Yeah, ordinarily I'm kind of a control freak, but I want to be able to play it on a GMA450, so the shaders are out of my hands for now.  FWIW I'm trying to keep things as standardized as possible by using, for example, GLM for vector math (more on that can of worms later).

shaders are a huge amount of fun to write. I'd recommend this book.

Coming from you, I'll believe it!  Thanks for the info and I'll check it out.

I have heard bits and pieces of advice about deprecated OpenGL functions.  As I understand it, OpenGL at one time supported a few fixed shader functions, and then, some time around version 2.0, started to migrate to programmable shaders.  It was around this point that they added a whole bunch of shit to the API, like having to create vertex buffer objects instead of calling draw functions.

I'd like to use the old (OpenGL 1.x) API because it's more intuitive to me, it's supported on netbooks and older laptops, and my game doesn't need programmable shaders.  Will this prevent my game from running on newer video cards and drivers?  Or is GL 1.x going to become one of those archaic features like the <blink> tag that are supported long after they've fallen out of fashion?

So I'm making a randomized racing game (details at my homepage URL), and I'd like to allow users to make their own tracks by piecing together chunks of roads & buildings, like a model train set.  It seems like the easiest way to do this would be RTS-style grid placement.  Does anyone know of any resources I could browse through before diving in?

Also, take a look at this:



I have heard that the collision geometry was all done with Chipmunk physics.  The graphics may be vector art or just hi-res textures (not sure), but they're not tile-based.

That being said, there are other reasons to have a tile-based world, like the simplicity of writing pathfinding, lighting and AI.

These are all great suggestions.  I notice that a lot of them are Nintendo titles - they must have a policy of not too much gore/grimdark in their games.

Serious Sam might actually be good.  The violence is really goofy and doesn't take itself too seriously.  I think it might be too hard for her since her reflexes haven't been honed by years of Quake.

That Zelda archery game looks awesome.  Almost makes me want to get a Wii U just to play it.  I'll see if I can find a clone somewhere.

Don't mean to clutter up the board, but I'm looking for a specific type of game.  My girlfriend loves playing Minecraft, but only the mob shooting parts - she'll farm for hours to make arrows and kill more mobs.  She'd really like a FPS, but she doesn't like all the blood and gore (me either, really).  I was wondering if there's a cartoonish FPS, preferably multiplayer, that has the same gameplay but isn't as violent or grimdark.

[Cafe Society]

Cafe Society
Video Clip:

I've been working on this project for the past year or so.  I figured this was a good time to post since I'm going to be making some major changes.

I got the idea after watching the "Substrate" screensaver based on Jared Tarbell's algorithm.  It generates realistic cityscapes by "growing" roads from a few initial seed points.  I kept thinking how fun it would be to drive through those cityscapes.


I wrote a similar algorithm, but then added the ability to turn the generated raster image into 3D geometry.  The algorithm is based on the streets of the Eixample district of Barcelona: every pixel on the bitmap corresponds to an octagonal street intersection.  Neighboring intersections are linked together to make a street network.


Buildings are made by finding all the polygons that represent areas completely surrounded by roads.  I offset them by about 2m and extrude by .25m to create a sidewalk, then extrude again to create a building.  Building heights are based on an exponential distribution: there are a few skyscrapers and many short apartment blocks.


I put different models on sidewalks for decoration.  One of these is the jukebox model.  If you get close enough and hit space, it allows you to pick a song from any that you dropped into your asset folder.  Then you have as long as the song lasts to drive through the city and "tag" one or more other jukeboxes.  There's a time counter that shows how many seconds are left in the song, and a distance counter that shows the distance to the next checkpoint. 

I keep track of every octagonal intersection and road, and use them to make a road connectivity graph.  I can then A* pathfind through this graph and figure out the optimal route to to a checkpoint and how long it should take to get there at max speed.  I use this information to procedurally generate races by setting a realistic time to complete them.  It also lets me mark the shortest route directly, instead of just pointing to the general location of the checkpoint as in "Crazy Taxi".

I'm looking for jobs at the moment, so I've postponed work on the project until my schedule is more predictiable.  My next steps are gonna be revamping the city algorithm (it's way too complicated and bug-prone now), and porting the game to C++.  Right now it's written in Panda3D, which is a Python interface to a precompiled engine.

The whole 'cover being blown' thing reminds me of The Ship

Oh man, I totally forgot about that game!  I saw it once on a European publishers' site.  Spy Party is kind of similar except for the "asymmetric multiplayer" aspect.

As the guy who made the thing, I'd like to hear what you mean with not very organized, critiques are good

/OT

Documentation documentation documentation.  I use tiddlywiki.  [example]

Yeah, look at Evan's site.  2 seconds after I load the page, I see a link to "Tutorial".  This leads to an example program that demonstrates plaid in 16 lines.  There's also a link, on the front page, to "API Reference", which lists member functions and variables for each class, and input parameters and return values for each function.  The whole library is divided into three modules, which can be used separately (using encapsulation appropriately) and have mostly one-word members (using name space appropriately).

