|
CaptainPoopsock posted:Good point I guess. I'll give it a few tries. I don't have much interest in Flixel or Game Maker, but I do have familiarity with C++ so I am diving into SDL to see how it is even though I realize this probably isn't going to help me get a game stood up in any short amount of time. That's really most of my problem. Plus, it's also good enough for making a 'real' game, so you don't feel like you're wasting your time learning some junk system that you'll never use outside of making tiny contest prototypes.
|
# ? Mar 13, 2012 03:08 |
|
|
# ? May 18, 2024 06:44 |
|
CaptainPoopsock posted:Good point I guess. I'll give it a few tries. I don't have much interest in Flixel or Game Maker, but I do have familiarity with C++ so I am diving into SDL to see how it is even though I realize this probably isn't going to help me get a game stood up in any short amount of time. That's really most of my problem. Maybe try SFML. SDL is built for C and is a little less user-friendly. You can even use SFML as a windowing thing for OpenGL just like SDL if you want to. Or maybe just do what tuna said. He's less of a noob than me.
|
# ? Mar 13, 2012 04:09 |
|
ambushsabre posted:Ok since I'm impatient and bored I threw together the thread. You can look at it here. I choose friday as the start / end date because that gives the weekend and friday night for people to work on their games, and it's also something to kind of look forward to at the end of the week. If you have any suggestions please let me know! I'll probably just put the week's theme at the top with 's all around it. I'd love see FlashPunk get equal billing with Flixel; they are pretty feature-comparable, but have a different enough style that some people might find one or the other more intuitive. FlashPunk just made more sense to me when I was an AS3 noob, plus there are a ton of good tutorials and screencasts on the FlashPunk site. quote:If you're ok with learning how to program, use Flixel or FlashPunk
|
# ? Mar 13, 2012 05:34 |
|
http://www.33rdsquare.com/2012/03/ai-creates-video-game.html Well pack it up. It's been a good run everyone.
|
# ? Mar 13, 2012 11:31 |
|
I'd like to put my hat into the ring if anyone needs sound/music. Also, since these are small indie titles (i.e. - not commercial) I would absolutely LOVE it if someone would implement FMOD into their game. That way, I can pretty much program all of the audio from my end, rather than just design it and send the ogg files off to you. But I'm up for doing it that way, too. PM me or email. I don't think I'm ready yet to be developing something on my own in a week.
|
# ? Mar 13, 2012 13:19 |
|
PalmTreeFun posted:Maybe try SFML. SDL is built for C and is a little less user-friendly. You can even use SFML as a windowing thing for OpenGL just like SDL if you want to. Hey, thanks for this. SFML looks right up my alley and super easy to use. I really have nothing against using any of the engines such as Unity (I actually am in between working on some Unreal projects) but I have always wanted to put something cheesy together from scratch. Looks like SFML is going to be a great starting point.
|
# ? Mar 13, 2012 15:18 |
|
SynthOrange posted:http://www.33rdsquare.com/2012/03/ai-creates-video-game.html Actually, that's a really cool project and props to the developer. I just hope technology like this isn't used as a shortcut to develop even more mediocre and lazy games. Until AI is as sophisticated as humans (we've got a ways to go), it always degrades the quality of the game.
|
# ? Mar 13, 2012 16:10 |
|
Orzo posted:Ah yes, running around in huge barren procedurally generated worlds with no design/purpose/soul, like Skyrim or Fallout, wasn't enough. Let's automate the entire process! Simple example - Infamous. Did you know that city was procedurally generated? The quests, not so much, but for the world, they used hex-based Wang tiles. Similar examples abound in anything where you're trying to add reasonable detail on a massive scale (throwing down vegetation on a height map for instance). If you had a similar tool that could pepper down NPCs and basic interactions between them? That would be pretty neat. Especially if it could pepper them down in a way that could be later influenced via player interaction by tweaking the NPC procedural basis. Bandit camps folding up and moving on as you take out their leaders, towns rising and falling, all kinds of neat potential there.
|
# ? Mar 13, 2012 16:17 |
|
I actually haven't played Infamous, so it's hard for me to respond properly. All I know is that for every game I've ever played that had some procedurally generated aspect, that part of the game has felt boring, lazy, and soulless. The world in Fallout. The weapons in Borderlands. The dungeons in Torchlight. Perhaps I'm alone in this and I'm showing my age. But did you read that article that breaks down how precisely crafted everything is about Super Metroid? I doubt you would ever get anything like that from an AI at the current level of AI sophistication.
|
# ? Mar 13, 2012 16:28 |
|
Shalinor posted:Procedural generation is an excellent tool, it's just all in how you use it. http://www.amazon.com/Texturing-Modeling-Third-Procedural-Approach/dp/1558608486/ref=dp_ob_title_bk This book is pretty awesome if you are looking for a good discussion of 'procedural stuff'.
