next »

[bateleur]

Right now I am programming the same thing I programmed two months ago and apparently got so completely wrong last time that I'm too amused by it to be horrified.

Aww, look at cute little previous-version-of-me, it's trying to code things! Bless!

[Prinsessa]

Fixed up a few bugs in my level editor. Now I'm going to make sure it stores which layers are locked in the project files, so that they are still locked when I re-load the level. After that, I'm going to polish the paste mechanism a little, and then it's on to custom text fields.

[Ludophonic]

Gamepad support under windows using Raw Input / Hid

[MikeSw]

First wannabe iteration of a community management SaaS.

[SplinterOfChaos]

I found out that the fallowing code creates a segfualt when compiled (with gcc) 64-bit, but runs fine otherwise.

Code:

struct S
{
    std::string s1, s2;
};

S s = { "abc", "def" };

I hadn't worked on my project for over a month and all-the-sudden it turns out my problem just had to do with initializing a variable. I now use ctors, which maybe i should have in the first place.

[Prinsessa]

I found out that the fallowing code creates a segfualt when compiled (with gcc) 64-bit, but runs fine otherwise.

Code:

struct S
{
    std::string s1, s2;
};

S s = { "abc", "def" };

I hadn't worked on my project for over a month and all-the-sudden it turns out my problem just had to do with initializing a variable. I now use ctors, which maybe i should have in the first place.

Yeah. This method is only safe with POD, as far as I know.

[Taiko]

Just finished documenting/cleaning up an implementation of the Boids algorithm.  It's a neat little piece of code - I'm constantly amazed at how complex behaviors can result from simple rules.  I want to put this in a game sometime (Half-Life 1 used it for the flying guys in the Xen homeworld).

If anyone here runs MacOSX you can find the XCode source code here and the binary here.  Hopefully everything in the source code makes sense!

[st33d]

Interface.

So exciting...

 

[st33d]

Okay. Now I'm doing a level in the game which is going to be a game within a game.

[JMickle]

I get really annoyed that computers in games are never anything exciting, so I have made a (pretty simple really, but complex by comparison) "proper" computer in my new game, complete with a filesystem and memory slots. Works nothing like a real computer, but is far more interesting than the computers you find in games like fallout3 etc.

out of interest, are there any games which DO have interesting computer systems?

[Pishtaco]

I get really annoyed that computers in games are never anything exciting, so I have made a (pretty simple really, but complex by comparison) "proper" computer in my new game, complete with a filesystem and memory slots. Works nothing like a real computer, but is far more interesting than the computers you find in games like fallout3 etc.

out of interest, are there any games which DO have interesting computer systems?

The IF game Jigsaw has an (I think simplified) implementation of an enigma machine.

[Ludophonic]

Now  gamepad support under OS X with the HID Manager 

[tergem]

Perlin noise... my first attempt with it so I expect to learn a bit.

[iffi]

I get really annoyed that computers in games are never anything exciting, so I have made a (pretty simple really, but complex by comparison) "proper" computer in my new game, complete with a filesystem and memory slots. Works nothing like a real computer, but is far more interesting than the computers you find in games like fallout3 etc.

out of interest, are there any games which DO have interesting computer systems?

Call of Duty: Black Ops has an easter egg that lets you interact with a terminal of sorts. I'm not sure how complex it is, having never played it before, but it might be interesting.

[Dacke]

There's always Introversion's hacking "simulation" game Uplink. As well as their upcoming game Subversion that promises to have some pretty complex systems in it (if it ever gets finished), I think actually running the computers that exist inside the game.

Then there are games like Kohctpyktop where you design integrated circuits (simulating small-scale computers)

[JMickle]

Uplink is fairly interesting, but Subversion so far is blowing my mind. The levels they are going to to make the world run itself are... astounding.

[SplinterOfChaos]

I'm working on creating waves and scaling the difficulty based on the player's score. First, the player accumulates score points based on the score and assign each enemy type a cost. I spawn a random enemy (right now using a lottery so i can fix the probabilities), decrement the spawn points by the cost of the enemy spawned, and if more points remain, i spawn another random enemy.

I get really annoyed that computers in games are never anything exciting, so I have made a (pretty simple really, but complex by comparison) "proper" computer in my new game, complete with a filesystem and memory slots. Works nothing like a real computer, but is far more interesting than the computers you find in games like fallout3 etc.

I share your pain. I think there was actually one game that inserted a browser and based the game around google. It was mentioned once on Xplay (a game review show on G4) three to five yeays ago.

But yeah, that really bugs me too. When a game limits your actions on physical things, it seems to make more sense. But limiting your actions of digital things is especially obtuse to our perception of reality since we're playing a video game, thus prooving the limitations are arbitrary.

I wonder if someone could make a game that has a computer running linux, perfectly full-featured, running through a virtual machine.

[madbranch]

I'm working on my C++ JSON reader/writer that we want to use in our game engine.
https://bitbucket.org/madbranch/jsonbox

I'm currently trying to give it ridiculous JSON input to make it bug lol

[ThemsAllTook]

I'm currently trying to give it ridiculous JSON input to make it bug lol

Neat. I wrote a json parser in C not too long ago. In case it's useful to you, here are the strings I used for my unit tests (one per line):

Valid json:

Code:

[]
{}
[0]
[-2.753]
[1e+2]
[0.2e-1]
[""]
["hello"]
[false]
[true]
[null]
[false, true]
[true, 3, null]
{"a": false, "b": true}
{"foo": 1, "bar": null, "baz": "test"}
[{}, []]
{"1": [false, 2, {}], "2": {"": []}, "3": null}
["\\"]
["\""]
["\b\f\n\r\t"]
["\u0020\u26A0\uD834\udd1e"]
["a\u0000b"]
["\/"]

Invalid json (comments for explanation, not part of input):

Code:

// <- Empty input
null
[
{
[1,]
{1: 2}
{"a"}
{"a": }
{"a": null, }
["a": null]
["]
{": 0}
{"": "}
[a]
{1, 2}
["\u012"]
["
"]// <- One entry with literal newline
[0xFF]
[1.]
[.1]
[01]
[-.1]
[-01]
[1.1.1]
[]a
{}[]
[, 0]
{ , "a": 0}
[1 2]
{"a": 0 "b": 1}

[madbranch]

That's nice thanks.
Trying them right now.
Among my tests were really big files with a lot of nested arrays and objects to see if I could make my algorithm crash with a stack overflow. It can go quite deep in the inception levels lol, but I'm looking to make my algorithm non-recursive.

I went with the philosophy to warn the user of the incorrect input with error messages, but I interpret what I can from the incorrect input. It has its advantages and disadvantages, but I wanted something simple to use and didn't want anything to do with exception and crashing management lol.