say this as someone who was beaten as a child.

...

To the person making a strategy rpg. Good luck. May you succeed where I have utterly failed. I'll be around if you want to bounce ideas.

How would you go about implementing a time-stop in terms of game code?

Same, on all accounts.

...

Make one using RPG Maker with him. If he complains about 2D beat him up for being a graphix whore

Stop executing the code that advances the game/physics simulation.

You do have your code nicely segmented into a game loop and a render loop, right?

Can anyone give me an idea of how fighting works in beat-em-up games like final fight?

...

i made some placeholder dodging animations, squeaky sneaker sounds included
i still need to get the AI tree to start checking if it can actually dodge in that direction though, like the hardcoded one did

It gets me every time, without fail.

i wanted to put more ed, edd n eddy sound effects, but this is the most popular one out there, can't find others unless i bother watching every episode to extract them myself
i did want that one where it sounds like a doll saying "mama" for when they get stunned

I need some help, if some artist want to create some tiles for me.
Here is an video showing the currently used tiles and their rotations, the very start of the video shows the X+,Z+ and Y+ lines.
Not only the current ones have between 200 to 500 triangles (made with boolean ops, it works but the result is super shitty), but they also are solid, so I can't see the inside when making the map.

The red square is to show how they are meant to fit inside an .25f sided cube, the second video better illustrates that by rotating some of the pieces

The original tiles: dropbox.com/s/6lb917gpo393p8t/tiles.7z?dl=0

Gonna follow this post with some explanations on how some tiles needs to fit with others

(The rotation webm will be uploaded in the next post due to filesize restrictions)

...

These show the links between the tiles.
The original tiles would also not match with each other sometimes, because I made them quickly to start prototyping

The edges of the same colors needs to match, besides that, their format matters not.

There is bound to be some symmetry constraints that I am not aware of (due to the nature of 'randomly rotating tiles until they fit' focused development)

Given the simple shapes of the tiles, you might be able to write a program that generates the model for each variation.
The only downside is you'll have to go through the fuckery that is model file formats.

one example of tiles that don't fit but should

nah, I want to be able to make many tile types, like cave tiles, aztec-temple-like tiles, scy fy tiles.

They need to be custom-made-meshes.

One thing I've been thinking of making is some sort of 'auto fit' for different tile types. But that some complex vertex fuckery I don't want to deal with for now

I *might* look into this later. Right now I'm tired as fuck and need to go to bed for two hours or so.
One thing you need to know is that if you have separate tiles and want them to alight, shadows and reflections WILL be off. What I mean by that is that you will still see the seams.
You're going to get something like pick related.

"align", of course.

Oh, and what about UV-mapping? You're going to want to texture these, don't you?

that's up to the artist making the normals correctly, or rather 'correctly for this specific not common need'

the test was made with calculated normals with 0 smoothing, all 4 sides have the same normal data, but they fit correctly.

the same goes for texture, not only the UV needs to be made thinking about tiling, but the texture should also tile correctly

The biggest problem is that since I have not made the tiles (besides adding and removing meshes from meshes with no regards to quality), I don't know all of the constraints, and the only thing I know for sure is that this seems like some very hard thing to do without being able to preview the final result quickly

unga bunga where the Tetrachrome dev at.

not much to show today.

Someone's making an SRPG?

someone tell me what i'm doing wrong
my AI is supposed to check if he can dash in a certain direction, if so do it, otherwise check the other possible directions
but in my case, for some reason he thinks that if he's right on the edge of the NavMeshMap he can go just about anywhere
so i force him onto the ropes of the ring, and he's still trying to dodge towards them, because he's right on the edge
bool CheckBackwards() { return NavMesh.CalculatePath(owner.nav.transform.position, owner.nav.transform.position + owner.nav.transform.forward * (-2), NavMesh.AllAreas, owner.path); } bool CheckLeft() { return NavMesh.CalculatePath(owner.nav.transform.position, owner.nav.transform.position + owner.nav.transform.right * -2, NavMesh.AllAreas, owner.path); } bool CheckRight() { return NavMesh.CalculatePath(owner.nav.transform.position, owner.nav.transform.position + owner.nav.transform.right * 2, NavMesh.AllAreas, owner.path); }

Never used NavMeshes on unity, but I'd assume that 'NavMesh.AllAreas' is enabling it to attempt to use the NavMesh outside of the ring

no, that means the Walkable, Jumpable , Not Walkable layers, and whatever other custom ones you have

well, try with only the walkable ones enabled then

nevermind, got it working with raycasts, i was using them before but i was retarded and forgot to return the OPPOSITE result, since i'm checking if there isn't anything in the way

he hasnt posted for a few threads. he was probably murdered by gypsies, because he didnt want to milk waifufags via patreon to pay his depts.

Polish government repossessed his organs to pay for his denbts
RIPIP

Looks neat. What kind of game will this be?

Hmm I guess the Abrudium Stone can output some Abrudium Chunks for further processing.

Regular Abrudium Ore extraction will also output some Abrudium Chunks.

Also working on a 5x5 Warehouse with all the usual chest variants… Expensive but provides large capacity.

You're using flat shading there, while I was using smooth shading. What those seams break is smooth shading. With flat shading you don't see the seams, because…well, you have seams everywhere.
If you want spheres to look round, you have to use smooth shading though. That might be something you can fix with normalmaps, but I'm not sure. I'll give it a shot though.
If that doesn't work (which is the outcome I expect) then you'd have to use merged meshes in those places. That means either merging those parts of the meshes after you're done with your map, or making those tiles consist of several segments and do some clever replacing.
The former is something I wouldn't do. It's simpler, but you'd have to use your own proprietary 3d format to define which vertices should be merged and which shouldn't. You still want some edges to be hard after all. The latter puts the workload to the artist, but should be easier.
In my opinion, the best solution would be a hybrid. You make the model in segments and (for example via their names) define which should be merged with their neighbors and which should be left alone.
More work for the artist, more work for the programmer, but at least split somewhat.

Yes please. Also is it possible to create a PvP gamemode? Having warfare PvP would be great, would basically be a test of autism of who can pump out more ordnance and units.

Might as well post my progress, even if it's basically nothing.

Text will be replaced with symbols later on, and so will the look of the thing (especially the ugly as sin indicator will be gone), but the functionality should be there, apart from the fact that it leaves a tiny indicator when I close it, which I have no idea why but I'll find out.

I have no idea why people consider pie/wheel menus to be a console thing, a mouse is so incredibly fast for selecting stuff from them, don't see how they are not a much more desirable alternative to the slough of a scrolling menu (like the ones in Arma).

Doing god's work

(checked)
hmm I've seen at their forum there is some few PvP mods but I never tried those things so I dont know how good they are.

Also they are about to add MP lobby which I am fucking surprised that it took for them so long to add one.
looks like some nice amount of trolling could be made :^)

Oh if only Factorio has a support for faction related technologies to make things more interesting.

Like low-tech guerilla vs hi-tech brute force or something like that.

Is that okay?

way better, but that is still one of the tiles that don't fit when it should. See:
Expected is shown here:

Technically didn't do much myself, the standard 3rd person animations in UE4 seem to work relatively well if you attach the camera to the head bone. They could use some adjustment with the headbob (sometimes it's a bit excessive). That said I do intend to keep it like this so it will be full body (though models and animations will probably come quite late in development).

Maybe once they release the game they could do what some other games have done and release the source code. Space Engineers did this (badly, since the terms basically make it impossible to share any source modifications, but still).

Won't happen though.

Well that surprised me that Jew Engineers in Space got open sourced, so if I understand it right basically it only allows for some glorified mods and shit, that's fucking it.

