[FoRK] insights on Angstro blog

Stephen Williams sdw at lig.net
Wed Oct 28 22:38:47 PDT 2009


Lucas Gonze wrote:
> Through provoking stuff from Rohit on the Angstro blog at
> http://www.angstro.com/node/61 .  The Marc Davis quote is the one that
> really gets me thinking.
>
> ===
> ...
>
> In fact, this led Tom Coates to suggest a seeming paradox: “the only
> way to address information overload is with more information: To
> determine what’s most relevant, algorithms need even better knowledge
> of your geographical and social context.”
>
> When the conversation turned to privacy, Joseph Smarr pointed at the
> difference between announcing ”I am here,” and bugging each of your
> friends individually to tell them. Sometimes, privacy comes from good
> filtering, not simply preventing the information from being shared at
> all (egress rather than ingress, so to speak).
>   

Privacy is secondary to good socializing and/or accomplishing something 
for most people most of the time.
> The weak link in a centralized aggregation service is sharing or
> syndicating over its own APIs. If I report my position to Twitter,
> that location is stuffed into the hidden “resource fork” of a Tweet in
> their API. If I report that I'm attending this symposium to LinkedIn,
> that fact is stuck in a “roach motel,” because they have no API access
> to event data, as Dave McClure mentioned. Davis concurred that “large
> companies have not create identifiers to make this easier at scale,”
> along the lines of Yahoo! Placemaker WOEIDs.
>
> Ångströ’s own Salim Ismail identified a particularly trenchant problem
> with social network interoperability: “Our identities are locked in
> walled gardens. It's impossible for me to comment about or contact
>   

This seems obvious to me: The spam hounding was inevitably going to 
drive most people from common, open email to private email with their 
groups.  I had been thinking this for a while.  MySpace was a mess, 
however Linked-In and especially Facebook have a reasonable model.  My 
main problem with it is lack of easy archiving that I get with email.  I 
already have that problem with chat systems and always have.
> someone without going through those gatekeepers. And even if I want to
> refer to someone I know on another social network, which of 18
> different contexts do I want to refer to him as?”
>   

Not only that, but Facebook weirdly doesn't really make a unique 
identifier obvious.  You often connect to people with multiple 
attributes to perform a search to select the right person from a list.  
Try finding me by name...  My ID seems to be "sdwlig", but if you search 
for "sdwlig" rather than using it in the URL, you don't get me, but this 
guy is close to the top of the list:

http://www.facebook.com/search/?q=sdwlig&init=quick#/profile.php?id=1602980654&ref=search&sid=752757461.2148172085..1
Probably not me.
> Ultimately, the people in your community define what’s salient for
> you, because that’s the shared context that marks your membership in
> that community. Adam Hertz, a founder of TuneIn based their product on
> that insight: “We believe people are the best filter, the people you
> pay attention to. We pull out all the media in your (inbox), and order
> it by popularity within your graph and, more and more, based on how
> you respond to it (your engagement).”
>   

Doesn't seem useful to me, unless it analyzes FoRK, Slashdot, etc.

Stephen





More information about the FoRK mailing list