Oh! Scope

Manos wants to get SPDY, and you can help

2011-06-27T00:00:00-04:00

Introduction

This summer, I've been fortunate enough to be accepted into the Google Summer of Code program to work on implementing Google's SPDY web protocol on top of the Manos Evented IO/HTTP Application platform. As of today, I'm at the point where I'd like to share what I've built so far and hopefully get enough testing and eyes on it to get pulled into master. After reading this post, you'll have no excuse not to contribute and test out this new functionality.

Requirements

Manos installation instructions and requirements can be found here. Additionally, you will need gnutls if you plan to try it over SSL (Linux only at the moment).

Building

In order to build with SPDY, you'll have to clone from my fork and checkout the 'spdy' branch:

git clone git://github.com/kersny/manos.git
cd manos
git checkout spdy

Then, you can simply build:

./autogen.sh (--with-gnutls if you want ssl)
make
(optional) sudo make install

If you don't want to clobber an already installed manos, you can skip the make install and invoke manostool.exe directly to try spdy.

Using it

To get your first manos app up and running check out the tutorial. Then to serve with spdy, simply set a port with -n:

manos --server -n 8081

Outputs:

Running (your manos app) on port 8080.
Running (your manos app) with SPDY on port 8081.

Now, you have to force Chrome to use SPDY because manos currently does not support protocol negotiation. To do so, invoke chrome from the command line with the switch '--use-spdy=no-ssl' This will make any non-spdy site fail, but allow us to test our own server. Then, simple head to localhost:8081 to navigate your Manos app.

Next Steps

My goal with this post is to generate discussion and intrest around manos's spdy support and more importantly get people critiquing and poking holes through my code. If you'd like to help, head to the discussion on the pull request and tell me how terrible my code is.

Even Easier S3 Jekyll deployment

2011-02-20T00:00:00-05:00

Simply deploying your Jekyll site to S3

Inspired by this Hacker news post, I decided to share my new blog deployment system based on Amazon's new static site support on S3.

Set up an S3 bucket for your site, as per this guide.
Install s3cmd, if you use homebrew on OS X, it's as easy as brew install s3cmd
Configure s3cmd: simply run s3cmd --configure and provide your AWS Access key and secret. You can also configure GPG at this time if so inclined.
Setup a push script: I have a bash script located in my jekyll root directory simply named "push" that goes as such:

jekyll

s3cmd sync --delete-removed _site/ s3://{bucketname}

Now you can simply access your site from the website endpoints listed on the developer site. Of course you don't get the benefits of git like Sören's method, but its dead simple to set up.

GPSd with Ruby

2011-01-20T00:00:00-05:00

GPSd

GPSd is a Linux daemon that can read and parse data from various GPS recievers, and it provides a standard interface for obtaining said data. GPSd provides JSON data over a socket connection in an easy to consume stream. In this example, I'm going to read the data with cool.io and parse it with Ruby bindings to yajl. Cool.io brings events like Node.js to Ruby with the simplicity of Sinatra. Yajl, on the other hand, is a streaming JSON parser that's fast and reliable.

First, install the required gems:

$ (sudo) gem install yajl-ruby
$ (sudo) gem install cool.io

Code:

require 'rubygems'
require 'cool.io'
require 'yajl'

ADDR = '127.0.0.1'
PORT = 2947

print = lambda { |data| puts data.inspect }
parser = Yajl::Parser.new
parser.on_parse_complete = print

cool.io.connect ADDR, PORT do
    on_connect do
        puts "Connected"
        write "?WATCH={\"enable\":true,\"json\":true}"
    end

    on_read do |data|
        parser << data
    end
end

puts "Running"
cool.io.run

PageRank for Ranking College Basketball

2010-12-13T00:00:00-05:00

Something I have been working on lately, after obtaining a large database of NCAA college basketball games, has been ranking teams algorithmically. One such ranking method that I recently implemented is PageRank, Google's method for ranking webpages.

PageRank

The PageRank algorithm itself is covered in Google's White Paper on the subject. I, however, learned much more from this article by Michael Nielsen. It covers the math in a way that even I can understand, and provides a concise, ~6 line example in python using matrices. This will serve as the basis for my Ruby implementation of ranking College Basketball teams.