So Standalone mods like Shadow of Chernobyl OGSE won't be allowed.
blog.marekrosa.org/2015/05/space-engineers-full-source-code-access_40.html

Oh well not that I really care much for this game anyway since it runs like shit on my rig, too bad it could probably satisfy muh tonk autism.

Fucking steam cucks, I bet sucking on Gaben fat jewish dick must be tasty for those faggots.

Nah worse than that, as far as I know it straight up doesn't allow you to share jack shit which means you can basically only fix bugs for the devs so they can slack off more. I think it's basically something along the lines of "you can share modifications but you can't share the entire sourcecode, but we also aren't giving you a way to share these modifications without sharing the entire source code and having the downloader compile the game from it". My point was just how games that aren't free sometimes still give source access. SE is probably the worst example.

If you want military vehicle building autism From the Depths is the only way to go right now.

Also in that link:

Who the fuck would want that buggy laggy shit, if someone wanted voxel planets they'd steal it from Outerra or something.

Shit this is indeed worse then I thought, fuck I can't even find a fitting reaction picture for it.

Alright I will look into that then.

k so i did more on the defensive stuff
made an AI for it
first he checks if the enemy is attacking
if not, he just stands still
otherwise, he checks if he has more than 50% energy
if so dodge
otherwise weave
if he did dodge (he could be cornered) he does nothing more
but if he didn't he weaves again
if he did weave, he does nothing
if he didnt (he doesn't have enough energy) he blocks
for the player to dodge, he just has to double tap A/S/D
ill post a video later, takes forever to finish one that doesn't look like crap
especially when the fps suddenly starts tanking out of nowhere (atleast while recording)

Well, my updated ones fit perfectly.
The lower one is 2.5 lower on the y-axis (not 0.25, because I scaled the models up by factor 10 in Unity to make them easier to work with).

However, you have the problem the exact problem I predicted earlier. I doubt that you can fix that with normal mapping. I *might* be able to work something out that does not require you to merge them in the editor or in the game. You shouldn't hold your breath though.

Hopefully that can be solved with docs.unity3d.com/ScriptReference/Mesh.CombineMeshes.html
It was part of the plan anyways, but I am unsure of what to do when I start making tiles with more than one material.

By fitting I did not meant fitting with itself, but with the pieces it will be constructed with.

Using the first picture in
That piece (4th left to right, 1st top to bottom) should fit with [1st left to right|1st top to bottom], [2nd left to right|1st top to bottom], [3rd left to right|1st top to bottom], [4th left to right|1st top to bottom] and [1st left to right|2nd top to bottom]
The parts where I drew the blue line should all match perfectly.

...

Reminds me that I had an idea for a run-of-the-mill medieval/fantasy game that beats the shit out of Skyrim with it's mechanics. You could sidle against walls like in a dungeon, and throw rope bombs or something from behind it.

But then again, that sounds like a lot of games these days. Basically, I'd want to have whatever people like in TPS/FPS medieval-fantasy games, and not skimp on the details.

That's what I meant.
You said that you're working with cubes of 0.25^3. When you take mine and place one at (0, 0, 0) and another one at (0, -0.25, 0) (so one cube underneath), they fit perfectly. In your old ones the centers are off. For them to fit perfectly you need to offset them by 0.2819751 (at least this one).

I have some other good news. See pic related.

...

I want it to look nice, damnit
If I have each terrain type able to fade off into nothingness, then I can stack them and arrange and layer them however I want to

Though, for some reason, a 48x48 hexagon doesn't tile correctly. Moving down to the next row fucks up the spacing by 3 pixels. I correct it, it's no longer constrained to the unit size of 8 pixels per subtile, but then it works fine

Fuck me

They are meant to be off center, so that when rotated, that offset makes them be where they should, check . It is hard to explain since I don't know the proper names of these particularities

I am happy to know that custom normals solve the seam problems

There is a grave misunderstanding, and I can't better demonstrate what I mean right now, I will be able to do so in about 5 hours.

The original tile that you are improving is not geometrically correct, the tip where it curves upwards should not be square, but round.

Oh cool never, I just fixed it. The hexagons work fine if their slope on the sides was 3:1 - I was using 2:1 and it didn't quite work, it was introducing non-tiling gaps. Now it works just fine.

I can't wait to come up with all sorts of cool ideas at work and never use them because I'm a lazy shit

VIDEOGAMES

Ahhh. Now I see.
I should intersect with a torus instead of two cylinders.

There are some details you need to know when working with the custom normals, but you should be fine.

I always imagined a VR rts should be something where you are a general with a map and soldier figures in front of you, being closer to grand strategy than a normal rts.

Just so I don't waste any more time.

Yep, exactly that

Okay.
I'll remake your tiles. Won't do the UV mapping though. At least not now since I got other shit to do.
Consider it a birthday gift.

I've been working on using the ninja build system, and it's kind of a dream. Most of the annoying shit I had to pretty much force to get working in Scons, GNU Make, qmake, Cmake, and premake, is pretty fucking easy when you swich your paradigms and separate your interests. When you get a basic build.ninja file generator up in Python (or your scripting language of choice), it's amazingly easy to get it to do whatever you want, because anything extra you need to do, you already know how to program in (as a programmer, anyway). It's like switching from any other template system to a mustache-based one. All about the logicless templates and logicless build systems. Consolidate the logic to languages that are built to handle logic, and save yourself a lot of headaches. Stop fighting with template languages and build systems, and just use ones that are predictable, and don't have their own languages for you to learn the ins and outs of. Just use the languages you want to use.

These two are also geometrically wrong, all others are as they should
Thanks, but It's jan 21

Test

Noted.
Wasn't talking about your birthday there, user.

So are you making a smooth surface Minecraft or what?

I am making what

He said he couldn't pay for internet, maybe they finally cut him out

Being a programmer is pain when you need real assets, especially specific assets. I suppose I can always try to get something from opengameart.org working.

I think the reason this happens is because you're a far more experienced programmer than you are artist (just making assumptions here). An experienced artist can doodle serviceable assets quickly because he's used to his tools and has a workflow. He already has all of the steps outlined in his head when starting and know where all the buttons are, meaning he doesn't have to go back and redo stuff because he got it right from the start or consult tutorials or documentation to remember how that one vital step was performed or what that one function was called. He can maintain his flow.

Almost done, user. The second one on your pic will take a bit longer though. Need to make sure the topology stays sane.

This

Good assumption. I am a professional programmer and I am not an artist in the least.

Still painful though. I need to actually work with a real artist or something.

I wanted to make a silly remark and ended up having to leave my computer for a couple of hours.

I am making some tile-based map editor, those pieces are the pieces for a auto-tiling tile (sorta like RPGmaker does, but in 3D)

They are round because if they were to be flat, the inwards and outwards variations would not have been 'detected', so to reduce the number of bugs in the final version, the prototype is already using ""'"'complex'""'"' tiles.

Gotta go again though, I will be back in about 2 hours

what are you trying to do?

In general, or in respect to this model?

In general, I'm making a simple open source (MIT licensed) Castlevania clone for fun. 2.5D, using C++11 and glfw with vanilla OpenGL (probably Box2D for physics). It might be extensible with some scripting language like Squirrel or Lua, but that depends on how much work it is to get it up and running in the first place.

In particular right now, I'm trying to get a simple humanoid model up and rigged that works well enough for the initial programming phase, so I can worry about emplacing a better model when the physics are up. Most of the models on opengameart are either too complex or badly rigged, and feel like more work getting them into a form that I need them than if I just do it from scratch.

I just need something relatively mannequin-like so I can get my skeletal animation system finished. I'll probably give up on trying to get something even partially passable in Blender and just do what looks like lego-people for now, which will be more than good enough at this phase of development.

Done.
Let's see, if we can share the files like this.