|
# ? Mar 13, 2012 16:39 |
|
Orzo posted:I actually haven't played Infamous, so it's hard for me to respond properly. All I know is that for every game I've ever played that had some procedurally generated aspect, that part of the game has felt boring, lazy, and soulless. The world in Fallout. The weapons in Borderlands. The dungeons in Torchlight. The problem with an AI developing a game is that you have to quantify "fun" and then create heuristics that create more fun. It's kind of like trying to teach a robot how to paint. You can tell it how to make the strokes, but you'll never be able to convey the concept of "art" to a machine.
|
# ? Mar 13, 2012 16:43 |
|
poemdexter posted:The problem with an AI developing a game is that you have to quantify "fun" and then create heuristics that create more fun. It's kind of like trying to teach a robot how to paint. You can tell it how to make the strokes, but you'll never be able to convey the concept of "art" to a machine. Your heuristics for quantifying fun can be an art all their own, though.
|
# ? Mar 13, 2012 16:46 |
|
Orzo posted:I actually haven't played Infamous, so it's hard for me to respond properly. All I know is that for every game I've ever played that had some procedurally generated aspect, that part of the game has felt boring, lazy, and soulless. The world in Fallout. The weapons in Borderlands. The dungeons in Torchlight. This is one of the things which kept me playing nethack for like 15 years. If you somehow haven't played nethack, most of the floors are procedurally generated, but every couple of floors you would run into floors that had a fixed layout (or several fixed layouts it would pick from randomly). But even the procedural floors had enough bits and pieces to use that they didn't feel soulless. Between shops, altars, beehives, throne rooms, etc, you could usually recognize specific floors when you have to come back up. And that's not even taking into account the different themes of the procedural floors (regular dungeon, classic dungeon, the gnomish mines, gehenna, etc). So if you really make an effort, procedural is great. If all you do is generate random terrain and mash trees and bushes on it, then, like you say, it kind of sucks, though.
|
# ? Mar 13, 2012 18:28 |
|
SynthOrange posted:http://www.33rdsquare.com/2012/03/ai-creates-video-game.html Re: Procedural generation, the challenge is that procedural generation is a great tool for keeping things from being overly repetitive, but that isn't the same as making them interesting. It's useless to create a bunch of content if that content is just going to be filler between the player and points of interest, you have to make sure that the content has something to do in it. Random events might be a good idea. poemdexter posted:The problem with an AI developing a game is that you have to quantify "fun" and then create heuristics that create more fun. It's kind of like trying to teach a robot how to paint. You can tell it how to make the strokes, but you'll never be able to convey the concept of "art" to a machine. OneEightHundred fucked around with this message at 18:53 on Mar 13, 2012 |
# ? Mar 13, 2012 18:41 |
|
Some games - roguelikes, mainly - do it by tuning the mechanics so that any possible output is sufficiently interesting content. Not all types of game can work well with the sort of super-dense style that implies, however.
|
# ? Mar 13, 2012 18:44 |
|
Orzo posted:I actually haven't played Infamous, so it's hard for me to respond properly. All I know is that for every game I've ever played that had some procedurally generated aspect, that part of the game has felt boring, lazy, and soulless. The world in Fallout. The weapons in Borderlands. The dungeons in Torchlight. The entire game of Minecraft.
|
# ? Mar 14, 2012 00:14 |
|
poemdexter posted:The problem with an AI developing a game is that you have to quantify "fun" and then create heuristics that create more fun. It's kind of like trying to teach a robot how to paint. You can tell it how to make the strokes, but you'll never be able to convey the concept of "art" to a machine. Why not?
|
# ? Mar 14, 2012 00:46 |
|
John Von Neumann posted:You insist that there is something a machine cannot do. If you tell me precisely what it is a machine cannot do, then I can always make a machine which will do just that.
|
# ? Mar 14, 2012 00:57 |
|
^^^^ That's awesome, and since I frequently find myself making that argument I will have to use that quote.
|
# ? Mar 14, 2012 01:33 |
|
http://vimeo.com/36579366 This is a pretty interesting presentation on "reactive" development, that is, dev tools which give immediate feedback and tweakability, and aren't beholden to a write -> compile -> adjust loop. There are some very cool Javascript demos, and even if you don't watch the whole thing, check out the part at around 10:30-16:00, which is game design specific.
|
# ? Mar 14, 2012 01:46 |
|
h_double posted:http://vimeo.com/36579366 Bret Victor is one of those guys who makes me feel like I'll never do enough in my life. "Oh, no bigs, I basically designed the UIs for some of the most popular devices ever created. I fill my time with rad research projects, hack hardware, and also happen to be incredibly musically talented. Whatevs." I can't for the life of me think of a use-case for the demo'd tool, but it's awesome all the same.
