I’ve been working on implementations of various Java programs and the like in my Data Structures and Algorithms class, and since I really like how this turned out, I figured it’d be worth sharing.

This is the first time I’ve used multiple buttons and the default UI layout – I think it’d be nice to share an implementation of it. This program also has a delightful side effect of showing how close heap sort works to O(n log(n)). Code follows:

main.class

// thanks to ping for heapsorter as well as rolf's colorbox
// Heapsort was originally published by J.W.J. Williams as
// Algorithm 232 in the journal Communications of the ACM [Wil 64].

package lab10;

public class Main{

    private static int[] a;
    private static int n;
    static int[] w1;

    static int w = 2000;
    static ColorBox b1 = new ColorBox(w);

    public static void main(String[] args) {
     //   int w = 600;
     //   ColorBox b1 = new ColorBox(w);
    }

    static int dh = 0;
    static int sc = 0;

    public static void sort(int[] a0)
    {
        a=a0;
        n=a.length;
        heapsort();
    }

    private static void heapsort()
    {
        buildheap();
        while (n>1)
        {
            n--;
            swap (0, n);
            downheap (0);
        }
    }

    private static void buildheap()
    {
        for (int v=n/2-1; v>=0; v--)
            downheap (v);
    }

    private static void downheap(int v)
    {
        int k =  2*v+1; //k = root's children
        while (k<n)
        {
            if (k+1<n)
                if (a[k+1]>a[k]) k++;
            if (a[v]>=a[k]){
                return;
            }
            swap(v, k);
            v=k;
            k=2*v+1;
            dh++;
        }
    }

    private static void swap(int i, int j)
    {
        int t=a[i];
        a[i]=a[j];
        a[j]=t;
        sc++;
    }
}

/*
 * Okay, these are the numbers I'm getting for the swaps needed for each job
 * When I say "should be" number x, I'm speaking of the following formula for
 * the O(n logn) amount of swaps required...
 *
 * x | x is a side of the array
 * swaps = x^2 * lb(x^2)
 * where lb is binary log.
 *
 * Since big O notation reperesents an >upper< bound after a certain nominal value
 * none of the trials should be larger than O(n log n).
 *
 *
 * size: 100
 * swaps: 123891
 * Should be ~130k
 * http://www.wolframalpha.com/input/?i=100%5E2+%2A+lb%28100%5E2%29
 *
 * size: 200
 * swaps: 575216
 * Should be about ~611k swaps
 * http://www.wolframalpha.com/input/?i=200%5E2+%2A+lb%28200%5E2%29
 *
 * size: 300
 * swaps: 1399854
 * Should be about ~ 1.4 million swaps
 * http://www.wolframalpha.com/input/?i=300%5E2+%2A+lb%28300%5E2%29
 *
 * size: 400
 * swaps: 2,619,194
 * Should be about 2.7 million swaps
 * http://www.wolframalpha.com/input/?i=400%5E2+%2A+lb%28400%5E2%29
 *
 * size: 500
 * swaps: 4,248,474
 * Should be ~ 4.4 million swaps
 * http://www.wolframalpha.com/input/?i=500%5E2+%2A+lb%28500%5E2%29
 *
 * size: 1000
 * swaps: 18,986,697
 * Should be 19 million swaps
 * http://www.wolframalpha.com/input/?i=1000%5E2+%2A+lb%281000%5E2%29
 *
 * size: 1500
 * swaps: 45,343,618
 * should be about ~ 47 million swaps
 * http://www.wolframalpha.com/input/?i=1500%5E2+%2A+lb%281500%5E2%29
 *
 * size: 2000
 * swaps: 83914749
 * should be about ~ 87 million
 * http://www.wolframalpha.com/input/?i=2000%5E2+%2A+lb%282000%5E2%29
 *
 */

next, we have the ColorBox method provided by the Professor and it includes my added buttons. This is the first time I’ve ever implemented multiple buttons on the UI. I know, minor victory. But that, to me, is really what programming is all about – little victories until one day, bam, you are half competent.