If you're just sharing models, the easiest way is to export them as collada files and share them through pastebin or as a git repository.
Here is the corner_diagonal_one_connection.blend file in .dae (collada) format, which is importable in Blender:
pastebin.com/0ACVuCkF

You can also do the stanford .ply format, but it's much less flexible and doesn't allow many things that collada does, like animations, armatures, textures, and the like. It is smaller though, and is probably as compact a format as you'll need if all your meshes are is vertices and normals:
pastebin.com/rAccUrxi

If you want to share a bunch at a time, a git repository of collada files on Github or gitgud is pretty good too.

Stuffing them in an archive embedded in an image works well enough, though, for just throwing a small set of models through an imageboard. I'm surprised hotwheels hasn't disabled that functionality like moot did years ago.

That isn't an option. I have various .blend and .fbx files and a Readme attached.
Can't just convert it to collada

Yeah, I didn't think so, just thought I'd list some options I'm aware of (and was also hoping somebody might point out that there's a better way). Sharing arbitrary files on anonymous imageboards can be a pain in the ass.

I'm honestly amazed you can just append archives onto images and Holla Forums doesn't catch that.

I could throw a simple mannequin placeholder model on the rig I'm using for humans in my WIP game and upload it for you (when I get home).

I appreciate that, but I'll just deal with something shitty looking for now, since it's all placeholder anyway. It's better off being an eyesore so I'll get around to fixing it earlier.

low poly it is.
also what's the easiest program for animations?

Guy who posted music in the last thread. Someone Requested a melodic piece similar to a song he posted (finnish folk-metal)

I'm almost done with it, I just need to create the instruments (just composed the whole thing in piano) and I'll post it when I'm done.

It worked, just don't expect me to know how to make an zipgif thingy.

Gonna check them tomorrow, today I am way too tired. Thanks in advance though

I didn't understand half of the steps, however I got the files before I even started reading your post, took me 30 seconds tops

The basic gist of it is that you have 2 files, the gif and the 7z archive. The 7z archive is appended directly onto the gif, so that it's one file that is just the other two put directly together.
When you specify a non-7z file to be opened by 7z, 7z sees it as a "drive", or some other character device, and then scans the drive for a 7z archive, which it finds after the gif data (this is for compatibility, to be honest. This is so that 7z can be used to store data on a tape).

Note that this can fail (though it's extremely unlikely), if the 7z magic number is inside the image somewhere by random chance.

You can make these files in Linux or Mac OS X with the following command (using your image file instead of foo.gif and your archive instead of bar.7z, will produce imgarchive.gif with the image and archive):

cat foo.gif bar.7z > imgarchive.gif

Or in Windows:

copy /b foo.gif+bar.7z imgarchive.gif

Some legit news for other unity devs:
source: blogs.unity3d.com/2016/05/19/make-your-daydreams-a-reality/

niceee

so what are Vulkan and Daydream exactly?

Vulkan is a low-level graphics API created by the Khronos group. It's essentially the next generation of OpenGL. It offers increased performance along the same lines as DX12, but it support all platforms that can run OpenGL.

I think it's some VR platform, not entirely sure… but that's not the important part of the press release info

This user sums it up pretty well

sounds awesome.

It's not really for direct use. It's way lower level than OpenGL, and needs rightly

Fucking phone.

Needs roughly 3 times the code for the same programs.

Hi /agdg/, i have an idea, on which i'd really like to start working.

I want to recreate the core mechanics of Dark Souls, as in combat system, movement, inventory and so on, but for an open "engine" or toolkit, im not sure what's the terminology there.

As a result, it would basically be a level editor for Souls games, and someone could create an entire world with it. I know C, Python, Javascript, and Java by proxy, but im guessing that this isn't all that relevant.

My question being: What's a good approach? Just start fucking around in Unity/Unreal/Cryengine with youtube tutorials? Start from scratch? Is there any fundamental difference when creating a "game" with a level editor? Could you possibly use the native Unity/Unreal/Cryengine editor even for a "final game"?

This Mantle hello world is pretty much what you can expect the same in Vulkan to look like: github.com/Overv/MantleHelloTriangle/blob/master/src/main.cpp
And here's the breakdown: open medium.com/@Overv/implementing-hello-triangle-in-mantle-4302450fbcd2

No you don't. None of those languages have very much in common with Java at all, other than being garbage collected. Javascript is one of the farthest C-like languages you can possibly get from Java.

but seriously, starting from scratch is not a good idea for something like that, and there's no particular "fundamental differences" when making a game that is mostly level editor other than the recommended practices of making your code modular and not hardcoding stuff unnecessarily.

meh…

o-okay

So, you're saying i could just go the fuck around in Unity with youtube tutorials route and end up a serviceable result? I am kind of doubtful because almost none of the good projects i've ever come across are made in those kind of engines.

There are at least half a dozen decent games made in Unity

He was too much of a nonjew for this world

You should probably focus on tiny sections of that first imo, and I'm saying that in context of you just getting into gamedev.
Having big plans is fine, but overwhelming yourself (with a level editor f.e.) when you're just beginning is something that's best to be avoided.

So, like open source engine (code is open source, usually under MIT, apache, or BSD license)?
Or, do u mean one where anyone could use to edit your game (download for free, but the code is closed source)?

yup, with the correct tools made by you of course.
Afaik, kerbal space program uses the unity editor for that, but they also have a package w/the editor related mod tools to do so (importing models, other stuff).
You can find the source code for that if you look, and see what it takes to add that type of support.

Not really, no.
Although, the majors differences come into play for how you could implement this type of system, and what the requirements are to make that happen.

That should help.

I'd probably decide if I wanted to make an engine, or use a preexisting one.
Steps are pretty straightforward from there… learn your tools, and get workin' with what you can do atm with said tools… i.e. learn, experiment, practice, rinse repeat.

I'm bias as I use Unity, but every engine has its flaws… it's the one whose flaws you can stand that you use.
One of unity's "flaws" is that it's so easy to use, and people who don't know what they're doing hop in and make shitty games that aren't optimized.
they're in it for the shekels… not for the joy of making games.

I realize that. I'd be happy to just get core mechanics right, without any level editor at all. Or do you mean, start with something else entirely?

Open as in, some user could go ahead and build a level or an entire game using it, share it in a thread and so on. An example of this would be the spring engine for RTS creation. Didn't think much about the license.

Im assuming that the editor part would be by far the largest and most complicated part of this project. I'll take a look at how KSP does it.

Preexisting, definitely. Creating a level editor completely from scratch would probably take at least a year i imagine, as long as one doesn't do it full time.

I've heard that a lot. However, what are the arguments against the other major players, Unreal and Cryengine for example?

Not gonna give up yet, but god damn making games is complicated as hell.

You should get used to the engine first before jumping into this (at least)3 years project you've got.
Lest 2 years from now you find yourself having to rewrite all the code you made because you weren't good at using the engine back then and now's everything held together by duct tape and barely works.

Do all the tutorials first, then try to add new stuff.
Odds are, one tutorial will explain how to add scores in the simplest and most optmized way.

One thing I've got to say, get good at the engine, then make the project public.
You've an idea with huge potencial in your hands, and lots of people would be glad to help somehow, but you've gotta be able to create something presentable first.

That sounds reasonable. What's a good starting point for actually getting to know the engine? Just go by the most popular tutorials i happen to find?

Do tutorials, then do small shit with 3D(think small) that doesn't require many art resources.

Also, I must say, I don't think pre-made engines like Unity are well-suited for having level-editors at all, principally if you want to make it some kind of "everyone can do his own level with no programming experience"

I've got a raging headache and am exhausted, so excuse my bluntness.

