25. Januar 2004

[ English ]

Playing around with Orkut

I have been playing around with Orkut a lot in the last 24 hours. One thing I don't understand about these Social Networking Platforms is why they ask for useless information and why they don't ask for the interesting things.

Orkut's Karma System Example: Orkut gives me five large textareas to explain things like what I have learned from my past relationships. It lets me give Karma-Points to Phil Wolff if I think he is sexy. It lets me upload photos of my cat.

But it does not ask me if I have ever met Phil in person. (I have.) It does not ask me for how long I have known him. It does not distuinguish between "friend" and "acquaintance".

Orkut's Yes/No-ButtonsIt asks "Is xxx your friend?" and I am supposed to give a "thumbs-up" or a "thumbs-down". (And if I say "no" I cannot even add an explanation to that. I hope I haven't hurt anybody.)

I don't get it. It would be so easy to gather this information and it would be so valuable. A simple "have met in person"-flag or something like "How often have you met X?" - "never - once - several times - often - meet regularly" would do it and give so much more depth to the network.

What is it that I am missing? Why is Orkut not gathering more data about the relationships of its members? (Like most of the information it could be optional - I bet many people would share it.) Please enlighten me.

Further reading on Social Network Platforms: My Interview with Stefan Smalla, Founder of Friendity, a German Friendster-Clone.

[Update January 26th] See also Marc Canter's Comment on Many-to-Many.

[Update January 27th] And do follow the TrackBack over to Lee Bryant.

[Update January 30th] apophenia: venting my contempt for orkut (via Joi Ito), Jeremy Zawodny: Why Google needs Orkut (thanks, Garvin)

[Update January 31st] apophenia: orkut pissyness, round 2

[Update February 1st] Christopher Allen: Insecurity at Orkut

[Update February 3rd] Arve Bersvendsen: Why you should read Terms of Services and Privacy Policies (via Haiko Hebig)

Life With Alacrity: Confirmed Email Privacy Hole at Orkut with some more remarks on Orkut and Feedback to the service. (thanks Anders)

[March 8th] Orkut has introduced a feature a little bit like what I asked for. However it is undocumented. There is no explanation what the rating is good for.

Trackbacks sind Links von anderen Weblogs auf diesen Eintrag.

Martin Roell reflects on the shortcomings of Orkut, and utters reservations I wholeheartedly agree with. The same goes for Ryze, LinkedIn and others. In an attempt to do something with the suggestions Martin gives to Orkut, I have added a...

Ton's Interdependent Thoughts: Orkut and the failure of social networking platforms (25.01.04 14:05)

It's funny to see

Mathemagenic: Orkut and a bit of thinking on on-line networking tools (25.01.04 15:40)

Ton's Interdependent Thoughts: Orkut and the failure of social networking platforms hit my aggregator at the same time that David Galbraith's Should sites like Orkut own your profile did. Tom's entry pointed me at friend Martin's entry Playing around w...

KnowProSE: Orkut (25.01.04 21:36)

Orkut launched on Thursday (see CNET) By Saturday morning I had quite a few invites, so I joined up. Is it just another social networking service? Did the implied Google link help? Maybe, but I'm inclined to believe it is...

Unbound Spiral: Orkut - Mushrooms (26.01.04 08:44)

Orkut was down. Fortunately, I was invited several hours before it went offline. Thanks, stuart! There are quite a lot comments on why Orkut was down. Just read Many2Many, and Ross blogged "Why Orkut doesn't work". I am quite agree...

REORIENT: Why Orkut down? (27.01.04 04:23)

After I read Martin Roell's "playing around orkut" and Ton Zijlstra's "Orkut and the failure of social networking platforms", I am starting to think what are the key shortcomings of currnt social network service and how could we improve it....

REORIENT: To be improved: Social Networks (27.01.04 06:02)

The first flight of the Orkut was brief, but generated a lot of interest. What does this tell us about the state of online social networking today?

Headshift: The flight of the Orkut (27.01.04 14:24)

Martin Roell reflects on the shortcomings of Orkut, and utters reservations I wholeheartedly agree with. The same goes for Ryze, LinkedIn and others. In an attempt to do something with the suggestions Martin gives to Orkut, I have added a...

Ton's Interdependent Thoughts: Orkut and the failure of social networking platforms (27.01.04 19:03)

Alle schreiben in diesen Tagen von sozialen Netzwerken. Und übersehen dabei Amazon, die - immer die Nase imWind, den Blick in fremde Einkaufswagen zulassen: "Amazon Launches New Social Network Called "Pricekut - Customers Can Now See and Comment o...