ColorBox.class

/*
 * ColorBox.java
 * Created on November 1, 2007, 11:08 AM
 * by Rolf Lakaemper
 * Zach added the buttons and actions
 */

package lab10;

import java.awt.*;
import java.awt.event.ActionEvent;
import java.awt.event.ActionListener;
import javax.swing.*;

public class ColorBox extends JPanel implements ActionListener{

    private static final int DEFAULTSIZE = 400;
    private int []colors;
    private int size;

    //zach's buttons
    public JButton j1;
    public JButton j2;
    public JButton j3;

    /* Creates a new instance of ColorBox */
    public ColorBox(int inSize) {

        size = inSize;
        colors = new int[size*size];
        initRandom();

        // Visuals
        setPreferredSize(new Dimension(size,size));
        JFrame frame = new JFrame("ColorBox");
        add(j1 = new JButton("sort"));
        add(j2 = new JButton("randomize"));
        add(j3 = new JButton("downheaps and swaps"));

        j1.addActionListener(new ActionListener() {
            public void actionPerformed(ActionEvent ae) {
                Main.w1 = Main.b1.getColorValues();
                Main.sort(Main.w1);
                Main.b1.setColorValues(Main.w1);}
        });

        j2.addActionListener(this);

        j3.addActionListener(new ActionListener() {
            public void actionPerformed(ActionEvent aw) {
                System.out.println("size: " + size);
                System.out.println("downheaps: " + Main.dh);
                System.out.println("swaps: " + Main.sc);}
        });

        frame.setDefaultCloseOperation(frame.EXIT_ON_CLOSE);
        frame.getContentPane().add(this);
        frame.pack();
        frame.setVisible(true);

    }

    /* Creates a new instance of ColorBox */
    public ColorBox() {
        this(DEFAULTSIZE);
    }

    public void initRandom(){
        for (int i=0;i<colors.length;i++){
            colors[i]=(int)(Math.floor(Math.random()*256));
        }

    }

    public int[] getColorValues(){
        // return a COPY of colors, keeps colors private
        int []c=new int[colors.length];
        System.arraycopy(colors,0,c,0,colors.length);
        return(c);
    }

    public void setColorValues(int []c){
        if (c.length!=colors.length){
            System.out.println("Error, arrays of different size.");
        }
        else{
            System.arraycopy(c,0,colors,0,colors.length);
        }
        repaint();
        return;
    }

    public void paintComponent(Graphics g){
        int count = 0;
        for (int row=0;row<size;row++){
            for (int r=row;r>=0;r--){
                g.setColor(new Color(colors[count],0,0));
                count++;
                g.drawLine(row-r,r,row-r,r);
            }
        }
        for (int col=1;col<size;col++){
            for (int c=col;c<size;c++){
                g.setColor(new Color(colors[count],0,0));
                count++;
                g.drawLine(c,size-c+col-1,c,size-c+col-1);
            }
        }
    }

    public static void main(String []args){
        ColorBox cb = new ColorBox(500);
        int []c=cb.getColorValues();
        java.util.Arrays.sort(c);
        cb.setColorValues(c);
    }

    public void actionPerformed(ActionEvent e) {
    // randomize button
       initRandom();
       repaint();

    // get ready for another run by setting counters to zero
    // downheap set back to zero
       Main.dh = 0;
    // swap count set back to zero
       Main.sc = 0;

    //  Junk drawer from other implementations/versions...

    //  long time1 = System.currentTimeMillis();
    //  Main.w1 = Main.b1.getColorValues();
    //  Main.sort(Main.w1);
    //  Main.b1.setColorValues(Main.w1);
    //  long time2 = System.currentTimeMillis();
    //  System.out.println(time2-time1);
    }
}

Which wasn’t on when polled.

This meant that every other time it was asked, it was waking up a processor to say, eh, I’ve been awake for 287/24000000000 of a second.

I re-ran the results connected to a Unix server and got the right results.

So it goes. Click on the image below and experience the magic for yourself.

int strcmp (char *s, char *t)
{
    for ( ; *s == *t; s++, t++)
        if (*s == '\0')
            return 0;
        return *s - *t;
}

So first we look at the input: we have two char pointers. Since we don’t have to change either s or t in order to compare them, pointers are the most efficient way to refer to them.