Do what you're comfortable with, do some tutorials, and become proficient with the tools.
If you don't mind rewriting your code entirely (which you'll definitely do if u start on what u actually want to do now), then go for writing what you actually want now.

Big difference between those two, the latter being a huge fucking project… you're talking an entire framework with all the tools necessary to create an entire game here, this is no small task (I'm talking years).
If you do it in-game, then it doesn't matter what engine you use.
If you do it in-engine, it depends on if you need access to the engine's source code.

never used it, so I have no idea, not looking into it tonight

Their documentation is shit in comparison, and they don't have an as active community compared to unity; i.e. they have a higher entry barrier.

use unity tutorials, then do random YT tuts on stuff that's sorta similar to what u wanna do, and then experiment with stuff you're interested in.

depends entirely on how you do it.

I'd say that the baseline for every single level editor is having a few objects/geometry/enemies pool that you chose off and put them on a map, I just don't really see a good way to implement something like this in unity.
But I'm a dumbfuck anyways so theremight be a good way.

welp, basically I know it's possible because I've already accomplished that

Can I see it?

Note, I'm saying this in a more "oh shit I'm curious now" way than a pedantic "bet you didn't" type.

no, I'm refactoring my code into compute shaders atm, and a lot of functionality is broken due to this.
I'll post screenshots in another thread once i get it functioning well again though.
I've posted how I do it other threads, but I use dual contouring w/a signed distance field for objects/geometry (no enemies as I'm making it fast), and procedural generation; the same idea applies to level editors, but myn is on the fly.

well, here's some screenies for awhile back when i was using dual marching cubes, but I didn't demo me changing the mesh except LOD

...

normal marching cubes in this screenie, not DMC or DC

some artifacts from my triplanar shader

Alright.

Im working as a programmer, refactoring shit legacy code is what i do on a daily basis.

Man you're right. Maybe this is way too ambitious from the get go. It's probably a pipe dream-tier plan. I just love it when these sorts of projects actually come together and a community forms around them, and there would likely be a lot of people who would have fun with this.

Maybe one day someone will somehow reverse engineer one of the Games. But now that i think through what would all be necessary, even for the "some user creates a level" kind of scenario, it sounds ridiculous. Enemies who need models, animations, hitboxes, movesets. The level itself, geometry, textures. A game really is a gigantic project.

Thanks for the advice either way, it's a refreshing perspective. I was just kind of thinking about this idea without the proper reality checks.

So dual marching cube makes more holes/caves?

Here's your current plan

That's the noise I'm generating (3D perlin noise) for the signed distance field; which creates holes/caves.
2D noise would create no holes, but dips/valleys/mountains and such depending on the seed/frequency/amplitude (plus lunicarity/persistence per iteration of noise octants) I use as inputs.

Dual marching cubes is an isosurface (same-surface) contouring algorithm… basically how the polygons get put together to form a surface.
Also those pics are of two different methods, one being the "dual" method (DMC/DC), and the other being the "primal method" (MC).
Example pic provided for differences.

few more screenies I found lying around… I think this one is normal marching cubes

Good stuff, you'll be right at home then heh.

my pleasure user
Though, having an overall goal, or dream you're working towards isn't bad to keep in mind.
It can be the drive to learning, and doing everything necessary to getting closer to achieving said dream.

Hope u make something great though, and put that passion for vidya to good use.

Thanks man, im off to sleep. I should probably concentrate more on the Uni in the coming months, but i'll play around with it from time to time. This is a good general you've got going.

Anyone have any experience with panda3d?

Just continue to program at a lower and lower level, to avoid ever needing assets.

When you're hunched over the kitchen table with a soldering iron, you'll know it's working.

First time devposting, be gentle. I've been making a platformer that's based around having lots of movement options. I have a tech demo working.

Questions:
*Can I make and call methods (As you would call them in java) in gml?
*I recently read that updating an object's x and y values to move it is not good. A. why and B. Do I need to completely rework my movement system which is based on updating variables for x and y movement per frame?
*If I posted the actual game would anyone play it, and/or how do I make a webm demonstration for easy sharing?

crossover between fire emblem and batallion wars

i think scripts serve the same purpose, otherwise i don't know

Thanks, I'll look it up in the manual. And this is a retarded question but is there some sort of reference on how to perform actions by writing it in code instead of actually drag n dropping the action itself? I've been having moderate success with just googling it but I think I'm making it harder on myself.

Did you do their tutorials yet?

There's a drag and drop option to run a script.
Move of your game should be this way, rather than using drag and drop. Each D+D option has a corresponding script command.

Also, using scripts as method calls works, but they're actually kind of slow. Even if you only ever pass 1 or 2 arguments in, last time I checked, their compiler pushes all sixteen allowable arguments per function call.

every event should have it's code counterpart, for example you can write

if place_meeting(self.x, self.y, obj_potato){
shitpost on Holla Forums
}

and it should replace selecting the collision event and then selecting the obj_potato object

No should I?

I know they have counterparts, I was asking: is there some comprehensive list of said counterparts?

I'd say that it's always good pratice to watch the tutorials of any given engine, even if the first 2-5hours are going to be boring ass bullshit.

here's the index
docs.yoyogames.com/source/dadiospice/002_reference/index.html
you'll find the functions by navigating the options

Ok I'll give it a shot. Thanks m8

Thanks!

...

ayy, the polish government didn't eat your liver, congrants on still being alive

A little bit. It was made for making animations IIRC, and it shows. It's slow to start up, and the collision cehcking was ass (e.g. "spheres" are actually truncated cubes). I ended up dropping it for OGRE.
Also its handling of file names is apparently case-sensitive.

Damn, that's very nice and in-depth. If only there existed such tutorials for every kind of gameplay formula.

Added 5 types of warehouse, smart and reqular one have 260 capacity, while the drone ones have 240. I will plan on trying if its possible to make a large drone that only lands on a specific port and making a quick charge drone port.

The large drone will be only used for logistic part but as the name implies it will have larger capacity and is bit faster too.

Then next comes a liquid storage tank with 5k (2x more) capacity, the 4 cubes you see here are the connection things.

Also is there a tutorial somewhere how to paint/texture a model without the usage of UV-mapping? And yes I am using Blender since it is currently the only best all libre all round graphic editor.

You now can pick up items, carry them around, manipulate them, then drop them somewhere and everything is stored. It also works across levels.

...

Would you play a lewd GSG/4X style game that revolved around slaves?

Only if you can play as a slave :^)

(checked)
You've mistaken me for someone else, but I guess thanks for the sentiment heh

i remember tetrachrome dev posting those before, might have been wrong

...

I do. It's actually pretty nice, mainly because of python. So if you hate python and everything it stands for, panda3d is not for you.

Scroll through panda3d.org/manual/index.php/Main_Page to get a good overview of what it can do.

The model format is reasonably competent, but the blender plugin doesn't support all features, so you may have to occasionally edit some material properties by hand with a text editor, with the command-line tools or in-engine with code. The format is plaintext, so it's easy to work with (can be compiled to a more efficient format).

You can roll your own shaders, which is made easier by panda providing a bunch of automagic shader inputs and it's trivial to push whatever data you'd like, both in the form of uniforms or buffers with custom layout.

There's a couple of physics engines available: the builtin (very simple), open dynamics engine (never tried it) and bullet physics (very good). They integrate very nicely into the rest of panda.

Note that panda doesn't have an IDE like Unity/Unreal has. You write your code with your editor of choice, then run it. There's no drag'n drop, no scene editor or any such. You'd have to make some stuff on your own or browse the forums to see if anyone's made something you can use. Concerning level editor in particular, I think there's a guy making an editor intended for large outdoor scenes.

