Spending some time for research, I found out that people most fear the transparency in the web. At the same time, this doesn’t prevent them from publishing private details without thinking in the blink of an eye.

So I was curious how much information really is accessible. In the context of my bachelor thesis entitled Super Social Everybody I started an experiment which was to examine how easy the profiling of users truly is.

• profiler.mi.hdm-stuttgart.de Online Profiler
• http://github.com/southdesign/FacebookProfiler Download and Customize

Part A: Blowing the friends graph

As a network I used Facebook with the Facebook API. The first thing I had to notice was, that the information one can get out of a foreign profile is almost useless. It’s just the name, maybe the city and sex – strongly depending on the users privacy settings. So the first step was to create a shill account, which I used to befriend with the target. For that I faked an interesting Facebook profile – fitting the imaginary job description. With this account I started to request several friendships.

Part B: Profile

As soon as the friendship was accepted I was able to profile the user. For this I created a profiler that allows to login with the shill account. Once logged in, a friend list is displayed enabling the profiler to profile the target with a single click.

Try it yourself

The profiler is online, all you need is a Facebook user to log in. Go to profiler.mi.hdm-stuttgart.de and try it!

You can also customize the profiler to fit exactly you needs. Just go to http://github.com/southdesign/FacebookProfiler and clone your own copy or fork it.

Live Demo – Digital Rights Day@HdM Stuttgart

I was invited to make a speech and live demo at the Digital Rights Day of the HdM Stuttgart on June, 18th.

It’s a great honour for me – so feel free to come and see the profiler live in action.

Think about they way people stay in touch with each other. Once up on a time we used to have a handful of good friends, like 5 or so plus we knew the people they hung out with. So all together thats about 30 to 40 persons we more or less stayed in touch with. And I think that’s a good manageable number of contacts. When you met them you had something to tell and they were easily able to support you in life because they pretty much knew what was going on in yours.

But then there was the social network.

All of a sudden you had 300+ friends and the cost of telling someone something went almost down to zero. Just hack it in your status and 300+ friends know that you’ll be on holiday the next week. I remember times as a kid when I was not supposed to tell anybody about going on holiday because burglars would come and rob our house. Does anyone really feel the responsibility for what he is writing to them 300?

Distributed system of friends

Lot’s of things have changed since we connected all our friend graphs. Think about organizing a friday night. Back in the old days someone knew a great party or concert, called his closest 5 friends and told them about it. Then they checked up on their best friends and just like this the whole information spread through the graph. But if someone in the – say 3rd level – friends knew that this party definitely would suck he told it back to the friends who told him and they propagated it back to their source until it reached you again. Maybe it never reached you. Thats exactly what IT guys call a distributed system with a multi-phase commit using transactions. The problem here is the coordination and the human willingness and mood. In an IT system such transactions do work – but still they are one of the most complicated things one could imagine when it comes to organizing and structuring IT systems.

But computers don’t get in a bad mood when they are told to cancel the plan for the 30. time. People do. And after maybe the third trial they will start to compromise and simply leave out some friends that don’t fit in the “transaction”.

Multicasting replaced the distributed system

So why do those social networks work so good? We know now that the old way of organizing or just telling something was really hard work. The social networks served us the possibility to set up that graph of contacts once and simply multicast our informations to all nodes. Compared to the distributed system this is way more effective and less expensive for the individual. Instead of starting the whole transaction thing we now simply ask a question or state were we’ll go and all the other 300 can easily decide whether to join or not.

And although this is a huge improvement in how we interact I see some major risks coming up.

Because of the multicast event organizing we do have contact to a lot of different persons. But when we meet them it’s hard to talk about something really new because they already know so much about each other through Facebook. The next time you’ll meet someone chances are high that it won’t be the same persons you’ve met recently so there won’t be topics to catch up on and the relationship is going to be very platonic.

In my opinion all that leads – or can lead – to less give-and-take because you’ll actually always will find someone within those 300 friends that will join you. You just have to ask all of them – which doesn’t cost you anything.

So where does this lead?

Don’t get me wrong I don’t say I like it – actually I love the fact that you can reach a ton of people with no struggle. I just say that this will totally change  how the world behaves. Say in the year 2050 you can simply write a message to all of your friends from kindergarten. You think that won’t changes us?