There is three possible input situations, first s could be greater than t, t could be larger, or they could be the same. Implicitly, we are comparing s to t. The for loop at the first line is made to go until s and t become not equal or to report they’re equal if the end of line \0 enters.

Otherwise, the difference between the two strings are reported as the result. This will be positive if s is longer than t, or negative if t us larger.

Therefore we have a discrete output for each of the three conditions. Note that since this comparison (in the for loop) is done until the end of similarity, the worst-case order of magnitude for this function is O(n). All other cases are O(1).

colortest.java

import javax.swing.*;
import java.awt.event.*;</code>

public class Colortest
extends JFrame
{
public static void main(String[] args)
{
new Colortest();	// create Window
}

private Colortest()
{
super ("Colortest");
addWindowListener(          // don't care about this
new WindowAdapter(){    // it only closes the application
public void windowClosing(WindowEvent event)    // when you close
{System.exit(0);}                           // the window
}
);
setContentPane(new ColorPanel()); // the actual drawing pane
pack();                             // organize, size etc.
setVisible(true);                   // show the window to the world !
}
}

colorpanel.java

import java.awt.*;
public class ColorPanel extends javax.swing.JPanel
{
// Fields ------------------------------------------------------------------
private final int numberOfRows    = 32;
private final int numberOfColumns = 32;
private final int rectSize = 16;
private Color[][] colorField; </code>

// Methods -----------------------------------------------------------------
public ColorPanel() {
initColorArray(numberOfRows, numberOfColumns);
setPreferredSize(new Dimension(rectSize*numberOfColumns, rectSize*numberOfRows));
}

protected void paintComponent(Graphics g)
{
// get array-size
int rows = colorField.length;			// number of rows
int columns = colorField[0].length;		// number of columns
for (int r=0;r &lt; rows; r++){
for (int c=0;c &lt; columns; c++){
g.setColor(colorField[r]);	// color for next rectangle
int positionX = c * rectSize;	// position
int positionY = r * rectSize;
g.fillRect(positionX, positionY, rectSize, rectSize);
}
}
}

//vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
/*
* initColorArray
* initialize 2 dimensional array 'colorField'
*/
private void initColorArray(int rows, int columns)
{
// A 2 dimensional array is an array of arrays
// hence it could be defined row-wise with different
// column - lengths ! However, the most usual case is a
// rectangular shape, i.e. every column has the same dimension.
// These arrays could  be declared in a less complicated way
// by new arrayname[sizeRow][sizeCol].
// But to make clear, that we really deal with nested arrays we define
// them beyond in a more universal way.

colorField = new Color[rows][];
for (int i = 0; i < rows; i++){
colorField[i] = new Color[columns];	// every column with same size here
	}

	// now define different colors
	for (int r=0; r < rows; r++){
            for (int c = 0; c < columns; c++){
		int red = (int)(255.0 * r / rows);      // compute portions of red
		int green = (int)(255.0 * c / columns);	// and green (max = 255)
		int blue = 0;				// no blue today
		colorField[r]1= new Color(red, green, blue);
}
}
}
//^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
}

I also owned a nice hardcopy copy of Gibbon’s decline and fall and had 2 vintage IBM computers, complete with printer.

I would play Warcraft II for the first time in a few months.

Will I ever be as cool again?

Anyhoos, here’s the pdf . Enjoy.

Sarah, our cities are teeming with millions of fatherless welfare thugs, welfare super-breeders, illegals and their anchor babies, all parasites on the taxpayers. Do you refuse to acknowledge that??

We used to trust politicians to figure this crap out for us, at least on a micro-level. Maybe that, not the media, not the current national situation, makes our time and political circlejerking so interesting and unique.

As I was born in the early 1980′s I wasn’t alive when Dallas Egbert went missing in August 1979. The story of his disappearance, and the wild theories that followed were based around his playing a LARP (Live Action Role-playing) version of D&D in the nine miles of unguarded and unmonitored tunnels that ran beneath the campus of MSU. Egbert was a genius, in the pure sense of the word, and was sixteen years old at the time of his disappearance.