|
# ? Mar 14, 2012 01:50 |
|
The Radix posted:Why not? Because what's fun to you might not be fun to me. However 5 to you will more than likely be 5 to me (please don't be pedantic to prove me wrong). But I guess that begs the question "Why does it have to be fun for everyone?" So maybe I'm wrong and you can program "fun" but it just seems weird to be able to program an adjective that is a sum of the whole piece of software. h_double posted:http://vimeo.com/36579366 http://www.ncrunch.net/ Unit tests without having to compile and run!
|
# ? Mar 14, 2012 01:58 |
|
I think Victor's tools are really cool, but all of the ideas he shows off for making coding more visual and providing immediate feedback seem extremely case-specific. Ok, I can easily see the path of an object through time and thus tune my platformer physics. What if I need to track several objects at once? I probably need to be able to specify a subset of the active game objects. How do I expose non-visual information like the state machine driving enemy NPCs? When you start trying to make a general purpose tool that can express everything code creates the complexity of your UI and visualizations shoots through the roof. For me, at least, the most important aspects of programming take place in my skull long before I touch a keyboard or mouse.
|
# ? Mar 14, 2012 01:59 |
|
poemdexter posted:Because what's fun to you might not be fun to me. However 5 to you will more than likely be 5 to me (please don't be pedantic to prove me wrong). But I guess that begs the question "Why does it have to be fun for everyone?" So maybe I'm wrong and you can program "fun" but it just seems weird to be able to program an adjective that is a sum of the whole piece of software. Specifically though, you said it wouldn't be possible to convey the concept of art to a machine. But I think you're missing that it already exists in a quantified form in your own mind, or else you wouldn't be able to process it either. "Art" and "fun" are fuzzy concepts, sure, but we humans manage to perpetrate them sometimes anyway.
|
# ? Mar 14, 2012 04:17 |
|
Internet Janitor posted:von Neumann quote
|
# ? Mar 14, 2012 04:18 |
|
The point of the quote is not to imply that replicating creativity is a simple matter, but rather that the difficulty lies in precisely quantifying what we mean by "creativity". I don't think it's a binary yes/no capability, but rather a continuous landscape. Programs that can achieve the nuance and depth of human level designers will need to be staggeringly complex and incorporate all sorts of domain knowledge. Some procedural content generation techniques are just warmed-over RNGs, but if you look at games like Dwarf Fortress, a huge amount of effort goes into physical simulation at varying levels of detail to create worlds with consistent geological history- that's working in the right direction.
Internet Janitor fucked around with this message at 05:23 on Mar 14, 2012 |
# ? Mar 14, 2012 04:40 |
|
No problem can be solved without first defining the problem in such a way that you can measure the effectiveness of the solution. Creativity hasn't been quantified, but there are several widely-used proxies for it; a machine can optimize for those proxies, crude as they are. Machine learning hasn't been widely applied to game development, but it's going to become more common as game metrics increase in prominence. That particular field is mostly just really advanced statistics, so you need large datasets to use it effectively.
|
# ? Mar 14, 2012 13:09 |
|
Internet Janitor posted:
I've heard people using that quote in some way before, and noone ever was able to build me a machine that's capable of genuinely enjoying drinking a steaming hot cup of coffee. On a more serious note, what's up with the jam plans? Are you supposed to work in teams or are you able to go solo if you want? Any idea's on how to form teams?
|
# ? Mar 14, 2012 13:47 |
|
PirateMathew posted:I've heard people using that quote in some way before, and noone ever was able to build me a machine that's capable of genuinely enjoying drinking a steaming hot cup of coffee. I will program a robot that pours coffee all over itself and then give a delightful "ding!" sound effect to signal that it is genuinely enjoying it. I'll even create a variable called genuinely_enjoying and set it to true when the coffee is poured on itself. As for the teams, if it works anything like the SA game dev competitions, it's completely up to you. Solo, groups, some with multiple devs, some with a single dev and single artists, it's whatever. I don't think they'll be any hard and fast rules besides stick to theme and you have 1 week.
