Install VarisCite PoC. Now go to an article on IGN to test it.

If you want the source, right-click save as. Crx files are zip files with the extension change from zip to crx.

Beware of all content scripts. Even ones you wouldn’t expect to be modifying your DOM may be doing it anyways for the hell of it. When an extension asks if you want to allow it access to “Your data on all websites”, ask yourself if this is really necessary, and do you trust the extension’s author?

To be more technical, the extension works by injecting JavaScript into the actual website. The JavaScript then stores its results into a meta variable, “chromesniffer_meta”, inside the web page’s DOM. The exploit works by modifying this element before the extension has time to read it. This coupled with little to no data cleansing means we can inject both HTML and JavaScript into the extension.

I’ve created a simple DoS exploit located at chromesnifferexploit.html. The page has a warning on it on what to do engage the exploit. Simply clicking the link will not engage the exploit, blow up your computer, or do anything.

This is a Google Chrome Extension for the English Wikipedia. Wikipedia Watchlist checks your watchlist and updates the badge on the browser action icon to reflect new changes. Clicking on the badge opens up a popup where you can highlight different changes to view both the edit summary and the time the change was made. Clicking the little ‘d’ will take you to the last difference for the page and clicking the page name will take you to the page itself.

Click Here. Wikipedia Watchlist at the Google chrome extensions site.

/* Trapeze RSS parser in 7 lines */
function parseRSS(d) {
    return $(d).find('item').map(function() {
        var i = new Object();
        $(this).children().each(function(){i[this.tagName]= this.textContent;});
        return i;
    });
}

Called with a return document from an ajax request, this function returns an array of items which are associative arrays from the feed’s items.

// Here's an example
$.get(your-rss-feed, function(d){
    parseRSS(d).each(function() {
      $('<a></a>').text(this.title).attr('href', this.link).appendTo('body');
      $('<br/>').appendTo('body');
    });
});

The biggest problem with the mouse is that the scroll wheel will only work for about six months. As I said in the beginning I had to RMA my first mouse because the scroll wheel died after six months. Six months more, and the casualty count increases. It just doesn’t scroll at all now. I only used it for scrolling in the browser and I’d imagine if I liked using it for video games it would have gone even faster. So now I am left with a useless scroll wheel, having to use the pinkie button to go forward, and no ability to remap buttons. Currently on the lookout for a new mouse.

The library includes support for RTSP, RDT, SDPP, ASMRP, and RMFF, and it is all in pure Python and Pythonically coded where possible (see sdpp.py). It requires Python 2.6 and the Twisted Python library. The library has only one known bug, and I can’t for the life of me find it. Perhaps another set of eyes will effect a solution.

To use the library to simply download an RTSP stream: python rdt.py -u <URL> -f <FILENAME>

If you like this post, please comment. Commenting is fast, anonymous, and greatly appreciated. Download the Twisted Python RDT-RTSP Library here

Update After doing some searching, I’ve found a FireFox plugin caled Video DownloadHelper which performs much the same function as the program I’ve written, but is more stable and integrated directly into FireFox. Its only downside is that it doesn’t automatically download the songs as you listen, you need to start downloading the ones you want. But it is more stable than my program and much more user friendly. Be sure to use Mp3tag (listed below) to quickly rename all the files you’ve downloaded.

Download and install the following pre-requisites:

1. Python 2.6.5 – http://www.python.org/ftp/python/2.6.5/python-2.6.5.msi
2. dpkt 1.7 – http://dpkt.googlecode.com/files/dpkt-1.7.win32.exe
3. pypcap 1.1 – pcap-1.1.win32-py2.6.exe

Save this Python program into a folder of its own: groovegrab.py

Run the program using python groovegrab.py

As you listen to songs on Grooveshark, the folder will populate with mp3 files. Each are named Track_X.mp3. But there’s a really neat, easy way to rename them. Download Mp3tag and in the program go to File > Change Directory and choose the directory where the mp3 files are. Select all the mp3 files in the main pane and go to Convert > Tag to Filename.

Enjoy your perfectly named mp3 files from a Grooveshark listening session.

Thanks to dugsong for dpkt and pypcap, and thanks to Andrew Brampton for his tutorial on dpkt/pypcap!

If you liked this post, please comment. Commeting is fast, anonymous, and greatly appreciated.

The simplest method is to just record whatever the computer plays. This can be done using Audacity, or anything which will record the audio the computer is currently playing. I found an application which uses this method to record audio from streaming websites. Streaming Audio Recorder records streaming audio by all audio coming from the computer. This is a very naive approach and suffers from quality degradation if the file was not originally going to be re-encoded, but it has the advantage of working in all cases.

Another approach is to capture the stream urls as they are played. Examples of this are WmRecorder, Jaksta, and Orbit Grab++. The stream is then requested a second time and the program records the response. This method downloads each stream twice, once from the original site, then again using the capture program. This works for many sites, but Grooveshark has a handshake challenge which breaks this method.

A final approach, and perhaps the best, is to capture the stream the first time. Unlike the previous approach, it downloads the stream only once. To accomplish this method on Grooveshark, I used WireShark.

Downloading the mp3 from Grooveshark:

1. In WireShark, go to Capture > Options. Under Capture Filter, enter in port 80 and start the capture.
2. Begin listening to the song on Grooveshark and many new lines should begin scrolling in WireShark. At the top of WireShark is a Filter: box. Enter in http.content_type == "audio/mpeg", press enter, and only one line should appear (more lines if more than one song was played).
3. Click the line once and if it comes up processing, hit Stop. Then in the middle pane click the Media Type line (again, Stop if necessary). Right-click the Media Type line and choose Export Selected Packet Bytes… Then save the file somewhere as an mp3 file, choosing All Files as the filetype.