The Data

For this application, I am using 141,523 College Basketball Games from 1981 forward, stored in a MySQL database locally. These games have the following data:

| id                | int(11)      |

| date              | datetime     |

| team1             | int(11)      |

| team1score        | int(11)      |

| team2             | int(11)      |

| team2score        | int(11)      |

| team1home         | int(1)       |

| team1inconference | int(1)       |

| year              | varchar(255) |

Note:

team1 and team2 are Foreign Keyed to another table of teams
I have a separate 'year' column in order to make year-by-year querying easier
I realize team1 is unnecessary in front of 'in conference', but I don't really feel like changing it

Gems

Since I'm doing this in Ruby, there are some (2) gems necessary to get up and running. The first is the 'mysql gem', a simple way to get my data out of the database and into code. This can be installed simply with:

gem install mysql

Next is the wonderful Narray Library, which handles arrays, matrices, and vectors with a C backend, allowing for not native Ruby speeds (read: fast). For this to work on my MacBook, I had to install version 0.5.9p7:

wget http://rubyforge.org/frs/download.php/61995/narray-0.5.9p7.tar.gz

tar -xvzf narray-0.5.9p7.tar.gz

cd narray-0.5.9p7

ruby extconf.rb

make

sudo make install

Now let us write some code...

Implementation

Require

First, we simply pull in our libraries so that we can actually do stuff

require 'rubygems'

require 'mysql'

require 'narray'

Initialization

Next up, we initialize variables needed for the script. First up in this process is the database class, used to query MySQL. If you want to actually use this, you will have to put in your own host, username, password, and database. Alternatively, you can just build up an array of hashes and calculate PageRank on this. Next, s & t are used to decide whether to follow a link or to pick a random page. This is described in more detail in Michael's post. After this, we use MySQL to get the number of teams in order to create our matrices. Using this value, we create a (square) identity matrix the with a dimension of the number of teams. Finally, we create the probability distribution for the random jump, setting each to an equal value, 1/(the number of teams).

dbh = Mysql.real_connect("host", "username", "password", "database_name")

s = 0.85
t = 1.0-s

query = "SELECT COUNT(*) as count FROM Teams"
res = dbh.query(query)
numteams = res.fetch_hash["count"].to_i

ident = NMatrix.float(numteams, numteams)
ident.diagonal!(1.0)

p = NMatrix.float(1, numteams).fill!(1.0)
p = (1.0/numteams.to_f) * p

Getting Games into a Matrix

The next step is to pull the game data out of a database and into a matrix. For this example, I am only doing games in the 2009-2010 year. I just loop through each of these games and increment the count in a matrix depending on which team won.

rawgames = NMatrix.int(numteams, numteams)

query = "SELECT * FROM Games WHERE year='2009-2010' ORDER BY date"
res = dbh.query(query)
res.each_hash do |row|
  team1 = row["team1"].to_i
  team2 = row["team2"].to_i
  if (team2 != 0 && team1 != 0)
    team1 -= 1
    team2 -= 1
    if (row["team1score"].to_i > row["team2score"].to_i)
      rawgames[team1, team2] += 1
    else
      rawgames[team2, team1] += 1
    end
  end
end

Making a Probability Matrix

Now that I know of all of the wins/losses for a team, I can create the probability matrix for the chance of one team "linking" to another. In this case, a loss counts as a link, so the highest ranked teams are those with the most links from other high ranked teams. To do this, I just have to get the total of all of the losses of a team, and then divide the losses to each other team by this number.

prob = NMatrix.float(numteams, numteams)

0.upto(numteams - 1) do |i|
  sum = 0.0
  0.upto(numteams - 1) do |j|
    sum += rawgames[j, i].to_f
  end

  0.upto(numteams - 1) do |j|
    val = rawgames[j,i].to_f / sum
    prob[i, j] = val.nan? ? 0 : val
  end
end

Calculating

With all of the information collected so far, we can calculate the PageRank for each team, allowing us to rank each team. You can see this calculation below, in the first line. The code that follows simply inserts the rank data into the database, allowing me to access and query it at my will.

fin = (t*((ident - (s*prob)).inverse)*p)

0.upto(numteams - 1) do |i|
  val = i + 1
  year = '2009-2010'
  rank = fin[0,i]
  dbh.query("INSERT INTO page (teamid, year, rank) VALUES (#{val},'#{year}', #{rank})")
end

All Together Now

Here's the final code, all put together and available for your usage. (MIT License)