The only documents I look at, when learning a library, are example programs and API references.  If the library uses concepts from another field of study, like physics or geology, I'll read the user manual (the Bullet user manual is a good example of this).  But most libraries don't need to introduce any new concepts, so a manual is a red flag that they are too poorly designed or documented. 

As far as OpenGL goes I found wrapping OpenGL itself to be mostly futile. OpenGL just doesn't wrap well in a object oriented way imho and if you try to preserve its structure you generally find that it is already pretty reasonable and that there isn't all that much to simplify. So you either have to wrap for your specific purpose or not at all imho.

I say this because I tried obviously: https://github.com/progschj/glp
But now all I do when I want to quickly try something in OpenGL is copypaste from my example collection and iterate from there.

Yeah, 3D rendering in general doesn't seem like something you should do in an OO manner, and it's a mental red flag for me if someone tries to do it. 

I see a lot of the same problems in C++ libraries as in Python libraries - the language is so powerful that it's easy to use features you don't need and make the project too complicated.  My rule of thumb is that I go to the maintainer's site and look for code samples.  If I can't find one, in under 5 minutes, that demonstrates some basic functionality in 25 lines or less, I look somewhere else.

They aren't hard to write but it takes a lot of time to write and debug properly. If you don't really need something special which cannot be done with an existing engine, then it's a bad idea to write your own engine (unless you don't want to make games but engines, then it's ok ofc).

I'm looking for a lightweight engine with few features that I can use for game jams or maybe even Ludum Dare.  I'm willing to do things radically differentl in order to make programming faster.  Most engines, though, are aiming for as many common features as possible.  Also, most engines:

• Don't encapsulate their classes or functions, which leads to bloat, more boilerplate code, and more complicated builds.
• Don't manage their namespace, which leads to overly verbose and confusing names.

Yes.

That sounds awesome.  Would run on my netbook too.  Downloading!

A well-played traitor game can end with the cast of players genuinely surprised to realize the traitor's identity, if they even realized a traitor was present. Sometimes, even, the traitor is the station's AI and then you get a Space Odyssey situation where the players are fighting against the ship's systems.

Also awesome.  That's exactly the feeling I'm trying to replicate.  But yeah, I agree it's difficult to get that kind of environment without other human players.

Not the easiest solution and you might not want to invest the time, but it isn't terribly hard to write your own things to make OGL easier.

Yeah, I'd rather write something myself than use a badly written library.  I'm starting to realize that if there are so many 3D engines available, they can't be terribly hard to write.

Instead of a gun getting pointed in the protagonist's face

No.

James Bond is OK, but I always liked John Le Carre movies.  The action is more psychological than physical.  Instead of a gun getting pointed in the protagonist's face, he might have the threat of getting his cover blown, and instead of seduction & assassination, he might have to concentrate on recruiting moles and figuring out people's loyalties.

Are there any games like this?  I think Tom Clancy produced one once, but it wasn't that great. 

Thanks, that's exactly what I'm looking for.  The "Dojo" engine posted earlier in this board is also a step in the right direction, although it doesn't look very organized.

On GLFW's site they recommend SFML as an alternative.  That's an excellent idea that I hadn't considered because I had this idea that SFML=2D.

I've been dicking around with various C++ libraries to find a good one for experimenting.  I'm looking for something similar to LWJGL: higher-level than raw OpenGL, but lower-level than a full game engine.  I don't need a scene graph, for example.

Two examples I've found are OGLplus and GLT.  I'll try them both out, but I was hoping someone one here has done it before, or found another library I've missed, and could give me some pointers.

Hi,

I couldn't really find an existing C++ Tweening framework that was both tiny and simple, so I build my own.

It's inspired by the Flash Tweener and CPPTweener and comes with all of Robert Penner's easing equations as defaults.

It's great.

I've had a hard time finding good code snippets in Java & C++.  I'm usually stuck with some 500mb spaghetti-code library that tries to do everything.  The best code I've read lately has been AS3...maybe they have a bigger community, or are into the demo scene or something.

I started using Java back when Minecraft was first becoming popular.  I looked at some of the developer's old code samples, and was just blown away by how well written they were.  I didn't think it was possible to write Java apps that were that concise, small and robust.  It made a huge difference in my programming skill, partly because I could reuse some of the techniques I learned, but mostly because I could recognize when the developer of a 3rd party library knew what he was doing.

Since then I've kept my eyes open for other projects that are similarly well-written so I can learn from them.  Some of the ones I've found are:

• Flashpunk
• SFML
• Almost any of the Xscreensaver hacks
• wxPython (probably wxWidgets in general)
• An AS3 GUI library called Minimal Comps ( http://blog.soulwire.co.uk/code/actionscript-3/simple-flash-prototype-gui-tool )
• "Blades of Exile" source (http://code.google.com/p/openexile/source/browse/trunk/Win32/Blades%20of%20Exile/boe.items.cpp )

Anyone have favorites of their own? I'm trying to get into C/C++ programming , so those would be particularly appreciated.