If you want a free software version of Unity, Godot is more in the right direction.

nah, I posted those a few months ago.
He also doesn't even use an isosurface contouring algorithm either… I only know of one other person on agdg that does this actually.
For reference… I'm the procgen dev who's making a game based on my custom cosmology.

anyways, here's some new ones from just now, but using my old marching cubes approach (from an old project that I finished a few months ago); not the one (DC) I'm converting to compute shaders (which looks a lot better imo).

god dang, getting the pipes to properly align is going to be a bitch. Welp looks like I need to model a pipe close to Factorio one as possible so that it doesn't look way too off. That and getting the window liquid thing to properly work somehow…

Is it a good idea to use Javascript instead of C# in Unity, yes or no?

I know JS better.

Unity is so bloated and laggy that JS can't make it much worse.

git gud with C#. Even the Unity team is moving their assets over to C#.

It all gets compiled into unity script or some shit anyways.
But I'd assume that they will phase it out with Boo script to focus on pure C#
I might be wrong or outdated on that

It all gets compiled into either IL or native code (with IL2CPP and only on iOS IIRC).
Boo has already been phased out some time ago.

this, use C#

How in the good god damn do I blend a forward walking animation with a sidestepping animation without the character doing some shuffle dance?

Maybe blend the forward with a 90° walk animation and just transition it to a sidestep at >80° or something?

Does anybody know if there was anything special about how Shadow of the Colossus handles the main character's animation?

It always looked weird, but really good, and I can't quire put my finger on it. Is it just normal mocap and I'm just crazy?

they used IK for whenever the player was climbing a collosus, that's all i know

probably simulating turning rate, if you are familiar with dota, each character have an forward vector that slowly changes to your desired destination, and they won't walk unless the forward vector is at least XXº of the desired destination.

In your case, you could have it play the animation based on the angle difference or something.

or something, not even I really get what I am saying, I hope it gives you some ideas though

yeah definintely using IK for the climbing as this user said
Also, their running animation has an extenuated step, and when running with a weapon it has some IK so it seems weighted.
It gives a more organic feel to the animations.

The script I have so far has 100% working movement/rotation targets that the actor does their best to follow. I even have a working system to blend between forward, left, right, and backwards walk animations. The animating just look like sit when blended together. So I think I need a bit better solution than 4 animations (walk forward, walk back, Strafe left, Strafe right)

What are your thoughts on SHOP and HTN planners?

github.com/oubiwann/shop2
cs.umd.edu/projects/shop/description.html

Find a game series or game genre that people love but has gone relatively shit over the past few years.

Fix the shit elements and add some elements that you always thought would improve the games that they just never implemented for one reason or another.

Halfway through development, let people know what you're doing and let them follow along with your dev process (DO NOT let them see everything, only progress).

Your reward will be success and dosh. Every major success story I've seen that HASN'T been a "first of its kind" has gone this route. Super Meat Boy (Super Mario World with improved handling), Stardew Valley (Harvest Moon, naturally), Undertale (if you take away the inane story bullshit you're left with an Earthbound clone that is passable), Binding of Isaac (success from Super Meat Boy mixed with combining roguelikes with Zelda),

Good observations, but I'm only making my game for fun and the satisfaction of it. If it's not popular that's fine. I'd rather not have the majority of screaming YouTube manchildren making videos about it.

I have yet to experiment with planners (so far I've only really experimented with behavior trees, and neural networks), but I have done some research in the past.

The premise is really interesting, and the application seems pretty well suited to games; although they have their own pros/cons compared to other styles of AI: aigamedev.com/open/review/planning-in-games/.
Although as it says in the paper (pdf attached to your post), "a full knowledge of the world" being necessary, isn't actually a good application for vidya (as you probably already realized).
As in halo, and fear, it has become apparent that you must only give the AI agents the necessary information through a type of perception system (i.e. only what each agent would know about at the current frame); then it makes for much more realistic behaviors; without the "god like omniscience" that a lot of game AI exhibit.
I know that F.E.A.R. uses a STRIPS style HTN planner, and achieved some "emergent" behaviors due to no intention of their own.
You can actually look at some of the fear's AI code, and learn from them aigamedev.com/open/article/fear-sdk/.

currently I'm reading C# 6.0 in a nutshell before diving into monogame, but I'm really exited about getting started with it soon. Anyone else working on something in monogame? Would love to have someone to talk to about developing with it, since at least to me it seems monolithic

what kind of game and what are you making it in?

There is at least one other guy working with monogame. iirc it's one of the guys making a roguelike

The models are working fine, everything worked, but I am remaking a lot of stuff and have nothing to show, I was expecting to post this sooner with something to show, but since I am going to sleep soon, I am just gonna report that the models worked

That's it, I'll make a new Paper Mario.
ORIGINAL IDEA DO NOT PIRATE IT'S PATENTED I SWEAR!!!!

Hey guise

I mostly use GML in gaymaker to maek my shit, and I'll probably stay like this for a while, but at the same time I want to explore new territory, what are some good 2d engines/frameworks to work with?

I heard Love2d and xna are gud

pls no construct or stencyl or any shit like that

Godot and Love2D

Try everything you can get your hands on.
Worst case scenario, you learn something new.

Love 2D is good for small scaled projects, i wouldn't build a super large game in it.

XNA is dead, microshaft stopped supporting XNA a looooong time ago, but monogame is basically the new XNA, being almost the exact same.
if you're understanding GML then try a light read through a book or two on C#, or just a crash course, then see if you can get whats going on in some monogame tutorials.
good luck on whatever you do

I am still fucking mad with the fact that Blender doesn't got PTEX support despite the "Blender in 2015" page was saying it was about only 90% done which looked like it would be put in, all my research about it seems to look like that they just abandoned it silently for some reason.

JUST FUCK MY SHIT UP FAM

Turns out I live in the same city as the fags that make Gamemaker. I'm gonna go try going there on Monday and beg for a summer internship.

Any tips? How do I ask for an internship?

Help me user

Haha, look at this fag

Hello, I'd like to apply for an internship
unzips Curriculum
I've been using your program for X time and I love it, I'm looking forward to helping with it.

learning is for nerds

MASZ PROBLEM KOLEŻKO? CYKA!

What's wrong sergei, can't torrent a good program?

With which modeling program should I spend time learning, rabbi?

So what? Blender is at least libre, you fedora tipping cunt.

At least I am not using Java Script like a pleb you are :^)

SPICS GET OUT

Got a link to the FEAR source code download that isn't broken?

PUTA BLYAD
HABLAS ES ESPANOL

lmgtfy.com/?q=F.E.A.R. public tools version 108

Thanks, I was hoping to get the one with just the codebase that was linked on the aigamedev forum which is now borked.

BIIIIIIIIIIIIG SOOOOOKA COMRADE, FUCK THESE AMERICAN SCUM AND THEIR "SUPERIOR SOFTWARE", AMIRITE?

SLAV PRIDE WORLD WIDE.

GAS THE AMERICANS

here's a good link i found with a refined search, it includes source code for a GOAP planner (bottom of the page)
it has a lot of links/resouces, but some are broken, but you can find non-broken links by just searching. but the SDK should have the full fear AI source code…

woops, forgot link, polite sage for double post
alumni.media.mit.edu/~jorkin/goap.html

I think you mean, gratis, not libre.

Gratis is free as in referring to price.
Libre is free as in referring to freedom.

wew lad

Thanks m8

...