Problems

I have 341 teams... with too many, memory could become an issue. For this case, I would look into a MapReduce method
If you're having trouble understanding this code, once again I highly reccommend this blog post
I'm not going to give you my game data, that you'll have to find for yourself

Results

Top 25 according to ESPN:

Kansas
Kentucky
Duke
Syracuse
Ohio State
West Virginia
Kansas State
New Mexico
Villanova
Purdue
Butler
Temple
Michigan State
Georgetown
Tennessee
Wisconsin
Brigham Young
Pittsburgh
Baylor
Maryland
Vanderbilt
Gonzaga
Texas A&M
Richmond
Xavier

My Top 25:

Duke
West Virginia
Georgetown
Tennessee
Syracuse
Kentucky
Kansas
Michigan State
Purdue
Butler
Pittsburgh
Wisconsin
Georgia Tech
Villanova
Kansas State
Notre Dame
Northern Iowa
Maryland
Louisville
Ohio State
Oklahoma State
Minnesota
Baylor
Vanderbilt
NC State

Note: My data includes March Madness games, where ESPN rankings are before the tournament.

iPad for School?

2010-05-09T00:00:00-04:00

What I need

As you may or may not know of me, I am finishing up my junior year in high school and will soon be moving into my senior year and then college. For my career path, which will be Electronics Engineering or Computer Science, computing will be essential. I will have to make a decision in the very near future regarding purchasing a computer that will likely have to last throughout my high school and college career (5 years) and allow me to take notes, write papers, and give presentations. I'm posting some of my ideas in order to hopefully illicit discussion regarding my choices, and maybe gain some further ideas into my situation. If you have any ideas for me or anything that I missed, please comment or tweet me @kersny.

What I've got

Right now, I am running on a Dell Inspiron 600m laptop with a broken and removed screen on either a 17" or 23" monitor. This has a 1.6Ghz processor, 512mb of RAM, and a 60gb hard drive... so yeah, its pretty weak. Seriously though, I dual boot XP and Ubuntu 10.04 and it manages to handle many tasks well, such as Visual Studio 2010 Express, but it has its problems and slows considerably for other common tasks or any sort of multitasking.

Options

1. MacBook

Purchasing a MacBook was my first thought when I considered buying a computer. I have always wanted one, and they offer many advantages. I'm sure that a MacBook would be plenty durable to last 5+ years, and I know that I could run OSX, Windows 7, and Linux to handle anything required of me in college The 8-10 hour battery life would also be great for note-taking and college classes. The only problems I see with this options are price and power. Obviously, a Macbook is not cheap and would require a considerable chunk of change. The other downside, power, is not so concrete. I was quite disappointed in the latest refresh of the MacBook line by Apple, and although I have read many of the reasons and justifications for doing so (one such), I don't really want to buy a $1200 laptop with a Core 2 Duo chip to last my entire college career. I just think that this would be too underpowered for the things that I could be doing throughout college. The next step up, a 15", is too expensive now but would have the lowest specs that I would probably consider purchasing.

tl;dr: I'd love to get a Macbook, but the power of the 13" model and the price of the 15" model are whats holding me back.

2. Regular Laptop

It's hard for me to look at the MacBook and not see the huge "Mac Tax" imposed on such an item. I could buy an i7 laptop with 4 Gigs of RAM from any number of PC companies for much less than even the low end MacBook. I know that this should be plenty powerful to handle most anything I would throw at it throughout my college years, but it has its own set of cons. The first of such would be durability, which is important for something that has to last 5 years. I've seen many a laptop fail hardware or case wise, and that could be a huge problem for a college kid. Another smaller concern is the fact that it is not a Mac, meaning that I could not develop apps for the iPhone or iPad.

tl;dr: A Windows Laptop would be plenty powerful but I'm scared about durability and I couldn't do iPhone development.

3. The iPad/Mac Mini power combo

This is an interesting option that I have been looking into over the last couple of weeks. For much less than the price that I would have to spend on a sufficiently powerful MacBook, I could get both an iPad and a Mac Mini. I would buy an iPad now, and use it through my senior year and onto college. I could write papers (with an external keyboard of course), give presentations, and do all the research required of me in my senior year. Add to this the ease of note taking and portability in college, and it could make a great platform. I would, however, have to buy a complementary Mac Mini in order to support coding and other programs that won't run on an iPad. This could be pushed into college, after a refresh or two, and give me a powerful box to anything not supported on an iPad. The only problems in this situation would be the known issues with content creation on the iPad and any labs or such in college that would require real programs to be run in the classroom, such as maybe an IDE or Compiler. I do believe, however, that a college cannot force you to have a computer if that don't provide you with one, and will likely have them available in any such lab situations.

tl;dr: An iPad + Mac Mini combo should let me do everything I need throughout high school and college, unless I am required to run special programs during class.