As you probably want to integrate the durchblick app in your current server infrastructure or the web you might want to have Apache2 support. Phusion created a fantastic apache mod that enables you to simply deploy rails applications with a default apache2 setup – the mod_passenger.

The installation itself is very easy and well documented on their website. The only thing you need to setup for your environment is a vhost. I had some problems with rake firing up the message queue in the beginning so I had to add the path to the vhost – finally I ended up with this vhost that should work for you as well:

1
2
3
4
5
6
7
8

<VirtualHost *:80>
    ServerName durchblick
    DocumentRoot /home/southdesign/apps/durchblick/public
    RailsEnv production
    SetEnv PATH /var/lib/gems/1.8/bin:/usr/local/sbin:
    /usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
    RailsBaseURI /
</VirtualHost>

I recently set up the Sierra Partyfloss Website because I’m so excited about the upcoming festival summer. Great people, great music and great weather.

I’m also going to present my new EP that I just signed at the Ministry of Sound Germany.

Tickets

Hurry up, the tickets will be sold out soon:

• http://www.sierra-partyfloss.de

I recently discovered something really weird about Facebook.

I was at a friend’s and he was about to sign up for Facebook. As soon as he finished the sign up procedure he was asked to be friends with some guys he truly knew in real life.

How can that be? He was totally new to Facebook and nobody could have accidentally stumbled upon him. He was kind of surprised and so was I.

So the only possible solution I could make up is the “Import Contacts” feature of Facebook. It let’s you connect your Yahoo or GMail account and search your contacts there for registered Facebook users. The new thing for me here is that Facebook not only looks them up, but stores them somehow to suggest you to connected contacts as soon as some of the contacts may sign up. For me, that is sick and impressive at the same time.

If you simply start a very rough calculation you’ll see what I mean:

• According to Facebook Stats they have like 400 million active users
• Say about a fourth of them used the “Import Contacts” feature, thats 100 million users
• Say the average contact import is “only” 5 contacts – although I’m sure it’s more – we have 500 million adresses to be stored

So that means, every time a new user signs up his mail address is being searched for in more than 500 million addresses within a few seconds? And I didn’t even count double or triple relationships in here, there is not only one person in the world having your mail address in his mail account, right?

I’m so proud to announce this semesters work - my new version of Durchblick!

The first thing to say is that lot’s of things have changed now. The app completely moved into the browser because I think that is the future. So I completely moved from Java to Ruby on Rails and I didn’t regret it for one second. The next major issue is that the app now totally syncs up with the Google GData services. And by that I mean your emails, contacts and events! So you create your project in Durchblick and get a email label created in your Google Mail account, an own project calender created on Google Calendar and a project contact group created in your Google Contacts.

And as most mobile devices have build in support for the googles services now – iPhones and Androids and of course the Nexus – you can even have all your project information with you out on the streets. And there is no need for a special Durchblick app, it totally includes into your current mobile OS.

The next decision I made is putting the app up on github and releasing it as open source. This actually has two major reasons. First of all I want to share the project and see if there is anyone interested in forking and improving it. And secondly I’m really interested in what other people do think about my coding style and patterns. The money I wanted to make with it i’ll simply offer as an installation and maintenance service, because my target group – small businesses – mostly won’t have an own administrator or computer geek that is able to set up such a system. So I’m kind of like into the old open source idea.

Features

So what does the app currently feature:

• Create projects, emails, medias, contacts and events
• Sync projects, emails, contacts and events with the Google services
• Store the medias on the rails server or the Amazon S3 Cloud
• Sync up your mobile with Google Mobile
• Upload a new version of a file and don’t lose the old one as history
• Download a project or your whole company as a zip archive to backup

What is still to come:

• Employee login the see who has done what
• Improved email sending
• Paging of all results
• Add jQuery interface optimizations and AJAX support
• Preview medias with thumbnails
• Internal optimizations: Add caching

Where can I get it?

As mentioned I uploaded it as open source on github:

• http://github.com/southdesign/durchblick
• and
• durchblick.southdesign.de is the project website

Durchblick Logo

◄ Back

Next ►

Picture 1 of 12