This is just me screwing around, really…. This implements a counting bloom filter in native php. I read about bloom filters this morning, and wrote this tonight in between playing Command and Conquer 3 campaign missions. The hashing function is configurable (and easily extensible,) rehashing the key multiple times to reduce the chance of collisions is also configurable (though I’m not entirely sure how needed this is.) Frankly this all might not even be (or work) properly… seems to… we’ll see…

< ?php

class bloom {

	/**
	 *
	 * A counting bloom filter
	 *
	 *	$f = new bloom(10, 'md5');
	 *
	 *	$f->add('foo');
	 *	$f->add('bar');
	 *	$f->add('foo');
	 *
	 *	$f->exists('foo');        // true
	 *	$f->exists('bar');        // true
	 *	$f->exists('baz');        // false
	 *	$f->exists('foo', false); // 2
	 *	$f->exists('bar', false); // 1
	 *	$f->exists('baz', false); // false
	 *
	 **/

	var $number_of_hashing_functions;
	var $map = array();
	var $hashlen = 32;
	var $hashfunc = 'md5';

	function __construct( $functions=1, $hashtype='md5' ) {
		$this->bloom($functions, $hashtype);
	}

	function bloom( $functions=1, $hashtype='md5' ) {
		$this->number_of_hashing_functions = (int)$functions;
		if ( !$this->number_of_hashing_functions || $this->number_of_hashing_functions < 1 )
			$this->number_of_hashing_functions = 1;
		$this->set_hashing_function($hashtype);
		$this->initialize_bitmat($this->number_of_hashing_functions);
	}

	function set_hashing_function($method) {
		switch ( $method ) {
			default:
			case 'md5':
				return false;
				break;
			case 'sha1':
				$this->hashlen = 40;
				$this->hashfunc = 'sha1';
				break;
		}
	}

	function hash($key, $n=0) {
		return call_user_func( $this->hashfunc, $n.$key );
	}

	function add($key) {
		for ( $i=0; $i< $this->number_of_hashing_functions; $i++ ) {
			$k = $this->hash($key, $i);
			for ( $n=0; $n< $this->hashlen; $n++ ) {
				$this->map[$i][$n][$k{$n}]++;
			}
		}
		return true;
	}

	function remove($key) {
		for ( $i=0; $i< $this->number_of_hashing_functions; $i++ ) {
			$k = $this->hash($key, $i);
			for ( $n=0; $n< $this->hashlen; $n++ ) {
				$this->map[$i][$n][$k{$n}]--;
			}
		}
		return true;
	}

	function exists($key, $bool=true) {
		$max = 0;
		for ( $i=0; $i< $this->number_of_hashing_functions; $i++ ) {
			$k = $this->hash($key, $i);
			for ( $n=0; $n< $this->hashlen; $n++ ) {
				if ( !$v = $this->map[$i][$n][$k{$n}] )
					return false;
				else
					$max = max($v, $max);
			}
		}
		if ( $bool )
			return true;
		else
			return $max;
	}

	function initialize_bitmat($n) {
		$empty_bitmap_line = array(
			'0' => 0, '1' => 0, '2' => 0, '3' => 0, '4' => 0, '5' => 0, '6' => 0, '7' => 0,
			'8' => 0, '9' => 0, 'a' => 0, 'b' => 0, 'c' => 0, 'd' => 0, 'e' => 0, 'f' => 0  );
		$this->map=array();
		for ( $i=0; $i< $n; $i++ ) {
			$this->map[$i]=array();
			for ( $l=0; $l< $this->hashlen; $l++ ) {
				$this->map[$i][$l] = $empty_bitmap_line;
			}
		}
	}

}

?>

That lisp looks like it might be an interesting language to learn?

There’s been a lot of hooting and hollering in the media about copy protection and piracy.  DRM seems to be a rallying cry for the younger generation.  Which makes sense.  I figured I’d throw my own $.02 out there to the wind (for all the good its worth) But I’d first like to say that (as is true with all generalizations) this doesnt apply to all people all the time.

You really have 3 classes of people who might buy a game.  You have the people with enough disposable income not to care, they just buy their game, and use it as directed.  (please not I’m not discussing the as directed part, simply the who’s yelling part. I fully agree that as directed should come outside the bos — or be returnable — and should also be reduced to lay speak since most of us dont speak asshole lawyer.)  Those people really arent who’se yelling about DRM.

You have the gaming middle class who, more or less, wants to go straight.  They buy what they can when they can, but at $60 a pop they cant afford *everything* they want.  To varying degrees they have a couple of options.  They can pick what they *think* they’ll like best, and abstain from everything they cant afford, or they can “test-run” the game and see whether they like it first.  As someone who has bought a car before I know which one of those options makes the most sense to me… This stage is definitely not a plateau, though, it slopes down from the previous upper class to the next… lower class.

Finally there’s the poor people.  They have a computer.. maybe handed down.. maybe cannibalized…  Maybe they’re preteen to mid/late twenties.  Maybe they’ve got a job, maybe they dont. Maybe they’re in school, maybe not.  But the common denominator here is that you arent going to see a red penny from these people no matter what you do.  One does not squeeze blood from a stone.  These people have WAY more time than they have money, and so spending 2 days downloading, and 2 days searching for an install key, then a crack, is NOT a problem for them… It’s cheaper than buying the thing.  What, I think, game companies are missing is that a good number of these people WANT to be in the middle and upper gamer class, and when they get there they *will* spend money, because they’ll have more of that then time.

What I think *IS* interesting is the time to money ratio.  You develop a game that might take 10-30 hours to beat (in the case of non mmorpgs) or last forever (in the case of mmorpgs.)  Essentially you’re making something that appeals TO the people who have more time than money… Much like payday loans your target audience is pretty dangerous in general, fiscally.  And as people progress to upper gamer class you have way more money than time.  And you wonder why people are willing to fork over $30 to get a bunch of WoW gold so they can kill something besides “soft fluffy bunnies, which pose no threat to humanity” without having to neglect their wife and kids for 400 straight hours.

Crucify the young for being young, crucify the old for being old.  It makes you wonder… in the minds eye of the video gaming industry… does their ideal customer base actually exist?  I bet it lies dangerously, and precariously, in the same kind of balance that the tobacco industry lies: hook em while they’re young and bleed em dry as they age.  Except the tobacco company does it better, they dont actively vilify their young “patrons” they simply wait.