So that is my situation and those are the options that I am currently considering. If you have any experience, advice, opinion, or idea, please let me know.

iTunes & Music code Redux

2010-01-01T00:00:00-05:00

It has been brought to my attention on my first iTunes COM API post that some of my information in that post is currently invalid:

Lyrics: LyricWiki decided to cut off their API because of copywright concerns, so that function will no longer work. Link
Album Art: Amazon has implemented authentication in their album art API, if you would still like to use this service, I recommend this post

My Segway Works!

2009-08-26T00:00:00-04:00

Its not 100% yet, but I got my Segway ride-able. Here are some videos:

Riding my Segway! from Kerry Snyder on Vimeo.

Riding My Segway - With Steering from Kerry Snyder on Vimeo.

I should be able to post more details and code soon.

Segway Video and TI Pong

2009-08-08T00:00:00-04:00

Yes, its been over a month, but I have made quite a it of progress on my Segway, and, without further adieu:

Item #1:

Initial Segway Test from Kerry Snyder on Vimeo.

It balances itself! Now all I need to do is adjust the PID constants. And put in a big red E-Stop switch.

Item #2: TI calculator Pong

I made this while I was bored in math class last year, all by myself, programmed in a BASIC-like language on our school's graphing calculators, which I believe are TI-83+'s.

Here is the calculator manual, with programming information: Chapter 16

And here is the code:

Lbl F
ClrHome
Menu("PLAY PONG","PLAY",Z,"HIGH SCORE",K,"QUIT",S)
Pause
Lbl Z
1→B
0→G
1→O
1→P
0→J
Lbl U
getKey→A
If A=34
Then
B+1→B
Then
8→B
End
ClrHome
Output(B,1,"[")
End
If A=25
Then
B-1→B
If B<1
Then
1→B
End ClrHome
Output(B,1,"[")
End
If G=0
Then
randInt(3,6)→X
randInt(6,10)→Y
Output(X,Y,"+")
G+1→G
Else
Output(X,Y,"+")
If P=1
Then
X+1→X
Else
X-1→X
End
If O=1
Then
Y+1→Y
Else
Y-1→Y
End
If X>8
Then
8→X
0→P
End
If X<1
Then
1→X
1→P
End
If Y>16
Then
16→Y
0→O
End
If Y<2
Then
If X=B or B-1=X or B+1=X
Then
2→Y
1→O
J+1=J
Else
ClrHome
Disp "GAME OVER"
Pause
Output(5,4,"SCORE:")
Output(5,11,J)
Pause
If J>E
Then
J→E
ClrHome
Output(5,2,"NEW HIGH SCORE")
Pause
End
Pause
Menu("GAME OVER","PLAY AGAIN",H,"QUIT",S)
End
End
ClrHome
Output(B,1,"[")
Output(X,Y,"+")
End
Goto U
Lbl H
1→B
0→G
1→O
1→P
0→J
Goto F
Lbl S
0→J
ClrHome
Stop
Lbl K
Output(5,2,"HIGH SCORE:")
Output(5,14,E)
Pause
Goto F

iTunes COM API C#: Playlists

2009-06-27T00:00:00-04:00

After a comment on my original iTunes COM API post inquiring about Playlists, I decided to revisit that part of the project. I had originally intended to include this in the first post, but it was more difficult than I had first though. Here are some code snippets to get you started:

Create a ComboBox listing the names of all the playlists:
```
foreach (IITPlaylist pl in app.LibrarySource.Playlists)
{
    comboBox1.Items.Add(pl.Name);
}
```
Notes:
- Retrieves all playlists in iTunes

Playing One:

private void button1_Click(object sender, EventArgs e)
{
    IITPlaylist pl = app.LibrarySource.Playlists[comboBox1.SelectedIndex + 1];
    pl.PlayFirstTrack();
}

Notes:

Could also be done on ComboBox1.SelectedIndexChanged()
+1 Accounts for differences in indexes

This is just some basic examples, for tons more information, check out this link. You have to login in with your iTunes account, but look for "iTunes COM for Windows SDK (ZIP)" If you have trouble locating this file, email me and I can get it to you. It includes iTunesCOM.chm, which is the de-facto resource for all of this information. Between my example code (Here and here), you should be able to preform many of the iTunes functions. As always, comment if you have any troubles.