“The Dungeon Master” is the 1984 book written by the flamboyant Texas-based private eye hired by Dallas’ parents. As this book was a single-serving social issue book that was subsequently never reprinted and ignored in lieu of the “cautionary tale” Mazes and Monsters, book, then movie with a damned-loose interpretation of the case, was released in 1982.

As who-done-it cases go, William Dear does a pretty thorough job of describing his process, albeit with some obligatory run-troughs of his super-cool gear. Many twist and turn lead investigators through many fruitless avenues and literal dead-end tunnels abound. The movie, and even the Newsweek review of the book printed on the back cover want to make so much hay of Dear trying a game of Dungeons and Dragons – either I must be totally desensitized to the idea of make-believe or again, the media looks for the best angle of attack. What happened to Dallas Egbert between his rescue and his tragic death is perhaps the least fleshed-out portion of the story and the most interesting.

By the end of this case, it becomes obvious that the boogey-man isn’t D&D, or role-playing; especially in the seen-through-the-minivan-windshield portrayal of it that William Dear gives us. Instead, it was the lack of support for a too-young person going to a large, impersonal university, drugs, and the hands-off, high-pressure treatment of gifted individuals. (Dear gets credit from me for posing this issue early and often)

Let us say for a moment that Egbert’s disappearance was due to his role-playing compulsion or addiction. Adult and child alike are known to descend into fantasy when reality becomes too much. The causes of these compulsions or fantasies aren’t to be blamed on the fantasy, but instead of a person unable to deal with the reality they’ve been placed in. I can’t blame the underclassmen that LARPed with Egbert for not seeing this as a mental illness, or at least a pressure valve stuck on “release” – the media is not the message. Novels didn’t destroy Victorian England, television didn’t destroy the children of the 1960s, and the Beatles didn’t cause the Manson murders. How much would of an enlightened mentor providing emotional support would have helped Egbert, an odd child with hardly any social skills? A lot more than D&D, and the drugs that Egbert used, thats for sure.

“They show you how things used to work back in the day. Like with the gays and the lesbians – look at what he did to Sodom and Gomorrah.”

The Mirodi Traveler’s Notebook is a device that has some traction in the neigh-unimaginable delve that is Stationary Blogging. For various reasons (the most compelling I’ve seen that its hard to buy in the US and anything Amazon doesn’t sell directly is for PROs, of course) this notebook has captured the imaginations of many weird people that are “travelers.” I do not need a teardown of components (like the kind routinely done to Apple products) to figure out I can find a square foot of leather and a decent stack of paper for less than $50.

Thankfully, if you have $50, and thinking “hey, I was going to make sweet consumer love over this product” – allow me to provide a useful alternative for the American traveler:

• An off-brand track suit. Thankfully, Uncle Sam (not that one) has us taken care of for just $26! (Wal*Mart)

• A Moleskine traveler’s notebook. (Moleskine Direct, $11 retail) No, it won’t work very well with your $300 fountain pen, but that Bic you swiped from Motel 8? Magic. Also, this guy has more than twice as many pages. Have more memories!

• A cheeseburger ($1-3, various) can be had, because you are a US American! $1 BK double pictured.

• Ten to Twelve Dollars.

I’ve been interested in the hacking movement to portabalize classic (and not-so-classic) games consoles that are reincarnated into huge handhelds. Since I am really interested in making an all-in-one system, and have been for a while, I’m going to recase this Game Boy Advance SP into something like a alarm clock case and also wire it for external control, internal amplified speakers, and fabricate a really neat controller to work the whole thing.

