« Dean: (Re)stating the obvious |
Main
| Point Tippi »
January 29, 2004
Eric Gradman: Distributed Social Software
Posted by Seb Paquet
From USC computer science student Eric Gradman comes a paper titled "
Distributed Social Software".
This is an ambitious, high-level description of how social software should really work in order to scale, preserve consistency, provide flexibility, and prevent fragmentation of the user base. The design could be summarized as "center the architecture on the individual user throughout". While I think it seriously needs fleshing out, the underlying philosophy seems right. I'm not convinced that preventing fragmentation follows directly from the scheme, though, because different open standards compete against one another and there's no guarantee that users will all embrace the same standard. Here's the abstract:
For many years email and usenet news constituted the majority of the Internet's use as a tool to facilitate communication among individuals. The last five years have given rise to a number of novel applications in this domain--which has come to be known as ``social software.'' Notable among these are instant messaging systems, weblogs, and services like Friendster and Tribe which exploit the concept of ``six-degrees of separation.''
These services generally employ centralized client-server architectures. These architectures are failing to adequately scale with the growing user-base. These services do not rely on open protocols; the user-base is fragmented among competing service providers. Users use numerous service providers to get the features they want, but have no easy way to maintain the consistency of their information on each.
This paper summarizes the ever changing state-of-the-art in social software, and presents an alternative to this ``service-centric'' view of social software. The novel user-centric distributed social software model outlined in this paper overcomes many of the limitations of the current model by drawing from ideas from the Semantic Web.
I think making things happen in this way might require many more well-coordinated, idealistic developers than are available right now. But one can always hope...
Compare:
Leonard Lin's "
Next-Generation Distributed Social Software Networks: Designs and Applications" presentation.
Comments (10)
+ TrackBacks (0) | Category: social software
- RELATED ENTRIES
- › Spolsky on Blog Comments: Scale matters
- › "The internet's output is data, but its product is freedom"
- › Andrew Keen: Rescuing 'Luddite' from the Luddites
- › knowledge access as a public good
- › viewing American class divisions through Facebook and MySpace
- › Gorman, redux: The Siren Song of the Internet
- › Mis-understanding Fred Wilson's 'Age and Entrepreneurship' argument
- › The Future Belongs to Those Who Take The Present For Granted: A return to Fred Wilson's "age question"
1. Ash Maurya on January 29, 2004 9:40 PM writes...
Decentralized social networking is not as far off or as ambitious as you may think. In fact, its here already...
Its uncanny, but we could have written a very similar paper almost 2 years ago when we started work on our social networking platform WiredReach. In ours, we would have added end-to-end security, real-time voice/video interactions and placed less emphasis on ontologies and more on plain simple RDF - allowing the data model to evolve naturally...
Great job Eric!
Permalink to Comment2. Ross Mayfield on January 30, 2004 12:25 AM writes...
Centralize the service, decentralize the control.
Permalink to Comment3. eric gradman on January 30, 2004 2:27 AM writes...
Ash,
I agree that RDF is a natural way to express this sort of data model. Whatever tools people choose to use to interact with the Semantic Web, it is desirable to have a convergence of data types.
"<Alice> <is friend of> <Bob>" may be comparable to "<Bob> <is buddies with> <Alice>".
Suggesting "ontologies" for these relationships keeps data types consistent.
And Ross, can you clarify that just a little? ;)
Permalink to Comment4. Seb on January 30, 2004 8:55 AM writes...
Ross: well said.
Permalink to Comment5. Frank Ruscica on January 30, 2004 9:38 AM writes...
Centralize the service, decentralize the control?
At an *actionable* level of detail, what does this mean?
Thanks,
Permalink to Comment6. Bill Seitz on January 30, 2004 2:02 PM writes...
"standards" must be (a) designed, and (b) agreed to.
They are generally designed around needs. And those needs are not clear yet.
So lots of standalone efforts make sense for now, don't they?
Permalink to Comment7. eric gradman on January 30, 2004 2:29 PM writes...
Needs aren't generated spontaneously. They evolve from experimental features that some programmer throws in on a whim. People discover they cannot live without that feature. It becomes a "need."
It would be nice if these "experimental features" could be foisted on as large a userbase as possible.
This distributed social software idea makes it possible to foist these features on anyone who's interested. Develop an RDF predicate that describes a relationship (sexiness) and an interface to manipulate it, and people will find it indispensible. Eventually it will become a standard.
Permalink to Comment8. Ed Bilodeau on January 31, 2004 10:26 PM writes...
Real social networks are the ones that you keep in your mind and in your heart. If you need a computer to keep track of your social ties, are you still talking about real social networks? Are they real relationships, or something the software has conjured up based on a whatever metadata we happen to have at our disposal?
Somehow, the notion of a "Click here to see who your friends are/who you know" button stikes me as a misguided effort.
Permalink to Comment9. eric gradman on February 1, 2004 6:24 PM writes...
Yes. It is a conjuration of whatever metadata we happen to have at hand.
And Ed, I am removing you from my friends list.
Permalink to Comment10. Zbigniew Lukasiak on February 2, 2004 3:49 PM writes...
What I would like to add to this framework is to put more weight on implicite Social Networking instead of going only the explicite way: http://zby.aster.net.pl/kwiki/kwiki.cgi?SocialRouting
Permalink to Comment