Getting an Angle

2009-06-16T00:00:00-04:00

So probably the biggest part of constructing a balancing robot, or scooter, is knowing what angle it is in relation to the ground. From this, you can determine how much you need to compensate. The balancing system in my Segway consists of two sensors:

ADXRS614 Gyro 50 °/s

ADXL203CE Accelerometer +/- 1.5g

*Above pictures from Sparkfun Electronics *

As I have explained in a previous post, both of these sensors are required to determine an accurate angle measurement. In my code so far, these are combined using a Complementary Filter. A great resource for coding a complementary filter on the Arduino can be found here. The hardest part that I found using this code is the Coefficients, which are not explained. I have determined, however, how to calculate these and get a working angle output. The 2 Coefficients in the program are 28.783 for the Gyro and 10.78 for the Accelerometer.

To calculate this for the Gyro:

Get the mv/°/s value for your Gyro from the Datasheet

-----25

Divide by 1000 (Milli volts in a volt)

-----0.025

Divide by 5 (Voltage of the Analog Reference)

-----.005

Divide by .0174532925 (Radians in a Degree)

-----0.2864789

Take the Inverse of that (1/x)

-----3.49

And there is your answer, 3.49. You would replace the 28.783 in the formula for this.

To calculate this for the Accelerometer:

Get the mv/g value for your Accelerometer from the Datasheet

-----1000

Divide by 1000 (Milli volts in a volt)

-----1

Divide by 5 (Voltage of the Analog Reference)

-----0.2

Take the Inverse

-----5

I'm not so sure on this one, but it seems to work. You would replace the 10.78 with this.

You can also replace the Voltage, in this case 5v, or the 1024 in the Formula, depending on your supply voltage or Analog Port Resolution.

When I put this all together in my test code, I get a solid, seemingly accurate angle measurement.

Next up - PID!

iTunes COM API in C#

2009-04-02T00:00:00-04:00

I decided that I would post a little guide to some of the things that I have learned in using the iTunes COM API in C#. It's not very well documented, and I though I could help.

Initialization:

First, you have to add a Reference and Using

Right click on References, and click "Add Refrence"

Switch to the COM tab and find "iTunes *.** Type Library, and click add.

At the top of your code, add "Using iTunesLib"

Finally, add these two lines of code under the class of the form, outside of any function:

delegate void Router(object arg);
iTunesApp app = new iTunesAppClass();

The Router helps in handling the iTunes Events, and "app" is what you will use to do anything with iTunes.

Events:

You'll need to attach at least one event for a program that controls iTunes, to see when something changes in iTunes, and here is some code for that:

app.OnPlayerPlayEvent += new _IiTunesEvents_OnPlayerPlayEventEventHandler(delegate(object o)
  {
      this.Invoke(new Router(app_OnPlayerPlayEvent), o);
  });

This should be put under InitializeComponent(), and will call app_OnPlayerPlayEvent when it is activated.

**Controls:

**These are pretty simple, just call them to control iTunes:

app.Play();

app.Pause();

app.NextTrack();

app.PreviousTrack();

app.Stop():

There are a few others, but these are the big ones.

**Track Information:

**iTunes provides pretty much all of the information about a playing track, there is a lot more than I mention here, I reccomend experimentation.

IITTrack track = app.CurrentTrack;

Albumlbl.Text = track.Album;

Artistlbl.Text = track.Artist;

Tracklbl.Text = track.Name;

These will populate the labels with the track (at the time) that is playing or paused. It must be refreshed to update the information.

