Wow, calculating shit with CUDA and rendering it to OpenGL is puzzling to say the least
[IMG]http://i.imgur.com/cacj7Ab.png[/IMG]
It's supposed to be nice and smooth, I bet I indexed something incorrectly or whatever. At least it doesn't crash and burn.
[Editline]May 5th[/Editline]
[URL="http://i.imgur.com/Y75ensy.png"]Much better[/URL], forgot that the shader needs to be aware it's being fed float4 values instead of float3.
Possibly the shader itself that wraps the color values or something like that, I'm expecting a fully continuous image...
[IMG]http://i.imgur.com/zuhfDsI.png[/IMG]
[IMG]http://i.imgur.com/eRVnmns.gif[/IMG]
i'm writing a third party server.
So I was wondering if you folks would be interested in my 24 hour Game Dev competition similar to Ludum Dare. I'd have a vote on the theme along with 1st, 2nd, and 3rd place prizes, with 1st place being something like a $40 Steam game.
I've always loved looking at the things people create on WAYWO and I figured this could be a fun way to give back to the community a bit, along with seeing what people can come up in 24hrs using a certain theme.
Any engine, language, and/or software could be used obviously.
[QUOTE=sarge997;47664526]So I was wondering if you folks would be interested in my 24 hour Game Dev competition similar to Ludum Dare. I'd have a vote on the theme along with 1st, 2nd, and 3rd place prizes, with 1st place being something like a $40 Steam game.
I've always loved looking at the things people create on WAYWO and I figured this could be a fun way to give back to the community a bit, along with seeing what people can come up in 24hrs using a certain theme.
Any engine, language, and/or software could be used obviously.[/QUOTE]
What would the policy on pre-existing code and/or frameworks, etc? For example, could I use the FSM library I already wrote?
[QUOTE=Sidneys1;47664567]What would the policy on pre-existing code and/or frameworks, etc? For example, could I use the FSM library I already wrote?[/QUOTE]
I'm not really sure just yet. But I know that base frameworks could be used, but most of your work should be from scratch.
[QUOTE=cartman300;47664149]Interesting project. What came to my notice some time ago is that there's no proper 3D "generic" map editors, will it be open source? It would be awesome if it had some sort of plugin API so people can write exporters for their own formats. I could help you with that.[/QUOTE]
I haven't actually thought of any of that yet.
The plan right now is to make a general purpose 3D game engine, which I could use for maybe making a game of my own.
This whole project is heavily influenced by the source engine and Hammer, as well as a few other editors that I've used in the past.
So perhaps once I get the bulk of the work done I will attempt to add some basic plugin support and even a vmf importer or something.
Working on organization skills, I think. I've compiled my project into a library and a header file instead of having a 250 line C file when it really doesn't have to be. Now I just rebuild the library whenever I add functionality. Im not good at make files though
Improved my [URL="https://github.com/Sidneys1/GFSM"]Finite State Machine[/URL] thanks to the input from <PsY>, it's now at least partially non-deterministic in that it's stack-based, and the best transition chosen is valid transition nearest to the current stack position. It's a lot easier to set up now though :)
[img]http://imgur.com/Zo8Qf4M.png[/img]
For example, in this case if the stack looks like this:
[code]
SettingsMenuState
PauseMenuState
InGameState
MainMenuState
[/code]
And the "return" transition is triggered, it will transition to PauseMenuState, not MainMenuState, even though both are valid. Then, PauseMenuState + "main menu" will pop off both PauseMenuState and InGameState to return to the nearest MainMenuState.
[url=http://www.wolframalpha.com/input/?i=most+calories+food]uhh wolfram you ok[/url]
Can I get some love from any HN or DN readers? [url]https://twitter.com/andrewmcwatters/status/595657658813784064[/url] Trying super hard to get some outreach going, and it's picking up! Grid now has [url=https://github.com/Planimeter/grid-sdk/issues]some issues[/url] contributed by non-Planimeter devs now.