I don't understand adding in keyframes,

Hey I was just asking, stop bullying me, nikolai ;_;

It was just a prank, bro :^)

Going back to the discussion.
How are you finding yourself with blender? I tried modeling some more complicated natural stuff with modo and 3DS max but it always ends up looking blocky and meh

I think I'll finally have to give in and learn Zbrush

(checked)
I use Blender for secondary purpose, so texturing, uv-mapping which I hate and everything else. So I can't tell much about the modelling procedure here

All of my models are made in Wings3D because Blender seems to be missing a few function that Wings3D offers, like snapping, swepping (useful for making pipes), scaling by Uniform/Radical/Axis etc… The 3D cursor plugin for Blender was really handy but now for some reason it is fucking broken like it doesn't recognize certain polygons (i.e. it ignores them) etc… So I am stuck with Wings3D for ages.

That and the fact that Blender is the biggest like really big compared to other free gratis/libre/whatever 3D programs out there. Which means Blender has alot of developers, they make quite some nice progress and so on.

Wings3D in the other hand is not that popular and it uses Erlang and not many people out there knows this language, Erlang has the advance that makes the program more robust as it doesn't crash so often like Blender does, due to it's programming structure.

In other words Wings3D progress slowly and doesn't offer that many options however for being a pure modelling program with some rendering options (actually sending the information to other rendering programs) it's not that bad. Through it doesn't offer any sort of Animation like one of it's inspired programs could do.

To recap: Wings3D is easy to get into for being a pure modelling program compared to Blender which has some steep learning curve but it is not so bad anymore like it was in the past.

Ah Enhanced 3D cursor was it… Well I updated it and it still fucking broken, see mp4.

SLAV MAEK BEST GAEM MADE OF SLAV

What version of openGL has the best level of compatibility? Is it worth learning about both 4.5 and vulkan?

Unless you're going to use newer OpenGL features, stick to 3.2 and turn on core profile.
OpenGL 3.2 is roughly 7 years old, so every GPU that is less than a decade old should support it. The only obvious exception is intel integrated shit, but those people don't play videogames anyway.

Except if you're actually going to use the advantages Vulkan has to offer, I wouldn't bother. You end up writing way more code to do roughly the same thing.
That said, OpenGL and Vulkan are very similar in terms of how you can efficiently utilize them. If you git gud at using OpenGL, then stepping over to Vulkan shouldn't be that hard. It's just that you suddenly have to write a fuckton of boilerplate to take care of all the implicit state OpenGL does for you which Vulkan doesn't.

As for which version of OpenGL offers what: See en.wikipedia.org/wiki/OpenGL
Note that OGL's API doesn't really change with newer versions since core profile was introduced. You just get more features. If your engine is coded using OpenGL 3.2, then there's usually nothing stopping you from just changing the version number of OGL to 4.5 and then using shader features from that version without a single change in code.

Thanks that all sounds very cool

Help me fam,

I'm unsure of whether or not to release my game engine right now, or implement the last remaining big features that I planned to make it great again.
Would it be very detrimental to the project if it came out in a less-than-optimal state?

And the thing is, I'm working on getting my game implemented as well (i.e not the engine itself, but the game I want to make) and improving game engine features that make it a better game engine are slowing down progress on the game. Because, I mean, the game could be implemented on a shitty engine, but when releasing the engine code under an open source license I want it to at least be good.

I think the problem is that I'm trying to do both though.

Will the engine be free and will there be tutorials and royalties ?

Holy shit I'm going to make this.

Are you cute?

Nah, fuck royalties, I really just kind of want some software out there under my name, and it'd be really nice to see people contribute with thoughts and code. Of course it will be both free and libre (I'm thinking BSD 3-clause.)

As for tutorials, I would make those before actually releasing it. I also have a bunch of test-code laying around.
Also working on documenting more of the code with doxygen (many parts are documented but still lack the doxygen ** syntax at the beginning.)

finish that shit up first then fam

That's pretty ambitious man, do you have a lot of experience?
Not doubting you, just curious.

I've never touched the VR API, I don't even have an oculus. I've got extensive C++ experience and can use opengl, though. Can't be that hard r-right? I mean it'd basically just be people moving around and the room and some piece moving around on a board.

I've been really torn on this issue. I know personally I played quite a few games on Linux laptops with integrated intel chips. So I don't know whether to go with OpenGL 4 or 3 when porting my old SDL_Renderer (hardware accelerated SDL) code to OpenGL, they can't work together for some odd reason so everything has to be converted.

I'd feel pretty bad if I couldn't play a game on my pc because of that kind of shit. I actually used SDL_Renderer in the beginning because I was developing on a shitty laptop which didn't support the newer OpenGL version.

But I might just go with OpenGL 4 because it seems like work is being done: phoronix.com/scan.php?page=news_item&px=Intel-FP64-Mesa-Close

If I'm lucky the drivers will be out by the time I release my game (prolly 1-2 years from now.)

Do you know of any place I can find some stats on how many people have cards that support which features? I know Steam publishes some statistics on this stuff.

Yeah but you also need to know 3d-modelling, sound design, music, art, ui, testing, knowledge of good game design and mechanics, etc.

Not trying to discourage you, but it sounds like a really big project which might require a few people. What I'm doing with my project personally is just implementing as much as I can on my own, then I plan to show it to someone who can do the things that I can't do and hopefully get them interested in working with me (I know a few people who can do art, but I don't want to approach them with just an idea, or a poorly executed idea.)

It sounds like a really cool idea though.

Glad to hear that.

I typically keep this to myself (for a reason), but here we go:
First person survival horror in Unity (also known as: How to make people loose interest in 6 words).
Well, I see it positively though. All those crap games made in Unity just lowered the standard to the point where not being terribly shit became remarkable. Not that that's enough for me. Being mediocre is worse than being bad after all.

That makes me wonder. How many nodevs with Slavic blood are posting in these threads?
Tetrachrome dev, me and those other two who posted earlier.

SDL's hardware rendering code doesn't work with OpenGL because the actual hardware renderer it uses depends on the system. It may use Direct3D if there's no OGL drivers, for example.

I haven't used SDL's new renderer functionality at all, only the old 1.2 one (which is software only afaik). The performance of it is pretty decent, but of course that's 2D only.
Taking a very brief glance at SDL2's source, I can tell you that the software blit code is actually optimized to use CPU extensions where possible. This makes it a pretty fast blit, and as long as you're only trying to blit things that are going to be on your screen, you should be able to get some decent performance off.

If you want to support intel integrated, you should ask yourself how far backwards you want to be compatible. If its a decade old intel integrated GPU, then I will tell you right now you are not going to get any fucking performance unless you will just use the software renderer. Those old GPU's only existed for the purpose of running the windows desktop & a youtube video, and you shouldn't expect anything more than that.
As for performance on newer GPU's, google some benchmarks.

Googling a bit, it seems like the support for the mesa driver is "mostly done". Some features are missing, but the ones that are missing are probably low priority features people don't care about.
fp64 is one of those features. If you're using double-precision floating point in your shader, then most of the time you're doing something wrong. There's some legitimate cases, but for games that little bit of accuracy really should not matter. The only case I can come up with is when you're calculating the WVP matrix in 3d rendering if your game world is very large, but you can (and should) precalculate the WVP on the CPU.

Aside from Steam, no clue.
You can simply see the % of each GPU used on the steam survey, and then use wikipedia to find out which version of OGL is supported on that GPU. (And of course, any newer GPU from the same vendor should support the same or newer OGL version.)

If you're sure that everything you want to support has support for OGL 3.2 or newer, I highly recommend to just go for 3.2 instead, even if they all support 4.0. Unless you're actually using features from newer OpenGL versions, it isn't worth it at all to limit your compatibility, since you get nothing in return for just bumping up the version number.

Fucking around with Sulfric Acid shell…

How do i do a camera like this, that views 2 combatants for the side, and adjusts for strafing/sidestepping (see badly drawn image)? i think Tekken has a camera llike that.