IITArtworkCollection Art1 = track.Artwork;

IITArtwork Art2 = Art1[1];

Art2.SaveArtworkToFile("c:\\temp\\Album.jpg");

Stream r = File.Open("c:\\temp\\Album.jpg", FileMode.Open);

Image temp = Image.FromStream(r);

r.Close();

pictureBox1.Image = temp;

SetImage(pictureBox1);

This is a big one, it will allow you to put the Album Art of the currently playing track in a Picture Box. I recommend checking the Art1[] or doing this in a Try...Catch to avoid errors when a song does not have art. You will also need these, also, for resizing:

Original Link

public Size GenerateImageDimensions(int currW, int currH, int destW, int destH)
{
    //double to hold the final multiplier to use when scaling the image
    double multiplier = 0;
    //string for holding layout
    string layout;
    //determine if it's Portrait or Landscape
    
    if (currH > currW) layout = "portrait";
    else layout = "landscape";
    
    switch (layout.ToLower())
    {
        case "portrait":
            //calculate multiplier on heights
            if (destH > destW)
            {
                multiplier = (double)destW / (double)currW;
            }
            else
            {
                multiplier = (double)destH / (double)currH;
            }
            break;
        case "landscape":
            //calculate multiplier on widths
            if (destH > destW)
            {
                multiplier = (double)destW / (double)currW;
            }
            else
            {
                multiplier = (double)destH / (double)currH;
            }
            break;
    }
    //return the new image dimensions
    return new Size((int)(currW * multiplier), (int)(currH * multiplier));
}
//Resize the image
private void SetImage(PictureBox pb)
{
    try
    {
        //create a temp image
        Image img = pb.Image;

        //calculate the size of the image
        Size imgSize = GenerateImageDimensions(img.Width, img.Height, this.pictureBox1.Width, this.pictureBox1.Height);

        //create a new Bitmap with the proper dimensions
        Bitmap finalImg = new Bitmap(img, imgSize.Width, imgSize.Height);

        //create a new Graphics object from the image
        Graphics gfx = Graphics.FromImage(img);

        //clean up the image (take care of any image loss from resizing)
        gfx.InterpolationMode = InterpolationMode.HighQualityBicubic;

        //empty the PictureBox
        pb.Image = null;

        //center the new image
        pb.SizeMode = PictureBoxSizeMode.CenterImage;

        //set the new image
        pb.Image = finalImg;
    }
    catch (System.Exception e)
    {
        //MessageBox.Show(e.Message);
    }
}

Another thing that you can do is a progress bar for Track Time. I have Timer1 set for 1 second intervals. This is not totally tested, your mileage may vary:

void Timer1Tick(object sender, EventArgs e)
{
    IITTrack track = app.CurrentTrack;
    progress.Maximum = track.Duration;
    progress.Value = app.PlayerPosition;
    if (track.Duration != oldtrack)
    {
        if (track.Name != oldname)
        {
            NewSong();
        }
        else
        {
            oldname = track.Name;
        }
    }
    else
    {
        oldtrack = track.Duration;
    }
}

REST Album Art and Lyrics

Finally, I have these few functions that will use Amazon and LyricWiki to find Album Art and Lyrics of a playing song. I would recommend running them in a BackgroundWorker, to prevent a program from locking up.

Lyrics:

private static string GetLyrics(string Artist, string Title)
{
    StringBuilder lyric = new StringBuilder();  //Build a new string
    lyric.Append("http://lyricwiki.org/api.php?func=getSong&artist="); //Add base URL
    lyric.Append(Artist);
    lyric.Append("&song=");
    lyric.Append(Title);
    lyric.Append("&fmt=xml");
    WebRequest requestlyr = WebRequest.Create(lyric.ToString()); //prepare web request
    StreamReader responseStreamlyr = new StreamReader(requestlyr.GetResponse().GetResponseStream()); //prepare responese holder
    string responselyr = responseStreamlyr.ReadToEnd(); //fill up response
    responseStreamlyr.Close(); //Close stream
    string lyrics = XmlParse_general(responselyr.ToString(), "lyrics"); //parse the XML
    return lyrics;
}

private static string XmlParse_general(string Url, string type)    //XML parse Function
{
    XmlTextReader xmlrt1 = new XmlTextReader(new StringReader(Url));
    while (xmlrt1.Read())
    {
        //Trace.Write("Node Type", xmlrt1.NodeType.ToString());
        string strNodeType = xmlrt1.NodeType.ToString();
        string strName = xmlrt1.Name;
        if (strNodeType == "Element" && strName == type)
        {
            xmlrt1.Read();
            return xmlrt1.Value; //Return output
        }
    } 
    return "";
}

Album Art:

private string AlbumURL(string artist, string album)