|
# ? Mar 14, 2012 14:13 |
|
poemdexter posted:I will program a robot that pours coffee all over itself and then give a delightful "ding!" sound effect to signal that it is genuinely enjoying it. I'll even create a variable called genuinely_enjoying and set it to true when the coffee is poured on itself. poemdexter posted:As for the teams, if it works anything like the SA game dev competitions, it's completely up to you. Solo, groups, some with multiple devs, some with a single dev and single artists, it's whatever. I don't think they'll be any hard and fast rules besides stick to theme and you have 1 week. Word. Stock sprites and ripped models it will be then
|
# ? Mar 14, 2012 14:19 |
|
PirateMathew posted:Word. Stock sprites and ripped models it will be then Allow me to introduce you to every game dev's sprite sheet: http://forums.tigsource.com/index.php?topic=8970.0
|
# ? Mar 14, 2012 14:26 |
|
poemdexter posted:Allow me to introduce you to every game dev's sprite sheet: http://forums.tigsource.com/index.php?topic=8970.0 Allow me also to submit the following http://sdb.drshnaps.com/ As long as no one tries to make money off of the sprites the majority of them seem to be free. I figure it is resource worth having.
|
# ? Mar 14, 2012 14:58 |
|
http://roguebasin.roguelikedevelopment.org/index.php/Finding_graphical_tiles Has a decent selection of tiles as well I don't know if the OP still posts here but updating it would be cool too. Links to tutorials for people, popular frameworks, an FAQ, etc.
|
# ? Mar 14, 2012 15:41 |
|
I guess I'm missing the spirit of it, but that von Neumann quote is kind of wrong.
|
# ? Mar 14, 2012 16:22 |
|
The discussion has strayed into academia. The point was that nobody is fooled by AI because AI still sucks compared to the human brain, and games that use procedurally generated assets where those assets could have been hand-designed suffer as a consequence. I think developers are so attracted to procedural generation because let's face it, it's elegant and cool from a coding/design perspective. Unfortunately, elegant or innovative code isn't what makes a game great.
|
# ? Mar 14, 2012 16:30 |
|
Orzo you're completely missing the point about procedural generation, which is that it enables genres that simply could not exist otherwise. People like roguelikes. People like games like Just Cause 2, with massive worlds that can only exist because they're procedurally generated and then fine tuned by hand. If you're going to compare every game to the standard set by something where every single pixel is carefully hand-crafted, then you're going to completely eliminate entire genres. At the end of the day Super Metroid is loving great but after you've beat it then there's nothing to do but beat it again and it's exactly the same. If it's a procedurally generated world then the experience isn't nearly as good, but you can play it forever and it'll be new every time. It's a tradeoff and anyone who has ever worked on a game with procedural elements is incredibly aware of that.
|
# ? Mar 14, 2012 16:38 |
|
EDIT: ^^^ Orzo posted:I think developers are so attracted to procedural generation because let's face it, it's elegant and cool from a coding/design perspective. Unfortunately, elegant or innovative code isn't what makes a game great. A game that could not have been made without procedural techniques doesn't "suffer" - it exists where it otherwise would not. Infamous is one such game. There are plenty of others, and especially in the indie sphere that becomes an important consideration. EDIT: VV Ah, gotcha. Shalinor fucked around with this message at 16:44 on Mar 14, 2012 |
# ? Mar 14, 2012 16:39 |
|
I'm not missing the point. I conceded in my original post that I'm probably just showing my age. I understand that people like games like that, it's just that I don't. I also understand that it enables entirely new genres and gameplay elements to exist, but I end up disliking those most of the time. I also would never compare every game to some standard as you suggested Paniolo--I was just using Super Metroid as an example, not as something to be emulated by every game.
|
# ? Mar 14, 2012 16:44 |
|
Orzo posted:I think developers are so attracted to procedural generation because let's face it, it's elegant and cool from a coding/design perspective. Unfortunately, elegant or innovative code isn't what makes a game great. Terrain generation is a huge one, having fine detail and realistic erosion is a major part of what makes terrain convincing, but adding hundreds of erosion lines to a mountainside by hand does not contribute significantly compared to just having World Machine do it and is a waste of your artist's time. As far as the fascination with the increasing scope of it, like basing gameplay on it, it's largely the same issue: Do players in your sandbox game care that Randomly Generated Side Mission isn't radically different from the other ones, or are they glad you've given them something nifty to do outside of the main storyline? Multiplayer is a similar concept, the amount of unique content you will see over a month of playing multiplayer is pathetic, but the number of unique ways that content will combine to give you a new experience is almost limitless. That's what procedural generation aims to achieve. OneEightHundred fucked around with this message at 17:11 on Mar 14, 2012 |
# ? Mar 14, 2012 17:08 |
|
|
# ? May 18, 2024 06:44 |
|
You've given good examples of another reason it's a useful tool and I don't disagree with what you've written, but to say 'this isn't true at all' is wrong. From reading tigsource, gamedev, here, and some random dev blogs, I've come across plenty of examples of developers who use random content generation because it's a neat thing to do, not because it necessarily makes their job easier. Whether or not those people are crazy or just really enjoy interesting programming challenges is a separate issue, but it does happen.
|
# ? Mar 14, 2012 17:22 |