So, the plan goes something like this:

Step 1: Buy all the junk I need to make it. (sources and status noted)

1. Game Boy Advance SP (ebay, purchased)
2. Replacement LCD screen for GBASP (ebay, not purchased)
3. Junky Computer Speaker (thrift store, not purchased)
4. AC/DC power converter (thrift store, not purchased)
5. Wires (ide/ribbon/rainbow cabling, maybe a DB25 cable)
6. Interfaces – controller plug (thinking a DB25 connector), power, headphone (Radioshack,  not purchased)
7. Nylon wire braided sleeve (for controller) (Digikey, not yet)
8. Heat shrink tubing (always need more than you think, amirite) (purchased)
9. Tact switches (haven’t sourced them yet)
10. Poly enclosure for controller (Polycase, not purchased)
11. Solder (have)
12. Bondo, Paint, Sandpaper, cut glass for casing. (Autozone/Hardware stores, not purchased)

Step 2: Create a wiring schematic for the project

1. Need to keep power, video, audio and control wires in mind.
2. Should have a good idea of the power needs by that point.
3. Unit test wiring schematic against parts in hand (is every connection accounted for?)

Step 4: Wire the Game Boy and associated parts to work together

1. Power for all devices
2. Extend the ribbon cable for GBA screen to front of device (hardest part I can forsee)
3. control wires for each GBA button

Step 4: Fabricate the enclosures for the project

1. Make a list of ports
2. Source all wire grommets/interfaces needed
3. Roughs of the controller tested and working 100%
4. Roughs of the GBA/audio/video case working 100%
5. Bondo and sand and sand and sand and sand
6. Paint and sand and sand and sand

So, that should cover the whole thing. This plan will probably change over time as I ask various online communities for input as I go.

If this goes really well, the next step would be to create something I’ve been dreaming about for days now: a portrait Game Boy Advance.

Outta be fun! I’m shooting for an August 1st ship date!

Then, while jaywalking across Richmond Street, I came up with it: create a story, fleshing out the world of a central character (the aforementioned Sergei Descartes), and take that feeling and world and exploit it using the mediums

The thematic conflicts, as such, I hope will be:

The plastic feeling of fanboyism and fetishistic collection vs. Making

Responsiblity as identity vs. self-determined persona

Anyways, the short-short story I made the other day will be the springboard of the world I’m imaginating. Look forward to it.

As for inspirations for this world, I’ve been thinking about:

• The history of monopoly
• Heavy Rain (the PS3 game)
• “Mission Earth” by L.Ron Hubbard
• Check Cashing Concerns
• “Free to Choose” by Thomas Friedman
• Cory Doctorow’s “Makers” (my review here)
• Robert Harris’ “Fatherland”
• Quitting smoking

Anyways, I already forgot what it was about, but I did put the typewriter in it.

Wal*Mart Backpack [pdf]

Holla, then.

So, that leads me to two conclusions:

• I’m actually going to have to buy an LCD for this thing.
• Since it’s already broken, I can certainly break it more and nobody’d be the wiser.

Portablizing home consoles is so in fashion (ala anything Ben Heck) – why not unportablize a handheld instead?

So, I now have an epic one-page sketch of what technology it’ll require (the target budget soup-to-nuts is $50) and even thought out a never-before-seen regime of control scheme (a very physical-layer solution is promised here). A hint: it’ll involve no less than 10 normally-closed switches and a DB25 connector.

Maybe once I get my bona fides in the console hacker world, I can finally inspire someone to create my dream console: the PSP stuffed into a GBP shell:

Once I have a proper project plan, I will update the three of you.

Today, I learned about the super-cool-but-never-mass-produced-and-well-named B-58 Hustler, which was the first in the proud heritage of American jet-powered aircraft that could end the world or at least ruin your day. Early jet pilots were either insane or crazy, depending on your point of view. Chuck Horner and Tom Clancy’s oddly-awesome Every Man A Tiger, in which Horner relates his mucho appreciated military careet, including flying some aircraft that look definately non-gravity defying.

Then, I read about two years worth of Overcompansating (for perhaps the third time) – I’ll be god-damned if it doesn’t speak to this man’s soul.

My doubleTwist experiment is going badly, as the program doesn’t know how to kill with fire earlier copies of podcasts and nobody needs 9 episodes of the History of Rome in triplicate on their phone.  Especially since the Evo’s battery wouldn’t last through 3 of them in the subway system, RF emitting frantically for a reply, slowly microwaving my left leg.

Still looking for an awesome manual typewriter in some thrift store in this city so I can fall in love with it, ignore it for six months and then sell it for a tidy profit on ebay come Chrismastime.

Good day, in all.