{
    StringBuilder art = new StringBuilder();  //Build a new string
    art.Append("http://ecs.amazonaws.com/onca/xml?Service=AWSECommerceService&Version=2005-03-23&Operation=ItemSearch&ContentType=text%2Fxml&SubscriptionId=09XAA8GD0PQVCC8KR5G2&SearchIndex=Music&Artist=");   //Add base URL

    art.Append(artist);
    art.Append("&Keywords=");
    art.Append(album);
    art.Append("&ResponseGroup=Images");
    WebRequest requestart = WebRequest.Create(art.ToString()); //prepare web request

    StreamReader responseStreamart = new StreamReader(requestart.GetResponse().GetResponseStream()); //prepare responese holder
    string responseart = responseStreamart.ReadToEnd(); //fill up response
    responseStreamart.Close(); //Close stream
    string url = XML_image(responseart.ToString()); //parse the XML
    return url;
}

private static string XML_image(string xml)
{
    XmlTextReader reader = new XmlTextReader(new StringReader(xml));
    reader.ReadToFollowing("LargeImage");
    reader.ReadToFollowing("URL");
    reader.Read();
    return reader.Value;
}

If anyone has any questions or needs help getting these going, especially the REST functions, or has any suggested improvements, contact me, I'd be glad to help.

Goodbye, and Good Luck!

Bit.ly C# API Update

2009-03-18T00:00:00-04:00

I decided to update my Bit.ly API DLL in C#, thanks to @codemypantsoff. The only new function that I added was the ability to get the User who shortened a URL from the Short URL or from the Hash. I figured it was also a good time to move all my stuff from my Google Sites page onto a blog post, so here it goes.

Calls:

MetaH - Retrieves the metadata from a hash

string a = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07", "epicdiggnation", "MetaH");
Console.WriteLine(a);

MetaU - Retrieves the metadata from a Short Url

string b = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","http://bit.ly/1RmnUT", "MetaU");
Console.WriteLine(b);

ExpandH - Returns the Expanded url from a hash

string c = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","epicdiggnation", "ExpandH");
Console.WriteLine(c);

ExpandU - Returns the Expanded url from a Short Url

string d = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","http://bit.ly/1RmnUT", "ExpandU");
Console.WriteLine(d);

Shorten - Returns a Shortened URL from a long URL

string e = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","http://cnn.com", "Shorten");
Console.WriteLine(e);

ClicksH - Returns number of clicks from a hash

string f = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","epicdiggnation", "ClicksH");
Console.WriteLine(f);

ClicksU - Returns number of clicks from a Short Url

string g = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","http://bit.ly/1RmnUT", "ClicksU");
Console.WriteLine(g);

UserU - Returns the User who shortened the URL from a Short Url

string h = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","http://bit.ly/1RmnUT", "UserU");
Console.WriteLine(h);

UserH - Returns the User who shortened the URL from a Hash

string i = API.Bit("bitlyapidemo", "R_0da49e0a9118ff35f52f629d2d71bf07","epicdiggnation", "UserH");
Console.WriteLine(i);

Console Application sample code:

using System;

using System.Collections.Generic;

using System.Linq;

using System.Text;

using Bitly;

namespace ConsoleApplication1

{

    class Program

    {

        static void Main(string[] args)

        {

            string a = API.Bit("bitlyapidemo","R_0da49e0a9118ff35f52f629d2d71bf07", "http://cnn.com", "Shorten");

            Console.WriteLine(a);

            Console.Read();

        }

    }

}

DLL Link

Source Link

If anyone would like me to update my WPF sample app or has a feature request, go ahead and contact me.

Quick Update

2009-03-08T00:00:00-05:00

Well, I haven't updated recently, mostly because I'm kind of frozen in the process. But some good things happened and I have quite a bit of money now. Expect updates soon where I can actually get to building and debugging this thing. I'm getting ready to order the motor controller, batteries, and a Gyro/Accel, so I'll be able to actually start.

Serial Port on the iPhone with Minicom

2009-02-24T00:00:00-05:00

This will be a little tutorial on how to set up Minicom on your iPhone or iPod Touch to view incoming serial data. I hope to continue this with Python or maybe even Obj-C. Click on (some) pictures to enlarge.