Aufgelesen: Dem Anderen in den Einkaufswagen spinxen... (28.01.04 16:21)

apophenia: venting my contempt for orkut (via Martin via Joi Ito)...

Anders Jacobsen's sideblog: Interesting perspectives on Orkut (30.01.04 17:11)


"What is it that I am missing? Why is Orkut not gathering more data about the relationships of its members?"

The same discussion has been going on surrounding FOAF. Why make the kind of relationship explicit, if you can describe it further by taking other metadata in account? If A and B have written a book together and state this in their description than you have an excellent description for their kind of relationship. But just for a subset of it. And that´s ok, because most relationships are multi-sided:

If you are linked to a person and that person is a dentist that will at least tell that you know A dentist. If you describe the relationship explicitely then what if you have more than one dentist? Or if your dentist is also your friend.
This would lead to all kind of power games, that also appeared with friendster: "Oh, I´m just a business acquaintance? I thought, I was your friend."...

An flexible ontology of describing human relationships with all it´s subcontexts and hidden meanings can in the end not be archieved.

I think, social software is not really a cause in itself, except maybe for dating purposes, but rather an implicit trust system. It´s part of the semantic web, only really powerful in combination with other metadata.

felix petersen am 25.01.04 13:36 #

Btw., being interested in the german side of things: does Friendity allow you to specify that? I don't think so... Stefan, Seyed - you hear that?

Frank Koehntopp am 25.01.04 13:45 #

Friendity does not classify relationships at the current moment. There are several problems with the taxonomy and thus doability: Does it even make sense to have one taxonomy for relationships of many people? And if you can't find a unified taxonomy, is it not better to have one simple type of relationship?

We have certainly been looking into this and will continue to do so, but so far, we have not come up with a way to classify relationships that fulfills the following criteria: (a) unified usage for all membrs, (b) non-overlapping (at least not too much), (c) complete (at least almost), and most importantly (d) understandable by the mainstream user as well as (e) not forcing the member to declare what he does not want to declare.

We do understand the usefulness of such a feature, but we're not yet sure yet it can be done at all in a good way. However, we'll keep looking.

P.S. Thanks, Heiko, for alerting me to the question.

Stefan Smalla am 25.01.04 14:16 #

By the way, as an addendum, especially to what Felix already hints at: There are of course various possibilities to implicitly find out about the subtleties of relationships in the personal network. That is much more of an area where Friendity sees possibilities in the mid-term and we're working on that.

Stefan Smalla am 25.01.04 14:25 #

I would like to see a "have met in person"-flag on networks such as LinkedIn or OpenBC. There, you often have more than one possibility to connect from one person to another through the network. When forwarding a request I would like to have an option like "prefer strong connections": The request should be forwarded preferably between people that have met in person. I think this would improve the quality of the message transfer on these kinds of platforms. What do you think?

Martin Röll am 25.01.04 14:31 #

Well, I can't speak for the others but only for Friendity:

The issue you are highlighting with requestion introductions can much easier be solved on that feature to ask for an introduction as compared to involving a relationship typology: When requesting the introduction, you are shown all connections to that person, and then you select one of those. This is simple and offers much more flexibility for the requester.

Stefan Smalla am 25.01.04 14:50 #

What about an idea, that might seem silly: Express the strength of the relationship in a number and take the avarage between the two numbers.
Thus you will not have to answers questions about semantics (there is a big difference between 'friend'(US) and 'Freund').
Maybe you could even calculate that number using questions but NOT disclose that number to either person but only tell them the 'strength' of their link.
The system will surely be very complex, but you might just use the 'mellow' term: friend.
I thought hard about whether to send some people I work with an invitation to LinkedIn, actually I only sent it to people who might know about how I think about privacy matters and whom I can expect to stand to their word and manners at all times ;)

Oliver Gassner am 25.01.04 15:04 #

Oliver: Yep, something like that.

Stefan Smalla am 25.01.04 15:10 #

in this stage of 'social networking service' development, i see offerings such as 'orkut' as declaratory vehicles - announcing your 'community' - an interactive 'blogroll' if you will.

i too believe it would be most valuable to be able to clearly delineate one's level of connection with each individual in these 'communities' - best friend, casual friend, acquaintance, business cohort, etc...

these services, in and of themselves - free floating, amorphous 'cartographers of connections in cyberspace' - do not necessarily add significant value to our 'social' lives without 'sticky,' easily adoptable and adaptable components that encourage us to a clearer focus on our purpose for gathering and for self-defining the levels of interaction in our groups or tribes or circles of friends, cohorts, and acquaintances.

judith am 25.01.04 16:28 #