If you liked this post, please comment. Commenting is fast, anonymous, and greatly appreciated.

If you like this post, please comment. Commenting is fast, anonymous, and greatly appreciated.

To install the script, save it as resize-match-dpi.scm into C:\Users\Odie.gimp-2.6\scripts. Open up GIMP and the script will appear under the Script-Fu menu at the top of an image window. If you are really adventurous, I’ve included a small wrapper allowing the script to be called from the console as follows:

"C:\Program Files\GIMP-2.0\bin\gimp-console-2.6" -i -b "(resizematchdpi \"input.bmp\" \"output.bmp\" 1280 0)" -b "(gimp-quit 0)"

resize-match-dpi.scm:

; Released under the GNU GPL 3.0 or later
(define (resize-match-dpi image drawable newwidth newheight)
  (let* (
	 (oldwidth (car (gimp-image-width image)))
	 (oldheight (car (gimp-image-height image)))
	 (newdpi 0)
        )
    (if (= newheight 0)
        (set! newheight (round (* (/ oldheight oldwidth) newwidth)))
	()
    )
    (if (= newwidth 0)
	(set! newwidth (round (* (/ oldwidth oldheight) newheight)))
	()
    )
    (set! newdpi (/ (* newwidth (car (gimp-image-get-resolution image))) oldwidth))
    (gimp-image-undo-group-start image)
    (gimp-image-scale image newwidth newheight)
    (gimp-image-set-resolution image newdpi newdpi)
    (gimp-image-undo-group-end image)
))
 
(define (resize_match_dpi filename output newwidth newheight)
  (let* ((image    (car (gimp-file-load RUN-NONINTERACTIVE filename filename)))
         (drawable (car (gimp-image-get-active-layer image)))
        )
    (resize-match-dpi image drawable newwidth newheight)
    (gimp-file-save RUN-NONINTERACTIVE image drawable output output)
    (gimp-image-delete image)
))
 
(define (resize_match_dpi_multi pattern outdir newwidth newheight)
  (let* ((filelist (cadr (file-glob pattern 1))))
    (while (not (null? filelist))
	   (let* ((filename (car filelist))
		  (image    (car (gimp-file-load RUN-NONINTERACTIVE
						 filename filename)))
		  (drawable (car (gimp-image-get-active-layer image)))
		  (outf (string-append outdir "/" filename)))
	     (resize-match-dpi image drawable newwidth newheight)
	     (gimp-file-save RUN-NONINTERACTIVE image drawable outf outf)
	     (gimp-image-delete image))
	   (set! filelist (cdr filelist)))))
 
(script-fu-register "resize-match-dpi"
  "<Image>/Script-Fu/Resize Match DPI"
  "Resize and match DPI"
  "odie5533"
  "odie5533"
  "2010-05-16"
  "RGB*"
  SF-IMAGE "Image"  0
  SF-DRAWABLE "Drawable" 0
 
  SF-VALUE  "New width" "1280"
  SF-VALUE  "New height" "0"
)

Example script usage:

$ awk -f tinypic.awk The_Grass_Snake_-_Natrix_natrix.jpg
http://i39.tinypic.com/dlnb6r.jpg

Let me start by saying I have no idea why I wrote this in AWK. I’ve only used AWK once before for a very short one-liner. I find AWK to be better than Perl, but worse than Python. It does make for a light and portable language: the Windows AWK executable (available here) is only 345 KB, combine with Windows cURL (available here) which is 405 KB and you’re still under a meg for both. My bloated Python install is over 100 MB and I just installed Strawberry Perl which is 174 MB.

If you like this post, please comment. Commenting is fast, anonymous, and much appreciated.

Here’s the script. Save as tinypic.awk

BEGIN {
    file = ARGV[1];
    # parse file extension
    split(file, arr, ".");
    file_ext = arr[length(arr)];
 
    url = "http://plugin.tinypic.com/plugin/index.php?popts=c,url\\|s,false";
    while (("curl -s " url | getline $0) > 0)
    {
        # parse necessary post variables from the HTML page
        if (match($0, "form action=\"http://s(.)", arr))
            server = arr[1];
        if(match($0, /id="uid" value="([^"]+)/, arr))
            uid = arr[1];
        if(!uid && match($0, /name="UPLOAD_IDENTIFIER" value="([^"]+)/, arr))
            uid = arr[1];
        if (match($0, /name="upk" value="([^"]+)/, arr))
            upk = arr[1];
    }
 
    # compile curl command to upload the file
    curl_cmd = sprintf("curl -s -L -F UPLOAD_IDENTIFIER=\"%s\" -F upk=\"%s\" -F domain_lang=\"en\" -F action=\"upload\" -F MAX_FILE_SIZE=\"500000000\" -F shareopt=\"true\" -F the_file=\"@%s\" -F file_type=\"image\" -F dimension=\"1600\" \"http://s%s.tinypic.com/plugin/upload.php\" -H \"Expect:\" -e \"http://plugin.tinypic.com/plugin/index.php\"", uid, upk, file, server);
 
    # upload the picture and parse the picture URL
    while ((curl_cmd | getline $0) > 0) {
        if (match($0, /name="pic" value="([^"]+)/, arr))
            img = arr[1];
        if (match($0, /name="ival" value="([^"]+)/, arr))
            ival = arr[1];
    }
    if (img && ival) {
        printf("http://i%s.tinypic.com/%s%s", ival, img, "." file_ext);
    } else {
        print "Unable to parse the return page.";
    }
}