Materials:

Jailbroken iPhone or iPod Touch

Charge Connector (MJS or Sparkfun)

Serial Data Source (Arduino in my case)

Computer/Wifi network

Putty

Install Minicom, Terminal OpenSSH, and Toggle SSH from Cydia

This step should be pretty self-explanitory, If you can't do this and can't google, then you probably shouldn't continue on.
SSH into your iPhone

Obtain your IP Address from Settings or the SBSettings App.

Start up the SSH Server

Go into Putty, enter the IP, and create a new SSH Session

On the "Potential Security Breach," click Yes

And Login as Root with Password "Alpine"

And Your In!
Make a Directory

Do a "CD .." twice, which whould put you in the root of the filesystem at /

Now do "CD usr"

And "MkDir etc"
Start up Minicom in configuration mode

Simple: "Minicom -s"
Setup iPhone Port

Arrow down 2, to Serial port setup and press enter

First, press A and change "/dev/ttyS1" to "/dev/tty.iap" and press Enter

Then press E and adjust to whatever serial source you are using. I used 19200 8N1 for my Arduino.
Save and Exit

Press Enter until you return to the main screen, arrow down to "Save setup as dfl," and press enter.

And press "Exit from Minicom"

You can now close the Putty Session.
Hardware Setup

This will vary depending on your Serial input and iPod connector, but here's mine:

In this case, I have the Serial TX from the Arduino connected to the RX of the iPod through a 1K resistor, for 5v to 3.3v. I also have the common ground connected. The Arduino is reading data from a Wii nunchuck through a Wiichuck Board and sending it out at 19200 Bps. Remember level changing for 5v and 12v sources!!
Read the Data

Power on your serial source.

Open up the Terminal on the iPhone, type in "Minicom," and press enter.

You should see serial data streaming in.

And your done!

PS. If anyone has experience with graphical Python on the iPhone, I would love to collaborate on an app. I have tested and the Python serial library will run on the iPhone. Contact Me.

Research Paper

2009-02-21T00:00:00-05:00

Well, here is a research paper I wrote for English on Open Source Hardware

Paper

Also- Google Docs sucks for this type of application.

Great Motor Controller Link

2009-02-19T00:00:00-05:00

Here is a great link summarizing ESC's for Battlebots. The only downside is that they all take RC Servo signals instead of PWM, which I believe will be more precise in my application. Right now I am mostly considering the RS80D or the IBC.

Battery Technology

2009-02-19T00:00:00-05:00

So now that I have motors, I will need to purchase batteries to power this thing. I am considering three different battery technologies for this system.

A123 Lithium

These are the batteries used by Daniel Fukuba in his scooter based on Trevor Blackwell's design. Dewalt now sells 36v batteries using this technology, and a matching charger. If I can use these batteries and charger while bypassing the internal current limiting, I should have a fairly simple battery setup with ample current capacity. For the final project I would plan on 3 of these packs in parallel.

Dissecting 36v Dewalt pack

Short video of Daniel Fukuba's

Sealed Lead Acid (SLA)

Although these batteries are large and heavy, they allow for large amounts of current draw and are the least expensive. They are also the cheapest to charge and maintain.

Example Battery

NiMH packs

This system for batteries would be easier to charge than A123 cells and provide ample current. This is also the type of batteries that Trevor Blackwell recommends.

Trevor Blackwell's page

These are the basic options that I am considering, I would love to hear any opinions on this!

My Motors Came!!

2009-02-18T00:00:00-05:00

Received the motors that I purchased off of Ebay Today

Pics:

The Box

The Whole System

Female Motor Connector

The Controller

Now onto the problem. The motors are great, and definitely worth the price. It also came with a motor controller though. This controller has 8 55V 110A [IRFP064N][9] HEXFET's. The only thing is that I have no idea how I would go about using this controller. Here is a Rundown:

[

H8/532 Main IC][10]

4 [IR2110S IC's][11]

2 GPS 35DAA010AKG ?

Any help would be appreciated in this matter. Not to mention that it would shave ~$300 off of the cost of this project. Thank You!

B-B-B-Balancing

2009-02-17T00:00:00-05:00

So first off, a Segway needs to know at what angle it is to the ground. This can be accomplished surprisingly easily.

Their are basically two types of sensor used for sensing tilt in a robotics application, the Gyro and the Accelerometer. Both have their advantages and disadvantages, and both must be used for reliable tilt data.

Accelerometer - Used in the Nintendo Wiimote and Nunchuck

Pros