Hi, I just saw that this blog has been linked on the German Wikipedia. Now I will continue writing about the rest of the files. So, stay tuned, I will be back after Christmas.

Happy Holidays to all of you.

Let’s take a look at the 2nd Episode. As pointed out here I was unable to decode the file with the correct md5sum. Nevertheless I was able to use it.

Barber.z5 is a story file for a z-machine. See wikipedia

Just download one of the many interpreters and load up the file. It’s a text adventure, set in a well known environment. For instructions on how to play, take a look here:

• A Beginner’s Guide to Playing Interactive Fiction
• Interactive Fiction – Getting started
• Interactive Fiction Player’s Guide

If you want a walkthrough, try using “strings” at the linux prompt. Otherwise, here’s one tip, without it you won’t even be able to get into the office: Take a look here

I haven’t solved the whole game myself yet, but when I do, I’ll inform you
Now, have a nice game-

After we extracted and decoded the files, we now want to use them. So, what are they, how to use them?

At this stage, we take a look at http://www.ilovewillies.com (Only awaylable in the Google-cache)

Edit: I fixed the links

1. EasterEgg1.jpg
2. Barber.z5
3. Eggdump
4. Sophie’s Choice
5. EasterEgg3.part.2
6. Dr. Mike Thorne’s flashing dot at the side of your screen

These are all the informations we need.

Episode 1 gives us an image of Moss, the interesting part is the exif-geolocation. See: BBCentral’s Blog

After my last post, you will end up with a lot of text files. We will now convert these text files to the actual binary files.

To do that we will first have to remove duplicated files. Transcode will sometimes extract one subtitle twice. Use your favourite image viewer (best with automatic search for duplicated files) to identify those. Do not delete them, as it could of course be possible (even if not very probable) that it wasn’t transcode’s fault but they are actually two identical subtitles.

The next step is to join all txt-files and to remove accidentally inserted spaces. A simple
sed -e “s/ //g” subtitles-*txt > base64.txt
entered on the console will do.

Now we have one file containing the base64-encoded source, the md5sum for that file and the original file name. As I didn’t find a program which could handle these files the way I wanted I wrote the following python script to extract the file and check the md5sum:


#!/usr/bin/python
import sys, base64, md5

def hexToString(md5bin):
result = “”
for ch in md5bin:
realCh = hex(ord(ch)).replace(“0x”,”")
result += realCh
return result

if len(sys.argv) < 2:
print “”"Usage: %s in_b64_enc_file

in_b64_enc_file – The Base64 encoded file to be converted
“”"%sys.argv[0]
sys.exit(0)

f = file(sys.argv[1], ‘rb’)
s = f.readline()
md5string = “”
while (s.strip()):
sarray = s.strip().split(‘:’)
if (sarray[0] == “Content-MD5″):
md5base64 = sarray[1]
md5bin = base64.decodestring(md5base64)
md5string = hexToString(md5bin)
print “MD5 should be: ” + md5string
if (len(sarray)<2):
sarray = s.strip().split(“=”)
filename = sarray[1].strip(‘”‘)
print filename
s = f.readline()

fout = file(filename, ‘wb’)
s = f.read()
decoded = base64.decodestring(s)
fout.write(decoded)
f.close()
fout.close()
m = md5.new(decoded)
print “MD5 is: ” + hexToString(m.digest())
print “MD5 matches: ” + str(hexToString(m.digest()) == md5string)

If you were not careful enough (like me) and entered a wrong letter, or enter the right letter with the wrong case, when pgm2txt asked you, you will notice that now. Well, it’s only a question of hours, the convert the image to text again . If any one of you gets the right md5sum for Episode 2 (Barber.z5), please let me know. I always get it wrong even though I am able to use the file.

The next post will talk about the extracted file, and what to do with them.

Now we get to the interesting part:
Converting the images to text.
I’m using pgm2txt for this purpose. Using database only mode (“-d”) ensures that we have full control. Using gocr intern heuristics gave many errors, e.g. “l” was read as “1″.

pgm2txt will ask you to enter the display text, whenever it cannot identify the characters. This means in database only mode, you will have to enter every character at least once. That would be a problem, but as the letters aren’t always clearly separated, you will often end up having to enter the text for combinations like “KWY”. Nevertheless, after having trained the db for some time, all you need is patience and cpu power. After some time (1000 pics => several hours) you will end up with a lot of text files, each containing one part of base64 encoded file.

Step one is easy.
You will need the following programs:

• transcode
• mplayer
• subtitleripper

Using with Linux just use the following script (originally taken from the Gentoo Wiki):

#!/bin/bash
lsdvd
echo “Please type in the stream number”
read DVDSTREAM;
mplayer -dvd-device /dev/dvd dvd://$DVDSTREAM -vo null -ao null -frames 0 -v 2>&1 | grep sid
# ask the user for sid…# the correct number is 0×20 + sid
echo “Please type in the subtitle SID, type in hexadecimal and add 0×20. Example: for sid 0, type 0×20″
read SID;

tccat -i /dev/dvd -T $DVDSTREAM -L | tcextract -x ps1 -t vob -a $SID > subs
subtitles2pgm -o subtitles-$DVDSTREAM -c 0,255,255,255
We will end up with a lot of pictures, each containing one piece of the final base64 encoded file. The next step will be to convert these pictures into text using an ocr-programm

It recently ordered the “The IT Crowd” DVD set from Amazon.uk. Season 1 contains some normal easter eggs, but the Season 2 DVD goes far beyond. The leet subtitles are actually base64 encoded files.
So, what to do now?

1. Extract subtitles from DVD
2. Convert subtitles to text
3. Decode text to files