[QUOTE=andrewmcwatters;47665451]Can I get some love from any HN or DN readers? [URL]https://twitter.com/andrewmcwatters/status/595657658813784064[/URL] Trying super hard to get some outreach going, and it's picking up! Grid now has [URL="https://github.com/Planimeter/grid-sdk/issues"]some issues[/URL] contributed by non-Planimeter devs now.[/QUOTE]
Maybe you should show some more pictures instead of expecting everyone that reads your website to download an executable.
Or maybe like a blog post or something talking about how awesome your thing is?
Aaand here's the above FSM implemented:
[vid]https://a.pomf.se/erggaj.mp4[/vid]
My two cents on custom engines:
[B]If you want to make a game: make a game, not an engine. There are mature options that weren't available a few years ago. Don't waste your time.[/B]
I've written my own engine! OpenGL ES 2.0 on Android with Java. It was an incredible learning experience, but it was not a productive experience. The first version was hobbled together about three years ago. Note: I've been mostly inactive over the last year so that's closer to two years; school, a new relationship, and some medical issues have taken almost all of my free time, but that's not a very interesting story :) The results? One game with 796 lifetime downloads.
Why? I wanted to port my Game Maker game to Android, and the original scope of my game was very limited. Marmalade was too big of a jump from Game Maker, Unity still had its huge paywall, and Game Maker Studio was only a rumor. I ended up using Java with the Android Canvas drawing methods. Eventually this wasn't enough so I learned OpenGL, and was on my merry way. Between graduating high school, starting college, and working summer jobs I didn't have much focused time to work. This lead to adding little features here and there until I had something surprisingly fleshed out. Not particularly user friendly, granted, but very functional. From the bottom up: a thread managed renderer, file IO, image/sound/font loader, [I]tons[/I] of drawing functions, particle system, GUI system, game state manager, and all the little strings in-between.
Great! Where does this leave me? I'm going to port my game to Unity instead of just reskinning.
My next project is going to be in Unity, and the reskinning process isn't going to play nice with the existing GUI. This means that it'll be easier to just rebuild it the new way from the ground up. Added benefit, I'll be on iOS too!After that process is finished, I think I'll open source the game engine. You can never have too many references. Positive and negative. I'm not sure where mine will be put on that spectrum, I've never really had anyone else look through it past the line count.
I feel that trying to throw a catch all "don't make an engine" clause for game development isn't going to be always true.
For instance in my case, I've been building my engine from the ground up for nearly a year so far (I've been attending lectures all year, so development time has been low).
People who have next to no knowledge of game development would benefit from this approach; they would learn [I]what [/I]systems are needed and then promptly build them.
By the end of it, you will know your own engine inside out, and should have less problems making a game of your own then jumping into someone else's engine and having to figure out their nonsense.
It has been a very educational experience so far, and I'm glad I started this project last summer.
Further, I'm still not certain on whether or not I actually want to make a game right away once this engine is complete; if I don't, then I at least have an engine that I could do something with - whether cleaning it up and making it open source, or whatever.
With that said, I'm sure there is a point where this isn't always practical. One could take this a step further and make their own rendering backend as opposed to OpenGL, or even make their own compiler, their own language, etc etc.
I always build engines and never the games that accompany them. Then I give up right before I start making the game and get another idea, to which I build an engine. Loop.
[QUOTE=Map in a box;47665854]I always build engines and never the games that accompany them. Then I give up right before I start making the game and get another idea, to which I build an engine. Loop.[/QUOTE]
Sounds like you need to build a more general multi-purpose engine then.
[QUOTE=Karmah;47665867]Sounds like you need to build a more general multi-purpose engine then.[/QUOTE]
My engines alternate the following languages: C++, LuaJIT, Java, and C. Ad infinitum.
[QUOTE=Karmah;47665809]I feel that trying to throw a catch all "don't make an engine" clause for game development isn't going to be always true.
For instance in my case...[/QUOTE]
No advice is universal and your motives belong to you.
This being said, you hold my exact opinion from two years ago.