Im using godot, but it's probably just some trigonometry and vector/matrix math, independent of programming language. to bad i cant into math.

if you're using unity, you can just use the default vector functions
cam.position=Vector3.Lerp(bug1.position+bug1.right,bug2.position-bug2.right,0.5f);cam.LookAt(Vector3.Lerp(bug1.position,bug2.position,0.5f);
that puts the camera in the middle position, slightly to the side of both of the bugs and makes the camera rotate towards the middle position of both bugs

So you basically want the camera to always look right through the middle between them, right?

so you basically want to have a line from each character to your camera and the angle should remain the same? Im pretty sure there is a way to calculate the angle between two vectors but i dont know/remember it because im bad at math. I also have no idea how you make sure it corrects the position to ensure that the angle is right

That's pretty cool.

What genre is this?

Amazing oekaki time baby

0. You got two entities.
1. Obtain the average position of the two. (Green)
2. Create a vector from the average position to the player. (Red)
3. Rotate the vector 90 deg. (Blue) Which direction you turn and how far you turn allows you to customize your camera a bit as desired. You can also scale the distance of your line to move the camera closer/further away.
4. Compute the camera position by adding your line vector to the camera position. (Pink)
5. Compute camera direction by obtaining your line's direction, and then inversing it. (Cyan) (So that it points towards the average point, rather than away.)

As for math:
typedef struct {float x,y;} vec2;vec2 pos[2]; // positionsvec2 avg = {(pos[0].x+pos[1].x) * 0.5, (pos[0].y+pos[1].y) * 0.5}; // averagevec2 vecToPlayer = {avg.x - pos[0].x, avg.y - pos[0].y}; // vector to playervec2 vecToCamera = {-vecToPlayer.y, vecToPlayer.x}; // vector to camera. This is a lazy way to do 90deg rot. Learn your trig, fgt.vec2 camPos = {avg.x + vecToCamera.x, avg.y + vecToCamera.y}; // camera posvec2 camDir = {-vecToCamera.x, -vecToCamera.y}; // camera direction. Depending on your engine you either use a normal (in which case you normalize this vector) or an angle (in which case you just atan2(camDir.y,camDir.x)).
Of course you should use the vector math stuff provided by your engine and not my handwritten math.

It's a mod for Factorio.

I'm doing some raw opengl stuff myself and I'm restricting myself to 2.1.

The way I look at it is this: My game is so simple graphically that there's no reason why it shouldn't work on literal toasters, but even with 3.0 there's no guarantee all hardware will support it. Graphics cards, sure. Intelshit, nope. For instance, I have a HP laptop from 2007 that won't create a profile higher than 2.1. Since I consider it to be a good benchmark for toaster-friendly and because I dev on it sometimes, I want to make sure my game runs on it. While I could implement some 3.2 stuff in cases where there are more efficient ways of drawing stuff with, that's only gonna help the more powerful harware, which isn't gonna have any problems running the 2.1 render anyway, since it's more powerful. So there's really no point.

Additionally, by limiting myself to 2.1, I have to learn less stuff and keep a smaller part of the API in mind. I only have two draw calls: DrawArrays and DrawElements. I don't have to think about if instanced rendering is more appropriate or if I can use PrimitiveRestartIndex. I just have my vertex arrays that I setup with my vertex pointers, pass the uniform the shader needs and then draw that shit. It's simple. Sure, sometimes it might even be easier to send a single vertex to represent something bigger and expand it with a geometry shader, instead of duplicationg a bunch of data, but again it's another thing to implement and another thing to keep track of. Simplicity is good.

But if you're doing anything remotely demanding, use a higher profile. You'll be quite happy to find out that making flashy particle systems or drawing forests becomes easier.

Is there any gud tutorial for Blender how to make gude particle/fire&smoke effects? Well I've seen there is quite a few fire&smoke tutorial but when it comes to particle there is none.

The only particle editor that I have is Timeline FX 1.26 or so which is old because there ain't newer on torrent sites. Which is ok for most purpose but well it is bit hard to get specific behavior out of the particles on that program that and there is not alot of special effects stuff to make the particles look a bit different.

Also the only blending method it has for some obscure reason is a additive one and you canno't change it's blending strength which sucks.

Obviously OGL1, although OGL2 is widely almost as widely supported by now, OGL3+ is somewhat well supported except on intel IGP older than haswell, OGL4+ is where shit gets complicated Nvidia has good support of everything AMD not so much and intel is totally SOL unless you're on skylake (and even then it's basic support)

Vulkan is only if you need OGL 4.5+ extensions but don't want to fuck over AMD people or really want to push it to the limit ( OGL4.5 with AZDO will still most likely be enough in that case if you don't care about pretty much excluding AMD from having decent performance)

I don't think that's the case actually. If memory serves, OGL 1 is so old and has such a different architecture, that a lot of the features were dropped from modern cards/drivers. They still work, but are emulated and thus incredibly slow.
OpenGL 2.1 is your best bet. It's pretty much the OGL alternative to D3D 9. It runs on just about anything.

Boxingdev looking for writefags, i need someone to give me a story for my game, and obviously for free untill game is actually published
shoot me an email

It's an underdog story, pay me now.

by writefag i mean person who actually thinks of dialogue n shiet
i could just say "it's rocky 8", doesn't mean it's gonna instantly create all the dialogue, does it?

I upgraded to substance painter 2 on their website but I still have a 50% off voucher for substance painter 2 on steam if anyone wants it.

yeah. the original OpenGL was for graphics workstations. It doesn't translate well to games at all

Awesome tool.
Don't need the voucher though, since I have a Substance Live subscription.

B2M is also a tool I couldn't live without anymore. It greatly helps making PBR versions of the old "Diffuse/Bump/Specular"-workflow textures.

Greetings my slavic friend, be you Lech, Czech or Cyka:^)

Unfortunately I cant answer your question, nor do I really have anything to offer in the makegame dept. other than a decent amount of BTC, delusions of grandeur, and a healthy disregard for intellectual property…

Allow me to at least offer my thoughts, for what they are worth:

I believe we will witness a real renaissance in game coming from the slav realms in the near future

part of this is of course the russian resurgence, whatever people may think of russia it has recuperated greatly and is already playing a much larger role in the world than say…10 years ago. Pop culture will be a huge factor down the road, and games are by far the most potent form of said

I for one yearn for a competent russian retort to all the calladuty blops that the US has been shitting out over the past decade, they could really score a home run if they made a game that was not only good, but mercilessly mocked all the motardedness we have been subjected to

the other factor is that despite some..european tendencies russia is a very different culture than the west, from their values to their philosophical outlook to their aesthetic sensibilities.

there is vast untapped potential here methinks, up until now slavs seem to be known for making interesting, compelling games whose main downside is that they are not very polished..

that seems to have changed, a podcast i will refrain from mentioning out of fear of being DEUS VULTed recently visited a russian indie game conference in moskva, and suprisingly one of the things they kept underlining was how impressed they were at the technical competence of the participants(they also talked about the soviet games museum which was fascinating)

If I were in charge of…public relations in russia I would invest heavily in vidya, it will become an even more potent cultural weapon as the technology advances, they should be taking advantage of the slump that western vidya is in as much as possible.

there is also the not small factor that life is still cheaper there, someone looking to find qualified people with a smaller budget would be well advised to look east..

unfortunately the russkis dont seem to give two fucks about exporting their work, so it is hard to follow this stuff as a europeon, I hope to start a thread soon where we can talk about russian/slavic vidya, after i have done some more research of my own, there seems to be a lot on the horizon albeit mostly smaller projects atm

just realized this whole rant was very russia-centric, probably somewhat unfair considering that the most impressive slavic game, the one that literally cuntpunched its direct western opponent, came from poland

That DevGAMM mascot is cute. Wish I was there so I could grab a t-shirt with it. Also participate in all the events I guess, that's important, but I can talk with people like Galyonkin and other cool dev folks on the internet anyways.

Also, russkies kinda like have a crisis right now due to having failed to develop any other than oil industry in the past 20 years, and with oil prices so low, so expect many of the Russian, and even more so Ukrainian devs, to move to other post-soviet countries like Baltics, Poland, Czech Republic, Slovakia and so on. Not sure about Belarus because it's very Russia, and Hungary isn't very good for business with government trying to put up bull like extra internet taxation.

Is "renaissance" the correct word? Was there ever a big Eastern-European games industry?
Putting semantics aside, I do think that we'll see more games from the non-weeb-east, but for different reasons. The guys behind the Unity engine have democratized game development, for the better and worse. A couple of years ago, if you wanted to make a game, you had to overcome a ridiculous entry-barrier. Unity Technologies gave people who wanted to make games the tool to do so.
Of course it's not without its downsides (see: post Greenlight Steam), but from the perspective of the (potential) developers, it's great.
Short version is that I think we'll see lots of games made by slavdevs, simply because independent development is far bigger now than it was in the past. Among the hundreds of shitty titles, you'll find your gems sooner or later.

Speaking as a gamer, I don't give a shit where a game comes from. Speaking as a developer, I'll simply make good games. I take too much pride in my own work to release some shit with my name attached to it.

you should try setting shading to smooth and adding a edge split modifier if thats blender
makes cylinders smoother and stuff

well, aside from the fact that i'm looking for a writefag, i did some minor stuff today
i condensed the "SetAIBehaviour" and "SetNextEnemy" functions into one function, which also immediately starts the next fight, rather than having you go near the ring. I'm doing this since i think it's just more convenient, + it makes the tutorial a little bit faster
and i've finally figured out how i want the defense system to work. Initially i wanted it to be like in MGR, but with the way animations and ragdolls work in here, i don't think i'd end up doing it that well
so instead you've simply got 3 tiers of defenses
1st tier is dashing - safest, since you're literally gonna be too far for the enemy to be able to hit you. but costs the most energy
2nd tier is weaving - if you successfully do it, you can completely save yourself from any enemy damage, but it still costs some energy, and you have to do the right weave, otherwise you don't get the invincibility bonus during it
3rd tier is blocking - if your opponent hits your arms, he only does 1/3 damage to them. but if he's doing an uppercut, you get stunned. doesn't cost energy though
i also added the controls in hitler's dialogue after you join the gym
i also fixed the "continue" button at the main menu, now it actually continues the last saved/loaded savefile
so for tomorrow, i'll probably just make the ingame menu, to have the settings menu and other buttons, rather than just a save/load one. I'll probably also make more AI behaviours for the fights

I'm glad I have finished that one at the very least. Will probably upload a webm of it in the next thread.

Right now I am fighting with a bug that seems to happen semi-randomly when loading items after they've been carried across scenes. I've spent a lot of time debugging it and I think I finally nailed it down. Constraint satisfaction problem. Fucking "RequireComponent" attribute for classes.
I just hope I'll be able to solve it without having to go all-out and implement a variation of the fucking Waltz algorithm.

Question I'm making a vn here, a dating sim that emphasizes the sim aspect, but I don't know what I should do about the resolution. The background art is planned to be in 1080p, but will that scale well for someone who wanted to change the resolution? And what size should the characters be? I am using ren'py as this is mostly a proof of concept work, but still wish to get the most out of it.

But GameMaker is shit

Aren't VNs typically at a fixed resolution anyways? If you allow changing the resolution I don't think you have to worry about downscaling anyways, especially if you have the art at 1920x1080

Here's more shit I'm going to force upon you.

And one more.

wut engine is gud

The game I'd like to make is going to be a 2d real time 1-10 player massive unit sort of 'rts'. The focus would be having lots and lots of units on screen

sounds pretty good huh

Pimp hand Plutonia Adventure Extreme

I'm really at a loss for what to do. It's frustrating because I can't see what's wrong.

copy it

don't copy it, the problem might be something that could come up later, just look at it real hard until you get it.
even if you feel like a retard once you find the problem, you'll be a better programmer in the long run.

I've spent two days debugging it to no avail. Maybe I should just get working on other shit instead.

Wouldn't be the first time I've fucked up something stupid and felt like a retard in the end

I'm saying copy it into your code, see if it works - if it does, then the code's the problem, if not, it's something else.

I did try that, and it did not help which has just confused me further.

perhaps your reload animations are missing a tag or key that makes the netcode realize what it's doing? Look over your other animations and the code that handles them and try and figure out what's missing.

You know what, whining here does help. I just worked it out. There was a fuck up in the code where it wasn't getting the reload animation duration of the server from the client properly. So it was playing after all, it just had an animation length of 0. Now I've got another strange problem on my hands but at least I know what my error was.

Sitting spots also new thread soon?

And that's fixed now too. For some retarded reason, the Montage_Play function takes in a play rate to play the montage with, and return the duration of that montage. But it only returns the default duration, rather than the duration / play rate.

There you go, random strangers who probably know less code than you do helped.

WHOOPS

>not just using CGPeers

wew

Thanks, /agdg/eeks.

Install Gentoo

only relevant image I have, heh

...

soundcloud.com/zawazawazawachameleon/misty-battlefield-pre-mastered

Finished this for the guy who asked, hadn't EQ'd anything yet, or added guitar (which I may just record myself) I'm also trying to synth some Celtic flute samples to no real avail, so I may just ask a buddy to record some.

Did you set up a networking subsystem in [projectfolder]/config/DefaultEngine.ini? Take a look at theirs and compare it to yours. That fucked me up at first, as well as a bug because of a missing windows update.

Yeah I've done that. It's all good now, the error was because I was using an unreal function and making assumptions about it rather than actually looking at the source code of it.

Shit you're pretty good. That last buzzy boppy track you shared did not do you justice.

Please make a track for my game

Send request over Soundcloud and I'll take a peak.

(Also, yo..you too)

fug i just realized
since my combat AI isn't hardcoded now, i can just set story variables from there, rather than hardcoding them.
so now i should have an easy time doing shit like tracking how many times you got hit, or stopping fights when a boxer reaches a certain health

opinions on going through "tricks of the windows programming gurus" in the current year? It's in the gentooman library. I'm attracted to it by the simple tools he uses and how far he manages to go with them, also there are zero other books that intense that have been made since. Already set up a windows XP VM with vc6 and dx8 and stuff is working gud

go ahead fam.

It's not SaaS, at least not in the fucked up Adobe-kind of way.
It's essentially a payment of installments. After you paid it off, you can keep the using the software indefinitely. You can pause and resume the payments at any time and, if you have a bit more disposable income in a month, just pay the rest off and be done with it whenever you feel like it. The money you paid is also added to your account over there so you can buy select pieces of software with it. Say you have spent enough money to afford B2M, you can then just use that money to buy a license and your monthly payments will be reduced by however much you paid for B2M before.

The reason I don't pirate this is because I don't want it to bite me in the ass later down the line. I mostly use software that is free (i.e. Blender) or that I've gotten for free. Substance is one of two exceptions, both of which are

But hey, to each their own…

Use multiplatform libraries and code in a platform-agnostic way. Use opengl instead of directx. Don't be a wangblows cuck.

I don't want to be, nothing I make myself would be windows dependent. According to the author moving from the book's materials on to openGL would be a "snap".

Fair enough then. Though afaik there are some differences. I'd say you're better off learning opengl from the get go, but I haven't read that book myself.

stop using this thread, it's almost dead

I hope I link this right

but it's just another shitty OP with no information in it?

yeah, but I don't even give a fuck, I just want to be praised by my progress
lucky me IDs are for threads not for boards