I'm not here to piss on anyone's parade. I'm just sharing my experiences and hoping that others can learn from them :)
[QUOTE=andrewmcwatters;47665451]Can I get some love from any HN or DN readers? [url]https://twitter.com/andrewmcwatters/status/595657658813784064[/url] Trying super hard to get some outreach going, and it's picking up! Grid now has [url=https://github.com/Planimeter/grid-sdk/issues]some issues[/url] contributed by non-Planimeter devs now.[/QUOTE]
Yeah, I've been havin' a go at GRID to learn and also improve it whenever possible :)
yeah i know, and if you guys are mentioning this because of grid, i am working on a game, and have been slowly going at it for a while, so yeah
so please spare me posts about engine development, i'm trying the best i can. i mean, i think that advice goes for games as equally as it goes for engines, or any other project.
if you want to make something used by [i]anyone[/i] tell at least [i]someone[/i]
i want to make a [url=http://www.andrewmcwatters.com/vadventure/]game,[/url] so i'm doing that
it's built on an [url=http://www.andrewmcwatters.com/grid/]engine[/url] i think is good enough to tell others about
powered by [url=http://www.andrewmcwatters.com/axis/]authentication[/url] for developers who want to control their own cloud storage
with great help from [url=http://www.andrewmcwatters.com/planimeter/]friends[/url] who have volunteered their time to projects they thought were interesting
so i do it, and i tell others about it
its clearly not good enough to get a lot of attention yet, but i'm going to keep trying and working on it until it is. but i don't want to just make a game, i want to make an engine, and a microplatform, and build a team, too
so if you want to do those things, do them all, but sure, if you just want to make a game, just do that
[editline]5th May 2015[/editline]
[QUOTE=ryankingstone;47666245]Yeah, I've been havin' a go at GRID to learn and also improve it whenever possible :)[/QUOTE]
people like you are the reason why open source communities and dev communities like facepunch's are so great
feedback is awesome, but contributions are stellar. thanks for your reports from earlier, it means a lot to me
[QUOTE=andrewmcwatters;47666268]
...
so please spare me non-directed posts about engine development, i'm trying the best i can
...[/QUOTE]
Woah there buddy, that wasn't directed at anyone. Especially not you. You make awesome stuff.
I mention this because I wasted 2 years of free time on something that should have taken 6 months, and am regretting it.
[QUOTE=chimitos;47666305]Woah there buddy, that wasn't directed at anyone. Especially not you. You make awesome stuff.
I mention this because I wasted 2 years of free time on something that should have taken 6 months.[/QUOTE]
:( but you did it right? i've got a lot on my plate and i'm barely doing it. i would say if you completed what you had in mind, you've accomplished more than i have
[editline]5th May 2015[/editline]
Grid is on the front page of DN right now. HN is nigh impossible to penetrate listings. [url]https://news.layervault.com/[/url]
[QUOTE=andrewmcwatters;47666316]HN is nigh impossible to penetrate listings.[/QUOTE]
no, its not
[QUOTE=chimitos;47665624]My two cents on custom engines:
[B]If you want to make a game: make a game, not an engine. There are mature options that weren't available a few years ago. Don't waste your time.[/B]
I've written my own engine! OpenGL ES 2.0 on Android with Java. It was an incredible learning experience, but it was not a productive experience. The first version was hobbled together about three years ago. Note: I've been mostly inactive over the last year so that's closer to two years; school, a new relationship, and some medical issues have taken almost all of my free time, but that's not a very interesting story :) The results? One game with 796 lifetime downloads.
Why? I wanted to port my Game Maker game to Android, and the original scope of my game was very limited. Marmalade was too big of a jump from Game Maker, Unity still had its huge paywall, and Game Maker Studio was only a rumor. I ended up using Java with the Android Canvas drawing methods. Eventually this wasn't enough so I learned OpenGL, and was on my merry way. Between graduating high school, starting college, and working summer jobs I didn't have much focused time to work. This lead to adding little features here and there until I had something surprisingly fleshed out. Not particularly user friendly, granted, but very functional. From the bottom up: a thread managed renderer, file IO, image/sound/font loader, [I]tons[/I] of drawing functions, particle system, GUI system, game state manager, and all the little strings in-between.
Great! Where does this leave me? I'm going to port my game to Unity instead of just reskinning.
My next project is going to be in Unity, and the reskinning process isn't going to play nice with the existing GUI. This means that it'll be easier to just rebuild it the new way from the ground up. Added benefit, I'll be on iOS too!After that process is finished, I think I'll open source the game engine. You can never have too many references. Positive and negative. I'm not sure where mine will be put on that spectrum, I've never really had anyone else look through it past the line count.[/QUOTE]
Bad advice. You learn a lot by making an engine
When I was banned I went through my old programming threads.
[url=http://facepunch.com/showthread.php?t=1017177&highlight=]Jesus[/url] fucking [url=http://facepunch.com/showthread.php?t=978909&highlight=]Christ.[/url]
[QUOTE=geel9;47666918]When I was banned I went through my old programming threads.
[url=http://facepunch.com/showthread.php?t=1017177&highlight=]Jesus[/url] fucking [url=http://facepunch.com/showthread.php?t=978909&highlight=]Christ.[/url][/QUOTE]
Try going through your old facebook chat messages.
I wish facebook would let me delete that shit
[QUOTE=Rocket;47666881]You also learn a lot by making a game. It comes right down to what you want to do: work on an engine or work on a game.[/QUOTE]
no work on everything because yes
yolo
[QUOTE=proboardslol;47666853]Bad advice. You learn a lot by making an engine[/QUOTE]
I said that? You do learn a lot in the process, but if that isn't your goal then its misplaced effort.
[QUOTE=Rocket;47666881]You also learn a lot by making a game. It comes right down to what you want to do: work on an engine or work on a game.[/QUOTE]
There's another starting point: Wanting to make a game when there's no decent engine for your workflow. At all.
Unfortunately there's no good LÖVE-style text centric modular engine (framework) for C# (or .NET) as far as I can tell.
Everything I've seen so far is heavily art- and/or editor-focused and usually has really annoying dependencies and quirks to work around.
I can't seem to get into that kind of creative process, so I'm making something that behaves differently and is a bit more flexible.
Anyway, Effektor progress: [code]#version 110
uniform sampler1D sampler1;
uniform sampler2D sampler2;
uniform sampler3D sampler3;
uniform samplerCube samplerCube;
uniform sampler1DShadow samplerShadow1;
uniform sampler2DShadow samplerShadow1;
// Fragment shader compiled using Effektor 0.
void main()
{
vec4 sample1 = texture1D(sampler1, 1.0);
vec4 sample2 = texture2DProj(sampler2, vec3(1.0, 2.0, 3.0), 10.0);
vec4 sample3 = texture3DLod(sampler3, vec3(1.0, 2.0, 3.0), 20.0);
vec4 sampleCube = textureCube(samplerCube, vec3(1.0, 2.0, 3.0));
vec4 sampleShadow1 = shadow1DProjLod(samplerShadow1, vec4(1.0, 2.0, 3.0, 4.0), 20.0);
vec4 sampleShadow2 = shadow2DProjLod(samplerShadow1, vec4(1.0, 2.0, 3.0, 4.0), 20.0);
gl_FragColor = sample1 + sample2 + sample3 + sampleCube + sampleShadow1 + sampleShadow2;
}[/code] As you can see I've implemented the texture sampling funtions and custom named variables.
I still have to check again, but I [I]think[/I] that makes this feature-complete for GLSL 1.10 (except for [I]break;[/I] and [I]continue;[/I] which I have to think about how to properly set up assignment barriers on and possibly array assignments which I may have postponed (They aren't directly supported in this language version anyway, so I'll probably start supporting them with the version where they first are expressions and eventually backport them once I have a fallback mechanism.)).
[editline]6th May 2015[/editline]
I should check for naming collisions on custom names...
What you see is a bug in my example code defining that shader, not Effektor.
Fun fact: I can write Logo, but just when i'm stoned. I think it might be because i read the manual while i was stoned.
[editline]5th May 2015[/editline]
I tested this twice.
Sorry, you need to Log In to post a reply to this thread.
